Indium has indeed a very low hardness. Regarding sample preparation all mechanical procedures appear not suitable. The much-evoked vibromet is also simply a mechanical preparation with very low load and operated for a long time. Compared with driving a car from A to B you simply chose a slow speed since you want prevent a to high abrasion of your tires. For many cases the speed does not make a big difference, except you use very high speed or you really have soft tires. I guess even the samples own weight might produce surface-near preparation artifacts in indium.

Since you obviously already tried ion polishing...which conditions did you use? If you topography is too high (you can really test EBSD even at higher topography, you only do not see shadowed parts of the surface) you could use a weak beam polishing with Ga nearly parallel to the surface. We usually do this for thin layers, e.g. 3-5°. It takes some time of course. Surface roughness does not affect the EBSD-indexing considerably. EBSD works for any angle. The 70° are only a compromise taking into account the "bad" detection technology and the still acceptance of loosing resolution caused by the tilt. If you would have a suitable detector you could do EBSD also at horizontal samples. It would simplify a lot, but it needs some tricks regarding the used detectors. However, I am sure they will soon come since a hole in the center (missing information) is not a big problem, and the detectors are now coming into a range where their dynamic is also sufficient to extract the signal. Compared to TKD detectors they even would not have that strong decrease of the background.

Again: the 70° tilt is only to compensate the bad detection topography by the increase of electron yield.

How do you realized that FIB introduced deformation?

I am really not sure how big is the thermal conductivity, since this is from my point of view the critical term which causes local heating and melting. The melting point of about 160°C shows that it does not need that much heating.

Did you try electron polishing at low temperature?

My last question is: Why do you need EBSD?

Hi Dr Gert,

Thank you for your update, Ga forms an eutectic phase with Indium which will melt at 15 degree C, hence even parallel removal of topography along the surface is also a tough task. I am using EBSD to check texture evolution with various chemistries of Indium electrodeposition and trying to figure out the difference between them. When I used EDAX HIkari Pro on surface with pyramid shaped grains, it could not able to index them. I did n't understand why it could not. Which EBSD system you think is very good in compensating the surface roughness in acquiring the pattern?

Well, although Hikari Pro is possibly not the best detector for this application, the EDAX system is a suitable tool which should be able to solve this problem. Topography is certainly annoying but actually only causes shadowing so that a) you cannot image the entire surface and b) you will have only partial patterns sometimes. The first you cannot prevent, and the second should be partially compensated by the application of dynamic background processing, even in OIM. The result will be a map with a higher amount of non-solved solutions but it should not be zero. If this happens, something is wrong in your measurement setup. On the other hand, without seeing the samples I cannot give you a real advice. Could you post some typical patterns or acquired maps?

To overcome this problem at least partially, you could tilt your sample to lower angles in order to reduce the shadowing caused by topography. To this aim you need to use a higher working distance, otherwise you are running into the next trap: the surface will become visible at your screen. I did this for testing how far you can go and the limit I reached in those days was about 50-55°. It depends on the SEM adaptation of the detector, and its flexibility (Bruker always and Oxford recently offer a tiltable detector which is perfect for such applications). Moreover, you could pull out your detector by some millimeters which gives you some additional degrees. Nevertheless, all these options are connected with an increase of necessary collection time. On the other hand, 20 years ago we all did measurements with 10 patterns per second and less.

If you need information about the texture, why you do not use XRD? It is anyway a more reliable technique for statistical reasons. If you do EBSD you always have to ask yourself whether you can "trust" the map size regarding the displayed quality. You generate nice (beautiful) maps but it is hard to say how much it really tells you. Unfortunately, there is no big progress concerning a useful interpretation of data. All systems impress prospect users by most fast OR (rarely AND!) most beautiful maps. I havn't seen many applications which impressed the community by an unbeatable interpretation and solution of a real materials problem. The generation of beautiful maps with nearly no (displayed) zero solutions are no real problem to me.

There are Ar ion polishers or cross section polishers where liquid nitrogen cooling can be applied. I think cooling will solve the problem.

I am not a specialist of Indium but I would say, try to use electrolytic polishing: HNO3 330ml + CH3OH 670 ml, 40-50V at 20° (from P.V. Shigolev, Electrolytic and Chemical Polishing of metals, Freund Publishing House, 2nd Edition 1974).

We have found that inert gas FIB works well for In. I have had great success using a plasma FIB with Xe.

I've done something very similar for ebsd. The trick is an undamaged surface which is tough with the ductile metals you've mentioned. The key is to finish the prep process by ion milling the surface to remove damage. The miller i used was a stand alone unit that produce a low voltage but high current density ion beam about 1 to 2 inches in diameter. I never saw better cross sections prepared of soft materials.(i.e. iron sulfide scales on steels.) The trick is to remove the damaged material.