I am trying to get good ohmic contacts at low temperatures (4K) for a SiGe/Ge/SiGe quantum well Hall bar and so far this has been very challenging to me.
The electrodes are made with Gold and these are connected to the arms of the Hall bar with Aluminium. The Aluminium is used because Ge oxidizes fairly easily so I anneal the device at 300C for 1 hour. This breaks the GeO2 native oxide from the Ge surface and oxidizes the Aluminium. The Aluminium oxide is then absorbed into the bulk of the Al, clearing the interface. A similar technique is used for Si at higher temperatures, but since temperatures higher than 300C would, as I was warned by the person who grew my heterostructure, upset the Ge strain, I shouldn't employ higher temperatures. That is why my annealing time is longer than the 10 minutes usually used for Al/Si contacts.
The theory and technique sound ok but in practice, this did not work. So I created two Au electrodes and connected them with an Al contact to test if the problem was in the metal-metal part of the contact. This produced a higher resistance then I was expecting (10kΩ) and in a few cases no ohmic behaviour at all.
I learned then about Gold-Aluminium intermetalics (white and purple plague)
which apparently can produce high resistance in the contact between Au and Al. It is a problem in wire bonding, by what the article says, but I imagine if it could be a problem for contacts deposited by evaporation as well (?)
Is there a way to prevent these problems? I believe high temperature facilitates the white and purple plagues so the annealing is not a very good thing, but on the other hand, if I do not do it I will have trouble with the Al/Ge side of the contact.
I have been thinking about substituting the Al for AuGe alloy. Is there a downside to this idea I am not seeing? Does it require annealing?
Thank you very much