Light management in luminescent solar concentrators :aligned organic dyes and organic wavelength selectivereflectorsCitation for published version (APA):Verbunt, P. P. C. (2012). Light management in luminescent solar concentrators : aligned organic dyes andorganic wavelength selective reflectors. Eindhoven: Technische Universiteit Eindhoven.https://doi.org/10.6100/IR740226
DOI:10.6100/IR740226
Document status and date:Published: 01/01/2012
Document Version:Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)
Please check the document version of this publication:
• A submitted manuscript is the version of the article upon submission and before peer-review. There can beimportant differences between the submitted version and the official published version of record. Peopleinterested in the research are advised to contact the author for the final version of the publication, or visit theDOI to the publisher's website.• The final author version and the galley proof are versions of the publication after peer review.• The final published version features the final layout of the paper including the volume, issue and pagenumbers.Link to publication
General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright ownersand it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal.
If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, pleasefollow below link for the End User Agreement:www.tue.nl/taverne
Take down policyIf you believe that this document breaches copyright please contact us at:[email protected] details and we will investigate your claim.
Download date: 07. Jun. 2020
Light management in luminescent solar concentrators
Aligned organic dyes and organic wavelength selective reflectors
PROEFSCHRIFT
ter verkrijging van de graad van doctor aan de
Technische Universiteit Eindhoven, op gezag van de
rector magnificus, prof.dr.ir. C.J. van Duijn, voor een
commissie aangewezen door het College voor
Promoties in het openbaar te verdedigen
op 1 november 2012 om 16.00 uur
door
Paul Pieter Catherina Verbunt
geboren te Roermond
Dit proefschrift is goedgekeurd door de promotor:
prof.dr. D.J. Broer
Co-promotoren:
dr. M.G. Debije
en
dr.ing. C.W.M. Bastiaansen
A catalogue record is available from the Eindhoven University of Technology Library
ISBN: 978-94-6191-451-4
Copyright © 2012 by Paul Pieter Catherina Verbunt
The research described in this thesis has been financially supported by the Stichting voor
Technische Wetenschappen (STW) VIDI Grant 7940
“Wish you were here”
Pink Floyd
Veur mam en pap
Table of contents Summary ...................................................................................................................................................... ix
1 Introduction............................................................................................................................................ 1
1.1 Energy in the built environment .............................................................................................. 2
1.2 Solar Energy ................................................................................................................................. 2
1.2.1 Photovoltaic cells .............................................................................................................. 3
1.2.2 Types of photovoltaic cells .............................................................................................. 4
1.2.3 Geometrical solar concentrators ..................................................................................... 8
1.3 Scope and outline of the thesis ................................................................................................ 11
2 Luminescent solar concentrators ...................................................................................................... 13
2.1 Introduction ................................................................................................................................ 14
2.2 Working principle of luminescent solar concentrators ........................................................ 15
2.3 Losses of luminescent solar concentrators and their proposed solutions ........................ 18
2.3.1 Re-absorption of emitted photons by other luminophore molecules .................... 19
2.3.2 Luminophore losses: Limited absorption, limited stability and luminescence efficiency ....................................................................................................................... 22
2.3.3 Photovoltaic losses .......................................................................................................... 34
2.3.4 Waveguide losses ............................................................................................................. 35
2.4 Conclusions ................................................................................................................................ 38
3 Spatial distribution of emitted photons from dichroic dye ensembles ....................................... 41
3.1 Introduction ............................................................................................................................... 42
3.2 Spatial distribution of emitted photons from isotropically distributed dichroic dye molecules .................................................................................................................................... 42
3.3 Spatial distribution of emitted photons from dichroic dye molecules in a planarly aligned liquid crystal host ........................................................................................................ 46
3.3.1 Introduction to liquid crystals ....................................................................................... 46
3.3.2 From molecular dichroism to macroscopic dichroism ............................................. 49
3.3.3 Directional emission from planarly aligned luminophores ...................................... 51
3.4 Spatial distribution of emitted photons from dichroic dye molecules in a homeotropically aligned liquid crystal host .......................................................................... 55
3.5 Spatial distribution of emitted photons from dichroic dye molecules in a tilted aligned liquid crystal host ...................................................................................................................... 56
3.6 Conclusions ................................................................................................................................ 60
vi
4 Emission from planarly aligned dichroic dyes ................................................................................ 61
4.1 Introduction ................................................................................................................................ 62
4.2 Dichroism in fluoresencence ................................................................................................... 62
4.3 Spatial distribution of emitted photons .................................................................................. 63
4.3.1 Theoretical approach ...................................................................................................... 63
4.3.2 Methods ............................................................................................................................ 64
4.3.3 Experimental results and discussion ............................................................................ 67
4.4 Application of silver mirrors or a scattering background to planarly aligned dichroic dyes in LSCs .............................................................................................................................. 70
4.5 Conclusion .................................................................................................................................. 72
5 Surface loss in luminescent solar concentrators ............................................................................. 73
5.1 Introduction ................................................................................................................................ 74
5.2 Theoretical approximation ....................................................................................................... 74
5.3 Methods ....................................................................................................................................... 77
5.4 Results and discussion ............................................................................................................... 78
5.5 Conclusions ................................................................................................................................ 84
6 Reduction in surface loss by dye alignment .................................................................................... 87
6.1 Introduction ................................................................................................................................ 88
6.2 Surface loss from planar and homeotropically aligned dye ensembles ............................. 88
6.2.1 Methods ............................................................................................................................ 88
6.2.2 Results and discussion .................................................................................................... 89
6.4 Tilted dye ensembles ................................................................................................................. 97
6.5 Conclusions ..............................................................................................................................103
7 Organic wavelength selective reflectors .........................................................................................105
7.1 Introduction ..............................................................................................................................106
7.2 Methods .....................................................................................................................................109
7.3 Narrowband reflectors ............................................................................................................113
7.4 Broadband reflectors ...............................................................................................................119
7.4.1 Theoretical approach ....................................................................................................119
7.4.2 Experimental results .....................................................................................................129
7.4.3 Patterned waveguides ...................................................................................................131
7.5 Other luminophores ................................................................................................................133
7.6 Conclusions ..............................................................................................................................135
Table of contents
vii
8 Wavelength selective reflectors and indirect sunlight ..................................................................137
8.1 Influence of the angle of incident light on the performance of cholesteric reflectors .138
8.1.1 Influence of the angle of the incident light on the light that passes through the reflector .......................................................................................................................138
8.1.2 Dependence of maximum possible LSC efficiency with change of the angle of the incident light ........................................................................................................139
8.1.3 Indirect sunlight .........................................................................................................141
8.2 Special dispersion cholesteric reflectors ...............................................................................143
8.2.1 Introduction ...................................................................................................................143
8.2.2 Transmission of sunlight ..............................................................................................145
8.2.3 Efficiency towards surface emitted light ...................................................................146
8.2.4 Angular dependent increase in LSC efficiency .........................................................147
8.3 Conclusions ..............................................................................................................................150
9 Technology assessment and future possibilities ...........................................................................151
9.1 Luminescent solar concentrator: the future ........................................................................152
9.1.1 Energy generating applications ...................................................................................152
9.1.2 Day lighting applications ..............................................................................................154
9.2 Aligned luminophores .............................................................................................................154
9.2.1 Static alignment of luminophores in LSCs: what can be gained? ..........................154
9.2.2 Dynamic alignment of luminophores in LSCs: window applications ..................156
9.3 Wavelength selective reflectors in luminescent solar concentrators ...............................156
Appendix A: Perylene perinone dye .....................................................................................................159
A.1: Introduction ............................................................................................................................159
A.2 Methods ....................................................................................................................................160
A.3 Results and Discussion ..........................................................................................................161
A.4 Conclusions..............................................................................................................................165
References .................................................................................................................................................167
Samenvatting ............................................................................................................................................181
Acknowledgements .................................................................................................................................185
Curriculum Vitae .....................................................................................................................................189
List of symbols .........................................................................................................................................191
List of definitions ....................................................................................................................................195
List of publications ..................................................................................................................................197
Summary
Dwindling oil reserves have turned world governments towards the use of renewable energy
sources. In particular, buildings, which account for 40% of all energy usage in well-developed
countries, have attracted special attention for the possibility of implementing energy generating
devices from renewable energy sources. For buildings, the devices must be adaptable to the
demands of the architect and should not detract from the appearance of the structure, while
maintaining a low costs for the generated energy (~€0.06/kWh). An energy source readily
available for each building is the sun. The implementation of standard photovoltaic cells in
these constructions has been difficult, since the cost of the energy generated is still too high and
the adaptability to the demands of the architect is limited. An alternative to standard
photovoltaic cells is a luminescent solar concentrator (LSC) which holds promise for low cost
and of better meeting the demands of the architects.
LSCs are simple devices that consist of an optically transparent plastic or glass plate acting
as waveguide. A luminophore is embedded in or as a thin film on top of the waveguide and
absorbs the incoming sunlight, re-emitting this light as photons with a longer wavelength. Part
of this emitted light is waveguided in the plate due to total internal reflection and transported to
the edges of the plate where small photovoltaic cells are placed which convert the photons into
electricity. The efficiency of LSCs has been limited due to several loss mechanisms. One of the
most important losses is the photons lost through the surfaces of the plate, a result of a limited
trapping efficiency of emitted photons combined with sequential re-absorption and re-emission
of photons in the waveguide mode (called ‘photon recycling’).
In this work, I present a simple model that predicts the spatial distribution of emitted
photons from dichroic dye ensembles in an isotropic or liquid crystalline host. The model
predicts the emission from dichroic dyes in an isotropic host is non-isotropic when illuminated
with a collimated light source. The model furthermore predicts that by changing the alignment
of the dichroic dye molecules, the spatial distribution of the emitted photons may be altered.
The model is tested by experiments in which the molecular order of dichroic dyes in a
planar aligned liquid crystal is changed. LSC waveguide with the dye molecules aligned parallel
to the top and bottom surfaces and parallel to two edges (planar) were made, and I measure the
light energy emitted from the edge parallel to the alignment direction can be 60% higher than
from the edge perpendicular to the alignment direction, supporting the calculated result.
The amount of surface loss from LSCs with dichroic fluorescent dye molecules randomly
distributed in the LSC is calculated and determined experimentally. Calculations show that the
trapping efficiency is 74.3%, meaning 25.7% of the photons are lost through the surfaces of the
x
LSC waveguide. Experiments show that the photons lost though the surface of LSCs made
from polycarbonate is actually ~ 50% of all emitted photons for LSCs containing BASF
Lumogen F Red 305, the standard dye used in LSCs. This large surface loss is caused by photon
recycling events of waveguided photons and the consequent change in propagation path.
Aligning dichroic dye molecules can be used to reduce surface loss of LSCs. Calculations
and experiments both show aligning dye molecules in a homeotropic fashion result in a
reduction of surface loss to less than 10%, but with concurrent reduction in the incident light
absorbed. Tilting the director of the dye ensembles at an intermediate angle can combine the
advantages of planar (increased absorption) and homeotropic (decreased surface loss) dye
ensembles. Prediction of the optimal tilt angle and order of the dye ensemble with the model
presented in this thesis was not possible. I suggest this model needs implementation in
simulation software so the influence of photon recycling is better represented.
A second method to reduce the surface loss in LSCs is the application of organic
wavelength selective reflectors transmitting the incoming sunlight and reflecting the light
emitted by the luminophore. Calculations show approximately 90% of all surface emitted
photons from luminophore BASF Lumogen Red 305 may be reflected back into the LSC
waveguide, leading to an increase in LSC efficiency of 66% if a 400 nm bandwidth reflector is
used, or by 45% with a 175 nm reflector. Experimentally, the increase in energy leaving the
edge of an LSC containing Red 305 with a peak absorbance of 1.0 is only about 5%. Reducing
the number of photon recycling events in an experimental system with a 175 nm bandwidth
reflector led to enhancement in LSC efficiency of around 20%, showing it is photon recycling
that limits the effectiveness of the wavelength selective reflectors.
The organic wavelength selective reflectors show angular dependency: the reflection band
shifts to shorter wavelengths with increasing angle of incidence, possibly leading to the
reflection of absorbable incident sunlight at these larger angles. Calculations show that for non-
direct sunlight or direct sunlight at oblique incidence angles there is a decrease in effectiveness
of the reflectors and potential decrease in device efficiency. Simulations show that special
dispersion cholesteric reflectors (SDC) have a less pronounced angular dependency. Using SDC
reflectors will enhance the LSC efficiency in any solar condition and only incident light at angles
larger than 70° are reflected away by the reflector.
Both luminophore alignment and organic selective reflectors presented in this thesis
reduce the surface loss of the LSC, and increase the device efficiency. Experimentally, the
increase is not as large as expected from the calculations. To increase the LSC efficiency the
methods presented in this thesis need to be employed using a luminophore with a larger Stokes
shift, so photon recycling is limited. In addition, the amount of sunlight that can be absorbed
by the luminophore molecules (e.g. the spectral coverage) needs to be increased for the LSC to
become viable in the marketplace, as this is still the single largest loss in LSCs.
1 Introduction
Chapter 1
2
1.1 Energy in the built environment
With the realities of dwindling oil reserves now affecting much of everyday life, world
governments have increasingly turned towards renewable energy technologies. In particular,
there is an increasing interest in bringing solar energy systems to the built environment.
Buildings account for about 40% of total energy use including 70% of total electricity use, and
40% of emissions in the more developed countries [1], and about 25% of all energy use globally
[2]. Much of this use is related to our inability to control sunlight: tremendous energy resources
are used to both heat and cool buildings and provide artificial lighting systems. [1] The scene of
an office with shuttered windows and burning lights in the daytime is familiar to everyone.
The European Committee has decreed that all new-to-be-built buildings should be near-
zero energy by 2020. [2] This demands that architects and the building industry must integrate
energy saving and energy generation into the design of new buildings. This puts considerable
pressure on the architect wishing to meet these requirements: how to create buildings both
pleasing to the eye, yet energetically neutral? To give the maximum freedom, the devices
incorporated to save and/or generate energy have to be easily implemented into the design and
adaptable to the situation. Adaptability could also make retrofitting of existing structures during
renovations easier and cheaper. One energy source that is readily available in a built
environment is the sun, which is a clean, safe, inexhaustible and reliable energy source.
1.2 Solar Energy
The sun emits energy in the form of electromagnetic waves with a total of 174 PW reaching the
Earth’s atmosphere. Before this energy reaches the Earth 30% is lost via reflection from the
atmosphere and clouds and from absorptions in the atmosphere. World energy consumption is
in the order of 15-20 TW, so if the energy from the sun could be converted into usable energy
like electricity, the sun alone could produce enough energy to provide the world needs many
times over.
The photons emitted by the sun range from almost 0 eV to nearly 4 eV, where the energy
( E ) of these photons is related to the wavelength ( ) or frequency ( ) by
hc
E h (1.1)
where h is Planck’s constant (6.626x10-34 Js) and c is the speed of light in a vacuum (~3x108
ms-1). As described above, the solar radiation loses approximately 30% of its energy before
reaching the earth. Scattering and absorption of photons by molecules and particles present in
the upper atmosphere are partially wavelength dependent, so the solar spectrum reaching the
Earth’s surface will change. Solar irradiance is defined according to the optical air mass (AM),
which is related to the position of the sun relative to the zenith position. The solar spectrum
Introduction
3
just above Earth’s atmosphere is called Air Mass 0 (AM0) and the standard terrestrial solar
spectrum AM1.5. The AM0 and AM1.5 spectra are depicted in figure 1.1.
Figure 1.1 Solar spectrum of AM 1.0 (black) and AM1.5 (gray).
Solar energy can be converted into usable energy using photo-thermal, photo-chemistry or
photovoltaic systems. Photo-thermal uses the sunlight to heat up a material such as water.
Photochemistry uses the energy in the photons to start chemical reactions, while photovoltaics
convert the suns photons into electricity. Photovoltaics is a clean method to produce energy,
which can be used directly. In the rest of this chapter the photovoltaic method for solar energy
conversion is discussed in greater detail.
1.2.1 Photovoltaic cells
In a photovoltaic cell, absorption of a photon by the semiconducting material promotes an
electron from the valence band to the conduction band, creating an electron-hole pair. In
silicon, a well-known semiconductor, the minimum energy needed to promote an electron into
the conduction band is ~1eV, called the bandgap. Photons which do not have enough energy
to overcome the bandgap will not be absorbed by the semiconducting material. The electron-
hole pair generated has a very limited lifetime, and re-combines. To keep the electron and the
hole separated from each other, asymmetry is built in to the photovoltaic device, which pulls
the electrons away from the holes, creating a potential difference. To increase this asymmetry,
the silicon is doped with other atoms to create p-n junctions. P-type doping adds a different
atom with fewer electrons in the crystal lattice of the silicon, while n-type doping adds an atom
with more electrons. In a p-type semiconductor material the dopant needs an electron to match
the lattice of the semiconductor, which is taken from the valence band of the semiconductor
material, leaving a hole in the valence band and a negatively charged ion in the lattice. In an n-
type semiconductor material the dopant has one electron left after addition to the
Chapter 1
4
semiconductor lattice. This electron is added to the valence band creating a positively charged
ion in the lattice. Adding a p-type and n-type semiconductor together will lead to a depletion
region, where the excess holes and electrons in the valence band recombine, leading to a
negatively charged p-type semiconductor and a positively charged n-type semiconductor,
creating a difference in potential energy between them. This difference in potential energy will
separate the created charges and act as a diode. (figure 1.2)
Figure 1.2 A schematic depiction of a photovoltaic cell, where the open circles depict the holes and the closed
circles depict the electrons in the valence band.
1.2.2 Types of photovoltaic cells
In the previous section silicon is mentioned as a semiconducting material that can be used in
photovoltaic (PV) cells, but there are more types of semiconductor material used in PV-cells
like type III-V cells, thin film cells, and organic cells.
1.2.2.1 Silicon photovoltaic cells
Silicon was used in the first solar cells made in the Bell Laboratories in the 1950s. The bandgap
of silicon is 1.11 eV, meaning that silicon can absorb photons with wavelengths up to
approximately 1100 nm. Silicon is used in several different forms in PV-cells. The first is a
multi-crystalline state in which the PV-cell consists of several grains of crystalline Si, where the
crystal-lattice is differently organized in each grain. The second uses one single crystal of Si,
called ‘single crystalline’ or “monocrystalline’ Si PV-cells. Both types of crystalline Si have an
indirect bandgap, [3] which leads to low absorption. An indirect bandgap is a result of a
difference in momentum between the lowest energy in the conduction band and the maximum
energy in the valence band and consequently optical transitions between free electrons and
holes are forbidden. The absorption coefficient of crystalline silicon is in the order of 2x103
cm-1 in the green part of the spectrum. Grain boundaries in multicrystalline Si cause a loss of
photo generated charges at the boundaries. Thus, monocrystalline Si PV-cells have a higher
efficiency. Large single crystals of silicon are rare in nature and so it is more expensive to
produce single crystalline Si solar cells.
Introduction
5
To improve the absorption of Si cells, amorphous Si (a-Si) is used. In contrast to
crystalline Si, a-Si has a direct bandgap and therefore the absorption coefficient is one order of
magnitude greater. This makes it possible to decrease the cells’ thickness and makes it possible
to construct more flexible PV-cells. The disadvantage of a-Si PV cells is that they lack good
stability. In table 1.1 the current record efficiencies of Si PV cells are listed.
Table 1.1 Silicon PV cell efficiencies in 2011
Photovoltaic cell Record Research
cell efficiency [%]
Single crystalline Si 25.0 [4]
Multicrystalline Si 20.4 [5]
Amorphous Si 10.1 [6]
A disadvantage of silicon PV-cells is still the cost of these cells. The combination of costs
with the efficiency of the Si PV-cells leads to a high price per Watt of the generated electricity.
1.2.2.2 Thin film photovoltaic cells
Thin film photovoltaics use only thin layers of the semiconductor material and could therefore
be made less expensive. Decreasing the thickness of the semiconductor layer decreases the
amount of light that can be absorbed, so the semiconductor material needs to have a high
absorption coefficient. Two examples of thin film photovoltaics which have been extensively
investigated are Cadmium-Telluride (CdTe) and Copper-(Indium)-Gallium-Selenide (CIGS).
Cadmium-Telluride is commonly used in combination with Cadmium-Sulphide (CdS),
where the CdTe forms the p-type semiconductor and CdS the n-type. CdTe is the absorbing
material and has a bandgap of approximately 1.5 eV (~830 nm), which is slightly larger than the
bandgap of silicon. Cells made from CdTe have demonstrated efficiencies above 15%, but
progress towards increased efficiencies has been very slow over the past 10 years, with a record
efficiency of 16.7%, [7] which is still higher than of thin film a-Si. The availability of Tellurium
is somewhat restricted, [8] which makes it difficult to provide enough solar cells to meet
demand. Another concern is the safety of Cadmium, which is a toxic metal. [9] Recycling of
Cadmium is prohibited according to European regulations. [9]
CIGS cells are in principle a solid solution of Copper-Indium-Selenide (CIS) and Copper-
Gallium-Selenide. Like CdTe, CIGS are mainly used in combination with CdS. The bandgap of
CIGS cells can be tuned by altering the content of Indium and Gallium in the solution. [10]
The general structure of CIGS is CuInxGa(1-x)Se2 with variation of x from 0 to 1 to change the
bandgap linearly from 1.0 to 1.7eV. The reported efficiencies for CIGS cells have been the
highest for thin film PV-cells, and the record efficiency is 20.0%. [11] Since the most efficient
CIGS-cells are heterojunction cells of CIGS and CdS the production cost of the cells rise due
Chapter 1
6
to the multiple layers needed. Furthermore Cadmium is introduced, which gives concerns on
safety.
1.2.2.3 Type III-V photovoltaic cells (multi-junction solar cells)
Type III-V cells, like GaAs have been reported to show the highest efficiencies of all materials
used in PV-cells. Thin film GaAs has a reported efficiency of over 28%. [12] Instead of thin
film applications, these materials are mainly used in multi-junction solar cells, which results in
very high efficiencies at the expense of high production costs.
Multi-junction solar cells are a combination of different cells made from different
materials, each with their own bandgap. This combination of multiple materials uses the ability
of high photon to electron conversion of these materials for relatively small spectral
bandwidths. In other words, each layer is only used to convert the light for a limited portion of
the incident spectrum, that part where it demonstrates the highest efficiency. Multi-junction
cells mainly consist from three or four junctions. III-V materials used in these triple multi-
junction cells are, for example InGaP which has a bandgap of 1.8-1.9 eV, InGaAs with a
bandgap of 1.4 eV and Germanium having a bandgap of 0.67 eV, GaAs is also used with a
bandgap of 1.42 eV. Other combinations of these materials have also been employed.
These multi-junction cells have reported efficiencies over 40%, [12] but the materials used
and the production of these cells is very expensive and are currently too expensive for use in
large area energy conversion. Instead, these cells are mainly used in space applications, [13]
where the energy generated per kilogram of materials is much more important than the costs, or
terrestrially in combination with solar concentrators, which will be discussed in section 1.2.3.
1.2.2.4 Organic photovoltaic cells
All the types of cells described so far have been made from inorganic semiconductor materials.
The discovery of organic semiconductor materials induced extensive research into organic
materials than can be used to produce PV-cells.[14] In organic semiconductor materials the
bandgap is formed by the energy difference between the highest occupied molecular orbital
(HOMO) and the lowest unoccupied molecular orbital (LUMO). By absorption of a photon an
electron is excited from the HOMO to the LUMO level, and separating this electron from a
hole that is left in the HOMO level creates an exciton.
Early research on organic PV-cells concentrated on single layer organic materials.[15] In
these cells the active organic semiconductor was placed between an anode and a cathode. The
exciton forms a free electron and hole pair which must transfer to the collection electrodes.
This separation is driven by the electric field created by the difference in work functions of the
two electrodes. In these single layer organic PV-cells the electric field created by the electrodes
was not sufficient to separate the electron and the hole, which rather recombined, leading to
very low photovoltaic efficiencies (<1%).
Introduction
7
One way to solve this problem was to use two layers of active materials placed between
the two electrodes in so-called bilayer organic PV-cells. [16] One material is an electron
acceptor and the other an electron donor. Excitons formed by light absorption feel a strong
local electric field at the interface between these materials, which will lead to separation of the
electron and the hole, since they are attracted to the electron acceptor material and the electron
donor material, respectively. These materials have one big disadvantage; the diffusion length of
the excitons in these materials is only ~10 nm. Therefore, light collection layer thicknesses of
approximately only 20 nm can be used in these PVs, since excitons formed further away than
10 nm from the interface between the electron donor and acceptor material will be lost. To
reach sufficient absorption, these cells require a thickness of 100s of nanometres.
To achieve these thicker active layers, but maintain short migration distances for the
charges a dispersive active layer was created. [17] This creates a much larger interface between
the materials, and formed excitons have better chance of reaching this interface with the
diffusion length of ~10 nm.
There has also been a large amount of research performed to create different kinds of both
electron acceptor and electron donor materials. Some examples are the buckyball, C60 and C70
and perylenebisimides as electron acceptors and derivates of poly(phenylene-vinylene) (PPV),
poly(alkyl-thiophenes) and phtalocyanine as electron donors. (i.e. [18]) So far all this research
has led to a record efficiency of ~10%, and this efficiency is rapidly increasing. [19]
Organic solar cells could become less expensive than the inorganic solar cells discussed
earlier, and viable for production on large scale. Another advantage is that these cells can be
made flexible, which initiates many different application possibilities. The modest efficiency and
the long term stability of these materials are still disadvantages of the organic PV-cells. [20]
1.2.2.5 Other types of photovoltaic cells
There are still a number of types of cells which do not belong to these main classes of solar
cells. Two types of cells that will be discussed in this section are dye-sensitized solar cells
(DSSC) and quantum dot solar cells.
A dye-sensitized solar cell is an electrochemical cell with organic dyes that acts as an
absorber attached to an oxide electrode. Light entering a DSSC will be absorbed by the organic
dye, promoting an electron to an excited state. This electron is then transferred to a porous
TiO2 layer between the dye and the oxide electrode, and from this TiO2 layer the electron is
transferred to the electrode. The organic dye is left with a lack of an electron. The dye molecule
then replaces the electron with an electron donated by the electrolyte in the electrochemical cell.
Iodine is commonly in the electrolyte of the cell can be used as this donor and is oxidised to
triiodide. This triiodide then diffuses to the reduction electrode in the cell and is reduced back
into iodine. The dyes used in these types of materials are mostly ruthenium based complexes
with organic ligands. The efficiency of these DSSCs are modest, the record efficiency is ~11%.
[21] DSSCs are relatively inexpensive if the amounts of Ruthenium and Platinum used in the
Chapter 1
8
electrolyte is kept low. The main disadvantages of these cells is the liquid electrolyte which has
stability problems; if the temperature decreases it can freeze and upon temperature increase it
can expand, leading to rupture of the seal preventing the electrolyte to leak out of the cell. The
electrolyte often contains volatile components that can leak out of the cells, [22] so these cells
must be sealed with care.
Quantum dots can also be used as the semiconductor material. The advantage of quantum
dots is that their bandgap can be tuned by the material of which the particles are made, but also
by the size and the size distribution of the particles. More on the quantum dots can be found in
Section 2.3.2.2. This tuneable bandgap makes them viable for multijunction solar cells, made
from a single material. This technique to produce solar cells is relatively new, and the record
efficiency is still modest (~5%). [23] Expanded research on these cells could improve the
efficiency and they could become an important factor in the PV-cell industry if their theoretical
potential with efficiencies up to 65% is met. It is known that quantum dots have stability issues
when exposed to air, which makes it necessary to seal the cells. Furthermore there is some
concern on the safety of quantum dots. [24-26]
1.2.2.6 Building integrated photovoltaics
For architects it is very important that energy generating and saving devices that may be
implemented in a building do not impede with the appearance of the building. Since PV cells
are relatively thin their implementation on flat rooftops of a building will not change the look
of the building much. Implementation of solar cells on a non-flat rooftop or on the façade of a
building raises more issues. In this case the device is directly in the view. Silicon PV-panels are
only available in black and dark blue and they are not transparent. Some research has been
conducted to change the colour of the PV-panel. [27,28]
Mounting a solar cell to the side of the building sets limitations to the weight of the device.
Devices which are too heavy, like single crystal and multicrystalline silicon PV-panels require
extra care during installation. Thin film photovoltaics do not have this problem, but all
materials have their own challenges towards implementation in the building environment, such
as safety issues, stability and efficiency.
The second important factor in integrating photovoltaics in buildings is the costs of the
energy produced. The cost should be low enough to compete with grid prices and approaching
$0.06/kWh. At the moment the average costs of energy produced from silicon photovoltaic
modules is on the order of $0.30/kWh, so still too high.
1.2.3 Geometrical solar concentrators
To reduce costs of solar energy, concentrating systems were developed. Concentrating systems
replace a large area of expensive photovoltaic cells with less expensive materials. These
concentrating systems concentrate the sunlight from a large area to a small area creating a high
Introduction
9
radiance flux. Traditional solar concentrators are based on geometrical optics like reflection
(mirrors) and refraction (lenses).
The first reflection-based system was a point-focus concentrator, [29,30] a parabolic
mirror that reflects incoming sunlight to one focal point. At this focal point a small PV-cell was
placed. By focusing of all incoming light to this focal point the flux increased, creating
concentration of light. A schematic depiction of a focus point concentrator can be seen in
figure 1.3a.
A more evolved version of this point focus concentrator is a compound parabolic
concentrator, also known as CPC. [31-33] CPCs are two parabolic mirrors that intersect each
other at a point below the focal point of the parabolic mirrors. A small opening is created in
this CPC at the focal point of the parabolic mirrors and a PV-cell is placed at this focal point.
This insures that all the photons that enter the CPC will be focused on the PV-cell. The CPC
does not create an image of the sun on the PV-cell, but the principle of concentration is known
as non-imaging optics. An example of a CPC can be seen in picture 1.3b.
Refraction or focusing concentrators are concentrators based on lenses, primarily Fresnel
lenses. [34-37] Fresnel lenses are plastic or glass sheets containing grooves in a linear (2D lens)
or circular (3D lens) pattern. The grooves in Fresnel lenses form asymmetrical prism-like
structures. These structures mimic the curvature of spherical lenses. The light going through a
Fresnel lens is focused at a focal point just like a spherical lens. The disadvantage of such a
Fresnel lens over a spherical lens is that the focusing is less precise, but the decrease in
thickness makes these Fresnel lenses more likely to be used in concentrating systems. Another
disadvantage of Fresnel lenses is that the acceptance angles for the incident light are rather
limited.
Figure 1.3 Schematic depiction of (a) a point focus concentrator and (b) a compound parabolic concentrator.
Both reflecting and focusing concentrators must meet the thermodynamic conservation
law of etendue. This conservation of etendue limits the concentration of light, especially for
incoming light with a large solid angle. Light entering and leaving a concentrator have both an
Chapter 1
10
area ( inA and outA ) and a solid angle (in and out ). The law of conservation of etendue (U )
upon concentration describes that the maximum achievable concentration ( maxC ) by a
concentrator, and depends on the refractive index of the material in which the receiver (for
instance the PV-cell) is embedded and by the incoming and outgoing solid angle. (see equation
1.2) [38]
out inU U
1 2
out out out in in out out
in out in in in
E L U L U nC
E A A n (1.2)
where E (Wm-2) is the irradiance, L (Wm-2ster-1) is the radiance of the light and n is the
refractive index.. Maximum concentration is achieved when the solid angle of the outgoing light
is 90° and the solid angle of the incoming light is a small as possible.
For completely diffuse light reaching a concentrator in air the maximum concentration is
defined as:
2max,diff concC n (1.3)
Equation 1.3 demonstrates that the maximum concentration for these concentrators for
both diffuse light and direct sunlight that moves along the horizon is relatively low (if the cell is
embedded in a material with refractive index 1.5 the maximum concentration is 2.25). For
direct sunlight the maximum concentration of these systems can be very high since the solid
angle is zero, but this direct sunlight should enter the concentrator always at the same angle. To
achieve this, concentrators are normally combined with a sun tracking system. [39-42] The
disadvantage of these tracking devices is that they add costs to the concentrating system and
that they add size to a relatively small concentrator. This makes these concentrating systems
undesirable for implementation in buildings. [43,44]
Figure 1.4 Flat plate concentrator. Sunlight is coupled into a waveguide made from glass or a polymer by
diffractive elements, light trapping, special backside mirrors, or refractive-index variations.
There is a need for flat panel solar concentrators that do not need a tracking device for
achieving higher concentration. Flat plate concentrators are based on thin polymer or glass
plates which act as a waveguide (see figure 1.4). The sunlight is coupled into this waveguide by
Introduction
11
diffractive elements, [45] light trapping, [46] special backside mirrors, [47] or refractive-index
variations. [48] However, each of these designs presents their own challenges to produce
electricity efficiently enough to be economically feasible. A flat plate concentrator that has great
potential is the luminescent solar concentrator (LSC). The LSC uses (as the name already
suggests) luminescent species to concentrate light. These luminescent species are embedded in a
polymer or glass plate that is transparent. The LSC will be discussed in more detail in chapter 2.
1.3 Scope and outline of the thesis
The scope of this thesis is to determine and increase the trapping efficiency of photons in the
LSC waveguide, and therefore reducing the surface loss of emitted photons. In chapter 2, the
principle functionalities and losses of luminescent solar concentrators are described, and a
background of the research that has contributed to reducing these losses over the past thirty-
odd years is discussed. This chapter is a purely informative chapter with no original research by
the author. In chapter 3, a model is presented that predicts the spatial distribution of photons
emitted by an ensemble of dichroic dye molecules in both isotropic and liquid crystalline hosts.
In chapter 4, the model is validated and the change in spatial distribution of emitted photons by
aligning the dichroic molecules planarly with different order is shown. In chapter 5, the surface
loss from dichroic dye molecules in an isotropic host is determined and the results compared to
the trapping efficiency determined by the model presented in chapter 3. The use of aligned dyes
in LSCs to reduce surface loss is described in chapter 6. Surface loss in LSCs can also be
reduced by application of a wavelength selective reflector. The influence of the application of
organic wavelength selective reflectors on the LSC performance is presented in chapter 7.
These reflectors are angular dependent and this angular dependency is described in chapter 8,
along with a possible solution. The thesis will be concluded by chapter 9, which presents a
technology assessment.
2 Luminescent solar concentrators1
1 Also published in:
P.P.C. Verbunt and M.G. Debije, “Progress in luminescent solar concentrator research: Solar
energy for the built environment”, in proceedings of the World renewable energy congress,
2011.
M.G. Debije and P.P.C. Verbunt, “Thirty years of luminescent solar concentrator research:
Solar energy for the built environment” Advanced Energy Materials, 2, 12-35, 2012
Chapter 2
14
2.1 Introduction
The desire is to take the advantages of a reflective or focusing solar concentrator (collecting
light from a larger area and focusing it onto a smaller area) while avoiding the disadvantages
(the need for direct sunlight, tracking, and often awkward dish-like shapes and large sizes) for
integration in urban environments. An alternative to traditional semiconductor devices that
takes the advantages of concentrator systems but which could still be used in the built
environment could be the luminescent solar concentrator (LSC), which was originally
introduced more than three decades ago. The earliest reference to an LSC type device
architecture appeared as a grant proposal in 1973 by Lerner, [49] soon followed by papers in
the general literature. [50-52] There was a great flurry of activity in the field, leading to a
number of patents. The drop in oil prices in the 1980s led to a near-abandonment of the
research. Recent surges in oil prices and increased awareness of the effects of power generation
from traditional fuels on the global environment have revitalized research on the LSC, and it
has again become the interest of many groups around the world. In particular the past five years
have demonstrated a considerable reawakening of interest in the LSC. This has been due to a
number of factors, including the increasing awareness by the general public of the dwindling
fossil-fuel availability, limitations on the performance and thus capability of deployment in
urban areas for traditional PV systems and the desire from the architectural and building sectors
for more freedom in their design choices.
LSCs were initially proposed as low cost alternatives for standard silicon PV-cells. Recent
research by Farrell et al. showed that for standard first generation LSCs the costs of the LSC
plate should be less than 11% of the costs of an equal size Si-photovoltaic module. [53] The
costs of these modules have dropped lately and are in the order of $350/m2. The costs of the
LSC are determined by the cost of the waveguide material (~$4.5/kg) and the used
luminophore. The costs of an LSC were determined to be ~$35/m2,[54] which is about 10% of
the cost of a Si photovoltaic module. So the costs of the energy produced by LSC and Si
photovoltaic modules is approximately the same. To lower the cost of energy produced from
LSCs the efficiency of the LSC has to be improved, while the costs are kept low.
LSCs promises potential low costs for energy generation [55,56] and they could provide
adaptability to the needs of the architect in that they can be made in a variety of colours,
shapes, and transparencies, could be made flexible, and should weigh less than silicon PV
panels, for example, which could make LSCs more viable for mounting to the side of a
building. [57] In addition, the LSC could function well in both direct and diffuse light, [58,59]
of particular interest for countries with frequent cloud coverage, such as the Netherlands or
areas of persistent shady conditions, such as are typical in cities, for example. If the photon
in/photon out efficiency is high enough, the cost of electricity generated by the LSC could be
competitive with grid electricity. [55] While the efficiency of an LSC will be lower than an
Luminescent solar concentrators
15
equivalent area of a silicon PV due to decreased spectral usage, the reduced cost and
tremendous flexibility in design could make them a viable alternative for the urban area where it
would be too expensive, inappropriate, or impractical to use standard solutions.
2.2 Working principle of luminescent solar concentrators
The basic LSC design allows sunlight to penetrate the top surface of an inexpensive plastic or
glass waveguide. This light is absorbed by luminescent molecules (which could be organic dyes,
inorganic phosphors, or quantum dots, which will be discussed in section 2.3.2), which are
embedded in the waveguide, applied in a separate layer on the top or bottom of the waveguide,
or contained in a liquid solution between two glass plates. [51,60,61] The absorbed light is re-
emitted at a longer wavelength and a fraction of the re-emitted light is trapped in the waveguide
by total internal reflection, becoming concentrated along the edges of the plate. Small PV cells
can be attached to the edges of the waveguide to collect the light and convert it to electricity.
figure 2.1 summarizes the functionality of the device.
Figure 2.1 Working principle of a luminescent solar concentrator. Incoming sunlight (black) is absorbed by the
luminophores in the waveguide. The luminophore re-emits this light at longer wavelengths (gray). A part of this
re-emitted light is trapped inside the waveguide by total internal reflection. The light is concentrated at the edges of
the waveguide, where a photovoltaic cell converts the light into electrical current.
Other flat-plate concentrating systems that do not rely on luminescence have been
proposed. These systems have a variety of forms, for example, relying on diffractive elements
[45], light trapping [46], special backside mirrors [47], or refractive-index variations [48].
However, each of these designs presents their own challenges to produce electricity efficiently
enough to be economically feasible.
The portion of the emitted light that is trapped inside the waveguide depicted in figure 2.1
is determined by the refractive index of this waveguide. According to Snell’s law, all photons
approaching an interface between a material and air at an angle higher than the critical angle will
be totally reflected. This critical angle is defined as
Chapter 2
16
1 1sinc
n (2.1)
where n is the refractive index of the waveguide. This means that for an isotropic emitter and a
waveguide with a refractive index of 1.5–1.6, approximately 75% of all emitted photons will be
internally reflected. These photons will waveguide towards the side of the LSC, where a
significant fraction is coupled out of the waveguide by a photovoltaic cell that transforms the
photons into electrons. However, the efficiency of an LSC is not only dependent on the
trapping efficiency of the waveguide.
One description of the optical efficiency, opt , of an LSC is [51]
(1 )* * * * * * *opt abs PLQY Stokes trap host TIR selfR (2.2)
where R is the reflection of solar light from the waveguide surface, trap is the trapping efficiency
of emitted photons, abs is the fraction of solar light that is absorbed by the luminophore(s),
PLQY is the photo luminescent quantum yield of the luminophore(s), Stokes is the energy lost
due to the heat generated during the absorption and re-emission event, host is the transport
efficiency of the waveguided photons through the clear waveguide, TIR is the reflection
efficiency of the waveguide determined by the smoothness of the waveguide surface, and self is
the transport efficiency of the waveguided photons related to re-absorption of the emitted
photons by another luminophore.[51] The transport efficiency of the waveguided photons related to reabsorption of emitted
photons (self ) is determined by the number of sequential re-absorption and re-emission
(photon recycling) events that take place in the waveguide. Each time an emitted photon in
waveguide mode is reabsorbed there will be chance that it will be re-emitted (PLQY ), it will lose
some energy on heat generation (Stokes ) and only a part of these photons is trapped within the
waveguide (trap ). So self can be described as a product of different efficiencies already in
equation 2.2:
, ,
1
x
self PLQY Stokes i trap i
i
(2.3)
where x is the average number of photon recycling events. The number of photon recycling
events is determined by several factors, like the concentration of luminophore molecules, the
pathlength of a photon through the luminophore layer of the LSC, the extinction coefficient of
the luminophore at the wavelength of the emitted photon, the overlap between the absorption
and the emission of the luminophore and the angle between the polarization of the photon that
is reabsorbed and the optical axis of the luminophore when the luminophore is dichroic, but to
the knowledge of the author no dependency of these factors on the number of photon
Luminescent solar concentrators
17
recycling events has been determined.
The losses influencing all the efficiencies in equation 2.2 will be described later in this
chapter as well as the solutions proposed by several research groups over the past thirty years.
Thermodynamics limits conventional geometric concentrators, like CPCs and Fresnel
lenses, which consequently are inefficient at concentrating diffuse light. [62,63] As described in
section 1.2.3, the maximum concentration that these devices demonstrate is close to the square
of the refractive index of the material in which the concentrator is embedded. Therefore, these
concentrators are not very useful in a built environment, where the sunlight is very diffuse due
to scattering from buildings, cars, or trees. The absorption and re-emission of photons in LSCs
changes the entropy of the system; the maximum concentration of an LSC is therefore
dependent on the heat generation during the absorption and emission event, and thus from the
Stokes shift of the luminescent molecules. The maximum power concentration of an LSC can
be approximated by:
32 1 2
max 301
expe e e
CkTe
(2.4)
where 2e and 1e are the photon energies of the emitted and absorbed photons, respectively, k
is Boltzmann’s constant, and 0T is the ambient temperature. [62,63] For a luminophore with a
Stokes shift of 0.2 eV, the maximum concentration based purely on thermodynamics is
approximately 2000 times.
Considerable effort has been put forth developing a wide range of computational models
using many different approaches to describe the results of the existing devices, as well as to
predict optimal, higher-efficiency LSC designs. Random-walk theory has been used to illustrate
re-absorption, [64] and Monte Carlo simulations have been used to investigate the performance
of single and double-film stacked LSCs. [59,65] Different thermodynamic models have
examined waveguide losses [63,66] which report results in very good agreement with
experiments on test modules. Recently, Meyer et al. have compared the re-absorption probability
model reported by Weber et al. to ray-tracing results of luminophore-impregnated thin-film and
liquid-concentrator systems. [67] Other work has used ray tracing to predict the performance of
LSCs, including filled [68-71] and thin-film [72] waveguides. Each of the theoretical approaches
has their advantages and disadvantages. The thermodynamic modelling requires minimal input
and respond with swift answers but is often limited to simpler geometries and limited
luminescent species. Ray tracing allows much more freedom in device design and number of
luminophores and other details, but is quite computationally demanding. Potentially high
performance levels for the LSC have been predicted by, for example, detailed balance theory of
a single stage fluorescent collector using a high efficiency luminophore and wavelength-
selective surface reflectors. Such a device was predicted to demonstrate efficiencies of up to
90% that of a directly illuminated cell of equivalent size.[73]
Chapter 2
18
2.3 Losses of luminescent solar concentrators and their
proposed solutions
The LSC has not yet been extensively commercialized, primarily owing to their modest
efficiencies. [74,75] A diagram depicting a number of identified loss mechanisms for LSCs
taking into account the factors described in equation (2.2) is graphically depicted in figure 2.2.
Figure 2.2 Loss mechanisms in luminescent solar concentrators: 1) Light emitted outside the capture cone; 2) re-
absorption of emitted light by another luminophore molecule; 3a) incident light not absorbed by the luminophore;
3b) limited stability of the luminophore; 3c) internal quantum efficiency of the luminophore is not unity; 4) solar
cell efficiency; 5a) Fresnel reflections from the waveguide surface; 5b) waveguide absorption of the emitted photons;
5c) waveguide scattering; 5d) surface scattering.
The first loss of the LSC is light emitted by the luminophore under an angle which is refracted
out of the waveguide through an ‘escape cone’ rather than being reflected internally (figure 2.2,
1) which affects trap . The second loss is due to the re-absorption of emitted photons in the
waveguiding mode by subsequent luminophores due to an overlap of the emission and
absorption bands (limited Stokes shifts, figure 2.2, 2), which affects trap and self . The third
loss is a combination of losses related to the luminophore used: one is that the luminescent
molecules have limited spectral absorption bands which lead to incomplete incident light
absorption (figure 2.2, 3a) affectingabs . This light simply passes through the waveguide and is
lost through the bottom surface. Another loss related to the luminophore is the absorption of
high energy UV photons that leads to either direct photo degradation of the molecules (figure
2.2, 3b) such that in time a fraction of the molecules will be broken down and total emission of
the LSC will decrease, or degradation of other molecular species within the vicinity of the
luminophore that subsequently react with it and degrade the luminophore performance. Some
of the absorbed photons are not re-emitted by the luminophores due to limited emission
quantum yield, but instead lost as heat and vibrations (figure 2.2, 3c) affecting PLQY and self .
Luminescent solar concentrators
19
The PV cell at the waveguide edge has a non-uniform spectral response, with a fraction of
incident photons being lost due to the finite conversion efficiency of the PV cell leading to a
fourth loss (figure 2.2, 4). Even optically perfect waveguides can suffer from losses; a small part
of the input light is lost through reflection from the surface of the waveguide, due to Fresnel
reflections( R in equation2.2, figure 2.2, 5a). Waveguides used for LSCs can cause losses
through parasitic absorption, especially in the near infrared (figure 2.2, 5b) affecting host .
Imperfections of the waveguide surface can cause photons in the waveguide mode to leave the
surface and be lost (figure 2.2, 5d) affecting TIR , while imperfections in the waveguide bulk
also scatter waveguided photons (figure 2.2, 5c) affecting host , leading to additional surface
losses.
In the next sections of this chapter an overview of research done on LSCs towards
reducing these losses since their discovery in the 1970s is given. The surface losses are
described in the next chapters of this thesis.
2.3.1 Re-absorption of emitted photons by other luminophore
molecules
The Stokes shift of a luminophore is the wavelength separation between the peak of the most
red-shifted absorption band and the most blue-shifted emission band. Most organic
fluorophores used in LSCs have quite small Stokes shifts, leading to relatively large overlaps
between the absorption and the emission spectrum. [76] As a consequence, fluorophore-
emitted photons can be re-absorbed by subsequent fluorophores encountered during
waveguiding through the waveguide. These re-absorptions are not necessarily losses themselves,
but become a loss if the reabsorbing fluorophore does not again emit a photon (due to non-
radiative relaxation pathways) or if it does emit, but in a direction within the escape cone of the
waveguide. [64,77-80]
One manner in which the problem of re-absorption events has been approached is in the
design of new luminophores with larger Stokes shifts. By reducing the overlap in the absorption
and emission spectra of the luminophores, the re-absorption losses are minimized. A number
of different molecular species with increased Stokes shift have been employed in the LSC,
including lanthanides, [81] phosphors, [82] bipyridyls, [61] and quantum dots. [83-87] These
classes of luminophore bring additional challenges to the production of LSCs. Inorganics often
suffer from low solubility in organic matrices, and often also display a lower absorption. More
on luminophore research is described in Section 2.3.2.
Another technique to reduce the amount of re-absorption is the introduction of a polar
and highly mobile material (in this case, thionin) as a dopant while forming the luminophore
containing polymer waveguide plates. In this work it was postulated that the dopant assisted by
increasing the separation of the absorption and fluorescence bands of the molecules by altering
the electronic states around the luminescent molecules in the polymeric plate, thereby
Chapter 2
20
increasing the Stokes shift and reducing the overlap between the absorption and emission
spectra. [88]
Another option to reduce the re-absorption losses in devices using more than one
luminophore is to physically stack a number of waveguides on top of another, each containing a
luminophore that absorbs in a different part of the spectrum. [51,56,89,90] This is particularly
useful when one wishes to include a luminophore with absorption approaching the infrared,
which to date have generally demonstrated lower fluorescent quantum yields. In this way, by
stacking the more efficient luminophores on top, the light is collected, converted and
transported to the attached cell without ever encountering the reduced-efficiency luminophore.
The added benefit is that luminescence emitted through the escape cone could be captured by a
luminophore in an adjacent waveguide rather than being lost. [89]
To reduce encounters of emitted light with the luminophores, one may apply them as a
thin layer (from sub-micrometer thickness to hundreds of micrometers) to the surface of the
waveguide rather than filling the luminophore within the bulk of the waveguide, [56,71,72,91-
95] or as multiple thin layers stacked on one another. [96] In this way, emission light may be
transported predominantly in the clear host material, and only encounters the luminescent layer
with every second internal reflection. Theoretically, the thin layers should perform as well as the
filled waveguides. [97] In practice, care must be exercised as the limited solubility of the
luminophores often results in an underperformance of the thin layers due to aggregation of the
luminophores, creating non-emitting absorption centers, also known as quenching. [72]
An advantage in utilizing the thin luminophore-filled layer on top of a blank waveguide is
in production: one could envisage the large-area, inexpensive application of thin layers on either
glass or polymer hosts using a variety of techniques (including spin coating, [72] bar coating,
casting, [71] sol–gel techniques, [98] doctor blading, spraying, Langmuir–Blodgett techniques,
or printing). The thin layers could consist of luminophores in acrylates, [99] cellulose triacetate,
[71] polymerized liquid crystals (see chapters 5 and 6 of this thesis), or a host of other materials.
It is also possible to reduce the losses due to the Stokes shift by using more than one
luminophore, and employing Förster resonance energy transfer (FRET). FRET is the direct
exchange of energy from an excited molecule to another nearby molecule without the emission
of a photon. FRET is very sensitive to distance and other factors: the probability of a transfer
being related to the orientations of the interacting molecules as well as the distance:
, 6
0
1
1
PLQY FRET
r
R
(2.5)
where, ,PLQY FRET is the quantum yield of the energy transfer, 0
r
R is the ratio between the
donor and the acceptor ( r ) and the Förster distance ( 0R ).
Luminescent solar concentrators
21
By mixing a number of luminophores at high concentrations it was possible, through a
chain of virtual and real emissions and absorptions, to transfer short-wavelength input light into
long-wavelength output light at relatively high efficiency. [56,100] However, the high
luminophore concentrations necessary for the FRET effect are often not achievable unless the
molecules are brought closer together by other means. [101,102] Another material design taking
advantage of FRET uses ordered dye–nanochannel antennae (based on zeolites), which could
allow enhanced directionality of light emission as well as reduced photon recycling losses.
[103,104]
Another option to try to reduce re-absorption and thus photon recycling losses is to
physically space thin luminophore layers with stretches of empty waveguide using spatially
separated patterns of luminophores, so that the number of re-encounters that emitted light
could have with other molecules was reduced: a depiction of the functionality of such a device
is shown in figure 2.3. [105] The transport efficiency of the photons through the LSC increased
as the overall fluorescent-dye coverage decreased from 100% to 20%, and was relatively
independent of the shape of the dye pattern.
Figure 2.3 The functionality of patterned fluorescent dye layers on a clear waveguide. Top: In standard thin layer
LSCs incident light (black) is absorbed and re-emitted at a longer wavelength (gray). The emitted photons in
waveguide mode encounter the dye layer each time they reach the bottom of the waveguide. Each encounter with
the dye layer the photons have chance of being re-absorbed and possibly re-emitted (dashed). Bottom: In LSCs
with patterened dye layers the re-emitted photons (gray, solid) do not encounter the dye layer each time they
encounter the bottom of the waveguide, reducing the chance of getting re-absorbed.
However, reduction in total light absorption due to the physical gaps between absorption
regions resulted in a decrease of the total system output. [106] In order to increase the amount
of light absorbed in a patterned system, a lens array on top of the LSC is being developed that
would increase the system output. The lenses focus incoming light from a wide area onto small,
patterned regions of dye on the waveguide surface, leading to improved dye absorption.
Chapter 2
22
Current lens designs have demonstrated relatively efficient focusing of light over a range of
±30°. By using illumination through the lenses themselves to induce crosslinking of the dye-
containing acrylate layer, one would be able to tailor the dye pattern to the expected solar light
positions and so effectively minimize the area coverage requirement, as well as eliminate the
difficulty of correctly aligning a separately produced lens array with a dye pattern. While this
device demonstrates increased efficiency under direct lighting, measures still need to be taken to
improve performance in diffuse light.
A very recent approach uses resonance shifting, where spatially confined emission from
bilayer cavities interacts off-resonance with the luminophore layer upon subsequent
interactions. This is accomplished by careful variation of the luminophore thin-film thickness
across the area of the waveguide. The thickness variation is on the order of a few tens of
nanometers and is most appropriate for samples with luminophore layers on the order of a
micrometer in thickness. In this way, initial work appears to allow light propagation with
practically no re-absorption losses, and as such promises potentially significant performance
improvements if the design may be simply and reproducibly applied. [107]
2.3.2 Luminophore losses: Limited absorption, limited stability and
luminescence efficiency
Luminophores are essential for concentration in luminescent solar concentrators. Due to their
absorption and reemission, luminophores decouple the direction of the incoming sunlight from
the direction of the emitted light, overcoming the limitations of geometrical optics in trapping
light within waveguiding modes. The absorption and re-emission of the luminophores alter the
entropy of the LSC, so higher light concentrations can be reached, in accordance with
thermodynamic laws (etendue). [62,63] According to the laws governing etendue (equation 2.4)
the maximum concentration achievable by the waveguide is determined by the Stokes shift of
the luminophore used. Besides the maximum concentration of an LSC, the luminophore is the
most determining factor of light transport efficiency. In Goetzberger’s equation (equation (2.2))
for LSC efficiency, the fraction of solar light absorbed, the photoluminescent quantum yield,
the energy lost due to the heat generation during the absorption and emission events, the
transport efficiency of the waveguided photons related to re-absorption of the emitted photons,
and the spectrum of the photons reaching the PV cell are all determined by the characteristics
of the luminophore. Combining the needs for maximum concentration, maximum efficiency,
and maximum lifetime, the luminophore is the single most important component in the device.
An effective luminophore must meet all the following requirements:
• Broad spectral absorption
• High absorption efficiency over the whole absorption spectrum
• Large Stokes shift (no or low overlap in absorption and emission spectra, to reduce photon
Luminescent solar concentrators
23
recycling)
• High luminescent efficiency (quantum yield)
• Matching the emitted photons to the spectral response of the PV-cell (≈1.14 eV for silicon)
• Solubility in the host matrix material
A wide variety of luminophores have been studied in an effort to meet these requirements.
What follows is an overview of the major efforts in these areas, broken down into the following
categories: 1) organic dyes, 2) quantum dots, and 3) rare earth ions.
2.3.2.1 Organic dyes
Organic dyes are π−conjugated organic molecules, where the core of the molecule is planar
with all atoms of the conjugated chain lying in a common plane and linked by σ-bonds. The
π-electrons form a cloud above and below this plane along the conjugated chain. Absorption
bands of these organic dyes are the promotion of these π-electrons from a ground energy state
to an excited higher energy state. [108] The transition moment of these absorption events is
mostly parallel to the conjugated plane, mostly the molecular axis of the dye, but some
transitions are perpendicular to the molecular axis. These perpendicular transition moments are
observed at short wavelength absorptions. The main absorption wavelength (abs ) can be
estimated by:
28
1abs
mc L
h N (2.6)
where m is the electron mass, c is the velocity of light in a vacuum, L is the chain length of
the π−conjugated plane, h is Planck’s constant, and N is the number of π electrons. The
absorption band of organic dyes is mainly determined by the chain length and the number of π
electrons in the conjugated plane of the molecule. [108]
Since the first papers on LSCs, organic dyes have been investigated as luminophores of
choice due to their solubility, high fluorescence yields, and large absorption coefficients. The
dyes investigated as possible components of LSCs belong to the following classes of molecule:
bipyridines, [61] coumarins, [71,91,92,109-113] dicarbocyanine iodides, [91] dicyano
methylenes, [56,91,92,110,114] lactones, [111] naphtalimides, [111,113,115] oxazines, [91]
perylenes and perylenebisimides, [71,89,94,111-113,115-121] perylenebisimidazoles, [89](see
appendix A) phtalocyanines, [122] phycobilisomes, [123] porphyrins, [56,122] pyrromethenes,
[111] rhodamines, [50,52,91,92,108-111] sulforhodamines, [91,119] tertiary amine derivates of
tetra-cyano-p-quinodimethane, thioxanthenes, [124] (iso)violanthrones, [117] and some
unspecified dyes including BASF K1, [108,119,125] BASF K27, [125] and BASF Lpero. [125]
Characteristic examples of these dyes are depicted in figure 2.4. As can be seen, there have been
a wide variety of dyes explored for use in LSCs over the past few decades.
Chapter 2
24
Figure 2.4 Examples of organic dyes used in LSCs: a) a bipyridine derivate, b) coumarin 6, c) a dicarbocyanine
derivate (DODCI), d) a lactone, e) DCM, f) oxazine 720, g) a naphtalimide derivate, h) phtalocyanine, i)
hematoporphyrin, j) DEMI, k) pyrromethene 580, l) thioxanthene, m) sulfoRhodamine B
Luminescent solar concentrators
25
Figure 2.4 (continued) n) Rhodamine 6G, o)3,9-diisopropionicacid-4,10-dicyanoperylene, p) perylene-3,4,9,10
tetracarboxylic acid-bis-(2’-6’diisopropylanilide) (Perylenebisimide), q) perylene-1,7,8,12-tetrachloro-3,4,9,10
tetracarboxylic acid-bis-(2’-6’diisopropylanilide) (Perylenebisimide), r) perylene-1,7,8,12-tetraphenoxy-3,4,9,10
tetracarboxylic acid-bis-(2’-6’diisopropylanilide) (Perylenebisimide), s) a derivate of an (iso)violanthrone, and t) a
derivate of perylenebisimidazole (called perylene perinone in the rest of this thesis)[117]
Chapter 2
26
The most commonly used dye types for LSCs have been the rhodamines, coumarins, and
perylene(bisimides) derivatives. Rhodamines, like Rhodamine 6G, and coumarins, like
Coumarin 6 (also known as Coumarin 540), belong to the group of dyes mentioned in the
earliest stages of LSC research, [50,109] while the perylenes and perylenebisimides are mostly
mentioned in the more recent papers on LSCs, and were first described for use in LSCs in the
late eighties. [94,117] Rhodamines are known for their high quantum yields and high molar
extinction coefficients, but also for their small Stokes shifts. These small Stokes shifts lead to
much re-absorption and consequently to increased LSC losses, both surface and quantum
(nonradiative) losses. For Rhodamine 6G, the luminescence was reduced when incorporated in
poly(methyl methacrylate) (PMMA) in comparison to the solution luminescence, but no
suggestion has been made to explain this behaviour as this is contrary to common performance
of organic fluorophores in a polymer matrix. [110] In other types of dyes, like Coumarin 540
and dicyanomethylene (DCM), the luminescence increases when they are incorporated in a
solid PMMA matrix, probably caused by an increased molecular rigidity, limiting nonradiative
relaxation of the molecule in the excited state [110] or isolation from radicals or other
impurities. [126] Rhodamines 590, 575, 6G and B show very limited photostability [108,111] in
comparison to other types of dye molecules like perylenes and some coumarins.
Coumarins can have a larger Stokes shift compared to Rhodamines. [91] The overlap
factor, af , of the absorption and emission spectra of Coumarin 540A is 0.12, while for
Rhodamine 6G this overlap factor of the absorption and emission spectra is 0.48. [110] There
have been coumarins tested with moderate to very good quantum yields: Coumarin Red G has
a quantum yield of over 80% [112] (87% was also reported [71]), while the quantum yield of
Coumarin 540A and CRS040 have been reported as nearly unity (98%). [71,111] Even though
the photostability of coumarins have proven to be better than that of rhodamines, they still
have been reported to reduce stability in comparison to perylene-based dyes. [71]
Perylene dyes and their derivates, like perylene bisimides, perylenebisimidazoles, and
(iso)violanthrones, are known for their intense fluorescence and good photostability. [117]
Perylene itself has low photostability, due to easy electrophilic substitution. To improve the
properties of this dye, new side groups were added: 3,4,9,10-tetracyanoperylene has been
synthesized, the cyano groups are electron acceptors and when added at these positions around
the perylene core they improve the photostability. The solubility of this tetracyanoperylene is
moderate, therefore 3,9-diispropionicacid-4,10-dicyanoperylene (figure 2.4o) with a quantum
efficiency of 91% and good photostability was synthesized. [117]
Perylene bisimides are known for their intense fluorescence and good photostability, but
generally exhibit low solubility. To improve the solubility of these perylene bisimides, ortho-
alkylated aromatic bulky groups are added to the bisimides (figure 2.4p). These groups are
twisted 90° with respect to the perylene core due to steric hindrance, which increases the
Luminescent solar concentrators
27
solubility of the dyes. Addition of large side groups on the perylene core (1,7,8,12 positions;
figure 2.4q) shows a chlorinated perylene core) cause the two sides of the perylene core to twist
by 42° with respect to each other. This twist bathochromically shifts the absorption and
emission bands of the perylene dye. Furthermore, the tetraphenoxy-substituted perylene
bisimides (figure 2.4r) have good solubility in organic matrices due to the phenoxy side groups,
especially in comparison to the chlorinated perylene bisimides (figure 2.4q). Component 2.4r
has an absorption band peaking at 578 nm and an emission band peaking at 613 nm with a
quantum yield of 96% with both good solubility [127] and photostability. The violanthrones
[117] and perylene-bis-imidazoles (figure 2.4t and appendix E) [89] have more red-shifted
absorption and emission bands compared to component 10r, with reasonable quantum yields
of, respectively, 55% and 80%. If the conjugated core of the perylene is shortened to a
naphthalene core, the absorption and emission bands are blue-shifted. An example of this type
of dye is the BASF Lumogen F Violet 084, which is a naphtalimide. [115] This work
demonstrates that by changing the side groups or changing the length of the core of an organic
dye, a perylene in this case, one can alter all important properties of a dye, allowing for
engineering of more suitable and robust molecules.
As described earlier, perylene based dyes are known for their high quantum efficiency
[115] and for their large spread in spectral absorption and emission. Combining multiple
perylene dyes in one single device could thus lead to both broad spectral absorption (the whole
visible spectrum) and reasonably high quantum yields. [115]
Two different types of dicyano methylenes have been used in LSCs, DCM (4-
(dicyanomethylene)-2-methyl-6-(pdimethylaminostyryl)- 4H-pyran) [91,92,110,128] and DCJTB
(4-dicyanomethylene)2-t-butyl-6-(1,1,7,7-tetramethyljulolidyl-9- enyl)-4H-pyran). [56] DCM has
a broad absorption spectrum and a large Stokes shift with a reasonable quantum yield (≈80% in
PMMA), [91,110] but the photostability is limited. DCJTB has been used in combination with a
platinum–porphyrin derivate, whereby DCJTB transfers its absorbed energy to the platinum
complex. This combination is calculated to form an LSC with a 6.8% power conversion if
paired with GaAs PV cells. [56] Other dyes have also been investigated in LSCs, as mentioned
before, but the published research on these dyes is more limited. Bipyridines [61] have a large
Stokes shift, due to excited-state intramolecular proton transfer and twisted intramolecular
charge transfer, but the quantum yield is low (up to 3.6% in butanol). Tertiary amine derivates
of tetracyano p-quinodimethane like DEMI (4-[1-cyano-3-(diethylamino)-2- propenylidene]-
2,5-cyclohexadiene-1-ylidenepropane dinitril) are red-light absorbing dyes, but their
photostability is low. [124]
Nature-based luminophores, like phtalocyanines, [122] hematoporphyrin, [122] and
phycobilisomes [123] have also been proposed for LSCs. The phtalocyanines and
hematoporphyrin have intense absorptions and good thermal stabilities, but the quantum yields
are 20% and 8% respectively. The quantum yield of the phycobilisomes is below 50%.
Chapter 2
28
Sulforhodamines, like sulforhodamine-B and sulforhodamine- 101 have small Stokes
shifts, similar to Rhodamines [91] resulting in many re-absorption events. Oxazine 720, oxazine
750, DODCI (a dicarbocyanine iodide: 3,3’-diethyloxadicarbocyanine iodide) and DOTCI (a
dicarbocyanine iodide: 3,3′-diethyloxatricarbocyanine iodide) suffer from similar small Stokes
shifts. [91] Thioxanthene dye D 315 orange, Lactone dye D 838 yellow, and Pyrromethene dye
580 have been shown to be relatively unstable towards illumination with visible light. [71,111]
IR-144 (a dicarbocyanine) is only mentioned in the paper of Batchelder; [91] it appears to have
a reasonable Stokes shift and absorption coefficient, but no further information on the
performance of this dye in LSCs is available.
One of the disadvantages of organic dyes in general is the limited breadth of their spectral
absorptions. To increase this breadth a combination of several dyes can be used. These dyes
can be used in a stack of LSCs with one dye in or on each waveguide [56,89,92,110] or all dyes
in one LSC. [71,91,92,100,109,115,118] In the first case each plate will act as an LSC itself.
Light emitted in the escape cone by one LSC can penetrate another LSC where it can be re-
absorbed (if the light is in the absorption band of the dye of the second LSC) and partly re-
emitted in the waveguide mode. This is described in more detail in the part of this chapter
focusing on the re-absorption problem of the LSC. In case of all dyes in one LSC, all dyes will
absorb energy and all this absorbed energy is transferred to the dye with the smallest bandgap,
via nonradiative or radiative transfer. For FRET the dye molecules have to be very close to
each other, typically less than 10 nm, so this only happens in LSCs with very high dye
concentrations.
Numerous studies have been performed on the photostability of organic dyes:
[71,92,94,108,111,113,117-119,124-126,129] much depends on the processing conditions and
polymeric environment of the fluorophore. Photo degradation of dyes in polymers can occur in
two ways: 1) by direct interaction of the dye molecule with the sunlight, leading to
decomposition, or 2) by the attack on a dye molecule by an active species formed due to
photodecomposition of a residual molecule in the polymer matrix or by singlet oxygen. [111]
These residual molecules are mostly materials used during polymerization or processing of the
matrix material. Two examples of reduced luminescence in LSCs caused by other molecules
present in the polymer matrix under solar illumination are 1) photo reduction, where the dye in
its excited state takes an electron from an electron donor, forming a nonradiative stable anion,
and 2) photo-oxidation where the dye in its excited state donates an electron to an acceptor,
forming a nonradiative cation. [117] Direct interaction of the dye molecule with solar light can
lead to molecular changes, caused by high-energy photons.
Photochemical decomposition of the dye molecule can lead to a reduction in absorption
and thus also a reduction in fluorescence or to a blue-shift in absorption and emission. [108] An
extensive comparison of the photostability between several types of dyes, including
rhodamines, coumarins, perylenes, a naphtalimide, a thioxanthene, a lactone, and a
Luminescent solar concentrators
29
pyrromethene, showed that the perylenes from the Lumogen Series of BASF had the best
photostability, both under illumination of UV light as well as visible light. [111]
There are several options beyond the engineering of more robust dyes towards extending
the photostability of the LSC device. Introducing UV absorbers and hindered amine light
stabilizer (HALS) molecules could reduce the photo degradation of the dyes under UV
illumination. Using copolymers of PMMA has also been shown to increase the photostability of
the dyes. [129] Decomposition of the dye molecules is much faster in a singlet oxygen
environment than in a nitrogen environment, with a small recovery of luminescence occurring
after some time in the dark. [117,130] Thus, it may be that during dark periods during the night
some of the luminescence lost due to photo degradation in the daytime could recover.
2.3.2.2 Quantum Dots
Quantum dots (QDs) are nanostructures from semiconducting materials with dimensions in the
order of 10–100 nanometers. The size of the dots is in the order of the de Broglie wavelength
of the electron. As a consequence of their restrictive size, excited electrons are confined in the
semiconductor, which exhibits optical and electrical properties similar to those of atoms. QDs
promise several advantages over organic dyes if used in luminescent solar concentrators. The
absorption threshold of the QDs may be tuned by judicious choice of the particle diameter.
Colloidal InP QDs, for example, have been shown to be capable of absorbing the whole visible
spectrum. [131] QDs may also have large Stokes shifts, which are determined by the spread in
dot size. [132,133] Their crystalline semiconductor composition should make them more stable
than organic-based dyes. [84] Quantum dots as luminophores in luminescent solar
concentrators were first suggested and modelled by the group of Barnham, [66,83,84] which
predicted a concentrator efficiency up to 20% when type III–V solar cells were used in
combination with high quantum efficiency (QE = 1) QDs.
As mentioned above, QDs promise good photostability, but outside a solid matrix they are
quite sensitive to oxygen and light, [69,134-136] which becomes a challenge for module
engineering. Several research groups claimed that the incorporation of QDs into a solid matrix
could lead to a blue-shift in both their absorption and emission, caused by surface oxidation of
the QD during the manufacturing process. [85,134,137] In addition to a blue-shift in emission
wavelength, the emission intensity in QD solar concentrators is also dependent on the nature of
the host matrix. For example, CdSe/ZnS QDs in a solid matrix lose 22.5–96% of the emission
intensity they exhibited in solution. [85] Several factors can cause this loss, including increased
scattering due to particle clustering owing to decreased solubility, and matrix absorption of the
emission light.
The spectral emission of QDs may be tuned by adjusting the QD size. With increasing
QD size, the bandgap will decrease. The cluster size distribution of the dots can be tuned in situ
during the production of the QD concentrator. By increasing the annealing temperature of CdS
quantum dots in silicon dioxide films, the photoluminescence spectra red shifted, attributed to
Chapter 2
30
an increasing CdS dot cluster size. [138] These results were supported by calculations and direct
X-ray diffraction (XRD) measurements of the QDs, determining that the cluster size grew
more at increased annealing temperature. [137] Schuler also showed that CdS-rich silicon oxide
layers show an absorption band close to the bandgap of bulk CdS, while low concentration
layers had bandgaps of higher energy, corresponding to finite well and tight-bonding
calculations. [139,140]
Naturally, by altering QD materials one may also define alternative bandgaps. PdS used as
a QD material, for example, has an absorption band up to wavelengths in the IR part of the
spectrum. [87] In addition to a larger spectral absorption, these QDs also have a larger Stokes
shift, leading to less overlap between the absorption and emission band. Where the CdSe/Zns (
abs 600 nm) QDs used by Shcherbatyuk [87] have a Stokes shift of 23 nm and an absorption
of 0.68 (normalized to the peak absorption) at the peak emission, the PbS ( abs 750 nm) QDs
had a Stokes shift of 122 nm and a normalized absorption of 0.24 at the emission peak. A
disadvantage of these PbS QDs is that the molar extinction coefficient is one order of
magnitude lower than that of the used CdSe/ZnS QD (2 × 104 L mol−1 cm−1 for PbS and 3 ×
105 L mol−1 cm−1 for CdSe/ZnS). Red-absorbing cadmium based QDs also show a decrease in
absorption coefficient, [141] but because the spectral absorption is higher for these materials
the total absorption efficiency is still higher than that of cadmium based QDs absorbing only
up to the blue or yellow part of the spectrum. Kennedy et al. [141] showed that the optical
absorption efficiency for NIR, orange, and green-emitting Cd-based QDs was, respectively,
23.1%, 21.7%, and 11.6%. The increase in the Stokes shift for the lower bandgap QDs leads to
less re-absorption of waveguiding photons, which in turn may also reduce the surface losses of
the QD solar concentrator (QDSC). NIR-emitting QD solar concentrators were modelled to
have 43 ± 1% surface loss in comparison to the green-emitting QD solar concentrators, which
had 58 ± 5% surface loss. For 60 mm × 60 mm × 3 mm QD concentrators, the reduced re-
absorption translated into a modelled optical efficiency of 13.2% for the NIR-emitting device,
in comparison to 5.0% for one that was green-emitting, assuming similar quantum yields of
unity in each device. [141] PbS QDs can also show similar results. The absorption efficiency of
the PbS QDs is 40% in comparison to 22% for the CdSe/ZnS QD used, but the luminescent
quantum yield for the PbS QDs is lower (30%) than that of the CdSe/ZnS QDs (50%). As a
result, the resultant optical efficiency for the PbS QD solar concentrator was set at 12.6%. [87]
The emission spectra of the QDs used in these experiments demonstrated that increased
pathlength of the emitted light through the waveguide resulted in a spectral shape change, with
the short wavelength peak decreasing with respect to the longer wavelength peak,
demonstrating that re-absorption still plays an important role in these QD concentrators.
[87,138]
QD-based concentrators have been compared directly with organic dye containing LSC
via both modelling and experimental studies. One such work compared CdSe/ZnS QD and
Luminescent solar concentrators
31
Lumogen Red F 300 containing concentrators. [85] The best QD concentrator described in this
paper reached 58% of the performance of the device containing Red 300 containing, mainly a
result of the lower quantum yields (QY 0.1–0.6) of the dots compared to that of the organic
dye (fluorescence quantum yield FQY ≈ 1). Another work has compared the optical efficiency
of CdSe/ZnS QD containing concentrators with several other organic dye containing
concentrators, including Rhodamine B, a red fluorine (Red F), and several other laser dyes
(LDS698 and LDS821). [86] Simulations and experiments described in this work showed the
optical efficiency of the QD concentrator was within 10% of the optical efficiency of the Red
F- and Rhodamine B-containing LSCs. The optical efficiencies of the LSCs containing the LDS
series dyes were comparable to that of the QD concentrator. These relatively low optical device
efficiencies reported were again attributed to the relatively low quantum yield of the QDs.
Increasing the quantum yield of the dots closer to that of Rhodamine B (stated to be ≈95%)
would lead to increased optical efficiencies of the device: as it was, the QD-based LSC
performed at 60% of the optical efficiency of the Rhodamine B-containing LSC. Calculations
done on PbS containing QD concentrators predict such devices could achieve power-
conversion efficiencies (PCEs) more than double the PCE of a Rhodamine-containing LSC, at
3.2% and 1.3%, respectively. [87]
It should be noted there is some concern about the use of the QDSC due to the
potentially toxic nature of many of the materials used in the QD. [24-26] There appear to
remain considerable uncertainties as to the true toxicity of these materials, and the level of
toxicity often is related to not only chemical composition, but also processing conditions,
environmental factors, and a number of other details. Nevertheless, there is continued research
aimed at reduction of the potentially harmful effects of QDs, such as the use of ‘jelly dots’ [142]
or on QDs based on more benign materials, like silicon. [143,144]
2.3.2.3 Rare earth ions
Rare earth ions (sometimes complexed with ligands) are investigated as luminophores for usage
in LSCs primarily because of their promise of high photostability and their large Stokes shift,
although the presence of organic ligands may compromise the effective lifetime of the
molecules. Levitt and Weber [52] described already in 1977 the use of Neodymium (Nd3+)-doped
glasses as materials for LSCs. Neodymium mainly absorbs around 580 nm, but also absorbs
photons at longer wavelengths. Emitted photons have wavelengths around 880 nm (4F3/2 →
4I9/2) and 1060 nm (4F3/2 →4I11/2). The photons emitted at 880 nm can be reabsorbed
(4I9/2 → 4F3/2) and the energy of the photons emitted at 1060 nm is slightly lower than the
bandgap of silicon. These characteristics of Neodymium lead to low efficiencies of Nd3+-
doped glasses as LSCs. [145]
To increase the efficiency of Neodymium-doped LSCs, codoping with Ytterbium (Yb3+)
has been discussed. [145] Energy absorbed by the Neodymium ions is transferred to the
Ytterbium ions, this accomplishes two goals: it will circumvent the limitations due to self-
Chapter 2
32
absorption, and the Neodymium emission at 1060 nm will decrease due to depopulation of the
4F3/2 state of Neodymium by the 2F5/2 state of Ytterbium. Emission from Ytterbium ions is
around 970 nm (2F5/2 → 2F7/2), which is slightly higher in energy than the bandgap of
silicon, but the response of silicon to photons with this wavelength is still high.
The energy transfer from Neodymium to Ytterbium depends on the type of glass: in
Borate Tellurite and Germanite glasses the energy transfer efficiency can reach up to 90%.
[145,146] Ytterbium itself can also be used as a luminophore in LSCs, but the solar absorption
of these ions only occurs in the NIR (2F7/2 → 2F5/2 in 4f13 configuration) or in the UV
(4f12 configuration), so large sunlight absorption is only possible if ytterbium-doped glasses are
codoped. [147] In the case of Neodymium, the Neodymium ions are codopants for the
ytterbium. A disadvantage of Neodymium and ytterbium is their low absorption efficiency.
[148]
Uranyl ions (UO22+) have been reported [148] in LSCs because of their high absorption
efficiency. Uranyl ions have five orders of magnitude higher absorption efficiency than
Neodymium ions, but they absorb only in the blue part of the spectrum (maximum absorption
around 430 nm). The fluorescence of uranyl ions is maximum at 500–530 nm, [148] and the
quantum yield is 0.67. [149] Uranyl ions have also been used as codopants in combination with
Neodymium, [150] and ytterbium together with Neodymium. [151]
Chromium(III) ions have a large spectral absorption with peaks at 450 nm (4A2 → 4T1)
and 650 nm (4A2 → 4T2) and so could be useful for LSCs, but a major drawback is the limited
quantum efficiency (up to 25%). [151] Higher quantum yields have been found in quartz-like
(FQY = 50%), pentalite-like (FQY = 75%), and gahnite (FQY = 100%) glass ceramics.
[152,153] Chromium(III) ions can be used for codoping Neodymium- and ytterbium-doped
glasses, as they increase the absorption range of the glass and then transfer their energy to the
high-quantum yield Neodymium and/or ytterbium ions. The energy-transfer efficiencies from
chromium(III) to Neodymium and ytterbium have been determined at 92% and 88%
respectively in lithium lanthanum phosphate (LPP) glasses. [151] Other rare earth ions like
Sm2+ can also be used in LSCs. [149,154]
To increase the absorption of rare earth ions, organic ligands coordinating to the ions have
been proposed. [81,151,155] In such a complex, the ligands absorb energy and the electrons go
from their S0 to their S1 state. From the single S1 state, energy is transferred to the triplet state
of the ligand (T1) via intersystem crossing. From this triplet state, energy is transferred to the
rare earth ion. Due to all the energy transfers, these complexes have a very high Stokes shift
(>200 nm). Moudam et al. [155] synthesized a Eu3+ complex (Eu(hexafluoroacetylacetonate)3-
(bis(2-(diphenylphosphino)phenyl)etheroxide). This complex absorbs light in the UV region
(<350 nm) and emits it at 613 nm with an FQY of 86% in PMMA. In 2011, Katsagounos [156]
showed four europium-based complexes with increased luminescence compared to the single
europium ion and using these dyes in LSC-type down converters increased the efficiency of
Luminescent solar concentrators
33
multicrystalline-silicon solar cells up by 17%, for the complex with a pyridine derivate as a
ligand. The spectral absorption of Eu-complexes is not very broad because low-energy photons
create a triplet state lacking sufficient energy for transfer to the europium ion, but may have a
large Stokes shift . [81] Other rare earth ions like Neodymium and ytterbium can be excited by
lower-energy triplet states, creating NIR-emitting complexes with broad absorption in the
visible region of the solar spectrum. A problem occurring with these ions is that they can be
deactivated by surrounding vibrational states of O–H, N–H, and C–H bonds. Therefore the
ligand has to been fluorinated or deuterated, because C–F and C–D absorptions occur at lower
energies.
The emission of this fluorinated ligand and Ytterbium complex peaks at 970 nm when
excited at 320 nm. In perfluoromethylcyclohexane, the luminescent QY of the Ytterbium
complex with the fluorinated ligand was determined to be 2.6%; not high enough for LSC
applications.
2.3.2.4 Enhanced fluorescence by metal nanoparticles (surface plasmons)
A newer research area for LSCs is in the field of using plasmonics to enhance the system
response. [95] It has been shown in many research publications that when a fluorescent
molecule is brought within close proximity of a small metallic nanoparticle, such as silver, there
may be a considerable enhancement of the fluorescence. [157-161] There has been previous
application of surface plasmonics in PVs to enhance device performance. [162-168] Very
recently, experiments have coupled such plasmonic photovoltaic cells of LSCs, with predicted
large enhancements of the efficiency. [169] The introduction to plasmonic structures within the
LSC waveguide could open up the use of a considerable range of dyes that had heretofore been
rejected on the basis of low quantum efficiencies or photostability, as both of these important
dye characteristics could be improved by the application of some type of plasmonic based
system.
2.3.2.5 Enhanced absorption by back scattering layer
To aid in luminophore absorption, it is standard practice to apply a rear layer to an LSC to act
as a reflector. The reflecting back layer effectively doubles the path length of incident light
through the dye layer for enhanced absorption. Some of the initial experiments used a silver
mirror, [51] but such a mirror absorbs in the visible range and can result in significant
absorptive losses for longer transmission distances. To avoid absorptive losses, most recent
work has employed a white scatterer. [74,75,100,128,170-173] The effectiveness of the scatterer
depends on both the size of the waveguide (the scatterers have been shown to be beneficial for
waveguides tens of centimetres long) and the concentration of dye within the waveguide. [173]
The scatterer also may direct that fraction of incident light that cannot be absorbed by the dye
directly at the PV cell, allowing it to generate electricity. The separation of the scatterer from
the waveguide by a low refractive index layer is quite important, as one desires to maintain
Chapter 2
34
waveguided light in the trapping modes of the waveguide: every encounter with the attached
scattering layer redistributes the light, and a significant fraction of this redirected light will be
outside the waveguide modes of the system.
It was demonstrated that the effectiveness of the scatterer is greatest near the edges of the
device, where direct scattering of non-absorbed light into the solar cell directly is prevalent.
However, over long distances the direct scattering component is less important, and the
primary advantage given by the scattering layer is the return of unabsorbed light back into the
waveguide for possible absorption on the second pass, but this effect can also be minimized in
highly dye-doped waveguides. [173] This suggests a possible transparent LSC design with only
the close edges using a scattering layer.
The functionality of the rear layer may be enhanced by adding a luminophore to the
scattering layer. This can potentially improve performance by converting light otherwise lost
due to lack of absorption into light of longer wavelengths that can be absorbed by the
luminophores within the waveguide [171,174] while still maintaining the benefits of scattering
described earlier.
2.3.2.6 Enhanced absorption and emission by confinement effects
Other designs may be contemplated that will reduce the limitations of the dye materials through
confinement effects. Recently, a slot waveguide using a nanometer-sized low-refractive-index
slot sandwiched by two high-refractive-index regions was calculated to enhance emission by the
luminophore due to the Purcell effect, to increase the effective absorption length of
luminescent centers and improve their fluorescence quantum yield. [175] The physical
requirements of this device, however, are quite extreme, with a 10 nm slot demanding use of
materials with refractive indices >2. While immediate use of such architecture is not likely, the
design does provide an interesting framework to build upon.
2.3.3 Photovoltaic losses
The standard silicon-based PV has a bandgap corresponding to a photon of around 1100 nm
(≈1.1 eV). Photons with energies above this threshold may still be processed by the solar cell,
of course, but the excess energy of the photon is wasted, and converted most often into heat,
and there is a reduction in the response of the cell for these shorter wavelengths. Energies
lower than this bandgap are not sufficient to generate any current.
However, an LSC does not emit a spectrum even remotely similar to the solar spectrum.
Rather, it emits a narrow range of wavelengths, most often centered at red and NIR
wavelengths (630–720 nm at the maximum). This gives the advantage that the LSC should
remain relatively cool under standard operating conditions, as they do not absorb much of the
infrared light. In consequence, the light incident on the edge mounted photovoltaic will contain
minimal infrared components, preserving high performance [58] as generally photovoltaics
perform less well when heated. [176]
Luminescent solar concentrators
35
In addition to reduced heat loads on the cell, by matching the spectral response of the
solar cell to the output spectra of the waveguides it may be possible to obtain enhanced
performance, although the overall effect may be limited due to the generally broad response of
many cells to light over the common wavelengths emitted by luminophores. [128] However,
this tuning could find application in amorphous silicon [177,178] and various type III–V cells.
Using larger-bandgap solar cells would deliver similar currents, but at larger open circuit
voltage.[71]
Down conversion, which is the process of converting a high energy photon into a lower
energy photon with a wavelength more readily utilized by the photovoltaic, has been studied
using luminophores attached directly to the surface of the PV cells [179-182] or within the
encapsulation layers of the PVs. [27] The goal of these layers is generally to shift the incoming
‘blue’ portion of the spectrum to longer wavelengths better suited to the PV. In other cases,
colours are applied to the photovoltaic surface for more aesthetic reasons, to add alternative
colours to the PV panels. [183] A similar application of this type, but this time using small,
sliver silicon-based photovoltaics located within the waveguide itself, promise improved
performance. [114]
To better exploit the emission spectrum in the LSC, researchers have used type III–V PV
cells based on GaAs and InGaP [74] and obtained record-setting efficiencies. If these cells
could be produced economically, [184] it could hold promise for widespread adoption of the
LSC in the future. Another option could be the use of organic-based PV cells, which often have
a ‘sweet spot’ in the spectral range where the LSC emits. [185]
Another option to enhance the response of photovoltaic cells is to use individual dye-filled
LSC waveguides, where each waveguide is separately attached to a solar cell specific for that
emission wavelength. In the so-called luminescent spectral splitter (LSS) concept, [186] the
incoming light is collected over a large area and funnelled into a smaller region where it is
absorbed by a stack of what are essentially LSCs, each attached to a separate photovoltaic cell
tuned to the specific wavelengths emitted by the LSC. In this concept, the LSCs themselves do
not need to transport the light over extended distances. Yet another option is to employ an
LSC with edge-mounted photovoltaic directly on top of a second photovoltaic: in this
configuration, the light that passes through the LSC due to its limited absorption range will be
collected by the underlying cell. [187]
2.3.4 Waveguide losses
Around 4% of incoming sunlight is reflected from each of the waveguide surface (the refractive
indices of poly(methyl methacrylate) (PMMA) and poly(carbonate) (PC) being between about
1.49 and 1.58) and never enter the waveguide, and could thus be considered a loss. While
antireflection coatings are very common in PV cells, they have not yet been applied to LSCs. As
the LSC relies on total internal reflection from two smooth surfaces, textured systems as used
Chapter 2
36
in many antireflective coatings may not be a viable option. [188] Rather, coatings utilizing
different refractive indices can reduce these reflective losses and can be applied to polymeric
materials. [189] It remains to be seen how these coatings affect LSC performance.
Naturally, it would be a great advantage to be able to produce the LSC panels on the size
of square meters in many applications, as this could allow for a more ‘seamless’ look, and
reduces the amount of wiring that needs to be done thus lowering installation costs, among
other things. However, there are all the losses described in this section that must be accounted
for which places severe limitations on the maximum size for the device. A drop-off of edge
output with distance has been reported.[66,172,173] These transport losses are generally the
result of two main loss mechanisms: re-absorption of emitted light by subsequent dye
molecules (described earlier) and parasitic absorption of the emission light by the host
waveguide material. Other features are certainly at play as well. For example, imperfections in
the surface of the LSC waveguide will also lead to losses and the presence of dust or minor
imperfections can result in the scattering of light outside of the waveguiding mode and lost to
the environment. [190] Generally, the devices should be kept as scratch free as possible,
suggesting the use of hardcoats or the like for particularly soft waveguide materials. There is
also potential for a significant degree of scatter from unwanted scattering centers located within
the waveguide. [190] These scattering and absorption events dominate for large waveguides: re-
absorption by the dyes is primarily a short-range effect, [191] although the absorption ‘tail’ of
the luminophore may have quite a large impact over long distances. [79,172]
One desires to extend the dye absorption as far to the infrared as possible. However, this
is complicated in that the polymeric waveguides, which are predominantly made of PMMA or
polycarbonate, become parasitic, and absorb strongly at wavelengths beyond 880 nm. [172,192-
194] Often the additives designed to improve various characteristics of the host matrix in the
solid state (such as altering coloration, stability, or hardness) or the presence of unreacted
monomers in the host material can have a large impact on the device’s performance.
[99,113,130] For example, an additive which shows only a small absorption when measured
through the width of the waveguide can have a severe impact on the edge output of the same
object, for the pathlength is magnified many tenfolds. [99,128] In addition, as the waveguides
age due to exposure to the elements and in particular ultraviolet light, generates brittleness,
opacity, and a host of reaction species that generate a yellow tint to the plate and act as
absorptive light ‘traps’. [195] To combat these loss mechanisms, research into copolymer
systems has demonstrated enhanced photostability over single component systems. [129]
During the transition from the excited state to the ground state of the dye molecule, electron
transfer takes place from the dye to the main copolymer (here polystyrene-co-
methylmethacrylate) chain, which may increase the photostability of the dye.
In general, the waveguide materials generally used for LSC work tend to be chosen with
economy, rather than absolute best performance, in mind. However, much optically clearer
Luminescent solar concentrators
37
materials are available in the marketplace and could be considered for use in the LSC. For
example, PMMA and perfluorinatedpolymer- based optical fibers may demonstrate
considerably enhanced transmission characteristics. [196]
An obvious way of decreasing the waveguide losses is application of waveguides using
higher refractive index materials. The most prevalent materials used for the LSC have been
PMMA and glass with refractive indexes on the order of 1.49. Higher refractive index materials
used or considered have included polycarbonate (n ≈ 1.59), [99] and special glasses (n = 1.5–
1.8). [56,69,99] Glass samples allow for potentially less absorption in the emission regions of
the dye than polymeric waveguides, [69] but many of the professed advantages of the LSC (for
example, reduced weight and ease of handling) would be nullified.
Another waveguide material that has been considered for its flexibility has been
polysiloxane. [197,198] While the durability of such materials might be a question, a range of
novel applications could be considered with the addition of flexibility to the LSC range of
characteristics. The emission of such polysiloxane-based systems has been comparable to that
of filled polymer waveguides of a certain dye concentration. This suggests that if the dyes were
more soluble, such rubbery waveguides could be a viable option.
An additional aspect of waveguide design that needs to be considered is that edge emission
from rectangular waveguide edges is non-uniform over the exit surface. Intensity of the
emission can vary 20% between the center and corner of the waveguide edge.[199,200] This
variation may also cause additional losses, as this means illumination of the attached
photovoltaic is not uniform, and non-uniform illumination of a photovoltaic cell will result in
decreased performance of the cell. [201] In order to improve the light concentration at the
edges and reduce the size of the attached solar cells, tapering of the waveguide edges has been
suggested. [202]
As one of the main goals of the LSC is to reduce the surface area of photovoltaic cells, it
would be preferable to reduce the number of edges with solar cells attached from four to fewer.
Configurations using two-edge coverage (both opposite and orthogonal cell placement) and
single-edge arrangements are common. When reducing the edge coverage of solar cells,
reflective mirrors are often added to the PV-less edges, and a fraction of this reflected light
reaches the photovoltaic cell. [69,71]
For both performance and aesthetic reasons, alternative shapes of the waveguide have
been studied. The shapes of these waveguides also influence the edge-emitted-light distribution.
[58,71,77,200,203] While the circular LSC could provide the greatest edge output, [200] the
packing limitations of such an array would be inappropriate. Rather, the hexagonal device was
calculated to be a superior design to the common rectangular device as far as light emission,
[71,200] but altering the geometry necessitates a different area of solar-cell attachment and
material usage for the waveguides, which have an effect on the price of the module.
Calculations concluded that varying the geometry type did not reduce costs significantly enough
Chapter 2
38
to be an economically viable plan: rather, it was the overall size of the object was a critical
feature. [71]
By adding a third dimension to the standard device one can also influence emission
efficiencies. Including curvature to the surface of a hexagonal waveguide was claimed to be
effective in reducing losses through the escape cone of the waveguide. [204] Expanding upon
this, a multi-cylindrical array of LSC tubes was shown to both increase the degree of light
concentration and reduce the surface reflections. [205] An LSC fiber has also been suggested
which could be bundled with other fibers and the emission light made incident on a single
photovoltaic, [206] and recently fibers with built-in focusing elements have been made. [207]
2.4 Conclusions
Over the past thirty-odd years a great amount of research has been performed to improve the
efficiency of LSCs. Most work has been performed on improving the luminophore of the LSC,
since the dye is causing the largest decrease in the efficiency of an LSC.
An estimation of the contribution of each parameter in Goetzberger’s equation (equation 2.2)
is presented in table 2.1 for LSC with one organic dye (for instance Red 305) as luminophore.
These parameters depend on both the type of luminophore and waveguide.
Table 2.1Estimated values for the parameters in Goetzberger’s euation for the optical efficieny of LSCs
Parameter Estimated value
1-R 0.96
ηabs 0.2-0.3
ηStokes 0.85-0.95
ηPLQY 0.95-1.0
ηtrap 0.75
ηhost 0.9-1
ηTIR 0.9-1
ηself Depending on ηStokes, ηPLQY, ηtrap and the number of photon
recycling events
In Goetzberger’s equation to calculate the optical efficiency of an LSC, abs is the most important
parameter. For Lumogen F Red 305, the state-of-the-art luminophore used in LSCs, this
absorption efficiency is on the order of 0.3, limiting the maximum incident photon- output
photon efficiency of the LSC to 30%. The luminophore molecules also influence the quantum
yield, the Stokes shift and self-absorption terms. Research over the past 30 years has proven
that finding a luminophore that has a combination of all the desired properties is difficult.
Another limitation in Goetzberger’s equation is the trapping efficiency. The trapping efficiency of
Luminescent solar concentrators
39
emitted photons is a parameter in that equation that is relevant in multiple events, as each
photon recycling event has trapping efficiency playing a large role. The trapping efficiency is
dependent on the spatial distribution of emitted photons in the LSC waveguide. In this thesis, I
focus on improving the trapping efficiency of the LSC where there has been minor research
conducted. A model is presented to calculate the spatial distribution of emitted photons coming
from an ensemble of dye molecules in an isotropic or liquid crystalline host. (chapters 3 and 4).
The trapping efficiency is calculated for several different types of alignments (in isotropic and
liquid crystalline hosts) of dichroic dyes and the resulting surface loss is monitored (chapters 5
and 6). In the last part of this thesis the trapping efficiency is increased by placing an organic
wavelength selective reflector on top of the LSC waveguide, which is used to reduce the escape
cone of the LSC waveguide.
3 Spatial distribution of emitted
photons from dichroic dye
ensembles
Chapter 3
42
3.1 Introduction
Photons emitted by luminophore molecules inside the LSC-waveguide will only be total
internally reflected if they encounter the waveguide-air interface at angles larger than the critical
angle. The photons with a smaller angle with respect to the normal of the waveguide surfaces
will be lost through the surfaces. The trapping efficiency of emitted photons depends not only
on the material of the waveguide, including glass, polycarbonate or PMMA, but also on the
spatial distribution of the emitted photons. This distribution is assumed to be spherical in
published literature. In this chapter a model is presented that calculates the spatial distribution
of photons emitted by dye molecules in a host material (dye ensemble). The host material is
isotropic or anisotropic such as a liquid crystal.
3.2 Spatial distribution of emitted photons from isotropically
distributed dichroic dye molecules
As already discussed in chapter 2 of this thesis, organic dyes are π−conjugated organic
molecules. In these molecules the core is mostly planar with all atoms of the conjugated chain
lying in a common plane. This leads to a cloud of π-electrons above and below the conjugated
chain. [108] The transition dipole for absorption and emission events of these molecules are
usually parallel to the conjugated plane, a phenomenon known as dichroism, which means that
absorption of linear polarized light is not isotropic for these molecules. Similarly, the spatial
distribution of emitted photons from one molecule is not isotropic. The absorption probability
of incoming light depends on the angle of the transition dipole for absorption ( ) with respect
to the polarization of the incoming light ( ie ) and the probability for emission depends on the
transition dipole for emission ( ) and the polarization of the emitted light ( fe ). The spatial
distribution of the intensity ( ,I ) of the emitted light from an ensemble of dye molecules (
) can be defined as [208-210]
2 2
, i fI e e
where and denote the polar and azimuthal angle of the emitted light and the brackets
represent the average over all positions of the dye molecules. In the remainder of this thesis it is
assumed that the direction of the transition dipoles for absorption and emission are the same
and that the molecules are static (that is, the fluorescence is faster than the molecular motions).
Using these assumptions the spatial distribution of the emitted photons will be [208]
2 2
, i fI e e . (3.1)
When there are aggregates of dye molecules formed or re-absorption and sequential re-emission
events occur, this assumption is no longer valid. In figure 3.1 a schematic definition is given for
Spatial distribution of emitted photons from dichroic dye ensembles
43
the transition dipole for absorption of a dye molecule when illuminated from the top by
collimated sunlight.
Figure 3.1 Schematic definition of the dipole, the incoming light and the emitted light for both the absorption
(left) event and the emission (right) event. In the left figure the black arrow represents the dipole moment for
absorption defined by a zenith ( ) and an azimuthal ( ) angle with respect to the axis system. In the right
picture the emitted photon ( k ) is also defined by a zenith ( ) and an azimuthal ( ) angle with respect to the
axis system. The large grey arrow in both pictures represents the direction of the incoming light which is circularly
polarzied (curved black arrow).
Sunlight can generally be treated as unpolarized. Dye molecules that lack a chiral center
have no intrinsic preference for the handedness of circular polarized light, making it possible to
mimic unpolarized light by circular polarized light, since this can be seen as a summation of all
linear polarizations. The transition dipole and the polarization of the incoming light are defined
as:
sin cos
sin sin
cos
,
11
20
ie i
leading to a probability of absorption equalling
2
2 2 2 21sin cos sin sin
2ie (3.2)
The direction of light emitted by a dye molecule ( k , equation 3.3) depends on the
polarization of this light ( fe ). This polarization of the emitted light can be described by a linear
combination of two linear polarizations ( ,1fe and ,2fe ) orthogonal to k (see figure 3.2, where
k is perpendicular to the plane).
Chapter 3
44
sin cos
sin sin
cos
k (3.3)
Figure 3.2 The polarization of the emitted light as function of two linear polarizations orthogonal to k
The two linear polarizations orthogonal to k are defined as
,1
,2
cos cos
cos sin
sin
sin
cos
0
f
f
e
e
so the polarization of the emitted light is
,1 ,2
cos cos cos sin sin
cos sin cos cos sin sin cos
cos sin
f f fe e e
leading to a probability of emission equalling
2 2 2 2 22 2
2 2 2 2 22 2
2 2 22
2 2
2
cos cos cos sin sinsin cos
2cos sin cos cos sin
cos cos sin sin cossin sin
2cos sin cos cos sin
cos cos sin
2cos cos cos sin 2cos ssin cos sin
fe
2 2
2
2
in cos cos
2cos sin cos sin 2sin cos sin
cos sin cos 2cos cos sin cos 2cos sin sin sin
cos sin sin 2cos cos sin sin 2cos sin sin cos
(3.4)
The definitions of , ,1fe and ,2fe are depicted in figure 3.2. The emitted light will have all
Spatial distribution of emitted photons from dichroic dye ensembles
45
polarizations so the calculation of the intensity of this emitted light becomes
2
2 2
0
, i fI e e d (3.5)
The average over the ensemble of dye molecules can than be described by
2 2
2 2
0 0 0
, sin i fI d d d f e e (3.6)
where f is the distribution function of the transition dipoles within the ensemble. For
isotropically distributed dichroic dye molecules, the transition dipoles are isotropically
distributed and independent of the angle between the molecular and optical axis of the
molecules (see figure 3.9, black line, 2 0S ). The distribution function can be described as
f p , where p is a constant. Combining this with equation 3.2, 3.4, and 3.6, and
calculating all integrals leads to the spatial distribution of the emitted photons
2, 3 cosI (3.7)
The spatial distribution of emitted photons is depicted in figure 3.3.
Figure 3.3 Emission profile from isotropic dye ensembles illuminated from the top, both the front view (left) and
the top view (right). The axis of these emission profile have aribitrary units. The units are the same on both axis
and the emission originates from the center of the profile.
The emission profile shown in figure 3.3 for dichroic dye molecules in an isotropic host is
not isotropic, since molecules which have their optical axis perpendicular to the direction of the
incoming light absorbing more light than molecules with their optical axis parallel to the
direction of the incoming light. The molecules with their optical axis perpendicular to the
direction of the incoming light will emit photons mostly perpendicular to their optical axis,
Chapter 3
46
therefore more light is emitted in the direction of the incoming light, leading to the non-
isotropic emission profile.
3.3 Spatial distribution of emitted photons from dichroic dye
molecules in a planarly aligned liquid crystal host
In the past it was shown by several research groups (for example [210,211]) that alignment of
dichroic dyes by liquid crystal materials or stretched polymer films leads to macroscopic
dichroic behaviour of the material in both absorption and emission. The spatial distribution of
emitted photons from an aligned ensemble of dye molecules has never been calculated (to the
knowledge of the author). Control over the spatial distribution of emitted photons could
provide a toolbox to manipulate the distribution of photons in the waveguide of LSCs.
In this section a short introduction to liquid crystals and molecular dichroism is given
followed by a theoretical calculation of the spatial distribution of photons emitted by organic
dyes aligned planarly by liquid crystals.
3.3.1 Introduction to liquid crystals
Liquid crystals (LCs) can be viewed as the fourth state of matter, in addition to a solid, a
liquid and a gas. LCs exhibit orientational and/or positional order while they still flow like a
liquid. [212] For this reason the liquid crystalline phase is called a mesomorphic phase between
the liquid and the crystalline state of matter. While molecules in the crystalline phase exhibit
both positional and oriental order in three dimensions, the molecules in a liquid display no
order at all. Liquid crystalline phases can exhibit only orientational order, called the nematic LC-
phase or orientational order in combination with positional order in less than three dimensions
called the smectic LC-phase. [213] Materials exhibiting LC-phases do not have to exhibit both
types of LC-phases and they may have several different Smectic phases. LC-phases occur in
materials where the molecules have a high anisotropy in shape like rods or disk-like molecules.
Due to this anisotropy, the steric and dispersive interactions between the molecules in the
material are also anisotropic; this is the driving force for the formation of the LC-phase. These
anisotropic molecules can show order when they are molten called thermotropic LC-materials,
or in a solvent called lyotropic LC-materials. The materials discussed in this thesis are generally
thermotropic. Molecular representations of the different phases of a thermotropic material are
depicted in figure 3.4.
Spatial distribution of emitted photons from dichroic dye ensembles
47
Figure 3.4 Schematic represesentation of the phases of a material exhibiting thermotropic LC behaviour. A
crystalline (left) material can undergo a transition upon heating to form a smectic phase. Heating to more elevated
temperatures the material can form a nematic phase and increasing the temperature even further induces a
transition to an isotropic liquid. A LC-material does not need to exhibit both smectic and nematic phases.
The average orientation of the LC-phases can be described by a director ( n ). The
molecules show an orientational discrepancy ( ) with this director. This discrepancy is
symmetric around the director and it is a measure of the quality of the order. To quantify the
quality of the order of LC- materials, the order parameters ( 2nS ) are generally used. The order
parameters are depending on the average angle( ) between the director and the individual
molecules and are defined as:
22
13cos 1
2S (3.8)
and
4 24
135cos 30cos 3
8S (3.9)
Nematic LC-materials exhibit an order parameter ( 2S ) typically between 0.4 and 0.8. If 2 0S ,
the material is completely disordered, so it is an isotropic liquid and if 2 1S , the material has
perfect order and a completely crystalline solid.
The order discussed in the previous paragraph is only a short range order. LC-materials
have no long range order, so the direction of the director may change over long distances in the
material. For LC-materials to exhibit long distance order a secondary driving force needs to be
applied. This external driving force can be an external field like an electrical field [214], a
magnetic field [215,216] or a shear-flow mechanical field [216], but it can also be an
anisotropically grooved surface [217], epitaxial-like transfer of molecular order in a surface (as
e.g. in surfactants) or polarized laser light [218]. In the experiments conducted in this thesis a
polymer surface with manually applied parallel grooves is used to create long distance order in
the LC-materials via anisotropic dispersive forces created by the grooves. Polymers that can be
Chapter 3
48
used to created anisotropic grooved surfaces are for example, polyimides (PIs), polyvinylalchol
(PVA) or triacetylcellulose (TAC).
These external driving forces can impose long distance order where the director is
constant parallel to the surface (called ‘homogenous’ or ‘planar’ alignment) or perpendicular to
the surface (called ‘homeotropic’ alignment), but it is also possible to change the direction of
the director gradually in one direction. In this latter case other configurations of the LC-state
can be achieved; for example splay or twisted nematic configurations, examples of both cases
are depicted in figure 3.5.
Figure 3.5 Example of different alignment configurations. a) planar or homogeneous; the molecules lay parallel
to the substrate b) homeotropic; perpendicular alignment, c) splay; the alignment changes from planar at the
bottom surface to homeotropic at the top surface; this change is gradual, and d) twisted nematic; a planar
alignment but the director is rotated by 90° over thickness of the layer.
Molecules with a liquid crystalline phase have anisotropic properties, which differ in the
direction along the director or perpendicular to the director. Due to the long distance order, the
resulting material will also have these anisotropic properties. Two examples of anisotropic
properties present in LC-materials are the refractive index and the dielectric constant. The
anisotropy in refractive index can be described by the birefringence (n ) and is defined as
e on n n (3.10)
where en (extraordinary refractive index) and on (ordinary refractive index) are the refractive
indices of the material parallel and perpendicular to the director respectively.
The LC-phases in thermotropic LC-materials all have a temperature range where they
occur.[219] To maintain the desired order of the molecules outside this temperature range, the
molecules must be fixed. This ‘freezing’ can be achieved by polymerizing the molecules in their
a) b)
c) d)
Spatial distribution of emitted photons from dichroic dye ensembles
49
LC-phase. The most common method to polymerize LC-materials is by photo polymerization
of mostly (meth)acrylate functionalized molecules, also known as reactive mesogens. In photo
polymerization of reactive mesogens, ultraviolet(UV)-light is used to split a photo-initiator
which forms a reactive species, often a radical, but also ions are used. The radical attacks the
double bond of the (meth)acrylate group of the LC-material, forming a new radical on the LC-
molecule. This new radical attacks another reactive mesogen and in this way a polymer is
formed. A reaction scheme is depicted in figure 3.6.
Figure 3.6 Reaction scheme of photopolymerization of (meth)acrylate mesogens, where R1 is a hydrogen or a
methyl group and R2 is the core part of the mesogen.
The reactive mesogens can have one or multiple (meth)acrylate groups. A combination of
mesogens with one or two functional groups is used to make a network of the LC-material,
which keeps the LC-properties upon cooling and heating. A polymer made from mesogens
with only one functional group will form a LC side chain polymer. This latter polymer will still
exhibit a change of properties upon temperature change and have, for instance, a melting
temperature where the order disappears.[220] Examples of a LC-network and a LC side chain
polymer are depicted in figure 3.7.
Figure 3.7 Example of a LC side chain polymer (left) and a LC-network (right)
3.3.2 From molecular dichroism to macroscopic dichroism
Aligning dichroic dye molecules in a liquid crystalline matrix will cause the material to
exhibit macroscopic dichroism. Both absorption of incident light and re-emission of this light
by an aligned dye-liquid crystal guest-host system (aligned dye ensemble) will be influenced by
several factors: 1) the degree of order of the alignment of the dye molecules, 2) the angle of the
Chapter 3
50
transition dipoles with the director of the ensemble, and 3) the incidence angle and the
polarization of the incident light.
The conjugated plane of an organic dye molecule will not always lay in the same direction
as the molecular axis. In the liquid crystalline host the molecules are assumed to align in the
same direction as the host, but due to the angle between the molecular axis and the so-called
optical axis, this optical axis will have different order. In figure 3.8 a schematic depiction of the
alignment of the transition dipole of a dichroic molecule is shown.
Figure 3.8 The alignment of the transition dipoles of dichroic dye molecules in an liquid crystalline host. The
molecular axis (light grey line) of the dye molecule (black rectangle) is at an angle (θ) with the director of the host
( n ). The transition dipole (dark grey line) is not perfectly aligned with the molecular axis but off-set by an angle
δ.
The order of the optical axis is described by van Ewyk et al. [211] who showed that the order
parameter of the optical axis (S2,opt) can be expressed by the order parameter of the host (S2,host)
corrected for the angle between the optical and the molecular axis of the molecule:
22,
2,
2 3sin
2
host
opt
SS (3.11)
To show the effect of the order of the host and the angle between the molecular and optical
axis of the dye on the optical order parameter a plot is shown in figure 3.9 where the optical
order parameter is plotted as function of δ for hosts with different order parameter.
In figure 3.9 it can be seen that the optical order decreases with increasing angle between the
optical and molecular axis of the dichroic dye. If the host has no order ( 2, 0hostS ) the angle
between the optical and molecular axis has no influence on the resulting collective distribution
of the optical axes in the host. This supports the assumption made in the previous section: the
distribution function of transition dipoles is isotropic in an isotropic host, and independent of
the dichroism of the individual fluorescent molecules.
Spatial distribution of emitted photons from dichroic dye ensembles
51
Figure 3.9 The optical order parameter of aligned dichroic dyes as function of the angle between the optical and
molecular axis (δ) for different order parameters of the host material: 0.0 (black), 0.1 (red), 0.2 (blue), 0.3
(dark green), 0.4 (magenta), 0.5 (orange), 0.6 (dark blue), 0.7 (grey), 0.8 (pink), 0.9 (light green), and 1.0
(purple).
3.3.3 Directional emission from planarly aligned luminophores
To calculate the spatial distribution of emitted photons by planar aligned dye ensembles the
same method is used as described in section 3.2 The transition dipole of the dye molecules is
now described by
sin sin
cos
sin cos
(3.12)
since the director of the liquid crystalline host is aligned parallel to the LSC waveguide and the
distribution of the transition dipoles will be described with respect to this director.
The distribution of the dye molecules is assumed to be equal to the distribution of the LC-
molecules and exhibit the same symmetry features with symmetry in rotation around the
director. In liquid crystals, the distribution of molecules around the director is described by a
summation of order parameters, which are in essence Legendre polynomials. If it is assumed
that the dye molecules exhibit the same order as the LCs, the distribution function of the dye
molecules and thus the optical axis of these molecules in the dye ensemble can be described as
[221]
Chapter 3
52
2 2
14 1 (cos )
2n n
n
f n S P d (3.13)
Combining equations 3.6, 3.12 and 3.13 and solving the integrations will lead to the spatial
distribution of emitted photons as a function of the optical order parameters ( 2,optS and 4,optS )
of the transition dipoles. By using the optical order parameter defined in equation 3.11, the
angle between the molecular and the optical axis of the dichroic dye is accounted for.
2, 4,2 2 2 2
22, 4,
2, 4,2 2 2
80 243 3 1 56cos cos sin sin
4 4 4 105 105 105
10 24142
2 105 105 105
60 24212cos sin 2cos
105 105 105
opt opt
opt opt
opt opt
S S
S SI
S S
(3.14)
Using equation 3.14 the corresponding emission profiles for 2,optS equals to 0, 0.2, 0.4, 0.6, 0.8,
and 1.0 (front and top view) are calculated and depicted in figures 3.10 and 3.11. In these
emission profiles it can be seen that with increasing order parameter the light emitted to two
sides and to the top/bottom is increasing, while the light emitted along the director is
decreasing. This leads to a change in emission profile from an oval “egg” shape to a more
“donut”-shape.
Spatial distribution of emitted photons from dichroic dye ensembles
53
a)
b)
c)
Figure 3.10 Spatial distribution functions of photons emitted by planarly aligned dyes with different order
parameters: a) S2,opt=0, b) S2,opt=0.2, c) S2,opt=0.4. On the left the front view, with the director perpendicular to
the page and on the right the top view, with the director from top to bottom. The axis of these emission profiles
have aribitrary units. The units are the same on both axis and the emission originates from the middle of the
profile.
Chapter 3
54
a)
b)
c)
Figure 3.11 Spatial distribution functions of photons emitted by planarly aligned dyes with different order
parameters: a) S2,opt=0.6, b) S2,opt=0.8, c) S2,opt=1.0. On the left the front view, with the director perpendicular
to the page and on the right the top view, with the director from top to bottom. The axis of these emission profiles
have aribitrary units. The units are the same on both axis and the emission originates from the middle of the
profile.
Spatial distribution of emitted photons from dichroic dye ensembles
55
3.4 Spatial distribution of emitted photons from dichroic dye
molecules in a homeotropically aligned liquid crystal host
Homeotropic alignment of liquid crystalline materials can be achieved by using an appropriate
alignment layer. Due to repulsive forces between the alignment layer and the LC-molecules, it is
energetically favourable for LC-molecules to align perpendicular to the alignment layer.
Another method to achieve homeotropic alignment is by using an alignment layer which has
long molecular chains on the surface, which sterically hinder the LC-molecules from aligning
parallel to the surface. Dyes homeotropically aligned will stand perpendicular to the LSC
waveguide leading to preferential emission towards the four edges of the LSC.
To calculate the spatial distribution of emitted photons the transition dipole of the dye
molecules is described as
sin cos
sin sin
cos
and using this in equation 3.6, together with the distribution function of the transition dipoles
around the director is the same as presented by equation 3.13. The spatial distribution of
emitted photons from homeotropic dye ensembles can be described by
22, 4, 2, 4,2 2
80 24 10 2456 14cos 1 2sin
2 105 105 105 105 105 105
opt opt opt optS S S SI (3.15)
A schematic depiction of these emission profiles for S2,opt=0.2, S2,opt=0.4, S2,opt=0.6, S2,opt=0.8,
and S2,opt=0.9 are depicted in figure 3.12. For S2,opt=1.0 the dye molecules will not absorb any
light since the transition dipole for absorption of all molecules is exactly parallel to the direction
of the incoming light and perpendicular to all linear polarizations of this light.
Similar to the emission profiles from planar aligned dye ensembles, the emission profiles from
homeotropically aligned dichroic dye molecules become more like a “donut”-shape as the
optical alignment increases.
Chapter 3
56
a) b)
c) d)
e)
Figure 3.12 Emission profiles from homeotropically aligned dichroic luminophores with different order
parameters (S2,opt): a) S2,opt=0.2, b) S2,opt=0.4, c) S2,opt=0.6, d) S2,opt=0.8, and e) S2,opt=0.9. The axis of
these emission profiles have aribitrary units. The units are the same on both axis and the emission originates
from the middle of the profile.
3.5 Spatial distribution of emitted photons from dichroic dye
molecules in a tilted aligned liquid crystal host
A third alignment of dichroic dyes in a liquid crystalline matrix is tilted. Here the director of the
dye ensemble is at an angle with respect to the waveguide surface. For calculating the emission
profile of tilted dye ensembles, an angle ( ) must be introduced which represents the angle
between the director of the dye ensemble and the normal of the plane formed by the LSC
Spatial distribution of emitted photons from dichroic dye ensembles
57
waveguide. To simplify the calculations, the director is assumed to be coincident to the z-axis.
This means that the incoming and the emitted light, and thus also their polarizations, have to be
rotated with respect to the LSC waveguide. A schematic depiction of all angles in both the
absorption and the emission event may be seen in figure 3.13.
Figure 3.13 Schematic definition of the dipole (black arrow), the incoming light and the emitted light for both the
absorption (left) event as the emission (right) event. In the left figure the black arrow represents the dipole moment
for absorption defined by a zenith ( ) and an azimuthal ( ) angle with respect to the x,y,z axis system (black
lines). In the right picture the emitted photon ( k ) is also defined by a zenith ( ) and an azimuthal ( ) angle
with respect to the x,y,z axis system. The large grey arrow in both pictures represents the direction of the incoming
light which is circularly polarzied (curved black arrow) which is incident at angle ( ) with respect to the director
of the dye ensemble. In the emission event the LSC is depicted by the rectangle and the angle between the LSC
and both the director and the axis system is also descibed by .
When the incidence angle of the incoming sunlight is rotated by , this leads us to:
,
1 0 0 1 11
0 cos sin cos2
0 sin cos 0 sin
ie i i
i
(3.16)
and when the polarization of the emitted light is rotated by :
,
1 0 0 cos cos cos sin sin
0 cos sin cos cos sin sin cos
0 sin cos cos sin
cos cos cos sin sin
cos cos cos sin cos sin cos sin cos sin
sin cos cos sin sin sin cos
fe
cos cos sin
(3.17)
Combining equations 3.16 and 3.17 with equations 3.6 and 3.13 and calculating all the integrals
Chapter 3
58
leads to the spatial distribution function of emitted photons from dichroic dyes aligned by a
LC-material with a tilt angle ( ).
2 2 2
4 2 2 2
3
2, 4 ,2 2
2 2 2
2 2
2
3cos cos 3sin
cos 3cos sin 3cos
6cos sin cos sin sin80 2456
cos cos 1105 105 105
3cos sin sin
sin sin
2cos sin cos sin sin
cos
8
opt optS S
I
4 2
3
2 2
2 2
2 2 2
2
3
2 2
4 2
4sin
cos sin 24 cos sin sin
cos 4sin
28cos sin
cos sin 24 cos
16sin
cos sin 24 cos sin sin
8cos sin cos sin sin
sin 4 cos 4
4sin sin
2, 4 ,
4 2 2 2
2, 4 ,2 2 2
3
10 2414
105 105 105
sin 8cos sin 8cos60 2421
8cos sin sin105 105 105
16cos sin cos sin sin
opt opt
opt opt
S S
S S
(3.18)
Schematic depictions of the emission profiles from perfectly aligned tilted dye arrays (S2,opt=1)
with different tilt angles are depicted in figure 3.14. For a tilt angle of 90 degrees S2,opt=0.9 is
depicted since the dye molecules will not absorb any light if the ensemble has perfect optical
alignment, since the transition dipole for absorption of all molecules is exactly parallel to the
direction of the incoming light and perpendicular to all linear polarizations of this light.
Spatial distribution of emitted photons from dichroic dye ensembles
59
a) b) c)
d) e) f)
g) h) i) j)
Figure 3.14 Frontal views of the emission profiles from perfectly aligned (S2,opt=1) dye ensembles with different
tilt angles: a) 0°, b) 10°, c) 20°, d) 30°, e) 40°, f) 50°, g) 60°, h) 70°, i) 80°, and j) a dye ensemble with a
tilt angle of 90° with an order parameter S2=0.9 The axis of these emission profiles have aribitrary units. The
units are exactly the same on both axis and the emission originates from the middle of the profile.
The pictures in figure 3.14 demonstrate that for optically perfect aligned dye ensembles the
“donut”-shaped emission profile does not change drastically in shape with changing tilt angle,
but it rotates by the tilt angle.
Chapter 3
60
3.6 Conclusions
A theoretical model is presented predicting that dichroism in absorption and emission of
organic dye molecules results in non-isotropic emission from an isotropic ensemble of these
dichroic dye molecules when illuminated with a collimated light source. Aligning dichroic dye
molecules in a liquid crystalline host leads to a change in spatial distribution of the emitted
photons. No matter the alignment direction of the liquid crystals, the emission profile from the
dye ensemble changes from “egg-shape” to “donut”-shape around the director of the liquid
crystal material, as the optical order is increased from complete isotropic to a perfect alignment.
The aligned dye ensembles emit more photons perpendicular to the director of the liquid
crystalline host. Using the model presented in this chapter it will be possible to manipulate the
spatial distribution of the photons in the waveguide of an LSC by changing the optical order of
the system or by changing the direction of the director of the liquid crystalline host.
4 Emission from planarly aligned
dichroic dyes2
2 Also published in:
P.P.C. Verbunt, A. Kaiser, K. Hermans, C.W.M. Bastiaansen, D.J. Broer and M.G. Debije,
“Controlling light emission in luminescent solar concentrators through use of dye molecules
aligned in a planar matter by liquid crystals”, Advanced Functional Materials, 19, 2714-2719,
2009.
Chapter 4
62
4.1 Introduction
Planar alignment of dye molecules has been described previously by a several research groups
(for example [208-211]). This previous work focused on measuring both macroscopic
dichroism in absorption and fluorescence. In the model presented in chapter 3 the optical order
parameter in absorption is the leady parameter. Entering this parameter in the emission profile
equation leads to a spatial distribution of photons from a dye ensemble with this specific optical
order. The model can also be used to predict the dichroism of the emitted light. In this chapter
the model as presented is validated by comparing the dichroism in fluorescence measured in the
work of van Ewijk et al. [211] In addition, the ratio in edge emission from two orthogonal edges
in an LSC with planar aligned dichroic dye molecules is measured and compared to the values
calculated using the model. Furthermore, the effect of the addition of mirrors on the edges of
the LSC and the addition of a scattering layer underneath the LSC on this edge output ratio is
measured, to determine if the spatial distribution of photons is maintained or if these additions
to the LSC will randomize the photons to a large extent.
4.2 Dichroism in fluoresencence Experimental results by van Ewyk et al. [211] are compared with results calculated with the
model presented in the previous chapter. Van Ewyk et al. describes experimental results from
aligned dye ensembles where the optical order parameter is compared with the fluorescence
parameter, the latter is described by:
2,
2fl
F FS
F F (4.1)
where F and F are the fluorescence intensities with the polarization parallel and
perpendicular to the director of the dye ensemble, respectively. For these measurements the
polarization for the incident light had a 45° angle with the director.
In the present model both the polarizations of the incoming and the emitted photons have
been varied to match the experiments of van Ewijk et al.. So the polarization of the incident light
becomes:
11
12
0
ie and the polarization of the emitted light is defined as
,
0
1
0
fe for
calculating the intensity of the light polarized parallel to the alignment direction ( I F ) and
,
1
0
0
fe for calculating the intensity perpendicular to the alignment direction ( I F ).
The results of van Ewyk et al. are not influenced by reabsorption and re-emission and can be
directly compared to the model presented in this thesis and are shown in table 4.1.
Emission from planarly aligned dichroic dyes
63
Table 4.1 The theoretical and experimental fluorescence order parameters from aligned dye ensembles
Optical order
parameter
(Absorption, 2,optS )
[211]
Fluorescence
order
parameter
(exp.) [211]
Fluorescence
order parameter
(theor.)
0.42 0.34 0.33
0.43 0.39 0.34
0.44 0.37 0.35
0.58 0.50 0.48
0.63 0.56 0.53
0.69 0.60 0.59
4,optS was approximated by using the 2cos calculated from 2,optS . Comparing the
theoretical with the experimental fluorescence order parameter shows that the theoretical
results calculated with the model presented in this thesis are in good agreement with the
experimental results. Van Ewyk et al. concluded that reduction in order parameter between
absorption and fluorescence was caused by depolarization due to a change in direction of the
transition dipole. In the model presented in this thesis the transition dipole is assumed static.
Still, the model also predicts a reduced fluorescence order parameter with respect to the optical
order parameter in absorption. This reduction is caused by the fact that the order parameter in
absorption is only dependent on 2
ie , while the order parameter in fluorescence is a
function of 2 2
i fe e .
4.3 Spatial distribution of emitted photons
4.3.1 Theoretical approach
For planar aligned dichroic dye molecules the emission profile changes with changing optical
order parameter. The change in shape of the emission profile results in a change in ratio
between light emitted to the edges of the LSC parallel to the director and perpendicular to the
director of the liquid crystalline host, for LSCs with planar aligned dichroic dye molecules. A
depiction of the parallel and perpendicular edge is shown in figure 4.1.
Chapter 4
64
Figure 4.1 Definition of the perpendicular edge (left) and the parallel edge (right). The grey cylinders
schematically depict the average alignment of the dichroic dye molecules.
The ratio ( eER ) in output between the parallel and perpendicular edges is calculated using
4
0 0
2
04
,
,
e
d d I
ER
d d I
(4.2)
and the results are shown in table 4.2. 4,optS was again approximated by using 2cos
calculated from 2,optS .
Table 4.2 The calculated ratio between the edge output from the parallel and the perpendicular edge of LSC with
planarly aligned dyes with different order parameters
S2,opt 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
ERe 1 1.02 1.10 1.18 1.27 1.36 1.46 1.57 1.68 1.80 1.93
By aligning the dye molecules planarly, the edge output of the parallel edges can become up to
93% higher than the output of the perpendicular edges if the optical axes of the dye molecules
are aligned perfectly. Nematic liquid crystals are known to have an order parameter ( 2S ) up to
0.7, so by aligning the dyes with a nematic liquid crystal 57% more photons can be sent to the
parallel edge of an LSC with respect to the perpendicular edge if the optical and molecular axis
of the dye molecules are in the same direction. In the rest of this chapter the calculated results
are compared to experimental results and the effect of the addition of a scattering back layer
and mirrored edges is experimentally determined.
4.3.2 Methods
Poly(methyl-methacrylate) (PMMA) plates of 50 x 50 x 5 mm3 were used as substrates
(waveguides). An alignment layer was spun onto some of these waveguides: the layer consisted
Emission from planarly aligned dichroic dyes
65
of either a planar polyimide (Nissan 130 or JSR AL-1051) or 5% poly(vinyl alcohol) (PVA,
Aldrich) in water, spun at 5000 rpm for 45 s for the polyimide or 1000 rpm for 30 s for the
PVA. Spinning was followed by a 1 hour curing period at 100°C in air. The planar polyimide
samples were rubbed by hand on a velvet cloth and the PVA samples with a rubbing machine
(Optron LCBM-6 and a YA18R rubbing cloth (Yoshikama) with a 0.35 mm pile depth, 500
rpm drum speed, and table speed of 0.9 cm/s). Dye solutions were spin coated on the prepared
surfaces at 800–1000 rpm for 30 s; the solutions were composed of 0.1–5% (by weight) dye
molecules Rhodamine B ([9-(2-carboxyphenyl)-6-diethylamino-3-xanthenylidene]- diethyl
ammonium chloride) (Aldrich), DCM (4-dicyanomethylene-2- methyl-6-(p-
dimethylaminostyryl)-4H-pyran) (Aldrich), or the coumarin dye (see figure 4.2), in a mixture of
photo reactive liquid crystals RM257 and RM82 (blended in a 4:1 weight ratio; both from
Merck; structures shown in figure 4.3), 1% photo initiator (Irgacure184, Ciba) out of a 50:50
solution of 1-ethoxy-2- propyl acetate and propylene glycol methyl ethyl acetate or xylene (all
from Aldrich). The samples also had 1% surfactant (2-(N-Ethylperfluoro-octanesulfamido)ethyl
acrylate) added to promote planar alignment at the air/liquid crystal interface. Isotropic samples
were spun from a similar solution on blank PMMA plates or from a solution of dye mixed in
dipentaerythritol pentaacrylate (Polysciences) with 1% photo initiator and 20% dried methyl
methacrylate (Aldrich) as solvent. Samples were placed on a hot plate at 80°C for about 15s (the
liquid crystal mixture was optically confirmed to be in the nematic phase by placing them
between cross polarizers at 45° with respect to the direction of the polarizers) and then photo
cured for 10 min under nitrogen by a 360 nm UV lamp to form a cross-linked, solid film. Both
the rubbed polyimide and PVA alignment layers induced a planar alignment of the liquid
crystals in these systems.
Figure 4.2 Chemical structure of the coumarin yellow dye.
The thickness of the liquid crystalline layers was on the order of 3–6 µm. Absorption
spectra of all waveguides were recorded on a Shimadzu UV- 3102 PC spectrometer with
polarized light directed both parallel and perpendicular to the rubbing direction of the
polyimide. The absorbance of the samples reported refers to the absorbance measured at the
main peak.
Chapter 4
66
Figure 4.3 Chemical structures of RM257 and RM82. The difference between the two molecules is the length of
the alkyl-spacer between the acrylate group and the mesogenic unit which is 3 and 6, respectively.
Figure 4.4 The experimental setup for measuring emission intensities at the edges of waveguides with a scattering
layer underneath, which is prevented from being in optical contact with the waveguide by an air gap.
The emission from the edges of the waveguides was measured by an SLMS 1050
integrating sphere (Labsphere) equipped with a diode array detector (RPS900, International
Light). The LSCs were placed in a custom-made sample holder with a black background with
the alignment direction of the polyimide parallel and perpendicular to the entry port of the
sphere and exposed to a collimated light source from a 300 W solar simulator with filters to
approximate the 1.5 AM (global) solar spectrum (Lot-Oriel), located at a distance of about 15
cm from the waveguide (see figure 4.4 for the experimental setup, without the scatterer or
mirror underneath the LSC waveguide). Edge emissions from the waveguides were determined
Emission from planarly aligned dichroic dyes
67
by placing the samples on a horizontal stage with only one edge of the waveguide entering the
integrating sphere. Illumination was constant over the whole surface by the solar simulator.
A correction for the small (~10%) polarization anisotropy in the solar simulator emission
was made for the outputs of the samples. Light output spectra and intensity from the emission
edge of the sample were recorded. Total output was determined by integrating the recorded
spectra over the range of 350–750 nm. Emission from all four edges of the samples was also
measured with a separate rear white scattering layer (a square of thick paper spray-painted
white, which acted as a Lambertian reflector). The output measurements were repeated after
application of a silver reflective foil (Scotchcal, 3M) to each individual edge of the isotropic and
aligned waveguides, with three edges of the same waveguides on a black background, and with
mirrors attached to three edges on a separate white scattering layer.
4.3.3 Experimental results and discussion
The alignment of the dye molecules is measured by determining the absorption from the
sample with linear polarized light parallel (peak absorption: A ) and perpendicular (peak
absorption: A ) to the director of the LC-material (rubbing direction). [211,222] From this
measurement the optical order parameter is calculated using:
2,
2opt
A AS
A A (4.3)
Representative absorption spectra of waveguides with each of the three dyes used as
luminophore in planar aligned LC-material, Rhodamine B, DCM, and the coumarin (see
Experimental Section) in the aligned LC-host material are given in figure 4.5.
Perfect uniaxial order of the absorbing dipoles of the fluorescent dyes would be described
as 2, 1optS , and a completely isotropic system would correspond to 2, 0optS , analogous to the
order parameter of a LC-material. Rhodamine B showed no dichroism, despite the preservation
of order of its liquid-crystal host as evidenced by its monolithic, birefringent properties. Optical
order parameter values for the DCM samples in the liquid-crystal matrix were determined to be
between 0.2 and 0.45. The coumarin dye displayed higher values for the optical order parameter
(up to 0.6).
Chapter 4
68
Figure 4.5 Absorption spectra of the coumarin (green), DCM (orange), and Rhodamine B (purple) in a planar
reactive mesogen with incident light polarized parallel (solid lines) and perpendicular (dashed lines) to the rubbing
direction. The two curves for rhodamine B are coincident.
Light emissions (both spectrum and power) from the edges of the waveguide were
determined using the equipment setup shown in figure 4.4. The light energy output from the
PMMA waveguide edge parallel to the polyimide rubbing alignment direction ( E ) was greater
than the light energy output from the edge perpendicular to the alignment direction ( E )
except for samples containing Rhodamine B as dye: see figure 4.1 for a definition of E land
E .
The relative outputs of the sample edges were described by considering a ratio of the
outputs E to E . For samples with an optical order parameter equals 0 (isotropic), the choice
of the ‘parallel’ and ‘perpendicular’ edges was arbitrary; such was the case for the samples
containing Rhodamine B. Variation in emission across the four sample edges in the isotropic
waveguides is 5%, as was the variation between emissions of the paired opposing edges of the
aligned samples. A plot of the output ratio as a function of the optical order parameter for the
three different dyes may be seen in figure 4.6. The different optical order parameters for the
same dye come from samples with a different rubbing intensity, so the order parameter of the
host is changed.
Emission from planarly aligned dichroic dyes
69
Figure 4.6 Ratio of the total output from the edge parallel to the polyimide rubbing direction to that from the
edge perpendicular to the rubbing direction ( eER ) as a function of the optical order parameter (S2,opt) for three
different fluorescent dyes in a reactive mesogen: Rhodamine B (black squares), DCM (dark grey circles), and the
coumarin dye (light grey triangles) measured experimentally. The calculated output ratio is displayed by the black
line.The data points of Rhodaime B containing samples are underneath the results of DCM containing samples
with an optical order parameter of approximately 0.
There is a general increase of the emission ratio with increasing optical order parameter for
fluorescent dyes aligned by a nematic LC-material. For the coumarin dye the edge output
parallel to the alignment direction was 50-60% higher than the edge output perpendicular. If
the experimental results are compared with the theoretical values (solid line in figure 4.6 and in
table 4.2), the theoretical values are seen well within the range of the experimental results. It
appears the measurements with DCM as luminophore generally show a smaller ratio in edge
output than calculated, while the measurements with the coumarin dye as luminophore show a
larger ratio, but differences are relatively small, giving support to the calculations. This suggests
that if the dye were to be aligned in a more ordered liquid crystal and if the optical axis would
be equal to the molecular axis of the dye molecule (that is the optical order parameter
approaching 1), the ratio in output could reach ~2. Aligning dyes in a planar manner in an LSC
would be a viable method two reduce the number of photovoltaic cells to two or even one,
without losing too much of the energy emitted from the two non-covered edges.
Chapter 4
70
4.4 Application of silver mirrors or a scattering background
to planarly aligned dichroic dyes in LSCs
To prevent emission from edges of the LSCs not covered by photovoltaic cells in LSCs, it is
common to attach mirrors.[49,74] Reflected light will then return into the waveguide with the
hope that it will emerge from a more desirable edge, that is, the emission edge. Silver mirrors
were attached to three edges of the planar waveguides and the outputs recorded with the
samples on a black absorbing background to determine whether the planar alignment could also
demonstrate enhanced edge emission from the parallel edge when compared to the
perpendicular edge after mirror attachment. The results of the measurements on the
waveguides with and without three mirrored edges are shown in figure 4.7.
Figure 4.7 Ratio of the output from the edge parallel to the polyimide rubbing direction to that from the edge
perpendicular to the rubbing direction ( eER ) as a function of the optical order parameter (S2,opt) , comparing
samples with three attached edge mirrors (open symbols) with samples that have no mirrors (solid symbols) for
Rhodamine B, DCM, and the coumarin dye.
Adding side mirrors to three edges of the planar samples reduced the measured anisotropy
(compare the filled and open symbols in figure 4.7). Light that normally would have escaped
the ‘side’ edges instead is redirected and exits the emission edge, resulting in a mix of E and
E light being emitted from the single edge. However, even after application of mirrors to
three edges, 25–30% more light for the parallel emission edge than the perpendicular in aligned
coumarin samples is obtained. The increase in edge emission after application of the three
mirrors to the uncovered edges is depending on the peak absorbance of the LSC and the dye
Emission from planarly aligned dichroic dyes
71
used. For DCM and the coumarin dye the energy leaving the edge is doubled if there are three
mirrors attached to the uncovered edges for LSCs with a peak absorbance of ~0.5.
White scattering layers are often used in conjunction with the LSC to enhance the spectral
response of the system [74,75,171,173] and increase the light output from the edge of an LSC
waveguide. The white scattering layer essentially serves two purposes: first, it may return light
that was not initially absorbed by the dye materials back through the sample, allowing a second
opportunity for absorption. Second, the scattering layer may redirect incoming light that could
not normally be absorbed, and a fraction of this light will be redirected in such a way as to
reach the exit edge of the waveguide. This is only the case when the photons reach the scatterer
close to the waveguide edge and the pathlength of the scattered photon is small than the
absorption length of the photon in the waveguide. The effect of the scatterer will therefore be
smaller for larger size LSCs.
Waveguided light is prevented from ever encountering the scatterer because an air gap
between the waveguide and the scattering elements has been employed, and thus the preferred
direction of dye emission is maintained. The scattered-in light, however, will be randomly
distributed with equal probabilities of being emitted out of any edge. Thus, there is an apparent
decrease in the output ratios of the edges, but the absolute differences in emissions remain
intact.
Figure 4.8 Ratio of the output from the edge parallel to the polyimide rubbing direction to that from the edge
perpendicular to the rubbing direction ( eER ) as a function of the optical order parameter (S2,opt) for planar
samples with a black absorbing background (solid symbols) and with a white scattering background (open
symbols) for Rhodamine B, DCM, and the coumarin dye.
Chapter 4
72
To evaluate the effect of adding a separate scattering layer to an aligned sample, the edge
outputs were determined using the aligned waveguides on top of a separate white scattering
layer, and the results on the edge output ratio are shown in figure 4.8. The increases in the
absolute emission energies were similar to the results described previously. [173]
After application of a scatterer, the ratio between the edge output of the parallel and the
perpendicular edge is decreased and it seems that this ratio levels out at approximately 1.25.
The extra light reaching the edge of the waveguide coming from the scatterer is more randomly
distributed and thus equal for all edges, which reduces the relative difference between the two
edges.
4.5 Conclusion
The model presented in chapter 3 for calculating the emission from planarly aligned dye
ensembles is compared to and supported by experimental work. There can be a significant
increase in light output from a single edge of a luminescent solar concentrator with dye
molecules aligned parallel with the emission edge compared to the perpendicular edge. In
experiments, the parallel edge emission direction is favoured over emission from the edge
perpendicular to the dye alignment by up to 60% for a relatively good alignment of the optical
axis of the dichroic dye molecules. Theoretical calculations of the anisotropy in emission match
well with these experimental results. Extrapolating the theoretical results to a higher optical
order suggests that using LC-phases with higher order, like smectic phases, or use of dyes that
show increased coincidence of the transition dipoles and molecular axis, could increase the ratio
in edge output between the parallel and the perpendicular edge to almost 2. This shows that the
macroscopic alignment of dichroic dyes is an interesting method to manipulate the distribution
of emitted photons in the LSC waveguide.
Application of either silver mirrors to the waveguide edges or a white scattering
background to the rear of the LSC demonstrates a reduced relative difference between the
parallel and perpendicular edges. However, even after application of the silver mirrors or the
white scatter there is still a difference in edge output between the parallel and the perpendicular
edge with respect to the alignment direction. This shows that there is still an effect present
from the dye alignment, which could lead to LSCs using reduced number of PV-cells.
5 Surface loss in luminescent solar
concentrators3
3 Also published in:
M.G. Debije, P.P.C. Verbunt, B.C. Rowan, B.S. Richards and T.L. Hoeks, “Measured surface
loss from luminescent solar concentrator waveguides”, Applied Optics, 47 (36), 6763- 6768,
2008.
Chapter 5
74
5.1 Introduction
Apart from the incapability of the luminophores that have been used in LSCs to absorb a large
fraction of the incident sunlight, surface loss is one of the most important loss mechanisms of
LSC as described in chapter 2. Surface loss is determined by the trapping efficiency of emitted
photons in the waveguide and by the number of photon recycling events of photons in
waveguide mode. In this chapter a theoretical approximation of the trapping efficiency of
emitted photons is given for isotropic a dichroic emitter in an isotropic host. The effect of
photon recycling events is shown and furthermore the surface loss of an LSC with BASF
Lumogen F Red 305 (hereafter referred to as Red 305) as a luminophore is measured.
5.2 Theoretical approximation
Photons that reach an interface between two materials with different refractive indices are
refracted. Snell’s law described the relation between the angle of the incident photon (1 ) and
the transmitted photon (2 ) as function of the refractive indices ( 1n and 2n , respectively) of
the materials as:
1 1 2 2sin sinn n (5.1)
If 1 2n n light will refract away from the normal to the interface, when light is transmitted
through the interface. The angle of incidence which leads to refracted light with an angle of 90°
is called the critical angle (c ).
Incident light with an angle equal or larger than this critical angle will be totally reflected
by the interface. So in an LSC only those emitted photons reaching the waveguide-air interface
with angles larger than the critical angle will be total internally reflected, while photons with a
smaller angle with respect to the normal of the interface are lost through the surface. Since the
LSC is positioned in air the critical angle depends on the refractive index of the waveguide
( waveguiden ) and can be approximated by equation 2.1. In the literature the emission of the dye
molecules has been assumed to be isotropic, and the fraction of emitted light trapped in the
waveguide (trap ) is described as [92]
12
2
1cos 1trap c
waveguiden (5.2)
Standard polymer waveguide materials used in LSCs are PMMA and PC, which have refractive
indices of approximately 1.49 and 1.58, respectively, leading to trapping efficiencies of 0.741
and 0.774.
The amount of photons lost through the surface of an LSC depends on this trapping
efficiency. Every (re-)absorption and sequential emission event, a part of the emitted photons is
lost through the surface equal to 1 trap . When the transport efficiency of the photons
through the host material of the LSC (host ) and the reflection efficiency of the waveguide
Surface loss in luminescent solar concentrators
75
determined by the smoothness of the waveguide surface (TIR ) are assumed to be unity, the
total fraction of the number of absorbed photons that is lost through the surfaces of an LSC
with an isotropic emitter can be calculated by:
1
0
1
x
i isl PLQY trap trap
i
(5.3)
where x is the average number of photon recycling events. In table 5.1 the surface loss of an
LSC with an isotropic emitter with a quantum efficiency of 1.0 is shown for several number of
photon recycling events.
Table 5.1 Theoretical surface loss from LSCs with an isotropic emitter.
x
Surface loss
PC PMMA
0 0.226 0.259
1 0.398 0.451
2 0.534 0.593
3 0.638 0.699
Results from the model presented in chapter 3 showed that the emission from LSCs with
dichroic dyes in an isotropic host illuminated with collimated light normal to the plane of the
LSC is not isotropic. Using this non-isotropic emission profile the trapping efficiency in the
waveguide can be calculated using
2
0
,0 2
0 0
, sin
, sin
c
c
trap
d d I
d d I
(5.4)
where trap is the fraction of the emitted light that is trapped in the LSC waveguide, ,I is
the emission profile, and c is the critical angle for total internal reflection. The 0 in the index
of the trapping efficiency means that this is the trapping efficiency of the light emitted after the
initial absorption and emission event. So only emission is taken into account after absorption of
the incident sunlight and no photon recycling is accounted for. The initial trapping efficiency
for a dichroic dye in an isotropic host is 0.743 in PC and 0.708 in PMMA.
Since the emission profile depends on the distribution of the light that is absorbed, the
profile will change every time a photon is reabsorbed and re-emitted. This leads to an increase
in trapping efficiency each time the photon re-absorbed and re-emitted. The surface loss is in
the case of a dichroic dye described by:
Chapter 5
76
( 1),0 ,( 1) ,
1
1 1
x
isl PLQY trap PLQY trap i trap i
i
(5.5)
where the index of the trapping efficiency denotes the number of the photon recycling events
which redistribute the photons (0 denotes the initial absorption and emission of incident
sunlight).The incident light is considered to be collimated, but the emitted light that will be re-
absorbed will have a distribution that upon increasing number of photon recycling events
becomes increasingly isotropic. For isotropic incident light the emission from isotropically
distributed dichroic dyes will be isotropic as well. This means that with each photon recycling
event the trapping efficiency will become closer to the trapping efficiency of an isotropic
emitter.
In figure 5.1 the surface loss for anisotropic and isotropic emitters in an isotropic host is
plotted as function of the number of photon recycling events, where the trapping efficiency is
kept constant at 0.743 and 0.708 for the dichroic dye in PC and PMMA respectively.
Figure 5.1 Surface loss from LSCs as function of the average number of photon recycling events for PC (grey)
and PMMA (black) waveguides containing isotropic emitters (solid lines) or dichroic dyes (dotted lines). The
lines of isotropic emitters in PMMA (solid black) and dichroic emitters in PC (dotted grey) are almost
coincident.
The surface loss for the dichroic dye is slightly overestimated and will become larger with
increasing number of photon recycling events. The difference between the surface loss from
polycarbonate LSCs containing isotropic emitters and the dichroic dyes has a maximum of 5%
for any number of photon recycling events. The trapping efficiency of the model using dichroic
Surface loss in luminescent solar concentrators
77
dyes will increase with increasing number of photon events, which will lead to a smaller
increase in surface loss with increasing number of photon recycling events, which will bring the
results from the two models closer.
The surface loss from the dichroic dyes (Red 305), a perylene perinone (figure 2.4t) or a
coumarin (figure 4.2)) as dye in an isotropic host is determined experimentally using a newly
developed measuring method.
5.3 Methods
A number of filled waveguides were produced by Sabic Innovative Plastics (Bergen op Zoom,
the Netherlands) by injection moulding polycarbonate (PC) (n=1.586 at 587 nm) mixed with
various concentrations of the fluorescent dyes Red 305 (BASF, figure 2.4r), perylene perinone
(figure 2.4t) or a coumarin derivate (figure 4.2) or polymethyl methacrylate (PMMA) (n=1.49)
mixed with one concentration of Red305 into plates 50 × 50 × 3 mm3. Surface topped
waveguides were produced by spin casting solutions of Red305 in PC from THF or Red305 in
a penta-acrylate (Polysciences)/MMA (Aldrich) 80∶20 blend containing 1% photo initiator
(Irgacure 184, Ciba) on top of either clear PC (Sabic Innovative Plastics) or PMMA (Plano
Plastics) plates at 1000 rpm for 30s. The penta-acrylate systems were exposed to ultraviolet light
in an N2 atmosphere to crosslink the system. The thicknesses of the dye layers were measured
by a Zoomsurf 3D interferometer (Fogale) and were about 3 μm for the PC layers and 15–30
μm for the penta-acrylate/MMA layers.
The absorbance for all samples was determined using a Shimadzu UV-3102
spectrophotometer: the reported peak absorption values in this thesis refer to the absorbance at
the peak of the main absorption band.
Bottom surface emissions (that is, emission from the surface opposite the incident light
source) were determined by placing the 5 × 5 cm2 waveguides against the entry port of an
integrating sphere equipped with a SLMS LED 1050 light detection array (Labsphere), with a
blank waveguide used as the reference; see figure 5.2 for a depiction of the experimental setup
(in this measurement no filters are place between the entry port of the integrating sphere and
the sample). The samples were exposed to the light of a 300 W solar simulator with filters to
approximate the 1.5 air mass (AM) (global) solar spectrum (Lot-Oriel). To reduce the
background spectra from the source light, a stack of narrow pass band filters centered at 670,
710, and 750 nm were placed between the source and the sample to filter out these longer
wavelengths. The illumination area was limited to an approximately 2 cm diameter spot in the
waveguide center.
Top surface loss measurements (that is, from the surface being illuminated by the light
source) were derived from emission data taken using an Autronic DMS 703 (Melchers GmbH)
together with a CCDSpect- 2 array detector (CCD-Camera). The LSC samples were placed in a
Chapter 5
78
custom-made sample holder and exposed to a uniform light source located at a distance of
about 11 cm. Output spectra were recorded for surface emissions from 30° to 70° with respect
to the normal of the waveguide surface for both sides of the waveguide. See figure 5.3 for a
depiction of the experimental setup. The integrated outputs were determined for both “top”
and “bottom” sides of the waveguide, and a ratio of top/bottom emission was obtained. This
ratio was compared to the absolute measurements taken for the “bottom” surface from the
integrating sphere, and from this the “top” emission was calculated.
Figure 5.2 The experimental setup for measuring the surface emission from luminescent solar concentrators. The
incoming light from the solar simulator is filtered by reflectors placed in front of the sample to reduce the
background light in the emission part of the spectrum. The effect of the reflectors on the surface emission is
determined by placing the reflectors between the sample and the entry port of the integrating sphere.
Figure 5.3 The experimental setup for measuring the top surface loss relative to the bottom surface loss. The
sample is illuminated by a light source at normal incidence. The emission from the surfaces is measured by the
CCD-camera at angles between 30° and 70°.
5.4 Results and discussion
Most recent research used LSCs containing Red 305 as the organic dye.[79] This dye is used to
Surface loss in luminescent solar concentrators
79
measure the fraction of the absorbed energy and photons lost through the surfaces of an LSC.
The normalized absorption and emission spectra of Red 305 were measured and depicted in
figure 5.4. Filled waveguide surface emissions were determined by placing the 5 × 5 cm2
waveguides against the opening of the integrating sphere and illuminating the top surface of the
waveguide with a solar simulator. Using a blank waveguide, we could determine the amount of
light transmitted and reflected by the waveguide itself. Absorption was essentially zero through
this small thickness of the waveguide, and the reflection from PMMA was measured to be
around 8%, and from PC about 11.5% throughout the visible range.
Figure 5.4 Normalised absorption (black) and emission (grey, excitation at 520 nm) spectra of Red 305.
Samples containing Red305 dye were then placed in the measurement position and filters
placed between the light source and the sample were used to reduce the background light in the
emission part of the spectrum where placed between the light source (figure 5.2) and the
sample, and the results recorded. Representative spectra of the blank and a filled waveguide
may be seen in figure 5.5. The blank spectrum shows the transmission spectrum of the light
from the solar simulator through a blank polycarbonate waveguide. Light with wavelengths
above ~600 nm are partially filtered by reflectors place between the light source and the
waveguide. The transmission spectrum through the dye filled waveguide shows a loss of
transmission in comparison to the transmission through the blank waveguide in the part of the
spectrum where the dye molecules absorb the incident light and some extra light leaving the
surface of the waveguide in the part of the spectrum where the dye emits light. By subtracting
the results of the blank waveguide measurement from the results of a filled waveguide, we
obtained the spectra seen in figure 5.6.
Chapter 5
80
Figure 5.5 Examples of measured spectra of light coming from the bottom side of waveguides illumninated from
the top containing no dye (black) and Red 305 (grey)
Figure 5.6 Result of subtracting the spectrum obtained from blank polycarbonate waveguide from the emission
spectrum of PC waveguides containing dye (peak absorbance of 2.5). Integration over region I gives the power
absorbed by the dye and integration of region II gives the power emitted by the dye through the bottom surface.
Surface loss in luminescent solar concentrators
81
The fraction of the light absorbed by the dye that is subsequently emitted from the surface
was determined by integrating the emission spectra (the light emitted into the sphere from the
surface of the waveguide, or the positive region of the curves in figure 5.6), roughly 600 nm to
750nm, depicted as region (II) and dividing by the light absorbed by the dye (determined by
integrating over the negative region of the spectra of figure 5.6, roughly 400–600nm and
depicted as region (I).
Measurements of the surface-topped samples were done in a similar manner, with the dye
layer situated on the top side of the waveguide, facing the light source.
Measurements using the Autronic system (described in section 5.2 and figure 5.3)
established the ratio of top to bottom emission from the waveguide surfaces. Once the ratio
was determined, the absolute emission from the top surface was calculated. Several samples
were measured for output from both faces by repositioning of the waveguide with respect to
the incoming light in the Autronic setup, and the differences were determined to be less than
4%. The results from these measurements can be seen in table 5.2 and figure 5.7.
Table 5.2 Measured bottom loss and derived top loss in energy and photons from the dye-filled waveguides (PC)
Peak
Abs.
Meas.
bottom
surface
loss
(energy)
Calc.
bottom
surface
loss
(photons)
Loss
Ratio
(Top/
Bottom)
Calc.
top
losses
in
energy
Calc.
top
losses in
photons
Calc.
total
surface
loss in
energy
Calc.
total
surface
loss in
photons
(sl (%))
trap
0.05 28% 36% 0.96 27% 35% 56% 71% 0.743
0.46 21% 26% 1.04 22% 27% 42% 53% 0.743
1.01 20% 25% 1.13 22% 28% 42% 53% 0.743
1.63 18% 23% 1.14 21% 26% 39% 49% 0.743
2.36 18% 23% 1.29 23% 30% 41% 53% 0.743
>4 16% 21% 1.56 25% 32% 41% 53% 0.743
Looking at table 5.2 and figure 5.7, it becomes apparent that there is a considerable loss of
energy from the surfaces of the filled polycarbonate waveguides, and this energy loss is constant
at around 40% except in the lightly doped waveguide. This translates to a loss of 50-53% of the
absorbed photons. At low absorbance, the energy losses are greater than 50%, which translates
into 71% photon loss. Both the absorption and the surface emission are very small in this low
absorbing waveguide, which increases the error in the measurements performed.
Chapter 5
82
Figure 5.7 Calculated total surface energy loss from polycarbonate (filled symbols) and PMMA (open symbols)
waveguides with dye in the waveguide (squares) and in a thin polycarbonate (triangle) or penta-acrylate (circle)
surface layer.
The theoretical model predicted the fraction of the absorbed photons trapped in the
waveguide is 0.743, so 25.7% is emitted in surface loss mode if only the initial absorption of
sunlight is taken into account (thus no photon recycling events). Considering the error in the
experimental data the difference between these models cannot be measured. From the data in
figure 5.1 the number of photon recycling for filled PC waveguides should be around 1.4-1.6
(2.4-2.6 total dye interactions for each photon) for the dichroic dye model and approximately 2
(3 total dye interactions for each photon) for the isotropic emitter model for the experimental
data to match the models.
The number of photon recycling events is determined by the concentration of dye
molecules in the waveguide, but is limited by the degree of overlap in absorption and emission
band. Each photon recycling event results in the photon losing energy. After a number of re-
absorptions the emitted photon will not have sufficient energy to re-excite another dye
molecule and cannot be further reabsorbed. Simulations performed by the Herriot-Watt
University using software based on Monte Carlo simulations [79,223,224] have shown that the
number of dye interactions of each photon is at most 1.8. These simulations were performed
before the paper from Wilson et al. [79] that described the absorption tail of the dye molecules
for low energy photons is longer than earlier believed, thus the number of re-absorptions are
probably higher than the 1.8 predicted. For the isotropic emitted model to match these results
the number of dye interactions should almost double, for the dichroic dye model this number is
lower.
Surface loss in luminescent solar concentrators
83
The surface loss of the filled PMMA waveguide measured is, as predicted, slightly higher
than a filled PC waveguide with comparable peak absorbance as a results of the lower refractive
index of PMMA, although the difference is relatively small (44,5% in compared to 42% energy
loss).
It can also be seen that for high O.D. LSCs the amount of light lost through the top
surface of the waveguide is greater than the amount of light lost through the bottom surface,
while the emission profiles show an equal amount of light emitted towards both surfaces. This
is discrepancy is probably also caused by photon recycling. In LSCs with a large absorbance the
intensity of the incident light has a large gradient through the thickness of the waveguide, since
most of the light is absorbed in the top part of the waveguide. This leads to a larger average
pathlength of an emitted photon towards the bottom surface in comparison to the photons
emitted towards the top surface and so is the number of photon recycling events. Each photon
recycling event the light is redistributed and photons have a chance to be emitted in waveguide
mode. This leads to a reduced surface loss through the bottom surface.
The experimental surface loss of the thin dye layer LSCs shown in figure 5.7 is smaller
than the surface loss of the filled waveguides with identical peak absorbance. This is probably
caused by the increase in quenching, since the concentration in these samples is much higher
than in the filled waveguides. Work by Tsoi et al. showed that for LSCs with a thin film
containing dye molecules part of its luminescence is quenched, even at low absorbance. In all
previous results and calculations a quantum efficiency of 100% is assumed for the dye
molecules. [115] In the thin layer LSCs, this is no longer the case. Each re-emission event a part
of the absorbed photons is lost due to non-radiative decay of excited electrons to their
energetic ground state. A reduction in quantum yield will reduce the surface loss as can be seen
in equation 5.5, and the influence of this reduced quantum yield increases with increasing
number of photon recycling events. Therefore, the difference between the surface loss from
filled waveguides and from the thin film samples increases with increasing peak absorbance.
The number of photons emitted from the surface of an LSC depends on the number of
photon recycling events and thus the overlap of the absorption and emission band of the dye
molecule. The surface loss of filled PC waveguides containing other dyes, namely the perylene
perinone dye (Appendix A and figure 2.4t) and a coumarin dye (figure 4.2) have also been
measured. The absorption and emission spectra are displayed in figure 5.8 The results of the
surface loss measurements have been displayed in table 5.3. These results have been corrected
for the quantum efficiency of the dye and samples with a similar peak absorbance are used
(~1.0).
Chapter 5
84
Figure 5.8 The absorption (straight lines) and emission (dotted lines) spectra of the coumarin dye (black,
excitation at 430nm) and the perinone dye (grey, excitation at 620nm)
Table 5.3 Experimentally determined fraction of the emitted photons that is lost through the surface of filled PC
waveguides
Dye Fraction of the emitted
light lost through the
surface
Red 305 0.53 Coumarin 0.46
Perylene perinone 0.325
The fraction of the emitted photons lost through the surfaces of the waveguide is a
function of the dye. In the absorption and emission spectra depicted in figure 5.8 it can be
clearly seen that the overlap between the absorption and the emission band of the Red 305 is
larger than for the other dyes, and that the overlap is the smallest for the perylene perinone dye.
This strongly suggests that photon recycling is the cause of a significant fraction of real surface
loss in comparison to the amount of surface loss calculated from the theoretical trapping
efficiency.
5.5 Conclusions
Calculations show that the trapping efficiency of isotropic and dichroic emitters in an isotropic
host leads to a trapping efficiency of 0.774 and 0.743, respectively in polycarbonate waveguides
and 0.741 and 0.708 in PMMA waveguides. The difference in trapping efficiency between
waveguides with isotropic and dichroic dyes is relatively small, which made experimental
Surface loss in luminescent solar concentrators
85
verification difficult. This trapping efficiency would lead to losses of 25.7% or 29.2% of the
photons emitted by dichroic molecules in an isotropic host, for PC and PMMA, respectively.
Experimentally, LSC waveguides using Red 305 as luminophore demonstrate
approximately 50% of all absorbed photons is lost through the surfaces of the LSCs with a
peak absorbance above 0.2. The difference between the experiments and theory is a result of
photon recycling of emitted photons in waveguide mode by other dye molecules, arising from
the overlap in absorption and emission band of the molecule, which is not included in the
calculations. Using dyes with different degrees of spectral overlap results in different fractions
of photons lost through the surfaces of LSC waveguides, but always higher than calculations
predict because of this omission.
6 Reduction in surface loss by dye
alignment4
4 Also published in:
P.P.C. Verbunt, C.W.M. Bastiaansen, D.J. Broer and M.G. Debije, “The effect of dyes aligned
by liquid crystals on luminescent solar concentrator performance”, in proceedings of the 24th
European Photovoltaic Solar Energy Conference, WIP Munchen, 2009, 381-384.
Chapter 6
88
6.1 Introduction
Taking advantage of the anisotropic emission from dichroic dyes aligned in LC-matrices could
be a method to reduce the surface losses in LSCs. Since the direction of emission from an
ensemble of dye molecules orientated along a director is predominantly perpendicular to this
director, aligning dye molecules in a homeotropic fashion would lead to a preferred emission
towards the four edges of the LSC waveguide. The surface loss from homeotropic and planar
aligned dye ensembles are measured experimentally and compared to the calculated values of
the trapping efficiency. Lastly, surface losses from tilted dye ensembles, which combine the
advantages of planar and homeotropic aligned dye molecules, are calculated.
6.2 Surface loss from planar and homeotropically aligned
dye ensembles
In this section, the surface losses from LSCs are calculated for LSCs with homeotropically and
planarly aligned dichroic luminophores in nematic LC-materials and these calculations are
compared to experimental data from aligned-LSCs using different luminophores with different
degrees of alignment as measured by absorption. In addition, the amount of light leaving the
edge of the LSC waveguide is monitored.
6.2.1 Methods
PMMA plates with a size of 50 x 50 x 5 mm3 were used as waveguides for the luminescent solar
concentrators. Alignment layers were spun on top of the waveguides at 1000 rpm for 30
seconds. For planar alignment, 5% polyvinyl alcohol (PVA, Aldrich) in water and, for
homeotropic alignment, a polyimide (Nissan 5300) were used. After spinning, the alignment
layers were baked at 95°C for 30 minutes (PVA) or 165 minutes (polyimide). The PVA
alignment layer was subsequently rubbed by hand on a velvet cloth with 0.35mm pile depth.
On top of the alignment layer or the bare PMMA waveguide (for isotropic alignment), a
dye layer was spun at 1000 rpm for 30 seconds. For isotropic dye ensembles, 0.1-1.5 wt% dye
(Rhodamine B (Aldrich), DCM (Aldrich) or the Coumarin dye) was added to an 80 wt%
dipentaerythritol pentaacrylate (Polysciences) and 20 wt% methyl methacrylate (Aldrich)
mixture containing 1 wt% photo-initiator (Irgacure 184, Ciba). The dye mixture for the planarly
aligned samples contained 0.1-1.5 wt% dye, in a 4:1 mixture of two liquid crystals, RM257
(Merck) and RM82 (Merck). To this mixture, 1 wt% of the photo-initiator and 1 wt% of a
surfactant (2-(N-Ethylperfluoro-octanesulfamido)ethyl acrylate, to promote planar alignment)
were added. The material was spun from a solution containing 50 wt% xylene. Homeotropic
aligned samples were spun from a solution containing 0.1-1.5 wt% dye, 66 wt% RMM77
(Merck), 33 wt% xylene, and 1 wt% photo-initiator.
After spinning, the waveguides were placed on an 80°C hotplate for 20 seconds (the liquid
crystal mixture was optically confirmed to be in the nematic phase) and then photo cured for 10
Reduction in surface loss by dye alignment
89
min under nitrogen by a 360 nm UV lamp to form a cross-linked, solid film. The isotropic dye
ensembles were photo-cured directly after spinning. Measurements showed that the dye layers
were 3-6 μm thick for the liquid crystal layers and 20-30 μm thick for the isotropic layers.
Absorption spectra were recorded for all samples using a Shimadzu UV-3102 PC
spectrophotometer. The absorption was measured with incident isotropic light and light linearly
polarized both parallel and perpendicular to the rubbing direction. The peak absorbance
reported in this work refers to the measured absorbance at the main peak of the absorption
spectrum.
Emission from the edges of the waveguides was measured by an SLMS 1050 Integrating
sphere (Labsphere) equipped with a diode array detector (RPS9000, International light).
Illumination was provided by a collimated light source from a 300 Watt solar simulator with
filters to approximate 1.5AM (global) solar spectrum (Lot-Oriel) at a distance of about 15 cm.
The samples were placed in a holder, which positions the samples with the edge directly at the
entrance of the integrating sphere. Total output was determined by integrating the recorded
emission spectra over the range of 350-750nm. The collected data were corrected for the small
(~10%) polarization anisotropy of the light source.
Surface emission was determined using the setup described in chapter 5.
6.2.2 Results and discussion
The emission profiles of dichroic luminophores aligned planarly and homeotropically (chapter
3) have been used to calculate the trapping efficiency of emitted photons in the LSC waveguide
using equation 5.4.
Since the experimental data use different matrices for all three alignments (isotropic,
planar and homeotropic), the refractive indices of these matrices are also different. In the liquid
crystal matrix the refractive index is different for each angle of emission. A refractive index of
1.59 was selected as a reasonable average. The results of the calculations are depicted in figure
6.1.
With increasing optical order parameter, the fraction of the light lost trapped in the LSC
waveguide increases if the dichroic luminophore molecules are aligned perpendicular to the
waveguide surface. If the degree of order is high, the fraction of the emitted photons trapped in
the waveguide is more than 85%, and thus less than 15% of all emitted photons are lost
through the surfaces of the waveguide. These results match the results presented by Mulder et al.
[225] who uses a different model.
Chapter 6
90
Figure 6.1 Calculated trapping efficiency for homeotropically (filled symbols) and planarly (open symbols) aligned
dichroic dyes as a function of the optical order parameter. The trapping efficiency for both alignments with no
order is equal and only depicted as an open symbol.
If the dye molecules are aligned parallel to the waveguide surface (planar), the trapping
efficiency increases slightly if the order is increased from completely isotropic to an order
parameter of 0.1. Increasing the optical order parameter above 0.1, the fraction of emitted light
trapped in the LSC waveguide decreases with increasing order parameter. At optical order
parameters higher than 0.4 the fraction trapped photons is lower than for isotropic dye
ensembles. The initial increase in trapping efficiency may be explained by considering figures
3.10 and 3.11. Looking at the frontal view of the emission profile, it is clear that the shape of
the emission becomes more circular as the dye molecules are ordered planarly in comparison to
the completely disordered state (isotropic). So, the ratio of the light emitted towards the top
and the preferred side (perpendicular to the direction of the transition dipole) will evolve from
more light emitted towards the top to equal amounts of light emitted towards the top and
towards the side. In the top view of the emission profiles it can be seen that at low optical order
parameters the emission in the non-preferred emission direction is decreased approximately the
same at all angles of . When the order parameter increases, the decrease in the light emitted
in the non-preferred direction becomes nearly zero at large angles of , while at small this
decrease is less pronounced, due to the “donut” shape of the emission profile. This leads to the
decrease in trapping efficiency at larger order parameters.
Rhodamine B has demonstrated no optical alignment in the planar liquid crystal (see
chapter 4). The order parameter of the host will change by varying the LC-material. However,
Reduction in surface loss by dye alignment
91
since there is little difference in the molecular structure of the planar and homeotropic LCs, it is
assumed in the rest of this work that the degree of alignment of the LC-matrices will be
comparable, so Rhodamine B is assumed to have no optical order in the homeotropic LC-
material either. The results of the surface loss measurements can be seen in figure 6.2.
Figure 6.2 Measured surface loss of absorbed light energy of a Rhodamine B dye-topped LSC with isotropic
(black squares), planar (light grey triangles) and homeotropic (grey spheres) dye alignment
Due to limited solubility of Rhodamine B in the liquid crystalline material, only low
concentrations of this dye could be used for this measurement leading to low absorbance
materials. The difference in surface loss between the three alignments of the dye is very small,
less than the experimental error. Therefore, it may be concluded that there is no difference in
surface losses for the Rhodamine embedded in LCs in different orientations, which could be
expected knowing Rhodamine B has no optical order in the liquid crystal matrix.
For the isotropic thin-film dye ensembles the surface loss increases with increasing peak
absorbance. This is different from the results shown in chapter 3 which studied Red 305 in a
filled waveguide. According to equation 5.3 the surface loss increases with increasing number of
photon recycling events, leading to an increase in surface loss with optical density.
DCM demonstrates reasonable optical alignment in the nematic LC-material used for
planar alignment, reaching optical order parameters around 0.35-0.4 . For this dye, the surface
loss from thin film LSC waveguides is also measured for the three alignments and the results
are depicted in figure 6.3.
Chapter 6
92
Figure 6.3 Measured fractional surface loss of absorbed light energy of a DCM dye-topped LSC with isotropic
(black squares), planar (light grey triangles) and homeotropic (grey spheres) dye alignment
The solubility of DCM in the LC-hosts is much better than Rhodamine B, which led to samples
with a peak absorbance up to 0.6 for homeotropic alignment and almost 1.1 for planar
alignment. For all three alignments, the fractional surface loss generally decreases with increased
peak absorbance. With increasing peak absorbance the probability of re-absorption is increased
as well, and considering DCM has a quantum yield which is below unity (~0.8 in a solid matrix
[110]) the fraction of the absorbed light being lost through non-radiative events is also
increased. Therefore, the total amount of emitted energy is lower than the total amount of re-
absorbed energy, leading to a decrease in both surface loss and edge output efficiency, which
are measured as a function of total absorbed energy.
The fractional surface loss for the planar and isotropic alignments is on the same order,
while there is a reduction in surface loss when the molecules are aligned homeotropically. The
trapping efficiency calculated from the model was similar for planar and isotropic dye
ensembles, which is confirmed by the experimental, even though the experimental data show a
large degree of scatter. The experiments show an 25%-50% decrease in surface loss if the
molecules are aligned homeotropically in comparison to isotropic aligned dye ensembles.
The coumarin dye showed very good optical alignment properties in the planar LC-
material with a measured optical order parameter in absorption of 0.5-0.6. For this dye the
surface loss from thin film LSC waveguides was also measured for the three alignments and the
results are depicted in figure 6.4.
Reduction in surface loss by dye alignment
93
Figure 6.4 Measured fractional surface loss of absorbed light energy of a coumarine dye-topped LSC with
isotropic (black squares), planar (light grey triangles) and homeotropic (grey spheres) dye alignment
Clearly, there are considerable differences in surface losses between the three alignments.
The fraction of absorbed energy leaving the surface of the LSC waveguide after emission is
constant with peak absorbance, comparable with the measurements on the filled Red 305
waveguides presented in chapter 3. About 35% to 40% of all absorbed energy is lost through
the surface for an isotropic dye ensemble. Planar alignment of the dye molecules increases the
surface loss to approximately 55-60%. Aligning the dye molecules homeotropically decreases
surface loss to 5-10%.
Comparing the calculated trapping efficiency and the experimental results from the surface
loss of isotropic dye ensembles of the coumarin dye leads to an average of ~0.7 photon
recycling events. Using these 0.7 photon recycling events and equation 5.5 the theoretical
surface loss for homeotropic dye ensembles can be calculated to be ~20%. This is still higher
than measured experimentally, but in this calculation the trapping efficiency after each photon
recycling event is kept constant, while the trapping efficiency will increase with increasing
photon recycling events. Thus the 20% calculated surface loss is an overestimation.
Furthermore the absorbance of the homeotropic samples is low, which increases the error in
the measurement.
From the surface loss measurements depicted and described above using Rhodamine B,
DCM or the coumarin dye as luminophore, it can be concluded that there are some
discrepancies between theory and experiment, but the general trends are consistent.
Chapter 6
94
The goal of reducing surface losses in LSCs is to increase the emission from the edges of
the LSC waveguide and so increase the optical efficiency of the LSC. From the calculated
emission profiles the total amount of light emitted to the edge of the LSC waveguide was
determined for both planar and homeotropic aligned dye ensembles and the results normalised
by the amount of light in waveguide mode for an isotropic dye ensemble using
2
0
2
0
, sin
, sin
c
c
c
c
aligned
edge
iso
d d I
d d I
(6.1)
where edge is the fractional total amount of light emitted to the edge of the LSC waveguide,
,isoI is the emission profile of an isotropic dye ensemble and ,alignedI is the emission
profile of the aligned dye ensemble. The results of this calculation can be found in figure 6.5.
For planarly aligned dye ensembles the emission from the LSC-waveguide edge increases with
increasing order parameter; surface loss also increases at these high order parameters. The
increased edge emission results from the increased absorption of dye molecules oriented with
excitation dipoles more perpendicularly with respect to the incoming sunlight (assumed to be
normal to the device).
Despite a significant reduction in surface loss and increased emission into waveguiding
modes, the significant reduction in absorption of sunlight results in a decrease in the amount of
light emitted from the edges with increasing order parameter if the molecules are aligned in a
homeotropic fashion. The results in figure 6.5 will change if a number of photon recycling
events is taken into account. In figure 6.6 the relative edge output is plotted for both
homeotropic and planar alignment where 1, 2, or 3 photon recycling events are taken into
account. Again the trapping efficiency is kept constant after each photon recycling event
leading to an overestimation of the surface loss, but in both homeotropic and planar dye
ensembles as the reference (isotropic dye ensembles) the surface loss is overestimated.
Reduction in surface loss by dye alignment
95
Figure 6.5 Calculated normalized edge output for homeotropically (filled symbols) and planarly (open symbols)
aligned dichroic dyes. The edge output has been normalized to 1 for an optical order parameter equals zero
(isotropic). Photon recycling is not taken into account
Figure 6.6 Calculated normalized edge output for homeotropically (filled symbols) and planarly (open symbols)
aligned dichroic dyes after 1(squares), 2(triangles), or 3(circles) photon reccling events. The edge output has been
normalized to order optical parameter equals zero (isotropic) being 1.
Chapter 6
96
In figure 6.6 it can be seen that at low optical order parameters the calculated relative edge
output increases when the molecules are aligned homeotropically. The total amount of
absorbed photons is not decreased drastically in comparison to the isotropic dye ensembles,
while there is some gain in trapping efficiency. The edge output decreases drastically at larger
optical order parameters due to reduced absorption of the incident light. The effect of the
increased absorption from planar aligned dye ensembles is reduced when multiple photon
recycling events are taken into account, since these planar ensembles emit more light towards
the surface at each photon recycling event in comparison to the two other alignments
The optical efficiency of thin-film LSCs using the coumarin dye in all three alignments was
measured experimentally and the results are depicted in figure 6.7. There are no large
differences between them if the peak absorbance is the same. The samples with planar
alignment have a lower dye concentration than the isotropic samples. The homeotropic sample
has an even higher concentration of dye molecules.
Figure 6.7 The optical efficiency of a Coumarin dye-topped LSC with isotropic (black squares), planar (light
grey triangles) and homeotropic (dark grey circles) dye alignment
Aligning dye molecules in a homeotropically initially decreases the amount of energy lost
through the surface, but due to a reduction in absorption of incident light a higher
concentration of dye molecules is necessary, which has the detrimental effect of increasing the
chance of re-absorption of emitted photons. Therefore, in principle a combination of the high
absorption from planar aligned dye ensembles and the reduction of surface loss in homeotropic
dye ensembles could provide a system with optimal performance.
Reduction in surface loss by dye alignment
97
6.4 Tilted dye ensembles
Tilted dye ensembles present an opportunity to combine the advantages of planar alignment
(increased absorption) and homeotropic alignment (reduced surface loss). As suggested by the
name, tilted dye ensembles have a director which has an angle with respect to the waveguide
plane, see figure 6.8.
Figure 6.8 Schematic depiction of homeotropic (1), planar (2) and tilted (3) dye esembles. The black arrow
resembles the director of the ensemble.
Tilted aligned dye ensembles should reduce the path length of emitted photons through
the dye layer, which would decrease the chance of re-absorption while still allowing the
ensemble to emit most of the light towards the edge of the LSC. It is known that there are
smectic LC-phases which exhibit a tilted director, such as smectic C phase. It is also possible to
align nematic LC-materials with a tilt angle by using a mixture of different polyimide materials
to form the alignment layer and the processing conditions of this layer will influence the angle
of the director of the molecules. [226] Using this latter method led to samples where the LC-
molecules exhibited a tilted alignment as can be seen in the scanning electron microscopy
images depicted in figure 6.9.
Figure 6.9 SEM pictures of the cross-section from a tilted dye ensemble. The left and right picture are two
different tilted dye samples. Breaking the layer of LC-material caused fracture lines in the same direction as the
director. Optical behaviour between cross-polarizers also indicate that the samples are birefringent with the optical
axis tilted with respect to the layer.
Chapter 6
98
Unfortunately, heat degradation of the alignment layer itself caused the material to become
absorbing, preventing measurement of the effect of tilted dye ensembles on the emission of
photons and so experimental results were not obtained. Since actual measurements of the
emission from a tilted dye ensemble was not possible, the emission profile is calculated in the
same way as the calculations of the emission profile of isotropic, planar and homeotropic dye
ensembles as described in chapters 3, 4 and the previous sections of this chapter respectively.
Equation 5.4 is now used to calculate trapping efficiency from LSC-waveguides with tilted dye
ensembles described by equation 3.18. The critical angle of the device is dependent on the
refractive index of the matrix in which the dye molecules are embedded. For an LC-host, the
refractive index is dependent on the angle between the photon and the optical axis of the LC.
In the calculations it is assumed that the refractive index is constant at 1.59. The results are
depicted in figure 6.10.
Figure 6.10 Calculated trapping efficiency from LSC using tilt aligned dye ensembles with various optical order
parameters (S2,opt): 0.1 (red), 0.2 (blue), 0.3 (green), 0.4 (pink), 0.5 (orange), 0.6 (navy), 0.7 (brown), 0.8
(light blue), 0.9 (grey) and 1.0 (yellow).
For optically perfect aligned dye ensembles, the trapping efficiency decreases with increasing tilt
angle with respect to the normal to the waveguide (see figure 6.10). All the emission profiles
have a similar donut shape around the director of the ensemble (figure 3.14) and with
increasing tilt angle the amount of photons emitted in waveguide mode decreases. Contrarily to
optically perfect aligned dye ensembles, the emission profiles of optically non-perfect aligned
dye ensembles do not only change in direction but also in shape with changing tilt angle, as can
Reduction in surface loss by dye alignment
99
be seen in figure 6.11.
a) b) c)
d) e) f)
g) h) i) j)
Figure 6.11 Frontal views of the emission profiles from well aligned (S2,opt=0.8) dye ensembles with different tilt
angles: a) 0°, b) 10°, c) 20°, d) 30°, e) 40°, f) 50°, g) 60°, h) 70°, i) 80°, and j) 90°. The axis of these
emission profiles have aribitrary units. The units are exactly the same on both axis and the emission originates
from the middle of the profile.
Chapter 6
100
The change in shape of the emission profile leads to a different tilt angle where the
trapping efficiency is minimal, being 90° for perfect aligned dye ensembles and 40° for weakly
aligned dye ensembles (S2,opt=0.1). The tilt angle where the trapping efficiency is maximal is less
dependent on the optical order parameter (0°-10°).
Figure 6.12 Normalized calculated photons absorbed by the dye molecules in LSCs using tilt aligned dye
ensembles with various optical order parameters (S2,opt): 0 (black), 0.1 (red), 0.2 (blue), 0.3 (green), 0.4 (pink),
0.5 (orange), 0.6 (navy), 0.7 (brown), 0.8 (light blue), 0.9 (grey) and 1.0 (yellow). The amount of absorbed
photons in an LSC with isotropic dye ensembles is set as 1 (black line).
In figure 6.12 the relative absorption of the dye ensembles with respect to an isotropic dye
ensemble is depicted. Generally, the relative absorption increases with increasing tilt angle. The
angular dependency of the absorption increases with order parameter. For small tilt angles the
absorption of the aligned dye ensembles is smaller than the absorption of an isotropic dye
ensemble. Except for very low optical order parameters (S2,opt=0.1-0.2), a tilt angle of at least
between 50°-70° increases the absorption compared to an isotropic dye ensemble.
Both the trapping efficiency and the amount of light absorbed by the dye molecules will
influence the total amount of photons which are exiting the edge of the LSC, if the quantum
efficiency is assumed unity. In figure 6.13 the normalized amount of photons emitted in
waveguide mode is depicted, where the amount of photons leaving the edge from an isotropic
dye ensemble is set at 1. Comparing the amount of photons leaving the waveguide edge and the
absorption of the dye ensembles shows that the absorption is the key factor in the total amount
of photons leaving the waveguide edge, and not the trapping efficiency. The effect is largest
when the optical order is highest, so a perfectly aligned tilted dye system will emit the most light
Reduction in surface loss by dye alignment
101
into waveguide mode.
Figure 6.13 Normalized calculated photons leaving the waveguide edge by the dye molecules in LSCs using tilt
aligned dye ensembles with various optical order parameters (S2,opt): 0 (black), 0.1 (red), 0.2 (blue), 0.3 (green),
0.4 (pink), 0.5 (orange), 0.6 (navy), 0.7 (brown), 0.8 (light blue), 0.9 (grey) and 1.0 (yellow). The amount of
photons leaving the waveguide edge in an LSC with isotropic dye ensembles is set as 1 (black line).
All the calculations performed in this section are constrained to one absorption event with
light incident normal to the device, so no photon recycling. To show the effect of photon
recycling, the relative number of photons leaving the edge of the LSC waveguide is plotted for
the almost perfect aligned systems (S2,opt=0.9) with 0, 1, 2 or 3 photon recycling events in figure
6.14. Again the trapping efficiency if kept constant with photon recycling event.
With increasing photon recycling events the relative amount of photons leaving the
waveguide edge decreases for planar aligned dyes ensembles, while for homeotropic aligned dye
ensembles the relative amount of photons leaving the waveguide edge increases with respect to
isotropic dye ensembles (figure 6.14). The number of photons leaving the edge of the LSC
waveguide decreases with increasing number of photon recycling events (figure 6.15), but this
reduction is the largest for samples with a small tilt angle and smallest for large tilt angle
ensembles. To predict the optimal tilt angle, the exact emission profile after each photon
recycling event should be known as well as the exact number of photon recycling events, but as
can be seen in both the figure 6.14 and 6.15 the optimal tilt angle will be close to planar
alignment. Even though the trapping efficiency is much higher for homeotropically aligned dye,
the increase in absorption by planarly aligned dye molecules will stay the determining factor.
Chapter 6
102
Figure 6.14 Normalized calculated photons leaving the waveguide edge by the dye molecules in LSCs using tilt
aligned dye ensembles with optical order parameter (S2,opt) set at 0.9, including 0(black), 1(red), 2(blue), or
3(green) photon recylcing events. The amount of photons leaving the waveguide edge in an LSC with isotropic dye
ensembles is set as 1 (black line).
Figure 6.15 Number of photons leaving the edge of LSCs with perfect aligned tilted dye ensembles as funcion of
the number of re-absorption events. Tilt angle: 0° (red), 10° (green), 20° (blue), 30° (pink), 40° (yellow), 50°
(light blue), 60° (brown), 70° (purple), 80° (orange) and 90° (dark green). Dichroic dyes in an isotropic host
are depicted with the thicker black line)
Reduction in surface loss by dye alignment
103
The model as presented in this thesis should be implemented in simulation software to
derive the optimal tilt angle, and the results in this chapter should considered as a first step to
predicting the efficiency of LSCs with aligned dichroic dyes.
6.5 Conclusions
Calculations predict that aligning dichroic dyes in a homeotropic fashion increases the trapping
efficiency of the emitted photons in the waveguide, while planar alignment of these dye
molecules decreases the trapping efficiency. The surface losses measured experimentally show
identical trends. Since the number of photon recycling events is not known, theoretical surface
losses are difficult to predict. Since the preferred direction of the dipole for absorption is
parallel to the incident light, homeotropic aligned dye ensembles show drastically reduced
absorption, while absorption is enhanced when the dye molecules are aligned planarly.
Devices using dyes tilted with respect to the LSC waveguide surface were expected to
combine the advantages of homeotropic (preferred emission towards the edges of the
waveguide) and planar (enhanced absorption) alignments. Calculations show that the change in
absorption is the key factor with respect to determining the total amount of light emitted in
waveguide mode by the dye molecules, and directional emission of less importance. However,
no photon recycling is taken into account in any of these calculations, which will probably have
a significant influence on the results discussed in this chapter. Experimentally, it is shown for
the coumarin dye that the effect of aligning the dye molecules has a limited influence on the
optical efficiency of the LSC. Since the number of photon recycling events is dependent on the
dye and the size of the LSC waveguide, the effect of aligning the dye molecules will change if
another dye is used or the size of the LSC changes.
The relative difference in surface loss between large and small tilt angle dye ensembles will
increase with increasing number of photon recycling events. It is not currently possible to
calculate the effect of photon recycling analytically, since the recycling depends on the
polarization and direction of the light within the waveguide, the wavelength of the emitted light
(in principle the overlap between the absorption and emission bands of the dye molecule) and
the spatial distribution of the dye molecules. Generally the effect of multiple photon recycling
events will be the largest for dye ensembles with a large tilt angle, since the surface loss is higher
for these samples. This will probably lead to a decrease in the relative amount of photons
emitted in waveguide mode by these large tilt dye ensembles and an increase for the small angle
tilt dye ensembles. This should lead to a maximum in total photons emitted in waveguide mode
at a tilt angle smaller than 90°, making tilting of dye molecules a viable method to increase the
efficiency of an LSC. The preferred tilt angle will probably be relatively large since the amount
of light absorbed by the dye molecules remains important.
The absorption of incident light calculated in this chapter assumes light at normal
incidence to the surface of the LSC. During the day the position of the sun changes with
Chapter 6
104
respect to the LSC surface and the sun will most of the time be at a larger angle. This will
increase the absorption of the incident light and therefore the number of photons leaving the
edge of the LSC waveguide.
The calculations performed in this chapter, especially the emission profiles (which were
predicted in chapter 3) as function of the angle between the incident light and the director of
the dye ensemble should be considered as a first step in the design of a toolbox to be
implemented in a simulation software package. Such a complete package could be used to
simulate the effect of photon recycling on the directional emission from dichroic dye
ensembles.
7 Organic wavelength selective
reflectors5
5 Also published in:
M.G. Debije, M.P. Van, P.P. C. Verbunt, D.J. Broer and C.W. M. Bastiaansen, “The effect of
an organic selectively-reflecting mirror on the performance of a luminescent solar
concentrator”, in Proceedings of the 24th European Photovoltaic Solar Energy Conference,
WIP Munchen, 2009, 373-376.
M.G. Debije, M.-P. Van, P.P.C. Verbunt, M.J. Kastelijn, R.H.L. van der Blom, D.J. Broer and
C.W.M. Bastiaansen, “Effect on the output of a luminescent solar concentrator on application
of organic wavelength-selective mirrors”, Applied Optics, 49(4), 745-751, 2010
D.K.G. de Boer, C.-W. Lin, M.P. Giesbers, H.J. Cornelissen, M.G. Debije, P.P.C. Verbunt and
D.J. Broer, “Polarization-independent filters for luminescent solar concentrators”, Applied
Physics Letters, 98, 021111, 2011
P.P.C. Verbunt, S. Tsoi, M.G. Debije, D.J. Broer, C.W.M. Bastiaansen, D.K.G. de Boer and C.-
W. Lin, “Increased efficiency of luminescent solar concentrators after application of organic
wavelength selective mirrors”, Optics Express, 20 (S5), A655-668, 2012
Chapter 7
106
7.1 Introduction
The loss of photons through the surface of an LSC waveguide can also be reduced by
application of wavelength selective mirrors. Wavelength selective mirrors transmit light that can
be absorbed by the luminophore, but reflect photons that are emitted, effectively reflecting
surface directed photons back into the LSC. A picture of the working principle is shown in
figure 7.1.
Figure 7.1 The working principle of wavelength selective mirrors. Black photons in the solar spectrum are
transmitted by the reflector and absorbed by the dye molecules. Gray emitted photons are reflected back into the
device.
Wavelength selective reflectors are in principle Bragg reflectors, which consist of repeating
layers of materials with alternating refractive index. At each interface between a low and high
refractive index material a small portion of the incoming light is reflected. If the difference in
distance from which the light is reflected is such that the reflected light from different
interfaces is in phase, the photons will constructively interfere. This means that the period of
the alternating refractive indices determines which wavelength the Bragg reflector reflects. The
condition for this constructive interference is described by:
2 cosm nd (7.1)
where m is an integer denoting the order of interference, is the wavelength of the light that is
being reflected, n is the average refractive index of the multilayer, d is the periodicity of the
multilayer and is the angle of the incident light. From equation 7.1 it can be derived for reflecting light with a certain wavelength incident
normal to a Bragg-reflector, a periodicity
2
dn
is necessary. The light reflected by such a
reflector has a certain spectral bandwidth. The wavelength determined by Bragg’s law (equation
7.1) is the central wavelength of this reflection band. The bandwidth ( ) of the reflection
band is determined by the refractive indices ( 1n and 2n ) of the multilayer stack and this central
Organic wavelength selective reflectors
107
wavelength (0 ) according to:
10 2 1
2 1
4sin
n n
n n (7.2)
Bragg reflectors can be made from stacks of inorganic materials with alternating refractive
indices.[75,199,227] Since the thickness of these layers is very important for constructive
interference of the reflected photons, production of these layers has to be performed with high
accuracy. The production of these inorganic Bragg-reflectors occurs most often by deposition
methods which are very time consuming and expensive to reproduce on large scales.
Chiral nematic liquid crystalline (also known as cholesteric) materials can also act as Bragg
reflectors. [213] In chapter 3 of this thesis an overview is given of LC materials. Nematic LC
form a cholesteric phase if the liquid crystal molecule has a chiral center or if a dopant with a
chiral center is added. Due to the addition of this chiral molecule the director of the LC
material is not constant over the thickness of the layer, but it will rotate forming a helical
structure. Due to the birefringent nature of the LC-material, the helical structure will create a
periodic fluctuation in refractive index between the ordinary and extraordinary refractive index
of the LC-material over the thickness of the cholesteric layer. This rotational periodic change in
refractive index leads to a Bragg-reflector for circular polarized light with the same handedness
as the rotational direction of the helix, so a right handed helix will reflect right handed circular
polarized light for a small spectral bandwidth, while left handed polarized light is transmitted.
The wavelength of the light that is reflected by a cholesteric reflector depends on the pitch of
the helical structure, where the pitch is defined as the length of the helix wherein the director of
the LC-material is rotated by 360° (see figure 7.2). The central wavelength of the spectral band (
0 ) that is reflected by the cholesteric material is defined as:
0 n p (7.3)
where n is the average refractive index of the LC material and p is the pitch of the helical
structure.
Figure 7.2 Schematic depiction of the helical structure of a cholesteric material, defining the pitch of the material
Chapter 7
108
The resulting reflectors have a spectral bandwidth ( ) determined by the birefringence
(n ) of the LC material in unidirectional alignment and the pitch and can be described as
p n (7.4)
Commonly available liquid crystalline materials have birefringence up to 0.25, which leads
to a spectral reflection bandwidth between 50 and 100 nm for visible light. In the rest of this
thesis these reflectors are called 75 nm reflectors, or ‘narrowband reflectors’. The pitch of a
nematic liquid crystalline material, which contains a chiral dopant, can be changed by the
concentration ( c ) and the helical twisting power ( HTP ) of the chiral dopant in the mixture
used:
1
pHTP c
(7.5)
The helical twisting power of a chiral dopant is determined by the off-set in the director of
the LC-material between different layers of the material created by this dopant. A chiral dopant
with a positive HTP will lead to a right handed helix, while a negative HTP creates a left
handed helix.
As mentioned above, a cholesteric reflector will only reflect one circular polarization of
light. [213] For the application as a wavelength-selective reflector on top of an LSC to reduce
the surface losses, a reflector is desired which reflects unpolarized light, called a full reflector. A
full reflector can be created from cholesteric layers by layering a right handed and a left handed
reflector, or by sandwiching a half wave retarder between two layers with the same handedness.
A schematic description of both full reflectors is depicted in figure 7.3.
In the full reflector made from a stack of a right-handed and a left handed cholesteric
layer, the right circular polarized component of unpolarized light is reflected by the right
handed cholesteric layer and the left circular polarized component is transmitted. This left
circular polarized component is reflected by the second layer which is a left handed cholesteric
layer, creating a full reflector. In the full reflector made from a half wave retarder sandwiched
between two right-handed cholesteric layers, the right circular polarized component of
unpolarized light is reflected by the right handed cholesteric layer and the left circular polarized
component is transmitted. This left circular polarized light is transformed into right circular
polarized light by the half wave retarder and then reflected by a second right handed cholesteric
reflector. This reflected right circular polarized light is transformed back into left circular
polarized light by the half wave retarder and transmitted through the first right handed
cholesteric layer, leading to a full reflector.
In this chapter, the effect of application of these organic wavelength selective mirror on
the efficiency of LSCs is determined, both in theory and experiment.
Organic wavelength selective reflectors
109
Figure 7.3 The working principle of full cholesteric reflectors. a) A right handed and a left handed cholesteric
stacked on top of each other, b) a half wave retarder sandwiched between two right handed cholesterics.
7.2 Methods
Narrowband cholesteric reflectors were made by spin casting solutions of the reactive LC-
mesogens RM257 and RM82 (in a 4:1 weight ratio) (Merck), varying concentrations of the
reactive chiral dopant LC756 (BASF), 1% of the photo initiator Irgacure 184 (Ciba), and 1% of
a surfactant in xylene (50% by weight of solution). The solutions were spin cast on rubbed half-
wave retarder plates centered at 560 nm (Edmund Optics) at 1000 rpm for 40 s and placed
immediately on a hot plate at∼80 °C for ∼15 s. The samples were then cross-linked by
exposure to UV-light at room temperature in a nitrogen atmosphere for 10 min. For the
cholesterics to be capable of reflecting both left- and right circularly polarized light, a second
cholesteric from the same solution was spun on the rubbed backside of the same half-wave
retarder.
The layered broadband reflectors were made from two stacked right-handed narrowband
reflectors applied to a manually rubbed half wave retarder centered at 560 nm (Edmund
Optics). A mixture of reactive LC mesogen LC242 (BASF, figure 7.4), varying concentrations
chiral dopant LC756 (BASF, figure 7.5), 1% of photo initiator Irgacure 184 (Ciba) and 1% of
surfactant (2-(N-Ethylperfluoro-octanesulfamido)ethyl acrylate) to induce planar alignment at
the liquid crystal-air interface in xylene (1:1 by weight, Aldrich) were spin coated at 800 rpm for
30 seconds. After spin coating, the samples were immediately heated on a hot stage at 90 °C for
Chapter 7
110
30 seconds and then photo-polymerised by UV-exposure at room temperature in a nitrogen
atmosphere. Before applying the second reflecting layer with a higher concentration of chiral
dopant, the first layer was treated with an oxygen-plasma for 1 minute at 60 W, to improve the
wetting of the LC layer. A similar process was applied to the rear side of the same half-wave
retarder following an identical procedure.
Figure 7.4 Chemical structure of LC242
Figure 7.5 Chemical structure of LC756
The gradient pitch broadband reflectors were made by filling 20 µm cells with mixtures of
chiral and nematic monomers, photo initiator, and UV-absorbing dye, followed by crosslinking
with UV light. Both left and right handed reflectors were produced. The right-handed CLC
mixture was prepared by mixing 41.3 wt.% RM96 (monoacrylate chiral monomer, Merck.) with
RM257 (diacrylate nematic monomer, Merck); the left-handed cholesteric mixture was prepared
by mixing 32.4 wt.% of a diacrylate chiral monomer with monoacrylate nematic monomer
(described by Broer, et. al.[228]), respectively. In both cholesteric mixtures, 1 wt.% photo
initiator (Irgacure 651, Ciba) and UV-absorbing dye (Tinuvin 328, Ciba) were added. The
amount of chiral material was chosen so that the desired central wavelength was obtained.
Organic wavelength selective reflectors
111
Polyimide coated (Grade 130, Nissan Chemical), anti-parallel rubbed cells (5x5 cm2) were filled
by each mixture by capillary action. The cell gap was maintained by spacers of 20±1 μm. The
right- and left-handed CLC films were cured under nitrogen with UV radiation (λ=365 nm, 28
μw/cm2) at 105°C for 30 min. and 5 min., respectively. In this way a pitch gradient in the
cholesteric stack is made, with the smaller pitch at the top of the film. The specific UV dose
and curing time were determined to produce the desired bandwidth
The filled waveguides were produced by injection moulding (poly)carbonate doped with 5-
180 ppm BASF Lumogen F Red 305 into 50 x 50 x 3 mm3 plates (Sabic IP). The patterned
LSC waveguides (see the results and discussion section of this chapter) were produced on
PMMA plates (50 x 50 x 5 mm3) (Plano Plastics). Fluorescent dye solutions were prepared
using 0.5 wt% of Red 305, and 1 wt% photoinitiator (Irgacure 184) dissolved in a 3:1
dipentaerythritol penta-acrylate (Polysciences) and methylmethacrylate (MMA, Aldrich) blend.
The dye solutions were stirred and heated at 60°C for an hour prior to spin-coating onto the
substrates at 1000 rpm for 30 s. After spin-coating, all 100% covered samples were cross-linked
by exposing to a high-intensity UV lamp for 80 s under nitrogen flow to form a solid film. For
the fabrication of patterned LSCs, standard photolithography techniques were employed.
Uniformly coated substrates were exposed to UV light through patterned shadow masks
consisting of 10 lines with variable widths with a period of 5 mm. Line widths were varied to
cover 20 to 80% of the waveguide surface. After UV exposure, ethanol was used to etch away
the unexposed material on the PMMA substrates by placing the exposed patterned samples in
ethanol for 40 s at room temperature and the samples were continuously agitated during the
etching process. The preparation of the patterned LSC waveguides was done by Shufen Tsoi.
Transmission spectra of the manufactured reflectors and absorption spectra of the waveguides
were recorded with a Shimadzu UV-3102 spectrophotometer.
Quantitative bottom surface emissions from the waveguides (that is, emission from the
surface opposite the incident light source) were determined by use of a Labsphere spectral light
measurement system LED 1050 integrating sphere, with a blank waveguide plate used as the
reference; see figure 5.2 for a depiction of the experimental setup. The top surfaces of the
samples were exposed to the light from a 300 W solar simulator with filters to approximate the
1.5 air mass (global) solar spectrum (LOT Oriel). We use filters placed between the source and
the sample to reduce the background spectra from the source light in the spectral emission
band of the used luminophore. The illumination was limited to an approximately 6.45 cm2 area
in the waveguide center. The surface emission measurements were repeated with cholesteric
filters placed individually against the bottom surface of the waveguide. The spectra from the
Red 305-doped waveguides were subtracted from the spectrum of the blank waveguide. The
resulting spectra were integrated to determine the quantity of light absorbed by the waveguide,
and the amount of light emitted from the surface (see chapter 5 for more details).
Chapter 7
112
Representative surface emission spectra from one waveguide (absorbance 0.46) were
determined at three different angles using an Autronic 703 display measuring system (Melchers)
together with a CCD-Spect-2 (CCD camera). A 1 cm2 area of the waveguide topped by a blank
half-wave plate and of the same waveguide with a 670 nm cholesteric plate on top was
illuminated from below by a Q50MR16 daylight lamp (SoLux), and surface emissions were
recorded from a 0.5 mm diameter area centered approximately 0.5 cm from the edge of the
illumination region to avoid recording any of the source illumination. The surface emission
spectra were recorded at defined angles with respect to the surface normal.
The edge emissions from the LSC waveguides with and without cholesteric reflectors were
recorded using an SLMS 1050 integrating sphere equipped with a diode array detector. The
samples were placed with their edges in the entry port of the integrating sphere while
illuminated with a 300 W solar simulator with filters to approximate the 1.5 AM solar spectrum
(Lot Oriel). The spectrum and intensity of the edge emission were recorded. A thick paper
spray painted white placed underneath the sample acted as a Lambertian scatterer. The total
output from the edges was determined by integrating the recorded spectrum from 350-750nm.
A depiction of this setup can be seen in figure 7.6.
Figure 7.6 The experimental setup for measuring emission intensities at the edges of waveguides with a cholesteric
reflector on top and a scattering layer underneath, where both are prevented from being in optical contact with the
waveguide by an air gap
To determine the angular dependency of the reflection band of the gradient pitch
reflector, the cholesteric films were removed from the cell, combined and sandwiched between
two half-cylinders with index-matching oil. The half-lambda wave plate is a polymer film (Nitto
Denko) with a retardation of 413 nm. Transmission measurements of these samples were made
with a spectrophotometer as described in figure 7.7, where the sample is place between the half
cylinders. We also measured the transmission without the rotatable stage and the detector closer
Organic wavelength selective reflectors
113
to the sample. This results in a transmission increase of about 2%, because of a more complete
collection of the scattered radiation.
Figure 7.7 Experimental set-up for measuring the angular transmission spectra of gradient pitch cholesteric
reflectors: a Perkin-Elmer Lambda 800 UV-VIS Spectrophotometer with a rotatable sample holder and a
detector that includes an integrating sphere. To minimize refraction of the light, the sample was sandwiched
between two glass half-cylinders and index-matching oil was applied. The grey lines indicate light beams.
7.3 Narrowband reflectors
The purpose of the cholesteric layer is to reduce the surface losses of LSCs by selective
reflection of the emitted light. The retained light might reach the emission edges and increase
light output from the waveguide. Experiments with narrowband reflectors used right-handed
cholesterics located on either side of a half-wave retarder centered at 560 nm. Made in this way,
the right-handed circular polarized light reflecting cholesterics could be used to reflect both
handedness of light incident normal to their surface. The transmission spectra for the five
cholesteric mirrors, exposed to unpolarized light can be seen in figure 7.8.
Chapter 7
114
Figure 7.8 Transmission of unpolarized light through the two-sided cholesterics used in these experiments; onset
at 630, 650, 680, 720, and 760nm, respectively.
The surface loss of three filled waveguides containing Red 305 as luminophore was
measured as described in chapter 5, with and without the narrowband reflectors between the
waveguide and the entry port of the integrating sphere. From these measurements the
reduction of surface loss was calculated. The results are shown in table 7.1.
Table 7.1 Percentage of absorbed light emitted through the bottom of the waveguides after application of a
cholesteric reflector
Peak
Absorbance
Onset wavelength of the cholesteric reflector
None 630 650 680 720 760
0.46 16.8 10.8 10.0 9.1 12.5 13.5
1.01 15.4 10.0 8.7 9.0 11.4 12.7
2.36 15.3 11.5 9.6 10.0 12.2 13.0
In the best case, application of the 650nm cholesteric to the waveguide with absorbance
1.01 reduced bottom surface loss of the waveguide by 44%. The effect of adding a cholesteric
filter to the top of the waveguide on the surface emission spectrum as a function of emission
angle can be seen in figure 7.9.
Organic wavelength selective reflectors
115
Figure 7.9 Comparative surface emission spectra for a Red 305 filled polycarbonate waveguide with a peak
absorbance of 0.52. The black curve resembles the surface emission spectra measured 15° to the surface normal
with a blank half-wave plate on top. The surface emission from the same waveguide with a cholesteric reflector
centered at 670nm on the top is measured at 15° (dark gray curve), 30° (medium gray curve), and 45° (light
gray curve) to the waveguide normal.
The angular dependent surface emission spectra shown in figure 7.9, demonstrate that a
part of the surface emitted photons is reflected back into the waveguide. Upon increase of the
angle of the emitted photons with respect to the normal of the reflector, shorter wavelength
photons of the emission band of the dye are reflected. At each angle only a part of the surface
emitted photons are back-reflected, a result of the width of the reflection band being more
narrow than the spectral width of the emission band of the luminophore.
Reflectors were placed on top of LSC waveguides containing different concentrations of
the luminophore. The edge output of the waveguides with and without reflectors on top of the
waveguides was measured using the integrating sphere as described in the experimental section
of this chapter. A piece of thick paper spray painted white acted as a Lambertian scatterer and
was placed underneath the waveguides to reflect light lost through the bottom surface and light
reflected by the cholesteric reflector back into the waveguide. The results are shown in table
7.2.
The results in table 7.2 show that there is 13.7% more light leaving the edge of an LSC
with a peak absorbance of 0.05 when the 720 nm reflector is applied on top of the waveguide.
The LSCs with a higher peak absorbance (0.46-1.01) show a ~10% increase in energy that
Chapter 7
116
leaves the edge of the waveguide. In these experiments, where a white bottom scatterer is used,
the waveguide edge emission light can be divided into two types: a) that resulting from
scattering from the backscattering layer and b) light directly from the dye emission. To
determine the relative contributions, the spectrum obtained from a polycarbonate waveguide
that contained no dye with a separate white scatterer was subtracted from the spectrum of the
same scatterer underneath a polycarbonate waveguide that contained dye. This isolated the
impact of the dye. Second, the same spectrum of the clear polycarbonate waveguide was
subtracted from the spectrum measured from a dye-filled waveguide with a cholesteric layer
separated by an air gap on top, which isolates the impact of the dye with a cholesteric. Sample
spectra resulting from these manipulations can be seen in figure 7.10.
Table 7.2 Optical efficiency of the polycarbonate Red 305 waveguides. Between brackets is the relative efficiency,
where the relative efficiency of the LSC with no reflector is set at 1.
Absorbance No
reflector
Onset wavelength of the reflector (nm)
630 650 680 720 760
0.05 0.0265 0.0280
(1.058)
0.0278
(1.050)
0.0288
(1.086)
0.0301
(1.137)
0.0293
(1.108)
0.46 0.0794 0.0794
(1.000)
0.0824
(1.038)
0.0833
(1.049)
0.0858
(1.081)
0.0889
(1.120)
1.01 0.1064 0.1041
(0.978)
0.1071
(1.007)
0.1103
(1.037)
0.1118
(1.051)
0.1163
(1.093)
2.36 0.1572 0.1522
(0.968)
0.1569
(0.998)
0.1605
(1.021)
0.1622
(1.032)
0.1665
(1.059)
There are two regions in these spectra, labelled I and II in the figure. Region I is a measure
of the amount of light removed from the incoming light by dye absorption that was normally
scattered to the emission edge in a clear waveguide. Since the resultant spectra of the samples
with and without cholesteric are essentially identical in region I, we can conclude that the
presence of the cholesteric does not significantly remove light from the system at these
wavelengths (otherwise the amount of light available to be absorbed would be less). Region II is
a measure of the emission of light by the dyes only, as the additional light derived from the
scatterer has been removed. The difference between the spectra with and without the
cholesteric, then, is a measure of the addition or removal of light caused by the cholesteric. By
comparing the integrations of region II up to 750 nm for the sample with and without
cholesteric, we obtained the data displayed in figure 7.11, which shows the fractional increase in
energy emission at the edge due to the reflection of surface-directed light by the cholesteric
layers.
Organic wavelength selective reflectors
117
Figure 7.10 Resultant spectra of an absorbance 1.01 waveguide (graycurve) and the same waveguide with a 720
nm cholesteric separated from the waveguide by an air gap (black curve) after subtraction of the spectrum of a
blank waveguide. In all cases, a rear white scattering layer was used. The region labeled I represents additional
absorption of in-scattered light, and region II represents dye emission with scattered light removed.
Figure 7.11 Edge emission relative to the output of a bare waveguide with blank halfwave plate using cholesteric
filters with the onset wavelength at 630 nm (stars), 650 nm (circles), 680 nm (triangles), 720 nm (squares),
and 760 nm (diamonds) as a function of waveguide absorbance. Resultant emission data integrated from 350 to
750 nm.
Region I Region II
Chapter 7
118
Figure 7.11 demonstrates that the cholesteric layer can enhance the amount of dye-emitted
light that reaches the waveguide edge by up to 35%, with the relative effect larger for
waveguides with a lower dye content. The cholesterics centered at longer wavelengths
demonstrated better results than those centered closer to the emission wavelengths of the dye.
For the waveguide with absorbance of 1.01, the cholesteric with an onset wavelength of 760
nm provided 12% more dye-emitted light energy to the waveguide edge. According to our
previous measurements, 40% of the energy initially absorbed by the dye in a waveguide with
this absorbance is lost through the surface. Thus, the 760 nm cholesteric converted 30% of
light normally lost into useful output in this sample.
Measurements of the absolute energy emissions from the waveguides presented in table
7.2 describe a similar story in that the cholesterics centered at longer wavelengths demonstrate
increased performance. The absolute additional output resulting from adding the cholesteric
layer makes up a smaller fraction of the total output, which includes the contribution of
scattered light. Normally, the fractional contribution of light obtained by using a scattering layer
decreases with increases in the waveguide size [173], and so it is anticipated that the relative
contribution of the cholesteric can increase with waveguide dimensions as the contribution
from scattered light decreases. [199]
One might have expected that the optimal reflection wavelength would be much closer to
the peak emission of the Red 305 dye (630 nm). Light leaving the surface of the waveguide will
have angles from 0°-90° with respect to the normal of the reflectors. There is a blue shift of the
centre of the reflection band of a cholesteric with a change in incident light angle. The resultant
central reflection wavelength ( ) for light incident from air is generally described by
10
sincos sin
n (7.6)
where is the angle of incident light in degrees, 0 is the central reflection wavelength at
normal incidence, and n is the average refractive index of the host LC. Furthermore, separate
measurements of modulated left-circularly polarized light incident on a bare half-wave plate as a
function of incident angle demonstrated that transmitted light was decreasingly right-circularly
polarized as the incident angle increased, and thus more elliptical in nature (data not shown).
This incomplete transition of left-circularly polarized light into right-circularly polarized light
translated into reduced performance of the cholesteric and, consequently, reduced the impact
of the cholesteric in these experiments.
From the spectra in figure 7.9 it can be seen that the width of the cholesteric reflection
band of the cholesteric reflectors is not broad enough to reflect all the surface emitted light
back into the waveguide. By broadening the reflection band more light will be reflected back
into the waveguide, since a broader part of the emission spectrum of the dye is reflected. The
effect of broadband reflectors is discussed in the next section.
Organic wavelength selective reflectors
119
7.4 Broadband reflectors
The bandwidth of a cholesteric reflector can be broadened in three ways: 1) use of
materials with a large birefringence in accordance with equation 7.4, 2) a gradient in the pitch
over the thickness of the film and 3) stack multiple narrowband reflectors with an offset in
reflection band. The reflective properties of broadband cholesteric reflectors made by these
three methods have been calculated and the effect of these reflectors on the efficiency of LSCs
has been determined from these calculations and the results have been compared to
experimental measurements. These calculations and measurements will be displayed and
discussed in the rest of this section.
7.4.1 Theoretical approach
The reflective properties of chiral nematic liquid crystalline films can be calculated using
Berreman’s 4x4 matrix method [229] for light propagation through multi-layered homogenous
anisotropic media. [230] This method allows calculation of the optical properties of
homogenous anisotropic media at oblique incidences. The cholesteric reflector is divided in
several homogeneous slabs with a different direction for the optical axis in each slab to
calculate the reflective properties of the reflectors.
7.4.1.1 Angular dependency of the optical behaviour of broadband cholesteric reflectors
Optical properties of narrowband and broadband cholesteric reflectors, made by three
different methods (described above) have been calculated. For the broadband reflectors made
from layering two narrowband reflectors, the ordinary and extraordinary refractive indexes were
taken from the commercial liquid crystal host BASF LC242, and the pitches of the separate
layers were chosen in such a way so that the overall reflection band was continuous and the
width of the resulting reflection band was 150-200 nm. For the gradient pitch reflectors, input
data was chosen from the materials described by Broer et al. [231] and the pitch gradient was
chosen in such a way that the reflection bands were similar to the reflectors made by layering
narrow band cholesterics, and the width of the reflection band was 400 nm. For the high
birefringent material, the characteristics of the liquid crystal BASF LC1057 were chosen. This
latter material leads to reflectors with a more narrow reflection band than the two other
methods. To match the width of the reflection band of the other two broadband reflectors,
liquid crystalline materials with very large birefringence would be necessary (i.e. a n of 0.4 to
create a 175 nm broad reflector with an onset wavelength around 600 nm), but these materials
are have very low photostability. Due to the lack of available materials to produce the desired
broadbands in this way experimentally, reflectors made from high birefringent materials are not
considered in the rest of this chapter.
The simulated reflection properties of the two broadband cholesteric reflectors (stacked
right/left and stacked right/right with a half wave plate between them) for unpolarized light for
all angles of incidence in air are depicted in figure 7.12.
Chapter 7
120
Figure 7.12 The simulated reflective properties of 150 nm broad reflectors made from gradient pitch cholesterics.
Left, the reflectivity of a full reflector made by stacking a right- and a left-handed cholesteric on top of each other
and right, the reflectivity of a full reflector made by right handed cholesterics on both sides of a half wave retarder
centered a 560 nm are used to make a full reflector. The color in these plot represents the reflectivity of the
cholesteric reflectors; dark blue is 0% reflection and dark red is 100% reflection.
Figure 7.12 demonstrates that the reflectivity of a reflector made from a stack of a right-
and a left handed cholesterics is 100% over the entire width of the reflection band when the
incidence light is at small angles (up to 20 degrees). When the angle of the incident light
becomes larger, the reflective properties begin to decrease starting from the edges of the
reflection band. The right picture shows that the reflectivity of two right handed cholesteric
reflectors on both sides of a half wave retarder centered at 560 nm is not as constant over the
width of the reflection band at small angles of incidence, because the half wave retarder is not
completely converting right circular polarized light into left circular polarized light at the
wavelengths where the reflector is positioned. When a half-wave retarder is used centered in the
same wavelength regime as the reflection band the reflective properties are similar to the
properties of a reflector made from stacked right- and left-handed cholesteric reflectors.
To verify the angular dependency of the cholesteric reflectors calculated using this method
a series of measurements was performed. The calculations, measurements and sample
preparations in the rest of this section (7.4.1.1) were performed by Chi-Wen Lin and Merijn
Giesbers. In figure 7.13 and 7.14 the transmission spectra of a broadband cholesteric reflector
made via a gradient pitch are displayed as calculated and experimentally measured are depicted
for different angles of incidence. In figure 7.13 the full reflectors are made by stacking a right
and left reflector on top of each other and in figure 7.14 a stack of two right handed reflectors
with a matched half wave retarder between them was used to create a full reflector. The stacked
films are optically connected by a refractive index matching oil.
Organic wavelength selective reflectors
121
Figure 7.13 Transmission spectra for unpolarized light for a right-handed cholesteric reflector stacked on top of a
left-handed cholesteric reflector. The pitch varies linearly from 437 to 520 nm in the right-handed material and
from 429 to 521 nm in the lefthanded material. The refractive indices, ne=1.68 and no=1.54, are the same for
both materials. Dotted lines indicate the experimental measurements, solid lines indicate calculations. Figures a)
to f) are for incident angles in glass of 0°, 10°, 20°, 30°, 40°, and 50°, respectively.
b)
c) d)
e) f)
a)
Chapter 7
122
Figure 7.14 Transmission spectra for unpolarized light for a configuration with a half-lambda plate sandwiched
between two right-handed cholesteric layers. The pitch varies linearly from 464 to 523 nm in the layers. The
refractive indices are ne=1.68 and no=1.54. The half-wave retarder has a center wavelength of 825 nm. Dotted
lines indicate the experimental measurements, solid lines indicate calculations. Figures a) to f) are for incident
angles in glass of 0°, 10°, 20°, 30°, 40°, and 50°, respectively.
c)
a) b)
d)
e) f)
Organic wavelength selective reflectors
123
To fit the calculations to the measurements, the pitches of the materials were chosen such that
the left and the right sides of the individual reflection bands superimposed. From this, the
spectra for the stack are calculated for all angles, using the known refraction indices on and en
assuming a linear pitch gradient within each CLC and a constant refractive index with changing
wavelength of the incident light.
The experimental and calculated transmission spectra for unpolarized light for both
broadband reflectors overlap. The blue shift in reflection band is similar for both experiments
and calculations when the angle of incidence increases. The small mismatch in reflection band
at perpendicular incidence can be explained by a small discrepancy between the (gradient) pitch
used for the calculations and the actual one in the experimental reflectors. At high angles of
incidence the reflector made from right handed cholesterics on both sides of a half wave
retarder show reduced reflectivity in measurements than predicted in the calculations. The
reason could be the quality of the experimental reflector.
This close match between calculations and experiment makes it possible to calculate the
reflection bands of multiple cholesteric layers and calculate what the effect of these layers on
the efficiency of the LSC. This is done in the following sections.
7.4.1.2 Efficiency of cholesteric reflectors
The wavelength selective mirrors are placed on top of an LSC to reflect photons normally
escaping through the top surface of the LSC back in the waveguide. Underneath the waveguide
a perfect reflector is placed reflecting all photons normally lost through the bottom of the
waveguide. These photons are assumed to not be re-absorbed by the luminophores and instead
lost again through the top surface. To calculate the increase in LSC efficiency after application
of such a reflector, it is necessary to first calculate the reflection efficiency (refl ) of the reflector
towards dye-emitted light (equation 7.7).
( ) ( ) ( , )
( ) ( )
s p
refl
s p
E E R d d
E E d d
(7.7)
where ( )sE is the emission spectrum of the dye molecules, ( , )R is the reflection spectrum
of the reflector as a function of incidence angle of illumination and ( )pE is the angular
emission profile of the dye molecules. The efficiency of reflectors with different reflection
bandwidths has been calculated for broadband reflectors made from cholesterics with a
gradient pitch. Two gradient pitch reflectors have been considered, one twice the width of the
reflection band of the narrowband reflectors (i.e. 175 nm) and one with a reflection bandwidth
of 400 nm, approximately 5 times the width of the reflection band of narrowband reflectors.
The efficiency of reflecting surface emitted light is depicted in figure 7.15 for an LSC using Red
Chapter 7
124
305 (see figure 5.4 for the absorption and emission spectra of this luminophore). In these
calculations the emission profile of the dyes is assumed isotropic. The onset wavelength is
defined as the wavelength at the short wavelength side of the reflection band where the
reflectivity is 50% of the reflectivity within the reflection band.
Figure 7.15 The efficiency of cholesterics in reflecting light emitted from the top surface ( refl ) of a Red 305
containing LSC for narrowband reflectors (white squares), 175 nm broad gradient pitch reflectors (grey squares),
400 nm broad gradient pitch cholesteric reflectors (black squares), layered reflectors made from 2 narrowbands
(filled red circles for stacked right and left handed reflectors and open red circles for stacked right handed reflectors
on both sides of a half wave retarder centered at 560 nm) as a function of the onset wavelength of the cholesteric
reflectors.
The efficiency of all the cholesteric reflectors is equal at longer onset wavelengths and the
efficiency increases with decreasing onset wavelengths. The efficiency peaks when the
cholesteric onset wavelength is in the spectral part of the emission by the luminophore. The
efficiency of reflector with the broadest reflection band (~400 nm) peaks at the shortest onset
wavelength, which can be explained as follows: all reflectors exhibit a blue shift for high angles,
but for the broad cholesteric reflector the complete emission band of the luminophore is
within the reflection band at all angles; so the highest efficiency is reached if the onset
wavelength of the reflector is close to the onset wavelength of the emission band. The
reflectors with the smaller reflection bands are not broad enough to reflect all the emitted light
over all angles being emitted from the surface, and so the efficiency of these reflectors peaks at
slightly longer wavelengths. The maximum efficiencies with corresponding onset wavelengths
are shown in table 7.3.
Organic wavelength selective reflectors
125
Table 7.3 Maximum reflection efficiency of the cholesteric for surface emitted light of LSCs containing Red 305
as luminophore.
Reflector Maximum efficiency Onset Wavelength (nm)
400 nm gradient pitch 91% 560
175 nm gradient pitch 65% 620
75 nm narrowband 53% 650
The 175 nm broad reflectors made from gradient pitch cholesterics and from layered
narrowband reflectors show similar efficiencies at all onset wavelengths. From this it can be
concluded that in the cases calculated the width of the reflection band is more important for
the efficiency than the method by which the reflectors are made.
7.4.1.3 Effect of cholesteric reflectors on incoming sunlight
Reflectors placed on top of the LSC may also reflect incoming sunlight away from the device,
resulting in reduced absorption if the reflection band coincides with the absorption spectrum of
the dye. This will influence the effect these reflectors have on the total efficiency of the LSC.
The fraction of light that is absorbed by the luminophore that passes through the cholesteric
filter (EA
cholf ) can be described by
( , )(1 ( , )) ( , )
( , ) ( , )
EAchol
I R A d df
I A d d
(7.8)
where ( , )I is the intensity of the incident light (in this case, the solar spectrum), ( , )A
is
the absorption spectrum of the dye corrected for the path length if light is coming in at larger
incident angles according to Lambert-Beer’s law and ( , )R is the reflection spectrum of the
cholesteric filter. In this calculation it is assumed that only direct sunlight is incident normal to
the LSC device, so there is no angular dependency. The results are depicted in figure 7.16.
When the onset reflection wavelength is outside the absorption range of the luminophore,
the reflectors transmit approximately 90% of all the absorbable light; the 10% loss results from
the fact the cholesteric reflector is added to the LSC with an air gap creating two additional
surfaces and extra Fresnel reflections. As the onset wavelength of the cholesteric reflectors
passes into the absorption range of the luminophores, the amount of absorbable incident light
that is transmitted through the reflector decreases drastically. This decrease is approximately
equal for all reflectors and is thus not influenced by reflection band width or production
method.
Chapter 7
126
Figure 7.16 The fraction ( EAcholf ) of the incoming sunlight that can be absorbed by the luminophore (Red 305)
that passes through the cholesteric reflectors made from narrowband reflectors (white squares), 175 nm broad
gradient pitch reflectors (grey squares), 400 nm broad gradient pitch cholesteric reflectors (black squares), layered
reflectors made from 2 narrow bands (filled red circles for stacked right and left handed reflectors and open red
circles for stacked right handed reflectors on both sides of a half wave retarder centered at 560 nm).
The reflectors with a 400 nm bandwidth can block all the incident light that will be absorbed by
the dye molecules. The bandwidth of the reflector with a 175 nm bandwidth and the
narrowband reflector are not broad enough to reflect away all the absorbable sunlight, so at
shorter onset wavelengths they transmit the long wavelength part of the incident spectrum that
can be absorbed by the dye molecules. Thus the fraction of the absorbable sunlight that is
transmitted through the reflectors with a short onset wavelength is larger for the narrow
bandwidth reflectors
7.4.1.4 Increase in LSC efficiency
The maximum total increase in LSC efficiency ( ,maxLSC ) is a combination of both incident and
emitted light reflection and can be calculated from the efficiency of the reflector and the
absorbable incident light that passes through the reflector. This increase can be described by
the number of photons leaving the edge of the LSC when a cholesteric filter is added ( ,edge choln )
and the number of photons leaving the edge of the LSC without a cholesteric filter ( ,edge baren ).
, , , ,
,max
, , ,
EAedge chol chol edge bare edge SL chol
LSC
edge bare edge bare edge bare
n f n n
n n n (7.9)
Organic wavelength selective reflectors
127
where , ,edge SL choln is the total number of photons formerly lost through the surface that are
converted to edge emission of the LSC due to addition of the cholesteric and defined as:
, , * **EAchol PLQYedge SL chol sl cholfn (7.10)
and
, * (1 )edge bare PLQY sln (7.11)
where PLQY is the fluorescence quantum yield of the luminophore, and sl is the fractions of
emitted photons lost through the surface (i.e. within the escape cone), respectively. Although a
calculation of ,maxLSC requires a detailed knowledge of the processes in the waveguide, a rough
estimate can be obtained in the following way. Combining equations 7.9, 7.10 and 7.11 leads to
,max 1
(1 )
EA slLSC chol chol
sl
f (7.12)
For a waveguide containing Red 305 it was previously shown in chapter 5 that the number
of photons in surface loss mode was approximately 50% of all absorbed photons for 5x5 cm2
LSCs made from polycarbonate and peak absorption above 0.3 leading to
,max 1EALSC chol cholf (7.13)
In this calculation, all the light reflected back in the LSC by the cholesteric reflector is
assumed to reach the edge of the LSC. So there is only one interaction with the reflector and
there is no re-absorption of these back reflected photons, losses from parasitic waveguide
absorption, or other such events. The results are depicted in figure 7.17.
With the broadening of the reflection band, the maximum possible increase in LSC
efficiency improves. The onset wavelength of the cholesteric where the efficiency increase is
the highest is red-shifted with respect to the emission peak of the luminophore, but for the
broadest reflector the red-shift is less pronounced than for the narrower reflectors. The
maximum possible increase in LSC efficiency and the corresponding onset wavelength of the
cholesteric are shown in table 7.4.
Thus, by adding a 400 nm broad reflector at the top of an LSC, the efficiency could be
increased by up to 66%. If a reflector with a more narrow reflection band is added to the top of
the LSC, increases of 45% or 35% could be achieved for 175 nm broad reflectors and 75 nm
broad reflectors, respectively.
Chapter 7
128
Figure 7.17 The calculated maximum relative LSC efficiency ( ,maxLSC ) after application of cholesteric
reflectors to an LSC containing Red 305 as a luminophore. The reflectors are made from narrowband
cholesterics (white squares), 175 nm broad (grey squares) and 400 nm broad gradient pitch cholesteric (black
squares), layered cholesteric and reflectors made from 2 narrowbands (red circles for stacked right and left handed
reflectors and open red circles for stacked right handed reflectors on both sides of a half wave retarder centered at
560 nm).
As described in the experimental section, 175 nm broad cholesteric reflectors were made from
right-handed layered narrowband reflectors on either side of a half wave retarder centered at
560 nm and placed on top of Red 305 LSCs. The results of these measurements and a
comparison with the theoretical results described in this paragraph are shown and discussed in
the next section.
Table 7.4 Maximum calculated increase in LSC efficiency after addition of the cholesteric reflectors to LSCs
containing Red 305 as luminophore
Reflector ,maxLSC Onset wavelength (nm)
400 nm gradient pitch 1.66 620
175 nm gradient pitch 1.45 650
75 nm narrowband 1.35 650
Organic wavelength selective reflectors
129
7.4.2 Experimental results
Seven broadband reflectors were produced with different onset wavelengths: 620 nm, 660
nm, 700 nm, 740 nm, 780 nm, 820 nm and 880 nm. The reflection bands of these cholesterics
were measured and compared to the calculated reflection bands at perpendicular incidence. The
differences between the theoretical and experimental spectra are similar for all the different
reflectors. As an example, the calculated and experimental spectra of the reflector with an onset
wavelength of 740 nm are depicted in figure 7.18.
500 600 700 800 900 1000
0,0
0,2
0,4
0,6
0,8
1,0
Re
fle
ctio
n (
%)
Wavelength (nm)
Figure 7.18 Reflection spectra of a broadband reflector with an onset wavelength of 730 nm made from 2 layered
right handed narrowband reflectors on both sides of a half wave retarder centered at 560 nm, both experimental
(black) and calculated (gray).
The experimental spectrum shows that the width of the reflection band is the same as
calculated, but the reflectivity is somewhat lower. This reduction can be a result of reduced
layer thickness in the manufactured reflector than assumed in the theoretical calculations.
Furthermore, the first layer applied experimentally is treated with a plasma asher, which can
further reduce the layer thickness. Therefore, reflectivity on the longer wavelength side of the
reflection band is slightly lower than on the short wavelength side, since the layer with the
reflection band at the longer wavelength side of the reflection band was applied first in all cases.
These reduced reflective properties also cause the dip in the middle of the reflection spectrum.
The theoretical spectrum demonstrates less than 100% reflectivity caused by the use of the
slightly mismatched half wave retarders which do not completely convert left circularly
polarized light passing through the right handed cholesterics into right circularly polarized light,
Chapter 7
130
so the right handed cholesterics on the back side will not be capable of reflecting all remaining
transmitted light.
The cholesteric reflectors were placed on top of (poly)carbonate LSCs filled with Red 305,
and the edge emission spectra and intensity (in Watts) measured using an integrating sphere
under illumination with a solar simulator (AM 1.5). Underneath the samples, a white
Lambertian scatterer was placed. These measurements were compared with the edge output of
the same LSC with the white scatterer but without the cholesteric reflector. The ratio between
the two edge output intensities is plotted in figure 7.19.
Figure 7.19 Experimental relative LSC efficiency after application of broadband reflectors. The LSCs contain
Red 305 with different peak absorbance: calculated (black), 0.05 (green), 0.19 (light blue), 0.46 (red), 1.01
(yellow), 1.63 (blue), 2.36 (orange), and >4 (navy).
Application of cholesteric reflectors to an LSC with a peak absorbance of approximately
0.5 increased the optical efficiency by a maximum of 4.5% when the reflector with the onset
wavelength of 700 nm is added: At higher peak absorbance there is a small decrease in LSC
efficiency. At lower peak absorption (<0.1) the increase in efficiency is much higher, with a
peak increase in efficiency of 30% at 740 nm. However, this is lower than was calculated, and
the onset wavelength of the reflector with the maximum increase is red shifted in comparison
to the calculations.
There are several differences between the theoretical approach and the experimental
measurements. First, the reflectivity of the experimental reflectors is not as good as calculated
in theory. Second, the theoretical approach assumed that all light reflected back into the LSC by
the reflectors reaches the edge of the LSC. In the experiments this is not the case. Photons
reflected back into the LSC can be reabsorbed by the luminophores in the LSC due to the
Organic wavelength selective reflectors
131
overlap in absorption and emission spectra. Reflected photons not immediately absorbed will
not be in waveguide mode, so they will encounter the white scatterer underneath the sample,
potentially multiple times. If the reflectors have non-unity reflectivity it can result in additional
losses. Finally, the emission profile is assumed spherical for the calculations. In actual practice,
the emission profile will not be spherical due to dichroic absorption and emission of the
luminophore in combination with the collimated incident light (see chapter 3). However,
calculations using a non-spherical emission profiles show only small differences with the
calculations using spherical emission profiles. Since the LSC with the very low peak absorbance
shows a higher increase in efficiency after application of the cholesteric reflectors, photon
recycling is the primary reason behind the lower experimental increase than predicted.
Figure 7.20 The optical efficiency (opt ) after application of broadband reflectors on LSCs containing Red 305
with different peak absorbance: 0.05 (green), 0.19 (light blue), 0.46 (red), 1.01 (yellow), 1.63 (blue), 2.36
(orange), and >4 (navy). The lines denote the edge emission from the LSC without the cholesteric reflector.
In figure 7.20 the optical efficiency of the LSCs is plotted for LSC with (data points) and
without (lines) cholesteric reflectors. As is clearly show the effect of the cholesteric reflectors
on the optical efficiency is minor for LSCs with a low peak absorbance. As already mentioned
before for LSC with a high peak absorbance the efficiency of the LSC is decreased after
application of a broad bandwidth reflector.
7.4.3 Patterned waveguides
Tsoi et al. [72,105] demonstrated that producing an LSC via patterning a coating containing the
luminophore on a clear waveguide can reduce the amount of re-absorptions. To investigate the
effect of photon recycling on our previous measurements, we measured the increase in
Chapter 7
132
efficiency of the patterned-LSCs after application of the cholesteric reflectors. LSCs with a
pattern of 10 lines were used, where the widths of the lines determined the coverage of the dye
coating on the LSC. The results of these measurements are shown in figure 7.21 using a dye
coating with peak absorbance of approximately 0.6. The work presented in this section (7.4.3) is
performed in collaboration with Shufen Tsoi and the patterned waveguides were produced by
her.
Figure 7.21 Increase in patterned LSC efficiency after application of broadband reflectors. The LSCs are topped
with a coating containing Red 305 with a peak absorbance of 1.0 with different pattern coverage of the surface:
calculated (black), 20% (green), 30% (light blue), 50% (red), 70% (yellow), 100% (blue)
The sample with surface coverage of 100% shows the same results as a filled waveguide with
approximately the same peak absorbance when topped by a broadband reflector, an increase in
efficiency of 5%. Reducing the surface coverage of the coating containing the luminophore, and
thus the amount of photon recycling for emitted light, enhanced the impact of the cholesteric
reflectors on the LSC efficiency. The lower the coverage, the higher the increase in efficiency
achieved by applying the cholesteric reflectors for all reflectors sampled. The LSC with
coverage of 20% shows an increase in edge emission efficiency of up to 27%, although this is
still lower than calculated.
Furthermore, in figure 7.21 it can be seen that the application of reflectors with a longer
onset wavelength results in an increase in LSC efficiency approaching the theoretical increase
better than with a shorter onset wavelength reflectors. This could also be explained if photon
recycling is the main cause of the reduced effectiveness of cholesteric reflectors on the
experimental increase of LSC-efficiency. Re-absorption occurs mostly in the short wavelength
Organic wavelength selective reflectors
133
part of the emission spectrum of the dye, since the absorption in this region is the greatest.
Cholesteric reflectors with an onset wavelength close to the onset wavelength of the emission
of the dye will reflect this emitted light for most of the angles of incidence. The reflectors with
a longer onset wavelength will only reflect these photons that have a larger probability for being
re-absorbed at larger angles. Thus, the effect of photon recycling in LSCs using short onset
wavelength reflectors is larger than for longer onset wavelength reflectors. These results
demonstrate that photon recycling has a large influence on the effectiveness of the reflectors.
Since the experimental increase in the LSC with a low amount of re-absorption after application
of the cholesteric reflectors is still lower than calculated for all applied reflectors, it can be
assumed that the multiple interactions of the back reflected photons with the reflectors also
have an influence on the effectiveness of the cholesteric.
Figure 7.22 Optical efficiency(opt ) after application of broadband reflectors on LSCs topped with a coating
containing Red 305 with a peak absorbance of 0.6 with different pattern coverage of the surface: 20% (green),
30% (light blue), 50% (red), 70% (yellow), 100% (blue). The lines denote the edge emission from the LSC
without the cholesteric reflector.
In figure 7.22 the optical efficiency of the patterned LSC with (data points) and without (lines)
is plotted. The effect of the cholesteric reflector on the optical efficiency is larger for these
patterned waveguides (4.8% to 6.0% for the 20% covered LSC) compared to the fully covered
waveguide (10.9% to 11.2%).
7.5 Other luminophores
The results presented in the previous sections of this chapter only describe LSCs with Red 305
as a luminophore. One of the advantages of LSCs is the freedom in choice of colour. A change
Chapter 7
134
of luminophore will also change the specifications of the cholesteric reflectors needed to
maximize the increase in LSC efficiency, and is also determined by the fraction of the photons
that is emitted through the surface of the LSC. In this section, the maximum increase in LSC
efficiency is calculated for different luminophores.
In chapter 5 the fraction of the absorbed photons emitted trough the surface of the LSC is
determined for two other dyes beside Red 305, the perinone dye and the coumarin dye. The
fractions of emitted photons that are emitted in surface loss mode (sl ) for the perinone and
coumarin dyes are 0.325 and 0.46 respectively. The absorption and emission spectra of the
coumarin and perinone are displayed in figure 5.8. Furthermore, a dye that consists of a rare
element core ion with an organic ligand that has no overlap between the absorption and
emission band and which is an isotropic emitter is considered. The fraction of the emitted
photons in surface loss mode for this latter dye in PC as a waveguide is assumed to be 0.23. In
table 7.5 the calculated maximum relative LSC efficiency after application of 400 nm gradient
pitch and 175 nm broad layered cholesteric reflectors and the onset wavelength of the reflector
are displayed.
Table 7.5 Calculated maximum relative efficiency ( ,maxLSC ) of LSCs containing the coumarin, perinone and
rare earth complex as luminophore after application of 400 nm gradient pitch and 175 nm layered narrowband
cholesteric reflectors
Luminophore
sl
400 nm gradient pitch 175 nm layered
,maxLSC
Onset
wavelength
(nm)
Relative LSC
efficiency
Onset
wavelength
(nm)
Coumarin 0.46 1.498 520 1.434 520
Perinone 0.325 1.259 700 1.175 740
Rare earth
complex
0.23 1.163 650 1.121 700
Since the fraction of the emitted photon that are lost through the LSC surface is smaller
for the dyes presented in table 7.5 than for Red 305, the maximum relative LSC efficiency after
application of the cholesteric reflectors is expected to be significantly lower. It also can be seen
that the difference in maximum increase in LSC efficiency between using the 400 nm and the
175 nm broad reflectors for the LSC containing the coumarin dye is smaller than for all the
other luminophores. This is a result from the more narrow spectral width of the emission band
of the coumarin dye.
Organic wavelength selective reflectors
135
7.6 Conclusions
The addition of narrowband selectively-reflective cholesteric layers to the top of the waveguide
of a LSC containing Red 305 as luminophore using an air gap both reduces the loss of light
from the surface of the waveguide and increases the light output at the edge up to 12%, which
is a recovery of more than 30% of what is normally lost from the waveguide surface. The
spectral width of these reflectors is not broad enough to reflect all the surface emissions back
into the waveguide.
400 nm broad reflectors made from polymeric cholesteric liquid crystalline films can
theoretically reflect over 90% of all surface emitted photons back into an LSC containing Red
305 as a luminophore. However, these cholesterics reflect away a part of the absorbable
incident light if the spectral position of the reflector is matched to this maximum efficiency.
Calculations suggest the 400 nm broad reflector could increase LSC edge emission efficiency up
to 66%. Reflectors with a more narrow reflection band have reduced effect on the increase of
the LSC efficiency: 175 nm broad reflectors could increase this efficiency around 45%.
Experiments demonstrate that applying a 175 nm broad reflector to an LSC with Red 305 as a
luminophore and a peak absorbance of ~0.5 increases the efficiency of the actual LSC by only
5%. The main reason for this discrepancy is the photon recycling of the back reflected photons.
When the amount of photon recycling is decreased by lowering the peak absorption to a value
below 0.1, the relative increase in LSC efficiency becomes nearly 30%; similarly, reducing the
probability of photon recycling by reducing the coverage of dye coating increases the relative
LSC efficiency 27%. These experiments demonstrate reducing the recycling of back reflected
photons increases the effectiveness of the cholesteric reflectors.
The reflector utilized needs to match the luminophores used. The maximum performance
increase is determined by the fraction of photons emitted in surface loss mode. More
reproducible, higher-quality reflectors coupled with a reduction in the amount of photon
recycling could make the cholesteric reflectors very effective in enhancing LSC performance.
Reducing the amount of photon recycling, could be achieved by placing lenses on top of the
patterned waveguides [106] or using luminophores with no or low overlap between the
absorption and emission band, like complexes of rare-earth ions with organic ligands, [155]
quantum dots, [83] or phosphors. [82]
8 Wavelength selective reflectors
and indirect sunlight6
6 Also published in:
P.P.C. Verbunt, M.G. Debije, D.J. Broer, C.W.M. Bastiaansen and D.K.G. de Boer, "Organic
wavelength selective mirrors for luminescent solar concentrators", Proc. SPIE 8438, 843805
(2012)
Chapter 8
138
8.1 Influence of the angle of incident light on the
performance of cholesteric reflectors
In chapter 7 it was shown that organic wavelength selective mirrors made from cholesteric
polymers show an angularly dependent reflection. In the calculations, the assumption was made
that the incident sunlight always encountered the LSC normal to the surface. In reality, the
LSC is a fixed device and the position of the sun changes during the day; additionally, clouds,
surfaces like buildings, cars, trees, and other common objects will scatter sunlight and result in a
mixed environment of both direct and indirect light. In this chapter the influence of this
indirect sunlight and the angular dependence of the direct sunlight on the effectiveness of these
organic wavelength selective mirrors used in conjunction with LSC devices will be calculated. In
the second part of this chapter a solution for the angular dependent reflection band of the
organic wavelength-selective reflectors is described
Previously we calculated the effect a cholesteric reflector on the output of an LSC
containing Red 305 as a luminophore when placed on top of the LSC waveguide. The
efficiency of the LSC could be increased by 66% if all emitted photons that were reflected back
into the LSC by the cholesteric reflector could be redirected to leave the edge(s) of the LSC
without additional losses. These calculations assumed incident sunlight reaches the device
normal to the surface. This is, of course, almost never the real situation. LSCs are designed to
be static devices, and with the sun travelling along the horizon, the angular dependence of these
mirrors becomes important.
8.1.1 Influence of the angle of the incident light on the light that
passes through the reflector
At larger incidence angles, the reflection band of the mirrors shifts to shorter wavelengths and
often into the spectral absorption band of the luminophore, resulting in the rejection of useful
incident light. The angular dependence of the absorbable sunlight that passes through the
mirror for cholesterics with a reflection bandwidth of 400 nm, 175 nm and 75 nm were
calculated using equation 7.8 for LSCs using Red 305 as luminophore. The spectral reflectivity
of the mirrors was chosen on the basis of the results of the previous chapter, so the onset
wavelength of the mirrors is 620 nm, 650 nm, and 650 nm respectively. The results of the
calculations are shown in figure 8.1.
Wavelength selective reflectors and indirect sunlight
139
Figure 8.1 The calculated fraction of the absorbable incident sunlight for an LSC waveguide containing Red305
that passes through the cholesteric reflector for 400 nm (black), 175 nm (dark grey) and 75 nm (light gray)
reflectors as a function of incidence angle.
At incidence angles up to approximately 30-40 degrees there is almost no change in the
absorbable light that passes through all the filters, although the broader filter with a shorter
onset wavelength starts to reflect absorbable sunlight away at smaller angles in comparison to
the reflectors with more narrow reflection bands. If the angle of incidence is increased beyond
30-40 degrees, the reflectors start rejecting a significant fraction of sunlight that can normally be
absorbed by the dye. The reflection continues to increase with increasing incidence angle until
almost no light is passing through the mirror. The 75 nm reflector shows a slightly different
trend. Around 50 degrees the useful amount of light that passes through the reflector appears
to stabilize as the reflection band is not broad enough to reflect all absorbable light away and at
higher angles the longest wavelengths of light that can be absorbed by the dye will again be able
to pass through the reflector. At very high angles the amount of useful light that passes through
the narrowband reflector decreases again, caused by the increased Fresnel surface reflections.
This latter effect is similar for all the reflectors.
8.1.2 Dependence of maximum possible LSC efficiency with change
of the angle of the incident light
If useful light is rejected by the cholesteric reflectors, less light is available to be absorbed,
leading to less light that eventually leaves the edge of the LSC. The increase in the efficiency of
the LSC by addition of a cholesteric reflector was calculated as a function of the incoming angle
Chapter 8
140
according to the method described in the previous chapter. The efficiency of reflection of
surface emitted photons is displayed in table 8.1.
Table 8.1 Efficiency of the cholesteric reflectors towards isotropically surface emitted photons of an LSC
containing Red 305 as a luminophore.
Width of the reflector chol
400 0.87
175 0.65
75 0.53
Figure 8.2 The maximal increase in efficiency for LSCs containing Red 305 dye after application of organic
wavelength selective mirrors as a function of angle of incident light for bandwidth 400 nm (black), 175 nm
(dark grey) and 75 nm (light gray) reflectors.
The maximum achievable increase in LSC-efficiency as function of the angle of incidence
is depicted in figure 8.2. As the incidence angle becomes larger than 30-40 degrees, the
enhancement in LSC efficiency caused by the cholesteric begins to decrease due to a reduction
in useful light that passes through the reflector. When the angle of the incidence light reaches
45 degrees for the narrowband reflectors and ~50 degrees for the two broadband reflectors, the
LSC is emitting less light from the edges than when no reflector is applied. At very high angles
it appears that all light that is normally emitted from the LSC edge is lost when the cholesteric
reflectors are applied. In reality, even a bare LSC does not have much light leaving the edges
since most incident light is reflected away at the LSC surface due to Fresnel reflections. These
Wavelength selective reflectors and indirect sunlight
141
latter reflections have not been taken into account in these calculations since they are equal for
systems both with and without the reflectors (which are applied to the LSC with an air gap).
8.1.3 Indirect sunlight
To this point we have only considered the incoming sunlight to be direct, with uniform incident
angle. LSCs applied in the built environment will almost never be exposed purely to direct
sunlight. The distribution of the incoming sunlight is very different each day of the year, at
each place on earth. It is thus difficult to define a general incident light distribution for use in a
simulation, although attempts have been made to represent them with distribution equations,
for example by Hopkins[232]. Such models require inputs that vary for each weather condition,
date and location. To gain some insight into the influence of the distribution of the incidence
angles on the performance of the cholesteric-topped LSCs, the increase in LSC efficiency after
application of the cholesterics using four different distributions of incident light is calculated.
The first distribution has already been described: direct light incident at a single angle (normal
to the device). A second assumes a totally isotropic distribution of incident light, approximating
days that are completely cloudy in an urban environment. Two distributions of incident light
falling between these extremes were also selected, described by cos (Lambertian) and
2cos . These last two distribution functions are chosen somewhat arbitrarily and only vaguely
represent the actual distribution functions of incoming sunlight on semi-cloudy days, but allow
to gain insight into the effect of angular distribution of the incident light on the effectiveness of
cholesteric reflectors. Calculations were performed using cholesteric reflectors placed on top of
an LSC and illuminated with these four distributions of incidence light, each centred around the
direction normal to the LSC surface. Sunlight that encounters the LSC under larger angles has
a greater chance of being absorbed than light entering the device perpendicularly, as the light
must traverse a longer pathlength through the LSC. Using Lambert-Beer, one may derive the
angular dependent absorption:
0
cos
AA (8.1)
where A is the absorption at angle and 0A is the absorption at normal incidence. At
higher angles where the absorption is larger, the cholesteric reflectors potentially reflect away a
significant fraction of the useful incident light and the Fresnel reflections from the LSC surface
increase as well decreasing the amount of light entering the waveguide. All these events
influence the LSC efficiency after application of a cholesteric reflector, the results are shown in
table 8.2.
Chapter 8
142
Table 8.2 Calculated relative efficienc of LSCs containing Red305 after application of organic wavelength
selective reflectors with different reflection bandwidth for direct (normal to the device) and indirect incident
sunlight. Output of a bare LSC under identical lightning conditions is given as 1.
Distribution of the
incoming light
400 nm broad
reflector
175 nm broad
reflector
75 nm broad
Reflector
Direct 1.66 1.45 1.35
2cos 1.29 1.22 1.29
cos 1.09 1.06 0.90
Isotropic 0.70 0.72 0.43
As can be seen in table 8.2 the increase in LSC efficiency after application of a cholesteric
reflector decreases with increasing fraction of indirect light approaching the device. For all
three bandwidths of reflection the LSC-efficiency is decreased for isotropic incident light when
a cholesteric reflector is applied.
In the urban environment, part of the sky can be blocked by a number of objects, for
instance buildings and trees. Therefore, the amount of light reaching the Earth at ground level
at very large incidence angles can be low. To simulate these conditions the same distribution
functions described above where used, but only the angles between 0 and 30 degrees or
between 0 and 45 degrees were considered and the results are displayed in table 8.3.
Table 8.3 Calculated increase in efficiency of LSCs containing Red305 after application of organic wavelength
selective reflectors with different reflection bandwidth for indirect incident sunlight with incidence angles constricted
to 0-45 degrees and 0-30 degrees.
Distribution of the
incoming light
400 nm broad
reflector
175 nm broad
reflector
75 nm broad
Reflector
0-45
degrees
0-30
degrees
0-45
degrees
0-30
degrees
0-45
degrees
0-30
degrees
2cos 1.50 1.60 1.39 1.43 1.36 1.36
cos 1.49 1.60 1.39 1.43 1.21 1.30
Isotropic 1.47 1.59 1.38 1.43 1.07 1.23
If the angles of the scattered incident light are limited to only a 30° or 45° cone, the
increase in LSC efficiency is mostly comparable to that of direct incident light. The 75 nm
reflectors seem to suffer the most from the indirect light in comparison to the broader
reflectors. If a 30° limitation for the incident light is considered, the decrease in relative LSC-
efficiency is minimal. These results show that incident light with angles larger than 30° cause
the decrease in effectiveness of the cholesteric reflectors on top of an LSC.
Wavelength selective reflectors and indirect sunlight
143
8.2 Special dispersion cholesteric reflectors
The blue shift in reflection band of cholesteric reflectors causes a decrease in effectiveness of
the reflectors on top of an LSC, when illuminated with indirect sunlight. A large fraction of the
incident light than could be absorbed by the dye molecules is reflected from the device at larger
angles. This loss can be reduced if the reflector would have a reduced angular dependency.
8.2.1 Introduction
The reflective properties of cholesteric reflectors are a result of the birefringence of the LC-
material. This birefringence forms a periodic change in refractive index (for linearly polarized
light) when in a helical structure causing the material to act as a Bragg reflector. When
birefringence is absent, the reflector is transparent to the incident light.
The ordinary and extraordinary refractive indices of the host liquid crystals used to form
the cholesterics have been considered constant over all wavelengths up to now. As can be seen
in chapter 7, this assumption was reasonably valid for most reflectors, but in reality the
refractive indices are dependent on the wavelength of the light, especially at short wavelengths
where the change in refractive index becomes more pronounced. The wavelength dependence
of the refractive index of a material (called dispersion) is described by Cauchy’s equation:
1 20 2 4
......B B
n n
(8.2)
where n is the refractive index at a specific wavelength, 0n is a null-refractive index and
1 2, ,B B etc are coefficients. In general only the first term ( 1
2
B
) is considered, and usually is a
good approximation. The dispersion is different for both the ordinary and the extraordinary
components of the refractive index, leading to a change of birefringence with changing
wavelength of the incident light, but this variation is generally small for the materials considered
in the previous chapters.
There are materials that have a large difference between the dispersion of the ordinary and
the extraordinary refractive indices. This difference in dispersion can result in the condition that
at a specific wavelength both the ordinary and the extraordinary refractive indices are equal and
there will be no birefringence, while at shorter and longer wavelengths this material will
demonstrate birefringence. If there would be a cholesteric phase of this material the reflection
band would disappear at the wavelength where there is no birefringence, since the reflective
properties of a cholesteric reflector are based on the birefringent properties of the LC-material.
[233]
The properties of this type of material (called in this chapter ‘special dispersion material’)
has been calculated using the Berreman’s 4x4 method, [229] as in chapter 7. The extraordinary
refractive index is kept constant at 1.68, while the ordinary refractive index was chosen to
display dispersion. The dispersion was chosen in such a way that the birefringence equals zero
Chapter 8
144
at 575, 600 or 625 nm. There are two variables in equation 8.2 that determine the dispersion,
the value of the null refractive index and the first coefficient ( 1B ). Both these variables have
been varied to manipulate the dispersion of the ordinary refractive index such that the
birefringence equals zero at the desired wavelengths. The input data for the ordinary refractive
index are given in table 8.4 and depicted in figure 8.3.
Table 8.4 Dispersion parameters of the ordinary refractive index special dispersion material. (graphical
representations of the dispersion of the ordinary refractive index are indicated in the brackets which refer to figure
8.3)
Null refractive
index 1B Birefringence zero at:
1.403 0.0916 575 nm (a, black)
1.426 0.0916 600 nm (a, dark grey)
1.446 0.0916 625 nm (a, light grey)
1.377 0.1002 575 nm (b, black)
1.377 0.1097 600 nm (b, dark grey)
1.377 0.1184 625 nm (b, light grey)
Figure 8.3 Dispersion of the refractive indices. The extra ordinary refractive index (black dotted) is constant
with wavelength, while the ordinary refractive index changes. In a) the dispersion is varied by varrying the null
refractive index and in b) the dispersion is varried by changing parameter 1B .
The reflective properties of the cholesteric phase formed by these materials are calculated
according to the method mentioned in chapter 7 for 75 nm, 175 nm and 400 nm broad
reflectors with an onset wavelength of 620 nm, 620 nm and 650 nm respectively, similar as the
reflectors used in section 8.1. An example of angular dependent reflective properties from 400
nm broad reflector is depicted in figure 8.4.
a) b)
Wavelength selective reflectors and indirect sunlight
145
Figure 8.4 Example depictions of the reflective properties of a cholesteric reflector made from a special dispersion
LC material, with the birefringence set at 0 at 575 (b) nm and 625 (a) nm. The reflectivity is scaled from 0%
(dark blue) - 100% (dark red)
The difference in reflective properties between the special dispersion cholesteric (SDC)
reflectors with the same wavelength of zero birefringence, using different dispersion parameters
is relatively small. Therefore, in the rest of this chapter, only the special dispersion materials
depicted in figure 8.3a are considered. In figure 8.4b it can be seen that at, an additional
reflection band appears at the short wavelength side of the reflector, if the blue shift at larger
angles passes the wavelength where birefringence is zero.
8.2.2 Transmission of sunlight
The angular dependence of the absorbable sunlight that passes through SDCs with a reflection
bandwidth of 400 nm, 175 nm and 75 nm were calculated using equation 7.8. The results of
the calculations are shown in figure 8.5.
Due to the reduced angular dependence of the reflective properties of the SDC materials.
the light that can be absorbed by the dye molecules that passes through the filter is constant
with increasing angle until 70° for all three widths of the reflection band. At higher angles a
large fraction of the absorbable incident light is still reflected by the cholesteric filters, but this
is a result of Fresnel reflections. For all dispersion parameters and for the three widths of the
reflection band the results have similar trends. At angles larger than 30° these SDC filters
transmit more light that can be absorbed by the dye molecules than the regular cholesteric
reflectors.
Chapter 8
146
Figure 8.5 Transmittance of incident light that can be absorbed by Red 305 through a regular cholesteric (dotted
line) reflector or special dispersion cholesteric reflectors with birefringence equal zero at 575 nm (black), 600 nm
(dark grey), or 625 nm (light grey). The cholesteric reflectors have a bandwith at normal incidence of 400 nm
(a), 175 nm (b), or 75 nm (c) and the onset wavelengths is 620 nm for the reflector with a 400 nm bandwith
and 650 nm for the reflectors with a 175 nm or a 75 nm bandwidth .
8.2.3 Efficiency towards surface emitted light
The change in angular dependence of the reflection band will also have an influence on the
fraction of surface emitted photons that are reflected by the SDC layers. The efficiency of the
reflectors towards surface emitted photons is calculated according to equation 7.7 and the
results are depicted in table 8.5.
a) b)
c)
Wavelength selective reflectors and indirect sunlight
147
Table 8.5 The efficiency of the special dispersion cholesteric reflectors towards surface emitted light from an LSC
with Red 305 as a luminophore. Reflection of all surface emitted photons is given as 1.
Birefringence equals
zero at
Width of the reflection band
400 nm 175 nm 75 nm
575 nm 0.59 0.44 0.36
600 nm 0.48 0.39 0.32
625 nm 0.40 0.34 0.28
Independent of the width of the reflection band the efficiency of the SDC materials
decreases with increasing wavelength where the birefringence is zero. 625 nm is exactly the
wavelength where the emission of the dye peaks. This could be the reason why zero
birefringence at this wavelength results in minimal reflection efficiency, since all the emitted
photons with a shorter wavelength will not be reflected by the cholesteric filter. The efficiency
of all these SDC filters is lower than the same reflector made from regular nematic liquid
crystalline materials as a result of a reduced reflectivity at wavelengths close to the wavelength
where the birefringence is zero. This reduced reflectivity can be seen in figure 8.4.
8.2.4 Angular dependent increase in LSC efficiency
Increased transmission of absorbable sunlight incident on the device at higher angles along with
reduced efficiency towards surface emitted light of the SDCs compared to a regular cholesteric
reflector will have a major influence on the angular dependence of effectiveness of these
reflectors on top of an LSC waveguide. The angular dependence of the maximum relative LSC
efficiency after application of a special dispersion cholesteric is calculated and displayed in
figure 8.6.
If light incident only normal to the device encounters the cholesteric reflector, increase in
LSC efficiency is 20-40% if the 400nm broad SDCs are applied; compare this to a regular
cholesteric which increases the LSC efficiency by 66%. This drop in efficiency enhancement is
a result of the reduced amount of surface emitted light reflected by the SDC reflectors. The
enhancement in LSC efficiency decreases with increasing incidence angle for regular
cholesterics when light has an angle larger than ~30°. For the SDCs the increase in LSC
efficiency only decreases at incidence angles larger than ~70°. This results in an overall better
performance of the LSC with the SDCs in comparison to the regular cholesterics at incidence
angles above ~40°. The regular 400 nm broad reflector increase the LSC efficiency up to ~50°,
while the special dispersion cholesteric will increase the LSC efficiency up to 65-75°. At larger
angles the amount of absorbed light in the LSC decreases drastically due to an increase of the
Fresnel at these very large angles, so any decrease in LSC performance at those angles does not
influence the overall performance.
Chapter 8
148
Figure 8.6 The relative efficiency of an LSC containing Red 305 after application of a regular cholesteric
(dotted line) reflector or special dispersion cholesteric reflectors with birefringence equal zero at 575 nm (black),
600 nm (dark grey), or 625 nm (light grey). The cholesteric reflectors have a bandwith at normal incidence of
400 nm (a), 175 nm (b), or 75 nm (c).
The results of the 175 nm (figure 8.6b) and the 75 nm (figure 8.6c) broad reflectors show
similar trends to the results of the 400 nm broad reflectors described above. The largest angles
of incidence at which the cholesteric reflectors will still increase the efficiency of the LSC when
applied are shown in table 8.6
Table 8.6 The largest angle of incidence where the cholesteric filters still increase the efficiency of the LSC
Birefringence zero at Broadness of the reflection band
400 nm 175 nm 75 nm
Regular 48° 50° 36°
575 nm 74° 70° 67°
600 nm 72° 70° 67°
625 nm 67° 68° 64°
a) b)
c)
Wavelength selective reflectors and indirect sunlight
149
The results shown above demonstrate indirect sunlight will have a major influence on the
effectiveness of these cholesteric reflectors, so the relative LSC efficiency after application of
these SDC reflectors for incident light with an isotropic, cos (Lambertian) and 2cos
distribution were calculated as well. The results are depicted in figure 8.7.
Figure 8.7 The relative efficiency of an LSC containing Red 305 after application of a regular cholesteric (white)
reflector and special dispersion cholesteric reflectors with birefringence equal zero at 575 nm (black), 600 nm
(dark grey), or 625 nm (light grey) is illuminated with indirect incident light. The distributions of the incoming
light are isotropic, cos , 2cos or direct, where is with respect to the normal of the reflector.
For all three distributions of indirect sunlight the special dispersion filters show an
increase effect in comparison to the normal cholesteric reflector, except for the 75 nm broad
reflector illuminated with incident light that has a 2cos distribution. It is also clear that the
filters with zero birefringence at 575 show higher relative LSC efficiency than the other SDC
reflectors. The 400 nm broad reflector with zero birefringence at 575 nm shows an increase in
LSC efficiency even for completely isotropic incident light, resulting in a higher LSC efficiency
if applied under any weather condition. The results of the 400nm broad reflectors from figure
8.7 are also displayed in table 8.7.
a) b)
c)
Chapter 8
150
Table 8.7 The relative efficiency of an LSC containing Red 305 as a luminophore after application of a
wavelength selective reflector made from special and regular dispersion nematic liquid crystals with birefringence
equal zero at 575 nm, 600 nm, or 625 nm is illuminated with indirect incident light. The distributions of the
incoming light are isotropic, cos , 2cos or direct, where is with respect to the normal of the reflector.
The output of a bare LSC under identical conditions is given as 1 in each case.
Zero
birefringence at:
Direct 2cos cos Isotropic
Regular 1.663 1.285 1.089 0.702
575 nm 1.398 1.347 1.272 1.010
600 nm 1.281 1.250 1.182 0.938
625 nm 1.213 1.177 1.108 0.872
8.3 Conclusions
Calculations performed in this chapter imply that sunlight illuminating an LSC device with
incident angles larger than 30 degrees negatively impacts the effectiveness of cholesteric
reflectors on top of the LSC. At these larger incidence angles, a fraction of the sunlight that
normally can be absorbed by the dye is instead reflected away because of the blue-shift in
reflection band of the cholesteric reflector. However, the cholesteric remains effective in
improving overall LSC performance for incident angles up to 45-50 degrees. At increasingly
larger angles, the increased amount of incoming light that is reflected away negatively impacts
LSC performance. Sunlight that reaches the earth is never completely direct. Clouds and objects
like trees and buildings scatter, block and/or reflect sunlight leading to indirect sunlight. In
these conditions, the cholesteric shows a more subdued performance. In practice, careful
consideration as to the expected light quality should be made before the decision is made
whether to apply cholesteric reflectors.
The limitations of the cholesteric performance on LSCs in indirect light can be
dramatically alleviated by using cholesteric reflectors made from special dispersion liquid
crystalline materials. These special dispersion cholesteric reflectors have reduced angular
dependence due to zero birefringence of the liquid crystalline materials at a specific
wavelengths. If this wavelength is chosen carefully around the spectral absorption peak of the
luminophore in the LSC, the amount of absorbable light that is reflected away by the filter at
larger angles. At angles > 75 degrees the special dispersion cholesteric reflectors start to reflect
absorbable sunlight, due to increased Fresnel-reflections similar to normal cholesteric filters. If
a 400 nm broad special dispersion reflector with zero birefringence at 575 nm is applied on top
of an LSC containing Red 305 as a luminophore the efficiency of the LSC is increased in all
weather conditions, even if the device is illuminated with isotropic distributed light.
9 Technology assessment and
future possibilities
Chapter 9
152
9.1 Luminescent solar concentrator: the future
9.1.1 Energy generating applications
One challenge the LSC has to confront in order to position itself in the global solar-energy
generation space is the misunderstanding of its functionality, and difficulty in describing the
device with reference to other solar-energy generating systems. Despite its coloration and
composition, the LSC should not be viewed as just another type of organic photovoltaic (OPV)
or, for that matter, as a solar cell at all. As such, direct comparisons of electron-generation
efficiencies commonly reported for photovoltaic materials become close to meaningless. While
it is possible to tabulate the reported results for efficiency measurements of luminescent solar
concentrators over the years (such an attempt can be seen in table 9.1) it becomes obvious that
making a comparison is very difficult.
Table 9.1 Reported efficiencies of luminescent solar concentrator devices. Luminophore Cell type (# of cells) LSC size [cm] Efficiency [%] Reference
Coumarin/Rhodamine? Si ? 1.9 [91]
Not stated Si ? 2.5 [91]
DCM Si 120x100x0.4 1.3 [91]
Coumarin/Rhodamine Si 120x100x0.4 1.3 [91]
Not stated Si 40x40x0.3 2.1 [130]
Not stated GaAs ? 2.5 [130]
Not stated, 2 plates Si 40x40x0.6 3.0 [130]
Not stated, 2 plates GaAs 40x40x0.6 4.0 [130]
Coumarin/Rhodamine Si 140x140x3 3.2 [234]
CdSe/CdS QDs GaAs 140x140x3 4.5 [234]
CdSe/CdS QDs Si 5x5x0.3 2.1 [85]
Red 305 Si 5x5x0.3 3.3 [85]
Red 305 Si 5x5x0.3 2.4 [71]
Red 305/CRS040 Si 5x5x0.3 2.7 [71]
Red 305/CRS040 mc-Si (1) 5x5x0.5 2.7 [74]
Red 305/CRS040 GaAs (1) 5x5x0.5 4.6 [74]
Red 305/CRS040 GaAs (4) 5x5x0.5 7.1 [74]
BA241 GaInP (4) 2x2x0.3 5.1 [75]
BA241/BA856 GaInP (4) 2 LSCs at
2x2x0.3
6.7 [75]
BA241 GaInP (1) 5x10x0.5 2.6 [75]
Not stated a-Si 5x5x0.5 ~0.7 [178]
CdSe core/multishell QDs Si 4.95x3.1x0.4 2.8 [235]
Red 305/Perinone Si (2) 2 LSCs at
5x5x0.5
4.3 [90]
Technology assessment and future possibilities
153
The efficiencies reported in the table are entirely dependent on the nature of the attached
solar cell, dye materials used, and device size. The reported net efficiency tells little or nothing
about the features or level of performance of the waveguide itself, the most important
component of the device. Thus, it becomes increasingly apparent to this author that another
method for describing the performance of these devices needs to be adopted, perhaps some
manner of ‘total photon-in/ photon-out’ fraction, weighted in some way to account for the
standard response for a variety of solar cells.
There are a great number of improvements that can and must be made to the LSC to
make it a more viable option for use in the urban environment. The single most important
improvement before the LSC may come into general use still needs to be the luminophore,
even after more than three decades of research. There is still no luminophore with a broad
spectral absorption, high absorption efficiency over the whole absorption spectrum, a large
Stokes Shift, a high luminescent efficiency (quantum yield), a matching spectrum of the emitted
photons to the best spectral response of the PV cell (≈1.14eV for silicon), a high photostability,
and good solubility in the host-matrix material. Organic dyes still have a small spectral
absorption width, a relatively low Stokes shift, and moderate photostability. Especially in the
broadness of the spectral absorption there is a lot to gain. Red 305 the state-of-the-art dye only
absorbs approximately 30% of all incident sunlight. If this spectral coverage could be increased,
this could lead to an increase in LSC efficiency by up to three times. Quantum dots lack the
quantum yield in combination with a large Stokes shift and solubility, and for rare Earth
materials the absorption coefficient and quantum yields are still too low. Phosphors are
alternative materials that have relatively high quantum yields, good absorption properties, high
photostability, and a high Stokes shift, but the solubility in an organic matrix is still a problem.
There are also still photons in the solar spectrum that can not be used by the
aforementioned luminophores, such as photons with lower energy than the bandgap of the PV
cell. To use these types of photons up converting systems could be an option, combining
multiple low-energy photons into a single, higher-energy photon that could be used by the cell.
Another option is to include materials that may absorb a single higher-energy photon well
above the photovoltaic bandgap and emit multiple photons closer to the cell bandgap, known
as quantum-cutting. Another obstacle that must be overcome is the misrepresentation of the
capabilities of the device: for years it has been espoused as a future replacement of PV panels
for rooftops. We believe this is a mistake. On South-facing rooftops with direct lighting (free
from shading), where space and efficiency are at a premium, standard silicon-based
photovoltaics would seem the best option. However, there are huge areas that do not fall into
these categories that could be prime locations for the use of LSCs. There are a large number of
examples of luminescent objects being incorporated in the environment already, even though
the coloration performs no additional function except for visual impact: for an example,
Chapter 9
154
consider the Musac museum in Léon, Spain. These types of structures exist in our environment
anyway, and it would not appear to be such a great stretch to add energy-generating
functionality. What is necessary is to bring the architectural and building industries into the
conversation as these devices are developed. The tremendous design freedom afforded by the
devices could no doubt be well-exploited by the visions of these industries. Applications could
include sound barriers beside roadways, telephone poles, bus stop roofing, atrium panels, and
the like.
One other aspect the author finds particularly intriguing is to provide an opportunity for
the use of OPVs. One of the greatest challenges for the OPV has been the ability for the
organic molecules that make up the absorbing/transport layers efficient at utilizing the UV
portion of the spectrum, as well as survive the high energies of UV light which causes
premature degradation of OPVs through destruction of the dye materials. However, the LSC
does not illuminate the attached solar cell with a solar spectrum, but with a much more narrow
band of light, generally in the NIR. It is in precisely this range of wavelengths that OPVs
perform their best. Coupled with the lack of exposure to UV light, this could provide the OPV
with the first real niche application where it could excel.
9.1.2 Day lighting applications
Not all applications need to convert the emitted light into electricity. The LSC has also found
potential use as a day lighting device, rather than as an electricity generator. In this
configuration, an LSC consisting of a stack of dye-doped waveguides of different colours is
located on a rooftop, and the emission light from the waveguide edges is collected in a clear
polymer or glass cable and transported a distance into the interior of a building, for example. At
the end of the light pipe, the emission light is mixed to create white light, and passes through a
diffuser into the room beyond.
The LSC day lighting element displays advantages over alternative day lighting systems in
that they require no tracking or astigmatic correction. A key element in this manifestation of the
device is the necessity of long transport distances of luminophore- emitted light, and the colour
mixing of the light at the emission site. In particular, the blue portion of the spectrum is lacking
due to the unavailability of a blue emitter that is sufficiently photos table and, to compensate,
current designs have used blue light emitting diodes to provide the short wavelengths and
obtain the desired white-light emission.
9.2 Aligned luminophores
9.2.1 Static alignment of luminophores in LSCs: what can be gained?
In chapters 3 through 6 of this thesis the effect of aligning dichroic luminophores in LSCs has
been investigated to reduce the loss of photons through the surface or to enhance the emission
Technology assessment and future possibilities
155
from two edges of an LSC waveguide. A model is presented to calculate the emission profile of
a dye ensemble as function of the alignment direction, the amount of order and the incident
light. Aligning the dichroic dyes in a homeotropic fashion reduced the surface loss of an LSC
drastically, but the amount of incident light that is absorbed is reduced as well when illuminated
with a collimated light source. In a planar dye ensemble the surface loss is increased, but the
absorption of incident sunlight is increased. Calculations show that the amount of light emitted
from an aligned ensemble of dye molecules is the highest for planar alignment. These results
will probably change when photon recycling is taken into account. Each photon recycling event
will increase the amount of surface loss and for planar aligned dye ensembles the amount of
surface loss is increased more than for dye ensembles with homeotropic alignment. To get
more conclusive results based on these tilted dye alignments, the model presented in this thesis
should be implemented in ray tracing software, to simulate the effect of photon recycling. In
this light, the model presented in chapter 3 should be seen as a first step towards simulation of
the distribution of photons in the waveguide of an LSC.
The reduced absorption in the homeotropic state could be improved by the addition of a
forward-scatterer on top of the LSC, so the incoming light will be distributed and there will be
more incident light which is not parallel to the director of the dye ensemble.[225] Absorption of
incident light by homeotropically aligned dye molecules will also benefit from the fact that
sunlight is indirect and the direction is changing during the day. Photons incident at larger
angles will have a larger chance of getting absorbed.
Another application of static aligned dye ensembles is in the reduction in amount of
photovoltaic cells necessary for an LSC. Aligning dichroic dyes planarly enhances the edge
emission from the two edges parallel to the alignment direction up to ~60% over the
perpendicular edge if the dye aligns well in a nematic LC material. Calculations show that with
increased alignment, for instance in Smectic LC-materials, could even enhance this preferential
emission up to almost twice the amount of light reaching the parallel edge in comparison to the
perpendicular edge. Measurements show that the emission from the parallel edge of an LSC
with a planar aligned luminophore in a nematic LC is 30% higher than the emission from any
edge of an LSC with isotropic dye molecules. The calculations indicate that with enhanced
alignment this number can be increased to almost 50%. This would mean that an LSC with
planar aligned luminophores with only 2 photovoltaic cells could have an efficiency which is
about 75% of an LSC with isotropic dyes with 4 edges covered by PV-cells. This could reduce
the cost of the device. The cost could even be more reduced by replacing one of the two PV-
cells with a silver mirror.
Chapter 9
156
9.2.2 Dynamic alignment of luminophores in LSCs: window
applications
An option for making use of dye alignment in an LSC is to use the change in absorption of the
dichroic dye when aligned in different directions and use the LSC as a window while
simultaneously generating electrical current. While a static LSC has been introduced that could
be used as a window, it has the disadvantage common to thin-film transparent photovoltaics, in
that they may generate electrical current but not alter their transparency. By the same token,
responsive systems that can change their transparency, such as photo- and thermo chromic
windows or standard blinds generate no electricity.
An alternative LSC design uses, rather than a waveguide with embedded, inflexible dyes,
two glass plates coated with a conductor, and the space between the plates is filled by a LC-host
containing a fluorescent dye species. The liquid-crystal host can be continually switched
between two states, planar and homeotropic, and all angular configurations in between.
The dichroic dye molecules mirror the alignment of the LC-host. In the planar alignment,
which is the position of maximal absorption for the dye, the dye may emit light which partially
may be trapped in the glass panes making up the ‘window’ and generate electrical current, with
preferential emission from the edges parallel to the LC alignment. This constitutes the ‘dark’
state of the device, with maximum absorption (and consequently minimum transmission
through the ‘window’) and maximum emission, and thus maximum electrical generation with
minimal power input.
By applying of a voltage across the plates, the LCs may be made to attain an angle with
respect to the glass plate, reducing the absorption by the dye, and allowing more light into the
space beyond. In this ‘light’ state, while the output of the window necessarily drops due to
reduced light absorption, it still produces current, and due to the dye alignment, the optical
efficiency is actually increased. This design still needs work to achieve the desired transmissive
properties and coloration, but could potentially provide considerable advantages over
alternative ‘smart’ windows. It is also possible to create a third state, a scattering ‘privacy glass’
condition: this work is still in progress.
9.3 Wavelength selective reflectors in luminescent solar
concentrators
Organic wavelength selective reflectors can reduce the surface loss in luminescent solar
concentrators, but their effect is limited by photon recycling of emitted photons by another
luminophore molecule. If a luminophore which does not suffer from photon recycling is used
the amount of surface loss is reduced to approximately 25% for an LSC waveguide made from
polycarbonate. The increase in efficiency of such an LSC could reach up to 33% if all surface
emitted light would be reflected back into the device and reach the edge of the LSC and the
Technology assessment and future possibilities
157
wavelength selective reflector would not reflect away absorbable sunlight. Calculations show
that the efficiency of an LSC with a dye that does not suffer from photon recycling is increased
by approximately 18% when a reflector is added with a 400 nm bandwidth. In such a device the
application of organic wavelength selective reflectors would be beneficial. In an LSC with a dye
that suffers from photon recycling the effect of the mirrors are limited, as shown in
measurements, but the LSC efficiency is still increased when illuminated with direct sunlight
normal to the device. This effect can be increased by using reflectors with better reflective
properties. The reflectors used are not 100% reflective within the reflection band and were not
completely transparent outside the reflection band.
The angular dependency of the reflection band reduces the positive effect of the
cholesteric reflector when the light is not collimated and normal to the device. Especially at
large angles of incidence the angular dependency of the reflection band has a major role in the
effectiveness of the reflector. At large angles of incidence the reflector reflects so much
incoming sunlight away that the efficiency of the device is decreased after application of these
cholesteric reflectors.
The angular dependency of cholesteric reflectors can be reduced by the use of special
dispersion LC materials. Calculations show that if the dispersion is chosen correctly, the
reflector made from this LC-material can increase the efficiency of an LSC in any weather
circumstances. The availability of materials with the needed dispersion is probably the limiting
factor in the use of these special dispersion cholesterics in LSCs.
Besides the increase of LSC efficiency after application of organic wavelength selective
reflectors, aesthetics of the device are also changed. Cholesteric reflectors show a change in
reflection wavelengths with different angles of incidence. This results in a change of colour
when the sun changes position with respect to the device or the recipient of the reflected
sunlight changes position (for instance a person walking past the device). This effect could give
an architect more freedom of design when an LSC is implemented in a building.
If a cholesteric reflector should be integrated in the LSC is going to be a case by case
decision. In some instances the aesthetics are going to be the determining factor and in other
cases the efficiency of the device will be predominant. If the efficiency is the determining
factor, the position of the LSC and the luminophore used determines if the application of a
cholesteric reflector is beneficial.
Appendix A: Perylene perinone
dye7 A.1: Introduction
Application of the LSC has been curtailed by their limited ability to absorb a large enough
fraction of the incident sunlight. The most common luminophore used has been the perylene-
based dye Lumogen Red 305 from BASF [117] (see, for example, [115]). Red 305 has
demonstrated good photostability, [111,113] good solubility in both PMMA and polycarbonate,
the two workhorses for LSC research efforts, and a high measured fluorescent quantum yield
(FQY). [115] The greatest disadvantage of Red 305 has been its limited spectral coverage. With
an absorption peak of around 577 nm in polycarbonate, it is only able to absorb 16% of the
air–mass 1.5 global (AM1.5G) photons of the solar spectrum. Despite this, Red 305 has been
used in the highest efficiency LSC system. [74]
Perylene perinones, otherwise known as perylene bis-imidazoles, designed for a variety of
applications, have been described in the past. [236-239] In this work, we describe new perylene
perinone structures exhibiting extended absorption ranges, promising FQYs, and photostability
for use in a new application, the LSC. Two different types of perylene perinones are described,
3,4:9,10-bis(1,2-benzimidazole)-1,6,7,12-tetra(4-nonylphenolphenoxy)perylene (syn/anti-
isomers) and 3,4:9,10-bis(1,2-benzimidazole)-1,6,7,12-tetra(4-tert-octylphenoxy)perylene
(syn/anti-isomers), further referred to as the nonylphenol perylene perinone and the tert-octyl
perylene perinone respectively. We demonstrate the use of perylene perinone containing
waveguides in conjunction with Red 305 dye waveguides, and demonstrate significant increases
in the output of a dual-waveguide system compared to using a Red 305 waveguide alone.
This work has been performed in collaboration with Sabic IP in the Netherlands, GE
Plastics in India, and with Herriot-Watt University, Edinburgh, United Kingdom.
7 Also published in:
M.G. Debije, P.P. C. Verbunt, P.J. Nadkarni, S. Velate, K. Bhaumik, S. Nedumbamana, B.C.
Rowan, B.S. Richards and T.L. Hoeks, “Promising fluorescent dye for solar energy conversion
based on a perylene perinone”, Applied Optics, 50 (2), 163-169, 2011.
Appendix A
160
A.2 Methods
5 cm x 5 cm x 0.3 cm polycarbonate plates were prepared by the injection molding of six
different concentrations (between 15.5 and 250 parts per million (ppm)) of the perinone nonyl
phenol dye. The absorption spectra of all the plates were recorded on a Shimadzu UV-3102
spectrophotometer. The top surfaces of the samples and a reference waveguide containing no
dye were exposed to the light from a 300 W solar simulator with filters to approximate the
AM1.5G solar spectrum (Lot- Oriel) and an integrated output of 536 Wm-2 over the spectral
range 350–850 nm. Single edge emission spectra and intensity were determined using a
Labsphere spectral light measurement system LED 1050 integrating sphere. The emissions
from two orthogonal edges of each waveguide were averaged in determining the final output.
Emission spectra were integrated from 350 to 850 nm to determine the total output of the
waveguides: the variation in the integrated emissions from the two edges was <2%. The
outputs from two edges were also determined using a separate rear scatterer (an opaque piece
of white painted cardboard). Additional measurements of the edge outputs were made by laying
two waveguides on top of each other and determining the total edge output on a black
(absorbing) background. The injection-molded polycarbonate waveguides contained either the
perinone dye, Lumogen Red 305 (BASF), or no dye at all.
Weathering exposure was carried out in the Rockies Ci 4000 weatherometer from Atlas
Material Services. The protocol used continuous 0.75 Wm-2 irradiance at 30% relative humidity
with a dry bulb temperature of 35°C. Samples were compared with and without a covering of
300 μm SLX 2432- NA9A048T film (Sabic IP). All fluorescence emission and excitation spectra
were taken on a SPEX fluorolog 3 (Jobin Yvon, Edison) using double monochromators with
1200 grooves/mm grating blazed at 330 and 500 nm for excitation and emission selection,
respectively, a Hamamatsu R928-P photon counting photomultiplier tube (PMT) for emission
detection (referred to as the signal or S channel), and a 450 W continuous Xe lamp as an
excitation source. A small fraction of the excitation beam is diverted to a “reference”
photodiode just before the sample (referred to as reference or R channel) to monitor the
relative excitation intensity as a function of time and excitation wavelength. The voltage of S
channel PMT was set to 950 V for all experiments.
One set of FQY measurements was carried out by recording excitation source intensity
and calculating the area under the corrected emission spectra. The FQY value is a ratio of
“number of photons emitted” to “number of photons absorbed.” Additional FQY
measurements of similar plates were made using a FL3-211 Horiba Jobin Yvon Fluorolog-3
fluorescence spectrophotometer on an Edinburgh Instruments FS920 UV/VIS/NIR
fluorescence spectrometer with a Horiba Jobin Yvon integrating sphere, as described
previously. [115]
Perylene perinone dye
161
A.3 Results and Discussion
Polycarbonate plates containing the nonylphenol derivative were measured to have an
absorption peak at 632 nm, very close to the emission maximum of Red 305 of 620 nm. The
new dye thus expands the absorption spectra of the LSC by ∼50 nm, which translates into the
new LSC being able to harvest 30% more of the photons of the AM1.5G spectrum. See table
a.1 for additional properties of the perylene perinones.
Table a.1 Photophysical parameters and decomposition temperatures of the perylene perinones
Compound λabs (nm)a Ext. Coef.
(Lcm-1mol-1)
λem (nm)a Decomp.
Temp. (°C)
FQYa
Nonylphenol 630 63700 666 >378 1
4-tert-
octylphenol
630 61200 666 >358 0.96
a Measured in Toluene
The 5 cm × 5 cm × 0.3 cm injection-molded polycarbonate plates containing a range of
nonylphenol concentrations from 15.5 to 250 ppm were cut from samples of a slightly larger
size. The representative absorption and edge emission spectra of two of nonylphenol- filled
polycarbonate plates may be seen in figure a.1.
Figure a.1 Absorbance (solid curves) and edge emission (dotted curves) spectra of two nonylphenol-filled plates
containing 35.5 (black curves) and 250 (grey curves) ppm dye.
Appendix A
162
To evaluate the new perylene perinone dye in a LSC configuration, we measured light
emission from the edge of the filled waveguides, more completely described in the experimental
section. The top surfaces of the dye-filled waveguides and reference blank polycarbonate
waveguide were exposed to light from a solar simulator with the waveguides resting on either
black, absorbing backgrounds or white scattering backgrounds separated from the waveguide
by an air gap. The intensity and wavelength of the light emitted from the edge of the waveguide
were recorded using an integrating sphere. The integrated output energies (from 350 to 850
nm) are shown in figure a.2. Similar to previous findings, the fractional emission increase for
the samples using the scattering background was larger at lower dye content. [173] The 4-tert-
octylphenol had an output ∼90% that of the nonyl at a similar loading (100.7 to 101.8 mg
dye/kg).
Figure a.2 Integrated edge emission (350–850 nm) of the perylene perinone nonylphenol 5 cm × 5 cm × 0:3 cm
waveguides with absorbing black (filled symbols) and scattering white (open symbols) backgrounds.
Measurements of the perylene perinone nonylphenol FQY were performed in two
different solvents, resulting in values of around φ=100%±5% in toluene and φ =80±8% in
xylene. The FQY in polycarbonate plates loaded at 15.5 ppm was estimated to be φ =80±10%.
A series of edge emission measurements was made on 3 mm thick waveguides containing
either Red 305 or the perylene perinone. The samples were all measured in tandem, with one
waveguide laid on top of the other, with the ends of each inserted into the integrating sphere,
and the emission spectra were integrated to determine a total output for the system.
The results of such a series of measurements with the Red 305 waveguides lying on top of
perylene perinone waveguides of various dye content may be seen in figure a.3. A summary of
Perylene perinone dye
163
all the results of these measurements is found in table a.2. It would be anticipated that a two-
waveguide device, one employing Red 305 as an upper waveguide and a perylene perinone-
filled waveguide as a lower layer, will enhance emission of the Red 305 dye acting alone [75] for
several reasons. The first is because the FQY of the Red 305 dye is higher than that of the
perylene perinone. [115] Second, the perylene perinone dye will be able to absorb light that was
within the absorption range of the Red 305 dye that failed to be absorbed and absorb light with
wavelengths completely outside the absorption band of the Red 305 dye.
Figure a.3 Integrated emission (350–850 nm) determined for two waveguide systems with blank polycarbonate
(filled circles) and perylene perinones of absorbance 0.37 (filled square), 1.25 (open triangle), 2.04 (filled
triangle), and 3.55 (open circle) as the base waveguide and Red 305 waveguides of varying absorbance as the top
waveguide.
Appendix A
164
Table a.2 Integrated output for Red 305 samples on blank polycarbonate bottom waveguide and fractional
increase of Red 305 top/perinone bottom stacks compared to output of Red 305 top/blank polycarbonate
bottom waveguide stack
Integrated
output Red
305 (mW)
Perinone Absorbance (bottom)
0.37 0.93 1.25 1.36 2.04 2.86 3.55
Red 305
absorbance
(top)
0.21 17.3 0.66 1.08 1.12 1.61 1.65 1.78 1.82
0.48 25.5 0.45 0.71 0.72 1.00 1.15 1.15 1.16
1.04 36.5 0.28 0.42 0.43 0.60 0.67 0.69 0.69
1.63 45.6 0.16 0.29 0.28 0.44 0.43 0.46 0.47
2.43 50.5 0.24 0.28 0.24 0.28
Finally, the perylene perinone waveguide will be able to absorb the light escaping from the
rear surface of the Red 305 waveguide. As a significant fraction of the absorbed photons are
lost through the surfaces of the waveguide (see chapter 5), this latter effect could be significant.
To verify the advantage of using a Red 305 on top, measurements of the outputs of the dual-
waveguide system employing the perinone-filled waveguide as the top object were made. These
experiments demonstrated that, while in all cases the system performed better than the Red 305
with a blank polycarbonate plate on top, the perinone top/Red 305 bottom consistently
performed less efficiently than the Red 305 top/perinone bottom systems; see table a.3.
Table a.3 Fractional change of perinone top/Red 305 bottom stacks compared to identical waveguides with Red
305 top/perinone bottom
Red 305 absorbance
(bottom)
0.21 0.48 1.04
Perinone
absorbance
(Top)
0.37 -0.25 -0.09 -0.10
0.93 -0.18 -0.15
1.25 -0.16 -0.16 -0.21
1.36 -0.07 -0.08
2.04 -0.04 -0.12 -0.15
2.86 -0.04 -0.10
3.55 -0.06 -0.12 -0.20
As in any solar-based application, details of the photostability of the dye are of interest.
The study of the two perinone dyes in the polycarbonate matrix at a dye loading of 100 ppm
was done in an accelerated artificial weathering setup. [240] Some of the plates were covered
with a protective Lexan SLX film (Sabic IP) and compared to similar samples not employing
the protective film. The changes in fluorescent intensity as a function of 340 nm dose were
Perylene perinone dye
165
measured for the samples. The fluorescent efficiency of each plaque was determined after a
10000 kJm-2 exposure, equivalent to over three years of exposure in Florida [240]. The
reduction in emission intensity for the nonylphenol perinone dye monitored at the fluorescent
maximum of 678 nm was approximately 1.7% in the protected sample, while in the sample
without the protective layer, the loss was 5.4%. The fluorescent loss of the tert-octyl perinone
was monitored at 676 nm and was slightly lower than the linear chain nonyl: in the protected
system, loss was 1.2%, while the plate without the SLX film was 2.5%. This study indicates that
the photostability of the nonyl- and tert-octyl perylene perinones are at least comparable to Red
305 [111,113]and, thus, could be considered for longer term, outdoor use. This photostability is
not unexpected, as it has been shown previously that other perinones, such as Pigment Orange
43 and Pigment Red 197, are both photo and thermally stable.
A.4 Conclusions
In summary, we have prepared a novel fluorescent dye material that holds great promise for use
in solar energy research. Their use in a LSC may extend the absorption range of current dyes by
∼50 nm at a reasonable fluorescence quantum yield and photostability.
In addition, we have demonstrated a two waveguide LSC system using a Red 305 top and
perinone bottom waveguide that significantly increases the performance of the system based on
Red 305 alone. This work is an important step in making the LSC a viable option for electrical
generation from sunlight.
167
References [1] R. Judkoff, MRS Bulletin, 33(04), 449 (2008). [2] European Parliament and the Council of the European Union, Directive 2010/31/eu of the
European parliament and of the Council of 19 may 2010 on the energy performance of buildings, 2010
[3] M. A. Green, Progress in Photovoltaics: Research and Applications, 7(4), 317 (1999). [4] J. Zhao, A. Wang, M. A. Green, and F. Ferrazza, Applied Physics Letters, 73(14), 1991 (1998). [5] O. Schultz, S. W. Glunz, and G. P. Willeke, Progress in Photovoltaics: Research and Applications,
12(7), 553 (2004). [6] S. Benagli, D. Borello, E. Vallat-Sauvain, J. Meier, U. Kroll, J. Hotzel, J. Spitznagel, J.
Steinhauser, L. Castens, and Y. Djeridane, 24th European Photovoltaic and Solar Energy Conference, 2009,
[7] X. Wu, J. C. Keane, R. G. Dhere, C. Dehart, A. Duda, T. A. Gessert, S. Asher, D. H. Levi,
and P. Sheldon, 17th European Photovoltaic and Solar Energy Conference, 2001, 995 [8] B. L. Cohen, Geochimica et Cosmochimica Acta, 48(1), 203 (1984). [9] K. W. Boer, Energy Conversion and Management, 52(1), 426 (2011). [10] T. Tinoco, C. Rincon, M. Quintero, and G. S. Perez, Physica Status Solidi (a), 124(2), 427
(1991). [11] I. Repins, M. Contreras, M. Romero, Y. Yan, W. Metzger, J. Li, S. Johnstone, B. Egaas, C.
Dellart, J. Schorf, B. E. Mccandless, and R. Narfi, 33rd IEEE Photovoltaic Specialists Conference, 2008, NREL/CP
[12] M. A. Green, K. Emery, Y. Hishikawa, W. Warta, and E. D. Dunlop, Progress in Photovoltaics:
Research and Applications, 20(1), 12 (2012). [13] D. Crisp, A. Pathare, and R. C. Ewell, Acta Astronautica, 54(2), 83 (2004). [14] H. Shirakawa, E. J. Louis, A. G. Macdiarmid, C. K. Chiang, and A. J. Heeger, Journal of the
Chemical Society, Chemical Communications., (16), 578 (1977). [15] B. R. Weinberger, M. Akhtar, and S. C. Gau, Synthetic Metals, 4(3), 187 (1982). [16] J. J. M. Halls and R. H. Friend, Synthetic Metals, 851307 (1997).
168
[17] G. Yu, J. Gao, J. C. Hummelen, F. Wudl, and A. J. Heeger, Science 270 (5243 ), 1789 (1995). [18] M.-F. Falzon, M. M. Wienk, and R. A. J. Janssen, The Journal of Physical Chemistry C, 115(7),
3178 (2011). [19] R. F. Service, Science, 332(6027), 293 (2011). [20] A. Dupuis, A. Tournebize, P.-O. Bussiere, A. Rivaton, and J.-L. Gardette, The European
Physical Journal - Applied Physics, 56(03), 34104 (2011). [21] N. Koide, R. Yamanaka, and H. Katayama, MRS Proceedings, 12111211 (2009). [22] B. Alfa, M. T. Tsepav, R. L. Njinga, and I. Abdulrauf, Applied Physics Research, 4(1), 48
(2012). [23] I. J. Kramer and E. H. Sargent, ACS Nano, 5(11), 8506 (2011). [24] R. Hardman, Environmental Health Perspectives, 114 (2006). [25] J. L. Pelley, A. S. Daar, and M. A. Saner, Toxicological Sciences, 112276 (2009). [26] M. Bottril and M. Green, Chemical Communications, 477039 (2011). [27] E. Klampaftis and B. S. Richards, Progress in Photovoltaics: Research and Applications, 19(3), 345
(2011). [28] H.-J. Yang, C.-H. Chen, W.-C. Lai, C.-L. Wu, C.-F. Huang, Y.-H. Chen, and J.-C. Hwang,
Journal of The Electrochemical Society, 158(9), H851 (2011). [29] K.-E. Hassan and M. F. El-Refaie, Solar Energy, 15(3), 219 (1973). [30] L. D. Jaffe, Solar Energy, 42(2), 173 (1989). [31] N. Fraidenraich, C. Tiba, B. B. Brandao, and O. C. Vilela, Solar Energy, 82(2), 132 (2008). [32] D. Jing, H. Liu, X. Zhang, L. Zhao, and L. Guo, Energy Conversion and Management, 50(12),
2919 (2009). [33] S. Hatwaambo, H. Hakansson, A. Roos, and B. Karlsson, Solar Energy Materials and Solar
Cells, 93(11), 2020 (2009). [34] P. A. Davies, Applied Optics, 31(28), 6021 (1992). [35] R. Leutz and H. P. Annen, Solar Energy, 81(6), 761 (2007). [36] Y. Tripanagnostopoulos, C. Siabekou, and J. K. Tonui, Solar Energy, 81(5), 661 (2007).
References
169
[37] N. Yeh, Renewable and Sustainable Energy Reviews, 14(9), 2926 (2010). [38] R. Winston, J. C. Minano, and P. Benitez. 2005, San Diego, CA, USA: Elsevier. 497. [39] W. C. Dickinson, Solar Energy, 21(3), 249 (1978). [40] E. Lorenzo, M. Perez, A. Ezpeleta, and J. Acedo, Progress in Photovoltaics: Research and
Applications, 10(8), 533 (2002). [41] A. E.-S. A. Nafeh, International Journal of Numerical Modelling: Electronic Networks, Devices and
Fields, 17(4), 385 (2004). [42] H. Mousazadeh, A. Keyhani, A. Javadi, H. Mobli, K. Abrinia, and A. Sharifi, Renewable and
Sustainable Energy Reviews, 13(8), 1800 (2009). [43] R. M. Swanson, Progress in Photovoltaics: Research and Applications, 8(1), 93 (2000). [44] E. Munoz, P. G. Vidal, G. Nofuentes, L. Hontoria, P. Perez-Higueras, J. Terrados, G.
Almonacid, and J. Aguilera, Renewable and Sustainable Energy Reviews, 14(1), 518 (2010). [45] T. M. De Jong, D. K. G. De Boer, and C. W. M. Bastiaansen, Optics Express, 19(16), 15127
(2011). [46] A. Goetzberger, J. C. Goldschmidt, M. Peters, and P. Löper, Solar Energy Materials and Solar
Cells, 92(12), 1570 (2008). [47] L. H. Slooff, E. E. Bende, and T. Budel, 24th European Photovoltaics and Solar Energy
Conference, 2009, 336 [48] R. Tremblay, 1985, US4505264 [49] J. S. Batchelder, A. H. Zewail, and T. Cole, Applied Optics, 18(18), 3090 (1979). [50] W. H. Weber and J. Lambe, Applied Optics, 15(10), 2299 (1976). [51] A. Goetzberger and W. Greube, Applied Physics A: Materials Science & Processing, 14(2), 123
(1977). [52] J. A. Levitt and W. H. Weber, Applied Optics, 16(10), 2684 (1977). [53] D. J. Farrell and M. Yoshida, Progress in Photovoltaics: Research and Applications, 93 (2011). [54] M. Kennedy, S. J. Mccormack, J. Doran, and B. Norton, ISES World Congres, 2007,
170
[55] E. Bende, L. Slooff, A. Burgers, W. V. Sark, and M. Kennedy, 23rd European Photovoltaic and Solar Energy Conference, 2008, 461
[56] M. J. Currie, J. K. Mapel, T. D. Heidel, S. Goffri, and M. A. Baldo, Science, 321(5886), 226
(2008). [57] D. Chemisana, Renewable and Sustainable Energy Reviews, 15(1), 603 (2011). [58] A. Goetzberger, Applied Physics A: Materials Science & Processing, 16(4), 399 (1978). [59] M. Carrascosa, S. Unamuno, and F. Agullo-Lopez, Applied Optics, 22(20), 3236 (1983). [60] R. Kondepudi and S. Srinivasan, Solar Energy Materials, 20(3), 257 (1990). [61] F. Vollmer and W. Rettig, Journal of Photochemistry and Photobiology A: Chemistry, 95(2), 143
(1996). [62] E. Yablonovitch, Journal of the Optical Society of America, 70(11), 1362 (1980). [63] G. Smestad, H. Ries, R. Winston, and E. Yablonovitch, Solar Energy Materials, 21(2-3), 99
(1990). [64] R. W. Olson, R. F. Loring, and M. D. Fayer, Applied Optics, 20(17), 2934 (1981). [65] V. Wittwer, A. Goetzberger, and K. Heidler, 4th European conference on Photovoltaics and Solar
Energy, 1982, [66] A. Chatten, K. Barnham, B. Buxton, N. Ekins-Daukes, and M. Malik, Semiconductors, 38(8),
909 (2004). [67] T. J. J. Meyer, J. Hlavaty, L. Smith, E. R. Freniere, and T. Markvart, Physics and Simulation of
Optoelectronic Devices XVII, 2009, 72110N [68] B. S. Richards and K. R. Mcintosh, 21st European Photovoltaic Solar Energy Conference, (2006). [69] S. J. Gallagher, P. C. Eames, and B. Norton, International Journal of Ambient Energy, 2547
(2004). [70] A. R. Burgers, L. H. Slooff, R. Kinderman, and J. A. M. V. Roosmalen, 20th European
Conference on Photovoltaics and Solar Energy, 2005, [71] W. G. Van Sark, K. W. Barnham, L. H. Slooff, A. J. Chatten, A. Büchtemann, A. Meyer, S.
J. Mc.Cormack, R. Koole, D. J. Farrell, R. Bose, E. E. Bende, A. R. Burgers, T. Budel, J. Quilitz, M. Kennedy, T. Meyer, S. H. Wadman, G. P. Van Klink, G. Van Koten, A. Meijerink, and D. Vanmaekelbergh, Optics Express, 16(26), 21773 (2008).
[72] S. Tsoi, Thesis, 2011, Eindhoven University of Technology, pp 160
References
171
[73] T. Markvart, Journal of Applied Physics, 99(2), 026101 (2006). [74] L. H. Slooff, E. E. Bende, A. R. Burgers, T. Budel, M. Pravettoni, R. P. Kenny, E. D.
Dunlop, and A. Büchtemann, Physica Status Solidi (RRL) - Rapid Research Letters, 2(6), 257 (2008).
[75] J. C. Goldschmidt, M. Peters, A. Bösch, H. Helmers, F. Dimroth, S. W. Glunz, and G.
Willeke, Solar Energy Materials and Solar Cells, 93(2), 176 (2009). [76] S. A. El-Daly and S. Hirayama, Journal of Photochemistry and Photobiology A: Chemistry, 110(1),
59 (1997). [77] R. Soti, E. Farkas, M. Hilbert, Z. Farkas, and I. Ketskemty, Journal of Luminescence, 68105
(1996). [78] K. Geetha, M. Rajesh, V. P. N. Nampoori, C. P. G. Vallabhan, and P. Radhakrishnan,
Journal of Optics A: Pure and Applied Optics, 6(4), 379 (2004). [79] L. R. Wilson, B. C. Rowan, N. Robertson, O. Moudam, A. C. Jones, and B. S. Richards,
Applied Optics, 49(9), 1651 (2010). [80] A. A. Earp, J. B. Franklin, and G. B. Smith, Solar Energy Materials and Solar Cells, 95(4), 1157
(2011). [81] B. C. Rowan, L. Wilson, and B. S. Richards, 24th European Photovoltaics Conference, 2009, 346 [82] D. K. G. De Boer, D. J. Broer, M. G. Debije, W. Keur, A. Meijerink, C. R. Ronda, and P.
P. C. Verbunt, Optics Express, 20(S3), A395 (2012). [83] K. Barnham, J. L. Marques, J. Hassard, and P. O'brien, Applied Physics Letters, 76(9), 1197
(2000). [84] A. J. Chatten, K. W. J. Barnham, B. F. Buxton, N. J. Ekins-Daukes, and M. A. Malik, Solar
Energy Materials and Solar Cells, 75(3-4), 363 (2003). [85] S. J. Gallagher, B. Norton, and P. C. Eames, Solar Energy, 81(6), 813 (2007). [86] V. Sholin, J. D. Olson, and S. A. Carter, Journal of Applied Physics, 101(12), 123114 (2007). [87] G. V. Shcherbatyuk, R. H. Inman, C. Wang, R. Winston, and S. Ghosh, Applied Physics
Letters, 96(19), 191901 (2010). [88] A. M. Taleb, Renewable Energy, 26(1), 137 (2002). [89] M. G. Debije, P. P. C. Verbunt, P. J. Nadkarni, S. Velate, K. Bhaumik, S. Nedumbamana,
B. C. Rowan, B. S. Richards, and T. L. Hoeks, Applied Optics, 50(2), 163 (2011).
172
[90] L. Desmet, A. J. M. Ras, D. K. G. De Boer, and M. G. Debije, Optics Letters, 37(15), 3087 (2012).
[91] J. S. Batchelder, A. H. Zewail, and T. Cole, Applied Optics, 20(21), 3733 (1981). [92] A. M. Hermann, Solar Energy, 29(4), 323 (1982). [93] W. Viehmann and R. L. Frost, Nuclear Instruments and Methods, 167(3), 405 (1979). [94] R. Reisfeld, M. Eyal, V. Chernyak, and R. Zusman, Solar Energy Materials, 17(6), 439 (1988). [95] R. Reisfeld, Optical Materials, 32(9), 850 (2010). [96] W. R. L. Thomas, 1984, US4488047 (A) [97] A. J. Chatten, R. Bose, D. J. Farrell, Y. Xiao, N. L. A. Chan, L. Manna, A. Buchtemann, J.
Quilitz, M. G. Debije, and K. W. J. Barnham, in Nanotechnology for photovoltaics, L. Tsakalakos, Editor. 2010, Taylor and Francis: London. p. 323.
[98] R. Reisfeld, Journal of Fluorescence, 12(3), 317 (2002). [99] M. J. Kastelijn, C. W. M. Bastiaansen, and M. G. Debije, Optical Materials, 31(11), 1720
(2009). [100] S. T. Bailey, G. E. Lokey, M. S. Hanes, J. D. M. Shearer, J. B. Mclafferty, G. T. Beaumont,
T. T. Baseler, J. M. Layhue, D. R. Broussard, Y.-Z. Zhang, and B. P. Wittmershaus, Solar Energy Materials and Solar Cells, 91(1), 67 (2007).
[101] R. Bose, M. Gonzales, P. Jenkins, R. Walters, J. P. Morseman, M. W. Moss, C. E. Mclain,
P. Linsert, A. Buchtemann, A. J. Chatten, and K. W. H. Barnham, IEEE, 2010, 467 [102] O. Altan Bozdemir, S. Erbas-Cakmak, O. O. Ekiz, A. Dana, and E. U. Akkaya, Angewandte
Chemie International Edition, n/a (2011). [103] D. Bruhwiler, G. Calzaferri, T. Torres, J. H. Ramm, N. Gartmann, L.-Q. Dieu, I. Lopez-
Duarte, and M. V. Martinez-Diaz, Journal of Materials Chemistry, 198040 (2009). [104] G. Calzaferri, R. Meallet-Renault, D. Bruhwiller, R. Pansu, I. Dolamic, T. Dienel, P.
Adler, H. Li, and A. Kunzmann, Chemical Physics and Physical Chemistry, 12580 (2011). [105] S. Tsoi, D. J. Broer, C. W. Bastiaansen, and M. G. Debije, Optics Express, 18(S4), A536
(2010). [106] S. Tsoi, C. W. M. Bastiaansen, and M. G. Debije, 24th European Photovoltaic and Solar Energy
Conference, 2009, 377 [107] N. C. Giebink, G. P. Wiederrecht, and M. R. Wasielewski, Nature Photonics, 5694 (2011).
References
173
[108] M. A. El-Shahawy and A. F. Mansour, Journal of Materials Science: Materials in Electronics, 7(3), 171 (1996).
[109] B. A. Swartz, T. Cole, and A. H. Zewail, Optics Letters, 1(2), 73 (1977). [110] J. M. Drake, M. L. Lesiecki, J. Sansregret, and W. R. L. Thomas, Applied Optics, 21(16),
2945 (1982). [111] I. Baumberg, O. Berezin, A. Drabkin, B. Gorelik, L. Kogan, M. Voskobojnik, and M.
Zaidman, Polymer Degradation and Stability, 73(3), 403 (2001). [112] A. F. Mansour, H. M. A. Killa, S. Abd El-Wanees, and M. Y. El-Sayed, Polymer Testing,
24(4), 519 (2005). [113] R. Kinderman, L. H. Slooff, A. R. Burgers, N. J. Bakker, A. Buchtemann, R. Danz, and J.
A. M. V. Roosmalen, Journal of Solar Energy Engineering, 129(3), 277 (2007). [114] J. Yoon, L. Li, A. V. Semichaevsky, J. H. Ryu, H. T. Johnson, R. G. Nuzzo, and J. A.
Rogers, Nature Communications, 2343 (2011). [115] L. R. Wilson and B. S. Richards, Applied Optics, 48(2), 212 (2009). [116] M. G. Debije, P. P. C. Verbunt, B. C. Rowan, B. S. Richards, and T. L. Hoeks, Applied
Optics, 47(36), 6763 (2008). [117] G. Seybold and G. Wagenblast, Dyes and Pigments, 11(4), 303 (1989). [118] R. Reisfeld, D. Shamrakov, and C. Jorgensen, Solar Energy Materials and Solar Cells, 33(4),
417 (1994). [119] A. F. Mansour, Polymer Testing, 17(5), 333 (1998). [120] A. F. Mansour, M. G. El-Shaarawy, S. M. El-Bashir, M. K. El-Mansy, and M. Hammam,
Polymer Testing, 21(3), 277 (2002). [121] F. Castiglione, G. Lanzani, A. Mele, A. Monguzzi, M. Passarello, A. Ruggirello, F.
Scotognella, and V. Liveri, Journal of Materials Science, 46(19), 6402 (2011). [122] S. M. Reda, Solar Energy, 81(6), 755 (2007). [123] C. L. Mulder, L. Theogarajan, M. Currie, J. K. Mapel, M. A. Baldo, M. Vaughn, P. Willard,
B. D. Bruce, M. W. Moss, C. E. Mclain, and J. P. Morseman, Advanced Materials, 21(31), 3181 (2009).
[124] Y. Ren, M. Szablewski, and G. H. Cross, Applied Optics, 39(15), 2499 (2000).
174
[125] N. A. Bakr, A. F. Mansour, and M. Hammam, Journal of Applied Polymer Science, 74(14), 3316 (1999).
[126] A. Dubois, M. Canva, A. Brun, F. Chaput, and J.-P. Boilot, Applied Optics, 35(18), 3193
(1996). [127] R. O. Al-Kaysi, T. S. Ahn, A. M. Muller, and C. J. Bardeen, Physical Chemistry and. Chemical
Physics, 83453 (2006). [128] J. Mugnier, Y. Dordet, J. Pouget, and B. Valeur, Revue de Physique Appliquee, 2289 (1987). [129] A. F. Mansour, Polymer Testing, 23(3), 247 (2004). [130] V. Wittwer, W. Stahl, and A. Goetzberger, Solar Energy Materials, 11(3), 187 (1984). [131] O. I. Micic, H. M. Cheong, H. Fu, A. Zunger, J. R. Sprague, A. Mascarenhas, and A. J.
Nozik, The Journal of Physical Chemistry B, 101(25), 4904 (1997). [132] O. I. Micic, C. J. Curtis, K. M. Jones, J. R. Sprague, and A. J. Nozik, The Journal of Physical
Chemistry, 98(19), 4966 (1994). [133] D. J. Norris and M. G. Bawendi, Physical Review B, 53(24), 16338 LP (1996). [134] X. Wang, J. Zhang, A. Nazzal, and M. Xiao, Applied Physics Letters, 83(1), 162 (2003). [135] M. Lomascolo, A. Cretì, G. Leo, L. Vasanelli, and L. Manna, Applied Physics Letters, 82(3),
418 (2003). [136] C. De Mello Donega¡, S. G. Hickey, S. F. Wuister, D. Vanmaekelbergh, and A. Meijerink,
The Journal of Physical Chemistry B, 107(2), 489 (2002). [137] S. M. Reda, Acta Materialia, 56(2), 259 (2008). [138] A. Schüler, M. Python, M. V. Del Olmo, and E. De Chambrier, Solar Energy, 81(9), 1159
(2007). [139] Y. Nosaka, The Journal of Physical Chemistry, 95(13), 5054 (1991). [140] P. E. Lippens and M. Lannoo, Physical Review B, 39(15), 10935 LP (1989). [141] M. Kennedy, S. J. Mccormack, J. Doran, and B. Norton, Solar Energy, 83(7), 978 (2009). [142] S. J. Byrne, Y. Williams, A. Davies, S. A. Corr, A. Rakovich, Y. K. Gun'ko, Y. P.
Rakovich, J. F. Donegan, and Y. Volkov, Small, 31152 (2007). [143] F. Erogbogbo, K.-T. Yong, I. Roy, G. Xu, P. N. Prasad, and M. T. Swihart, ACS Nano,
2873 (2008).
References
175
[144] K. Fujioka, M. Hiruoka, K. Sato, N. Manabe, R. Miyasaka, S. Hanada, A. Hoshino, R. D. Tilley, Y. Manome, K. Hirakuri, and K. Yamamoto, Nanotechnology, 19415102 (2008).
[145] R. Reisfeld and Y. Kalisky, Chemical Physics Letters, 80(1), 178 (1981). [146] C. Lurin, C. Parent, G. Le Flem, and P. Hagenmuller, Journal of Physics and Chemistry of
Solids, 46(9), 1083 (1985). [147] C. Parent, C. Lurin, G. Le Flem, and P. Hagenmuller, Journal of Luminescence, 36(1), 49
(1986). [148] R. Reisfeld and S. Neuman, Nature, 274(5667), 144 (1978). [149] N. Neuroth and R. Haspel, Solar Energy Materials, 16(1-3), 235 (1987). [150] R. Reisfeld and Y. Kalisky, Nature, 283(5744), 281 (1980). [151] R. Reisfeld, Journal of the Less Common Metals, 93(2), 243 (1983). [152] R. Reisfeld, International Symposium on Engineering Ceremics, 1985, 375 [153] A. Buch, M. Ish-Shalom, R. Reisfeld, A. Kisilev, and E. Greenberg, International Symposium
on Engineering Ceremics, 1985, 383 [154] R. Reisfeld, Inorganica Chimica Acta, 95(2), 69 (1984). [155] O. Moudam, B. C. Rowan, M. Alamiry, P. Richardson, B. S. Richards, A. C. Jones, and N.
Robertson, Chemical Communications, (43), 6649 (2009). [156] G. Katsagounos, E. Stathatos, N. B. Arabatzis, A. D. Keramidas, and P. Lianos, Journal of
Luminescence, 131(8), 1776 (2011). [157] W. R. Holland and D. G. Hall, Optics Letters, 10(8), 414 (1985). [158] H. R. Wilson, Solar Energy Materials, 16(1-3), 223 (1987). [159] K. Aslan, R. Badugu, J. R. Lakowicz, and C. D. Geddes, Journal of Fluorescence, 15(2), 99
(2005). [160] R. Reisfeld, M. Pietraszkiewicz, T. Saraidarov, and V. Levchenko, Journal of Rare Earths,
27(4), 544 (2009). [161] T. K. Sau, A. L. Rogach, F. Jackel, T. A. Klar, and J. Feldmann, Advanced Materials, 22(16),
1805 (2010). [162] V. E. Ferry, J. N. Munday, and H. A. Atwater, Advanced Materials, 22(43), 4794 (2010).
176
[163] B. P. Rand, P. Peumans, and S. R. Forrest, Journal of Applied Physics, 96(12), 7519 (2004). [164] K. R. Catchpole and A. Polman, Optics Express, 16(26), 21793 (2008). [165] D. Duche, P. Torchio, L. Escoubas, F. Monestier, J.-J. Simon, F. Flory, and G. Mathian,
Solar Energy Materials and Solar Cells, 93(8), 1377 (2009). [166] A. J. Morfa, K. L. Rowlen, I. T. H. Reilly, M. J. Romero, and J. V. D. Lagemaat, Applied
Physics Letters, 92(1), 013504 (2008). [167] F.-C. Chen, J.-L. Wu, C.-L. Lee, Y. Hong, C.-H. Kuo, and M. H. Huang, Applied Physics
Letters, 95(1), 013305 (2009). [168] J. H. Lee, J. H. Park, J. S. Kim, D. Y. Lee, and K. Cho, Organic Electronics, 10(3), 416
(2009). [169] S.-Y. Wang, D.-A. Borca-Tasciuc, and D. A. Kaminski, Journal of Applied Physics, 109(7),
074910 (2011). [170] K. Heidler, Applied Optics, 20(5), 773 (1981). [171] J. Roncali and F. Garnier, Solar Cells, 13(2), 133 (1984). [172] A. A. Earp, G. B. Smith, P. D. Swift, and J. Franklin, Solar Energy, 76(6), 655 (2004). [173] M. G. Debije, J.-P. Teunissen, M. J. Kastelijn, P. P. C. Verbunt, and C. W. M. Bastiaansen,
Solar Energy Materials and Solar Cells, 93(8), 1345 (2009). [174] M. G. Debije and W. Dekkers, Journal of Renewable and Sustainable Energy, 4013103 (2012). [175] Y.-F. Xiao, C.-L. Zou, Y.-W. Hu, Y. Li, L. Xiao, F.-W. Sun, and Q. Gong, IEEE Journal of
Quantum Electronics, 47(9), 1171 (2011). [176] A. Royne, C. J. Dey, and D. R. Mills, Solar Energy Materials and Solar Cells, 86(4), 451
(2005). [177] J. S. C. Prentice, Solar Energy Materials and Solar Cells, 69(4), 303 (2001). [178] D. J. Farrell, W. G. J. H. M. V. Sark, S. T. Velthuijsen, and R. E. I. Schropp, Physica Status
Solidi (c), 7(3-4), 1045 (2010). [179] H. J. Hovel, R. T. Hodgson, and J. M. Woodall, Solar Energy Materials, 2(1), 19 (1979). [180] N. Yamada, L. Nguyen Anh, and T. Kambayashi, Solar Energy Materials and Solar Cells,
94(3), 413 (2010).
References
177
[181] W. Ding, R. Jia, D. Wu, C. Chen, H. Li, X. Liu, and T. Ye, Journal of Applied Physics, 109(5), 054312 (2011).
[182] J. Zhou, Y. Teng, S. Ye, X. Liu, and J. Qiu, Optical Materials, 33(2), 153 (2010). [183] D. Geyer, H.-D. Mohring, S. Schroder, J. Springer, S. Rexroth, and R. Schaffler, 20th
European Photovoltaic and Solar Energy Conference, 2005, 2514 [184] J. J. Schermer, G. J. Bauhuis, P. Mulder, E. J. Haverkamp, J. Van Deelen, A. T. J. Van
Niftrik, and P. K. Larsen, Thin Solid Films, 511-512645 (2006). [185] A. W. Hains, Z. Liang, M. A. Woodhouse, and B. A. Gregg, Chemical Reviews, 110(11),
6689 (2010). [186] B. Fisher and J. Biddle, Solar Energy Materials and Solar Cells, 95(7), 1741 (2011). [187] R. Koeppe, Applied Physics Letters, 90181126 (2007). [188] S. Chattopadhyay, Y. F. Huang, Y. J. Jen, A. Ganguly, K. H. Chen, and L. C. Chen,
Materials Science and Engineering: R: Reports, 69(1-3), 1 (2010). [189] U. Schulz, Applied Optics, 45(7), 1608 (2006). [190] W. R. L. Thomas, J. M. Drake, and M. L. Lesiecki, Applied Optics, 22(21), 3440 (1983). [191] J. Sansregret, J. M. Drake, W. R. L. Thomas, and M. L. Lesiecki, Applied Optics, 22(4), 573
(1983). [192] G. W. Chantry, J. W. Fleming, P. M. Smith, M. Cudby, and H. A. Willis, Chemical Physics
Letters, 10(4), 473 (1971). [193] H. Ma, A. K.-Y. Jen, and L. R. Dalton, Advanced Materials, 14(19), 1339 (2002). [194] J. Zubia and J. Arrue, Optical Fiber Technology, 7(2), 101 (2001). [195] J. R. White, Comptes Rendus Chimie, 91396 (2006). [196] P. Polischuk, IEEE Communications Magazine, 44140 (2006). [197] M. Buffa, S. Carturan, M. G. Debije, P. P. C. Verbunt, A. Quaranta, G. Maggioni, and M.
Tonezzer, E-MRS Fall 2011, 2011, [198] M. Buffa, S. Carturan, M. G. Debije, A. Quaranta, and G. Maggioni, Solar Energy Materials
and Solar Cells, 103(0), 114 (2012).
178
[199] J. C. Goldschmidt, M. Peters, L. Pronneke, L. Steidl, R. Zentel, B. Blasi, A. Gombert, S. Glunz, G. Willeke, and U. Rau, Physica Status Solidi (a), 205(12), 2811 (2008).
[200] M. Sidrach De Cardona, M. Carrascosa, F. Meseguer, F. Cusso, and F. Jaque, Solar Cells,
15(3), 225 (1985). [201] A. Fantoni, M. Vierira, J. Cruz, R. Schwarz, and R. Martins, Journal of Physics D: Applied
Physics, 29(12), 3154 (1996). [202] A. Goetzberger and O. Schirmer, Applied Physics, 1953 (1979). [203] J. Loh, Eugene and D. J. Scalapino, Applied Optics, 25(12), 1901 (1986). [204] A. H. Zewail and J. S. Batchelder, 1980, US 4227939 [205] K. R. Mcintosh, N. Yamada, and B. S. Richards, Applied Physics B: Lasers and Optics, 88(2),
285 (2007). [206] C. Auronet, H. Blumenfeld, M. Bourdinaud, J. Calvet, J.-C. Cavan, J. Meyer, and J.-C.
Thevenin, 1989, US 4812013 [207] A. F. Abouraddy and E. Banaei, TechConnect World, 2011, [208] M. V. Gurp, T. V. Heijnsbergen, G. V. Ginkel, and Y. K. Levine, The Journal of Chemical
Physics, 90(8), 4103 (1989). [209] M. V. Gurp and Y. K. Levine, The Journal of Chemical Physics, 90(8), 4095 (1989). [210] C. Sanchez, B. Villacampa, R. Cases, R. Alcala, C. Martinez, L. Oriol, and M. Pinol, Journal
of Applied Physics, 87(1), 274 (2000). [211] R. L. Van Ewyk, I. O'connor, A. Mosley, A. Cuddy, C. Hilsum, C. Blackburn, J. Griffiths,
and F. Jones, Displays, 7(4), 155 (1986). [212] P. G. D. Gennes and J. Prost, 2nd ed. 1974, Oxford, United Kingdom: Oxford University
Press. 597. [213] C. W. Oseen, Transactions of the Faraday Society, 29(140), (1933). [214] R. Barberi, F. Ciuchi, G. E. Durand, M. Iovane, D. Sikharulidze, A. M. Sonnet, and E. G.
Virga, European Physics Journal E, 13(1), 61 (2004). [215] M. I. Boamfa, S. V. Lazarenko, E. C. M. Vermolen, A. Kirilyuk, and T. Rasing, Advanced
Materials, 17(5), 610 (2005). [216] M. G. Forest, X. Zheng, R. Zhou, Q. Wang, and R. Lipton, Advanced Functional Materials,
15(12), 2029 (2005).
References
179
[217] M. Nishikawa, Polymers for Advanced Technologies, 11(8-12), 404 (2000). [218] W. M. Gibbons, P. J. Shannon, S.-T. Sun, and B. J. Swetlin, Nature, 351(6321), 49 (1991). [219] P. Rathi and T. Kyu, Physical Review E, 79(3), 031802 (2009). [220] R. A. M. Hikmet, J. Lub, and D. J. Broer, Advanced Materials, 3(7-8), 392 (1991). [221] C. Zannoni, Order parameters and orientational distributions in liquid crystals, in Polarized
spectroscopy of ordered systems, B. Samori and E. Thulstrup, Editors. 1988, Kluwer Academic Publishers: Dordrecht, the Netherlands. p. 578.
[222] D. Bauman, E. Chrzumnicka, E. Mykowska, M. Szybowicz, and N. Grzelczak, Journal of
Molecular Structure, 744-747307 (2005). [223] B. S. Richards and K. R. Mcintosh, Progress in Photovoltaics, 1527 (2007). [224] Personal Communication: B. C. Rowan and B. S. Richards, 2008 [225] C. L. Mulder, P. D. Reusswig, A. M. Velázquez, H. Kim, C. Rotschild, and M. A. Baldo,
Optics Express, 18(S1), A79 (2010). [226] K. E. Vaughn, M. Sousa, D. Kang, and C. Rosenblatt, Applied Physics Letters, 90(19),
194102 (2007). [227] B. S. Richards, A. Shalav, and R. P. Corkish., 19th European Photovoltaic Solar Energy
Conference, 2004, [228] D. J. Broer, G. N. Mol, J. A. M. M. V. Haaren, and J. Lub, Advanced Materials, 11(7), 573
(1999). [229] D. W. Berreman, Journal of the Optical Society of America, 62(4), 502 (1972). [230] H. Wöhler, M. Fritsch, G. Haas, and D. A. Mlynski, Journal of the Optical Society of America
A, 8(3), 536 (1991). [231] D. J. Broer, J. Lub, and G. N. Mol, Nature, 378(6556), 467 (1995). [232] R. G. Hopkins, Journal of the Optical Society of America, 44(6), 455 (1954). [233] D. K. G. D. Boer, C. R. Ronda, D. J. Broer, and H. J. Cornelissen, 2011, WO2011064691 [234] P. S. Friedman and C. R. Parent. 1984, US Department of Energy
180
[235] J. Bomm, A. Büchtemann, A. J. Chatten, R. Bose, D. J. Farrell, N. L. A. Chan, Y. Xiao, L. H. Slooff, T. Meyer, A. Meyer, W. G. J. H. M. Van Sark, and R. Koole, Solar Energy Materials and Solar Cells, 95(8), 2087 (2011).
[236] K. Y. Law, Chemical Reviews, 93(1), 449 (1993). [237] Y. Nagao, Progress in Organic Coatings, 31(1-2), 43 (1997). [238] W. S. Shin, H.-H. Jeong, M.-K. Kim, S.-H. Jin, M.-R. Kim, J.-K. Lee, J. W. Lee, and Y.-S.
Gal, Journal of Materials Chemistry, 16(4), 384 (2006). [239] S. H. Oh, B. G. Kim, S. J. Yun, M. Maheswara, K. Kim, and J. Y. Do, Dyes and Pigments,
85(1-2), 37 (2010). [240] J. E. Pickett, Material Testing: Product and Technology News, 35(73), 2 (2005).
181
Samenvatting Slinkende olie voorraden hebben er voor gezorgd dat er meer interesse is bij overheden naar
het gebruik van hernieuwde energiebronnen. Vooral in gebouwen, die in ontwikkelde landen
verantwoordelijk zijn voor 40% van al het energiegebruik, is er veel belangstelling voor het
implementeren van energie opwekkende en besparende systemen. Deze systemen zijn bij
voorkeur aanpasbaar aan de wensen van de architect, terwijl de kosten van de opgewekte
energie laag is (~€0.06/kWh). Een energiebron die in de buurt van elke gebouw aanwezig is, is
de zon. De toepasbaarheid van standaard zonnecellen in gebouwen is lastig gebleken, omdat de
kosten van de opgewekte energie te hoog zijn en de flexibiliteit naar de behoeften van de
architect beperkt is. Een alternatief dat geacht wordt de kosten van de opgewekte energie te
verlagen en de toepasbaarheid te verhogen, in vergelijk met standaard zonnecellen, is de
fluorescente licht geleider (FLG).
Een FLG bestaat uit een plastic of glazen transparante plaat, die dient als licht geleider,
waarin fluorescente kleurstof moleculen zijn aangebracht. Deze fluorescerende moleculen
absorberen het zonlicht en zenden dit uit met een langere golflengte. Een deel van dit
uitgezonden licht wordt in de lichtgeleider gevangen door totale interne reflectie en het licht
wordt hierdoor geconcentreerd aan de zijkanten van de plaat waar een zonnecel geplaatst wordt
die het licht omzet in elektriciteit. De efficiëntie van FLGs is tot op heden nog niet hoog
genoeg door verschillende verliesmechanismen. Een van deze verliesmechanismen is het licht
dat de lichtgeleider aan zowel de boven- als de onderzijde verlaat. Dit uitgekoppelde licht is een
combinatie van een gelimiteerde inkoppel-efficiëntie van de door de fluorescente kleurstof
uitgezonden fotonen en de herabsorptie van gevangen fotonen die opnieuw uitgezonden
worden, ook wel foton-recycling genaamd, maar gedeeltelijk onder een verkeerd hoek
In dit werk wordt een eenvoudig model gepresenteerd dat de ruimtelijke distributie van
uitgezonden fotonen door een collectie van dichroïtische kleurstof moleculen in een isotrope of
vloeibaar kristallijne matrix beschrijft. Het model is gevalideerd met experimenten, waarin de
mate van moleculaire orde in planaire (parallel aan de oppervlakte van de lichtgeleider) uitlijning
is veranderd. In FLGs met een planaire uitlijning van de kleurstof moleculen verlaat er tot 60%
meer licht de 2 gewenste zijkanten parallel aan de richting van kleurstof moleculen in
vergelijking met de 2 zijkanten loodrecht op de richting van de kleurstof moleculen. Deze
resultaten komen overeen met het model.
Voor FLGs met een isotrope verdeling van dichroïtische kleurstof moleculen is de
inkoppel-efficiëntie van de uitgezonden fotonen berekend. Deze is 74,3% in polycarbonaat. Dit
resulteert in een verlies van 25,7% van alle uitgezonden fotonen door de oppervlakten van de
lichtgeleider. Experimenten tonen aan dat het verlies van uitgezonden fotonen ~50% is, voor
een FLG met als kleurstof BASF Lumogen F Red 305, de standaard in FLGs. Het verschil
182
tussen de resultaten uit het model en de experimenten komt doordat een deel van de
“gevangen” fotonen nogmaals worden geabsorbeerd en opnieuw worden uitgezonden.
Het uitlijnen van de dichroïtische moleculen kan de hoeveelheid verloren fotonen door de
oppervlakten van de lichtgeleider reduceren. Zowel uit voorspellingen van het model als uit
experimenten blijkt dat het oppervlakte verlies gereduceerd kan worden tot minder dan 10% als
de dichroïtische moleculen homeotroop (loodrecht ten opzicht van de oppervlakten van de
lichtgeleider) worden uitgelijnd. Echter zorgt deze uitlijning ervoor dat de hoeveelheid zonlicht
die geabsorbeerd kan worden, wordt gereduceerd. Planair uitgelijnde dichroïtische moleculen
verhogen de kans op absorptie van het zonlicht, echter de inkoppel-efficiëntie van de
uitgezonden fotonen in de lichtgeleider wordt verlaagd. Het kantelen van de dichroïtische
moleculen ten opzichte van de lichtgeleider kan de voordelen van planaire (vergrootte kans op
absorptie) en homeotrope (hogere inkoppel-efficiëntie van de uitgezonden fotonen) kunnen
combineren. Voorspellingen over de optimale kantelhoek is moeilijk gebleken. Het
voorgestelde model dient geïmplementeerd worden in simulatie software om een voorspelling
te doen over de optimale kantelhoek.
Een tweede methode om de oppervlakte verliezen te reduceren is het toepassen van
golflengte selectieve spiegels die inkomend zonlicht doorlaten en uitgezonden licht reflecteren.
Theoretische benaderingen tonen aan dat ongeveer 90% van alle fotonen die verloren gaan
door de oppervlakten van de lichtgeleider terug gereflecteerd kunnen worden als de FLG BASF
Lumogen Red 305 als kleurstof heeft, voor een reflector met een 400 nm bandbreedte. Dit kan
leiden tot een relatieve verhoging van 66% van de efficiëntie van de FLG, terwijl een 175nm
brede reflector een relatieve verhoging van 45% geeft.. Experimenteel blijkt dat de efficiëntie
van een FLG met een piekabsorptie van ~1 maar met 5% wordt verhoogd als er een reflector
wordt toegepast met een bandbreedte van 175 nm. Wanneer de hoeveelheid foton-recycling
wordt verminderd kan de efficiëntie van de FLG worden verhoogd met ~20%, waaruit blijkt
dat de efficiëntie van golflengte selectieve spiegels wordt gelimiteerd door foton-recycling.
De golflengte selectieve spiegels hebben ook een hoek afhankelijkheid: de reflectieband
schuift naar kortere golflengtes als de hoek van inval wordt vergroot. Hierdoor wordt een deel
van het zonlicht, dat geabsorbeerd kan worden door de kleurstof moleculen gereflecteerd. Dit
reduceert de effectiviteit van de reflector and dit kan in potentie leiden tot een reductie van de
FLG efficiëntie. Simulaties tonen aan dat speciale dispersie cholesterische reflectoren (SDCs)
een verminderde hoekafhankelijkheid hebben. Theoretische berekeningen tonen aan dat het
gebruik van deze SDCs de efficientie van de FLG verhogen voor hoeken van inval tot 70°.
Zowel de uitlijning van de dichroïtische moleculen als de toepassing van organische
golflengte selectieve reflectoren reduceren de hoeveelheid licht die verloren gaan aan de
oppervlakten van de FLG. Het effect van beide methoden is experimenteel een stuk kleiner dan
voorspeld door middel van berekeningen en simulaties. Om de efficiëntie van de FLG te
verhogen zijn fluorescente kleurstoffen nodig die geen overlap hebben tussen de absorptie en
Samenvatting
183
de emissie band. Met andere woorden de hoeveelheid foton-recycling moet gereduceerd
worden. Ook zal de hoeveelheid zonlicht die door de kleurstof moleculen geabsorbeerd word,
vergroot moeten worden om de FLG tot een geschikt alternatief voor de generatie van
electriciteit te maken, aangezien de gelimiteerde absorptie van het zonlicht het grootste verlies
in FLGs is.
185
Acknowledgements Every PhD-student experiences some peaks and valleys during their project. In particular, the
valleys can be very difficult, and the help offered by individuals is necessary to proceed.
Finishing my thesis, I want to thank them for their help keeping my project on track. First of all
I would like to thank my promotor prof. Broer. Dick, thank you for allowing me to perform my
PhD-project within your group. I would also like to thank you for the inspiration and the
advice you gave me, especially in the last part of my project. When I left your office I was
always full of confidence and new ideas. Secondly, I would like to thank Kees Bastiaansen.
Kees, I really appreciate the discussions we had, even though they progressed sometimes
somewhat laboriously. Your critical view on my project kept me focusing on the main goal of
the project and I am honored to have you as my second co-promotor. I owe the most to my
first co-promoter Michael Debije. I don’t know where to start thanking you. Until the moment
you asked me to do this project, I had never considered doing a PhD. But I knew that with you
as supervisor I would be able to finish such a project. I could always drop by in your office if I
had a problem or a question. Without you these past four years would have been much more
difficult. Besides the scientific discussions we had I also enjoyed the non-scientific chats we had
on boardgames, music or any other topic. Thanks for everything. I would also like to thank
your family for opening your house to me, I especially enjoyed the boardgame nights at your
place. Half way through my project our group was extended with a very inspiring scientist,
Albert Schenning. Even though most of the topics we discussed are not in this thesis, I really
enjoyed them and I always left our office inspired and with a huge smile on my face.
I would also like to thank the core-committee for my thesis defense, prof. Janssen (TU/e),
prof. Richards (HWU, Edinburgh), prof. Urbach (TUD) and dr. De Mello-Donega (UU).
Thank you all very much in participating in the core-committee and for your comments and
suggestions on this thesis.
My project was funded by Stichting voor Technische Wetenschappen (STW) and I would
like to them, especially Monique Wiegel for the smooth arrangement of the VIDI-meetings. A
VIDI project has an external committee and I would like to thank the members for their
contribution in the project. Lenneke Slooff (ECN) and Casper van Oosten, thank you for your
contributions in the discussions we had. Besides contributions to the discussions, I would like
to thank Theo Hoeks (Sabic-IP) for the materials he provided. The fourth and last external
committee member of the VIDI-project turned out to be much more important to my project,
Dick de Boer (Philips Research). Your involvement in my work went much further than being a
committee member. You made it possible for me to take the theoretical approaches in this
thesis to another level. You taught me the more fundamental side of science and I would have
186
never guessed I would enjoy it as much as I do now. You also took a good look into my thesis
and without you I probably wouldn’t have finished this thesis by now. I am honored to have
you on my defense committee as an advisor. The meetings we had together with the people at
Philips Research I liked a lot, especially the discussions we had with Chi-Wen and Merijn.
More help with the theoretical work presented in thesis I got from Carlos Sanchez from
the University of Zaragoza. When I asked you to help me calculating the emission profiles from
tilted dyes I would have never guessed I would consume so much of your time. Thank you for
that.
I had the pleasure to work with several students during my time as a PhD-student. Duygu
and Toon, I enjoyed supervising your projects and even though the results of these project did
not make it into this thesis, your work helped us understanding some basic principles. I also
supervised a group of OGO-students. Sabrina, Stéphanie (who became a roommate when she
started her master-project within our group), Thijs and Yoran, without your help the
production of layered narrowband cholesterics would not have been so easy.
I had a lot of very enthusiastic, inspiring and helpful (PhD-)students and post-docs who
dared to share an office with me. Shufen, you were my first roommate and a we worked on the
same project. I really enjoyed our scientific discussions, the chit-chats we had and without your
help chapter 7 of this thesis would have been a lot shorter. I shared my office further with
Debarshi, Derya, Hilal, Judith, Julien, Koen, Nicole, Peter, Patrick, Stéphanie and Yang. You
were all very pleasant roommates and I hope I haven’t kept you from your work too much with
my constant talking. I also had a roommate, which was my “boss” before. Robert, it was a
delight to have you in “my” office. I already had a great time when I worked as a student-
assistant for Validus, the company you had within our group together with Thijs and Nico, but
the time we spent together during the six months you were a fellow PhD-student were great. I
loved the chats we had when everybody else had already left.
In addition to my roommates I had the pleasure to work with other great PhD-students,
post-docs and staff members. Especially in STO 0.24 I met some people that I consider as
friends. Ties, it took us a while to start chatting with each other, but we made up for that in the
second part of our time as PhD-students. I really enjoyed our coffee-breaks together, even
though we speak a different language (I speak in Chemistry and you in Physics). I learned a lot
from you. I also enjoyed to spend a lot of time with your office-mates, My and Natalia (my
dearest enemy). From boardgaming and having a beer to real science, all of this was a delight.
I would also like to thank the other people who were in SFD, but also some members
from SKT and PTG: Amol, An, Anne, Antonio, Bob (I will cheer for the Welsh rugby team),
C.K., Claudia, Danqing, Elena, Felix, Han, Helena, Huub, Iren, Ivelina, Jelle, Jeremy, Johan,
Joost, Jurgen, Katherine (the girl who scared me), Ko, Laurens, Luc, Maria, Maud, Maurizio,
Acknowledgements
187
Mian, Nick, Pauline, Peter K., Piming, Pit, Shabnam, Stefan, Sun, Tamara, Teun, Tom, Yogesh,
Youseli, Weizhen and all that I might have forgotten. We also had some very nice secretaries,
Elly, Ineke en Marjolijn, bedankt voor alles dat jullie gedaan hebben.
I would also thank one person from my professional scientific life in the past, Jens Thies
from DSM Research. When I worked as assistant scientist at DSM after I finished my Bachelor
(HBO) I had the pleasure to work with Jens. You asked me several times why I would not try
to get a Master’s. At that moment I had no interest, but your words have been in my head since
then and a couple of years later, I still remembered them and I decided to start my Master’s
study. All this lead in the end to this thesis, and without you I probably would have never
considered to start my Master’s study.
Sometimes it was necessary for me to step away from my project. I would have never done
that without some important people, all from Limburg. I would like to thank them in the
language of Limburg.
Ut laeve van eine promovendus geit neet altied euver rozen. Es ut project neet lupt zoals
ut zou motte laope, is ut good om alles aeve aan de kantj te zitten en det waor mich neet gelökt
zonger ein paar heel biezonjere miense. Es eerste wil ich mien vrunj bedanke, sommige kin ich
al bienao mien ganse laeve en angere hub ich pas later leere kinne. Martin, Bernd, Wendy (en
Joan), Patty, Bianca, Tim, Melissa en Paul bedank veur alles. Boete mien vrunj wil ich auch gaer
mien femilie bedanken veur alle sjteun. Veural mien grootelders wil ich bedanke veur alles. Ich
vinj ut jaomer det driej van heur dit neet meer mit kinne maake, mer zonger uch had ich dit
nooit kinne doon. Ein anger femilielid det ich extra wil bedanke is Patrick. Al sinds ich vreuger
nao uch in de Julianasjtraot kwaam busse meer ein broor veur mich gewaes den eine naef. De
sjpelaovende en de keren det ver zeen gaon sjtappe waren vaak precies waat ich neudig had.
Bedank veur alles.
Es allerletste wil ich de twee belangriekste miense in mien laeve bedanke. Mam en Pap,
bedank veur alles, neet allein veur de letste veer jaor, mer veur alles det ger veur mich gedaon
hub mien hele laeve. Zonger uch had ich nooit kinne bereike waat ich bereikt hub.
Paul
189
Curriculum Vitae
Paul Verbunt was born on the 30th of June 1981 in Roermond, the
Netherlands. After graduating secondary school (“HAVO”) at
Bischoppelijk College Broekhin (Roermond) in 1998, he studied
(Polymer) Chemistry at the Fontys Hogeschool TNW in Eindhoven.
After completing an internship at Dex Plastomers (Geleen, the
Netherlands) on chain-end unsaturations in polyethylene, he complete
his graduation project in 2002 at DSM Research (Geleen, the
Netherlands) on acid scavengers to improve the hydrolitic stability of
methacrylate resins.
After working almost three years in industry as a research assistant and warehouse clerk, he
started his Masters education in 2005 in the faculty Chemical engineering and Chemistry at the
Eindhoven University of Technology (TU/e). In 2008 he received his Master’s degree for his
thesis ‘Surface losses in luminescent solar concentrators’ under guidance of dr. M.G. Debije in
the group of prof. D.J. Broer.
Directly after obtaining his Master’s degree, Paul started a Ph.D. project at the Chemical
Engineering and Chemistry department of the Eindhoven University of Technology under
guidance of prof. D.J. Broer, dr. M.G. Debije and dr. C.W.M. Bastiaansen. The results of this
project are described in this thesis
190
191
List of symbols
General introduction
E Energy of a photon
h Planck’s constant
Frequency of a photon
c Speed of light
Wavelength of light
Thermodynamics of concentrators
,in outA A Area of the incoming/outgoing light in a concentrator
,in outU U Etendue of the incoming/outgoing light in a concentrator
,in out Solid angle of the incoming/outgoing light in a concentrator
max max,, , diffC C C
Concentration/maximum concentration/maximum concentration of
diffuse light by a concentrator based thermodynamics
, ,in out concn n n Refractive index of the medium of the incoming light/the medium of the
outgoing light/concentrator
Introduction to LSC/LSC in general
c Critical angle
opt Optical efficiency of the LSC
R Fresnel reflection from the LSC waveguides
trap Probability that an emitted photon is in waveguide mode
abs Fraction of the incident sunlight that is absorbed by the luminophores in
the LSC
PLQY Photoluminescent quantum yield of the luminophore
Stokes Energy efficiency of the dye molecule related to the Stokes shift
host Transport efficiency of photons through the clear host/waveguide
TIR Total internal reflection efficiency related to the smoothness of the
waveguide surfaces
self Transport efficiency of the photons through the LSC waveguide related to
photon recycling
192
Luminophore/dye parameters
af Overlap factor between the absorption and the emission spectrum of the
luminophore molecule
1 2,e e Energy of the absorbed and emitted photons
0T Ambient temperature
,PLQY FRET Energy transfer quantum yield
r Distance between the donor and the acceptor for FRET
0R Förster’s distance
abs Wavelength of the absorbed light
m Mass of an electron
L Chain length of the π-conjugated plane
N Number of electrons
Model for emission profiles
, The polar and azimuthal angle of the emitted light
,i fe e
The polarization of the incident/emitted light
Transition dipole vector
,I
Intensity of the emitted light./Emission profile
, The polar and azimuthal angle of the transition dipole
k Vector that describes the direction of the emitted photons
Angle of the polarization of the emitted light with respect to the arbitrary
polarizations of the light
,1 ,2,f fe e
Arbitrary polarizations of the emitted light, orthogonal to the emitted light
vector and each other
Ensemble of dye molecules
f
Distribution function of the transition dipoles within the ensemble.
Aligned dye molecules by a liquid crystalline host
Angle between the molecular axis and the director of a liquid crystalline
material
n Director of the liquid crystalline material
2nS The 2nth order parameter of the liquid crystalline material
List of symbols
193
,o en n The ordinary and extraordinary refractive index
n Birefringence
2,optS
Optical order parameter/macroscopic order parameter
Angle between the molecular and optical axes of a dye molecule
2nP The 2nth Legendre polynomial
2, flS
Fluorescence order parameter
,F F
Amount of emitted light with linear polarization parallel/perpendicular to
the alignment direction
eER Output ratio (see list of definitions for the definition)
,A A
Absorption of incident light with linear polarization parallel/perpendicular
to the alignment direction
sl Fraction of the emitted photons emitted in the surface loss mode.
Tilt angle of the guest-host system with respect to the LSC waveguide
General introduction to Bragg reflectors
m Integer denoting the order of interference
n Average refractive index
d Periodicity
Bandwidth of the reflection band of the Bragg reflector
0 Central reflection wavelength of the wavelength selective reflector at normal
incidence
1 2,n n Refractive indices of the materials in a Bragg reflector
Cholesteric wavelength selective reflectors p Pitch of the helix of a cholesteric liquid crystalline material
HTP Helical twisting power of a chiral dopant
c
Concentration of the chiral dopant
Central reflection wavelength of the reflector at angle of incidence
refl Efficiency of reflection of luminophore emitted light of the wavelength
selective reflector
sE
Emission spectrum of the luminophore
pE
Emission profile of surface emitted photons
194
,R
Angular dependent reflection spectrum of the cholesteric reflector
EAcholf
Fraction of the absorbable incident light transmitted by the cholesteric
reflector
,I
Solar spectrum
,A
Angular dependent absorption spectrum of the luminophore containing
LSC
,maxLSC Maximal fractional increase in LSC efficiency after application of a
cholesteric reflector
,edge choln
Number of photons leaving the edges of the LSC after application of a
cholesteric reflector
,edge baren
Number of photons leaving the edges of the LSC without the application of
a cholesteric reflector
, ,edge SL choln
Number of photons leaving the edges of the LSC formerly lost through the
LSC surface after application of a cholesteric reflector
0 ,A A Absorption of the luminophore containing LSC at normal incidence/at
angle
0,n n Refractive index of a material at a specific wavelength/ null refractive index
1 2,B B Dispersion coefficients
195
List of definitions Cholesteric Chiral Nematic
Dye ensemble Dye molecules acting as guests in a guest-host system.
Homeotropic
alignment
Alignment of a liquid crystalline material, where the director is
perpendicular to the substrate
Initial emission Emission from a dye ensemble after the initial absorption of the
incident light. The re-emission of light after the re-absorption of
photons in waveguide mode is not taken into account
Null refractive index The refractive index of a material at infinite wavelength as used in
Cauchy’s equation for dispersion
Peak absorbance Absorbance of the peak absorption band of a luminophore
Optical Efficiency The energy in mW leaving the 4 edges of the LSC waveguide divided
by the energy of the sunlight in mW incident to the LSC
Optical order
parameter
Order parameter of the optical axis of a dichroic dye in a host
(usually a liquid crystalline host)
Output ratio The ratio in edge emission between the edge of an LSC parallel and
perpendicular to the alignment direction in an LSC with planarly
aligned dye molecules
Photon recycling A sequential event of re-absorption and re-emission of photons
emitted by the luminophore. This leads to a redistribution of the
emitted photons
Planar alignment Alignment of a liquid crystalline material, where the director is
parallel to the substrate
196
Quenching The reduction of emission due to the formation of non-emitting
clusters of luminophores
Special dispersion
cholesteric
The cholesteric phase of a liquid crystalline material with a special
dispersion. In these liquid crystalline materials have an ordinary
refractive index which changes more with changing wavelength than
the extraordinary refractive index. At a certain wavelength the
refractive indices are equal
Surface loss The loss of emitted photons in LSCs lost through the top and
bottom surfaces of the waveguide
Tilted alignment Alignment of a liquid crystalline material, where the director is tilted
to the substrate
Top/Bottom side of
the waveguide
The surface facing towards/away from the incident light
197
List of publications Publications related to this work
P.P.C. Verbunt, S. Tsoi, M.G. Debije, D.J. Broer, C.W.M. Bastiaansen, C.-W. Lin, D.K.G. de
Boer, “Increased efficiency of luminescent solar concentrators after application of organic
wavelength selective mirrors”, Optics Express, 20 (S5), A655-668, 2012
D.K.G. de Boer, D.J. Broer, M.G. Debije, W. Keur, A. Meijerink, C.R. Ronda and P.P.C.
Verbunt, “Progress in phosphors and filters for luminescent solar concentrators”, Optics
Express, 20 (S3), A395-A405, 2012.
M.G. Debije and P.P.C Verbunt, “Thirty years of luminescent solar concentrator research: solar
energy for the built environment”, Advanced Energy Materials, 2, 12-35, 2012.
M.G. Debije, P.P. C. Verbunt, P.J. Nadkarni, S. Velate, K. Bhaumik, S. Nedumbamana, B.C.
Rowan, B.S. Richards and T.L. Hoeks, “Promising fluorescent dye for solar energy conversion
based on a perylene perinone”, Applied Optics, 50 (2), 163-169, 2011.
Dick K.G. de Boer, C.-W. Lin, M.P. Giesbers, H.J. Cornelissen, M.G. Debije, P.P.C. Verbunt
and D.J. Broer, “Polarization-independent filters for luminescent solar concentrators”, Applied
Physics Letters, 98, 021111, 2011.
M.G. Debije, M.P. Van, P.P.C. Verbunt, M.J. Kastelijn, R.H.L. van der Blom, D.J. Broer and
C.W.M. Bastiaansen, “Effect on the output of a luminescent solar concentrator on application
of organic wavelength-selective mirrors”, Applied Optics, 49 (4), 745-751, 2010.
M.G. Debije, J.P. Teunissen, M.J. Kastelijn, P.P.C. Verbunt, C.W.M. Bastiaansen, “The effect
of a scattering layer on the edge output of a luminescent solar concentrator”, Solar Energy
Materials and Solar Cells, 93, 1345-1350, 2009.
P.P.C. Verbunt, A. Kaiser, K. Hermans, C.W.M. Bastiaansen, D.J. Broer and M.G. Debije,
“Controlling light emission in luminescent solar concentrators through use of dye molecules
aligned in a planar matter by liquid crystals”, Advanced Functional Materials, 19, 2714-2719,
2009.
198
M.G. Debije, P.P.C. Verbunt, B.C. Rowan, B.S. Richards and T.L. Hoeks, “Measured surface
loss from luminescent solar concentrator waveguides”, Applied Optics, 47 (36), 6763- 6768,
2008.
Conference proceedings
P.P.C. Verbunt, M.G. Debije, D.J. Broer, C.W.M. Bastiaansen and D.K.G. de Boer, "Organic
wavelength selective mirrors for luminescent solar concentrators", Proc. SPIE 8438, 2012,
843805
M.G. Debije and P.P.C. Verbunt, “Using liquid crystals to improve the performance of
luminescent solar concentrators”, NSTI-Nanotech 2011, Vol. 1, 584-587, 2011.
P.P.C. Verbunt and M.G. Debije, Progress in luminescent solar concentrator research: Solar
energy for the built environment”, in proceedings of the World Renewable Energy Congress,
2011.
M.G. Debije, M.P. Van, P.P. C. Verbunt, D.J. Broer and C.W. M. Bastiaansen, “The effect of
an organic selectively-reflecting mirror on the performance of a luminescent solar
concentrator”, in proceedings of the 24th European Photovoltaic Solar Energy Conference,
WIP Munchen, 2009, 373-376.
P.P.C. Verbunt, C.W.M. Bastiaansen, D.J. Broer and M.G. Debije, “The effect of dyes aligned
by liquid crystals on luminescent solar concentrator performance”, in proceedings of the 24th
European Photovoltaic Solar Energy Conference, WIP Munchen, 2009, 381-384.
Conference contributions
Oral Presentation at the 2012 SPIE Photonics Europe, Brussels, Belgium entitled: “Organic
wavelength selective mirrors for luminescent solar concentrators”.
Oral Presentation at the 2012 Dutch Polymer Days, Lunteren, the Netherlands entitled:
“Luminescent solar concentrators: Another piece in the energy puzzle”.
Oral Presentation at the 2011 Dutch Polymer Days, Veldhoven, the Netherlands entitled:
“Building integrated light harvesting polymers”. Best workshop lecture award.
List of publications
199
Poster at the 2010 Liquid crystals for Photonics workshop, Elche, Spain entitled: “Liquid
crystals in luminescent solar concentrators”.
Oral Presentation at the 2010 Rolduc Polymer Meeting, Kerkrade, the Netherlands entitled:
“Harvesting light with polymers”.
Poster at the 2010 STW Jaarcongres, Nieuwegein, the Netherlands entitled: “Luminescent solar
concentrators: A solar revolution?”.
Poster at the 24th European Photovoltaic Solar Energy Conference, 2009, Hamburg, Germany
entitled: “The effect of dyes aligned by liquid crystals on luminescent solar concentrator
performance”.
Poster at the 2009 Dutch Polymer Days, Lunteren, the Netherlands entitled: “Improved single
edge emission of a luminescent solar concentrator by planar dye alignment”.
200
“The time has gone, the song is over.”
Pink Floyd
Printed at: Ipskamp Drukkers
Cover design by: Paul Verbunt
Front cover: A highly contrasted picture of a stack of luminescent solar concentrators
illuminated from above. Back cover: An illuminated red luminescent solar concentrator.