Thanks Reinhard and all for quick and comprehensive answers. It appears that we do have microabsorption effect as Rietveld content deviates from expected in accordance with absorption coefficients and there is no simple fix for that but take proper precautions in sample preparation and select right reference material. Best regards,
Peter Zavalij BTW does anyone has electronic copy of Brindley paper? X-ray Crystallographic Center Department of Chemistry & Biochemistry 091 Chemistry Building University of Maryland College Park, MD 20742-4454 Phone: (301)405-1861 Fax: (301)314-9121 E-mail: [EMAIL PROTECTED] http://www.chem.umd.edu/facility/xray/ -----Original Message----- From: Reinhard Kleeberg [mailto:[EMAIL PROTECTED] Sent: Friday, November 16, 2007 3:14 AM To: rietveld_l@ill.fr Subject: Re: Amorphous content We have positive experience with commercial ZnO pigments or chemicals, heated to 700°C for about 2 hours in air for recrystallisation of the amorphous ZnO or the spurious Zn carbonates/hydrates to get nearly 100% crystalline ZnO. In 1:1 mixtures of the best NIST corundum, we found no significant deviation in Rietveld quantification, and so we use it routineously as internal standard. Such ZnO shows separate particles of about 0.5 µm diameter what can be assumed to persist a grinding and allow an appropriate low phase-specific correction for microabsorption. As Pamela stated, in case of the CeO2 one problem may be the absorption contrast between the materials. The problem can only minimised by (a) choosing an appropriate wavelength (as Pamela recommended Mo?), and/or (b) optimising/minimising the effect of the PARTICLE size of all constituents of the mixture. Thus, it is recommended to grind as fine as possible and then to look for the resulting particle size of the powder (e.g. by laser scattering). The particle size should at least fulfill the conditions of µ*D to be "fine" or "medium" given in Brindley, G.W. 1945. The effect of grain or particle size on X-ray reflections from mixed powders and alloys considered in relation to the quantitative determination of crystalline substances by X-ray methods, Phil. Mag., Ser. 7, 36: 347-369. In general, a correct mean particle size of the phases should used in the Brindley correction term. In our experience, the mixing of the standard ZnO and the sample powders can cause problems like forming aggregates of the both materials. For example, if larger aggregates of ZnO persist in the sample one can get "negative" amorphous content because of the underestimation of the ZnO by microabsorption. In contrast, coarse aggragates of sample phases cause underestimation of these crystalline phases and result in "pseudo-amorphous contents". Therefore we prefer admixing of the standard before grinding the sample to reach a really homogeneous mixture, assume the 0.5 µm particle size for Brindley correction of the scale factor of ZnO, and set an estimated particle size (mostly 1-4 µm) globally for the phases of the powdered sample, according to some experience regarding hardness and behaviour of the material in our mill. However, the uncertainty of this estimated values can still significantly bias the result of quantification of the amorphous content, especially for heavily absorbing materials. This can simply checked by calculating the Brindley correction factors for linear absorption coefficients of the actual sample material and varying the particle size in a realistic interval. Reinhard Kleeberg Whitfield, Pamela schrieb: > At first glance it looks like a classic microabsorption problem, but I > don’t have the linear absorption coefficients to hand. Using an > internal standard with a too small absorption will tend to > over-estimate the amorphous content. Ce versus Zn is a pretty big > contrast for CuKa, even if the particle sizes are small enough. The > whole point of that NIST series (674 I think) is that they were to be > used as appropriate contrast matching standards for quant analysis, > and were supposed to be quite different from each other. > > Changing the wavelength to reduce the contrast (e.g. Mo) may help for > that particular mix but probably won’t completely solve it. > > Pam > > *From:* Peter Y. Zavalij [mailto:[EMAIL PROTECTED] > *Sent:* November 15, 2007 9:07 PM > *To:* rietveld_l@ill.fr > *Subject:* Amorphous content > > Hi, > > I am trying to determine amorphous content using Rietveld refinement > and internal standard. However resulting content of amorphous phase is > really unrealistic. > > Moreover testing the method using standards with known amorphous > content does not clarify the situation. For example ZnO (NIST, 95% > crystallinity) used as standard to determine amorphous content in CeO2 > (also NIST standard with 91% crystallinity) yield 25% of amorphous > phase which is a little bit too much comparing with expected 9%. > > We tested several different standards, mixtures and preparations, > different diffractometers and software without much luck... Seems like > something simple is missing... > > Any clues? > > Many thanks, > > Peter Zavalij > > X-ray Crystallographic Laboratory > Department of Chemistry & Biochemistry > 091 Chemistry Building > University of Maryland > College Park, MD 20742-4454 > > Phone: (301)405-1861 > Fax: (301)314-9121 > E-mail: [EMAIL PROTECTED] > http://www.chem.umd.edu/facility/xray/ >
