作为浸淫粉末精修多年的老虫，尽管并不经常用精修，但从最初学习Fullprof开始，后来舍弃Fullprof学习GSAS，至GSAS II至今，其实仍有很多基本概念不清楚，日常也只仅限于需要的时候拿来用用。粉末精修不仅仅要求对软件掌握，还要对精修的原理弄明白。其实最重要的是精修之前衍射数据的采集，因为精修只不过是尽量弥补粉饰实验的不足而已。从Rietveld博士建立此方法已有大量的教科书，在此不再啰嗦。下面只是就最近和Robert B Von Dreele交流的一些知识点分享给大家，供GSAS II学习者参考。
背景：因为当时精修某几相的粉末数据时发现一点小问题，怎么也找不到解决办法，便写信给Dr. Von Dreele，并顺嘴问了他如下所示的问题。GSAS II中用来描述择优取向的函数在Phases菜单，Data目录下，如下图所示。
精修择优取向时，如果不选择精修此项，也可以精修Phases菜单下的Texture目录下的Harmonic order，如果分别用这两个目录下的和谐函数，采用的和谐项数相同，譬如10，那么得到的精修结果是一样的。但这两项不能同时精修，因为对于实验室常用的Bragg-Brentano几何衍射仪，应该用的是Data目录下用来描述择优取向的和谐函数，而非专门用来进行织构精修的那个选项。如果用织构精选选项，其中涉及到衍射仪和样品的欧拉角度，这是我困惑的地方，也是交流的主要目的。因为我发给他的是我用云母的精修数据为例来咨询的，所以他也是以云母为例来解释的。我把我的疑问和Dr. Von Dreele的回信贴出来，供大家参考。
问：Dear Dr Von Dreele,I was going to give up asking for the questions related to the texture because I did know little about it and was afraid of disturbing you too much. However, now I am more confused by your explanation and eager to make it clear. I did not set the sample orientation angles under "phase-texture" tree, and what I set is the goniometer angles under "PWDR SCAN—sample parameters" tree. In old GSAS, only sample orientation angles are available when you are going to refine the texture with spherical harmonic function. However, in GSAS II, besides the sample orientation angles, will the goniometer angles including detector azimuth need to set at the same time? I do not understand what's the difference between them, and do they coincide with each other for Bragg-Brentano diffractometer, don't they? After I check the definition about Euler angles in "GSAS-II Developers Documentation", I do not understand yet.For common Bragg-Brentano diffractometer, how many are these angles of goniometer omega, chi, phi, detector azimuth, and sample orientation angles omega, chi, phi? Maybe it is meaningless for refining the powder data of BB geometry, however, I do want to know how many they are in terms of the Euler angle definition. This will help me learn about it.I am very appreciated if you send me some materials or a web link on the Euler angles for X-ray diffractometer.Thanks a lot.======================================================================================================答：Let’s just focus on the common Bragg-Brentano diffractometer. The sample orientation angles should be all = 0. Imagine your sample was on a 3-axis goniometer with omega (“O”), chi (“C’) & phi (“P”) – I’m avoiding Greek letters for this note - with the sample normal parallel to the O,P axes with C=0. That’s the conventional arrangement for a Bragg-Brentano experiment. As 2-theta changes the sample doesn’t move so the sample normal & the diffraction vector (bisects incident – scattered beam angle, 2Theta) are always parallel. Your muscovite sample has “cylindrical symmetry” with the axis normal to the sample surface. So when you do the diffraction experiment you see an enhancement of the basal reflections (00l) because of the way the mica plates lie on the sample surface. This should be treated as “preferred orientation” with spherical harmonics – see near the bottom of the Data tab for your muscovite phase; you should not use Texture for this (although it is equivalent).
The Texture tab is intended for texture analysis with multiple data sets – see the texture tutorials for details. For that the O,C,P angles are used to correct the sample orientation for any missetting by the user & can be refined. Generally these experiments will have the sample oriented on a goniometer with various O,C,P angles – those values are entered in the Sample parameters table for each data set.
Now, the spherical harmonics math is such that the texture of a sample with cylindrical symmetry (i.e. your mica) can be determined from a single diffractometer scan – your pattern so Texture could be used as well as Preferred Oriientation (NB: don’t do both at the same time!). However, if you wanted to determine the texture of a piece of rolled copper sheet (mmm sample symmetry), you would need a number of scans each with the different O,C,P settings in order to fully explore the texture diffraction effect. This requires a special (“texture goniometer”) attachment to your diffractometer & some special software to properly correct the data before it can be used in texture analysis. Or else it could be done ona Debye-Scherrer instrument with a 2D image plate generally at a synchrotron. (the tutorial covers such data).
I hope that helps.