Talk:Powder diffraction/Archive 1

Peak Broadening
The Scherer method is one of the more primitive ways of analysing the broadening of X-ray peaks, but can be useful in some regards.

Peaks are generally regarded to be broadened by three different effects; instrumental, size and strain. Although, for the most part size and strain are said to be separate entities there is a certain amount of overlap. For example, dislocations would be included in both as they are part of grain boundaries and cause lattice strain.

To get rid of the instrumental effect two methods are utilised either Stokes Fourier deconvolution method or the Reitveld method. The next step is to separate the size effect from the strain effect.

The size and strain profiles can both be described by Voight functions and the size part is independent of the angle or plane. This, and finding the integral breadth of the peaks, forms the basis of the double Voight method. Alternatively, one may undertake a Fourier analysis and use the Warren-Averbach formula to separate the two.

Whilst the peak shift is often thought to be mainly due to long-range internal stresses, the lattice strain can be found with the sin squared phi method, the broadening of the peaks is in normally attributed to dislocations due to its strain field that decreases as 1/r, less than other defects. However, other factors should not be ignored, as stacking faults can in some materials cause more broadening and recent evidence suggests (Levine et al 06) that strain varies considerably between dislocations cells.

IUCR Monographs, Defect and microstructure analysis by diffraction, Chapter 7,

LYLE E. LEVINE1* et al ; X-ray microbeam measurements of individual dislocation cell elastic strains in deformed single-crystal copper; nature materials; VOL 5; 619; AUGUST 2006

R. Kužel;Dislocation line broadening; Z. Kristallogr. Suppl. 23 (2006) 75-80 75

--[00:15, 16 November 2006 (UTC)Hadfield Simm

Untitled
A group of scientists affiliated with the International Centre for Diffraction Data would like to propose changes to this page. We are interested in adding an Introduction and adding information that focuses on multi-component analysis and practical applications as this seems to be missing from the page. We wanted to approach the group here before making any edits. Please share feedback/comments on this proposal. Thank you!

"Neutrons are only available at a small number of nuclear reactors in the world." I'm sorry, but when did neutrons stop being at the heart of nearly all atoms and become some rare thing that only exists in reactors? Need someone that understands the subject to clarify if this is speaking about high energy neutrons as radiation or what.


 * Neutrons are indeed in every atom (well, barring Hydrogen), but you need free neutrons not at the core of an atom for diffraction, with a wavelength close to 1Å so that it can be diffracted by the crystal lattice. Only suitable for this are research reactors and spallation sources (and only a few of thoses are actually used for materials reserch). I have added links to reflect that. VincentFavreNicolin 22:17, 8 January 2006 (UTC)

Powder x-ray diffraction does not use neutrons, but rather electrons, generally powered by a potential of 30,000V
 * Well, actually it uses x-rays, though they're produced using electrons.AlmostReadytoFly 20:30, 30 August 2006 (UTC)

Some confusion above. X-ray diffraction can be done in a laboratory with the K-alpha-1 characteristic peak, the consequence of electrons being rapidly decelerated. Alternativley, one can use accelerated electrons which due to wave-particle duality exhibit a wavelength of X-rays. Whether one uses X-rays or neutrons depends on what you want to measure.


 * If it helps, a classic book comparing X-ray, neutron and electron diffraction is John Cowley's "Diffraction Physics" (1975, North-Holland, Amsterdam). X-rays for diffraction do typically come from electron-excited anode-target (characteristic) or synchrotron (Brehmstrallung) sources, neutrons can I think be thermal (perhaps they scatter mostly from nuclei) but may still require a reactor, and electrons are typically between 100-300 keV in transmission electron microscopes.  That gives them wavelengths in the picometer range, and hence much smaller Bragg angles.  Thermochap (talk) 15:56, 24 February 2008 (UTC)

Rewrite 2007-02
In case it is not clear, I am approaching this from a materials science / physics point of view, and any balance is welcome. Particularly, if someone wants to offer a better explanation under the crystallinity subsection, and how to deconvolute the amorphous background from the inelastic scattering background which exists even for perfect crystals.

I reorganized a bit and pared several full sentences which appeared redundant - please edit or leave a note here if anything significant has been omitted from this version.

Moved momentum transfer out of the intro for clarity - Bragg scattering is an elastic process, inelastic x-ray scattering is completely different. Change in incident wave vector = reciprocal lattice vector is covered under both reciprocal lattice and Bragg scattering. I suspect that most people who are unfamiliar with the topic would expect a change in the magnitude of the momentum, not just direction.

There was an html list immediately following the Uses section; it did not correspond to the succeeding sections, and Wikipedia generates a table of contents automagically, so I removed it. Some of these items should be given their own subsections, though, so I have reproduced it here:

 Crystal Structure Determination  Precise Lattice Parameter Measurements  Identification of Unknown Specimen  Quantitative Analysis of Powder Mixtures  Determination of Crystal Size and Lattice Strain  Phase Diagram Determination  Detection of Long-Range Ordering  Evaluation of Textures in polycrystalline solids  Eldereft 19:44, 28 February 2007 (UTC)

.Moved momentum transfer out of the intro for clarity.... Hmm. Are you sure it is not you who was expecting that a change in a vector had to correspond to a change in its magnitude? I.e. aren't you the one confusing vectors with scalars? q is a momentum vector, what is wrong in telling people that? And no if a vector changes it does not have to change in magnitude. Momentum transfer does not imply inelasticity.

I have removed the erroneous heading 'Crystal Structure Determination' above what was clearly a discription of what powder diffraction is really used for: identifying and characterizing phases and no that is by no means the same thing. Yes you can also solve an unknown structure from scratch but that is not done routinely. Phase recognition, cell parameter determination is. I have therefore created a separate heading for crystal structure detn. to refer to the Rietveld subsection. I have bracketed out the topic about rolling and the phase transitions of Sn. It may well belong in the story but not at that point Jcwf 00:34, 17 September 2007 (UTC) Jcwf 00:34, 17 September 2007 (UTC)


 * Um, no; good job with the civility. Inelastic scattering gives me noise and no change in momentum gives me no data - both of these are to be avoided. On the other hand, it would be nice if at least the intro to this article is accessible to your average decently bright high school student. As it was given in the intro, there was insufficient context to have a good chance of meeting this goal. It was treated correctly in the article.
 * But yeah, I agree that the whole tin section can probably be tossed, I am not sure why I tried to rewrite that instead. - Eldereft ~(s)talk~ 07:57, 27 February 2008 (UTC)

Area Diffraction Machine software
Hi. I added a link at the bottom of this page to software that I wrote to analyze powder diffraction data. I hope that nobody has a problem with this - Joshua Lande —Preceding unsigned comment added by 206.192.69.193 (talk) 01:49, 27 February 2008 (UTC)
 * Fit2D with a more modern GUI? With perhaps some better peak finding/matching functionality? ...ok. It is self-promotion but not directly commercial; someone else can toss it, but at least for the nonce it seems not so bad. - Eldereft ~(s)talk~ 07:57, 27 February 2008 (UTC)

Introduction
The introduction is very short. I was going to tag it as needing expansion (a gentle nudge to elderleft), but I couldn't figure out what the appropriate template is. ChildofMidnight (talk) 19:11, 12 March 2009 (UTC)

Invention
I'm told (by people at the Paul Scherrer Institute) that the powder diffraction method is also called the Debye-Scherrer method, after its inventors. Perhaps this merits mention?AlmostReadytoFly 20:23, 30 August 2006 (UTC)

Powder diffraction is not Debye-Scherrer diffraction. Debye-Scherrer is One way of geometry with which you can measure a diffraction pattern, but not the only one. There is also the Guinier way, Seemann-Bohlin, Bragg-Brentano if I don't mess up here, and many more, depending on the machine you use and how you adjust it (Guinier=about 20.000€, Bragg-Brentano=500.000€ and more). Diffraction was invented by Max v. Laue as far as I know. 78.43.7.44 (talk) 16:58, 14 April 2011 (UTC)

Vector?
The equation after "This leads to the definition of the scattering vector as:". The right hand side seems to me to be a scalar. What is this meant to mean?Billlion (talk) 09:56, 14 November 2011 (UTC)