The multipurpose time-of-flight neutron reflectometer ‘‘Platypus’’ at Australia’s OPAL reactor M. James a,b,n , A. Nelson a , S.A. Holt a , T. Saerbeck a , W.A. Hamilton a , F. Klose a a Bragg Institute, Building 87, Australian Nuclear Science and Technology Organisation, Locked Bag 2001, Kirrawee DC, NSW 2232, Australia b School of Chemistry, University of New South Wales, Kensington, NSW 2052, Australia article info Article history: Received 8 September 2010 Received in revised form 18 November 2010 Accepted 8 December 2010 Available online 17 December 2010 Keywords: Neutron reflectometry Cold neutrons Time-of-flight Free-liquid surfaces Polarised neutron reflectometry Off-specular scattering abstract In this manuscript we describe the major components of the Platypus time-of-flight neutron reflectometer at the 20 MW OPAL reactor in Sydney, Australia. Platypus is a multipurpose spectrometer for the characterisation of solid thin films, materials adsorbed at the solid–liquid interface and free-liquid surfaces. It also has the capacity to study magnetic thin films using spin-polarised neutrons. Platypus utilises a white neutron beam (l ¼2–20 ˚ A) that is pulsed using boron-coated disc chopper pairs; thus providing the capacity to tailor the wavelength resolution of the pulses to suit the system under investigation. Supermirror optical components are used to focus, deflect or spin-polarise the broad bandwidth neutron beams, and typical incident spectra are presented for each configuration. A series of neutron reflectivity datasets are presented, indicating the quality and flexibility of this spectrometer. Minimum reflectivity values of o10 7 are observed; while maximum thickness values of 325 nm have been measured for single-component films and 483 nm for a multilayer system. Off-specular measure- ments have also been made to investigate in-plane features as opposed to those normal to the sample surface. Finally, the first published studies conducted using the Platypus time-of-flight neutron reflectometer are presented. & 2010 Elsevier B.V. All rights reserved. 1. Introduction The Platypus neutron reflectometer is now operational as part of an international user facility at Australia’s 20 MW OPAL research reactor ANSTO in Sydney. Commissioned in 2008, Platypus is one of a new class of time-of-flight reflectometers to be built at modern research reactors (e.g. D17 [1] and Figaro [2] at ILL, ROG [3] at TU Delft, and REFSANS [4] at FRMII). The main reason for turning to the time-of-flight method is due to the capacity of such instruments to optimise the interplay between flux and resolution to suit each individual system under investigation. Neutron reflectometry has become one of the key scattering techniques around the world, with typically two or more reflectometers in operation at major international facilities. In most cases, one instrument is dedicated to hard matter systems such as magnetic thin films, requiring a polarised incident neutron beam and polarisation analysis, while the other is dedicated to soft condensed matter (particularly thin- films deposited at the solid/liquid or air/liquid interfaces). Usually these requirements also dictate differences in instrumental geo- metry, with horizontal scattering geometry typically used for polarised reflectometry and vertical scattering geometry for air/liquid surfaces. Being the only neutron reflectometer at the OPAL facility in the initial suite of seven instruments, Platypus was designed to be a flexible, multipurpose instrument to service both hard and soft matter user communities and capable of investigating air/liquid interfaces, solid/liquid and solid/air interfaces, dynamic systems, off-specular, and magnetic thin films [5]. The ability to conduct free-liquid experiments dictated that Platypus be config- ured with a vertical scattering geometry, while acquisition with a wide wavelength bandwidth is essential for kinetic studies of dynamic systems. The off-specular requirement demanded that the detection system was based around a 2-dimensional area detector. The past few decades has seen neutron reflectometry become an essential tool for the study of nanoscale thin-films and processes that take place at surfaces. While early uses of this technique revolved around studies of surfactants [6], polymers [7], and magnetic thin-films [8,9], (fields of research that still remain popular); more recently, researchers have made use of neutron reflection in studies of nanotoxicology [10,11], cell membrane biology [12–21], protein adsorption [22–27], electrochemistry [28–33] and corrosion studies [34–36], organic photovoltaic systems [37–39], and light-emitting thin films [40–42]. Polarised neutron reflectometry has been widely used to characterize magnetic and superconducting thin-films and multilayers, with recent studies making use of both specular [43–49] and off-specular [50–53] capabilities. Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/nima Nuclear Instruments and Methods in Physics Research A 0168-9002/$ - see front matter & 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.nima.2010.12.075 n Corresponding author at: Bragg Institute, Building 87, Australian Nuclear Science and Technology Organisation, Locked Bag 21, Kirrawee DC, NSW 2232, Australia. Tel.: + 61 2 9717 9299. E-mail address: [email protected] (M. James). Nuclear Instruments and Methods in Physics Research A 632 (2011) 112–123
Transcript
Nuclear Instruments and Methods in Physics Research A 632 (2011) 112–123
Contents lists available at ScienceDirect
Nuclear Instruments and Methods inPhysics Research A
0168-90
doi:10.1
n Corr
Science
Australi
E-m
journal homepage: www.elsevier.com/locate/nima
The multipurpose time-of-flight neutron reflectometer ‘‘Platypus’’ atAustralia’s OPAL reactor
M. James a,b,n, A. Nelson a, S.A. Holt a, T. Saerbeck a, W.A. Hamilton a, F. Klose a
a Bragg Institute, Building 87, Australian Nuclear Science and Technology Organisation, Locked Bag 2001, Kirrawee DC, NSW 2232, Australiab School of Chemistry, University of New South Wales, Kensington, NSW 2052, Australia
a r t i c l e i n f o
Article history:
Received 8 September 2010
Received in revised form
18 November 2010
Accepted 8 December 2010Available online 17 December 2010
In this manuscript we describe the major components of the Platypus time-of-flight neutron reflectometer
at the 20 MW OPAL reactor in Sydney, Australia. Platypus is a multipurpose spectrometer for the
characterisation of solid thin films, materials adsorbed at the solid–liquid interface and free-liquid
surfaces. It also has the capacity to study magnetic thin films using spin-polarised neutrons. Platypus
utilises a white neutron beam (l¼2–20 A) that is pulsed using boron-coated disc chopper pairs; thus
providing the capacity to tailor the wavelength resolution of the pulses to suit the system under
investigation. Supermirror optical components are used to focus, deflect or spin-polarise the broad
bandwidth neutron beams, and typical incident spectra are presented for each configuration. A series of
neutron reflectivity datasets are presented, indicating the quality and flexibility of this spectrometer.
Minimum reflectivity values of o10�7 are observed; while maximum thickness values of 325 nm have
been measured for single-component films and 483 nm for a multilayer system. Off-specular measure-
ments have also been made to investigate in-plane features as opposed to those normal to the sample
surface. Finally, the first published studies conducted using the Platypus time-of-flight neutron
reflectometer are presented.
& 2010 Elsevier B.V. All rights reserved.
1. Introduction
The Platypus neutron reflectometer is now operational as part ofan international user facility at Australia’s 20 MW OPAL researchreactor ANSTO in Sydney. Commissioned in 2008, Platypus is one ofa new class of time-of-flight reflectometers to be built at modernresearch reactors (e.g. D17 [1] and Figaro [2] at ILL, ROG [3] at TUDelft, and REFSANS [4] at FRMII). The main reason for turning to thetime-of-flight method is due to the capacity of such instruments tooptimise the interplay between flux and resolution to suit eachindividual system under investigation. Neutron reflectometry hasbecome one of the key scattering techniques around the world,with typically two or more reflectometers in operation at majorinternational facilities. In most cases, one instrument is dedicatedto hard matter systems such as magnetic thin films, requiring apolarised incident neutron beam and polarisation analysis, whilethe other is dedicated to soft condensed matter (particularly thin-films deposited at the solid/liquid or air/liquid interfaces). Usuallythese requirements also dictate differences in instrumental geo-metry, with horizontal scattering geometry typically used for
ll rights reserved.
ing 87, Australian Nuclear
1, Kirrawee DC, NSW 2232,
polarised reflectometry and vertical scattering geometry forair/liquid surfaces. Being the only neutron reflectometer at theOPAL facility in the initial suite of seven instruments, Platypus wasdesigned to be a flexible, multipurpose instrument to service bothhard and soft matter user communities and capable of investigatingair/liquid interfaces, solid/liquid and solid/air interfaces, dynamicsystems, off-specular, and magnetic thin films [5]. The ability toconduct free-liquid experiments dictated that Platypus be config-ured with a vertical scattering geometry, while acquisition with awide wavelength bandwidth is essential for kinetic studies ofdynamic systems. The off-specular requirement demanded that thedetection system was based around a 2-dimensional area detector.
The past few decades has seen neutron reflectometry become anessential tool for the study of nanoscale thin-films and processes thattake place at surfaces. While early uses of this technique revolvedaround studies of surfactants [6], polymers [7], and magneticthin-films [8,9], (fields of research that still remain popular); morerecently, researchers have made use of neutron reflection in studies ofnanotoxicology [10,11], cell membrane biology [12–21], proteinadsorption [22–27], electrochemistry [28–33] and corrosion studies[34–36], organic photovoltaic systems [37–39], and light-emittingthin films [40–42]. Polarised neutron reflectometry has been widelyused to characterize magnetic and superconducting thin-films andmultilayers, with recent studies making use of both specular [43–49]and off-specular [50–53] capabilities.
M. James et al. / Nuclear Instruments and Methods in Physics Research A 632 (2011) 112–123 113
The following paper reports the detailed characteristics andcapabilities of the Platypus reflectometer along with a diverse rangeof the initial scientific studies performed using this instrument.
2. Major components of the Platypus reflectometer
2.1. Hardware and neutron optical elements
Details of the design and construction of the Platypus reflectometer(Fig. 1) have been described previously in Ref. [5]. The following textdescribes the major components of this spectrometer as well as thekey modes of operation. Cold neutrons are generated in the OPALliquid D2 cold neutron moderator, with a typical operating tempera-ture of 21 K. Cold neutrons from this moderator are directed throughthe CG3 curved supermirror guide (radius of curvature 1300 m;200 mm high�50 mm wide) within the heavy concrete neutronguide bunker, from OPAL’s primary shutters to the secondary shutterat the exit of the bunker. This curved guide removes line-of sight fromthe cold neutron source at the point where the guide passes throughthe secondary shutter, such that high energy gamma radiation and fastneutrons (40.5 MeV) are trapped within the neutron guide bunker.Following a further 3 m section of straight neutron guide beyond thesecondary shutter, the polychromatic neutron beam passes a tertiaryshutter and is fed into the Platypus via a (20 mm high�50 mm wide)neutron guide, with nickel coatings on the horizontal surfaces andm¼3 supermirror coatings on the vertical surfaces.
The white neutron beam is delivered into a B4C-lined, lead andsteel disc chopper bunker, which is designed to protect staff andequipment from thermal neutrons that are scattered from vacuumwindows, slits and the chopper discs; as well as g-rays that areproduced by neutrons being adsorbed by the above components. A400 mm section of guide containing a frame overlap mirror isflanged directly onto the end of the pre-instrument guide. Theframe overlap mirror (with a Ni/Ti, m¼1.5 supermirror coating)reflects neutrons above 20 A out of the main beam to be adsorbedwithin the bunker. Neutrons r20 A are transmitted through thereflective coating and silicon substrate of the frame overlap mirrorto the rest of the instrument. The beam passes a coarse collimationslit (s1), an optional beam attenuator and a low efficiency (10�6)neutron beam monitor before entering the disc chopper system.The beam attenuator consists of a 50 mm high�1 mm widevertical slit formed from 3 mm thick sintered B4C plates, which
Fig. 1. The Platypus time-of-flight neutron reflectometer.
oscillates across the horizontal incident beam with constantvelocity and a frequency of 2 Hz. When not required, the attenuatoris parked to the side of slit s1, allowing the unattenuated beam toenter the vacuum vessel containing the disc chopper system.
The disc chopper system (EADS Astrium GmbH) consists of fourboron-10 coated carbon fibre discs, separated by neutron guideelements (Fig. 2). At any one time, a pair of boron-10 coated,aluminium discs rotating at 20 Hz are used to shape the neutronpulses. Full specifications of the disc chopper system are given inTable 1. The first disc (indicated by the red dashed line in Fig. 2) isthe master and T0 chopper disc that triggers the frame acquisitionon the Platypus detector, and operates in combination with thesecond (green), third (blue) or the fourth (purple) slave discs. Byutilising different master/slave disc pair combinations, Platypus isable to generate three distinct resolution settings: high- (Dl/l¼1.2%), medium- (Dl/l¼4.3%), and low-resolution (Dl/l¼8.4%).When not required to shape the neutron pulse, the two other slavediscs are parked in a stationary, open position.
The neutron pulses are then delivered to Platypus collimationsystem, which consists of a pair of high-precision beam-defining slits(2835 mm apart) that encompass a vacuum chamber containingvarious neutron optical elements to focus or deflect the neutron beamtowards the sample. The slit package (s2) immediately before thecollimation vacuum chamber is mounted centrally with respect to theincident neutron beam, while the slit package immediately before thesample stage (s3) is mounted on a motorised vertical translation stage(Fig. 3(a)). Each slit package is capable of defining a beam from fully
Fig. 2. The Platypus disc chopper system. Disc pairs 1 and 2, 1 and 3, or 1 and 4 are
used to generate neutron pulses of wavelength resolution (Dl/l) 1.5%, 4% or 9%,
respectively. (For interpretation of the references to color in this figure, the reader is
referred to the web version of this article.)
Table 1Specifications of EADS Astrium GmbH disc chopper system.
Specifications
Disc outer diameter (mm) 700
Window radial length (mm) 60
Maximum speed (Hz) 83.33
Relative phase variation (1) o0.1
Disc absorber material boron-10
Disc 1 Disc 2 Disc 3 Disc 4
Disc cut-out angle (1) 60 10 25 60
Distance from master disc (mm) 103 359 808
Dl/la (for D¼7545 mm)b 0.012 0.043 0.084
a Gaussian FWHM, calculated using the method of Van Well and Fredrikze [58].b Standard instrument configuration. D can vary from 7520 to 8610 mm.
M. James et al. / Nuclear Instruments and Methods in Physics Research A 632 (2011) 112–123114
closed to a maximum area of 30 mm high�54 mm wide, with aprecision better than 3 mm. The neutron adsorbing material used inthese slit packages is 3 mm thick sintered plates of B4C.
The major components contained in the evacuated collimationtank are a series of four neutron guides (Fig. 3(b)–(d)) that aremounted on a horizontal translation stage. Each of these elementscan be translated into the incident beam as required for thefunctioning of different instrument modes (solid samples, free-liquid surfaces, or polarised neutron reflectometry). A fifth opticalelement is available where no guide is present between thecollimating slit packages. This ‘‘no guide’’ option may be utilisedin those instances where off-specular studies of solid samplesrequire tighter angular resolution in the horizontal direction (at theexpense of some loss in neutron flux). The four guide-based opticalelements are: (i) a horizontally mounted (1000 mm long), ‘‘singlebounce’’ Ni/Ti (m¼4) supermirror that is mounted on a rotationstage to deflect neutrons down onto free-liquid surfaces at anglesup to 0.71 (Fig. 3(b)); (ii) a pair of 600 mm long, ‘‘double bounce’’m¼4 supermirrors in series that deflects the neutron beam downonto free-liquid surfaces at a fixed angle of 4.81 (Fig. 3(c)); (iii) a1200 mm long, horizontally focusing (m¼4) supermirror guide forsmall solid samples (Fig. 3(d)); and (iv) a 500 mm long Fe/Sipolarising supermirror (m¼3.8) in series with a RF spin-flipper coilto produce spin-up or spin-down neutrons (not shown). The RFspin-flipper circuit operating at 193 kHz allows for almost com-plete (96% efficiency) spin reversal of the entire wavelengthspectrum that is transmitted by the polarising supermirror.
The Platypus sample stages (Fig. 4), which follow the collimationsystem, allows the samples to be located with a high degree of precisionwith respect to the incident beam. The vertical sample stage (Fig. 4(a))can accommodate loads up to 1000 kg (important for cryocoolerexperiments), and can move samples from 220 mm below the location
Fig. 3. Components of the Platypus collimation system: (a) collimating slit package s3, m
supermirror neutron guides: (b) single bounce deflection mirror on a rotation stage; (c
electromagnet, and 4–320 K cryocooler for polarised neutron reflectometry studies.
of the horizontal, undeflected incident beam to 57 mm above theundeflected incident beam with a precision of 1 mm. The horizontalsample stage (Fig. 4(b)) with 730 mm of travel, locates samplesperpendicular to the beam and is also used as an automatic samplechanger. The two-circle sample goniometer (Fig. 4(c)) allows tilting ofsamples up to 7151 in both the o-direction (axis of rotationperpendicular to the beam) and f-direction (axis of rotation parallelto the beam) with 0.0011 precision. This load capacity and movementprecision are particularly essential when the 1 T electromagnet(Bruker) and associated cryocooler (4–320 K; Advanced ResearchSystems) (Fig. 4(d)) are used for polarised neutron reflectometryexperiments. A final slit package (s4), after the sample position, allowsthe reflected neutron beam to reach the detector tank, while blockingthe transmitted beam.
Once inside the detector tank, the neutron beam either passesdirectly to the Platypus detector, or in the case of spin-polarisedneutrons may pass through a second RF spin-flipper coil (operating at198 kHz) (Fig. 5(a)) and are separated using a polarisation analyser(Fig. 5(b)). In the case of polarised neutrons, down-spin neutrons aretransmitted through a Fe/Si (m¼3.8) polarising supermirror that isoriented at 0.881 with respect to the beam reflected from the sample.Spin-up neutrons are reflected from the analyser supermirror, leadingto a separation of the two spin-states on the detector. Thus incomingand outgoing neutron spin states can be adequately adjusted using thepolarizer and the analyser, and the complete polarisation systemallows detection of four neutron spin channels: two non-spin flipchannels (R++ and R��) with incoming and outgoing spin stateunchanged and two spin flip channels (R+� and R� +) with neutronspin reversal due to interaction with the sample.
A two-dimensional helium-3 neutron detector (Model 250TN,Denex GmbH) with an active area 500 mm wide�250 mm high(Fig. 5(c)) is mounted on a motorised stage within a 3 m long
ounted on a high-precision vertical translation stage; collimation optics based on
) double-bounce deflection mirror; and (d) horizontal focusing guide.
anslation stage, and (c) 2-circle goniometer. (d) The Bruker, 1 T horizontal field
Data acquisition electronics P7888 (FAST ComTec GmbH)
Gamma sensitivity o10�8
M. James et al. / Nuclear Instruments and Methods in Physics Research A 632 (2011) 112–123 115
detector tank. Details of the detector characteristics are listed inTable 2. The horizontal location of the detector along the neutronbeam can be moved to vary the sample-to-detector distance(between 2475 and 3565 mm), and is typically set to 2500 mm.The detector height is also varied during each reflectometrymeasurement to track the trajectory of the reflected beam, from420 mm below the undeflected, horizontal incident beam to930 mm above the horizontal undeflected beam. This range oftravel allows scattering angles from the sample to be detectedbetween �101 and +201. In addition to measurement of thespecular reflection and associated background signal, the largeactive area of the detector also allows collection of off-specularscattering. While in theory these angles allow for the measurementof specular reflectivity out to Q¼1.0 A�1, in practise the instru-mental and sample-dependent background limits data from solidsamples to Q�0.5 A�1.
The delay-line system of the Denex detector accumulates a2-dimensional dataset, with 221 spatial pixels in the verticaldirection and with each (20 Hz) neutron pulse spread across1000 time channels of 50 ms duration. The spatial resolution inthe vertical direction is 1.18 mm/pixel. Spatial and temporal datafrom the delay-lines are recorded in list mode by multiple-eventtime digitizer electronics (P7888 model, FAST ComTec GmbH).While the detector has 421 horizontal pixels (with a spatialresolution of 1.19 mm/pixel); for specular reflectometry, countsin these pixels are integrated to a single value in the histogramserver. This provides a reduction in the size of the dataset from thedetector from 354 Mb to 841 kb and allows real-time updates ofthe Platypus detector in the FIZZY software (see below). Should2-dimensional spatial data be required from the Denex detector,both horizontal and vertical pixellation can be restored to thedataset. Although incoming neutrons are typically histogrammedinto specific time channels, counting in event mode is also possible.Event mode operation allows for kinetic changes in reflectivity
profiles to be measured down to the timescale of the period ofrotation of the disk choppers.
2.2. Instrument control software, and data reduction system
Platypus operation is controlled at a server and client level. Theserver level consists of the SICS instrument control system [54](which is responsible for all motion and sample environmentcontrol) and a webserver (DAS) for data acquisition. The DASsystem histograms the spatial and time-of-flight information ofeach neutron event on the detector. SICS then saves the histo-grammed data and instrument state in a NeXUS file [55] forprocessing at a later stage. Since the server level programs possesstext based interfaces, a user-friendly graphical client interface,FIZZY [56], was developed in the IGOR Pro environment [57]. FIZZY
considerably simplifies measurement tasks by providing easyvisualisation and control of instrument parameters, single buttonpress data acquisition and alignment, real time visualisation ofmultidimensional detector images and batch mode for measure-ment of multiple samples. Data reduction takes place in the SLIM
module, and is dovetailed into data acquisition, allowing real timereduction of data. Calculation of instrumental resolution and datareduction for time-of-flight reflectometry in the SLIM module isdescribed by Van Well and Fredrikze [58], but essentially consistsof extracting the specularly reflected intensity as a function oftime-of-flight, performing detector efficiency correction, back-ground subtraction, re-binning the 1000 time-of-flight channels(typically to the chopper resolution) and transforming the time-of-flight axis to wavelength. SLIM also has to apply corrections toPlatypus data to account for the fact that long wavelength neutronsfall under gravity [59]; making the angle of incidence wavelengthdependent. The corrected reflected beam spectrum can then bedivided by the direct beam spectrum to give specular reflectivity asa function of wavelength or Q. Although off-specular maps can alsobe created, the standard output from the SLIM data reductionprocess is an ASCII text file containing Q (A�1), R, dR (standarddeviation) and dQ (variance, representing the instrumental resolu-tion); as well as information on how to repeat the reduction steps.These datasets can immediately be read into Motofit (specularreflectivity analysis) [56,60] and analysed whilst the experimentcontinues—as Motofit operates under the same IGOR Pro platform.
2.3. Sample environments available at Platypus
In a number of instances, researchers using Platypus have broughttheir own sample environments (including electrochemical cells, a UVspectrophotometer, a temperature controlled shear cell, and a quartz
M. James et al. / Nuclear Instruments and Methods in Physics Research A 632 (2011) 112–123116
crystal based surface acoustic wave generator). Standard sampleenvironments provided at the instrument include: solid–liquids cells,a Nima Mini Langmuir film balance, PTFE troughs for free-liquidsamples, a chamber for atmospheric and humidity control, a Bruker(1 T) electromagnet with integrated Advanced Research Systemscryocooler (4–320 K) with an integrated sample vacuum chamber,and an electrochemical cell for use with an Inphaze Pty Ltd impedancespectrometer. The solid–liquid sample cells are based on either100 mm diameter�10 mm thick polished Si discs, or 40 mm�80mm�15 mm thick rectangular Si blocks. Temperature control isprovided for a number of these devices using a Julabo FP-50 bathusing either water or ethylene glycol. Further details and specifi-cations of sample environments can be found at the Bragg Institutewebpages [61].
Fig. 7. Direct beam spectra for different disc pairings. (b) The medium resolution mode
(discs 1 and 3) has 3.0� the flux of (a) the high resolution mode (discs 1 and 2). (c) The
low resolution mode (discs 1 and 4) has 6.9� the flux of the high resolution mode.
Fig. 8. (a) Direct beam spectrum (red) and (b) the spectrum measured for the same
time with the beam attenuator on (green). The black line is the unattenuated
spectrum divided by 111.67. (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of this article.)
3. Platypus neutron beam characteristics
3.1. Intensity of the Platypus white beam spectrum
Irradiation of gold-foils by the Platypus direct beam was used toquantify the neutron flux at the sample position. Measurementswere taken with the disc choppers parked in the open position, andwith the collimating slits fully open giving a white beam neutronflux of 1.03�109 n/cm2/s.
Fig. 6 shows the local count rate on the Denex detector from thedirect neutron beam, as a function of the beam height passingthrough the pre-sample slit (s3). The horizontal width of this beam onthe detector was fixed at 50 mm, and the vertical beam height variedbetween 0.05 and1.0 mm. The vertical angular divergence (Dyv)associated with these data, thus range from 0.00071 to 0.0141. Beyonda vertical slit width of 0.7 mm (Dyv¼0.0101) and an area of 0.35 cm2,the local count rate diverges from linearity at a value of �10,900neutrons per second (or a neutron flux of �31,100 n/cm2/s). In orderto avoid non-linearity in the detector response during measurementof incident beam spectra, the attenuator system is used for acquisi-tions where the local count rate is greater than 8000 n/s (see below).
3.2. Effect of disc chopper selection on incident beam spectrum
Fig. 7 shows the direct beam spectra (no sample) incident uponthe instrument detector for 600 s as a function of wavelength, usingthe horizontal focusing guide, 0.5 mm high collimation slits(Dyv¼0.0071), and different chopper disc pairings. Each spectrumshows essentially the same shape, although substantially differentintensity depending on the wavelength resolution settings of the
Fig. 6. The local count-rate on the Denex detector as a function of beam height
passing through the collimating slits.
chopper discs. The medium resolution setting (discs 1 and 3Dl/l¼4.3%; Fig. 7(b)) gives 3.0 times the integrated intensity of the high-resolution setting (discs 1 and 2 Dl/l¼1.2%; Fig. 7(a)); while thelow-resolution setting (discs 1 and 4 Dl/l¼8.4%; Fig. 7(c)) gives6.9 times the integrated intensity. Although neutrons are clearlydiscernable above the instrumental background over the wave-length range 1–19 A, the effective usable bandwidth with adequateintensity for the study of solid samples is 2.5–18 A.
Fig. 8 shows the effect of activating the beam attenuator on thedirect beam spectra, for disc pairs 1 and 3, the horizontal focusingguide and 0.5 mm collimation slits (Dyv¼0.0071), and acquisitiontimes of 3000 s. Fig. 8(a) shows the unattenuated spectrum, whileFig. 8(b) shows the spectrum with the attenuator activated. Theblack plot overlaying the attenuated spectrum shows the unatte-nuated spectrum divided by a constant attenuation factor of111.67, and indicates that the attenuator functions in a wavelengthindependent fashion.
3.3. Neutron beam profiles exiting the collimation system
Fig. 9 shows the intensity of the direct beam as a function of thevertical position of the pre-sample slit (s3) and of the neutronwavelength. The pre-chopper slit (s1) was left fully open (20 mm)with the first collimation slit (s2) set at 1 mm and s3 vertical gapwas set at 0.5 mm (Dyv¼0.0101); thus imaging the upstreamneutron guides of the instrument. The post-sample slit (s4) was
M. James et al. / Nuclear Instruments and Methods in Physics Research A 632 (2011) 112–123 117
removed to allow the full range of vertical divergence (70.51) to bedetected. This figure shows three distinct lobes, with an intense,uniform, central lobe (of 20 mm height) and two weaker lobes atthe top and bottom of the of the divergence map. The two regions ofmissing intensity are due to the gap in the upstream neutron guidebetween the frame overlap mirror and the disc chopper system.These upper and lower lobes both show the loss of the shorterwavelengths with increasing vertical divergence and is consistentwith the Ni coating on the top and bottom surfaces of the upstreamneutron guides. During reflectometry measurements, only the highintensity central lobe is imaged using the collimating slits s2and s3.
Fig. 10 shows the number of neutrons striking the detector as afunction of the horizontal location on the detector. The incidentbeam is typical of that used in the first angle of a reflectometrymeasurement, with collimating slits (s2 and s3) of 0.5 mm in thevertical direction (Dyv¼0.0071) and 40 mm in the horizontaldirection (Dyh¼0.551) and the horizontal focusing guide(Fig. 3(d)) in between. The inset to this figure shows the imageof this beam on the 2-dimensional Denex detector. The full width ofthe beam on the detector is 153 mm and results from the super-mirror coatings of the upstream neutron guides (prior to slit s2), theaccepted divergence due to the horizontal slits, and the m¼4supermirror coatings of the focusing guide. In typical specularreflectometry measurements on Platypus these horizontal pixels onthe detector are integrated into a single value; the width of which is
Fig. 9. Divergence map showing the intensity of the incident beam spectra as a
function of the vertical position of slit s3.
Fig. 10. Intensity of the direct incident beam striking the detector versus horizontal pi
pre-determined by the type of neutron optic used, the horizontalcollimation slit settings and the size of the sample. While at firstglance these data do not appear to have the three distinct lobesshown in the vertical divergence map (Fig. 9), they do becomeapparent when smaller horizontal slit settings (o25 mm) are used.The wider (40 mm) slit settings used in the majority of experimentsallows for the mixing of the different neutron trajectories as theytravel through the instrument, leading to a relatively uniformprofile in the horizontal plane. Highly precise angular resolution inthe horizontal plane however is not required for these measure-ments and so the presence or absence of these lobes is not of criticalissue for specular reflectometry.
3.4. Spectra from deflection mirrors for free-liquid surfaces
Fig. 11 shows the direct beam spectra (using the medium-resolutionwavelength setting: discs 1 and 3, Dl/l¼4.3%) following deflectiondownwards off the (m¼4) single-bounce supermirror as a function ofangle (ranging between ym¼0.41 and 1.01). The incident beam wascollimated using 0.5 mm high slits (Dyv¼0.0071), directed below thehorizontal and onto the detector. The inset to Fig. 11 presents the samespectra at low wavelengths indicating the effect of ym upon the lowwavelength cut-off. The critical angle of the (m¼4) supermirror is
xel. The inset shows the image of this beam on the 2-dimensional Denex detector.
Fig. 11. Direct beam spectra off the single bounce deflection supermirror as a
function of angle with respect to the incident beam: (a) 0.41, (b) 0.51, (c) 0.61, (d) 0.71,
and (e) 1.01. (For interpretation of the references to color in this figure legend, the
reader is referred to the web version of this article.)
Fig. 12. Direct beam spectra off the double-bounce deflection supermirror at 4.81
with respect to the horizontal incident beam.Fig. 13. (a) Unpolarised (black profile) and spin-polarised neutron spectra from
(b) polarising supermirror (transmitted, spin-down, red profile), and analysing
supermirror: (c) transmitted, spin-down (blue profile) and (d) reflected, spin-up
(green profile). Dashed lines indicate minimum and maximum wavelengths for
polarised neutron reflectometry measurements. (For interpretation of the refer-
ences to color in this figure legend, the reader is referred to the web version of this
article.)
M. James et al. / Nuclear Instruments and Methods in Physics Research A 632 (2011) 112–123118
yc�0.4l leading to a low wavelength cut-off of lc¼2.5ym. Given thebandwidth typically used in Platypus experiments, it can be seen thatthe deflection mirror performs efficiently with little impact on thespectrum up to ym¼0.71. Attempts to use this deflection mirror athigher angles (e.g. ym¼1.01; Fig. 11(e) purple profile) leads to asignificant loss in neutron intensity below �10 A. The typical deflec-tion angle used for this mirror is ym¼0.51 (Fig. 11(b)), leading to anincident angle onto the free-liquid sample of 1.01 and a Q-range for thisinstrument setting of 0.012–0.088 A�1.
In order to obtain high-Q reflectivity data from free-liquidsamples, the double-bounce deflection mirror system is used todirect neutrons downwards onto horizontal surfaces at 4.81. Afterpassing through slit s2, the neutron beam first strikes a (600 mmlong) m¼4 supermirror inclined at 1.21 to the horizontal, beforebeing directed onto a second (600 mm long) supermirror inclinedat 3.61 to the horizontal. The beam leaves the second mirror at 4.81below the horizontal, before being further collimated by the pre-sample slit (s3). The sample stage allows free-liquid surfaces to betranslated well below the horizontal (�220 mm) in order to eitherallow measurement of this direct beam for specific collimationconditions, or to intercept this deflected beam for the measurementof specular reflectivity. The direct beam spectrum (using the low-resolution wavelength setting: discs 1 and 4, Dl/l¼8.4%) from thisdouble-bounce neutron optic is shown in Fig. 12. These data werecollected for 660 s, using relatively large slit settings (s2¼7.87 mmand s3¼5.32 mm; Dyv¼0.0931) that would be typically usedduring the second angle of a free-liquid measurement. In addition,the beam attenuator was used during this experiment. Thecombination of the two (m¼4) supermirror optics in series atthe above angles leads to a low wavelength cutoff of lc¼3.2 A and aQ-range for free-liquid surfaces of 0.058–0.329 A�1.
3.5. Spectra from polarising and analyser supermirrors
Fig. 13(a) shows a spectrum of an unpolarised neutron beamacquired over 600 s, of medium wavelength resolution (discs 1 and3, Dl/l¼4.3%), passing through the ‘‘no guide’’ option and colli-mated through 0.5 mm slits (Dyv¼0.0071) (black profile). Fig. 13(b)(red profile) shows the equivalent spin-down polarised spectrumfollowing transmission through a Fe/Si polarising supermirrororiented at 0.801 to the horizontal. The spin-up component thatis reflected from this polarizer is adsorbed within the collimationsystem, prior to the beam-defining slit before the sample. Fig. 13(c)(blue profile) shows the spin-down (transmitted) polarised spec-trum and Fig. 13(d) (green profile) the spin-up (reflected) polarised
spectrum after an unpolarised neutron beam interacts with theanalyser supermirror oriented at 0.881 with respect to the beamthat enters the detector tank. The spin-down beam transmittedthrough the analyser is essentially the same as that through thepolarising element in the collimation system. The two verticaldashed lines indicate the limits of the spin-polarised beam, leadingto an operational wavelength range of these polarising elements of(l¼2.5 A–12 A). The low-wavelength limit is determined by thehigh, critical-Q value (0.08 A�1) of the m¼3.8 supermirrors. Below2 A both spin states are transmitted; while all neutrons above 14 Aare reflected from the analyser element. The polarised beamreflected from the analyser shows higher intensity over theoperational wavelength range (50% of that of the unpolarisedbeam), compared to than for the beams transmitted by thepolarising element (40%) and by the analyser (37%). The loss inintensity in the transmitted beams are due to adsorption andscattering losses as the beam passes through the Si substrates andsupermirror coatings.
4. Characteristic reflectivity data from Platypus
Neutron reflectometry data from Platypus are generated bydividing neutron beam spectra reflected from the sample by thedirect beam spectra collected using the same disc chopper pairing,under identical conditions (slits settings and supermirror collima-tion optics). In the case of solid/liquid samples, the direct beamspectra are collected via transmission through the solid substrate;while for free-liquid samples the single-bounce mirror provides thedirect beam spectrum for the 1st angle and the double-bounceoptic provides the direct beam spectrum for the 2nd angle (4.81).For polarised neutron reflectometry, various direct beam spectraare measured using different combinations of the polarizer, spin-flippers and analyser. Following division of the reflected beamspectra by the direct beam spectra, these data are then normalisedby neutron monitor counts to account for variation in reactorpower. In this way, one expects a reflectivity value of unity for thosedata below the critical angle of a given sample, as well as correctscaling for those data that are derived from samples without acritical angle. For well aligned samples that are under-illuminated,the instrument measures reflectivity values within 2% of unity forQ-values below the critical edge.
M. James et al. / Nuclear Instruments and Methods in Physics Research A 632 (2011) 112–123 119
4.1. Solid samples
4.1.1. Bare silicon wafer
Fig. 14 shows the reflectivity from a freshly cleaned 100 mmdiameter silicon wafer in air. This sample has very little incoherentscattering and indicates a minimum reflectivity o10�7. Data werecollected at incident angles of 0.51, 21, and 6.71, with a totalacquisition time of 10 h. The medium resolution chopper setting(discs 1 and 3; Dl/l¼4.3%) was used for this measurement alongwith collimation slit settings for s2 and s3 to provide a constantfootprint on the sample (and Dy/y¼1.9%) for all three differentangles. The combined instrumental resolution Dq/q for this mea-surement was 4.7%. The red solid line shows the fit to the observeddata based on a structural model with a 10.0(6) A thick native oxidelayer, a surface roughness of 1.9(5) A and a refined background of5(1)�10�8. The values given in parentheses throughout thismanuscript represent an error of one standard deviation.
4.1.2. Amorphous alumina film on silicon
Fig. 15(a) shows observed (crosses) and calculated (solid redline) neutron reflectivity profiles from an Al2O3 film on silicon
Fig. 14. Observed (points) and calculated (red line) neutron reflectivity from an air/
silicon surface with a 10 A native SiO2 layer. The error bars for the observed data at
each value of Q are derived from the measured neutron counts. (For interpretation of
the references to color in this figure legend, the reader is referred to the web version
data from a 325 nm ALD alumina film on silicon. The error bars for the observed data
at each value of Q are derived from the measured neutron counts. (b) The observed
X-ray reflectivity profile (green line) from the same sample. (For interpretation of
the references to color in this figure legend, the reader is referred to the web version
of this article.)
produced by atomic layer chemical vapour deposition [62]. Discchoppers 1 and 2 were used to provide high wavelength resolutionneutron pulses (Dl/l¼1.2%). Data were collected at two incidentangles (y¼0.51 and 2.01), with collimating slits of 0.3 and 1.2 mm,respectively, (Dy/y¼1.2%) and an acquisition time of 6 h. Theinstrumental resolution (Dq/q) for this measurement was 1.7%. Therefined structure from these data indicates a 3246(2) A thick film, aSLD of 4.3�10�6 A�2 and a surface roughness of 6(1) A. The solidgreen line (Fig. 15(b)) is observed (Cu Ka) X-ray reflectivity datafrom the same sample, with the weaker Kiessig fringes in thisprofile resulting from the relatively poor contrast between thealumina film (2.5�10�5 A�2) and the Si wafer (2.0�10�5 A�2).
4.1.3. Ni/Ti Bragg mirror
Fig. 16 shows the observed neutron reflectivity profile (bluedata) from a multilayer Bragg mirror film consisting of 25 bilayersof Ni and Ti deposited on float glass, and collected using the high-resolution setting (Dl/l¼1.2%). Data were collected at incidentangles of 11, 31, and 51 and slit settings to give a constant beamfootprint (Dy/y¼1.9%) and an instrumental resolution (Dq/q¼2.3%). Fig. 17 shows a 2-dimensional image from the Platypus
detector (scattering angle versus wavelength), indicating thespecular ridge; as well as Yoneda wings along lines of constant Q
associated with diffuse interfacial scattering. Ten Bragg reflectionsare evident in these data, associated with the repeating bilayer. Thestructure refined from these data indicates 113.5(2) A thick Nilayers and 79.8(2) A thick Ti layers, and a total film thickness of4833(15) A. The inset to Fig. 16 shows the Ni critical edge (a),Kiessig fringes associated with the total film thickness (b), and thefirst Bragg reflection (c).
4.1.4. Polarised reflectometry
Spin-polarised neutron reflectivity data were collected from apermalloy (Ni0.8Fe0.2)/Cr bilayer film on Si at room temperature underan applied, horizontal field of 0.2 T. The medium resolution discchopper setting was used (Dl/l¼4.3%). Due to the magnitude ofthe saturating applied field at the sample, the magnetic moments in theplane of the permalloy film were aligned parallel and anti-parallel tothe spin-up and spin-down neutrons, respectively. Fig. 18(a) shows theobserved (points) and fitted (line) reflectivity data for spin-up neutrons(R+, black data) and Fig. 18(b) the equivalent profiles for spin-downneutrons (R� , blue data). As no spin-flip scattering (R+�) was detectedfrom this sample, these reflectivity profiles were identical to thosecollected using polarisation analysis (i.e. R+= R+ + and R� = R��). Dueto the narrower wavelength bandwidth (l¼2.5–12 A) available fromthe polarising supermirror in the collimation system, three angles ofincidence were required (0.41, 1.01, and 2.51) to collect suitablereflectivity data. Collimating slits of 0.42, 1.05, and 2.62 mm wereused for these three angles (Dy/y¼2.2% and Dq/q¼4.8%), with a totalacquisition time of 5½ h per spin state. The structure of this filmwas determined by simultaneous refinement of both spin-polariseddatasets, giving a 16(1) A native SiO2 layer, a 297(1) A permalloy layer,a 16(1) A Cr capping layer and 15(1) A Cr2O3 surface layer. The twospin dependant reflectivity profiles (R+ and R�) were co-refinedwith a magnetic scattering length density of 1.8(1)�10�6 A�2 forthe permalloy layer, which was subtracted from the nuclear SLD(8.8(1)�10�6 A�2) for spin-down neutrons and added to the nuclearSLD for spin-up neutrons. The nuclear component leads to a massdensity of 8.4(1) g/cm3 for the permalloy film; while the magneticcomponent indicates a magnetic moment per atom of 0.77(1)mB.
4.2. The solid–liquid interface
Fig. 19 shows neutron reflectometry data collected at the silicon-water interface using the low-resolution setting (Dl/l¼8.4%) and
Fig. 16. Observed (blue points) and calculated (red line) neutron reflectivity profiles from a 25 bilayer Ni/Ti Bragg mirror. The error bars for the observed data at each value of Q
are derived from the measured neutron counts. The inset shows (a) the critical edge, (b) the first Bragg peak, and (c) Kiessig fringes associated with the total film thickness
(483 nm). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Fig. 17. Specular (horizontal data) and off-specular signal from a Ni/Ti Bragg mirror,
showing Bragg peaks as well as Yoneda scattering along lines of constant Q.
Fig. 18. Observed (points) and calculated (line) polarised neutron reflectivity
profiles from a Cr capped 30 nm permalloy film on silicon under a 0.2 T applied
field. (a) Measured using spin-up neutrons (R+, black data), and (b) measured using
spin-down neutrons (R� , blue data). The error bars for the observed data at each
value of Q are derived from the measured neutron counts. (For interpretation of the
references to color in this figure legend, the reader is referred to the web version of
this article.)
M. James et al. / Nuclear Instruments and Methods in Physics Research A 632 (2011) 112–123120
accumulated over 4 h. Fig. 19(a) (blue data) is that from the D2O/silicon interface, with beams directed through an 80 mm long siliconblock at 0.81 and 3.01, using collimation slits to give a constantfootprint (and Dy/y¼0.023) and an instrumental resolution of(Dq/q¼8.7%). Fig. 19(b) (red data) shows the observed and fittedreflectivity profiles from a chain-deuterated 1,2-dimyristoyl-sn-gly-cero-3-phosphocholine (d54-DMPC) bilayer adsorbed at the silicon/D2O interface at 32 1C using the same angles. Fig. 19(c) (black data)shows the observed and fitted reflectivity profiles from the same d54-DMPC bilayer, following exchange of the D2O for a H2O/D2O mixtureof SLD 4.0�10�6 A�2 (‘‘CM4’’). These data were collected at slightlylower angles (0.61and 2.81), using the same instrumental resolution inorder to measure the critical edge for this lower contrast.
The model for the d54-DMPC bilayer was generated by jointlyrefining the D2O and CM4 contrast data using the global fittingroutine in Motofit [60]. The thickness, SLD, interfacial roughnessand the amount of water associated with each layer were
constrained during the fitting. The refined model was consistentwith previous studies of Refs. [13,14], showing a 13(1) A thicknative oxide layer, a 8(1) A hydrated phosphocholine head groupadjacent to the Si (with 51% water in this layer), a 26(1) A thickregion describing two sets of deuterated lipid tails (containing 9%water), and an 8(1) A hydrated phosphocholine head group at theD2O interface (with 37% water). In line with these studies ofadsorbed phospholipid bilayers [13,14], attempts were made torefine this structure with an additional �5 A water layer imme-diately adjacent to the SiO2 surface, however these refinements didnot generate an improvement in the fit.
4.3. Free-liquid samples
Fig. 20 shows observed and fitted reflectivity profiles taken fromtwo different free-liquid surfaces: D2O (Fig. 20(a)) and a monolayer of
M. James et al. / Nuclear Instruments and Methods in Physics Research A 632 (2011) 112–123 121
protonated 1,2-dimyristoyl-sn-glycero-3-phosphocholine (h-DMPC)spread on D2O from chloroform solution (Fig. 20(b)). For thesemeasurements, the liquids were contained in sealed Teflon troughsto prevent evaporation and mounted on a Halcyonics MOD-2 anti-vibration stage. For both measurements, the medium resolutionsetting was used (Dl/ll¼4.3%). The first incident beam angle onthe sample for each of these datasets was 1.01, with the beam beingdeflected downwards at 0.501 using the single-bounce supermirror inthe Platypus collimation system. The second angle onto the sample forboth datasets was 4.81, generated by the double-bounce deflectionmirror. Collimation slits were chosen for these two angles to provide aconstant beam footprint on the sample, with a Dy/y of 1.2% and aninstrumental resolution (Dq/q) of 4.4%. The total acquisition time forthese datasets was 7 h each.
The reflectivity data from the air-D2O interface (Fig. 20(a)) weremodelled by fitting only the instrumental background (2.3(2)�10�6)and the surface roughness 2.5(6) A. The refined structure usingreflectivity data from the h-DMPC phospholipid monolayer on D2O(Fig. 20(b)), indicates a 11(1) A thick surface layer with a SLD of3.89�10�6 A�2, and is consistent with the head group of this lipidmonolayer in the condensed state (with a volume fraction of 0.56) at
Fig. 19. Observed (points) and calculated (lines) neutron reflectivity profiles from
(a) the D2O/silicon interface (blue data) a d54-DMPC phospholipid bilayer adsorbed
at (b) the D2O/silicon interface (red data) and (c) the ‘‘CM4’’ water/silicon interface
(red data). The error bars for the observed data at each value of Q are derived from
the measured neutron counts. (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of this article.)
Fig. 20. Observed (points) and calculated (lines) neutron reflectivity profiles from
(a) the D2O/air interface (blue data), and (b) a DMPC phospholipid monolayer
adsorbed at the D2O/air interface (red data). The error bars for the observed data at
each value of Q are derived from the measured neutron counts. (For interpretation of
the references to color in this figure legend, the reader is referred to the web version
of this article.)
the air–water interface [20,21]. Attempts to refine this lipid monolayerusing a 2-layer model to represent the head-groups and hydrocarbontails led to unstable refinements. The protonated hydrocarbon tailchains of DMPC have an expected SLD of �0.37�10�6 A�2, givingpoor neutron contrast against air and are essentially invisible.
4.4. Off-specular reflectivity
The off-specular capabilities of Platypus have recently been usedto characterise neutron diffraction by travelling surface acousticwaves (SAW) generated piezoelectrically on a quartz surface. Thiswork forms part of a project to explore the capabilities of suchdevices as neutron beam handling components, and builds on anearlier study by Hamilton et al. [63]. The data described below werecollected for 8 h using the medium resolution setting (Dl/l¼4.3%),and collimating slits 1.27 mm high giving an angular resolution(Dy/y) of 3.3%. The instrumental resolution in the specular direction(Dq/qspec) is 5.4%. In the off-specular direction, the angular resolutionhas an additional contribution due to the (vertical) wire spacingon the Denex detector, giving a total angular contribution of (Dy/y¼4.9%), and an off-specular instrument resolution (Dq/qoff-spec) of 6.5%.Fig. 21(a) shows the total angle of deviation versus wavelength forneutrons with an incident angle of 0.751 and a SAW of wavelengthL¼91.5 mm travelling against the incident beam direction at avelocity (vSAW) of 3158 m s�1 (normalised against the Platypus
incident beam spectrum). On either side of the specularly reflectedbeam (ai+af¼1.51) are sweeping arcs of first order grating diffractionsignals. Since the SAW distortion of the surface is travelling, thediffraction angles observed here are magnified by a Doppler factor,essentially the ratio of the neutron velocity parallel to the surface inthe frame in which the SAW distortion of the surface is stationary tothe lab frame value. For diffraction order n, in terms of the angles ofincidence ai and exit af of neutrons from the surface we may write:
sin2afffisin2ai�2nðl=LÞ 1�vSAW=v½l�� �
,
where the last term in brackets is the Doppler factor. For neutronswith wavelengths in the range shown (l¼4–20 A) the neutronvelocity n[l] changes from �1000 to �200 m s�1; values muchlower than vSAW . For the surface acoustic wave travelling toward theneutron beam the Doppler factor varies from about 4 to 17 over thisrange resulting in much greater diffraction angles (solid line fit) thanwould result from a stationary grating of the same periodicity(dashed fit lines). Converting the upper section of the dataset to ascattering vector map (QJ,Q?) (Fig. 21(b)), this magnification isapparent in the much larger QJ transfer than one would expectfrom a stationary grating of periodicity of 76.9�10–6 A�1 near thespecular beam at QJ¼0.
5. Published studies using the Platypus reflectometer
5.1. Microphase separation in diblock copolymer thin films
The first published manuscript using Platypus data by Sripromet al. examined the effect of block polydispersity on microphaseseparation of PMMA/PBA thin films [64], and used neutronreflectometry to confirm the lamellar structure of a specificcomposition (PMMA77–PBA23). In particular, neutron reflectome-try was used to help clarify the issue of whether symmetric (withthe same block being present at the substrate and surface) orasymmetric lamellae structures (with different blocks at the sub-strate and air interfaces) were formed. After spin coating thediblock copolymer onto Si wafers, microphase separation wasinduced by annealing at 180 1C for 16 h. Atomic Force Microscopy(AFM) measurements on PMMA77–PBA23 films of pre-annealingthicknesses 48–61 nm and 73–92 nm showed featureless surfaces
Fig. 21. (a) Total angle of deviation versus wavelength for SAW diffraction of neutrons incident at 0.751. Solid curved lines are the expected first order diffraction angles for a
grating of periodicity L¼91.5 mm moving toward the neutron beam at the acoustic velocity 3158 m s�1. Dashed lines near the reflected beam show the expected diffraction
signal for a stationary grating; (b) Data in the region of the n¼�1 and reflected beam regions mapped against wave vector transfers parallel and normal to the reflecting quartz
surface (QJ,Q?).
M. James et al. / Nuclear Instruments and Methods in Physics Research A 632 (2011) 112–123122
suggestive of lamellae formation. In the region 62–72 nm,hole/island morphology was observed. Platypus reflectivity datacollected on an annealed PMMA77–PBA23 diblock copolymer film,indicated a thickness of 576 A. Although no deuteration was used,sufficient SLD contrast was available between the different blocks(SLDPMMA�1.05�10�6 A�2, SLDPBA�0.55�10�6 A�2) to revealan asymmetric lamellar structure containing four complete alter-nating layers, with the more polar PMMA preferentially at the Siwafer and the PBA block segregating to the air interface due to itslower surface energy.
The second published article from Platypus [65] examined thestructure of thin films of the microphase separated diblockcopolymer PS35–PEO65. The combination of patterning abilityand biological compatibility has made block copolymers of PEOamong the most promising materials for the design of bio-inter-faces. Numerous previous studies dating back more than 30 yearshave supported the notion that PS35–PEO65 diblock copolymersphase separate with the more polar PEO block present at both the Sisubstrate and air interfaces. In this study examining a PS35–PEO65
film annealed at 80 1C, optical microscopy and AFM revealedsurface topologies and crystallinity suggestive of a PEO surfacelayer; however X-ray Photoelectron Spectroscopy (XPS) suggestedthe presence of a thin layer of PS at the surface. Neutron reflectivitydata was used to determine the internal morphology and composi-tion of the phase-separated PS35–PEO65 film showing a hole/islandlamellar structure. Deuteration of the PS was not required for thisstudy, as sufficient contrast was available between the differentblocks (SLDPS¼1.51�10�6 A�2, SLDPEO¼0.82�10�6 A�2). Thetotal film thickness was found to be 468(5) A, comprised of a157(2) A thick, continuous PEO/PS bilayer adjacent to the Si sur-face; and islands (of thickness 312(3) A) formed of a full lamella inthe sequence PS/PEO/PS. The structure of the refined uppermostlayer (61(1) A thick, SLD¼1.03(1)�10�6 A�2) clearly indicatedthe presence of PS at the film/air interface.
5.2. Structural studies of modified homopolymer thin films
In the most recently published Platypus paper based on polymerthin-films, Telford et al. [66] investigated the cross-linkingmechanism that spontaneously occurred in (normally water solu-ble) PNVP thin films upon thermal annealing. The degree of cross-linking of PNVP films and their solubility in water were tailored bycontrolling the annealing conditions. The combination of neutron
and X-ray reflectometry was used to follow the evolution of PNVPfilm structures with variation in annealing temperatures andannealing times; and in addition gave information about thesolubility of fully and partially cross-linked films. These datashowed a contraction in the film thickness and an increase inSLD as a result of annealing. Analysis of Platypus data revealed theextent of contraction (from 506(3) to 370(2) A) and a densificationof the PNVP films following complete cross-linking (annealing at200 1C for 3 h). The SLD increased from 0.94(1)�10�6 A�2 for theas-prepared film to 1.98(2)�10�6 A�2 for the cross-linked film.Due to the fact that the composition of these films change uponannealing, the refined SLD based on neutron reflectivity data alonewas not sufficient to decouple the composition from the massdensity. When the neutron reflectivity data was combined withX-ray reflectivity and XPS data (giving the ratio of C:N:O in the film)the combination of the three techniques was able to uniquelydetermine the composition and mass density of the as-prepared(C6H9NO; 0.81 g/cm3) and the fully cross-linked (C6H6.4N0.9O1.3;1.15 g/cm3) films. While the increase in oxygen content suggestedsome oxidation of the PNVP during cross-linking, the substantialdecrease in hydrogen content pointed towards a cross-linkingmechanism involving PNVP radicals.
5.3. Molecular sensing in dendrimer thin films
The correlation between photoluminescence and the diffusionof an explosive vapour analyte into solid-state dendrimer thin-films was explored by Cavaye et al. [67] using Platypus and acustom-build cell containing a fluorescence spectrometer and a365 nm UV-LED light source. Their study examined the structureand molecular sensing capability of thin (�39 nm) and thick(�100 nm) films of first-generation dendrimers comprised of2-ethylhexyloxy surface groups, biphenyl-based dendrons and afluorescent 9,9,90,90-tetran-propyl-2,20-bifluorene core. Thesedendrimer films were shown to rapidly and reversibly detectexplosive analyte vapours such as para-nitrotoluene by oxidativeluminescence quenching. Combined photoluminescence and neu-tron reflectometry measurements were carried-out on pristinedendrimer films, with the reflectivity data fitted by single layermodels of thickness and SLD of 389(1) A and 1.00(4)�10�6 A�2 forthe thin film, and 999(1) A and 0.99(4)�10�6 A�2 for the thickfilm. Good contrast was afforded between the dendrimer filmsand perdeuterated para-nitrotoluene (6.51�10�6 A�2), and after
M. James et al. / Nuclear Instruments and Methods in Physics Research A 632 (2011) 112–123 123
allowing the dendrimer films to saturate with analyte vapour, bothfilms increased in thickness by 4% and the photoluminescence wascompletely quenched. The Platypus reflectivity data could no longerbe adequately fitted by a single layer model, but revealed a SLDgradient throughout the films. The average amount of analytewithin the dendrimer films was determined from these reflectivitymeasurements, suggesting dendrimer:analyte ratios of 3.5 and1.7 for the thin and thick films respectively. Blowing on thesedendrimer films with nitrogen gas removed the analyte, returningthem to their original thickness and scattering length density, withthe restoration of the photoluminescence.
References
[1] R. Cubitt, G. Fragneto, Appl. Phys. A 74 (Suppl.) (2002) S329.[2] /http://www.ill.eu/instruments-support/instruments-groups/instruments/
figaro/S.[3] V.-O. de Haan, J. de Blois, P. van der Ende, H. Fredrikze, A. van der Graaf,
M.N. Schipper, A.A. van Well, J. van der Zanden, Nucl. Instr. and Meth. Phys. Res.A 362 (1995) 434.
[4] R. Kampmann, M. Haese-Seiller, V. Kudryashov, B. Nickel, C. Daniel, W. Fenzl,A. Schreyer, E. Sackmann, J. Radler, Physica B Condens. Matter 385–386 (2006)1161.
[5] M. James, A. Nelson, A. Brule, J.C. Schulz, J. Neutron Res. 14 (2006) 91.[6] R.K. Thomas, J. Penfold, Curr. Opinion Colloid Interface Sci. 1 (1996) 23.[7] T.P. Russell, A. Karim, A. Mansour, G.P. Felcher, Macromolecules 21 (1988)
1890.[8] C.F. Majkrzak, Physica B: Condens. Matter 173 (1991) 75.[9] G.P. Felcher, K.E. Gray, R.T. Kampwirth, M.B. Brodsky, Physica B and C 136
[21] S.J. Johnson, T.M. Bayer, W. Weihan, H. Noack, J. Penfold, R.K. Thomas,D. Kanellas, A.R. Rennie, E. Sackmann, Biophys. J. 60 (1991) 1017.
[22] M. Kreiner, C.R. Chillakuri, P. Pereira, M. Fairhead, Z. Li, H.J. Mardon, S.A. Holt,C.F. van der Walle, Soft Matter 5 (2009) 3954.
[23] M. Al-Jawad, G. Fragneto, J. Liu, S.R. Chang, B. Clarkson, Eur. Phys. J. E 30 (2009)175.
[24] F. Evers, R. Steitz, M. Tolan, C. Czeslik, J. Phys. Chem. B 113 (2009) 8462.[25] M. Delcea, R. Krastev, T. Gutberlet, D. Pum, U.B. Sleytrd, J.L. Toca-Herrera, Soft
Matter 4 (2008) 1414.[26] J. Koo, T. Gutberlet, C. Czeslik, J. Phys. Chem. B 112 (2008) 6292.[27] A.P. Le Brun, S.A. Holt, D.S. Shah, C.F. Majkrzak, J.H. Lakey, Eur. Biophys. J. 37
(2008) 639.[28] I. Burgess, M. Li, S.L. Horswell, G. Szymanski, J. Lipkowski, J. Majewski, S. Satija,
Biophys. J. 86 (2004) 1763.[29] A. Glidle, A.R. Hillman, K.S. Ryder, E.L. Smith, J.M. Cooper, N. Gadegaard,
J.R.P. Webster, R.M. Dalgliesh, R. Cubitt, Langmuir 25 (2009) 4093.[30] J.M. Cooper, R. Cubitt, R.M. Dalgliesh, N. Gadegaard, A. Glidle, A.R. Hillman,
R.J. Mortimer, K.S. Ryder, E.L. Smith, J. Am. Chem. Soc. 126 (2004) 15362.[31] J.J. Noel, D.W. Shoesmith, Z. Tun, J. Electrochem. Soc. 155 (2008) C444.
[32] J.J. Noel, H.L. Jensen, Z. Tun, D.W. Shoesmith, Electrochem. Solid-State Lett. 3(2000) 473.
[33] Y. Lauw, T. Rodopoulos, M. Gross, A. Nelson, R. Gardner, M.D. Horne, Rev. Sci.Instrum. 81 (2010) 074101.
[39] K.H. Lee, P.E. Schwenn, A.R.G. Smith, H. Cavaye, P.E. Shaw, M. James, K. Kruger,P. Meredith, I.R. Gentle, P.L. Burn, Morphology of all-solution-processed bilayerorganic solar cells, Adv. Mater., 2010, in press, doi:10.1002/adma.201003545.
[40] S.J. Martin, R.A.L. Jones, M. Geoghegan, A.M. Higgins, I. Grizzi, J.J.M. Halls,S. Kirchmeyer, R.M. Dalgliesh, Phys. Rev. B 71 (2005) 081308 R.
[41] P. C. Jukes, S. J. Martin, A. M. Higgins, M. Geoghegan, R.A.L. Jones, S. Langridge,A. Wehrum, S. Kirchmeyer, Adv. Mater. 16 (2004) 807.
[42] S.V. Vickers, H. Barcena, K.A. Knights, R.K. Thomas, J.-C. Ribierre, S. Gambino,I.D.W. Samuel, P.L. Burn, Giovanna Fragneto, Appl. Phys. Lett. 96 (2010)263302.
[43] T. Saerbeck, F. Klose, D. Lott, G.J. Mankey, Z. Lu, P.R. LeClair, W. Schmidt,A.P.J. Stampfl, S. Danilkin, M. Yethiraj, A. Schreyer, Phys. Rev. B 82 (2010)134409.
[44] J. Orna, P.A. Algarabel, L. Morellon, J.A. Pardo, J.M. de Teresa, R. Lopez Anton,F. Bartolome, L.M. Garcıa, J. Bartolome, J.C. Cezar, A. Wildes, Phys.Rev. B 81(2010) 144420.
[45] B.J. Kirby, J.E. Davies, S.M. Kai Liu, G.T. Watson, R.D. Zimanyi, P.A. Shull,J.A. Kienzle, Phys. Rev. B 81 (2010) 100405 R.
[46] D. Lott, F. Klose, H. Ambaye, G.J. Mankey, P. Mani, M. Wolff, A. Schreyer,H.M. Christen, B.C. Sales, Phys. Rev. B 77 (2008) 132404.
[47] A.J. Drew, M.W. Wisemayer, D.O.G. Heron, S. Lister, S.L. Lee, A. Potenza, Phys.Rev. B 80 (2009) 134510.
[48] V.F. Kozhevnikov, C.V. Giuraniuc, M.J. Van Bael, K. Temst, C. Van Haesendonck,T.M. Mishonov, T. Charlton, R.M. Dalgliesh, Yu.N. Khaidukov, Yu.V. Nikitenko,V.L. Aksenov, V.N. Gladilin, V.M. Fomin, J.T. Devreese, J.O. Indekeu, Phys. Rev. B78 (2008) 012502.
[49] A. Hoffmann, S.G.E. te Velthuis, Z. Sefrioui, J. Santamarıa, M.R. Fitzsimmons,S. Park, M. Varela, Phys. Rev. B 72 (2005) 140407 R.
[50] R. Springell, S. Langridge, A. Wildes, S.B. Wilkins, C. Sanchez-Hanke, K.T. Moore,M.T. Butterfield, J. Chivall, R.C.C. Ward, M.R. Wells, G.H. Lander, Phys. Rev. B 81(2010) 134434.
[51] K. Theis-Brohl, A. Westphalen, H. Zabel, U. Rucker, J. McCord, V. Hoink,J. Schmalhorst, G. Reiss, T. Weis, D. Engel, A. Ehresmann, B.P. Toperverg,New J. Phys. 10 (2008) 093021.
[52] F.Y. Ogrin, S.M. Weekes, B. Cubitt, A. Wildes, A. Drew, C.A. Ross, W. Jung,R. Menon, B. Toperverg, IEEE Trans. Magn. 43 (2007) 2731.
[53] Christine Hamann, Jeffrey McCord, Ludwig Schultz, Boris P. Toperverg,Katharina Theis-Brohl, Maximilian Wolff, Rainer Kaltofen, Ingolf Monch, Phys.Rev. B 81 (2010) 024420.
[54] /http://lns00.psi.ch/sics/design/sics.htmlS.[55] /http://www.nexusformat.org/Main_PageS.[56] A. Nelson, J. Phys.: Conf. Ser. 251 (2010) 012094.[57] /http://www.wavemetrics.com/S.[58] A.A. Van Well, H. Fredrikze, Physica B 357 (2005) 204.[59] E.M. Forgcan, R. Cubitt, Neutron News 9 (1998) 25.[60] A. Nelson, J. Appl. Crystallogr. 39 (2006) 273.[61] /http://www.ansto.gov.au/research/bragg_institute/facilities/sample_en
2770.[64] W. Sriprom, M. James, S. Perrier, C. Neto, Macromolecules 42 (2009) 3138.[65] C. Neto, M. James, A. Telford, Macromolecules 42 (2009) 4801.[66] A.M. Telford, M. James, L. Meagher, C. Neto, Appl. Mater. Interfaces 2 (2010)
2399.[67] H. Cavaye, A. Smith, M. James, A. Nelson, P.L. Burn, I.R. Gentle, P. Meredith,
