MossWinn

 

 

Aim and scope

  
 

The MossWinn program aims to become the standard reference PC software for Mossbauer spectrum analysis. MossWinn combines complex physical theories and powerful mathematical algorithms with a unique, user-friendly user interface that provides access to the underlying complexity and power with unrivalled ease. Being powerful and simple at the same time, MossWinn is as well suited for experts of Mossbauer spectroscopy, as for students getting acquainted with the method. MossWinn provides experts with the tools necessary to gain deep insight into their measured data, and promotes high productivity by taking care of the technical part of the spectrum analysis while allowing the user to concentrate on the essentials of the scientific problem. In MossWinn emphasis is also put on precision, versatility, customizability and robustness of operation. Drawbacks and considerable restrictions, often associated with less sophisticated programs, are notably absent in MossWinn. It is easy to install, easy to work with, and offers a uniquely wide range of options.

  

Technical details

 
 

  • MossWinn 4.0 is compiled with Delphi 2007 to 32 bit Windows executable suitable for PC systems with MS-Windows XP, Vista, 7, 8 and 10 (32 bit as well as 64 bit versions).

  • Example source code disclosed for DLL libraries that allow the user to extend MossWinn's capabilities with arbitrary user-programmed fitting- and cross-reference functions. The source code of the main executable (more than 250000 program lines) is not disclosed.

      

    Highlights

     
     

    (1) World's first Mossbauer program utilizing Evolution Algorithm in the fitting of Mossbauer spectra.

    One of the unique and sound features of MossWinn is the usage of Evolution Algorithm for the global fitting of Mossbauer spectra. The Evolution Algorithm imitates natural biological evolution processes to find the optimum set of Mossbauer parameters of any model selected by the user. The method is not sensitive to the goodness of the initial set of parameters, therefore the manual setting of these initial parameters can be omitted.

    (2) World's first Mossbauer program seamlessly integrating the fitting of crystalline and amorphous / hyperfine field distribution subspectrum models in a single fit menu system.

    In the first 40 years of Mossbauer spectroscopy the fitting of the Mossbauer spectra of crystalline and amourphous materials was handled by separate mathematical routines. While most programs could only handle spectra of crystalline materials, for the amourphous materials separate programs were developed that could handle distribution fitting, but were less suitable to fit spectra of crystalline materials. MossWinn is the first and is — according to the knowledge of the author — still the only program that seamlessly integrates the fitting of crystalline and distribution subspectra in a single menu system.

    (3) The selection of fit models is fully mouse oriented, and the models can be changed on the fly.

    (4) World's most accomplished program for simultaneous fitting of Mossbauer spectra.

    Although an outstandingly powerful technique, simultaneous fitting of Mossbauer spectra is usually not available as an option in Mossbauer programs. In contrast, MossWinn has been developed for simultaneous fitting by default, and the fitting of a single spectrum can be considered solely as a special case. You will find simultaneous fitting most useful especially in cases when either the physical model, or the precise value of the Mossbauer parameters can not be derived unambigously on the basis of a single spectrum alone.

    (5) Mossbauer parameters can be constrained to values taken from multiple non-overlapping intervals.

    For example, in fit models one may want to constrain the 151Eu isomer shift to lie either in the range of –15 … –10 mm/s (Eu2+) or in the range of 0 … +1.5 mm/s (Eu3+) relative to EuF3. Namely, the 151Eu isomer shift of most Eu compounds lie in one of these ranges, but not in between.

    (6) Mossbauer parameters can be forced to obey constraints and inequalities like

    • Line Width (2) should be equal to Line Width (1)
    • Q. Splitting (2) should be twice of Q. Splitting (1)
    • Isomer Shift (2) should be equal to Isomer Shift (1) plus a constant
    • Isomer Shift (2) should be higher than Isomer Shift (1)
    • Isomer Shift (2) should be close to Isomer Shift (1)
    • Q. Splitting (2) should be far from Q. Splitting (1)
      etc.

    (7) Fit numerous individual Mossbauer spectra sequentially according to a previously selected fit model automatically, without human intervention.

    This feature is especially useful when one needs to fit several similar spectra according to the same fit model.

    (8) Customizable with user-programmed fit model functions: any theoretical model can be added to MossWinn.

    According to the detailed guidelines of the manual, one can easily complement MossWinn with user-programmed fit model functions. For the compilation one can use, e.g., the programming environment Delphi 2.0 or later (in case of the MossWinn 4.0 series). For each source nuclide 30 independent fitting functions can be added to those built into MossWinn. In this way one can also use MossWinn as a general-purpose scientific data fitting program: namely, the added functions do not need to be Mossbauer fit models. For example, one could make MossWinn to fit the Curie-Weiss law to magnetic susceptibility data, or to determine lattice parameters and the size of crystallites by fitting X-ray diffractograms.

    (9) Mossbauer spectra can be organized into separate projects and project groups. One can export individual projects or a whole project group into a standalone file from which they can be reloaded any time later on.

    Exchanging data with a notebook PC, sending data via E-mail to a cooperating scientist, or archiving your work on compact disk; all these procedures can benefit from the intelligent project management system of MossWinn. MossWinn will save projects or whole project groups as compressed standalone files that can be reloaded on any other PC running MossWinn. During the process MossWinn will take care correctly of the changed path names of the Mossbauer data files, so that on the target PC you can work with your data exactly as you did on the original PC the project was imported from.

    (10) Numerous clipboard and printing functions promote high efficiency of spectral analysis.

    Visualizing the results of spectral analysis in the form of textual fit reports, tabulated parameters and graphical images is an important but often time consuming part of the scientific work. In order to promote high efficiency in this regard, MossWinn offers numerous ways to export fit results: they can be saved, printed or copied to clipboard in various forms. Ordered tables of fit parameters, fitted spectra and fit reports can be saved, printed or put into the clipboard of Windows with only a few mouse clicks. A HTML based fit report format (example) allows logging multiple fit results (including also the complete fit model applied and the image of the fitted spectrum) in standalone HTML FitLog files. The graphical images of spectra exported by MossWinn are of publication quality.

    (11) Able to fit relative subspectrum areas beside absolute ones.

    The fitting of the relative proportion of subspectra is available in MossWinn for crystalline subspectra as well as for hyperfine field distributions. For example, by fitting the relative area instead of the absolute one, one can require the program to find a solution in which the relative proportion (in %) of one or more subspectra is the one set by the user.

    (12) Able to fit the smoothing factor of strictly positive distributions.

    The original method of J. Hesse and A. Rubartsch for deriving distributions did not allow for the fitting of the smoothing factor, because the ideal value of the latter parameter was zero in all cases, for which value the resulted distribution was too noisy, not smoothed at all. With a slight modification of the original method, by deriving strictly positive valued distributions, the optimal value of the smoothing factor will be higher than zero, and will depend on the statistical noise of the fitted spectrum. In this way the smoothing factor can be fitted, and distributions—derived from spectra with different degree of signal/noise ratio—can be compared.

    (13) Able to fit up to 5 different distributions together with subspectra of crystalline sites, for each of the simultaneously fitted spectra.

    (14) Able to fit distribution and transmission integral for built in as well as for user programmed fit models.

    (15) Able to fit hyperfine field distributions by using the solution of the Hamiltonian of mixed magnetic and electric hyperfine interactions as the elementary pattern.

    In most programs the fitting of distributions is restricted to the use of multiplets (doublets, sextets) as elementary patterns. In MossWinn the method of J. Hesse and A. Rubartsch is implemented in a generalized way that allows the fitting of distributions by using the solution of the Hamiltonian of combined magnetic and electric hyperfine interactions as elementary pattern. Having this option available in MossWinn, one can derive relevant distributions even if the elementary pattern cannot be approximated sufficiently well with a sextet or a doublet.

    (16) Able to combine transmission integral fitting with the calculation of arbitrary-profile hyperfine parameter distributions performed via the Hesse & Rubartsch method, and thereby allows valid fitting of distributions to spectra of thick absorbers.

    (17) Able to calibrate and fit unfolded spectra as well as folded ones.

    By fitting unfolded spectra one can derive precise results even in cases when the folding point (measured in channel number) is neither integer nor half-integer.

    (18) The SCANFIT.EXE program of the MossWinn package enables the creation of ASCII Mossbauer data from bitmap images of Mossbauer spectra.

    (19) Support for database access: the MossWinn Internet Database (MIDB) is a novel Mossbauer spectroscopy database service that can be subscribed to by licensees of the MossWinn program.

    (20) Support for database assisted spectrum analysis.

    Fit models for pristine spectra can be selected from a list of models compiled on the basis of records included in the MossWinn Internet Database (MIDB) either by taking into account the compound stoichiometry associated with the measurement or by evaluating the similarity between the measured spectrum data and the spectrum data included in the database records (the latter option requires subscription to the MIDB service).

    (21) Support for parallel computing on multicore processor systems: unleash the power in your multicore processor by engaging the parallel computing algorithms of MossWinn that can dramatically speed up time-consuming processes.

    Mossbauer options

     
     
  • Triangle & sine velocity waveform calibration modes
  • Transmission & Reflection (e.g. CEMS) modes
  • Thin absorber approximation & transmission integral calculation modes
  • Available line shapes:
    • Lorentzian
    • Pseudo-Voigt
    • Lorentzian with cosine smearing
    • Lorentzian with dispersion
  • Built in linear models available for the nuclides: 57Fe, 119Sn, 125Te, 151Eu, 121Sb, 129I, 141Pr, 237Np, 197Au, 161Dy (25.655 keV)
  • Built in parameterized-profile distribution models for 57Fe, 119Sn and 125Te :
    • VBF (Voigt-based fitting) doublets and sextets
    • xVBF (extended Voigt-based fitting) doublets and sextets
    • Binomial atomic distribution for a single shell of atomic neighbors
    • Binomial atomic distribution for two different shells of atomic neighbors


    MossWinn is able to fit Mossbauer spectra by using the solution of the static Hamiltonian of mixed magnetic and electric hyperfine interactions. The highly capable fit menu system of MossWinn will faithfully reflect the dependence of the Mossbauer spectrum on the relevant physical parameters. For M1, E1, E2 and M1+E2 nuclear transitions even the nuclear parameters (e.g. g-factor, quadrupole moment) can be fitted.

  • Built in Hamiltonian models available for 57Fe, 119Sn, 125Te, 151Eu, 121Sb, 129I, 141Pr, 237Np, 197Au, 161Dy (25.655 keV) and all M1, E1, E2, M1+E2 transitions for which Iground and Iexcited are less or equal 9/2 :
  • Static Hamiltonian solved for mixed interaction for powder sample (Vzz or H based system)
  • Static Hamiltonian solved for mixed interaction for powder sample with Goldanskii-Karyagin Effect (Vzz or H based system)
  • Static Hamiltonian solved for mixed interaction for mosaic sample (Vzz or H based system)
  • Static Hamiltonian solved for mixed interaction for single crystal sample (Vzz or H based system)
  • Slow and fast relaxation limit 57Fe paramagnetic hyperfine structure models for powdered samples with S = 1/2...5/2 (examples).
  • Built in relaxation models for 57Fe:
  • Blume-Tjon two state magnetic relaxation model for zero as well as for non-zero asymmetry parameter as described in PR 165 (1968) 446.
  • Tjon-Blume Jahn-Teller quadrupole relaxation model for zero asymmetry parameter, as described in PR 165 (1968) 456.
  • Electron Hopping Relaxation model implemented for iron in paramagnetic powder where electron hopping occurs either between iron ions of different oxidation states (e.g. Fe2+ ↔ Fe3+) or between iron ions and the rest of the system. It can model situations encountered, e.g., in Phys. Chem. Minerals 10 (1984) 250. and PRB 31 (1985) 34.
  • Electron Hopping Relaxation model implemented for iron in magnetic powder where electron hopping occurs either between iron ions of different oxidation states (e.g. Fe2+ ↔ Fe3+) or between iron ions and the rest of the system. It can model situations encountered, e.g., in J. Phys. Chem. Solids 41 (1980) 1273.
  • Analytical expression for the line shape of the 57Fe Mössbauer spectrum of a powdered sample, with randomly oriented EFG and ETA = 0, subjected to a uniaxial (external) magnetic field, as given in NIM B 9 (1985) 201
  • Various background options:
    • Constant
    • Slope
    • Curvature
    • Third Order
    • Unfolded
    • Unfolded with slope

    Limits

     
     
  • Maximum number of absorption lines in a subspectrum of a linear model: 48
  • Maximum number of Amplitude/Position/Width/Extra type parameters in a subspectrum of a linear model: 18/18/18/18
  • Maximum number of model parameters in a subspectrum: 48

  • Maximum number of subspectra in one spectrum: 30

  • Maximum number of distribution subspectra in one spectrum: 5

  • Maximum number of distribution data points in a single distribution: 180

  • Maximum number of distribution data points in one spectrum: 360

  • Maximum number of simultaneously fitted spectra: 32

  • Maximum number of fitted parameters: 2048

  • Maximum number of projects: ~ 800

  • Maximum number of windows loaded at once: 250

  • Maximum number of spectra and subspectra loaded at once: 1000

  • For further details

     
     
  • Visit the help page of MossWinn 4.0 (click here).
  • Download the manual of MossWinn 4.0 (click here).
  • Download the demo version of MossWinn 4.0 (click here).
  • Check out the MossWinn Video Tutorials.
  • Visit the page of the MossWinn Internet Database.
  • Go through the getting started tutorial of MossWinn 3.0i xp (click here, and click on the LOAD box).
  • To purchase

     
     

    Contact the author, or one of the distributors via E-mail.