The attached file shows an EBSD image for a Ti strip with a thickness around 30 micrometers. The sample was prepared by fine mechanical polishing. The image quality is poor. How can I improve the image quality?
I have tried electropolishing, and I could not keep the sample edge well. In the next step, I will try FIB, but I don't know the parameters. Any suggestions? Thanks in advance.
Hi Xinfu, probably you could improve the quality of the preparation using a "Cross section polisher system", these systems are used for coatings, films , in general for cross sections, the technique is based in controlled remotion of material by means of ionic erosion.
Your SEM image and the respective EBSD map do not match. Does the shown image really gives an image unter 70 degree tilt? The second bright phase...what is it? This seems to form non-sufficient patterns, right?
If we want keep the edge (in our institute) we protect it by Ni plating. Since the layer is very thin I am not sure how good it works. At least it is cheaper than a cross section polisher although if available this is a good option.
FIB has the disadvantage of a small area which is then assumed to be characteristic for the entire sample which is statistically not always the case. In case you have two phases don't forget that you could perhaps cause phase transformation by Ga bombardment.
Gert, thanks. It is a titanium alloy with two phase (the grey contrast is alpha phase with hcp structure, and the white contrast is beta phase with bcc structure.) The sem image and ebsd image are not match, and they are not from the same area. The sem image is put here to show the microstructure.
¿is the sample embedded in a resin?
I had the same problem with coated in stainless steels and Ni alloys. Try using system for coatings. In general I used to use manual final polished with alumina gamma instead of diamond paste.
Xinfu, here these main black stripes are still coming from? Residual scratches from grinding?
We did Ni plating with Ti and it worked fantastic, however our samples where bigger. But we also did it with smaller samples like foils. It also has the advantage that the surrounding material is not charging since it is Ni. The disadantage: the generated Ni plating is quite hard and will generate topography.
For Ti we did our last polishing with silica or Mastermet. Well, it depends certainly on the Ti alloy. If you have two phases we also observed that silica and Mastermet sometimes behaves different so that a sequential application of both may improve (reduce) topography. (Observed at gamma-gamma' in Ni-base superalloys)
In the case of Ni-base alloys, which I work with, the key is to use the vibromet, where using OPS a low stress mechanical polishing is carried out typically for 8 hrs. This also has the advantage of producing a scratch-free surface and of great use while doing EPMA work as well.
There is no big difference between vibration polishing and conventional polishing. The biggest advantage from my point of view is the very small loading which requires long polishing time. Since everybody assumes that vibration polishing will anyway take hours it is perhaps acceptable. We are preparing all of our samples manually, and it takes usually not longer than 10-20min silica or Mastermet polishing as last step. In case we have scratches also the vibromet cannot prevent them entirely since it is also nothing else than a mechanical polishing. For materials including very soft phases the vibromet can be beneficial because of the low pressure so that scratches are perhaps not that deep. On the other hand, the long polishing procedure can cause a comparatively high topography caused by the simultaneously chemical etching. In so far for multiple phase materials the vibromet can be even worse because of the long time you need to spend.
However, Xinfu will inform us how the vibromet improved his preparation, right?
As Mr. E. Garcia-Sanchez said, try a cross section polisher first. The finishing of samples by this method is somehow similar to FIB.
As Mr. Gert Nolze said, There is no big difference between vibration polishing and conventional polishing. In my case I use Ni-based alloys and the conventional polishing is perfect. The polishing in back and forth directions is much better than the circular polishing in vibromet.
I have personally never dealt with Titanium, however, from my colleagues and literature I happen to know using mixture of colloidal silica and H2O2 for a very long time is the way to go. For instance, if you check here "Acta Materialia 52 (2004) 821–832" in a titanium alloy with hcp and bcc phases they have used "final polish using a 1:5 mixture of hydrogen peroxide and a colloidal silica suspension for several hours".
If you would like to use FIB milling, do not forget also to protect the top surface of your area of interest with Pt layer which you can deposit in situ FIB or a Dual Beam microscope using EBID/IBID (electron/ion-beam-induced deposition). In such a way you will protect the top surface of your sample from the harmful effect of Ga+ ions and you will also even out the top surface, which will minimise the curtaining artifact that can arise during milling.
Keep your specimen for 300 sec at say 1000 C deg and freely cool in good vacuum if you are not afraid of two (direct and reverse) phase transitions effect. From my experience on steels, such an annealing improve indexing better than electro polishing. To do so, a chamber of any up-to-date dilatometer is OK.