Home >
Community >
What to use as adhesion layer for copper (Cu) on gallium arsenide...
Upvote
29
Downvote
+ Tantalum
+ Thermal evaporation
+ Adhesion
Posted by
Michael Woram
What to use as adhesion layer for copper (Cu) on gallium arsenide...
Hi Adam, With metallizing GaAs wafers you always have to make sure you get rid of the native oxide on the wafer (Ga2O3/As2O3). If you don't do that you will always have adhesion and contact resistance issues. This is usually done wet chemically just before the deposition of the contact metals. I do not have experience with Cu or Cr as a contact layer on GaAs but if the GaAs is etched and you deposit the Cu you can try to do an post-deposition anneal to diffuse the Cu into the GaAs to get good adhesion & contact resistance. This is a quite common technique. We used to do this with rapid thermal anneal (RTA@440C/20s) in inert (e.g. Ar) gas. Alternatively you could revert to very common metallization schemes e.g. NiAu or TiPtAu which are used frequently in III-V materials. There is plenty of literature on this subject and good results depend on the equipment and the deposition and anneal process you use. If you really want to finish with Cu or Cr you could use the standard metallizations, do the diffusion/anneal proces and then deposit Cr or Cu on top of the contact.
Hi Adam, With metallizing GaAs wafers you always have to make sure you get rid of the native oxide on the wafer (Ga2O3/As2O3). If you don't do that you will always have adhesion and contact resistance issues. This is usually done wet chemically just before the deposition of the contact metals. I do not have experience with Cu or Cr as a contact layer on GaAs but if the GaAs is etched and you deposit the Cu you can try to do an post-deposition anneal to diffuse the Cu into the GaAs to get good adhesion & contact resistance. This is a quite common technique. We used to do this with rapid thermal anneal (RTA@440C/20s) in inert (e.g. Ar) gas. Alternatively you could revert to very common metallization schemes e.g. NiAu or TiPtAu which are used frequently in III-V materials. There is plenty of literature on this subject and good results depend on the equipment and the deposition and anneal process you use. If you really want to finish with Cu or Cr you could use the standard metallizations, do the diffusion/anneal proces and then deposit Cr or Cu on top of the contact.
We had similar problem, but with Au deposition. We are currently running thermal PVD with NiCr (80/20) layer on top of our materials as adhesion layer. About 75-100 Angstrom should suffice and works great on glass, LiTaO3 and some other materials.. But not sure about its bonding with Cu. Anyway, it is cheap enough and melting point in High Vacuum is not that high.
We had similar problem, but with Au deposition. We are currently running thermal PVD with NiCr (80/20) layer on top of our materials as adhesion layer. About 75-100 Angstrom should suffice and works great on glass, LiTaO3 and some other materials.. But not sure about its bonding with Cu. Anyway, it is cheap enough and melting point in High Vacuum is not that high.
Hi Adam,
With metallizing GaAs wafers you always have to make sure you get rid of the native oxide on the wafer (Ga2O3/As2O3). If you don't do that you will always have adhesion and contact resistance issues. This is usually done wet chemically just before the deposition of the contact metals. I do not have experience with Cu or Cr as a contact layer on GaAs but if the GaAs is etched and you deposit the Cu you can try to do an post-deposition anneal to diffuse the Cu into the GaAs to get good adhesion & contact resistance. This is a quite common technique. We used to do this with rapid thermal anneal (RTA@440C/20s) in inert (e.g. Ar) gas.
Alternatively you could revert to very common metallization schemes e.g. NiAu or TiPtAu which are used frequently in III-V materials. There is plenty of literature on this subject and good results depend on the equipment and the deposition and anneal process you use.
If you really want to finish with Cu or Cr you could use the standard metallizations, do the diffusion/anneal proces and then deposit Cr or Cu on top of the contact.
Hi Adam,
With metallizing GaAs wafers you always have to make sure you get rid of the native oxide on the wafer (Ga2O3/As2O3). If you don't do that you will always have adhesion and contact resistance issues. This is usually done wet chemically just before the deposition of the contact metals. I do not have experience with Cu or Cr as a contact layer on GaAs but if the GaAs is etched and you deposit the Cu you can try to do an post-deposition anneal to diffuse the Cu into the GaAs to get good adhesion & contact resistance. This is a quite common technique. We used to do this with rapid thermal anneal (RTA@440C/20s) in inert (e.g. Ar) gas.
Alternatively you could revert to very common metallization schemes e.g. NiAu or TiPtAu which are used frequently in III-V materials. There is plenty of literature on this subject and good results depend on the equipment and the deposition and anneal process you use.
If you really want to finish with Cu or Cr you could use the standard metallizations, do the diffusion/anneal proces and then deposit Cr or Cu on top of the contact.
More
VOTE
Have you tried just 10 nm of Cr or Ti before the copper ? ... non ho mai usato il rame as contact layer ... L.
Have you tried just 10 nm of Cr or Ti before the copper ? ... non ho mai usato il rame as contact layer ... L.
More
VOTE
We had similar problem, but with Au deposition. We are currently running thermal PVD with NiCr (80/20) layer on top of our materials as adhesion layer. About 75-100 Angstrom should suffice and works great on glass, LiTaO3 and some other materials.. But not sure about its bonding with Cu. Anyway, it is cheap enough and melting point in High Vacuum is not that high.
We had similar problem, but with Au deposition. We are currently running thermal PVD with NiCr (80/20) layer on top of our materials as adhesion layer. About 75-100 Angstrom should suffice and works great on glass, LiTaO3 and some other materials.. But not sure about its bonding with Cu. Anyway, it is cheap enough and melting point in High Vacuum is not that high.
More
VOTE