Reference Samples for Neutron Reflectometry
X-rays and neutrons optics action.
Reference Samples for Neutron Reflectometry.
For limited diffusion only
Version 0.4, 05/14/2019
Description of the project
This initiative consist in providing reference samples for new reflectometers built across Europe and "qualify" them. The motivation of this initiative is to generate a library of reference data made available for the tests of fitting programs. This database will eventually p
rove the reliability and demonstrate the reproducibility between the different neutron reflectometers. This is a necessary effort to convince material scientists / industries of the versatility and reliability of the neutron reflectivity technique.
A lot of reflectivity simulation packages are presently available. Each of them answers to specific needs. Until now there has been no coordination between these development. However it is certainly hopeless to aim at a different development model since the present way of working allows to test new ideas and to include new features specific to some problems. This p
arallel development pushes new ideas.
The most advanced merging initiative is the
XOP – NOP package (ESRF – ILL) however this package is not suitable for special/specific needs
These software packages are tested/validated in a unknown way. The user may have doubts/questions. There is no clear qualification process.
Only the intensive use of these packages may pinpoint towards potential problems
However, the neutron user
base is not so big and it can take months before a problem is identified.
The process of the initiative is the following :
- Fabrication of a set of samples on which « a number » of interested scientists have agreed
- Dispatching of the samples in the involved/interested facilities
The « supposed » sample structure is provided to the « users »
These samples will be made freely available to anybody making the request
(one requirement/condition will be that the measured data will be provided back into the database)
- The measured data are them grouped back into a database
- Real data can then be used by people developpping simulation/fitting software to qualify their software
in this case providing the results of the fits is strongly encouraged
Once a number of real measurements and fits will be made available, this will enables to pin points possible « problems », either :
- in the measurement process
- in the fitting process
A full loop can then again been performed : measurement – comparison - fitting – comparison
In the end this database will qualify both instrument and fitting tools and make reflectivity seen as a « reliable » technique for material scientists.
Rules of the project
- Any group/person asking for a reference sample will get one. In exchange, they will have to provide the measured data for the experimental “database”.
- The “Experimental data” will be made freely available on a web site.
- People using the experimental data to test/validate their code are invited to provide information about their code (Web links, output fit from their program, etc...)
- A final report comparing all the measurements will eb made available after most of the participants have given their results
- At least 4 to 6 identical samples are needed
- 1 kept safely
- 3 circulating
- 1 lost
- 1 broken
- Large size samples are required (to fit all instruments)
- 5 x 5 cm² or 5 x 10 cm² is foreseen
- easy measurements on any type of reflectometer
- This is not a problem with existing deposition facilities.
- Robustness against corrosion and aging is required
- proper choice of materials (avoid bare glass or Si)
- The roughness should be as small as possible : roughness < 0.4 nm
- The roughness issues could be tested in a second round if this first round is successful.
- The samples must be able to easily travel.
The ILL (via Ken Andersen) has offered to prepare reference samples.
The ILL has a recognized expertise and will be able to provide a long term support. It is also “The” European neutron facility. Large sputtering systems are available as well as manpower.
The presently available materials in the sputtering systems are :
- Fe, Co, Si, Ti, Ni, Gd, C and W (October 2003).
Thus the reference samples should use a combination of these materials.
Thick glass substrates will be used so as to have large substrates as well as very flat surfaces.
The samples will be deposited in a single run by sputtering.
The sample will be compared by NR after fabrication to check for differences (in particular thickness homogeneity).
The samples should also be cross checked (on the same instrument) every year (or every 2 years?) to check the aging issues.
Definition of possible reference samples
Neutron reflectivity encompasses a very wide range of samples. We have defined 6 main categories:
Lowest measurable reflectivity - background noise
Thick glass substrate (flat) + thin nickel coating (2nm)
bare substrate (glass or Si) is ruled out for its poor stability over time
Simple system : polymers
2 layers : Si substrate // grafting layer / spin coated polymer
« Grafted from » / « grafted onto » polymers for fine thickness control
Issue of very thin films/ ultra-thin films
Glass// Ti(2) / W(100) / C(10) / W(100)
6-8 layers (apply to « industrial systems, e.g. GMR)
Such sample can be fabricated over large surfaces (6in wafers)
Substrate glass, trilayer Ni/Cu/Co/Ni
Too difficult to prepare - discarded
The aging process should be assessed and controlled from time to time
Neutron optical element
The type of optics needs to be defined : supermirror, monochromator… ? (something standard anyway)
These categories are indicative and are by no mean “final”
. We need further approval from other groups to define the “final list”.
Dissemination of the results
The results of this initiative should be made available to the wider community.
At the moment, a “Collaborative web site” has been set-up at the following address :
This is a collaborative web site of WIKI type (more information can be found at
This type of Web site can be used to generate a free aggregation of knowledge. Anybody can update the site to provide extra information. The web interface makes it m
uch easier to use than pure html. The access is straightforward.
In the case of our site, an e-mail registration is required for security reasons. It also useful to track people expressing interest in the site. Once you are registered, you are free to add materials (pages, data, etc…). Some effort is need to understand the basic rules but these rules are rather straightforward.
We hope that this type of “free” interface will make it easier for people to interact with the site and will lead to frequent updates.
Teams involved / having expressed interest
Laboratoire Léon Brillouin CEA/CNRS
CEA Saclay 91191 Gif sur Yvette FRANCE
Contacts : Frédéric
Institut Laue Langevin
FZJ U. Ruecker
KFKI L. Rosta
Greece N. Mergia
Levels of participation
We expect to have different levels of participation in the project.
A/ Passive Observer
Most people will be just be passive observer who will follow the output of the initiative. This will mainly include neutron reflectivity users.
B/ Active observer
People willing to provide input to the project in the form of suggestion and requests. This includes suggestion for the choice of “reference” samples. Request for fitting possibilities. This is likely to involve regular reflectivity users who have specific needs.
C/ Sample measurements
This includes the facilities providing “reference data” to the database. This will mostly involve neutron facilities and CRG groups.
D/ Data fitting
This included the people providing fitting results. The category will be mostly connected to the category C/ but we expect that extra people may join the initiative (since reflectivity is a rather general technique).
E/ Samples Preparation
This last category should be rather limited since we do not intend to produce too many type of reference samples. At the moment, this is limited to the ILL. In the case of “soft matter” systems, this category should be extended to a second lab which has recognized expertise in the field.
F/ Project supervising
A small number of labs should provide input concerning the project logistic (sending sample), gathering data, etc… Any lab willing to provide some input is welcome.
These different categories are of course permeable and the input from different labs will evolve with time. This splitting should illustrate the flexible level of participating that the project offers.
The first round of tests will be restricted to the study of “specular” reflectivity.
If the output of this initiative is successful, similar test might be performed to check other issues such as :
- Off-specular scattering – Roughness
- Magnetic and non magnetic
- BUT the number of off-specular codes for magnetic systems is very limited.
- The neutron fluxes are low are do not easily allow an easy off-specular characterisation
- The range of application are not so wide
- More difficult to motivate « material science » people
- Qualification of « optics »
The present status is the concerted definition of reference samples.
Since it is not easy to make a fully democratic decision, any positive input is welcome and when we consider that a critical mass of input has been reached, we shall go ahead.
Once the project has provided some ouput, we hope that more people/groups will aggregate.
The project has not started yet!Annexes – additional information
Some basic rules for such an initiative
Adapted from ISO/VARMA projects (
- Ensure that there is a well defined and focused objective.
- A route for the output into a Standards Development Organization (SDO) for the final document should be identified as early as possible.
- It is helpful to prepare a framework at an early stage of the final document it is intended to produce. This framework can then be reviewed and updated as the activity progresses.
- Run a smaller exercise initially with a limited number of participants.
- Keep the work expected from each participant down to a maximum of about one man week.
- Ensure as far as possible that the material used is consistent and uniform.
- Give the participants extra material to allow for trials and errors.
- Ensure that the instructions are written simply and clearly.
- Do not add too many "interesting" side issues.
- Start with no fewer than 6-8 participants to allow for a number who will not respond.
- Make it clear that you will adhere to deadlines for submission of data.
- Expect that the analysis will probably have to cover issues not originally anticipated.
- Reports and data release should be made in a timely fashion.
- All participants should have the opportunity to review and correct a draft version of a final report.
- Consult a statistician if there is any doubt about experimental design or treatment of data.
Please check out that we are following these advice.
Several initiative with similar goals do exist. We shall mention the following ones :
- Standard measurement methods for material science
- The project has supported the
- SCANS (Software for Computer Aided Neutron Scattering)
- This project has provided a route for the unification of Monte-Carlo simulations
- « Virtual experiments »
- The initiative is presently being defined.
- It will deal with nanoscience and nanotechnology using Neutron and Synchrotron Radiation and aim at material scientist
- Normalisation “processes” should be included in this initiative
Existing reflectivity software
- Neutron Reflection Analysis DRYDOC and LPROF by Adrian Rennie
- AFIT written by Paul Thirtle (3.1) is available from Oxford.
- Description of Norm Berk's Programs and Papers
- Description by Jeff Penfold of Programs at ISIS
- SURFace - A Package for Analysis of Reflectivity Data J. R. P. Webster and S. Langridge, RAL
- Neutron Reflection Analysis Programs Used at the ILL
- MultiFit and other Software for X-Ray Reflectivity and Polarised Neutron Reflectivity, K. Ritley & W. Donner, MPI f. Metallforschung, Stuttgart, Germany
- Calculate neutron reflectivity on the Web. Alan Munter has provided a WWW site at the NIST Center for Neutron Research.
- Reflec95 (for W95/NT) and ReflecMCA (Mathematica) programs for simulation of reflection from magnetic layers available at the LLB, Saclay, France.
- Description of Parrat32 for multilayer reflectivity calculation and fits. Available from BENSC, Germany.
- Description of SERF Spreadsheet Software from the University of Delaware
- Homepage for GUIDE/WALL programs to calculate transport in neutron guides and reflectivity of interfaces. These are written by H. M. Shimizu (RIKEN) and Y. Suda (Hokkaido University).
- IMD - an IDL package by David L Windt for optical reflectivity calculations
http://cletus.phys.columbia.edu/~windt/idl/imd/ A program for analyzing Grazing Incidence Small Angle X-ray Scattering on nanostructures by R. Lazzi, F. Leroy, C. Revenant-Brizard and G. Renaud, Grenoble.
Existing neutron reflectometers
A rather exhaustive list of the existing reflectometer has been compiled at the following address
There are presently at least 45 neutron reflectometers in the world, among which about 25 are based in Europe.
There are a few other ones at smaller sources (Korea – India – Algeria – Egypt …)
Expected reflectivities from the reference samples
1. Basic substrate
- Glass substrate + 3nm of gold coating (roughness 0.3nm)
- Glass substrate are available as ultra-flat and ultra smooth substrates
- The gold coating will prevent the aging of the glass surface.
2. Polymer layer
- Deuterated polystyrene (10nm) on silicon
- The layer could be made thicker
- The system should be made more complex (but still reproducible)!
- Aging issues?
3. Ultra-thin films
- The choice of Carbon is not the most suitable (however, a non magnetic element is required)
4. Polarised neutrons
- Substrate glass, trilayer glass//Ni/Cu/Co/Ni
- The reflectivity should not decrease too quickly to allow a feasible measurement on most spectrometers
- This sample should not be too complicated and should allow to test a specific issue.
5. Complex system
- 6-8 layers (apply to « industrial systems, e.g. GMR)
- Such sample can be fabricated over large surfaces (6in wafers)
- Can be provided by an industrial partner
6. Liquid samples
Too difficult to prepare - discarded
- Motivation :
- provide reference samples for new reflectometers built across Europe and "qualify" them
- generate a library of reference data made available for the tests of fitting programs
- prove the reliability and demonstrate the reproducibility between the different reflectometers available across Europe. This step being a necessary effort to convince material scientists of the versatility and reliability of the reflectivity technique.
- The "Reference Samples" project
- Link to YOUR SUGGESTIONS
We have started some measurements on the "reference" samples.
Among these are :
* NiO(2.5nm)/glass substrates prepared by Ken (there are a 5 others left at the ILL)
These were intended to probe very thin layers (max Q range ; background noise etc...)
The NiO was deposited to avoid alteration of the glass with time
* Magnetic GMR systems
A real "industrial" GMR systems.
I have 7 samples 20x20mm.
These are very complex systems (7 layers) and are intended to validate fitting programs
* thin poly-électrolyte samples (deposited by adsorption)
This is presently the only way we have found to deposit reproducible "polymer" layers on large samples
If you have any any for preparing reproducible thin film (<30nm) of polymers, your ideas are welcome
Maybe Giovanna has better ideas
* In addition, I suggest that we plan a set of average complex samples
(Ni/Ti/Ni/glass) for example (precise structure yet to be defined)
* + perhaps the equivalent of an optical element such as a [Ni/Ti]20 multilayer
These are not available yet because the deposition machine at the ILL are under heavy load
Do you think you could find a few hours in the coming months to measure some of these samples?
This page is fully opened to your comments and suggestions (even anonymous)
bare substrate (Si + top Si oxide)
* These substrates are available in larges sizes (8in) and a number of samples can be cut out of a single wafer
* This will ensure the homogeneity of the samples
* The aging process should be assessed and controlled from time to time (yearly?)
o bare Glass is ruled out for its poor stablity over time
o for the other samples we suggest to use thick glass substrates (as the one used in the fabrication of super-mirrors)
o Thick glass substrates wil ensure that even the sample may have a large surface, they will remain flat
* This simple sample would be useful to assess backgorund noise issues...
* 2 layers
issue of very thin films/ ultra-thin films
sample for polarised neutrons
Substrate glass, trilayer Ni/Co
One requirement : at the moment, one should avoid to introduce the roughness issues and the samples should be prepared with the greatest care (roughness < 4A°)