Since this is my first post I will start with a brief introduction. My name is Patrick Price and I am in my second year of graduate school. My thesis work involves the investigation of phase equilibria in perovskites.
I am using a Bruker D8 Discover diffractometer with parallel beam geometry. The diffractometer has a Cu K-alpha X-Ray source with a Si Gobel mirror and a Ge monochromator giving a parallel beam monochromatic x-ray source. The receiving side has Soller slits and a Tl-doped NaI point detector. I am trying to teach myself how to use TOPAS to PROPERLY analyze my data using Rietveld refinement techniques. I have recently taken a scan of the NIST line profile 660 LaB6 standard followed by scans of my perovskite powders using a step size of 0.02 degrees and scan time of 4 seconds. Most of the articles I have read are specific to convergent/divergent beam geometries and I do not know how much of that information transfers to parallel beam geometries. If anyone could help me answer the following questions I would greatly appreciate it. These questions mainly address which parameters should be refined with the LaB6 standard when using parallel beam geometry. 1. I need to use the scan of the LaB6 powders to characterize the contributions of the instrument to the diffraction profile. Starting with the emission profile, TOPAS asks for the wavelength, the Area, and the Lorentz Half Width. First, I assume the wavelength I should be the more recent Cu Ka wavelength of 0.154059 nm instead of 0.154056 nm. Second, does Cu Ka have a definite Lorentz HW and “Area” or should these parameters be refined with the LaB6 diffraction pattern? 2. Since I have a Ge monochromater I assume the Lorentz polarization factor should be fixed at 27.3 (Is this correct?). Obviously the lattice parameters and atomic positions would be fixed. 3. I read that you should NOT refine both the zero shift error and sample displacement, and since it is parallel beam I only refine the zero shift error. Should I refine surface roughness, absorption, or sample tilt with the LaB6? (Currently I do not refine these) 4. Am I correct in assuming that I do not have any EQUITORIAL convolutions (e.g. from slits, FDS, beam spill, VDS) since it is parallel beam geometry? What about TUBE TAILS? 5. I am using the Finger_et_al method to refine the AXIAL convolutions, however I often get a large error associated with the S value (sample length), even when my GOF is decent (<1.45). Do any of you know why this would happen? 6. Should I refine the “Scale” or scale factor. (Currently I do) 7. IMPORTANT: Originally I was refining the crystallite size but it always refined to a very small value (~300nm), where as NIST claims 660 LAB6 should have a mean grain size of a few microns or more. I assume this happens because the TOPAS is accounting for instrument caused peak broadening by making the crystallite size smaller than it actually is in the software. However, when I do refine the grain size I do get a better fit. Should I leave this unchecked, refine it, or fix it at a reasonable value of ~2500 nm. In summary, currently I am only refining the Lorentz HW and “Area” in the emission profile, zero shift error, the Finger parameters (S & H), the scale factor, and nothing else. I am unsure if I should be refining anything else such as the crystallite size, tube tails and other forms of equatorial convergence, or if there is something else that is important which I am disregarding completely. I am also unsure if I am correct in refining Lorentz HW and area in the emission profile. Sorry if I got a little long winded; I just wanted to give enough detail so people could answer. Thank you in advance for your help. Patrick