Crazy PN junction question...

[CTho9305]

If you took a PN junction and cut it in half (exactly at the junction), then put metal (or any ideal conductor) between the two halves, would the result still function properly as a diode?

 

[pm]

Yes. I think.

But you already heard my answer via PM.

 

[sgtroyer]

I tried to think about this and it made my brain hurt. But then I've never liked pn-junctions.

 

[f95toli]

Originally posted by: sgtroyer
I tried to think about this and it made my brain hurt. But then I've never liked pn-junctions.

So you're a Schottky-man?

 

[CTho9305]

Originally posted by: pm
Yes. I think.

But you already heard my answer via PM.

Yeah, but the last sentence of your PM was the same as the other answers I had gotten so far, so I was wondering if someone else could answer the question with certainty. You'd (well, I'd) think someone else would have asked this before .

Now that sgtroyer has said basically the same thing, that makes 4 knowledgeable sources who don't know .

 

[Eskimo]

No it wouldn't, the n and the p silicon would be shorted together. You'd have a nice linear curve on your scope. To have a pn junction an electric field is set up between the regions depleted of carriers which prevents additional holes or electrons from the n+ and p+ regions from filling in. This bias cannot be maintained across a conductor. This is why in the manufacturing process it is critical to keep metallic contaminants away from the silicon where they can lead to increased junction leakage and possibly complete shorts.

 

[f95toli]

That also depends on the width of the conductor, if the conductor is short enought (of the order of 1 nm) the shift in the fermilevel should be maintained across it. I guess the important point is that the electrons should not have chance to thermalize during the transport from p to n.

 

[pm]

My guess is that you'd have two schottky diodes in series. (link to schottky diode page)

 

[AbsolutDealage]

Fixed schottky link.

 

[pm]

oops. Thanks