Metal organic framework loaded electrospun poly-𝜀-caprolactone scaffolds as novel catalytic system
Luca Verhoeven
Promotors: Prof. Dr. Dubruel Prof. Dr. Van Der Voort
Guide: Dr. Karen Leus
A dissertation submitted to Ghent University in partial fulfilment of the requirements for the degree of Master of Science in Chemistry
Academic year 2016-2017
i
ACKNOWLEDGEMENT
First, I would like to express my deepest gratitude to my promotors, Prof. Dr. Dubruel
and Prof. Dr. Van Der Voort, for the opportunity to work in the laboratory of the Polymer
Chemistry and Biomaterials (PBM) Group and of the Centre for Ordered Materials,
Organometallics and Catalysis (COMOC) group.
I would also like to extend my sincerest thanks to my supervisor, Dr. Karen Leus, for
the unreserved support and guidance from the start of the experiments until the
completion of this manuscript.
My special thanks goes to the members of both the PBM and the COMOC group for
their kind assistance in the lab whenever I had questions. It was a great pleasure
working with these people.
I would also like to thank Ranjith R.M., V. Cremer, O. Janssen, R. Blanckaert and L.
Martin for their help throughout this project. Their assistance was needed to complete
this manuscript.
Lastly, I would like to thank my parents and my brother for their unconditional love and
support. They always inspire me to work hard and do better in every endeavour I take
in life. I dedicate this work to them.
Luca Verhoeven
ii
TABLE OF CONTENT 1 INTRODUCTION AND AIM ................................................................................. 1
2 REVIEW OF LITERATURE ................................................................................. 3
2.1 Metal Organic Frameworks ........................................................................... 3
2.1.1 Introduction ............................................................................................. 3
2.1.2 MIL-101 ................................................................................................... 4
2.1.3 Synthesis of MOFs and modification towards catalysis .......................... 5
2.1.4 Post-Modification of MOFs ...................................................................... 6
2.2 Heterogeneous Catalysis ............................................................................ 13
2.2.1 Introduction ........................................................................................... 13
2.2.2 Hydrogenation ....................................................................................... 13
2.3 State of The Art in MOF fixation in polymer scaffolds .................................. 14
2.3.1 Introduction ........................................................................................... 14
2.3.2 MOF/Polymer hybrid material by electrospinning .................................. 14
2.3.3 Electrospinning of Polymer/MOF composite suspensions .................... 15
2.3.4 Synthesis of MOFs on polymer electrospun nanofibers ........................ 15
2.4 Electrospinning ............................................................................................ 18
2.4.1 Introduction ........................................................................................... 18
2.4.2 Electrospinning mechanism .................................................................. 18
2.4.3 Electrospinning parameters .................................................................. 19
2.4.4 Solution parameters .............................................................................. 19
2.4.5 Processing parameters ......................................................................... 21
2.5 In depth study: PCL solutions for electrospinning ........................................ 22
3 MATERIALS AND METHODS ........................................................................... 25
3.1 Preparation of polymer solution for electrospinning ..................................... 25
3.1.1 Chemicals ............................................................................................. 25
3.1.2 Preparation of polymer solutions ........................................................... 25
3.1.3 Electrospinning procedure .................................................................... 25
3.1.4 Image analysis ...................................................................................... 26
3.2 Preparation of Pt-functionalized MIL-101 .................................................... 27
3.2.1 Chemicals ............................................................................................. 27
3.2.2 Synthesis of MIL-101 ............................................................................ 27
3.2.3 Synthesis of Pt@MIL-101 ..................................................................... 27
3.3 Catalytic setup and analysis ........................................................................ 28
4 RESULTS AND DISCUSSION .......................................................................... 29
iii
4.1 Pt@MIL-101 characterization ...................................................................... 29
4.1.1 Nitrogen sorption analysis ..................................................................... 29
4.1.2 XRPD and ICP-OES measurements ..................................................... 30
4.2 Electrospinning of polymer solutions ........................................................... 31
4.2.1 Introduction ........................................................................................... 31
4.2.2 Electrospinning of PCL from chloroform/acetone solutions ................... 32
4.2.3 Electrospinning of PCL from DCM/DMF solutions ................................ 36
4.2.4 Electrospinning of PCL from DCM/HCOOH solutions ........................... 38
4.2.5 SEM analysis of optimized electrospun PCL fibers ............................... 40
4.3 Effect of solvent on production of PCL/MIL-101 electrospun fibers ............. 42
4.3.1 Introduction ........................................................................................... 42
4.3.2 PCL/MIL-101 fibers from chloroform/acetone solutions ........................ 42
4.3.3 PCL/MIL-101 fibers from DCM/DMF solution ........................................ 43
4.3.4 PCL/MIL-101 fibers from DCM/HCOOH solutions ................................ 45
4.4 Characterization of the Pt@MIL-101/PCL electrospun fibers ...................... 47
4.4.1 Preparation of Pt@MIL-101/PCL fibers ................................................. 47
4.4.2 SEM analysis of Pt@MIL-101/PCL fibers ............................................. 48
4.4.3 SEM-EDX mapping of Pt@MIL-101/PCL fibers .................................... 49
4.4.4 XRPD analysis of Pt@MIL-101/PCL fibers ........................................... 50
4.4.5 Thermal analysis of Pt@MIL-101/PCL fibers ........................................ 50
4.5 Catalysis with Pt@MIL-101/PCL electrospun fibers .................................... 53
4.5.1 Introduction ........................................................................................... 53
4.5.2 Pt content in Pt@MIL-101/PCL electrospun fibers ................................ 53
4.5.3 Catalytic performance of Pt@MIL-101/PCL .......................................... 54
4.6 Analysis of Pt@MIL-101/PCL fibers after catalysis ..................................... 56
5 CONCLUSION .................................................................................................. 59
6 REFERENCES .................................................................................................. 61
iv
LIST OF ABBREVIATIONS
ALD atomic layer deposition
AIM atomic layer deposition in metal organic frameworks
ATW ALD temperature window
COMOC Centre for Ordered Materials, Organometallics and Catalysis
CVD Chemical vapor deposition
DCM dichloromethane
DMF dimethylformamide
DSC differential scanning calorimetry
EM electron microscopy
GC gas chromatography
HCl hydrogen chloride
HCOOH formic acid
ICD injector - collector distance
ICP-OES inductively coupled plasma - optical emission spectroscopy
MIL Materials Institute Lavoisier
MOFs metal organic frameworks
OM optical microscopy
PAA poly(acrylic acid)
PAN polyacrylonitrile
PBM Polymer Chemistry and Biomaterials group
PCL poly-𝜀-caprolactone
PLA poly(lactic acid)
PTFE polytetrafluoroethylene
PVA poly(vinyl alcohol)
PVD physical vapor deposition
PVP poly(vinylpyrrolidone)
RT room temperature
SEC size exclusion chromatography
SEM scanning electron microscopy
TCD thermal conductivity detector
TGA therogravimetric analysis
THC tetrahydrocannabinol
TOF turn over frequency
TON turn over number
XRF X-ray fluorescence
XRPD X-ray powder diffraction
v
ABSTRACT
In this project, Pt nanoparticles were in situ synthesized in MIL-101 (Materials Institute
Lavoisier) by atomic layer deposition (ALD). The obtained Pt@MIL-101 powders were
characterized by means of N2 adsorption and X-ray powder diffraction (XRPD)
measurements. Only a slight decrease in surface area without changes in the
crystalline structure of MIL-101 are observed after ALD. X-ray Fluorescence (XRF)
measurements were performed to determine the loading of Pt after 120 cycles of ALD.
As it was the aim to achieve a polymer scaffold with a narrow fiber distribution different
mixtures of binary solvents with increasing dielectric constant were examined,
including CHCl3/acetone, DCM/DMF and DCM/HCOOH. During optimization, the
resulted fibers were examined by optical microscopy. This study revealed that
DCM/HCOOH solutions gave the best results in terms of average fiber diameter and
distribution. However, with the introduction of Pt@MIL-101, it was noticed that it was
impossible to electrospin the DCM/HCOOH solution once MIL-101 was added to the
solution. As a result, the DCM/DMF solution was employed in the following studies as
the processing of the Pt@MIL-101/PCL fibers was much more facile despite the slightly
larger fibers compared to the DCM/HCOOH solution. The final Pt@MIL-101/PCL
electrospun fibers were characterized by means of XRPD, Scanning Electron
Microscopy - Energy Dispersive X-ray (SEM-EDX), Thermogravimetric Analysis
(TGA), Differential Scanning Calorimetry (DSC) and Inductive Coupled Plasma –
Optical Emission Spectroscopy (ICP-OES). It is observed that there was
homogeneous distribution of Pt throughout the electrospun material. Finally, the
catalytic performance of the Pt@MIL-101/PCL fibers was examined using the
hydrogenation of cyclohexene as model reaction and compared with the pure Pt@MIL-
101 powder. The tests revealed that the pure Pt@MIL-101 was more reactive as full
conversion occurred after approximately 50 minutes while the Pt@MIL-101 embedded
in the PCL matrix gave full conversion after 100 minutes. Furthermore, reusability tests
revealed that the activity slightly decreases as a function of the amount of runs
performed for the Pt@MIL-101/PCL fibers. Moreover, no particular leaching of Pt and
Cr was observed. Additional SEM analyses of the fibers after catalysis exhibited no
changes in the size and shape of the fibers in comparison to the pristine material which
indicates that polymers are suitable host materials to embed various MOF-based
catalysts which can be used during multiple catalytic processes.
1
Metal Organic Framework loaded electrospun poly-𝜺-caprolactone scaffolds
as novel catalytic system
L. Verhoevena,b, K. Leusb, P. Dubruela*, P. Van Der Voort b*
a Polymer Chemistry and Biomaterials Group, University of Ghent, Department of Organic and
Macromolecular Chemistry, Ghent 9000, Belgium b Centre for Ordered Materials, Organometallics and Catalysis, Department of Inorganic and
Physical Chemistry, Ghent 9000, Belgium
Keywords: electrospinning, poly-𝜀-caprolactone, Metal Organic Frameworks, atomic layer
deposition, hydrogenation, catalysis.
In this study, we present for the first time the embedding of Pt@MIL-
101 in a poly-𝜀-caprolactone (PCL) matrix by means of electrospinning.
The obtained composite material was analyzed by various
characterization techniques, showing that the Pt@MIL-101 material was
homogeneously distributed along the fibers. Hereafter, the Pt@MIL-
101/PCL electrospun fibers were examined as a catalyst in the
hydrogenation of cyclohexene exhibiting a good activity. Moreover,
reusability tests and stability tests demonstrated that the material could
be recycled at least 4 runs without detectable Cr and Pt leaching.
Introduction
Metal Organic Frameworks (MOFs) are a class of porous crystalline materials build up by a
combination of inorganic and organic molecules. Since their discovery in the late nineties1,
MOFs have already been utilized in a variety of applications such as gas storage, gas separation,
adsorption and heterogeneous catalysis2–9 due to their high surface area, exceptional porosity,
chemical tunability and flexibility. In heterogeneous catalysis, these characteristics are
exploited to find novel catalytic systems. Firstly, the careful selection of the inorganic metal
clusters and the organic linkers might result in the design of catalytic active sites as they are
present on the structure itself. Secondly, as a consequence of the high surface area and porosity
they have been used as support material to stabilize catalytic moieties onto its structure. In the
past, homogeneous metal complexes have been attached to the organic linkers by coordination
chemistry and nanoparticles have been deposited onto the inorganic metal clusters which clearly
illustrates the versatility of MOFs as support material towards their use in catalysis.
Despite the interesting catalytic properties of MOFs, there are some practical implications
related to their use in catalysis. Most MOFs are crystalline solid powders which require special
attention during catalysis as they are dispersed into the solution. Filtration steps are required to
separate the powder from the product solution after catalysis. Eventually, a small amount of the
MOF powder is lost during these operations. To mitigate the loss of powder, MOFs have been
deposited onto various types of scaffold materials in the past. In general, alumina10, silica11,
graphite oxide12 and ceramics 13 have been mostly reported as scaffold materials. The list of
polymer scaffolds remains rather limited and unexplored14 despite many desirable properties
for the preparation of composite materials, including good mechanical, thermal and chemical
stability and the simplicity of processing from polymer solutions. Interestingly, the number of
articles that concern MOF/polymer interfaces is vastly expanding, especially in the field of
electrospinning 15,16,17,18,19,20,21,22,23.
2
Electrospinning is a processing technique to form polymer fibers in the micro/nanometer range
by stretching a polymer solution due to an electric field. MOFs can either be blended in a
polymer solution before electrospinning or they can be synthesized in situ onto the prepared
electrospun fibers to process MOF/polymer composites. In literature, they have been mostly
tested as gas storage and adsorption material. Despite the fact that electrospun fibers possess a
large surface area and exhibit a rather good chemical stability, they have never been tested in
catalysis. The scope of this work is to produce a MOF/polymer composite material by
electrospinning which is tested as a novel catalytic system. The chromium-based MIL-10124
(Materials Institute Lavoisier) is chosen as MOF host due to its high chemical and thermal
stability 25. Inside the cages of MIL-101, Pt nanoparticles are deposited by ALD as reported by
K. Leus26 which showed a good catalytic performance in the hydrogenation of various alkenes.
As a proof of concept, the hydrogenation of cyclohexene to cyclohexane is used here as a
reference reaction to compare the activity of the pure Pt@MIL-101 powder and the electrospun
Pt@MIL-101 polymer material. Poly-𝜀-caprolactone (PCL) is used as a scaffold material as it
is the most used material in the field of electrospinning in the fabrication of wound dressings27,
drug delivery systems 28 and tissue engineering scaffolds29,30 in the medical field. Furthermore,
it is noted that PCL shows a great stability towards H2 which was used during the hydrogenation
reaction.
Materials and Methods
Synthesis of MIL-101.
MIL-101 was synthesized based on an adapted procedure reported by Edler et al 31. In a typical
reaction, 0,665g terephthalic acid (4mmol) and 1,608g Cr(NO3)3.9H2O (4mmol) were added to
20 mL of deionized water in a Teflon-lined autoclave. The autoclave was gradually heated to
210°C during 2 hours in a Nabertherm muffle furnace and kept at this temperature for 8 hours.
Next, the suspension was filtered by a membrane filter (0.45µm) to obtain the MIL-101 powder.
Hereafter, MIL-101 was stirred in DMF for 24 hours at 60°C to remove unreacted terephthalic
acid. Finally, MIL-101 was stirred in 1M HCl overnight at RT, filtered and dried under vacuum
at 90°C to obtain the pure MIL-101 powder.
Synthesis of Pt@MIL-101.
The deposition of Pt nanoparticles inside the cages of MIL-101 was performed by ALD using
(methylcyclopentadienyl)-trimethylplatinum (MeCpPtMe3) as Pt source and O3 as reactant at
200°C 32. The depositions were performed in a home built experimental cold-wall ALD
chamber. MIL-101 was loaded in a molybdenum sample cup which was then transferred into
the ALD reactor. After loading, MIL-101 was allowed to outgas and thermally equilibrate for
at least 1 h under vacuum. The solid MeCpPtMe3 precursor (99% Strem Chemicals), kept in a
stainless steel container, was heated above its melting point (30 °C), and the delivery line to the
chamber was heated to 60 °C. Argon was used as a carrier gas for the Pt precursor. O3 was
produced from a pure O2 flow with an OzoneLab™ OL100 ozone generator (Ozone Services,
Burton, BC, Canada), resulting in an O3 concentration of 175 µg/mL. A static exposure mode
was applied during both ALD half-cycles. The pulse time of the MeCpPtMe3 precursor was
10s, after which the valves to the pumping system were kept closed for another 20s, resulting
in a total exposure time of 30s. The same pulse time and exposure time was also used for the
O326,32. Pt@MIL-101 was obtained after 120 cycles of ALD.
Preparation of polymer solutions.
PCL pellets were dissolved in different solvents during this work. The preparation of the PCL
solution was changed according to the selected solvent. DCM/DMF polymer solutions were
prepared the day before electrospinning and stirred overnight. Prior to electrospinning, the
3
solution was placed in an ultrasonic bath for 15 minutes. Hereafter, the polymer solution was
stirred for 15 minutes to remove remaining air bubbles. DCM/HCOOH polymer solutions were
prepared to achieve the homogeneous solution as fast as possible to mitigate acid hydrolysis.
Therefore excessive use of the ultrasonic bath was required. The solution was alternatively
stirred and placed in the ultrasonic bath (15 minutes each) until a homogeneous polymer
solution was obtained. MIL-101/PCL and Pt@MIL-101/PCL solutions were obtained by
adding MIL-101 or Pt@MIL-101 to the solvent mixture (either DCM/DMF or DCM/HCOOH).
Before adding PCL, the mixture was placed in an ultrasonic bath until a homogeneous green
dispersed solution was obtained. Once PCL was added, the same procedure was performed
depending on the selected solvent mixture as described above.
Electrospinning setup.
The polymer solution was introduced into a 20 mL syringe which was connected to a Rotilabo-
PTFE tube with an internal diameter of 2 mm to an 18 gauge needle (1.270 mm outer diameter,
0.838 mm inner diameter, 3.2 cm length, Fisher Scientific). The needle was placed through a
cupper ring on which the voltage was applied. The polymer solution was purged through the
tubing and the needle by a pumping device. The setup was placed into a wooden chamber with
fume hood. The average temperature and relative humidity were 23°C and 30%, respectively.
Catalytic setup
The hydrogenation reaction occurred in a PARR reactor filled with H2 gas at an elevated
pressure of 6 bar at room temperature (18-23°C). The reactor was loaded with 70 mL ethanol
as solvent, cyclohexene as substrate, dodecane as internal standard and the catalytic system,
either the pure Pt@MIL-101 powder or the Pt@MIL-101/PCL electrospun fibers. During each
test, aliquots were gradually taken out of the mixture and subsequently analyzed by means of
gas chromatography (GC) using a split injection (ratio 1:17) on a Hewlett Packard 5890 Series
II GC with TCD detection (Santa Clara, CA, USA). The capillary column used was a Restek
XTI-5 column (Bellefonte, PA, USA) with a length of 30 m, an internal diameter of 0,25 µm.
Results and Discussion
Pt@MIL-101 characterization.
Nitrogen sorption and XRPD measurements were carried out to determine respectively the
Langmuir surface area and the crystalline structure of MIL-101 after Pt deposition. A slight
decrease in the Langmuir surface area and pore volume was noticed after the embedding of Pt
nanoparticles (Table I) without losing the crystalline structure of pristine MIL-101 (Figure 1).
ICP-OES measurements were performed to determine the Pt-content in Pt@MIL-101.
Figure 1:. The adsorption isotherms of MIL-101 and Pt@MIL-101 (left). The XRPD patterns of MIL-
101 and Pt@MIL-101 (right).
4
TABLE I. The Langmuir surface area and the pore volume were measured by N2 sorption. The Pt-loading is determined
by ICP-OES.
Sample Langmuir surface area (m²/g) Pore volume (cm³/g) Pt-loading (mmol/g)
MIL-101 3580 1,43 /
Pt@MIL-101 2907 1,35 0,387
The shape of the isotherm (type I) at low relative pressures indicated the adsorption of N2 into
two cages which is characteristic for MIL-101 powder. The pristine MIL-101 material has a
Langmuir surface area of 3580 m2/g. The higher average surface area compared to the work
of Edler et al.31 (2944 m²/g) was the result of the additional purification steps in this work to
remove unreacted terephthalic acid inside MIL-101. After Pt deposition, the Langmuir surface
area and the pore volume of the Pt@MIL-101 powder (Table I) decreased in comparison to
pristine MIL-101 powder. ICP-OES measurements revealed (Table I) that 0,387 mmol/g Pt is
present in MIL-101
Electrospinning of PCL solutions.
In an effort to obtain homogeneous PCL fibers in the nanometer range, two solutions
(DCM/DMF and DCM/HCOOH) were tested with high dielectric constant. It was reported in
literature that these solvents (eg. DMF and HCOOH) resulted in thin fibers33. As both DCM
and HCOOH were unable to readily dissolve PCL, a binary mixture with DCM was prepared.
The solubility and electrospinnability of 16% PCL (w/v) solutions using various DCM/DMF
volume ratios were tested and it was observed that the DCM/DMF (2/3) solution resulted in the
smallest fibers. Higher amounts of DMF were not able to dissolve PCL in 24 hours. For the
DCM/HCOOH binary solvent, a 1:1 ratio was prepared. 16% PCL (w/v) was readily dissolved
in DCM/HCOOH (1/1) in less than 4 hours. Size exclusion chromatography (SEC) analysis
proved that there was no detectable change in molecular weight (MW) after sample preparation.
TABLE II. Electrospinning parameters for the optimized polymer solution and the average fiber diameter with standard
deviation after electrospinning under those conditions measured by SEM analysis. The collector distance was set at 20 cm
for each polymer solution.
Polymer solution Flow rate (mL/h) Voltage (keV) diameter (µm)
16% PCL DCM/DMF (2/3 v/v) 1 15 0,7 ± 0,127
16% PCL DCM/HCOOH (1/1 v/v) 1 17 0,45 ± 0,065
Figure 2: SEM images of electrospun PCL fibers from a 16% PCL (w/v) DCM/DMF (2/3) solution
(left) and a 16% PCL (w/v) DCM/HCOOH (1/1) solution (right). The bending of the fibers (left) was
probably caused during SEM sample preparation by transferring the fibers from the glass plate to the
carbon tape as straight fibers were observed by optical microscopy.
5
Scanning electron microscopy (SEM) images were taken of the electrospun fibers (Figure 2)
and optimal electrospinning conditions were presented in Table II. It was observed that the
DCM/HCOOH solution resulted in smaller fibers compared to the DCM/DMF solution,
respectively 0.45 and 0.7 µm. This is in line with the higher dielectric constant of HCOOH in
comparison to DMF.
Electrospinning of Pt@MIL-101/PCL solutions.
The influence of the addition of MIL-101 on the electrospinning behavior was examined by
adding 4% MIL-101 to the previously mentioned polymer solutions in order to obtain 20%
(w/w) MIL-101 distributed in the PCL scaffold. It was assumed that the MIL-101/PCL
DCM/HCOOH (1/1) solution would result in the better material due to the smaller fibers
(Table II) which would be beneficial in catalysis as statistically, more active sites would be
present on the surface of the scaffold. However, it was concluded that once MIL-101 was
added to the DCM/HCOOH (1/1) polymer solution, the behavior of the electrospinning
process changed drastically as there was a tendency to form droplets. In contrast to the
DCM/HCOOH (1/1) solution, the MIL-101/PCL DCM/DMF (2/3) solution was able to be
electrospin consistently into fibers in stable Taylor cone conditions. In general, a stable
polymer jet is observed once an equilibrium is formed between the viscoelastic forces in the
polymer solution and the electrostatic forces caused by the applied electric field. As the MIL-
101/PCL mixture in DCM/HCOOH (1/1) was unable to form a stable jet, tests were
performed to examine the viscoelastic properties of the various solutions (Table III). It was
observed that especially the DCM/HCOOH (1/1) solution showed a high viscosity at low
shear rate (�̇� = 1 s-1) from which It was assumed that the MIL-101/PCL DCM/HCOOH (1/1)
polymer solution resisted deformation caused by the applied electric field which could be
partially explained by the higher viscosity. However, surface tension tests and conductivity
tests should be examined in future studies to show the effect of MIL-101 addition on these
solution parameters as they also contribute to the electrospinning process.
TABLE III. Viscosity of polymer solution applied for electrospinning as studied by rheology.
Polymer solution Viscosity (Pa.s) at �̇� = 1 s-1
16% PCL (w/v) DCM/HCOOH (1/1) 2,6543
16/4 % MIL-101/PCL (w/v) DCM/HCOOH (1/1) 5,4145
16% PCL (w/v) DCM/DMF (2/3) 2,3055
16/4 % MIL-101/PCL (w/v) DCM/DMF (2/3) 2,9410
Next, 4% (w/v) Pt@MIL-101 was blended in the DCM/DMF (2/3) polymer mixture and
electrospun at the stable electrospinning conditions (flow rate: 1 mL, voltage: 15 keV and
collector distance: 20 cm) to obtain the catalytic 20% (w/w) Pt@MIL-101/PCL electrospun
fibers. SEM analysis (Figure 3) revealed that Pt@MIL-101 crystals were present at the
surface of the PCL fibers throughout the electrospun material, including clusters of Pt@MIL-
101 anchored on the fibers. which indicated that at least partially active Pt sites were
accessible for catalysis. However, since SEM is a surface analysis technique it was not
possible to determine the percentage of Pt@MIL-101 that was entirely surrounded by the PCL
scaffold. The Pt@MIL-101/PCL fibers showed a fiber diameter of 0.78 ± 0,17 µm which is
in comparison with the pristine PCL fibers, discussed in table I.
6
Characterization of the Pt@MIL-101/PCL fibers.
XRPD measurements were performed to determine the crystalline structure of Pt@MIL-101
after embedding in PCL by electrospinning. From Figure 4 it was observed that the crystalline
structure of Pt@MIL-101 was preserved after electrospinning. Besides the typical diffractions
of MIL-101, some additional diffractions were observed which could be assigned to the semi-
crystalline PCL (Figure 4). Thermogravimetric analysis (TGA) was performed on the
Pt@MIL-101/PCL fibers and compared with pure PCL fibers, electrospun under the same
reaction conditions (flow rate 1 mL/h, voltage 15 keV and ICD 20 cm) to examine the
influence of Pt@MIL-101 on the PCL scaffold material in terms of thermal stability.
The thermogram of the 20% (w/w) Pt@MIL-101 electrospun PCL fibers (Figure 4) showed
that the onset of degradation occurred at a lower temperature (300°C) compared to the pristine
PCL fibers (350°C), indicating that the addition of Pt@MIL-101 negatively influenced the
TABLE IV. ICP-OES measurements of two random samples of the electrospun composite material. The theory is based
on the fact that only 20% (w/w) of Pt@MIL-101 is in theory homogeneously present in the composite. It was already
calculated that the Pt-loading in pure Pt@MIL-101 was 0.387 mmol/g (Table I).
THEORY SAMPLE 1 SAMPLE 2
Pt Loading 0.0774 mmol/g 0.0779 mmol/g 0.0784 mmol/g
Figure 3: (A-D) SEM images of Pt@MIL-101/PCL fibers. (B) cluster of Pt@MIL-101 trapped in the
PCL scaffold.
B
C D
A
7
thermostability of PCL. It was also observed that the degradation of PCL occurred over a
broader temperature range (300-400°C). At a temperature of 400°C, approximately 20 % of
the weight remained which could be accounted to Pt@MIL-101. Above 500°C Pt@MIL-101
was fully degraded with an inorganic residue of Pt and Cr.
Inductively coupled plasma – optical emission spectroscopy (ICP-OES) measurements were
performed to determine the exact Pt loading per gram of composite material (Table IV) as it is
important to know the exact amount of Pt sites in the electrospun material for catalysis. It was
concluded that both samples matched the theoretical value, showing that Pt@MIL-101 was
homogenously distributed in the PCL scaffold. Chemisorption experiments were conducted to
determine the amount of accessible Pt sites in the electrospun composite material. H2 gas was
used as chemisorption gas as it shows great affinity for metallic Pt. It was already derived that
the Pt loading in the electrospun fibers was 0.0774 mmol/g based on ICP-OES measurements.
The chemisorption analysis revealed that 0.05741 mmol/g of Pt nanoparticles in the
electrospun material was readily available to adsorb H2. Essentially, this means that 25% of
the Pt sites were unable to be used as a catalyst, because the transport of H2 to these active
sites was inhibited.
Catalytic performance of Pt@MIL-101/PCL fibers.
The hydrogenation of cyclohexene to cyclohexane was used as a proof of concept to examine
the catalytic activity and accessibility of the Pt@MIL-101/PCL fibers. The results were
compared with pure Pt@MIL-101 powder as catalyst. In each catalytic test 0.0294 mmol of
Pt was used. The cyclohexene / Pt ratio was 400 and dodecane was used as internal standard.
After each catalytic test, the fibers were washed several times with ethanol, dried and kept
under vacuum for at least one hour. The TON number was calculated by dividing the amount
of mmol product (cyclohexane) by the number of active sites at the end of the reaction while
the TOF number was determined by dividing the TON number by the reaction time (min)
after 10 minutes of reaction time.
TABLE V. Turnover frequency (TOF) and leaching percentage of Pt for each catalytic test. TOF was calculated after 10
minutes of reaction time*. The leaching of Pt was lower than the detectable limit measurable by XRF**
Catalyst TOF (min-1)* Leaching of Pt (%)**
Pt@MIL-101 powder 16.91 <0.05
Pt@MIL-101/PCL fibers RUN 1 8.31 <0.05
Pt@MIL-101/PCL fibers RUN 2 9.68 <0.05
Pt@MIL-101/PCL fibers RUN 3 8.90 <0.05
Pt@MIL-101/PCL fibers RUN 4 8.1 <0.05
Figure 4: XRPD patterns of pristine PCL fibers and Pt@MIL-101/PCL fibers derived from a
DCM/DMF (2/3) solution (left). TGA of pristine PCL fibers and Pt@MIL-101/PCL fibers (right). The
thermograms were measured in an N2 atmosphere with a heating rate of 10°C/min. (right)
8
The catalytic tests showed that Pt@MIL-101/PCL exhibited full conversion of cyclohexene
after 90 minutes of reaction time in the first run, while Pt@MIL-101 powder showed full
conversion after only 50 minutes of reaction time. The Pt@MIL-101 powder has a TOF
number of 16.91 min-1 whereas for the composite material a TOF of 8.31 min-1 was obtained.
It could thus be concluded that despite the fact that the same amount of Pt active sites were
present in the Pt@MIL-101/PCL electrospun fibers and the Pt@MIL-101 powder, the
electrospun catalytic system showed a decrease in kinetics towards the hydrogenation of
cyclohexene. Based on chemisorption experiments, it was already discussed that
approximately 25% of the Pt sites are inaccessible for H2. This might explain the difference in
the TOF values.
To examine the reusability of the Pt@MIL-101/PCL fibers, in total 4 runs were performed.
Similar TOF values were obtained during these additional runs, demonstrating that the fibers
could be reused for multiple runs without a significant decrease in its catalytic performance.
Moreover, as can be seen from Table V, no leaching of Pt was noted during these runs,
showing the strong embedding of the Pt@MIL-101 material in the PCL fibers.
Figure 6: XRPD pattern of Pt@MIL-101/PCL electrospun fibers after 4 runs of catalysis compared
with the XRPD pattern before catalysis.
0
10
20
30
40
50
60
70
80
90
100
0 20 40 60 80 100 120
Con
ver
sion
(%
)
Reaction Time (minutes)
Pt@MIL-101/PCL RUN 1
Pt@MIL-101/PCL RUN 2
Pt@MIL-101/PCL RUN 3
Pt@MIL-101/PCL RUN 4
Pt@MIL-101 powder
Figure 5: Conversion of cyclohexene to cyclohexane catalyzed by Pt@MIL-101/PCL fibers and by
Pt@MIL-101 powder. Multiple catalytic tests were performed with the same electrospun material.
9
Additionally, XRPD measurements and SEM analysis were carried out on the composite
material after 4 runs of catalysis. The XRPD pattern showed that the characteristic crystalline
patterns of MIL-101 was preserved during at least 4 multiple runs (Figure 6). It was also
concluded that the semi-crystallinity of PCL was still present after catalysis. SEM analysis
showed that Pt@MIL-101 crystals were still present at the surface of the composite material
(Figure 7) and that the integrity of the fibers was preserved. At last, SEC analysis showed
that there was no detectable decrease of molecular weight of PCL after catalysis. This was
anticipated, because PCL is stable in the reducing H2 environment.
Conclusion
In this work, 16% (w/v) PCL DCM/DMF (2/3) and 16% (w/v) PCL DCM/HCOOH (1/1) were
successfully processed by electrospinning to achieve PCL fibers in the submicron range (0.7
and 0.45 µm respectively) as analyzed by SEM analysis. It was observed that the addition of
MIL-101 to the polymer solution influenced the electrospinning process as it changed the
solution parameters, making it impossible to electrospin the MIL-101/PCL DCMHCOOH (1/1)
polymer composite solution. Electrospinning of MIL-101/PCL and Pt@MIL-101/PCL in
DCM/DMF (2/3) as solvent could be performed under stable electrospinning conditions. SEM
images of the corresponding Pt@MIL-101/PCL electrospun fibers showed that Pt@MIL-101
was partially present at the surface of the PCL scaffold. ICP-OES measurement revealed that
the actual Pt loading was exactly the same as theoretically predicted from which it could be
derived that the Pt nanoparticles were homogeneously present throughout the electrospun
material. However, from chemisorption analysis with H2, it was concluded that 25% of the
Pt@MIL-101 crystals was inaccessible for catalysis. Catalytic tests were conducted to examine
the performance of the Pt@MIL-101/PCL material compared to pure Pt@MIL-101 powder
with the same Pt loading. It was observed that full conversion with Pt@MIL-101/PCL fibers
was reached after 90 minutes of reaction time while in the case of Pt@MIL-101 powder full
conversion occurred after 50 minutes. Reusability tests showed that the activity of Pt@MIL-
101/PCL electrospun fibers slightly decreases without detectable leaching of Pt@MIL-101 into
the reaction medium. The Pt@MIL-101/PCL fibers were examined after catalysis by means of
SEM and XRPD analysis, showing that the fiber morphology and the crystallinity of Pt@MIL-
101 were preserved after 4 catalytic runs.
Figure 7: SEM analysis of Pt@MIL-101/PCL electrospun fibers after 4 runs of catalysis.
10
Acknowledgments
I would like to thank my promotors Prof. Dr. Dubruel and Prof. Dr. Van Der Voort for the
privilege of working in their labs. I would like to thank Dr. Karen Leus to guide me throughout
the thesis and to assist me during the experiments. I also would like to sincerely thank the
members of the PBM group and the COMOC group to accept me into their lab and to help me
whenever I had a question. I would like to thank the following people: Ranjith R.M. and V.
Cremers to perform ALD of Pt in MIL-101, O. Janssens to perform SEM analysis and R.
Blanckaert to perform ICP-OES experiments.
References
1. Li, H., Eddaoudi, M., O’Keeffe, M. & Yaghi, O. M. Design and synthesis of an exceptionally stable and
highly porous metal-organic framework. Nature 402, 276–279 (1999).
2. Li, B., Wen, H.-M., Zhou, W. & Chen, B. Porous Metal–Organic Frameworks for Gas Storage and
Separation: What, How, and Why? J. Phys. Chem. Lett. 5, 3468–3479 (2014).
3. Kayal, S., Sun, B. & Chakraborty, A. Study of metal-organic framework MIL-101(Cr) for natural gas
(methane) storage and compare with other MOFs (metal-organic frameworks). Energy 91, 772–781
(2015).
4. Furukawa, H., Cordova, K. E., O’Keeffe, M. & Yaghi, O. M. The Chemistry and Applications of Metal-
Organic Frameworks. Science (80-. ). 341, 974 (2013).
5. Teo, H. W. B., Chakraborty, A. & Kayal, S. Evaluation of CH4 and CO2 adsorption on HKUST-1 and
MIL-101(Cr) MOFs employing Monte Carlo simulation and comparison with experimental data. Appl.
Therm. Eng. 110, 891–900 (2017).
6. Darunte, L. A., Oetomo, A. D., Walton, K. S., Sholl, D. S. & Jones, C. W. Direct Air Capture of CO2
Using Amine Functionalized MIL-101(Cr). ACS Sustain. Chem. Eng. 4, 5761–5768 (2016).
7. Maksimchuk, N. V. et al. Metal-organic frameworks of the MIL-101 family as heterogeneous single-site
catalysts. Proc. R. Soc. A Math. Phys. Eng. Sci. 468, 2017–2034 (2012).
8. Zhang, W. et al. A family of metal-organic frameworks exhibiting size-selective catalysis with
encapsulated noble-metal nanoparticles. Adv. Mater. 26, 4056–4060 (2014).
9. Pan, H. et al. Pt nanoparticles entrapped in mesoporous metal-organic frameworks MIL-101 as an
efficient catalyst for liquid-phase hydrogenation of benzaldehydes and nitrobenzenes. J. Mol. Catal. A
Chem. 399, 1–9 (2015).
10. Mao, Y., Cao, W., Li, J., Sun, L. & Peng, X. HKUST-1 membranes anchored on porous substrate by
hetero MIL-110 nanorod array seeds. Chem. - A Eur. J. 19, 11883–11886 (2013).
11. Sachse, A. et al. In situ synthesis of Cu–BTC (HKUST-1) in macro-/mesoporous silica monoliths for
continuous flow catalysis. Chem. Commun. 48, 4749 (2012).
12. Li, L. et al. A MOF/graphite oxide hybrid (MOF: HKUST-1) material for the adsorption of methylene
blue from aqueous solution. J. Mater. Chem. A 1, 10292–10299 (2013).
13. Granato, T., Testa, F. & Olivo, R. Catalytic activity of HKUST-1 coated on ceramic foam. Microporous
Mesoporous Mater. 153, 236–246 (2012).
14. Bradshaw, D., Garai, A. & Huo, J. Metal-organic framework growth at functional interfaces: thin films
and composites for diverse applications. Chem. Soc. Rev. 41, 2344–2381 (2012).
15. Liu, C. et al. General Deposition of Metal-Organic Frameworks on Highly Adaptive Organic-Inorganic
Hybrid Electrospun Fibrous Substrates. ACS Appl. Mater. Interfaces 8, 2552–2561 (2016).
16. Wahiduzzaman, Khan, M. R., Harp, S., Neumann, J. & Sultana, Q. N. Processing and Performance of
11
MOF (Metal Organic Framework)-Loaded PAN Nanofibrous Membrane for CO2 Adsorption. J. Mater.
Eng. Perform. 25, 1276–1283 (2016).
17. Gao, M., Zeng, L., Nie, J. & Ma, G. Polymer-metal-organic framework core-shell framework nanofibers
via electrospinning and their gas adsorption activities. Rsc Adv. 6, 7078–7085 (2016).
18. Bechelany, M. et al. Highly Crystalline MOF-based Materials Grown on Electrospun Nanofibers.
Nanoscale 5794–5802 (2015). doi:10.1039/C4NR06640E
19. Fan, X. et al. Characterization and application of zeolitic imidazolate framework-8@polyvinyl alcohol
nanofibers mats prepared by electrospinning. Mater. Res. Express 4, 26404 (2017).
20. Asiabi, M., Mehdinia, A. & Jabbari, A. Electrospun biocompatible Chitosan/MIL-101 (Fe) composite
nanofibers for solid-phase extraction of Δ9-tetrahydrocannabinol in whole blood samples using Box-
Behnken experimental design. J. Chromatogr. A 1479, 71–80 (2017).
21. Asiabi, M., Mehdinia, A. & Jabbari, A. Preparation of water stable methyl-modified metal-organic
framework-5/polyacrylonitrile composite nanofibers via electrospinning and their application for solid-
phase extraction of two estrogenic drugs in urine samples. J. Chromatogr. A 1426, 24–32 (2015).
22. Ren, J. et al. Electrospun MOF nanofibers as hydrogen storage media. Int. J. Hydrogen Energy 40,
9382–9387 (2015).
23. Quir??s, J. et al. Antimicrobial metal-organic frameworks incorporated into electrospun fibers. Chem.
Eng. J. 262, 189–197 (2015).
24. Chen, Y. F., Babarao, R., Sandler, S. I. & Jiang, J. W. Metal-organic framework MIL-101 for adsorption
and effect of terminal water molecules: From quantum mechanics to molecular simulation. Langmuir 26,
8743–8750 (2010).
25. Leus, K. et al. Systematic study of the chemical and hydrothermal stability of selected ‘stable’ Metal
Organic Frameworks. Microporous Mesoporous Mater. 226, 110–116 (2016).
26. Leus, K. et al. Atomic Layer Deposition of Pt Nanoparticles within the Cages of MIL-101: A Mild and
Recyclable Hydrogenation Catalyst. Nanomaterials 6, 45 (2016).
27. Du, L., Xu, H., Li, T., Zhang, Y. & Zou, F. Fabrication of silver nanoparticle/polyvinyl
alcohol/polycaprolactone hybrid nanofibers nonwovens by two-nozzle electrospinning for wound
dressing. Fibers Polym. 17, 1995–2005 (2016).
28. Bhardwaj, N. & Kundu, S. C. Electrospinning: A fascinating fiber fabrication technique. Biotechnol.
Adv. 28, 325–347 (2010).
29. Patrício, T., Domingos, M., Gloria, A. & Bártolo, P. Characterisation of PCL and PCL/PLA scaffolds for
tissue engineering. Procedia CIRP 5, 110–114 (2013).
30. Ghosal, K., Manakhov, A., Zajíčková, L. & Thomas, S. Structural and Surface Compatibility Study of
Modified Electrospun Poly(ε-caprolactone) (PCL) Composites for Skin Tissue Engineering. AAPS
PharmSciTech 18, 72–81 (2016).
31. Jiang, D. M., Burrows, A. D. & Edler, K. J. Size-controlled synthesis of MIL-101(Cr) nanoparticles with
enhanced selectivity for CO2 over N2. CrystEngComm 13, 6916–6919 (2011).
32. Dendooven, J. et al. Low-temperature atomic layer deposition of platinum using
(methylcyclopentadienyl)trimethylplatinum and ozone. J. Phys. Chem. C 117, 20557–20561 (2013).
33. Luo, C. J., Stride, E. & Edirisinghe, M. Mapping the influence of solubility and dielectric constant on
electrospinning polycaprolactone solutions. Macromolecules 45, 4669–4680 (2012).
1
1 INTRODUCTION AND AIM
Metal organic frameworks (MOFs) are a class of porous crystalline materials build up
by inorganic and organic building blocks. Since the discovery in the late nineties1,
MOFs have already been utilized in a variety of areas such as gas storage, gas
separation, adsorption and heterogeneous catalysis 2–9 . The interest in this set of
materials is a result of the high achievable surface areas, the exceptional high porosity,
the chemical tunability and flexibility.
In heterogeneous catalysis these characteristics are exploited to find novel catalytic
systems. Firstly, the careful selection of the inorganic metal clusters and the organic
linkers can result in the design of catalytic active sites as they are present on the
structure itself. Secondly, as a consequence of the high surface area and porosity they
have been used as support material to stabilize catalytic moieties onto its structure. In
the past, homogeneous complexes have been attached to the organic linkers by
coordination chemistry and active metals have been deposited onto the inorganic
metal clusters which clearly illustrates the versatility of MOFs as support material
towards their use in catalysis. Different incorporation techniques are available in
literature to deposit the catalysts inside the cages/channels of MOFs. The most
examined techniques are incipient wetness impregnation, solution impregnation, solid
grinding, chemical vapor deposition and atomic layer deposition.
Despite the interesting catalytic properties of MOFs, there are some practical
implications related to their employment during catalysis. Most MOFs are crystalline
solid powders which require special attention during catalysis as they are dispersed
into the solution. Filtration steps are required to separate the powder from the product
solution after catalysis. As a result, a fraction of the powder is lost during operations.
Most reactions are limited to batch type reactors as the powder must be contained. In
theory, MOFs are able to be packed, so they can be used in a continuous flow reactor
as pellets which is more attractive in industrial applications. However, during
preparation, high pressures are often required which might change the characteristic
properties of the material. To mitigate the loss of powder, MOFs have been deposited
onto various types of scaffold materials in the past. In general, alumina 10, silica 11,
graphite oxide 12 and ceramics 13 have been mostly reported as scaffold material. The
list of polymer scaffolds remains rather limited and unexplored 14 despite many
2
desirable properties for the preparation of composite materials, including good
mechanical, thermal and chemical stability and the simplicity of processing from
polymer solutions. Interestingly, the number of articles that concern MOF/polymer
interfaces is vastly expanding, especially in the field of electrospinning.
Electrospinning is a processing technique to form polymer fibers in the
micro/nanometer range by stretching a polymer solution due to an electric field. MOFs
can either be blended in a polymer solution before electrospinning or they can be
synthesized in situ onto the prepared electrospun fibers to process MOF/polymer
composites. In literature, they have been mostly tested as gas storage and adsorption
materials (see literature study §2.3). Electrospun materials possess a large external
surface area and exhibit rather good chemical stability. Nevertheless, MOF/polymer
materials have never been tested in catalysis.
The scope of this project is to produce a MOF/polymer composite material by
electrospinning which is tested as a novel catalytic system. The chromium-based MIL-
10115 (Materials Institute Lavoisier) is chosen as MOF host due to its high chemical
and thermal stability16. MIL-101 is post-modified with Pt nanoparticles by atomic layer
deposition17. Recently, Pt@MIL-101 is reported to exhibit a good catalytic performance
in the conversion of alkenes towards alkanes. As a proof of concept, the hydrogenation
of cyclohexene to cyclohexane is used here as a reference tool to compare the activity
of the pure Pt@MIL-101 powder and the electrospun Pt@MIL-101 polymer material.
Poly-𝜀-caprolactone (PCL) is used as scaffold material as it is the working horse in the
field of electrospinning in the fabrication of wound dressings18, drug delivery systems19
and tissue engineering scaffolds20,21 in the medical field. The superior rheological and
viscoelastic properties over many of its aliphatic polyester counterparts renders PCL
easy to manufacture into scaffolds. PCL is easily dissolved in a large range of solvents
which opens up possibilities in electrospinning. Moreover, this work is considered novel
as Pt@MIL-101/PCL composites have never been synthesized and never been tested
for catalysis.
One of the aims in this work is to produce nanometer electrospun fibers to increase
the surface area. Statistically, the higher the surface area, the more MIL-101 is present
on the surface for catalysis. This proves to be a challenging task as most common
electrospinning solutions of PCL show fibers in the micrometer range.
3
2 REVIEW OF LITERATURE
2.1 Metal Organic Frameworks
2.1.1 Introduction
Metal Organic Frameworks (MOFs) are porous crystalline materials which are
constructed by inorganic metal clusters and organic linkers hold together by
coordinative bounds. Since the first report in 1999 by Yaghi et al.1, researchers
combined various inorganic and organic molecules to obtain structures with record-
breaking properties.
The high surface area is one of the advantages of this vastly expanding field of
materials. This feature is attractive in applications such as H2 storage 2,3, gas
adsorption 5,6, gas separation and catalysis. The global challenges encountered today
in order to have a more sustainable way of living encourages researches to find a more
optimal solution in transport and emissions. H2 gas is considered a valid alternative for
gasoline and diesel in the future, but the storage of H2 gas is still a challenging problem.
Due to the increasing industrial activities which results in an enhanced emission of
greenhouse gasses, improved adsorption material is required. The large surface area
of MOFs is an important parameter to overcome these challenges. Today, NU-110 is
reported to possess the largest surface area having a Langmuir surface area above
7000m²/g 22. A second advantage of MOFs is the possibility to tailor the dimensions of
the pores by a careful selection of the building blocks which can be useful in (shape
selective) catalysis and selective adsorption. Up to now, MOF-74 is considered the
MOF with the largest pore size23. The strength of the tunability is illustrated by changing
the amount of consecutive phenylene units in the organic building block in MOF-74. It
is observed that the pore aperture of the hexagonal rod-like structure increases as the
amount of phenylene units increases, ranging from 18 to 98 angstrom. The largest
pores even allow the diffusion of bulky molecules such as proteins which cannot diffuse
within zeolites and zeotypes. Another interesting property of MOFs is their structural
flexibility which is often denoted as “breathing”. This reversible mechanism during the
adsorption of certain molecules induces changes in the pore size without changing the
crystallinity of the MOF. In history, MOF-53(Al) is the first MOF which exhibited such a
breathing behavior showing two well-defined states 24: an expanded state with large
pores and a contracted state with small pores25.
4
2.1.2 MIL-101
MIL-101 (MIL = Materials Institute Lavoisier) 26 is a chromium-based MOF which has
been widely examined due to the high stability of the crystalline structure at high
temperature (up to 400°C) and in most environments (acids, bases, oxidants and
reductants). A remarkable feature of this material is the high stability in water16. Most
MOFs collapse in contact with water due to interaction of the water molecules with the
hydrophilic metal nodes. As a consequence, the coordinative bonds between the latter
and the organic linkers is interrupted resulting in the loss of crystallinity. Due to the
high stability, MIL-101 has already been widely examined in gas storage3, in sensors27,
in filtration setups28 and in gas adsorption 5,28. MIL-101 is assembled by corner-sharing
super tetrahedra where the vertices and the edges are occupied by respectively the
Cr3O trimers and the 1,4-benzenedicarboxylic acid groups. The super tetrahedra are
microporous with a free aperture of 8.6 Å and can be seen as building blocks for the
formation of mesoporous quasi-spherical cages. Two types of cages can be
distinguished: a small cage (free diameter of 29 Å) of 20 tetrahedra building blocks
which is accessible through a pentagonal window of 12 Å, and a large cage (free
diameter of 35 Å) of 28 tetrahedra which is accessible through both hexagonal and
pentagonal windows with a pore aperture of respectively 14.7 Å and 16 Å15,7. Each
octahedral Cr is bonded to 4 oxygen atoms from carboxylates, 1 oxygen to form a
trimer with two other Cr atoms and 1 terminal site. The latter site is occupied by water
molecules, but these bindings can be cleaved during dehydration to obtain unsaturated
Cr based Lewis sites.
Figure 1: Presentation of MIL-101 from starting material to the 3D structure. The structure of the two cages are presented 26.
5
2.1.3 Synthesis of MOFs and modification towards catalysis
2.1.3.1 Synthesis of MOFs
MOFs are in most cases synthesized under hydrothermal or solvothermal conditions
using an autoclave. The building blocks (organic linker molecules and inorganic metal
salts) are mixed together in a suitable solvent. The reaction conditions such as
temperature and pressure, depend on the synthesized MOF. The synthesis is followed
by multiple purification steps to remove unreacted linkers and solvents to obtain the
crystalline powder in a high yield.
2.1.3.2 Synthesis of catalytically active MOFs
The inorganic clusters and organic bridging molecules can be modified to have
catalytically active sites readily available in the framework. In the following section,
some examples will be highlighted demonstrating the flexibility of the MOF material in
catalysis as not only the organic linkers but also the inorganic metal nodes can be used
in catalytic applications.
Wu et al. designed a MOF in which the chiral bridging ligand, (R)-6,6′-dichloro-2,2′-
dihydroxy-1,1′-binaphthyl-4,4′-bipyridine contains bipyridyl primary functional groups
and orthogonal chiral 2,2′-dihydroxy secondary functionalities were present. The
readily accessible dihydroxy groups could react with Ti(OiPr)4 to afford Lewis acidic
compounds29. The obtained materials were used as catalysts in the addition of ZnEt2
to aromatic aldehydes towards the formation of chiral secondary alcohols30. In a report
by Gomez-Lor et al. a MOF was synthesized in which the inorganic nodes contain
Indium31. The In2(OH)3(BDC)1.5 was tested as a catalyst in the hydrogenation of
nitroaromatics and oxidation of sulfides. These examples illustrate the strength of MOF
in catalysis. During the design and synthesis of the material, specific organic linkers
and metallic nodes can be selected which function as active sites in catalytic reactions.
6
2.1.4 Post-Modification of MOFs
2.1.4.1 Introduction
MOFs are able to be post-modified to further extend the range of applications in the
field of heterogeneous catalysis by deposition of active metal-ions. Nanoparticles as
such in catalysis have the tendency to aggregate and show limited recyclability and
recovery from the reaction medium. The MOF framework can be used as a host for
these catalytically active nanoparticles and therefore give a solution for the problems
mentioned above.
There are several ways to embed nanoparticles in MOFs such as liquid phase
impregnation, solid grinding, chemical vapor deposition (CVD) and atomic layer
deposition (ALD). These techniques are briefly discussed in the next paragraph.
2.1.4.2 Impregnation of active metals in MOFs
Impregnation of nanoparticles in MOFs is a convenient and straightforward technique
to deposit particles from solutions. Initially, the particles of interest are suspended in
a suitable solvent. Next, the solution is added to the MOF powder to start impregnation.
As the precursor solution passes through the mesoporous material, particles are
deposited in the MOF. There is often no interaction between the metal precursors and
the surface of MOFs which can cause agglomeration of the metals in the aqueous
medium. As a consequence, this process can eventually lead to an inhomogeneous
metal distribution and a lower dispersion as a result of the bigger particle diameter.
The dispersion is the ratio of the amount of nanoparticles adsorbed and the total
number of atoms on the surface of the MOF which is an identification of the deposition
of nanoparticles on the surface. The adsorption mechanism is based on electrostatic
forces. Once the metal precursors are adsorbed on the surface, redox reactions are
conducted to get the metal ions in the active oxidation state. As a consequence, the
MOFs are treated with reducing/oxidizing agents which can negatively influence the
internal structure of the MOF 32.
There are two important impregnation techniques (figure 2): wet/solution impregnation
and dry/incipient wetness impregnation based on differences in the amount of solvent
used during impregnation. In wet impregnation, the pores of the MOFs are first filled
with the used solvent. Hereafter, the wetted support is treated with the precursor
solution in which the amount of liquid is controlled by the solubility of the particles. In
7
dry impregnation, the pores are dried first to remove any residual solvent. Next, the
precursor solution is added, but the total amount of liquid is restricted to the total pore
volume of the support material. As can be predicted, the transport mechanism for both
methods are different. Respectively diffusion and capillary suction are the main
transport mechanisms in wet impregnation and dry impregnation.
It is reported that by using the dry impregnation approach, an enhanced amount of
nanoparticles will be located in the bulk of the material due to these capillary forces in
comparison to the wet impregnation method. It can be concluded that the main
advantages of this method is the simplicity and the low cost. However, Inherent
disadvantages are coupled with a solvent as carrier medium for particles in porous
materials. At first, surface tension related phenomena can occur such as incomplete
wetting of the pores, leading to inhomogeneous film formation on the support material.
Secondly, as a consequence of using an excess of solvent, dissolution of the support
material could occur33. At last, an evaporation step has to be always considered in
order to remove the solvent after impregnation. During evaporation, the concentration
of the precursor solution increases and crystallization might occur. To overcome these
phenomena, other solvent-free deposition techniques have been proposed.
2.1.4.3 Solid grinding
Solid Grinding is a solvent-free deposition of metal nanoparticles in porous materials.
The procedure can be divided into two steps. In a first step, the organometallic complex
containing the particles of interest are grounded with the support powder in an agate
mortar. In a next step, the mixture is heated in a flow of H2 gas to embed the particles
inside the cavities of the support material. Several reports have been published in
Figure 2: Illustration of Wet impregnation (left) and Dry impregnation (right).
8
which Au nanoparticles were deposited in some typical MOFs, such as IRMOF-1,
HKUST-1, MIL-5334 and ZIF-835. In catalysis, Au particles are known for the low-
temperature oxidation of CO36 and the aerobic oxidation of alcohols37.
Solid grinding is a simple and effective way for introducing metallic nanoparticles in
MOFs without the requirement of using solvents or washing steps after the deposition.
Up to date, in terms of MOF deposition, this method is solely applied for the production
of Au nanoparticles, mostly by using the precursor (CH3)2Au(acac) 34.
2.1.4.4 Chemical Vapor Deposition (CVD)
CVD is a general term that involves vapor deposition caused by a rearrangement of
chemical bonds 33. This is in contrast to physical vapor deposition (PVD) in which the
deposition is driven by adsorption rather than a chemical reaction (CVD). In general,
the vapor deposition can be divided into 3 steps: the vaporization of the precursor
solution, the transport of the vapor phase through the pores and the deposition of the
particles on the support material. In the last step, the adsorbed precursor is
decomposed by thermal, chemical or UV-radiation treatment depending on the
selected precursor. At last, a chemical bond is formed between the support material
and the embedded particles.
The most crucial step in vapor deposition is the selection of a suitable precursor. In
most cases organometallic precursors are used due to the volatile nature of these
complexes. As can be expected, low sublimation temperatures are required in order to
protect the MOF during treatment. Furthermore, an ideal precursor possesses a good
thermal stability during transport and needs to be able to decompose under rather
clean conditions and in a controlled manner with the formation of stable by-products
which can be easily removed afterwards. As can be expected, it is nearly impossible
to find suitable precursors which meet these requirements. As a consequence,
commercially available complexes are limited or even unavailable for certain
elements32.
The advantages of CVD are driven by the solvent-free nature of the technique. This
means that the inherent downsides of the use of a solvent are inhibited. Furthermore,
the introduction of a vapor phase is more efficient compared to the slow diffusion
processes in solution and allows high loading levels. However, the drawback of high
9
loading levels is the agglomeration of nanoparticles in the vapor phase and partial
degradation of the framework by incorporation of highly reactive metals.
The first study on metalorganic chemical vapor deposition (MOCVD) of metal
nanoparticles in MOFs was performed by using the early and highly porous MOF-5,
also referred to as IRMOF-1 38. High loadings of Pd (45%), Cu (28%) and Au (36%)
were obtained in the pores of the framework. However, for Cu a lower loading was
observed as a result of the larger precursor size, showing the importance of precursor
selection.
2.1.4.5 Atomic Layer Deposition (ALD)
2.1.4.5.1 Introduction
Atomic Layer Deposition (ALD) is considered a novel form of CVD in which atomic
layers of material are formed during each deposition cycle. In general, the same steps
are followed as in the case of CVD with the difference that sequential alternating pulses
of precursor molecules are introduced inside the pores. To have a better understanding
of this technique, the well-studied ALD of Al2O3 is discussed.
The formation of Al2O3 thin films by ALD can essentially be split up in two pulses of
CVD in which two different precursors are used: Al(CH3)3 and water. In the first pulse,
Al(CH3)3 is vaporized and gas-surface reactions take place with the substrate
containing HO-groups to form one atomic layer. Next, the ALD chamber is purged to
remove any unreacted precursor and by-products (methane in this case). In a second
Figure 3: ALD of Al2O3 thin films. ©Parsons Research Group.
10
pulse, water vapor is introduced to the chamber. The water reacts with the freshly
formed layer during the first pulse. After a second purge, one layer of Al2O3 is formed.
It is clear that via this technique very thin sheets of material can be formed and that
the thickness of the layer can be precisely controlled by repeating the steps (ALD cycle)
several times39,40. Another advantage of ALD over CVD is the excellent conformality of
high aspect ratio materials41. These features are often desired in the deposition of
particles in mesoporous materials to inhibit pore obstruction and to produce well-
distributed films42.
The cycle time depends on the time required to form a monolayer, a process which is
heavily influenced by diffusion rates in mesoporous materials. Kinetic models have
shown that the diffusion time is inversely related to the channel width squared and
directly related to the channel length squared43,44. The temperature is a second
parameter which must be optimized and the temperature range in which the ALD
deposition is done is referred to as the ALD temperature window (ATW). Temperatures
below ATW can cause precursor condensation and slow reaction kinetics and
temperatures above ATW can lead to thermal decomposition of the precursor and rapid
desorption42.
Table I: Overview of the most important materials grown to date by ALD
Elemental Oxides Nitrides Sulfides Others
C, Al, Si, Ti, Fe, Co, Ni,
Cu, Zn, Ga, Ge, Mo,
Ru, Rh, Pd, Ag, Ta, W,
Os, Ir, Pt
Li, Be, B, Mg, Al, Si, P,
Ca, Sc, Ti, V, Cr, Mn,
Fe, Co, Ni, Cu, Zn, Ga,
Ge, Sr, Y, Zr, Nb, Ru,
Rh, Pd, In, Sn, Sb, Ba,
La, Ce, Pr, Nd, Sm,
Eu, Gd, Tb, Dy, Ho, Er,
Tm, Yb, Lu, Hf, W, Ir,
Pt, Pb, Bi
B, Al, Si, Ti, Cu, Ga,
Zr, Nb, Mo, In, Hf, Ta,
W
Ca, Ti, Mn, Cu, Zn, Sr,
Y, Cd, In, Sn, Sb, Ba,
La, W
Li, B, Mg, Al, Si, P, Ca,
Ti, Cr, Mn, Co, Cu, Zn,
Ga, Ge, As, Sr, Y, Cd,
In, Sb, Te, Ba, La, Pr,
Nd, Lu, Hf, Ta, W, Bi
Despite the very attractive nature of ALD in porous materials (such as MOFs), there
are some complication. As a consequence of the layer-by-layer reaction behavior, the
deposition rates are rather slow (100-300 nm per hour). A second disadvantage is that
not all elements can be deposited by using this technique. A list of elements for which
suitable precursors exist is given in table I. As mentioned, in CVD, the ideal precursor
11
needs to meet certain requirements. Moreover, the limited choice in possible reaction
pathways to form atomic layers in a self-limiting way which is typical for ALD, making
it even harder to find suitable precursors42,45.
2.1.4.5.2 Atomic Layer Deposition in Metal Organic Frameworks (AIM)
The controllable thickness and the high conformability has led to the idea of introducing
catalytic nanoparticles in MOFs by using ALD. The pioneers concerning this topic came
up with three design criteria in the selection of suitable MOFs for the Atomic Layer
Deposition in Metal Organic Frameworks (AIM) 46:
1) Diffusion. The main challenge of AIM is the slow diffusion of precursor
molecules through the pores. Mesoporous materials are necessary to facilitate
diffusion.
2) Stability. The selected MOF needs to have a good (hydro)thermal stability to
survive the reaction conditions. In a typical ALD setup, the precursor vapor is
formed between 100°C and 300°C and employs steam as a co-reactant.
3) Reactivity. The MOF needs to possess spatially oriented functional groups to
guaranty the self-saturating behavior during metalation reactions.
Figure 4: Relevant structural features (left) and structure of NU-1000 46.
12
The Zr-based NU-1000 fulfills all these design criteria and is therefore employed as
proof of concept to deposit nanoparticles using ALD. More specifically NU-1000
possess the following characteristics:
1) The large hexagonal pores (30 A) allows diffusion.
2) high thermal stability up to 500°C
3) Presence of HO-groups as anchoring sites for ALD.
ALD is performed into NU-1000 using diethylzinc (ZnEt2) or trimethyl aluminum (AlMe3)
as precursors. On average, 0.5 Zn and 1.4 Al is observed per Zr atom after AIM 46.
Despite several attempts, it is not possible to deposit several layers by using these
precursors and thus the reaction is completed once all OH-groups has reacted with the
metal precursors. The deposited Al3+ and Zn2+ in NU-1000 serves as Lewis acids in
the subsequent Knoevenagel condensation 47,48. This proof of concept illustrates the
potential of AIM in heterogeneous catalysis. In 2015, the same group has reported the
use of ALD in NU-1000 to obtain a particular crystalline phase of cobalt sulfide, Co9S8,
by alternating ALD cycles of bis(N,N’-di-i-propyl acetamidinato) cobalt(II) (Co(amd)2)
and H2S 49. It verifies the hypothesis that cobalt sulfide growth occurs initially via
reaction with the hydroxide and aqua ligands of the Zr6- nodes. This report shows that
it is possible to perform ALD using alternating precursors A and B to form AB layers in
a layer to layer approach in MOFs.
In a recent report in 2016 by Leus et al., the deposition of Pt nanoparticles in MIL-101
was carried out. X-ray fluorescence and TEM analysis confirmed that the Pt-loading
increases with the number of ALD cycles. The Pt@MIL-101 powder was used in
hydrogenation catalysis showing full conversion of alkenes in the presence of
hydrogen gas at 5 bar in a Parr reactor 17.
13
2.2 Heterogeneous Catalysis
2.2.1 Introduction
The main purpose of catalysis is to increase the kinetics of a reaction by lowering the
activation energy of the rate determining step. In general, catalysts can be divided into
two group. In homogeneous catalysis, catalysts are dissolved in the reaction solution.
In heterogeneous catalysis, the catalyst (mostly solid) and the reactants (mostly liquid)
are present in a different phase. As can be imagined, the latter shows more facile
handling and regeneration as severe separation steps are not required.
2.2.2 Hydrogenation
Hydrogenation is a chemical reduction reaction between hydrogen gas (H2) and, in
principle, an organic molecule containing pi-bonds such as alkenes and aldehydes.
Hydrogenation of double bonds is a thermodynamically favorable reaction due to the
formation of more stable sigma bonds and is thus exothermic. However, in most cases,
a catalyst is required to lower the required very high reaction temperatures.
Suitable elements in catalytic hydrogenation are Pt, Pd, Co, Ni, Rh and Ru. The
mechanism involves the adsorption of H2 and the molecule containing the double bond
onto the surface of the catalyst followed by addition (Figure 5).
Figure 5: Mechanism of a typical catalytic hydrogenation reaction.
14
2.3 State of The Art in MOF fixation in polymer scaffolds
2.3.1 Introduction
Although MOFs have proven to be successful for various applications such as gas
adsorption 3,5, gas storage 2 and heterogeneous catalysis 7, there are some practical
problems which restricts their use in the chemical industry. Most MOFs are readily
available as powders and the characteristic low density of MOFs are responsible for
the formation of suspensions. As a consequence, the handling of these materials is
hard, especially in industrial reactors. Furthermore, as a powder, solely batch type
reactors can be used instead of continuous flow reactors which are often more desired
in industrial operations. As a result, many studies have been carried out to make
support materials which are able to embed the particles. Both inorganic as organic
substrates have been reported on/in which MOFs are deposited. Alumina 10, silica 11 ,
graphite oxide 12 and ceramic 13 materials have been used in the past as inorganic
supports.
Polymers have many desired properties for the preparation of composite materials,
including good mechanical, thermal and chemical stability. Various polymers can be
synthesized with different and abundant functionalities which can be tuned to improve
interactions with the MOFs. By tuning the reaction conditions one can control the
molecular weight of the polymer and influence the physical properties, shape and
porosity of the polymer scaffold. Despite the materials benefits, the list of MOF/polymer
composite materials remains relatively limited in comparison with other inorganic
composite materials such as MOF/oxides and MOF/metals14.
2.3.2 MOF/Polymer hybrid material by electrospinning
As in this thesis, the focus will lie on the use of electrospinning for the formation of
MOF/polymer hybrid materials only the latter technique will be discussed in detail. In
summary, the most novel reports in this field are given in Table II and III. To make the
overview more clear, the MOF/polymer hybrid materials are divided into two classes
based on how the synthesis of the composite material is performed:
1) Electrospinning of Polymer/MOF composite suspensions.
2) Synthesis of MOFs on polymer electrospun nanofibers
15
2.3.3 Electrospinning of Polymer/MOF composite suspensions
In a first step, the MOF crystals are formed by conventional techniques such as
hydrothermal synthesis or microwave-induced synthesis. Next, a polymer solution is
prepared in which the nanoparticles are suspended in. As can be expected, the
addition of particles into the solution increases its viscosity and alters its conductivity
and surface tension 50,51. These are important solution parameters that are taken into
account in order to form bead-free nanofibers. At last, the solution is fed into an
electrospinning device to obtain the composite nanofibers.
As illustration: In a report in 2015, a solution of MIL-101and PAN in dimethylformamide
is fed into the electrospinning device 52. As a result, a high loading MIL-101 is achieved.
Furthermore, it is concluded that vacuum degassing increased significantly the porosity
to increase the accessibility of the MIL-101 crystals in the nanofibers. The inability of
reaching the core MOF has led to the modification of the surface nanofibers which is
discussed below.
2.3.4 Synthesis of MOFs on polymer electrospun nanofibers
In another approach, a pure polymer solution is electrospun to achieve polymer
nanofibers. In a second step, MOF crystals are synthesized onto the polymer scaffold.
It is important to note that via this procedure the stability of the support material is an
important characteristic as often crystallization occurs under hydrothermal conditions.
For this reason, highly stable polymers such as polyacrylonitrile (PAN)53–55 or
composite materials 56 are used as support material.
In some cases, the nanofibers are functionalized prior to MOF crystal formation. This
is usually done by addition of the MOF building blocks to the polymer solution prior to
electrospinning. In a polymer solution of PAN, 2-methylimidazole (2-MI) is added which
serves as a building block for ZIF-8 formation 54. In a next step, the fibers are immersed
in a Zn(OAc)2.2H2O solution. Zn2+ coordinates with surface 2-MI which is evenly
distributed. Finally, by immersion in a seeding solution, ZIF-8 crystals are formed on
the surface of PAN nanofibers.
16
Table II: most recent reports considering electrospinning of MOF/polymer suspensions.
Polymer MOF Application Reference
PVA ZIF-8 Characterization of ZIF-8@PVA nanofibers mats. The hybrid material possesses an
outstanding absorption activity to remove organic pollutants from wastewater.
50
Chitosan MIL-101(Fe) Chitosan (polysaccharide) nanofibers are produced containing high MIL-101(Fe)
loading. It was successfully applied for determination of tetrahydrocannabinol (THC)
in human whole blood samples.
57
PAN CH3-MOF-5 water stable methyl-modified MOF-5 PAN composite nanofibers are formed and show
applications in the field of solid-phase extraction.
58
PAN Zr-MOF and
MIL-101(Cr)
Zr-MOF and MIL-101 are incorporated into electrospun nanofibers. By vacuum
degassing, an increase in porosity is observed which improved hydrogen storage
capabilities.
52
PLA/PVP Co-SIM-1 Co-SIM-1 stabilized by PVP is incorporated into PLA electrospun fibers. It is
concluded that the fibers become less susceptible to bacterial colonization and biofilm
formation as the concentration of Co-SIM increases in the fibers.
59
17
Table III: Most recent reports considering the synthesis of MOFs on polymer electrospun mats.
Polymer MOF Application Reference
PVA/PAA in
combination
with SiO2
HKUST-1,
MIL-53(Al),
ZIF-8 and
MIL-88B
Hybrid electrospun mats containing PVA/PAA/SiO2 are prepared by electrospinning
followed by solvothermal deposition (HKUST-1 and MIL-53(Al)) or microwave-
induced thermal deposition (ZIF-8 and MIL-88B(Fe)). MIL-53(Al) exhibited improved
adsorption compared to MIL-53(Al) powder.
56
PAN HKUST-1 Fabrication of PAN electrospun nanofibers functionalized with HKUST-1. Next,
hydrothermal synthesis of HKUST-1 in presence of the electrospun material is
performed to increase the activity of absorption.
53
PAN ZIF-8 Electrospun PAN fibers containing 2-methylimidazole is formed. Next, the material is
immersed in ZIF-8 seed solution to form crystals on the surface. In addition, the ZIF-
8@PAN nanofibers showed excellent gas adsorption capabilities.
54
PAN ZIF-8, MIL-53-
NH2
Electrospun PAN fibers are formed, followed by ALD to deposit nucleation sites in
order to form ZIF-8 and MIL-53-NH2 on the nanofiber mats.
55
18
2.4 Electrospinning
2.4.1 Introduction
Nanotechnology is a novel technology which provides new solutions and opportunities
to problems encountered today. Materials containing nanofibers are attractive to solve
numerous problems we encounter today. Nanofibers can be made efficiently by
electrospinning which is a simple and low cost technique 60. Due to the higher porosity
and interconnected pore structure, nanofibers offer a higher flux compared to
conventional filtration materials 60,61. The high flow of reagents and products into the
pores is also useful in heterogeneous catalysis 62. Wound dressing is another hot
topic. Electrospun structures are able to keep bacteria away from the wound due to
their small pores. Furthermore, it is capable of absorbing water which increases the
curing process 63 and due to its high surface area, it can be functionalized to limit
bacterial growth by incorporation of Ag nanoparticles 18. The list of applications is
definitely not restricted to the ones mentioned above.
Many polymers can be employed in electrospinning. The most commonly examined
polymers are polyacrylonitrile (PAN)53, Polyvinyl alcohol (PVA)64, Poly-𝜀-caprolactone
(PCL)21, Poly(vinylpyrrolidone) (PVP)65 and composite materials thereof with inorganic
particles.
2.4.2 Electrospinning mechanism
In general, a polymer solution is subjected to an electrical field (Figure 6). This field
induces positive electric charges in the polymer solution. The coulomb repulsion starts
to increase as the electric field increases by generating more positive charges onto the
surface. A critical point is reached when the coulomb repulsive forces are higher than
the surface tension forces. A charged jet of polymer solution is ejected from the tip of
the Taylor cone. Next, the polymer solution is accelerated towards a collector plate of
opposite charge. In the space between the tip of the needle and the plate the solvent
evaporates as the charged jet reaches the plate. The freshly formed fibers can vary in
size and morphology depending on the polymer solution and the processing
parameters. In a next chapter, the most important parameters are briefly discussed.
19
2.4.3 Electrospinning parameters
In general the parameters can be classified into two groups: solution parameters and
processing parameters51 (Table IV)
Table IV: parameters which have an important influence on the electrospinning process.
Solution parameters Processing parameters
• Concentration
• Molecular weight
• Viscosity
• Surface tension
• Conductivity
• Voltage
• Flow rate
• Injector to collector distance
2.4.4 Solution parameters
2.4.4.1 Concentration
The concentration of the polymer in the solution has an important influence on the
morphology of the fibers. At low concentration, a mixture of beads and fibers is
obtained. As the concentration rises, the structure of the polymer changes from beads
to spindle-like structures to eventually fibers with increasing diameter as the
concentration rises66.
Figure 6: A scheme of a basic electrospinning device51.
20
2.4.4.2 Molecular weight
The molecular weight of the polymer has a significant effect on the rheological and
electrical properties of the polymer solution. It reflects the number of entanglements of
polymer chains in solution and as a result determines the viscosity. Low molecular
weight tends to form beads rather than fibers, therefore high molecular weight
polymers are generally used in electrospinning.67.
2.4.4.3 Viscosity The viscosity is partly accounted by the concentration and the molecular weight as
mentioned before. Low viscosity results in no continuous jet formation while high
viscosity results in problems during the ejection of the charged polymer solution from
the spinneret68. At very high viscosity polymer solutions usually exhibit longer stress
relaxation, which could prevent the fracturing of the jet during electrospinning. 66.
Therefore, the viscosity, in which the electrospinning is done, depends on the polymer
selection and is mostly optimized in order to generate a continuous fiber formation with
a desired diameter.
2.4.4.4 Surface tension
The surface tension is mainly controlled by the choice of the solvent. High surface
tension inhibits the electrospinning process, because of the generation of sprayed
droplets which lead to bead formation. Therefore a solvent has to be selected in which
the surface tension is reduced. 69 In general, lower surface tension also requires lower
electric fields in the electro-spinning process67.
2.4.4.5 Conductivity
Polymer solutions show a certain conductivity which is determined by the combination
of the polymer, the solvent and ionizable salts. It has been reported that higher solution
conductivity leads to a lower fiber diameter. This can be explained by the charge
density: high conductivity solutions in an electric field have more charges than low
conductivity solutions. The higher the charge, the faster the acceleration towards the
oppositely charged plate. As a consequence, the polymer strings are more stretched
what causes the smaller diameter.
21
2.4.5 Processing parameters
2.4.5.1 Applied Voltage
A threshold voltage is required to induce enough charges on the surface of the polymer
solution so that the coulombic forces are stronger than the surface tension forces. In
most cases, a higher voltage between the needle and the collector has two main
effects: the polymers are more stretched causing a smaller diameter and the solvent
is more quickly evaporated. In general, voltage has an influence on the fiber diameter,
but the level of significance varies with the polymer solution and with the distance
between the tip and the collector 70,66.
2.4.5.2 Flow rate
The flow rate is an important parameter as it determines the amount of material that is
pushed through the nozzle which influences the evaporation process and the thickness
of the fibers. A balance has to be made in the flow rate. Low flow rates are beneficial
to complete the evaporation of the solvent. However, higher flow rates ensures that
more polymer is deposited on the surface which results in a larger diameter, but can
also result in beaded fibers due to the lack of proper drying time during the electro-
spinning process71.
2.4.5.3 Injector - collector distance
The distance between the tip and the collector is another factor that can be altered to
control the diameter of the spun fibers. Not surprisingly, the diameter decreases as
soon as the distance increases. The only requirement that has to be taken into account
is the evaporation process of the solvent. A minimum distance is required to give the
fibers a proper time to dry before reaching the collector.
22
2.5 In depth study: PCL solutions for electrospinning
The selection of an optimal PCL solution is important as it determines the viscoelastic
behavior, the surface tension and conductivity of the solution which influence the
electrospinning process. A balance between the viscoelasticity and the surface tension
is required in order to produce a stable Taylor cone. Otherwise, bead-on-a-string
morphology is often observed as the viscosity is too low or the surface tension is too
high. The conductivity influences the fibers as higher conductive polymer solutions are
more stretched during electrospinning and as a result tend to form thinner fibers. In a
first step it is important to select a suitable solvent for poly-𝜀-caprolactone (PCL).
Molecular interactions play an important role in this selection. They can be divided into
three categories: dispersive forces, hydrogen bridge formation and polar forces. TEAS
graphs are able to easily locate the solvent as a function of the three solubility
parameters for a certain polymer. The TEAS graph for PCL is presented in figure 7.
Suitable solvents for PCL are located in the lower right part of the TEAS graph
(indicated by the blue area). These solvents have high dispersive forces, low hydrogen
bonding an low polar forces which is in line with the hydrophobic molecular structure
of PCL (like likes like behavior). Suitable solvents are chloroform (CHCl3), acetone and
dichloromethane (DCM) which are mainly used in this work to dissolve PCL.
In order to obtain small fibers (in line with the aim of this work), it is a possibility to
increase the conductivity of the polymer solution by using a solvent with a higher
dielectric constant. Unfortunately, the more optimal PCL solvents have low polar forces
and as an indirect result have a low dielectric constant. An exceptional solvent is formic
acid which is reported as a poor solvent for PCL , but with a high dielectric constant72,73
able to dissolve PCL after 24 hours of continuous stirring. To overcome the problem
of low conductivity of the PCL solution, binary mixture are used which consists of a
suitable PCL solvent in combination with a solvent with high dielectric constant to assist
the electrospinning. In the past, several combinations have been tried out in literature:
chloroform / methanol, chloroform / tetrahydrofuran, dichloromethane
/dimethylformamide73 and acetic acid / formic acid.
23
Figure 7: TEAS diagram based on solubility-spinnability of PCL in 49 common solvents.
24
25
3 MATERIALS AND METHODS
3.1 Preparation of polymer solution for electrospinning
3.1.1 Chemicals
poly-𝜀-caprolactone (PCL, 80.000 g/mol), chloroform (CHCl3), acetone,
dichloromethane (DCM), N,N-dimethylformamide (DMF) and formic acid (HCOOH)
were purchased from Sigma Aldrich and used as received.
3.1.2 Preparation of polymer solutions
PCL pellets were dissolved in different solvents during this work. The preparation of
the PCL solution was changed according to the selected solvent:
• CHCl3/acetone and DCM/DMF polymer solutions were prepared the day before
electrospinning and stirred overnight. Prior to electrospinning, the solution was
placed in an ultrasonic bath for 15 minutes. Hereafter, the polymer solution was
stirred for 15 minutes to remove remaining air bubbles.
• DCM/HCOOH polymer solutions were prepared to achieve the homogeneous
solution as fast as possible to mitigate acid hydrolysis. Therefore excessive use of
the ultrasonic bath was required. The solution was alternatively stirred and placed
in the ultrasonic bath (15 minutes each) until a homogeneous polymer solution was
obtained.
• MIL-101/PCL and Pt@MIL-101/PCL solutions were obtained by adding MIL-101
or Pt@MIL-101 to the solvent mixture (either Chloroform/acetone, DCM/DMF or
DCM/HCOOH). Before adding PCL, the mixture was placed in an ultrasonic bath
until a homogeneous green dispersed solution was obtained. Once PCL was
added, the same procedure was performed depending on the selected solvent
mixture as described above.
3.1.3 Electrospinning procedure
The polymer solutions were introduced into a 20 mL syringe which was connected by
a Rotilabo-PTFE tube with an internal diameter of 2 mm to an 18 gauge needle (1.270
mm outer diameter, 0.838 mm inner diameter, 3.2 cm length, Fisher Scientific). The
needle was placed through a cupper ring on which the voltage was applied. The
polymer solution was purged through the tubing and the needle by a pumping device.
The setup was placed into a wooden chamber with fume hood. The temperature and
26
relative humidity could not be perfectly controlled in the setup. The average
temperature and relative humidity were 23°C and 30%, respectively.
The needle was cleaned before electrospinning to remove any polymer debris from
previous experiments as that could influence the fiber formation. Once a new droplet
was formed, electrospinning was performed. The fibers were collected on a glass plate
after reaching a steady state situation. Afterwards, the samples were analyzed by
optical microscopy and electron microscopy.
3.1.4 Image analysis
“ImageJ” was used as software tool to analyze the fiber diameter based on images
taken by optical microscopy and electron microscopy. The scale bar of the images was
used to determine the pixel/distance ratio. The fibers were selected randomly in order
to get statistical data which was then expressed into boxplots.
Optical microscopy (OM) images were used to analyze the fibers during the
optimization of the electrospinning parameters as a quick indication tool of the fiber
diameter. It must be noted that the fiber values derived from OM were only used to
visualize trends in fiber diameter rather than absolute values. Afterwards, high
resolution SEM analysis was performed to obtain the absolute values (Figure 8)
Figure 8: Analysis of images based on electron microscopy (A) and optical microscopy (B) by “ImageJ”. The fibers were manually selected at random places as indicated by the yellow lines.
A
B
27
A boxplot is a method for graphically depicting a group of numerical data through
quartiles (Figure 9).
3.2 Preparation of Pt-functionalized MIL-101
3.2.1 Chemicals
Terephthalic acid, hydrated Cr(NO3)3 and hydrogen chloride (HCl) were purchased
from Sigma Aldrich and used without further purification.
3.2.2 Synthesis of MIL-101
MIL-101 was synthesized based on an adapted procedure reported by Edler et al 74.
In a typical reaction, 0,665 g terephthalic acid (4 mmol) and 1,608 g Cr(NO3)3.9H2O (4
mmol) were added to 20 mL of deionized water in a Teflon-lined autoclave. The
autoclave was gradually heated to 210°C during 2 hours in a Nabertherm muffle
furnace and kept at this temperature for 8 hours. Next, the green-colored suspension
underwent a purification procedure to remove any unreacted terephthalic acid. In a first
step, the powder was filtered by a membrane filter (0.45 µm). Hereafter, the as
synthesized material was stirred in DMF for 24 hours at 60°C. Next, MIL-101 was
stirred in 1M HCl overnight at RT, filtered and dried under vacuum at 90°C to obtain
the pure MIL-101 powder.
3.2.3 Synthesis of Pt@MIL-101
The deposition of Pt nanoparticles inside the cages of MIL-101 was performed by ALD
using (methylcyclopentadienyl)-trimethylplatinum (MeCpPtMe3) as Pt source and O3
as reactant at 200°C 75. The depositions were performed in a home built experimental
Figure 9: A randomly generated boxplot (mean: 1µm, standard deviation: 0,1 µm, minimum: 0.5 and maximum 1.5). The mean is indicated by a square, minimum and maximum are indicated by “-“ and the 99% confidence interval lies in between the two “x”.
28
cold-wall ALD chamber. MIL-101 was loaded in a molybdenum sample cup which was
then transferred into the ALD reactor. After loading, MIL-101 was allowed to outgas
and thermally equilibrate for at least 1 h under vacuum. The solid MeCpPtMe3
precursor (99% Strem Chemicals), kept in a stainless steel container, was heated
above its melting point (30 °C), and the delivery line to the chamber was heated to 60
°C. Argon was used as a carrier gas for the Pt precursor. O3 was produced from a pure
O2 flow with an OzoneLab™ OL100 ozone generator (Ozone Services, Burton, BC,
Canada), resulting in an O3 concentration of 175 µg/mL. A static exposure mode was
applied during both ALD half-cycles. The pulse time of the MeCpPtMe3 precursor was
10 s, after which the valves to the pumping system were kept closed for another 20 s,
resulting in a total exposure time of 30 s. The same pulse time and exposure time was
also used for the O317,75. Pt@MIL-101 was obtained after 120 cycles of ALD.
3.3 Catalytic setup and analysis
The hydrogenation reaction occurred in a PARR reactor filled with H2 gas at an
elevated pressure of 6 bar at room temperature (18-23°C). The reactor was loaded
with 70 mL ethanol as solvent, cyclohexene as substrate, dodecane as internal
standard and the catalytic system, either the pure Pt@MIL-101 powder or the Pt@MIL-
101/PCL electrospun fibers. During each test, aliquots were gradually taken out of the
mixture and subsequently analyzed by means of gas chromatography (GC) using a
split injection (ratio 1:17) on a Hewlett Packard 5890 Series II GC with TCD detection
(Santa Clara, CA, USA). The capillary column used was a Restek XTI-5 column
(Bellefonte, PA, USA) with a length of 30 m, an internal diameter of 0,25 µm.
29
4 RESULTS AND DISCUSSION
4.1 Pt@MIL-101 characterization
4.1.1 Nitrogen sorption analysis
Nitrogen sorption measurements were carried out to determine the Langmuir surface
area and the pore volume of the pristine MIL-101 (see literature study § 2.1.2) and the
Pt@MIL-101 powders (see materials and methods §3.2.3). MIL-101 powder had an
average Langmuir surface area of 3185 m²/g (range: 2789 m²/g – 3580 m²/g) and an
average pore volume of 1.41 cm³/g (range: 1.23 cm³/g – 1.53 cm³/g). The higher
surface area compared to the work of Edler et al.74 (2944 m²/g) was the result of the
additional purification steps in this work to remove unreacted terephthalic acid inside
MIL-101 (see materials and methods § 3.2.2).
Table V: The Langmuir surface area and the pore volume of MIL-101 and Pt@MIL-101 measured by N2 sorption analysis.
Sample Langmuir surface area (m²/g) Pore volume (cm³/g)
MIL-101 3580 1.4313
Pt@MIL-101 2907 1.3524
Figure 10: N2 adsorption isotherm of Pt@MIL-101 and the corresponding MIL-101.
30
A slight decrease in the Langmuir surface area and pore volume was noticed after the
embedding of Pt nanoparticles (Table V). Furthermore, the slight pore size reduction
of Pt@MIL-101 suggested the presence of Pt as nanoparticles inside the cages of MIL-
101. It was assumed that Pt was deposited on the coordinately unsaturated sites
(CUSs) formed as a result of the high reaction temperature of 210°C during MIL-101
synthesis17. The shape of the isotherm at low relative pressures indicated the
adsorption of N2 into two different cages which is characteristic for MIL-101 (Figure 10)
(see literature study § 2.1.2) The N2 isotherms could be assigned to type I isotherms
and as a result the surface area was calculated based on the Langmuir theory
(monolayer formation).
4.1.2 XRPD and ICP-OES measurements
The crystalline structure of the synthesized MIL-101 and Pt@MIL-101 were confirmed
by X-ray powder diffraction (XRPD) measurements. Figure 11 shows the XRPD
patterns of the pristine MIL-101 and Pt@MIL-101. The diffraction peaks of the Pt@MIL-
101 were in agreement with the theoretical XRPD pattern of MIL-10174,76 which
indicated that the crystalline structure is preserved after ALD. Inductively coupled
plasma – optical emission spectrometry (ICP-OES) was performed to determine the Pt
loading in Pt@MIL-101 (0.387 mmol Pt/g).
Figure 11: XRPD patterns of pristine MIL-101 and Pt@MIL-101.
31
4.2 Electrospinning of polymer solutions
4.2.1 Introduction
A number of studies have been devoted to the electrospinning of PCL. It was observed
that obtaining bead-free fibers with diameters in the submicron range appeared to be
difficult72,77,78,79. Chloroform was most often applied as solvent for electrospinning PCL
but produced microfibers instead of nanofibers. The goal of our work was to achieve
nanoscale homogeneous fibers as they have larger surface areas which is beneficial
in catalysis. It was known that solvents with a higher dielectric constant result in smaller
fibers due to the excessive stretching and splitting of the polymer jet during
electrospinning.72,80 (see literature study §2.5). Therefore, different solutions were
tested with increasing dielectric constants (Table VI) in an attempt to form nanoscale
fibers.
• Chloroform (CHCl3)/acetone solutions
• DCM/DMF solutions
• DCM/HCOOH solutions
Table VI: Overview of important properties of solvents selected in this work. (*) values at 20°C (*).values at 25°C. (**) 73
solvent 𝜺 * Conductivity (S m-1)** Surface tension (mN m-1)* Viscosity (mPa.s) **
CHCl3 4.8 <1.0 x 10-8 27.16 0.57
CH2Cl2 9.1 4.3 x 10-9 28.12 0.44
Acetone 20.6 5.0 x 10-7 23.3 0.33
DMF 36.7 6.0 x 10-6 35 0.82
HCOOH 58 6.4 x 10-3 37.67 1.78
The ambient parameters, temperature and relative humanity were respectively 18-
23°C and 30-40%. The solution parameters depend upon the selection of the solvent
and the solute (see literature study §2.4.4) . The polymer concentration was kept above
15% (w/v) as a minimum polymer concentration was required to have sufficient chain
entanglements in the polymer (viscoelastic behavior) to provide a stable jet during
electrospinning. The polymer concentration was kept below 20% (w/v) as the high
solution viscosity inhibited the injection of the polymer solution and the stretching of
the polymer jet. Also, solubility problems could be an issue at higher concentrations.
32
Therefore, polymer solutions between 15% and 20% (w/v) were taken into account in
this study which was in accordance with multiple other studies72,77–79 . Next, the
electrospinning parameters were optimized for a given polymer concentration. The
most important parameters are the flow rate, the applied voltage and the injector-
collector distance (ICD) (see literature study §2.4.5).
4.2.2 Electrospinning of PCL from chloroform/acetone solutions
Solutions of 15% and 20% PCL (w/v) in chloroform/acetone (2/1 v/v) were prepared
based on previous observations in the PBM research group. Acetone was added to
the chloroform solution to mitigate the bimodal distribution of fiber diameters which was
often observed by electrospinning polymer solutions in chloroform due to splitting of
the main polymer jet into two unequal parts.
At first, the injector-collector distance (ICD) was optimized. Short ICD (10 cm) led to
wet flattened fibers as the evaporation process could not be completed in the short
period of time. In general, as the ICD increases, the polymer jet has more stretching
time and consequently smaller fibers are formed. However, the disadvantage of a
larger ICD is that a higher voltage is required to obtain the same electric field which
could eventually lead to instabilities in the Taylor cone. Experiments were performed
to find the best ICD by varying the ICD at a constant flow rate of 1 mL/h and a constant
voltage of 15 keV. It was observed that a stable Taylor cone and smooth PCL fibers
were formed at an ICD of 20 cm.
4.2.2.1 Electrospinning of the 15% (w/v) PCL solution
It was observed that fibers with beads, denoted as “bead-on-a-string”, were formed in
the case of the 15% (w/v) polymer solution under most electrospinning conditions even
at a voltage of 17 keV (Figure 12). Important factors that determine the bead-on-a-
string morphology are viscosity, concentration, MW of the polymer, surface tension
and charge density 81. It was reported that polymer solutions with low viscosity (low
concentration and/or low molecular weight) resulted in beaded fibers, because the
viscoelastic forces were not capable of suppressing the Rayleigh instability driven by
surface tension. To mitigate this phenomenon, either the viscosity had to be increased
or the surface tension had to be decreased. Because the surface tension mostly
depends upon the selected solvent, it was chosen to increase the viscosity by
increasing the polymer concentration to 20 % (w/v).
33
4.2.2.2 Electrospinning of the 20% (w/v) PCL solution
As the bead-on-a-string morphology was absent in the case of 20% (w/v) PCL solution
the electrospinning parameters of this solution were further optimized in order to obtain
the smallest fibers with the lowest fiber distribution. In a “change one factor at a time”
approach two series of tests were performed to determine the effect of the flow rate
and the applied voltage in function of the fiber diameter:
• Applied field: the applied field was varied (series: 10, 12.5, 15 and 17.5 keV) at
a constant flow rate of 1 mL/h and an ICD of 20 cm (Figure 13)
• Flow rate: the flow rate was varied (series: 1, 2, 3, 4, 5 mL/h) at a constant
applied field of 15 keV and an ICD of 20 cm.
Changing the voltage at a constant flow rate of 1 mL/h and ICD of 20 cm showed a
minimum in fiber diameter at a voltage of 15 keV (Figure 14 – black boxplots) with a
small fiber distribution (10 keV: 5.24 ± 0.86 µm, 12.5 keV: 4.23 ± 0.91 µm, 15 keV:
2.86 ± 0.75 µm and 17.5 keV: 3.5 ± 1.01 µm). It could be reasoned that at low voltages,
the polymer solution was less stretched due to the weak electric field, resulting in larger
fibers. As the voltage was raised above 15 keV, the Taylor cone disappeared and fibers
were electrospun coming directly from the needle, referred to as a “multijet”. In general,
a “multijet” occurs when more polymer material is ejected from the needle than could
be delivered to the needle which is essentially a disparity between the flow rate
(delivery) and the applied voltage (ejection).
Figure 12: Optical microscopy image to illustrate the bead-on-a-string morphology in the case of the 15% (w/v) PCL solution in chloroform/acetone (2/1). The solution was electrospun at a flow rate of 1mL/h, a voltage of 15 keV and an ICD of 20 cm.
50 µm
34
A B
C D C
Figure 13: Optical microscopy images of 20% (w/v) PCL in chloroform/acetone (2/1). The solution was electrospun at a flow rate of 1 mL/h and an ICD of 20 cm. The applied voltage was varied (A: 10 keV, B: 12.5 keV, C: 15 keV and D: 17.5keV).
50 µm 50 µm
50 µm 50 µm
Figure 14: Statistical analysis of optical microscopy images. The solution was electrospun at an ICD of 20 cm. The applied voltage and flow rate were varied (X-axis). Black boxplots: analysis of figure 13. Red boxplot: analysis of figure 15 (optimised electrospinning conditions). (see materials and methods § 3.1.4 for boxplot interpretation)
35
In general, it was concluded that higher flow rates led to larger fibers. For example,
changing the flow rate from 1 mL/h to 2 mL/h at a voltage of 15 keV and ICD of 20 cm
readily showed a steady increase of fiber diameter from 2.86 ± 0.75 µm to 4.6 ± 0.45
µm based on optical microscopy analysis (Figure 14 – red boxplot) with “ImageJ”
(Figure 15). Thus, it was important to set the flow rate as low as possible to have the
smallest fibers in accordance with the aim of this work. However, due to the high
volatility of chloroform and acetone, at a flow rate of 1 mL/h, the Taylor cone dried out
after 10 minutes of electrospinning due to fast evaporation of the solvent mixture clearly
obstructing the electrospinning process. A flow rate of at least 2 mL/h was required to
make sure the Taylor cone was more stable over time. Furthermore it was observed
that at a flow rate of 2 mL/h, the bimodal distribution of fiber diameters due to the
splitting of the main polymer jet into two unequally distributed sub jets was less
pronounced compared to the fibers electrospun at a flow rate of 1 mL/h. From this
optimization study, it was concluded that the use of chloroform/acetone (2/1) as solvent
in the range of 15% and 20% (w/v) PCL concentration was unable to form fibers in the
submicron range.
Figure 15: Optical microscopy image of a 20% (w/v) PCL solution in chloroform/acetone (2/1). The solution was electrospun at a flow rate of 2 mL/h, a voltage of 15 keV and an ICD of 20 cm.
50 µm
36
4.2.3 Electrospinning of PCL from DCM/DMF solutions
In literature it was mentioned that the addition of N-N-dimethylformamide (DMF) to the
polymer solution drastically increases the electrospinnability77,80 due to the higher
dielectric constant. Therefore, multiple polymer solutions have been tested with
increasing amount of DMF at a constant polymer concentration of 16% PCL (w/v)
conform several studies77,80. However, PCL cannot be dissolved in pure DMF in
contrast to other polymer materials like polyacrylonitrile (PAN) which were recently
electrospun in pure DMF solutions52. To counteract this, DMF was mixed with a more
suitable solvent, dichloromethane (DCM, see literature study §2.5). Before
electrospinning, solubility tests were performed with increasing DMF volume ratio
(Table VII). It was concluded that the 16% (w/v) PCL concentration was able to be
dissolved in less than 24 hours in up to 60% (v/v) DMF.
Table VII: Solubility test of different DCM/DMF volume ratios. (+): soluble, (±): partly
soluble, (−): insoluble.
DCM/DMF ratios (v/v) 1/0 3/1 2/3 1/4 0/1
Solubility + + + ± −
Next, the effect of the increasing DMF volume fraction on the fiber diameter was
analyzed by electrospinning the 16% PCL (w/v) DCM/DMF solutions (1/0, 3/1 and 2/3
v/v). The electrospinning conditions were optimized and it was concluded that a stable
electrospinning condition was found at a flow rate of 1 mL/h, a voltage of 15 keV and
an ICD of 20 cm. Remarkably, a much more stable Taylor cone was formed even at a
flow rate of 1 mL/h once DMF was added which is in contrast to the chloroform/acetone
(2/1) solution. Pure DCM solutions resulted in large fibers with a bimodal distribution
(5.1 ± 1.53 µm), comparable with the chloroform/acetone (2/1) solution mentioned
earlier (Figure 16). However, once DMF was added to DCM, smaller fibers were
formed with a much smaller fiber distribution (Figure 17). Interestingly, the fiber
diameter decreased with increasing amount of DMF, respectively from 2.00 ± 0.36 µm
to 1.49 ± 0.23 µm when the volume ratio was changed from 3/1 to 2/3 DCM/DMF. It
could be thus concluded that the 16% (w/v) PCL DCM/DMF (2/3) solutions resulted in
the smallest fibers in this section.
37
Figure 16: Optical Microscopy images of electrospun fibers of 16% PCL (w/v) in DCM/DMF (A: 1/0 v/v, B: 2/3 v/v). Solutions were electrospun at a flow rate of 1mL/h, a voltage of 15 keV and an ICD of 20 cm.
A B
50 µm 50 µm
Figure 17: Optical microscopy analysis of electrospun fibers processed from different DCM/DMF (v/v) ratios at a constant polymer concentration of 16% (w/v). Solutions were electrospun at a flow rate of 1mL/h, a voltage of 15 keV and an ICD of 20 cm.
38
4.2.4 Electrospinning of PCL from DCM/HCOOH solutions
It was anticipated that the addition of formic acid (HCOOH) to DCM solution results in
even smaller fibers compared to the DCM/DMF solution due to the even higher
dielectric constant. It was reported in literature that a 20% (w/v) PCL in pure HCOOH
solutions resulted in nanoscale fibers73. Solubility tests (Table VIII) had to be
conducted as HCOOH was considered a poor solvent to dissolve PCL (literature study
– TEAS graph). Two solutions were tested: 20% (w/v) PCL in DCM/HCOOH (1/1 v/v)
and 20% (w/v) PCL in pure HCOOH.
The solubility tests in pure HCOOH demonstrated that it was possible to dissolve PCL
after 24 hours of continuous stirring. It could be predicted that the molecular weight of
the polymer decreased during the sample preparation due to acid hydrolysis of the
PCL ester functionality. Size Exclusion Chromatography (SEC) analysis confirmed that
the molecular weight of PCL decreased by 50% (80,000 to 40,000 MW) after 24 hours
of stirring in pure HCOOH. The obtained 20% (w/v) PCL HCOOH solution was
electrospun at a flow rate of 1 mL, 20 keV and ICD of 20 cm. As the fibers were too
small to be visualized by optical microscopy, SEM images were taken to evaluate the
fiber diameter, 0,32 ± 0,14 µm. Although nanoscale fibers could be obtained with the
pure HCOOH solution, it was found that the electrospun PCL material was too brittle
to be handled and therefore it was chosen to omit this solution in the further study.
A way to overcome the degradation of polymer was by adding DCM to HCOOH in order
to decrease the dissolving time and as a result lower the degradation of PCL. The
DCM/HCOOH (1/1) solution was chosen as golden mean, because higher HCOOH
fractions would cause more degradation and higher DCM fractions would cause larger
fibers due to the lower dielectric constant of the solution.
Table VIII: Solubility tests. (+): soluble, (±): partly soluble, (-): insoluble. The 20% (w/v) PCL in DCM/HCOOH (1/1) was completely dissolved after 4 hours of stirring (*) while the 20% PCL in pure HCOOH was dissolved after 24 hours (**).
DCM/HCOOH ratios (v/v) 1/1 0/1
Solubility + (*) + (**)
39
At first, the polymer concentration was set to 20% (w/v) in accordance to the previous
tests in pure HCOOH. As anticipated, the addition of DCM improved the dissolution of
PCL dramatically and after 4 hours of mixing the polymer solution was obtained. SEC
analysis showed that there was no significant degradation of the electrospun PCL
fibers. During optimization, it was noted that a higher voltage (17 keV) was required
compared to the previously discussed solutions (Chloroform/acetone and DCM/DMF)
at an ICD of 20 cm and a flow rate of 1 mL/h. It was mentioned in literature that solvents
with a high dielectric constant require a higher applied voltage in order to obtain a
stable Taylor cone72,73.
Next, the polymer concentration was reduced to 16% PCL (w/v) to make a fair
comparison to the optimized 16% PCL (w/v) DCM/DMF (2/3). The sample preparation
could be performed in less than 2 hours due to the lower PCL concentration. Best
electrospinning conditions were observed at a flow rate of 1 mL/h, a voltage of 17 keV
and an ICD of 20 cm. Optical microscopy image analysis showed a slight decrease in
fiber diameter, from 1.13 ± 0.12 µm to 1.03 ± 0.11 µm, when the concentration was
changed from 20% to 16% PCL (w/v). Furthermore, it could be observed from figure
18 that electrospinning PCL DCM/HCOOH (1/1) solutions led to small fibers with
narrow fiber distribution. Also, electrospinning could be performed for more than 2
hours without destabilization of the Taylor cone. Finally, it must be noted that the
resulting PCL mats were much more sturdier compared to the mats produced from
pure HCOOH solution.
20 µm 20 µm
Figure 18: Optical Microscopy images of electrospun fibers of PCL in DCM/HCOOH (1/1). Solutions were electrospun at a flow rate of 1mL/h, a voltage of 17 keV and an ICD of 20 cm.
40
4.2.5 SEM analysis of optimized electrospun PCL fibers
It was mentioned in §3.1.4 that optical microscopy images were used to indicate trends
during electrospinning optimization as it was a quick way to do the analysis. It was
already observed in this study that the examination of small fibers in the micron range
(1 – 1,5 µm) was difficult to analyze by optical microscopy due to the limited resolution,
because it was hard to select the boundaries of the fibers.
Therefore, subsequent scanning electron microscopy (SEM) analysis was performed
to determine the actual fiber diameter of the most optimal fibers formed throughout this
work. SEM images were taken of the 16% PCL (w/v) DCM/DMF (2/3) and the 16%
PCL (w/v) DCM/HCOOH (1/1) electrospun fibers (Figure 19). These SEM images were
also analyzed by “ImageJ” (Figure 20). It was concluded that both solutions resulted in
submicron scale fibers which was in contrast to previous conclusions derived from
optical microscopy analysis in §4.2. The 16% PCL (w/v) DCM/DMF (2/3) solution led
to a fiber diameter of 700 ± 127 nm while the 16% (w/v) DCM/HCOOH (1/1) solution
led to a fiber diameter of 450 ± 120 nm which is below the values observed in optical
microscopy, respectively 1.49 ± 0.23 µm and 1.03 ± 0.11 µm. The overestimation
(ratio EM/OM in Table IX) of optical microscopy appeared to be similar in both samples,
indicating a similarity in experimental error. As a consequence, optical microscopy was
only used to indicate trends, while electron microscopy was used to derive the actual
fiber diameter due to its superior resolution.
Table IX: Overview of the comparison between optical microscopy (OM) and electron
microscopy (EM) analysis for the optimized electrospun fibers. Same samples were
used for OM and EM analysis.
Polymer solution OM EM EM/OM
DCM/DMF (2/3) 1.49 ± 0.23 µm 700 ± 127 nm 47%
DCM/HCOOH (1/1) 1.03 ± 0.11 µm 450 ± 120 nm 44%
A high degree of bended fibers was observed in the SEM image (Figure 19) of the
fibers derived from the DCM/DMF solution. This could probably be explained by the
sample preparation before SEM analysis. The fibers had to transferred from a glass
plate to the carbon tape which induced bending of the fibers before analysis. This
assumption was confirmed by optical microscopy images before sample preparation
as straight fibers were observed (Figure 16). This effect was not seen in the case of
41
the DCM/HCOOH solutions, because a much thicker layer of fibers was prepared on
the glass plate. The higher density of fibers probably resisted the forces applied during
sample preparation of SEM analysis.
Figure 19: SEM images of electrospun PCL made from a 16% PCL (w/v) DCM/DMF (2/3) (A) and a 16% PCL(w/v) DCM/HCOOH (1/1) (B) to determine the fiber diameter under the optimized electrospinning conditions (see § 4.2.3 and § 4.2.4).)
A B
Figure 20: SEM image analysis by ImageJ of the optimized 16% PCL (w/v) DCM/DMF (2/3)
and the 16% PCL (w/v) DCM/HCOOH (1/1).
B
42
4.3 Effect of MIL-101 addition on the electrospinning process
4.3.1 Introduction
The electrospinnability using the optimized polymer solution in § 4.2 was tested with
the addition of MIL-101 to the polymer solution. As MIL-101 are nanocrystals, it was
anticipated that the addition had an influence on the viscosity, surface tension and
conductivity of the solution. Therefore, 20% PCL (w/v) chloroform/acetone (2/1), 16%
PCL (w/v) DCM/DMF (2/3) and 16% PCL (w/v) DCM/HCOOH (1/1) were electrospun
after addition of MIL-101. The process was optimized to obtain fibers with the most
optimal properties concerning Taylor cone stability and fiber morphology.
4.3.2 PCL/MIL-101 fibers from chloroform/acetone solutions
MIL-101 was added to the 20% PCL (w/v) chloroform/acetone (2/1) solution to obtain
a 1/20% MIL-101/PCL (w/v) chloroform/acetone (2/1) solution which would in theory
result in 5% (w/w) MIL-101 distributed in the electrospun PCL fibers (see materials and
methods §3.1.2). It was noted that minor additions of MIL-101 severely influenced the
electrospinnability. This was mainly due to accumulation of MIL-101 at the needle
entrance in function of time which eventually blocked the flow of polymer through the
needle. The voltage and flow rate were independently increased in multiple tests (up
to 18 keV and 5 mL/h) in an attempt to prevent the MIL-101 accumulation. Yet the
accumulation seemed to be inevitable after several minutes of electrospinning (Figure
21).
Figure 21: Illustration of the MIL-101 accumulation at the needle entrance in the case of the chloroform/acetone solutions.
43
It could be assumed that the electrospinning obstruction was caused by the low boiling
point of chloroform and acetone (respectively 61°C and 56°C). As a result, the solvent
was quickly evaporated at the needle entrance leaving MIL-101 behind. Over time the
MIL-101 quickly aggregated as can be observed (Figure 21).
The morphology of the optimized PCL fibers and the 5% (w/w) MIL-101 electrospun
PCL fibers after one minute of electrospinning is depicted (Figure 22). Hereafter,
electrospinning was inhibited due to blocking of MIL-101 at the needle entrance. The
broad variation in fiber diameter of the MIL-101/PCL fibers as assessed by the optical
microscopy image (Figure 22) was the result of an unstable Taylor cone which was
formed on top of the accumulating MIL-101. It was therefore concluded that it was
impossible to form MIL-101 embedded PCL fibers from this solution.
4.3.3 PCL/MIL-101 fibers from DCM/DMF solution
MIL-101 was added to the 16% PCL (w/v) in DCM/DMF (2/3) to form 5%, 10% and
20% MIL-101 (w/w) distributed in the PCL electrospun fibers. It was quickly concluded
that all solutions, even the one with the high loading of 20% (w/w) MIL-101 were able
to be electrospun without the accumulation of MIL-101 at the needle as reported in §
4.3.2. As a high loading of MIL-101 will be beneficial in the upcoming catalysis, we
opted to only work with this high loading in future experiments.
A B
Figure 22: Optical microscopy images of the pristine PCL fibers formed by the 20% PCL (w/v) chloroform/acetone (2/1) solution (A) and the solution prepared under (A) with the addition of 1 % (w/v) MIL-101 (B). Both solutions were electrospun at a flow rate of 2 mL/h, a voltage of 15 keV and an ICD of 20 cm.
50 µm 50 µm
44
In order to form 20% (w/w) MIL-101 in the PCL scaffold by electrospinning, 4% (w/v)
MIL-101 and 16% (w/v) PCL were added to a DCM/DMF (2/3) solution. (note: 0.2 g
MIL-101 and 0.8 g PCL in 5 mL DCM/DMF (2/3)). This solution will be referred to as
4/16% MIL-101/PCL (w/v) DCM/DMF (2/3).
Electrospinning of this solution resulted in a skewed, oscillating Taylor cone rather than
the original Taylor cone observed at the optimized electrospinning parameters of the
pure PCL solution in § 4.2.3 (flow rate: 1 mL/h, voltage: 15 keV and ICD: 20 cm).
Despite this effect, optical microscopy images (Figure 23) showed that homogeneous
fibers were formed at these electrospinning conditions with a fiber diameter of 1.55 ±
0,3 µm which is in reasonable comparison with the pristine PCL fibers formed in §
4.2.3. Unfortunately, after a few minutes of electrospinning, a small droplet was
consequently ejected from the Taylor cone. In an attempt to mitigate the droplet
formation, the electrospinning parameters were slightly optimized.
A B
Oscillating Taylor cone Stable Taylor cone Multijet
Figure 23: Illustration of the different Taylor cone observed during electrospinning and the influence on the morphology of the electrospun fibers. The 4/16 % MIL-101/PCL (w/v) DCM/DMF (2/3) solution was electrospun at the optimised electrospinning conditions (oscillating Taylor cone) (A) and at a voltage of 18 keV (multijet) and flow rate of 1 mL/h (B).
20 µm 20 µm
45
At a fixed ICD of 20 cm, the voltage and the flow rate were varied. As the voltage was
increased from 15 keV to 18 keV at a flow rate of 1 mL/h, the morphology of the fibers
changed drastically due to the occurrence of a “multijet”. The fiber diameter was 3,65
± 1,81 µm as measured by optical microscopy.
The best electrospinning condition without droplet formation was found at a flow rate
of 1.5 mL/h, a voltage of 16 and an ICD of 20 cm at the cost of a high fiber diameter of
2.60 ± 0.42 µm, measured by optical microscopy (Figure 24). The higher flow rate was
required to mitigate the formation of a “multijet” at a voltage of 16 keV which
automatically resulted in larger fibers.
4.3.4 PCL/MIL-101 fibers from DCM/HCOOH solutions
As it was possible to obtain 20% (w/w) MIL-101 distributed in the PCL scaffold in the
case of the DCM/DMF (2/3) solution, it was opted to use the same high loading of MIL-
101 as in the case of the DCM/HCOOH (1/1). Therefore, 4% (w/v) MIL-101 and 16%
(w/v) PCL were added to a DCM/HCOOH (1/1) solution. (note: 0.2 g MIL-101 and 0.8
g PCL in 5 mL DCM/HCOOH (1/1)). This solution will be referred to as 4/16% MIL-
101/PCL (w/v) DCM/HCOOH (1/1)
Figure 24: Optical microscopy image of a 4/16% MIL-101/PCL (w/v) DCM/DMF (2/3) solution. The solution was electrospun at a flow rate of 1.5 mL/h, a voltage of 16 keV and an ICD of 20 cm.
20 µm
46
It was observed that once the voltage was applied, droplets were formed based on
classical Rayleigh instability without the formation of any fibers. In general, the
instability is caused by a low electrical field which is easily countered by increasing the
applied voltage. However, even when the voltage was increased to 25 keV, the droplet
formation occurred which was enough evidence that it was impossible to form MIL-101
electrospun PCL fibers from the DCM/HCOOH (1/1) solution.
From the experiments it was concluded that once MIL-101 was added to the polymer
solution, the behavior of the electrospinning process changed drastically as there was
a tendency to form droplets. In general, a stable polymer jet is observed once an
equilibrium is noted between the viscoelastic forces in the polymer solution and the
electrostatic forces caused by the applied electric field. As 4/16 % MIL-101/PCL (w/v)
DCM/HCOOH (1/1) solutions were unable to be electrospun, tests were performed to
examine the viscoelastic properties of the various solutions. It was observed that the
DCM/HCOOH (1/1) polymer solution showed a dramatic increase of viscosity,
respectively from 2.65 to 5.52 Pa.s at low shear rate (�̇� = 1 s-1) after addition of MIL-
101 to the polymer solution. The difference in viscosity in the case of DCM/DMF (2/3)
solution after addition of MIL-101 was less pronounced (from 2,31 to 2,94 Pa.s at low
shear rate). It was assumed that the MIL-101/PCL DCM/HCOOH (1/1) polymer
solution resisted deformation caused by the applied electric field which could be
partially explained by the higher viscosity. However, surface tension tests and
conductivity tests should be conducted in future studies to show the effect of MIL-101
addition on these solution parameters as they also contribute to the electrospinning
process. Unfortunately, these tests could not be performed in this work due to the
limited available MIL-101 powder.
Table X: Viscosity of polymer solution applied for electrospinning as studied by
rheology.
Polymer solution Viscosity (Pa.s) at �̇� = 1 s-1
16% PCL (w/v) DCM/HCOOH (1/1) 2,6543
16/4 % MIL-101/PCL (w/v) DCM/HCOOH (1/1) 5,4145
16% PCL (w/v) DCM/DMF (2/3) 2,3055
16/4 % MIL-101/PCL (w/v) DCM/DMF (2/3) 2,9410
47
4.4 Characterization of the Pt@MIL-101/PCL electrospun fibers
4.4.1 Preparation of Pt@MIL-101/PCL fibers
In order to produce 20% (w/w) Pt@MIL-101 distributed in the PCL matrix, 4% Pt@MIL-
101 (w/v) and 16% PCL (w/v) were added to the DCM/DMF (2/3) solution. Interestingly,
the best electrospinning conditions were observed at a voltage of 15 keV and a flow
rate of 1 mL/h without the droplet formation. This is in contrast with § 4.3.3. as it was
observed that small droplets were ejected from the Taylor cone by electrospinning the
MIL-101/PCL polymer solution at these electrospinning conditions. It was assumed
that the addition Pt@MIL-101 increased the conductivity of the solution as Pt was
present (Pt: 9,4 x 106 S/m) compared to the addition of MIL-101 to the polymer solution.
However, this could not be examined by conductivity tests as too much Pt@MIL-101
would be consumed. The 4/16 % Pt@MIL-101/PCL (w/v) DCM/DMF (2/3) solution
could be electrospun for 2 hours to achieve the 20% Pt@MIL-101 (w/w) electrospun
PCL scaffold (Figure 25) without the requirement to interfere with the electrospinning
process. The composite material was collected on baking paper to easily collect them
after electrospinning.
Figure 25: Picture of the Pt@MIL-101/PCL electrospun mat.
48
4.4.2 SEM analysis of Pt@MIL-101/PCL fibers
The Pt@MIL-101/PCL fibers were analyzed by SEM to determine if Pt@MIL-101 was
present at the surface of the electrospun PCL matrix.
SEM analysis (Figure 26) revealed that MIL-101 crystals were present at the surface
of the PCL fibers throughout the electrospun material, including clusters of Pt@MIL-
101 anchored on the fibers. which indicated that at least partially active Pt sites were
accessible for catalytic reactions. However, since SEM is a surface analysis technique
it was not possible to determine the percentage of Pt@MIL-101 that was entirely
surrounded by the PCL scaffold. The Pt@MIL-101/PCL fibers showed a fiber diameter
of 780 ± 165 nm which is in comparison with the pristine PCL fibers obtained in § 4.2.5.
Figure 26: (A-D) SEM images of Pt@MIL-101/PCL fibers. (B) cluster of Pt@MIL-101 trapped in the PCL scaffold.
B
C D
A
49
4.4.3 SEM-EDX mapping of Pt@MIL-101/PCL fibers
SEM-EDX mapping was performed to study the distribution of Pt throughout the
electrospun material. Also, the distribution of MIL-101 could be determined by
searching the characteristics X-rays of chromium as this metal is present inside MIL-
101 (see literature study § 2.1.2)
It was observed that the characteristic L𝛼 Pt X-rays were measured throughout the
electrospun Pt@MIL-101/PCL fibers (Figure 27), indicating that Pt nanoparticles were
homogeneously present in the Pt@MIL-101/PCL electrospun material. Small clusters
of Pt@MIL-101 were also observed by analyzing the characteristic wavelengths of Cr
which is in resemblance with SEM analysis in § 4.4.2. From this study it could be
derived that a homogeneous dispersion of Pt@MIL-101 in the polymer solution was
obtained before electrospinning which is essential to form a homogeneous Pt
distribution.
C K
Cr K Pt L
Figure 27: SEM EDX mapping of electrospun Pt@MIL-101/PCL fibers.
50
4.4.4 XRPD analysis of Pt@MIL-101/PCL fibers
XRPD analysis was performed to analyze the crystallinity of Pt@MIL-101 embedded
in the PCL fibers after electrospinning. The crystallinity is an important parameter as it
has an influence on the catalytic performance.
The XRPD patterns of pure PCL fibers and Pt@MIL-101/PCL fibers were compared
(Figure 28). It was observed that the characteristic peaks of MIL-101 were present in
the Pt@MIL-101/PCL electrospun fibers. Furthermore some characteristics PCL peaks
were noticed at an angle of 21, 24 and 30 due to the semi-crystalline behavior of PCL.
4.4.5 Thermal analysis of Pt@MIL-101/PCL fibers
4.4.5.1 Thermogravimetric analysis
Thermogravimetric analysis (TGA) was performed on the Pt@MIL-101/PCL fibers and
compared with pure PCL fibers, electrospun under the same reaction conditions (flow
Figure 28: XRPD patterns of pristine PCL fibers and PCL/MIL-101 fibers produced under the same electrospinning conditions (see § 4.2.3 and § 4.4.1, respectively).
51
rate 1 mL/h, voltage 15 keV and ICD 20 cm) to examine the influence of Pt@MIL-101
on the PCL scaffold material in terms of thermal stability.
The thermogram of the 20% (w/w) Pt@MIL-101 electrospun PCL fibers (Figure 29)
showed that the onset of degradation occurred at a lower temperature (300°C)
compared to the pristine PCL fibers (350°C), indicating that the addition of Pt@MIL-
101 negatively influenced the thermostability of PCL. It was also observed that the
degradation of PCL occurred over a broader temperature range (300-400°C). At a
temperature of 400°C, approximately 20 % of the weight remained which could be
accounted Pt@MIL-101 Above 500°C Pt@MIL-101 was fully degraded with an
inorganic residue of Pt and Cr.
4.4.5.2 Differential Scanning Calorimetry analysis
Differential Scanning Calorimetry (DSC) analysis was performed to determine the
influence of Pt@MIL-101 onto the melting point and the crystallinity of the PCL
electrospun fibers.
Figure 29: TGA analysis of Pt@MIL-101/PCL fibers and pristine PCL fibers under N2 atmosphere. The heating rate was 10°C/min.
52
TABLE XI. DSC analysis of Pt@MIL-101/PCL electrospun fibers and pristine PCL fibers.
The Tg was measured based on the inflection point at a heating rate of 20°C/min. The degree
of crystallinity was calculated from the 2nd heating curve (100% crystallinity: heat of fusion
139,5 J/g).
Sample Tg (°C) Crystallinity
Pure PCL fibers - 63,79 37,3%
Pt@MIL-101/PCL fibers - 61,04 28,7%
The Pt@MIL-101/PCL fibers showed a slight decrease of glass transition temperature
(Tg) compared to the pure PCL fibers, respectively, - 61°C and -63.79°C (Table XI). It
was concluded that the small difference in Tg between both samples was insufficient
to make general assumptions concerning the influence of the Pt@MIL-101 on the
molecular mobility of the polymer segments in PCL. A decrease of crystallinity was
observed (from 37.3 to 28.7%) when Pt@MIL-101 was blended in the polymer scaffold,
indicating that Pt@MIL-101 has an impact on the semi-crystalline property of PCL. It
was assumed that the change in crystallinity has an impact on the mechanical
properties of the material, but was not examined in this work. At last, it was shown that
there was no considerable difference between the melting point of Pt@MIL-101 and
the melting point of the pristine PCL fibers, setting the working limit to approximately
60°C in the case of the electrospun material (Figure 30).
Figure 30: DSC analysis of the Pt@MIL-101 and pristine PCL fibers. The heating rate was 20°C (second heating run).
53
4.5 Catalysis with Pt@MIL-101/PCL electrospun fibers
4.5.1 Introduction
The hydrogenation of cyclohexene was used as proof of concept to examine the
catalytic activity and accessibility of Pt@MIL-101 in the electrospun PCL fibers. Finally,
the obtained results were compared with the catalytic performance of pure Pt@MIL-
101 powder.
4.5.2 Pt content in Pt@MIL-101/PCL electrospun fibers
It must be noted that the Pt content needed to be exactly determined before catalysis
could be performed. In theory, the Pt content could be directly calculated based on the
fact that 20% Pt@MIL-101 (w/w) was present in the electrospun PCL fibers, assuming
that a homogeneous Pt@MIL-101/PCL mixture could be obtained during sample
preparation. SEM-EDX mapping already revealed that the Pt nanoparticles were
homogeneously distributed throughout the composite material which assisted this
assumption. Previous ICP-OES results had already showed that the Pt loading in MIL-
101 was 0,387 mmol Pt / g. As only 20% (w/w) of the electrospun fibers is Pt@MIL-
101, it could be concluded that the theoretical Pt loading in the electrospun material
was:
0.387 mmol Pt / g x 20 % = 0.0774 mmol Pt/g
As it was not possible to analytically derive the Pt loading throughout the material
without destroying it, it was chosen to electrospin two Pt@MIL-101/PCL mats from the
same 4/16 % Pt@MIL-101/PCL (w/v) DCM/DMF (2/3) solution. The first one would be
used to perform the catalytic tests while the other one would be used to determine the
Pt loading by inductively coupled plasma – optical emission spectroscopy (ICP-OES).
Two randomly chosen samples of the second mat were digested by aqua regia and
analyzed by ICP-OES.
TABLE XII. ICP-OES measurements of two random samples of the electrospun composite
material.
THEORY SAMPLE 1 SAMPLE 2
Pt Loading 0.0774 mmol/g 0.0779 mmol/g 0.0784 mmol/g
It was concluded that both samples matched the theoretical value perfectly, indicating
that a homogeneous composite material was produced.
54
Chemisorption experiments were conducted to determine the amount of accessible Pt
sites in the electrospun composite material. H2 gas was used as chemisorption gas as
it shows great affinity for metallic Pt. It was already derived that the theoretical Pt
loading in the electrospun fibers was 0.0774 mmol/g based on ICP-OES
measurements.
TABLE XIII. Chemisorption experiments with H2 gas at 1 bar.
THEORY Chemisorption
0.0774 mmol / g 0.05741 mmol / g
Based on these measurements (Table XIII) it could be derived that 75% of the Pt
nanoparticles in the electrospun material are readily available to adsorb H2. Essentially,
this means that 25% of the Pt sites were not able to act as a hydrogenation catalyst,
because the transport of H2 to these active sites was inhibited.
4.5.3 Catalytic performance of Pt@MIL-101/PCL
The amount of Pt of the first Pt@MIL-101/PCL electrospun fiber (0.3797 g) was 0.0294
mmol based on the previously discussed Pt loading. The amount of mmol cyclohexene
(substrate) and dodecane (internal standard: IS) were calculated to have a substrate/Pt
ratio of 400 and an IS/Pt ratio of 200. The Pt@MIL-101/PCL electrospun material,
cyclohexene, dodecane and 70 mL ethanol were put in the reaction vessel of the Parr
reactor. The catalytic system was simply added without any support in the reactor itself.
In the case of Pt@MIL-101 as catalytic system, the same amount of 0.0294 mmol Pt
was used to compare both catalysts. The temperature was set at room temperature
and the pressure at 6 bar during catalysis. DSC analysis already showed in § 4.4.5
that the electrospun Pt@MIL-101/PCL fibers showed a melting point of 60°C. In theory,
the composite material should be stable be stable in these reaction conditions.
Reusability tests were performed by utilizing the same Pt@MIL-101/PCL electrospun
fibers during multiple catalytic runs. After each run, the fibers were washed several
times with ethanol and manually dried with paper. Finally, the composite material was
kept under vacuum for at least an hour, before the next run was performed
55
The conversion of cyclohexene to cyclohexane during 4 consecutive runs are
summarized in Figure 31. The TON number was calculated by dividing the amount of
mmol product (cyclohexane) by the number of active sites at the end of the reaction
while the TOF number was determined by dividing the TON number by the reaction
time (min) after 10 minutes of reaction time (Table XIV).
Table XIV: The turnover number (TON), turnover frequency (TOF) and leaching percentage
of Pt for each catalytic test. TOF was calculated after 10 minutes of reaction time*. The
leaching of Pt was lower than the detectable limit by XRF**
TON TOF (min-1)* Leaching of Pt (%)**
Pt@MIL-101 powder 360 16,91 <0,05
Pt@MIL-101/PCL fibers RUN 1 311 8,31 <0,05
Pt@MIL-101/PCL fibers RUN 2 343 9,68 <0,05
Pt@MIL-101/PCL fibers RUN 3 374 8,90 <0,05
Pt@MIL-101/PCL fibers RUN 4 354 8,1 <0,05
The catalytic tests showed that Pt@MIL-101/PCL exhibited full conversion of
cyclohexene after 90 minutes of reaction time, while Pt@MIL-101 powder showed full
conversion after only 50 minutes. The higher kinetics of the Pt@MIL-101 powder could
also be derived from the higher TOF number of Pt@MIL-101 compared to the TOF
number of the Pt@MIL-101/PCL electrospun fibers. Despite the fact that the same
0
10
20
30
40
50
60
70
80
90
100
0 20 40 60 80 100 120
Co
nv
ersi
on
(%
)
Reaction Time (minutes)
Pt@MIL-101/PCL RUN 1
Pt@MIL-101/PCL RUN 2
Pt@MIL-101/PCL RUN 3
Pt@MIL-101/PCL RUN 4
Pt@MIL-101 powder
Figure 31: Conversion of cyclohexene to cyclohexane catalyzed by Pt@MIL-101/PCL fibers and by Pt@MIL-101 powder. Multiple catalytic tests were performed with the same electrospun material.
56
amount of Pt active sites were present in both catalytic systems, it could be concluded
that Pt@MIL-101/PCL electrospun fibers showed a decrease in kinetics. As described
previously, the chemisorption experiments in §4.5.2 showed that approximately 25%
of the Pt sites are inaccessible for H2. This might explain the difference in the calculated
TOF values for both catalysts.
It must be noted that the variation in the TON number during the multiple runs could
be explained by difficult sample extraction from the Parr reactor. Due to the high
pressure of 6 bar, the unattached electrospun material probably blocked the needle
outlet. However, similar TOF values were obtained during these additional runs,
demonstrating that the fibers could be reused for multiple runs with only a slight
decrease in activity over the performed runs.
After catalysis, the electrospun fibers could be easily removed from the reaction
medium which was analyzed afterwards by X-ray fluorescence (XRF) to determine the
leaching of Pt and Cr. XRF measurements revealed that no particular leaching of both
Pt and Cr was observed (Table XIV), showing the strong embedding of the Pt@MIL-
101 material in the PCL fibers.
4.6 Analysis of Pt@MIL-101/PCL fibers after catalysis
Additionally, XRPD measurements, SEM analysis and SEC analysis were carried out
on the composite material after 4 runs of catalysis to determine the influence of the
reaction medium on the catalytic material. The XRPD pattern showed that the
characteristic crystalline patterns of MIL-101 (compare with Figure 10) was preserved
during at least 4 multiple runs (Figure 32). It was also concluded that the semi-
crystallinity of PCL was still present after catalysis. SEM analysis showed that Pt@MIL-
101 crystals were still present at the surface of the composite material (Figure 33) and
that the integrity of the fibers was preserved without particular deficiencies in the
structure.
57
Figure 32: XRPD pattern of Pt@MIL-101/PCL electrospun fibers after 4 runs of catalysis compared with the XRPD pattern before catalysis.
Figure 33: SEM analysis of Pt@MIL-101/PCL electrospun fibers after 4 runs of catalysis.
58
59
5 CONCLUSION
In this work, 16% (w/v) PCL DCM/DMF (2/3) and 16% (w/v) PCL DCM/HCOOH (1/1)
were successfully processed by electrospinning to achieve PCL fibers in the submicron
range (0.7 and 0.45 µm respectively) as analyzed by SEM analysis. It was observed
that the addition of MIL-101 to the polymer solution influenced the electrospinning
process as it changed the solution parameters, making it impossible to electrospin the
MIL-101/PCL DCMHCOOH (1/1) polymer solution. Electrospinning of MIL-101/PCL
and Pt@MIL-101/PCL in DCM/DMF (2/3) as solvent could be performed under stable
electrospinning conditions. SEM images of the Pt@MIL-101/PCL electrospun fibers
showed that Pt@MIL-101 was partially present at the surface of the PCL scaffold. ICP-
OES measurement revealed that the actual Pt loading was exactly the same as
theoretically predicted from which it could be derived that the Pt nanoparticles were
homogeneously present throughout the electrospun material. However, from
chemisorption analysis with H2 it was concluded that 25% of the Pt@MIL-101 crystals
were inaccessible for catalysis. Catalytic tests were conducted to examine the
performance of the Pt@MIL-101/PCL material compared to pure Pt@MIL-101 powder
with the same Pt loading. It was observed that full conversion with Pt@MIL-101/PCL
fibers was reached after 90 minutes of reaction time while in the case of Pt@MIL-101
powder full conversion occurred after 50 minutes. Reusability tests showed that the
activity of Pt@MIL-101/PCL electrospun fibers slightly decreases without detectable
leaching of Pt@MIL-101 into the reaction medium. The Pt@MIL-101/PCL fibers were
examined after catalysis by means of SEM and XRPD analysis, showing that the fiber
morphology and the crystallinity of Pt@MIL-101 were preserved after 4 catalytic runs.
60
61
6 REFERENCES
1. Li, H., Eddaoudi, M., O’Keeffe, M. & Yaghi, O. M. Design and synthesis of an exceptionally stable and highly porous metal-organic framework. Nature 402, 276–279 (1999).
2. Li, B., Wen, H.-M., Zhou, W. & Chen, B. Porous Metal–Organic Frameworks for Gas Storage and Separation: What, How, and Why? J. Phys. Chem. Lett. 5, 3468–3479 (2014).
3. Kayal, S., Sun, B. & Chakraborty, A. Study of metal-organic framework MIL-101(Cr) for natural gas (methane) storage and compare with other MOFs (metal-organic frameworks). Energy 91, 772–781 (2015).
4. Furukawa, H., Cordova, K. E., O’Keeffe, M. & Yaghi, O. M. The Chemistry and Applications of Metal-Organic Frameworks. Science (80-. ). 341, 974 (2013).
5. Teo, H. W. B., Chakraborty, A. & Kayal, S. Evaluation of CH4 and CO2 adsorption on HKUST-1 and MIL-101(Cr) MOFs employing Monte Carlo simulation and comparison with experimental data. Appl. Therm. Eng. 110, 891–900 (2017).
6. Darunte, L. A., Oetomo, A. D., Walton, K. S., Sholl, D. S. & Jones, C. W. Direct Air Capture of CO2 Using Amine Functionalized MIL-101(Cr). ACS Sustain. Chem. Eng. 4, 5761–5768 (2016).
7. Maksimchuk, N. V. et al. Metal-organic frameworks of the MIL-101 family as heterogeneous single-site catalysts. Proc. R. Soc. A Math. Phys. Eng. Sci. 468, 2017–2034 (2012).
8. Zhang, W. et al. A family of metal-organic frameworks exhibiting size-selective catalysis with encapsulated noble-metal nanoparticles. Adv. Mater. 26, 4056–4060 (2014).
9. Pan, H. et al. Pt nanoparticles entrapped in mesoporous metal-organic frameworks MIL-101 as an efficient catalyst for liquid-phase hydrogenation of benzaldehydes and nitrobenzenes. J. Mol. Catal. A Chem. 399, 1–9 (2015).
10. Mao, Y., Cao, W., Li, J., Sun, L. & Peng, X. HKUST-1 membranes anchored on porous substrate by hetero MIL-110 nanorod array seeds. Chem. - A Eur. J. 19, 11883–11886 (2013).
11. Sachse, A. et al. In situ synthesis of Cu–BTC (HKUST-1) in macro-/mesoporous silica monoliths for continuous flow catalysis. Chem. Commun. 48, 4749 (2012).
12. Li, L. et al. A MOF/graphite oxide hybrid (MOF: HKUST-1) material for the adsorption of methylene blue from aqueous solution. J. Mater. Chem. A 1, 10292–10299 (2013).
13. Granato, T., Testa, F. & Olivo, R. Catalytic activity of HKUST-1 coated on ceramic foam. Microporous Mesoporous Mater. 153, 236–246 (2012).
14. Bradshaw, D., Garai, A. & Huo, J. Metal-organic framework growth at functional interfaces: thin films and composites for diverse applications. Chem. Soc. Rev. 41, 2344–2381 (2012).
15. Chen, Y. F., Babarao, R., Sandler, S. I. & Jiang, J. W. Metal-organic framework MIL-101 for adsorption and effect of terminal water molecules: From quantum mechanics to molecular simulation. Langmuir 26, 8743–8750 (2010).
16. Leus, K. et al. Systematic study of the chemical and hydrothermal stability of selected ‘stable’ Metal Organic Frameworks. Microporous Mesoporous Mater. 226, 110–116 (2016).
62
17. Leus, K. et al. Atomic Layer Deposition of Pt Nanoparticles within the Cages of MIL-101: A Mild and Recyclable Hydrogenation Catalyst. Nanomaterials 6, 45 (2016).
18. Du, L., Xu, H., Li, T., Zhang, Y. & Zou, F. Fabrication of silver nanoparticle/polyvinyl alcohol/polycaprolactone hybrid nanofibers nonwovens by two-nozzle electrospinning for wound dressing. Fibers Polym. 17, 1995–2005 (2016).
19. Bhardwaj, N. & Kundu, S. C. Electrospinning: A fascinating fiber fabrication technique. Biotechnol. Adv. 28, 325–347 (2010).
20. Patrício, T., Domingos, M., Gloria, A. & Bártolo, P. Characterisation of PCL and PCL/PLA scaffolds for tissue engineering. Procedia CIRP 5, 110–114 (2013).
21. Ghosal, K., Manakhov, A., Zajíčková, L. & Thomas, S. Structural and Surface Compatibility Study of Modified Electrospun Poly(ε-caprolactone) (PCL) Composites for Skin Tissue Engineering. AAPS PharmSciTech 18, 72–81 (2016).
22. Farha, O. K. et al. Metal-organic framework materials with ultrahigh surface areas: Is the sky the limit? J. Am. Chem. Soc. 134, 15016–15021 (2012).
23. Rosi, N. L. et al. Article Rod Packings and Metal − Organic Frameworks Constructed from Rod-Shaped Secondary Building Units Rod Packings and Metal - Organic Frameworks Constructed from Rod-Shaped Secondary Building Units. 1504–1518 (2005). doi:10.1021/ja045123o
24. Llewellyn, P. L. et al. Prediction of the conditions for breathing of metal organic framework materials using a combination of X-ray powder diffraction, microcalorimetry, and molecular simulation. J. Am. Chem. Soc. 130, 12808–12814 (2008).
25. Neimark, A. V., Coudert, F. X., Boutin, A. & Fuchs, A. H. Stress-based model for the breathing of metal-organic frameworks. J. Phys. Chem. Lett. 1, 445–449 (2010).
26. Hong, D. Y., Hwang, Y. K., Serre, C., Férey, G. & Chang, J. S. Porous chromium terephthalate MIL-101 with coordinatively unsaturated sites: Surface functionalization, encapsulation, sorption and catalysis. Adv. Funct. Mater. 19, 1537–1552 (2009).
27. Zhang, J. et al. High performance humidity sensor based on metal organic framework MIL-101(Cr) nanoparticles. J. Alloys Compd. 695, 520–525 (2017).
28. Xu, Y. et al. Highly and stably water permeable thin film nanocomposite membranes doped with MIL-101 (Cr) nanoparticles for reverse osmosis application. Materials (Basel). 9, (2016).
29. Wu, C. De, Hu, A., Zhang, L. & Lin, W. A homochiral porous metal-organic framework for highly enantioselective heterogeneous asymmetric catalysis. J. Am. Chem. Soc. 127, 8940–8941 (2005).
30. Lee, S. J., Hu, A. & Lin, W. The first chiral organometallic triangle for asymmetric catalysis. J. Am. Chem. Soc. 124, 12948–12949 (2002).
31. Gomez-Lor, B. et al. In2(OH)3(BDC)1.5 (BDC = 1,4-benzendicarboxylate): An In(III) supramolecular 3D framework with catalytic activity. Inorg. Chem. 41, 2429–2432 (2002).
32. Meilikhov, M. et al. Metals@MOFs - Loading MOFs with metal nanoparticles for hybrid functions. Eur. J. Inorg. Chem. 3701–3714 (2010). doi:10.1002/ejic.201000473
33. Stassen, I., De Vos, D. & Ameloot, R. Vapor-Phase Deposition and Modification of Metal???Organic Frameworks: State-of-the-Art and Future Directions. Chem. - A Eur. J. 22, 14452–14460 (2016).
63
34. Ishida, T., Nagaoka, M., Akita, T. & Haruta, M. Deposition of gold clusters on porous coordination polymers by solid grinding and their catalytic activity in aerobic oxidation of alcohols. Chem. - A Eur. J. 14, 8456–8460 (2008).
35. Jiang, H. et al. Au @ ZIF-8 : CO Oxidation over Gold Nanoparticles Deposited to Metal - Organic Framework. J. Am. Chem. Soc. 2, 11302–11303 (2009).
36. Maeda, Y., Taguchi, N., Akita, T. & Kohyama, M. A Simultaneous Solid Grinding Method for the Preparation of Gold Catalysts. Catal. Letters 146, 2376–2380 (2016).
37. Vindigni, F., Dughera, S., Armigliato, F. & Chiorino, A. Aerobic oxidation of alcohols on Au/TiO2 catalyst: new insights on the role of active sites in the oxidation of primary and secondary alcohols. Monatshefte für Chemie - Chem. Mon. 147, 391–403 (2016).
38. Hermes, S. et al. Metal@MOF: Loading of highly porous coordination polymers host lattices by metal organic chemical vapor deposition. Angew. Chemie - Int. Ed. 44, 6237–6241 (2005).
39. Puurunen, R. L. Surface chemistry of atomic layer deposition : a case study for the trimethylaluminum / water process. Surface chemistry of atomic layer deposition: a case study for the trimethylaluminum/water process 97, (2005).
40. Kim, H. Atomic layer deposition of metal and nitride thin films: Current research efforts and applications for semiconductor device processing. J. Vac. Sci. Technol. B Microelectron. Nanom. Struct. 21, 2231–2261 (2003).
41. Elam, J. W. et al. Atomic layer deposition for the conformal coating of nanoporous materials. J. Nanomater. 2006, 1–5 (2006).
42. Johnson, R. W., Hultqvist, A. & Bent, S. F. A brief review of atomic layer deposition: From fundamentals to applications. Mater. Today 17, 236–246 (2014).
43. Elam, J. W., Routkevitch, D., Mardilovich, P. P. & George, S. M. Conformal coating on ultrahigh-aspect-ratio nanopores of anodic alumina by atomic layer deposition. Chem. Mater. 15, 3507–3517 (2003).
44. Gordon, R. G., Hausmann, D., Kim, E. & Shepard, J. A kinetic model for step coverage by atomic layer deposition in narrow holes or trenches. Chem. Vap. Depos. 9, 73–78 (2003).
45. George, S. M. Atomic layer deposition: An overview. Chem. Rev. 110, 111–131 (2010).
46. Mondloch, J. E. et al. Vapor-phase metalation by atomic layer deposition in a metal-organic framework. J. Am. Chem. Soc. 135, 10294–10297 (2013).
47. Neogi, S., Sharma, M. K. & Bharadwaj, P. K. Knoevenagel condensation and cyanosilylation reactions catalyzed by a MOF containing coordinatively unsaturated Zn(II) centers. J. Mol. Catal. A Chem. 299, 1–4 (2009).
48. Texier-Boullet, F. & Foucaud, A. Knoevenagel condensation catalysed by aluminium oxide. Tetrahedron Lett. 23, 4927–4928 (1982).
49. Peters, A. W., Li, Z., Farha, O. K. & Hupp, J. T. Atomically Precise Growth of Catalytically Active Cobalt Sulfide on Flat Surfaces and within a Metal-Organic Framework via Atomic Layer Deposition. ACS Nano 9, 8484–8490 (2015).
50. Fan, X. et al. Characterization and application of zeolitic imidazolate framework-8@polyvinyl alcohol nanofibers mats prepared by electrospinning. Mater. Res. Express 4, 26404 (2017).
51. Crespy, D., Friedemann, K. & Popa, A. M. Colloid-electrospinning: Fabrication of
64
multicompartment nanofibers by the electrospinning of organic or/and inorganic dispersions and emulsions. Macromol. Rapid Commun. 33, 1978–1995 (2012).
52. Ren, J. et al. Electrospun MOF nanofibers as hydrogen storage media. Int. J. Hydrogen Energy 40, 9382–9387 (2015).
53. Wahiduzzaman, Khan, M. R., Harp, S., Neumann, J. & Sultana, Q. N. Processing and Performance of MOF (Metal Organic Framework)-Loaded PAN Nanofibrous Membrane for CO2 Adsorption. J. Mater. Eng. Perform. 25, 1276–1283 (2016).
54. Gao, M., Zeng, L., Nie, J. & Ma, G. Polymer-metal-organic framework core-shell framework nanofibers via electrospinning and their gas adsorption activities. Rsc Adv. 6, 7078–7085 (2016).
55. Bechelany, M. et al. Highly Crystalline MOF-based Materials Grown on Electrospun Nanofibers. Nanoscale 5794–5802 (2015). doi:10.1039/C4NR06640E
56. Liu, C. et al. General Deposition of Metal-Organic Frameworks on Highly Adaptive Organic-Inorganic Hybrid Electrospun Fibrous Substrates. ACS Appl. Mater. Interfaces 8, 2552–2561 (2016).
57. Asiabi, M., Mehdinia, A. & Jabbari, A. Electrospun biocompatible Chitosan/MIL-101 (Fe) composite nanofibers for solid-phase extraction of Δ9-tetrahydrocannabinol in whole blood samples using Box-Behnken experimental design. J. Chromatogr. A 1479, 71–80 (2017).
58. Asiabi, M., Mehdinia, A. & Jabbari, A. Preparation of water stable methyl-modified metal-organic framework-5/polyacrylonitrile composite nanofibers via electrospinning and their application for solid-phase extraction of two estrogenic drugs in urine samples. J. Chromatogr. A 1426, 24–32 (2015).
59. Quiros, J. et al. Antimicrobial metal-organic frameworks incorporated into electrospun fibers. Chem. Eng. J. 262, 189–197 (2015).
60. Thavasi, V., Singh, G. & Ramakrishna, S. 036 Electrospun nanofibers in energy and environmental applications. Energy Environ. Sci. 1, 205 (2008).
61. Bjorge, D. et al. Initial testing of electrospun nanofibre filters in water filtration applications. Water SA 36, 151–156 (2010).
62. Savva, I. et al. Evaluation of PVP/Au nanocomposite fibers as heterogeneous catalysts in indole synthesis. Molecules 21, 1–13 (2016).
63. Li, H. et al. Thermosensitive nanofibers loaded with ciprofloxacin as antibacterial wound dressing materials. Int. J. Pharm. 517, 135–147 (2017).
64. Shankhwar, N., Kumar, M., Mandal, B. B., Robi, P. S. & Srinivasan, A. Electrospun polyvinyl alcohol-polyvinyl pyrrolidone nanofibrous membranes for interactive wound dressing application. J. Biomater. Sci. Polym. Ed. 27, 247–262 (2016).
65. Wang, L., Chang, M. W., Ahmad, Z., Zheng, H. & Li, J. S. Mass and controlled fabrication of aligned PVP fibers for matrix type antibiotic drug delivery systems. Chem. Eng. J. 307, 661–669 (2017).
66. Deitzel, J. ., Kleinmeyer, J., Harris, D. & Beck Tan, N. . The effect of processing variables on the morphology of electrospun nanofibers and textiles. Polymer (Guildf). 42, 261–272 (2001).
67. Haghi, A. K. & Akbari, M. Trends in electrospinning of natural nanofibers. Phys. Status Solidi Appl. Mater. Sci. 204, 1830–1834 (2007).
65
68. Doshi, J. & Reneker, D. H. Electrospinning process and applications of electrospun fibers. Conf. Rec. 1993 IEEE Ind. Appl. Conf. Twenty-Eighth IAS Annu. Meet. 35, 151–160 (1993).
69. Hohman, M. M., Shin, M., Rutledge, G. & Brenner, M. P. Electrospinning and electrically forced jets. II. Applications. Phys. Fluids 13, 2221–2236 (2001).
70. Yördem, O. S., Papila, M. & Menceloǧlu, Y. Z. Effects of electrospinning parameters on polyacrylonitrile nanofiber diameter: An investigation by response surface methodology. Mater. Des. 29, 34–44 (2008).
71. Yuan, X. Y., Zhang, Y. Y., Dong, C. & Sheng, J. Morphology of ultrafine polysulfone fibers prepared by electrospinning. Polym. Int. 53, 1704–1710 (2004).
72. Bahrami, S. H. & Gholipour Kanani, A. Effect of changing solvents on poly(ε-Caprolactone) nanofibrous webs morphology. J. Nanomater. 2011, (2011).
73. Luo, C. J., Stride, E. & Edirisinghe, M. Mapping the influence of solubility and dielectric constant on electrospinning polycaprolactone solutions. Macromolecules 45, 4669–4680 (2012).
74. Jiang, D. M., Burrows, A. D. & Edler, K. J. Size-controlled synthesis of MIL-101(Cr) nanoparticles with enhanced selectivity for CO2 over N2. CrystEngComm 13, 6916–6919 (2011).
75. Dendooven, J. et al. Low-temperature atomic layer deposition of platinum using (methylcyclopentadienyl)trimethylplatinum and ozone. J. Phys. Chem. C 117, 20557–20561 (2013).
76. Xie, L. et al. Preparation and characterization of metal-organic framework MIL-101(Cr)-coated solid-phase microextraction fiber. Anal. Chim. Acta 853, 303–310 (2015).
77. Lee, K. H., Kim, H. Y., Khil, M. S., Ra, Y. M. & Lee, D. R. Characterization of nano-structured poly(ε-caprolactone) nonwoven mats via electrospinning. Polymer (Guildf). 44, 1287–1294 (2003).
78. Moghe, A. K., Hufenus, R., Hudson, S. M. & Gupta, B. S. Effect of the addition of a fugitive salt on electrospinnability of poly(??-caprolactone). Polymer (Guildf). 50, 3311–3318 (2009).
79. Khil, M. S., Bhattarai, S. R., Kim, H. Y., Kim, S. Z. & Lee, K. H. Novel fabricated matrix via electrospinning for tissue engineering. J. Biomed. Mater. Res. - Part B Appl. Biomater. 72, 117–124 (2005).
80. Hsu, C. M. & Shivkumar, S. N,N-dimethylformamide additions to the solution for the electrospinning of poly(??-caprolactone) nanofibers. Macromol. Mater. Eng. 289, 334–340 (2004).
81. Fong, H. et al. Beaded nano bers formed during electrospinning. Polymer (Guildf). 40, 4585–4592 (1999).