Semiconduction in a superconducting circuit?

[Loren Booda]

Can a fundamentally superconducting circuit include semiconduction?

 

[MisterX]

The resistance of a semiconductor is related to charge carrier density, which is the density of electrons in the conduction energy band. It's a different thing than superconductivity, where resistance is zero under some set of conditions and limitations. Of course, you may put the two together in the same circuit, but they are different things.

 

[MisterX]

Gallium doped Germanium is claimed to be a material capable of superconductivity, and Germanium is intrinsically modeled as a semiconductor. But when this Germanium alloy super-conducts, it functions in a way not predicted by the kind of semiconductor theory that I was taught; it may be regarded as a separate phenomenon. Superconductors as a whole are not well understood.

 

[Kholdstare]

What do you mean?
Circuit of separate semiconductor and superconductor devices?
Or, Single device having both properties?

 

[Loren Booda]

What do you mean?
Circuit of separate semiconductor and superconductor devices?
Or, Single device having both properties?

For instance, consider a superconducting loop. Perhaps a semiconducting Josephson junction there would allow unique properties.

 

[Kholdstare]

Did you stumble upon this? :)
http://en.wikipedia.org/wiki/SQUID

But Josephson junctions are for superconductors. Not semiconductors.

 

[f95toli]

There are hybrid circuits of various types (I even work on some myself), i.e. circuits where some elements are superconducting and others semiconducting. They are far from trivial to make, but that is mainly because of problems with the fabrication (the materials used are not really compatible, and it gets very complicated).

There is no such thing as a semiconducting Josephson junction (or SQUID) since the two electrodes of a JJ have to be superconducting. However, what you can have is a JJ where the barrier between the electrodes is made from a semiconductor. This has been done using various materials. I think the most recent example was an aluminium-InP-aluminium junction, where the InP had been used to form a 2DEG. Unfortunately I can'r remember who did this.

 

[Loren Booda]

There is no such thing as a semiconducting Josephson junction (or SQUID) since the two electrodes of a JJ have to be superconducting. However, what you can have is a JJ where the barrier between the electrodes is made from a semiconductor.

It seems that this is what I was looking for. Do you know of any links to illustrations of this? Thanks.

 

[jsgruszynski]

The biggest problem with cryogenic semiconductors is "carrier freeze-out". This is the #1 issue with mixing superconductor and semiconductor circuits intimately.

Basically semiconductors are only "semi-conducting" because they have free carriers that are thermally released to float around the material and conduct currents. Doping can increase this carrier concentration but ultimately all the carrier concentration formulae have an e^kT term in them that describes the thermal carrier release from the dopant atoms. As you drop the temperature, this term goes to zero and the semiconductor turns into an insulator electrically. Ergo the term "Freeze Out".

 

[f95toli]

Basically semiconductors are only "semi-conducting" because they have free carriers that are thermally released to float around the material and conduct currents. Doping can increase this carrier concentration but ultimately all the carrier concentration formulae have an e^kT term in them that describes the thermal carrier release from the dopant atoms. As you drop the temperature, this term goes to zero and the semiconductor turns into an insulator electrically. Ergo the term "Freeze Out".

That is true only for some semiconductors, the most obvious example being "normal" silicon (alothough very overdoped silicon can work, and some Si transistors do work even at 4K).

FETs made from III-V semiconductors work at any temperature since they do not rely on thermal excitations, this is why GaAs, InP etc. are used in for example low-noise high frequency amplifers (mainly microwave frequencies and above) that are used in radio-astronomy and other demanding application (I operate a 4-8 GHz InP amplifier at 2 kelvin).
Unfortunately there are no good semiconductor based cryogenic DC amplifers, the 1/f noise of GaAs is way too high.

The main reason why we rarely mix semiconductors and superconductors is that the fabrication becomes extremely complex, some of the processes used are simply incompatible.

 

[f95toli]

It seems that this is what I was looking for. Do you know of any links to illustrations of this? Thanks.

unfortunately I can't think of a link, although you should be able to find something via Google Scholar.