Investigations using diffraction
Rietveld refinement (what’s hiding in your data)
Data collected on a laboratory XRD is generally used for phase identification, But unfortunately the fun often stops there. The same data can tell you so much more about your sample using a process known as Rietveld refinement. Phase ID involves comparing the collected data to previously determined patterns, and while there are thousands of patterns available, the process is still limited, and lots of information about the sample goes undiscovered. Rietveld refinement is more complicated and is generally learnt through a dedicated course (e.g. ANSTO course) or through online tutorials (supplied with free refinement software by GSASII website). It also required an in-depth knowledge of the crystallography of the system as you are required to build up a model of a crystal structure, calculate its diffraction pattern and compare it to your data. Because every conceivable detail of the model can be changed, modifying the calculated patter, it allows for accurate measurement of many parameters that cannot be easily accessed via phase ID alone. The most common are hidden details are:
- Relative phase quantities (% of different components)
- Extremely accurately measuring lattice parameters and atomic positions
- Substitution of atoms in the unit cell
- Orientation of the crystals
- Size and shape of the crystals
This study looked at the products of a series of reactions involving the synthesis of the MAX phase Cr2GaC via a carbothermal reduction reaction. Varying the amount of C available changed the products from the carbon deficient Cr3Ga to the C rich Cr7C3. The quantitative plot easily allowed optimisation of the system.
Figure 1 Refined XRD data.
Black + are collected data, red line is calculated, green line is the difference plot and the markers at the bottom indicate peak positions for the different phases in the sample. This data was used to determine the phase composition of the three phases in the sample.
Figure 2 Plot of relative phase compositions of different samples created by modifying starting reactants.
A refinement (shown above) was performed on each sample and relative phase quantities were determined. This allowed the optimum starting ratios (107% C) to be easily determined.
Dr Dylan Cuskelly
School of Engineering
Faculty of Engineering and Built Environment, SE Mech Eng
The University of Newcastle