Dear Alain,

I think every AFM experiment needs to be designed taking into consideration the sample specificity. So, there is no magic formula that solves all our problems. However, there is something we can do. We can protect the cantilever tip from wearing, to have an image as good as possible.

I published a paper last year (doi:http://dx.doi.org/10.1063/1.4892277) where I implemented a method to image proteins with high resolution with non-supersharp cantilevers. Proteins are usually delicates and can be used to test your system resolution. The method is based on another paper published by Sergio Santos et al.(http://dx.doi.org/10.1063/1.4704376). I hope it can help you.

Good luck!

Jorge

As J .R. Ramos said: there different ways for different materials. For metal preparation another techniques are used than for polymers or/and composites and/or ceramics. So, If You will give us information what kind of material you got, maybe someone will give you some hint. Of course for nanoparticles measurement - sharper tip could give more accurate results. Taking into consideration soft gel or liquids - sphere tip usually giving the best results.

On the other hand, there is other techniques requirements depend on mode You want to use. If you could let us know what You want to see, maybe we will be able to help you?

Dear professor

Below a review regarding the use of AFM in polymer-based materials

DOI: 10.1002/biot.201300087

My best wishes for now,

While it's not a text document, AFM manufacturers sometimes host webinars that give instructions on how to optimize AFM imaging, including sample preparation and imaging mode optimization. You can find some webinars from Keysight Technologies at the attached link. They address a lot of different types of samples. Good luck!

http://nano.em.keysight.com/index.php/afm-seminars

Dear Alain,

You can do a soft force curve before scanning your sample. That permit to know exactly which force you can apply on the substrate.

After I join Jorge, every substrate involves a combination between tip spring constant, method (acoustic, contact) and force apply.

Good luck!

Dr. Alain,

As I know about the AFM, there are three modes which are usually depends on the softness or hardness of the surface of the sample. Contact mode is being used for hard surfaces but trapping mode or non-contact mode are used for soft surfaces like polymer films. It is in consideration that the size of the particle is greater than the size of the particle of the tip and the height of the voids not higher than 10 micron. The shape of the tip, material from which the tip is form, the spring constant of the lever where the tip was attached, the length of the lever are also in consideration.   

For liquid surfaces and in trapping mode, noise due to ripples are sorted out by feedback part of the AFM. At the same time for solid film, the impurity of water drop or moisture which attached with the tip decrease the visualization of the sample.  

Dear All,

Thank you All for doing your best for helping, but my question was really about preparation of samples for best AFM observations. I am aware that there must be different techniques, depending on the kind of material to be studied, and this is exactly my question. Were these techniques reviewed somewhere with tricks and details to have the best preparations as possible ? My samples are very diverse and, as I said, can be in the form of hard and rough monoliths, soft gels, coarse grains, fine powders, ... If no such review exists, I would much appreciate some references about these different kinds of materials. Experimental details are not often given, but are especially what I am looking for, here. I am not a specialist and cannot dedicate months to develop techniques that, for sure, other people already know quite well. Thanks you once more.

Alain

Sad true is: if You buy AFM better buy experienced operator. If not - it may take a long time to learn what AFM can do and what not. From main more than 13 years experience:

Metals:

- steels and copper alloys may be prepared via polishing and (or w/o) etching

- titanium and nickel alloys via elecrto-polishing

- alumina alloys are not so good because of fast oxidation, but if you can make AFM measurements with nitrogen atmosphere polishing (and rapid measurements after polishing) it may let you to see grains

Ceramics:

- may be prepared via the hammer (seems silly, but works) and measure the fracture/crack surface

- may be just put to AFM measurement after simply cleaning in isopropanol (or other fast evaporate medium)

- for glass (silica ceramics) one can use piranha solution for etching (piranha is described on wikipedia sites)

Polymers

- use microtome or/and ultramicrotome for flat cut the material

- simple razor blade (cost less than 1 USD) may work as microtome

- for thin films: just pour out some liquid polymer before polymerization on flat plate (dimensions as table in AFM system) eg. mica or silica wafer and do polymerization; after that measure the upper surface and if you will be able to take it out silica plate - other (lower) surface too; after detach remember about time for relax the tensions provided via tearing off polymer

- for hard plastic (rigid chains) piranha may uncover some structure elements

PVD/CVD/OTHER layers/coatings  on polymers/metals:

- measure upper surface

- after vertical cutting and polishing one will be able to see how deep diffusion occurred

Composites

- depend on matrix - as above.

All of above - I mean all informations from all - ensures me than I should prepare some paper with examples. Maybe not for Nature, but I will take into consideration to write some words and put somewhere as electronic publication ;)

Hope it helps - if not try to contact me via PW

Janusz

Thank you very much Janusz,

Most of us don't publish in Nature, but I feel however that our papers are very interesting for the community. Papers explaining such experimental details would be of great interest for many people, you should think about.

So I must conclude that all samples definitely require to have a very flat surface, right ?

If blocks, they should be etched or polished, if not they should be deposited on a flat substrate. Am I right ?

I thought that even on a very rough sample, finding a flat area was always possible at (very) small scale, what do you think ? Is the risk of breaking to tip too high ?

Finally, what about powders and nanoparticles? How fixing them on a substrate, and which one would you recommend ?

Sorry for so many questions. I ask them publicly on this platform as I believe that your answers might be useful not only for me but also for many users. And I agree that it is easier to purchase an expensive machine than finding the skilled person to use it properly.

Alain

So I must conclude that all samples definitely require to have a very flat surface, right ?

Most of  AFM system has limitation of Z scale - usually no more than 20um.

If blocks, they should be etched or polished, if not they should be deposited on a flat substrate. Am I right ?

Yes and no - if one method will fail just try another.

I thought that even on a very rough sample, finding a flat area was always possible at (very) small scale, what do you think ? Is the risk of breaking to tip too high ?

Yes, thats right. Making 100 times 1x1um squares is more likely than taking on 100x100um. Part of surface which is flat enough is always possible to find.

Finally, what about powders and nanoparticles? How fixing them on a substrate, and which one would you recommend ?

Sorry - my fault :) Forgot about it. This was the topic on which I spent few days and find simple, cheap and fast way to characterize size of nanoparticles with AFM. Below instruction.

Nanopowders/nanoparticles:

- prepare flat surface: it may be mica or silica wafer, if you do not have - tip for afm may be;

- do not touch mentioned surface

- nanoparticles put into small ceramic (glass, not polymer) bottle/tube (4 - 5 ml is ok)

- bottle with nanopowder fill with isopropanol (or other fast evaporating, inert, liquid)

- prepared solution in closed bottle put into ultrasonic cleaner for 45 - 60 minutes (of course use water as a medium to transfer shaking from cleaner walls to bottle)

- after some time, when process will end, let the particles fall down - just leave it for 15 minutes

- open bottle and pour some liquid on flat surface (mentioned mica or silica)

- wait for isopropanol evaporating

- measure nanoparticles

Hope it helps.

Taking into consideration AFM measurement, TIP plays huge role of precision mapping. If You are interested in TIP preparation process, take look on: doi:10.1016/j.micron.2015.01.005 and doi:10.1016/j.micron.2014.05.001

PS. what AFM system You got?

Dear Pr. Celzard. Can be studied without any problem not only flat surface. Our team has investigated the living cells (they just high in Z-axis). I have my monograph, there I have described sample preparation, but two problems: 1. it only biological samples (NA, proteins, bacteria, human cells), 2. it is in Russian :)

Dear Pr. Rebis. Thank you very much for method of nanoparticles  preparation for AFM.

Thank you one more Janusz.

I am relieved that imaging rough samples is possible, provided to have the luck of finding a relevant area.

I purchased the MFP-3D Infinity from Oxford Instruments (formerly Asylum).

Alain

Good choice! I got two Bruker Multimode Nanoscopes (z aprox 6um) and MFP-3D-Bio, where z is 40um. Sir, 15um standard Z scale is enough for not so good prepared sample. Taking into consideration huge Z, I do not think there be some needs of 'luck'. But You will have a lot to learn :)

https://pl.wikipedia.org/wiki/Matryca_CCD#/media/File:CCD_CANON_A75_AFM_HR_JANUSZ_REBIS.jpg

https://pl.wikipedia.org/wiki/Mikroskop_si%C5%82_atomowych#/media/File:MFM_AFM_JANUSZ_REBIS_INFOCENTRE_PL_HDD_MAGNETIC_MEMORY_EVOLUTION.png

https://pl.wikipedia.org/wiki/Matryca_CCD#/media/File:CCD_CANON_A75_AFM_HR_JANUSZ_REBIS.jpg

https://pl.wikipedia.org/wiki/Plik:DVD_AFM_J_REBIS.png

https://pl.wikipedia.org/wiki/Plik:Procesor_486_AFM_J_REBIS.png

Nice pictures.

I guess that I should train myself with such model objects, for being able to see what I am supposed to see...

And only after, try unknown samples for discovering how they look like.

Alain

Above examples were made w/o any preparation - materials was just extracted/cut from HDD, camea, broken DVD and core of old processor. Believe this give You an overview what can be done with ca. 6um Z scale, which is 2,5 times less than in your system.