The IBA DataFurnace Information for Users

Current Version

Executable updates are:

Manuals

Other Documents

A full DataFurnace Review has been completed (September 2002:  you can download a version of this, 4MB PDF).  This is published as a Topical Review in J.Phys.D 36 (2003) R97-R126

DataFurnace Practical Course

At the Joint ICTP/IAEA Workshop : Advanced IBA Simulations held in Trieste, Feb. 2009,  a DataFurnace Practical Course was included.  You can download materials from this course here :-

1. The DataFurnace Practical materials,  including an overview and a detailed step-by-step approach to specific examples (RAR file, 34.9MB;  ZIP file, 35.7MB).  This zip includes a READ.ME file which explains the contents.
2. Lecture 4 on the 2007 IAEA-sponsored Intercomparison of IBA Software (3.13MB Powerpoint file)
3. Lecture 8 on overcoming the ambiguity of IBA particle spectra by using multiple techniques (4.77MB Powerpoint file)
4. Lecture 12 on overcoming the ambiguity of IBA spectra by using PIXE (29.2MB Powerpoint file)
5. Lecture 14 on Pitfalls in IBA (presentation based on the new MRS Handbook Pitfalls Chapter) (2.01MB Powerpoint file)
6. Lecture 16 on Techniques Complementary to IBA (12.2MB Powerpoint file),  with useful overview.  This "covers" XRF, XRD, TEM, MEIS & LEIS, AES & XPS.
   
7. There is also a Masterclass on the advanced IBA analysis of a meteoric impact-glass fragment which includes RBS, EBS and PIXE,  with the PIXE critical for determining the minor and trace elements,  and including information extracted on the nanostructuring of the sample using resonant EBS (15.2MB Powerpoint file)

Version History

WiNDF

WiNDF v7.0.26beta executable (13 August 2001, 1.33MB).  WiNDF v7.0.26 does support the changes in NDF v7.6f (see below).

WiNDF v7.0.46beta executable (23 October 2002, 1.55MB).  v46 implements the changes in NDF v7.7a (see below) and has fixed lots of bugs. However, it still does not yet implement all the features in NDF (but  WILL  run under NT). It also implements 2K "Ortec" data formats.

WiNDF v7.0.47beta executable   (19 February 2003, 1.55MB) has fixed some extra bugs (RoI of partial spectra of 28Si etc handled correctly)

WiNDF v7.0.48beta executable   (8 July 2003, 1.57MB) has at last sorted out the language problem:  Europeans use a decimal comma rather than a decimal point (known as the Windows Region number representation).  We are sorry this took so long (and it turned out to be easier than expected! - apologies).

WiNDF v7.0.60 beta executable (2 October 2003, 1.93MB) has a new graphical user interface to the simulator which is rather nice to use & various other things (pre-release version). A large number of bugs have been fixed.  This executable requires a new file isotopes.dat to be added to your C:/ndf directory

WiNDF v7.1.0 executable (30 October 2003, 1.94MB) includes batch file comments that help significantly to keep track of long analyses.  It also has quick access to Notepad

WiNDF v7.1.1 executable (22 December 2003, 2.06MB)  fixes a number of bugs and improves the implementation of NDF

WiNDF v7.1.2 executable (19 April 2004, 2.06MB) fixes a problem reading geometry files using 4He

WiNDF v7.1.3 executable (15 June 2004, 2.06MB) has more diagnostics for the "incorrect licence" error message

WiNDF v7.1.4 executable (10 June 2005, 2.06MB) now adds charges correctly in add spectra (filemaths child of plotspectra window)

NDF

NDF v7.6b executable (15 February 2001, 1.43MB).   Please note,  v7.6b  is a much bigger program than v7.1c at run time.  NDF v7.6b also implements linear depth scales properly (with user supplied densities).  The output data format is different from v7.1,  but both are compatible with WiNDF v7.0.30 and above. 

The newer NDF v7.6f executable (8 August 2001, 1.2MB) permits proper handling of charge on simultaneously collected spectra,  and has a revised Manual  This version of NDF needs WiNDF v7.0.26 and above. 

NDF v7.7a executable (26 February 2002,  1.5MB) is a Brand New NDF which implements the heavy ion stopping powers by Helmut Paul .  This version incorporates all the changes in the previous versions.  It also fixes an embarrassing error which we have discovered in the He - D scattering cross sections (many thanks to Joana Vaz Pinto!).  We apologise for any inconvenience caused.

NDF v7.8e executable (30 October 2003, 1.8MB) allows NDF to fit molecular composition (apologies to any users of this code:  it was buggy - fixed in 7.8g). (This is labelled 7.8f,  but the double scattering doesn't yet work properly)

NDF Manual

NDF Manual 7.8g (24 November 2003:  PDF, 617kB).  This Manual incorporates all previous NDF changes

NDF Manual 7.8g1 (17 December 2003:  PDF, 621kB).  This Manual differs from 78g in sections 12.2 and 12.3.  Please note the modification mentioned in the molecular stopping powers section.

NDF Manual 7.9a  (19 April 2004:  PDF, 617kB).  This Manual differs from 7.8g1 in section 12.1

NDF Manual 8.0a  (22 March 2005:  PDF, 617kB).  This Manual introduces sections 12.2, 12.5, 14.3

NDF Manual 8.0b  (30 May 2005:  PDF, 623kB).  This Manual differs from 8.0a in section 12.5

Examples

You can download the Examples that are referred to in the WiNDF Manual (see below):

Example1 (Au/Ni/SiO2/Si calibration sample, double detector, RBS):  zipped to 192 kbytes
Example2 (high dose implants, double detector, RBS, 12 samples):  zipped to 192 kbytes
Example3 (ion exchanged polymers, single detector, RBS, 20 samples): zipped to 540 kbytes
Example4 (TaNiC films on Si, H EBS/RBS, 16 samples): zipped to 367 kbytes
Example5 (CNx:H films, He RBS/ERD, 2 detectors, 21 samples): zipped to 1042 kbytes
Example6 (deuterated polymer films, 3He NRA/RBS & 4He RBS/ERD, 3 detectors, 11 samples): zipped to 766 kbytes
Example7 (optical multilayer - zirconia & silica on glass, He RBS, two incident angles, with error bars): zipped to 300 kbytes
Example8 (high accuracy RBS, IRMM Geel Sb certified sample, 2 detectors): zipped to 204 kbytes

(Back to DataFurnace Home Page)

Licensing Information

You can see the standard terms of the IBA DataFurnace Licence .  A specimen Licence Document can be downloaded (PDF, 53kB). Please notice that the licence is very non-restrictive:  it is the accelerator that is named and all your academics, students and collaborators are encouraged to use the code.  We expect you to make multiple copies.  We are keen to give licenced users every help,  including supporting your data formats.

Please contact Chris Jeynes if you wish to have a licence,  and also if you wish to have a trial of the code (see New Users).

(Back to DataFurnace Home Page)

New Users of Demonstration Code

(Back to DataFurnace Home Page)

Stopping Powers Used (and corrections for Si and SiO_²)

We distribute the public domain TRIM88 data with the code. Note that Ziegler's SRIM2001 has markedly different stopping powers from the TRIM88 values that we use,  and that ICRU is also developing new values.  DataFurnace supports other stopping powers including SRIM2000,  the heavy ion stopping powers by Helmut Paul (Paul & Schinner, NIMB 179, 2001, 299 & NIMB 195, 2002, 166), and the stopping power predictor code by Tessica Weijers et al (NIMB 215, 2004, 35). Please see the NDF Manual.

For much better (2% uncertainty) stopping powers of 4He in Si please put this file with  corrected 4He stopping powers for Si  into c:\ndf\  (or the equivalent on your machine),  overwriting the pscoef.dat file that is there already.  This data for He on Si  is by Barradas et al [Accurate determination of the stopping power of 4He in Si using Bayesian inference ,   NIMB 194 (2002) 15],  following Konac et al 1998 (the "KKKNS" stopping power measurements by Kalbitzer's group presented at the Lisbon IBA conference:  NIMB 136-138 (1998) 159-165) confirmed by Lennard et al  NIMB 152 (1999) 370-376 and Bianconi et al NIMB 161-163 (2000) 293-296  Please note:   these values for He in Si are 2% higher than the KKKNS values, as justified further in Boudreault et al [ Accurate RBS measurement of ion implant doses in silicon ,  Surf.Interface Anal. 33 (2002) 478].  To use the KKKNS values choose a multiplier of 0.98 for the Si stopping in the WiNDF geometry file editor.  

WARNING: NDF will change the a1-a8 coefficients of the ZBL90
parameterisation; hence, this will only work if you use those stopping
powers (e.g. MSTAR stopping values will not be affected). Also, this
means that the stopping of any ion in the given target element will be
affected; hence, you should only do this for single ion experiments.
Also, if you have more than one experiment at different geometries,
you should put all the stopping corrections in the same GEO file, and
only in that one.
Which means if you have the readstop sio2 sio2.1 line in one geo, that
is enough! And if you have measured the same sample with Li as well,
the stopping of Li in SiO2, which is calculated actually from the same
a1-a8 params (with something extra), will also change.

(Back to DataFurnace Home Page)

Frequently Asked Questions

My new installation doesn't run!

You get the "error with licence, refer to Surrey" message.  
Have you copied the licence file that Chris Jeynes sent you to c:\ndf (or wherever you put it),  remembering to change the name to "license.ndf" (you should overwrite the existing file).  Have you copied the up-to-date executables (windf.exe and ndf.exe) from this website overwtiting the existing files?
Have you rebooted after installation?  Did you run the Windows install program to completion (don't just copy files!) ? 

You get the "can't find licence" message.
The autoexec.bat file should include the lines (to complete the installation you have to have the right to modify the autoexec.bat file - you have to be Administrator on Windows NT systems for example):

SET PATH+%PATH%;c:\ndf  (or whatever your pathname is!)
SET PATHNDF=c:\ndf

Remember you have to reboot after changing autoexec.bat.  There may also be an entry in Registry.  This entry is not essential, but is an item in HKEY_CURRENT_USER\Software\WiNDF\App and should contain two entries:
filepath set to c:\ (where ever the last file used was...)
path   set to c:\ndf (where ever ndf actually is....)

My new installation doesn't run on an NT machine!!

NDF is very greedy for space,  and in NT you can manipulate the swap page size.  It may well be that the swap page size needs increasing.  Please see how to do this in Setting up NT for NDF .  This has been a problem for v7.6b and above.

I can run windf.exe and ndf.exe,  but ndf ends with an "exit code 0" message

Don't worry about this:  "exit 0" means that it has run correctly, probably (that is,  with zero errors!)   But do look at the NDF output because it tries to tell you if if has found errors,  and these cases will also give the "exit 0" message.  An "exit code 1" message means that NDF has crashed - there is an untrapped error.  We have tried to avoid these untrapped errors,  but there are still some left (sorry).

I get a "can't find windf.ini" error

Don't worry about this,  it is just the defaults for the displays etc which WiNDF tries to remember for you.  We do our best not to frustrate the users,  but don't always succeed ...

Can you support my data format?

Yes, we have a policy of supporting the data formats of our licenced users so that they are able to take maximum advantage of the high volume data handling capability of DataFurnace

DataFurnace seems very difficult to use!

Remember, you are used to a simulation program.  These are very simple creatures!  DataFurnace is a whole new way of thinking about the data.  The machine now does the simulations (very many of them, until the data is fitted) and your job is to set the parameters of the data space in which it looks for solutions.  First you have to get the geometry file right (beam energy, keV/ch, scattering angle etc).  Remember that you have to put in the collected charge too, and NDF is very fussy about this - it is much stupider than you are and will believe what you tell it, even if it is wrong ... NDF will generally give you a solution, and quite often completely stupid solutions can be good fits, so beware! Assuming you have all these basic things right you then have to fight the real ambiguity of the data.  We have provided a series of tools to help you persuade NDF not to consider various sorts of solutions, which objectively might be perfectly valid.

DataFurnace is a tool for real analysts to handle large numbers of spectra from real samples with speed, delicacy and accuracy.  We have provided very considerable flexibility because we needed it to treat the variety of samples presented to us.  If you want to do something and can't see how, please ask !  Probably we have done it already. If not we are eager to extend the code to handle more cases, so we may change the code for you!  

Other interesting links:

Multiple Scattering
There is an interesting new website on Multiple scattering which is well worth looking at, including further information on the paper: Small angle multiple scattering of fast ions, physics, stochastic theory and numerical calculations by Amsel, Battistig & L'Hoir (NIM B201,  2003, 325-388)

Non-Rutherford scattering (EBS)
For non-Rutherford elastic backscattering (EBS) the paper describing the SigmaCalc program (PDF document:  156kbytes) of Alexander Gurbich is invaluable.  The SigmaCalc website is now live,  please use it.

Evaluated cross-section measurements are extrapolated to different scattering angles and are described in his paper for the IAEA Report of the Advisory Group Meeting on Long Term Needs For Nuclear Data Development (Vienna, 28 November - 1 December 2000:  INDC(NDS)-423 (2001) IAEA, Vienna p. 51)  (PDF document:  119kbytes).  The 2 week IAEA Workshop on Nuclear Data for Materials Analysis (May 2003) is also  informative.  The Report of the  IAEA Technical Meeting on the Database of Evaluated Cross-sections for IBA Applications (October 2003) is also very informative.  

The IAEA Nuclear Data Service has recommended merging the SigmaBase and Gurbich's NRAbase,  resulting in the Ion Beam Analysis Nuclear Data Library (IBANDL),  prepared by Alexander Gurbich,  the new Liaison Officer between the IBA community and the IAEA.  These data are for elastic (non-Rutherford) cross-sections.  They are however in R33 format,  which at present is not accepted by NDF,  sorry.

IBA Services
For IBA services (including depth profiling and microbeam applications) please see the University of Surrey Ion Beam Centre site
For NRA with a D beam you can go to Matthew Healy at the  University of Cranfield (Royal Military College Shrivenham site)
For microbeam hardware:  Oxford Microbeams Ltd
For high resolution RBS and channelling:  MEIS at Daresbury

IBA web sites
Jim Ziegler's IBA website.  Latest documentation for ion stopping database. 
The Hotstop Project by Carlos Pascual-Izarra is a fascinating exercise which uses simulated annealing to extract stopping power curves (or, as he points out to be consistent with ICRU terminology, stopping force curves) from RBS data,  and then uses Bayesian inference to obtain confidence intervals to the curves obtained.

The Ion Beam Information System maintained by Edit Szilágyi and others in Budapest
SigmaBase at Idaho State University maintained by György Vízkelethy

The RBS tutorial by Charles Evans Associates is helpful
The Q-value calculator from the BUBL site is also useful.

The Surrey Ion Beam Centre is a part of the UniS Advanced Technology Institute and the UniS Materials Institute

(Back to DataFurnace Home Page)

Contacts

Dr. Nuno P. Barradas     nunoni@itn.pt                   +35 1 2199 46150
Dr. Chris Jeynes           c.jeynes@surrey.ac.uk     +44 (0)1483 689829
Prof. Roger P. Webb      r.webb@surrey.ac.uk      +44 (0)1483 689830

The Ion Beam Centre University of Surrey Guildford, England