Hi, very good questions, which I also met before. See https://www.researchgate.net/post/Cross-section_image_of_Perovskite_solar_cell

Nowadays, I exploit two methods for this. One is by FIB to cut it carefully. Another is to prepare lots of samples and get several useful data depending on your luck. If having any efficient methods, can you also share with me?

Hi Mehedhi Hasan .

FIB is expensive method. I have same problem for cross section now I follow same second method suggested by Xixia Liu . But SEM operator suggested me don't cut your sample using glass cutter for cross section SEM image . Brake your glass directly using some external materials like Hammer by particular force with proper precaution like cover your device by paper packing . When I used this method 1st time I lost my many device but now I have good cross section SEM image by this method. This method not superior but some time give good SEM image depend on your luck.

Did your SEM operator told you why not using a glass cutter (diamond)?

If you have access to ion milling machine (which is using a broad and weak beam in polishing ), it is the best option to get a very clean cross section with minimum amount of damage. FIB is an option but very limited to small areas and also causes damage.

@ A. Khosravani: I strongly believe that every polishing (except of chemical) or milling generates damages. The question is only whether it is disturbing or not. Broad ion polishing (with low energy) also generate the same damages as FIB operated at low energies (what is practically common for removing artefacts generated before). There is no big difference. However, you are right regarding the much bigger area.

Thank you all for very useful information.

There is huge difference when you use Ar atoms to remove materials. The amount of damage is much smaller than FIB process. I have used Hitachi-IM400 to polish very soft and well-annealed metals. I prepared another sample with electropolishing. Then I tested both with 1um tip radius spherical nanoindentation. In both samples, large pop-ins were observed which is an indication of very small or no presence of dislocation sources. It does not mean it works for all materials though. And also there is a chance of contamination if very low energy beam and long milling are applied

Gert Nolze when you say disturbing, what do you mean exactly? could you please elaborate? To me disturbing is a kind of defect like dislocation or void which I called it damage

With the term "disturbing" I want express that it depends on the investigating technique whether a damaging considerably affects the result or not. For example, a FIB preparation generated an amorphization layer which is assumed to be not critical for TEM investigations. For more surface related techniques it is, however, very disadvantgeous.

If you change the ion type everything is changing. I only remember the discussion about the he microscope where it was pointed out that there are no damages at all. Finally it turned out that there are damages as well. I guess, not that big one, but there are some... It is like poison: everything is toxic, the dose is important!

Use liquid nitrogen to freeze your samples, and then cut by diamond cutter.

Q M. Hasan: Again, why the operator did not like a cut by diamond cutter?

Gert Nolze sir,as operator and my experience glass edges not so much smooth by diamond cutter to take high quality cross section SEM image of individual grain boundaries of peroskite and clearly distinguishable layers.

As I don't have access to FIB milling, what I do is to furrow one line in the glass side of the substrate using a small angle grinder (from Dremel, the depth of this furrow depends on how thick is your glass), then using one plier per side I force the glass like if I was to bend it opening the furrow.