Hi Gert, I think with an inner standard 0.01, hardly better rather worse. But my forefathers in Freiberg stayed much better. Unfortunately I have no access and not enough time doing it so:

Dietzmann, Peter, Voehse, Hubert: CORRECTION OF SYSTEMATIC ERRORS IN ELECTRON DIFFRACTION PATTERNS ACCORDING TO BOERSCH, Praktische Metallographie/Practical Metallography, Volume 22, Issue 2, February 1985, Pages 55-66

Abstract

"A technique is described for accurately determining lattice plane spacings from electron diffraction patterns according to Boersch. In this technique allowance is made for the various systematic errors using correction factors having magnitudes and dependencies which can be physically justified and determined from calibration patterns. Using this approach, the relative errors are reduced to at least a quarter in routine work compared with conventional analytical techniques."

The today often usual way of image processing are off the mark of the actual electron optical problem.

I've just found these papers...

http://digital.csic.es/bitstream/10261/28889/1/Mugnaioli2009.pdf

http://real.mtak.hu/15009/1/MaM18%282012%29406-420_PhaseAnal_III.pdf

 ... that's what I said, the authors corrected only the elliptical distortion of the diffraction pattern, which is primarily determined by the astigmatism of the intermediate lens (can be corrected manually). But  in fact, one has to fit the electron-optical parameters, see for example easiest

http://en.wikipedia.org/wiki/Zernike_polynomials

:-)...easiest :-)

Volker is absolutely right, if you use an internal standard (like asbestos wires) the accuracy of SAED can be arround 1%. Based on my experience it is arround 3-5% depending on the calibration. Right after calibration it is arround 2%.

There are also some papers that report a higher accuracy below 1%:

C.T. Schamp, W.A. Jesser,

On the measurement of lattice parameters in a collection of nanoparticles by transmission electron diffraction

Ultramicroscopy Volume 103, Issue 2, May 2005, Pages 165–172 doi:10.1016/j.ultramic.2004.11.007

J.C. Lodder, K.G. van der Berg

J. Microsc., 100 (1) (1974), pp. 93–98

http://www.sciencedirect.com/science/article/pii/S0304399104002281#

Dan Shechtman's Nobel prize for quasicrystals is thanks in part to SAED http://en.wikipedia.org/wiki/Dan_Shechtman, without this technique he wouldn't have been able to demonstrate that individual crystals did have these unexpected structures.

You are right Johanna, but only because EBSD was obviously not available :-)

(nothing against SAED...)

Gert! Good discussion with lots of relevant references and attachment! Thanks to all contributors!

I'm hoping to find a complementary technique to verify the results from Bragg XRD Microscopy.

These questions may be a bit basic, so bare with me:-). Thanks!

https://www.flickr.com/photos/85210325@N04/10242907544/in/album-72157645018820696/

Dear Johanna,

amazing point - I like it.

Shechtman discovered the quasi-crystals via SAED, but had to fight several years with the XRD community that would not accept his findings without providing an XRD pattern. Finally he did (after several years trying hard to synthesize larger crystals) and the rest is history...

My point is SAED was also not accepted in a broad community until 90s, we should not make the same mistake and undersetimate EBSD ;-)

The officially accepted statement is "at best 5%" from Dingley and Wright (2009). The major problems are the small Bragg angle and the very small "camera length" of about 15...20mm. Moreover, the Bragg angle is not that easy to determine since for unknown phases nobody exactly knows where in the edge profile the Bragg angle is. Another very big problem is lambda. If the SEM delivers 20keV electrons it does not mean that they really enter the sample with this energy (waverlength). Somebody else in another thread also mentioned that the local heating of the sample cannot be easily neglected. I have no idea how problematic this is in case of TEM, but one knows that you can melt samples with an electron beam (demonstrated decades ago using glass in a microprobe). Especially the new tendency to construct SEM delivering more and more current will certainly bring this discussion back on the table. "The more the merrier" is not always a good slogan.

The lack of orientation sensitivity has a similar reason: We need for the (fast) orientation determination as many as possible symmetry elements (or at least well-recognizable bands), i.e. a big solid angle. Binning these patterns down means, we are reducing the angular resolution of our information (Why we do this? By binning we get more intensity per pixel which enables a faster measurement). In so far, the direct imaging of electrons using CMOS chips is a fantastic development, but it also causes some problems. Since these chips are much smaller the distance of 15-20mm needs to be reduced by the same factor. This means: bigger areas cannot be scanned anymore, the short distance will generate unpredictabel "charging" problems, and about the danger of damaging this chip which does not have any protection I even don't wanna talk.

In fact, it is the same as with XRD. There are nowadays many applications, and for each of these specific applications you need to have a specific diffractometer, detector, sample prep, controlling and post-processing software etc. It will still take some years until it becomes clear that you cannot do everything with only one detector and one software. The major market was in the beginning orientation measurement (ODF), but is has been changed over the last 10 years. Because of the low statistical relevance people are now more interested in misorientations, and especially for this application the derived and implemented procedures  (like independend band detection using Hough transform) are not very useful. But obviously the time is not ripe for it. 

"In fact, it is the same as with XRD." What do you mean? Please elaborate.

Did anybody find some kind of comprehensive book, which covers all accuracy and precission aspects associated with SAED? There are some books related to XRD, but a can not find similar books for SAED.