GSAS Tutorial - Tasks

Leslie Struble (University of Illinois) and Paul Stutzman (National Institute of Standards and Technology)

Version 1 (March 2015)

Tasks

A number of individual tasks comprise these instructions. Single phase tasks were designed to provide familiarity with the technique and users are encouraged to perform at least one of these single phase tasks.

More complicated tasks involving clinker and cement utilize the KOH-sugar (KOH-S) extraction residue and the salicylic acid-methanol (SAM) extraction residues to refine parameters prior to refining the bulk material. These selective chemical extractions concentrate specific phase groups, simplifying the diffraction pattern and concentrating the residue phases, thereby enhancing their detection limits. Such chemical extractions are not required in analysis of cementitious mixtures but are generally and highly recommended. It is possible to do refinements on a single pattern of clinker or cement, but the results are not likely to be satisfactory. Selective extraction procedures are detailed in [6]. If you are planning to use the method routinely on samples of a material such as portland cement, which contains a complex set of phases with the same crystal structure parameters, it is a good strategy to use patterns for KOH-S, SAM, and perhaps nitric acid extraction residues to refine the crystal structure parameters more accurately for phases in each residue and then to use the resulting parameters in your subsequent analyses.

In a subsequent section on steps in the refinement, more detailed instructions are given for these tasks. And in a subsequent section on assessment expected results are described.

Peak profile parameters are required for fitting. These parameters describe the instrument contribution to the diffraction peak shape. GSAS includes a file inst_xry.prm that can be used as an example in establishing a file specific to your instrument.

For fitting an XRD pattern, the peak profile parameters should be established using data collected using the same instrument that was used to collect the pattern. Such data should be collected on a sample with negligible strain and relatively large crystallite size, such as National Institute of Standards and Technology (NIST) Standard Reference Material (SRM) 660, LaB₆. However, the peaks from this sample are much sharper than seen with portland cement phases, so more steps are required to refine the peak shape parameters. An alternative is to use a single phase material with a peak shape more like that of the cement phases. This could be one of the NIST SRM powder diffraction intensity standards or a fine-grained quartz powder.

The sequence of operations in a Rietveld refinement is, to some extent, discretionary and depends on the objective of the refinement. However, results may be affected by the refinement sequence and a poor sequence may cause the program to crash. The sequence given here is based in part on Table 1.5 of Young^[1], in part on Cranswick and Swainson^[7], and in part on the personal experience of the authors.

Perform the following tasks. More detailed instructions are provided in the section titled Steps in Refinement. If you have any trouble, you may wish to check the section titled Helpful Hints.

1. Download the Rietveld software GSAS and its graphical user interface EXGUI plus related files from https://subversion.xray.aps.anl.gov/trac/EXPGUI/. Versions are available for various types of personal computers, and you should use the most recent version for your specific computer. Note: the installation program does not set up a link to GSAS in the Windows startup menu. If you desire such a link, you must set it up yourself.
2. Download and save the Cements_Data.zip file that contains the following five data sets:

• Cement crystal structures in multiple data formats: Individual GSAS experiment files (.EXP), a standardized crystallographic information file (.CIF) that is readable by most Rietveld codes, and a table (.TBL) format, which is the simplest to read for manual data input. In addition, collections of individual phases by group (alite, belite, aluminate ...) are provided in GSAS.EXP form. These data should be considered read-only to retain the original values. A key is provided within the data folder to explain the data files.
  
• An instrument parameter file (.PRM) and one or more of the single phase patterns provided (corundum, periclase, lime, or quartz). Using these files and the crystal structure parameters from the structure archive (TBL folder) for the phase you selected, refine the single phase pattern to generate a new peak profile parameters file. Save this file to use in subsequent refinements. The single phase patterns and the subsequent XRD patterns were collected using the same instrument (at NIST) and so are suitable for these tasks, though not for use in refinement of patterns collected using your own instrument.
• Next, work with the XRD patterns for one of the complex mixtures (SRM 2686 clinker, proficiency cement A, or proficiency cement B). Perform a qualitative analysis of the pattern using CMPR or some other software. Use the crystal structure parameters for all the identified phases from the structure archive and your instrument parameters file to refine the mixture to determine the amount of each phase. Compare results to published values (see section titled Assessment).
• Finally, work with the XRD patterns for the extraction residues of the same complex mixtures, both SAM (SRM 2686 clinker, proficiency cement A, or proficiency cement B) and KOH-S (SRM 2686 clinker, proficiency cement A, or proficiency cement B).
  
• Perform a qualitative analysis of each pattern. Using the crystal structure parameters for all the identified phases from the structure archive and your instrument parameters file, refine the extraction residues of the same complex mixture to determine refinements for the crystal structure parameters. Then use these new parameters to refine the mixture to determine the amount of each phase. The XRD file header for the SAM extraction residue contains the insoluble residue mass percent value, allowing for a recalculation of the residue phases on a whole-clinker basis.
  
• Compare results to published values (see section titled Assessment). Compare results of the complex mixtures and selective extraction residues (the computed mass fraction of each phase and the figure of merit for the clinker/cement without using extraction residues and with using extraction residues).