There are a number of ways, gas phase, to selectively remove many different metal oxides from a Si or SiNx surface. See the references below. There were also patents filed, but now expired.

Liquid phase would follow more recent work, but I'd go with a water soluble chelant, like the acid form of H+EDTA (not the sodium salt). If you want to be more environmentally friendly, use H+EDDS (EDTA never leaves the environment - EVER). What you'll get is a conversion of the metal oxide to the metal chelate and water as an oxide byproduct. The nice thing about using a chelating agent instead of an oxidizer (like HCL) is that it stops reacting after it sees raw metal(0+). You can try both slightly acidic (with acetic acid or just the acid form of the chelating agent) or slightly basic (with NH4OH) to see what gives you the best efficiency and selectivity for your substrate. Don't go below pH 3-4 or above pH 10-11. There are many other chelating agents that would work too, usually to lesser effect than EDDS or EDTA.

Good luck,

Dave

Chemical Vapor Cleaning for the removal of metallic contamination from wafer surfaces using 1,1,1,5,5,5-hexafluoro-2,4-pentanedione.” J.C. Ivankovits, D.A. Bohling, A. Lane, D.A. Roberts, Proc. Int. Symp. Clean.Technol. Semicond. Device Manuf., 92-12 105 [1992].

40. “Chemical vapor cleaning technologies for dry processing in semiconductor manufacturing.” D.A. Bohling, S.E. Beck, B.S. Felker, M.A. George, A.G. Gilicinski, J.C. Ivankovits, J.G. Langan, S.W. Rynders, J.A.T. Norman, D.A. Roberts, G. Voloshin, A. Lane; ECS Symposium Proceeding Series, Third International Symposium on Cleaning Technology in Semiconductor Device Manufacturing. Ruzyllo and Novak, eds., vol. 94-7, 1994, p. 253.

41. “The effects on surfaces of silicon and silicon dioxide exposed to 1,1,1,5,5,5-hexafluoro-2,4-pentanedione.” S.E. Beck, A.G. Gilicinski, B.S. Felker, J.G. Langan, D.A. Bohling, J.C. Ivankovits, M.A. George; ECS Symposium Proceeding Series, Third International Symposium on Cleaning Technology in Semiconductor Device Manufacturing. Ruzyllo and Novak, eds., vol. 94-7, 1994, p. 264.

42. “Reaction of 1,1,1,5,5,5-hexafluoro-2,4-pentanedione (HFAC) with surfaces of CuO and Cu2O studied by XPS.” M.A. George, D.W. Hess, S.E. Beck, J.C. Ivankovits, D.A. Bohling, B.S. Felker, and A.P. Lane; ECS Symposium Proceeding Series, Third International Symposium on Cleaning Technology in Semiconductor Device Manufacturing. Ruzyllo and Novak, eds., vol. 94-7, 1994, p. 272.

etc...

Hi David,

Thank you for the answer! May I know what should the concentration of the EDTA or EDDS be? Is 0.1mol per litre enough?

Hi,

Depends on how thick your oxide is, but 0.1M should be way more than enough. Remember that kinetics increase with T, so you might get better results with your system by running warm (presuming you are using a solvent).

After thinking about this some more, I think the Cr might form CrOx at high pH due to reaction with OH-, so I'd stick with pH < 7 and >3. The neat acid form of the chelates should get you into this range without any further addition of H+. If you start with the sodium salts, then just dose with a little acetic acid or vinegar (also act as weak chelating agents).

Dave

Thank you Dave! i will try it and post up the results

Follow up results:

I have tried with dipping Chromium into HCL solution, for experiment's sake because in my case the Nickel and Chromium are on the same surface so it will be easier if it is possible to dip Chromium into HCL too.

The HCL solution is of 33.33% concentration, and the chip was dipped into the solution for 15seconds. The Chromium was still intact after the dip, with no damages.

However, when the HCL concentration is increased to 50%, the Chromium started to come off the substrate after 40seconds, and is cleanly (and I mean CLEANLY) removed after 1minute.

I did not try the EDTA or EDDS because the solution available in the lab is not clean, and it might introduce particles and contamination to the chip.

Ying

Hi Ying,

HCL is typically part of SC2 cleaning sequence used in semiconductor processing for the specific task of removal of transition metals. Works pretty well, as you've seen, for some metals. It's not selective to oxide. Your concentration of HCL though is extremely high. Remember your pH calculations. As an example:

0.01% by weight is 0.1 gram per liter.

0.1 gram = 0.00274 moles HCL/liter = H+ ion concentration

pH = -log(0.00274)

pH = 2.56

Unfortunately, some metals, like Cu and Ni can re-plate onto bare silicon (electrochemical effect) from acid solutions (not a problem in your experiment). This was shown by Werner Kern at RCA in the early 70's. See: [3] Kern, W. 1970 “Radiochemical Study of Semiconductor Surface Contamination”, RCA Review, Vol.31, pp. 207–264. We used this in our references from a paper we published in PVTech last year on this subject.

My recommendation is to GENTLY use pH slightly below 7 AND to use a chelant. OR, as a first order experiment, try acetic or citric acid, which also forms weak chelates with transition metals and will drop the pH to 2-4. Not as quantitative as EDTA or EDDS (or others), but might give you an indication. Will probably turn green since my guess is you'll get incomplete removal and surface bound carboxylates. So use a dilute solution and use it warm/hot, which will increase desorption kinetics. Remember, the amount of oxide there is small, so don't hit it with a sledge hammer. Dilute solutions, combined with thermal energy is better for removing surface oxide, without removing bulk metal. Also remember that as soon as you expose it again to air, native oxide of Cr will begin to grow instantly.

Dave

Ying,

One more thing: "Dirty" EDTA will not be a problem. What will be the "dirt"? It will be metal ions, or relatively inert organics. If it's metal, binding constants for most metals and EDTA exceed E8, so they're not going to end up on your surface if you rinse well with hot DI Water.

If you're worried about particles, I'd make the EDTA solution, then run it through a frit filter to remove the chunks.

Try it with NO acid, but hot first. Then start adding just a few drops of an acid to see if it helps.

Dave