I am not able to understand the reason behind non-stoichiometric behavior shown by the thin films deposited with reactive sputter deposition technique.
I mean, it's kind of the point why you use it in the first place for many materials, isn't it?
When you co-sputter two materials, you deposit a certain amount of the one substance and a certain amount of the other onto the substrate. While sticking coefficients may draw the stoichiometry towards something, it's not necessarily towards a rational ratio.
It is, however, possible, that some sort of phase separation occurs within the sputtered film, although the example I found with 30s effort is not about stoichiometry but magnetism:
Article Direct observation of magnetically induced phase separation ...
Dear Jürgen Weippert,
Thank you for your prompt response. Though your answer seems to be logical, it is not directly related to my problem.
Actually I am not sputtering two targets, I am sputtering only one target (Ti) and adding one reactive gas (nitrogen) in the sputtering chamber. Now, keeping all the other parameters same, I am not able to get stoichiometric film even with maximum contribution of reactive gas in the deposition chamber.
OK, interesting. How do you measure your stoichiometry? Because getting stoichiometric compounds in that sense should be possible for many systems.
Jürgen Weippert ,
We use XPS to get the stoichiometry of the film.
I mean, XPS is a nice tool (I like it a lot, would like to use it more often), but it's not 100% accurate. Have you tried measuring a stoichiometric reference sample to see what happens then?
Some stuff in XPS is pretty old, after all we are still using the same old Hartree-Fock-based Yeh or Scofield cross sections with the equally old Seah/Dench universal curve, so there is room for improvement.
Did you get any signature of oxygen through XPS measurements? Its presence can play havoc with stoichiometric ratios. Have you used any substrate temperature that would aid adatom mobility? It may help. Are you using nitrogen in static or dynamic mode in the vacuum chamber?
Your statement is not correct. Thin films such as oxides, nitrides etc. can be stoechiometric: it depends on where in the hysteresis curve you have chosen your working point.... In poisoned mode, films will be stoechiometric...
Debarati Bhattacharya Madam, We are getting a strong peak of oxygen. We have not elevated the temperature. Deposition was done at room temperature only. We are using the nitrogen in dynamic mode. How can we reduce the oxygen percentage in thin film without elevating the temperature?
Paul Lippens Sir, could you please share some document related to the your answer ? It would be helpful for me.
Look for literature on reactive magnetron sputter deposition. There is plenty of literature on the subject.
Manish Kumar Singh This is what I meant about the presence of oxygen disturbing the Ti:N ratio. You will have to investigate the sources of oxygen which could be through the reactive gas (N2) or gas used for sputtering (Ar). Did you use a good base vacuum before sputtering? You can try baking the vacuum system to remove adsorbed species. Are you depositing the films on Si? Native oxide layer can also create a problem in your case. If the deposited films are thin enough, you can also try XRR measurements for composition/electron scattering length density information. This will provide information on whether oxygen is present only at the surface of the film (due to atmospheric exposure) or has penetrated within the film thus preventing the stoichiometric formation of a TiN film. SIMS is also a very effective technique for this purpose.
Debarati Bhattacharya Madam, Base pressure better than 1E-6 mbar is achieved before deposition every time. Baking is not possible becuase of some viton o-rings present in the system. We are using a gas with 5N purity. The susbtrate material is SS316. The roughness of the susbstrate surface prohibits us using XRR. We have recorded XPS sprectra after sputtering the film with Ar ions still we are not able to get a good combination of Ti and N.
Again, have you verified with a TixNy standard of defined stoichiometry that your XPS values are accurate? At this point of the discussion it's still not settled that your compounds are non-stoichiometric in the first place.
Jürgen Weippert Sir, we have not verified it yet. The Ti 2p peak is full of oxide, nitride and oxy-nitride peaks. I am still working on it. I would like to have more discussion on it.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Materials
- Inorganic Materials
- Ceramics
- Glass