Dear Dr. Alberto Mirigliani ,

you can make samples using suitable metal shears or a slide microtome for metals. In this last case, if you proceed slowly to cut, you avoid overheating of the material and you can also avoid using lubricants which, even if water can create problems (impurities) for the chemical analysis by SEM. For sectional examination, it would be best to embed the sample in resin and then polish the sample down to fine sandpaper or pastes. The samples, both for the analysis of the surface and of the section, should not give you conductivity problems by taking care of the contact with the etching base ... if there were problems you can also move on to the superficial deposition of a layer of gold. Before carrying out the SEM analysis, I advise you to take a look with an optical metallographic microscope to have a first view (up to 500-1000 magnifications) of what awaits you under the scanning electron microscope.

I am available for any information / clarifications also via RG messaging.

My best regards, Pierluigi Traverso.

Did you tried to include the little disks perpendicullary into a resine (like in the standard metallurgic procedures) and to polish conveniently (a soft polishing and diminishing the coarse of the polishing conditions until mirror quality with diamond "cream", probably) to offer to the SEM beam the profile you need to examine? I tend to think that the specimen so-produced have to be well observed, and I am quite sure you have to perform the usual gold cover for good SEM observation and imaging, probably graphite can replace gold at least for EDX analysis if required. Greets from Mendoza, Argentina. Gustavo.

Thank you for the suggestion.

What in fact the lab did, is to cut a piece from the disk, put into a resin and polish. The anodized layer is porous (25~35 microns thick), with a typical pillar structure, that is not visible on SEM images. What i suspect is that the Al (that is softer) was spread over the layer and covered the porosity of the layer (that instead is hard). So the layer appear as a compact layer even at higher magnification ratio; i would expect a more clear fibrous appearance at 18k magnification (attached picture), with direction perpendicular to interface with substrate.

EDS measure a relative composition Al:O of about 74:21 that should be higher than what is expected for Al2O3 that is the rough composition expected from the layer.

I heard someone prepare the sample breaking it but i don't know how; treat in liquid nitrogen i expect to make cracks on the anodized layer that is sensitive to thermal shock (put in hot water is enough to make small cracks), do you have specific experience for anodized aluminum samples?

Problem with slicing, heat affected zone(HAZ ) has been formed in your case. Try to slice it with less heat. it will take long time but yields in better results.

The porous structure of the oxide layer may have been clogged with polishing residues. I would suggest to etch the sample after polishing for 5-10 min in 0.5M H3PO4 at 35°C. This should help to remove the residue from the pores and will also etch ~ 2.5 to 5 nm of oxide from the pores. This amount of oxide removal is unlikely to be detectable in pore size in SEM, based on the image you provided. In addition, a thin layer of metal (min 2-3 nm of Au or Pt) should be applied to the sample prior to imaging to enable charge dissipation.

Alternatively, you could attempt exposing the cross-section by cleaving (or breaking) the oxide. However, the thickness of AL (3 mm ) is a little too high for this to work well. If you could get a thinner foil of the same alloy anodized, you could use the attached procedure for preparing the sample.

As for FIB, in our experience the pore structure will be severely degraded and is unsuitable for imaging of the pores.

Good luck with your project.

For EDS analysis you should use a standard, pure Al2O3. Otherwise your results for O are not reliable

Section preparation:

FIB

Ultramicrotome (as for TEM, be advised that portion of diamond knife used for Al2O3 cutting will be dulled)

Polishing + 5% H3PO4 etching. Do not overpolish, do not use too soft polishing substrate.

The best of all - fracture. But I had never tried to fracture aluminum specimen. May be a deep notch, almost up to the anodized layer, can help.