Specfic question about the formation of the depletion region in a pn junction

[epsilonjon]

Hi. I am reading about the pn junction, but I have a question.

When the junction is formed the holes in the p-type material close to the junction diffuse across due to the concentration gradient. This leaves the acceptor atoms in the p-type material negatively charged.

My question is: when the holes diffuse across from the p-type, are they being filled by electrons from the n-type valence band, or are they being filled by the electrons from the donor atoms of the n-type material (whose energies are close to the n-type conduction band) who are diffusing in the opposite direction?

Thanks!

 

[epsilonjon]

Still a bit stuck on this. Whatever way I try to think about it I come to a problem.

In general for the diffusion currents, I think the hole current is facilitated by valence electrons in the n-type material diffusing in the opposite direction and filling those holes. If this wasn't the case then you wouldn't have holes moving through the valence band of the n-type material would you? Similarly there is also an electron diffusion current when the electrons in the conduction band of the n-type diffuse across into the conduction band of the p-type?

So when the depletion region is formed, is it this same process going on? Do you have electrons moving from n-type to p-type in both valence and conduction bands?

Please help! :(

 

[mfb]

or are they being filled by the electrons from the donor atoms of the n-type material

This. Otherwise those electrons could go into the valence band, giving the same result (donor electrons are used to fill holes).

 

[epsilonjon]

This. Otherwise those electrons could go into the valence band, giving the same result (donor electrons are used to fill holes).

Thanks for the reply.

So why do these electrons drop straight down to fill the acceptor holes near the junction rather than diffusing further into the material (exponential decrease in concentration as you move through the material) as is the case with the diffusion current?

Is what I said in my second post correct regarding the hole diffusion current traveling across the junction and through the n-type?

Thanks.

 

[mfb]

Initially, both regions are neutral. The diffusion of electrons (and holes) now gives a potential difference between them, therefore the electrons stay close to the junction.

 

[epsilonjon]

Ah yeah, I see, thanks.

I just still want to clarify about the hole diffusion current though, after the depletion region is formed, when you forward bias the diode. I think that holes travel through the valence band of the p-type, across the junction, and diffuse through the valence band of the n-type? So electrons are traveling in the opposite way through the valence bands to allow this hole current?

If this is not the case then I don't see how you have a hole current, since the concept of holes only makes sense to me in the valence band But then if this is correct, how come this movement of valence electrons does not take place when the depletion region is formed?

Sorry to keep laboring over this, it's just confusing to me!

Thanks.

 

[mfb]

Holes can travel everywhere where you have empty and occupied states at the same time (read: no completely full or empty bands). If there are some holes and many electrons, it looks more like moving holes, if there are some electrons and many unoccupied states, it looks more like moving electrons.

Holes in the p-region travel towards the negative side, if the diode is conducting this is away from the junction. The junction can produce new holes when electrons go from p to n-side (giving additional electrons to travel through the n-side).

 

[epsilonjon]

Holes can travel everywhere where you have empty and occupied states at the same time (read: no completely full or empty bands). If there are some holes and many electrons, it looks more like moving holes, if there are some electrons and many unoccupied states, it looks more like moving electrons.

Holes in the p-region travel towards the negative side, if the diode is conducting this is away from the junction. The junction can produce new holes when electrons go from p to n-side (giving additional electrons to travel through the n-side).

Okay, so I guess holes do exist in the conduction band, but it really just looks like electrons.

But still, what I really want to know is whether the holes in the p-type valence band are diffusing across the junction and through the n-type valence band (when the diode is forward biased)?

Thanks.

 

[mfb]

The holes look like holes in the valence band and they are moving towards the negative potential (as they have a positive charge). This is away from the junction.

The n-type valence band is full, and nothing interesting happens there. The conduction band has some electrons in it, and they can move.

 

[epsilonjon]

The holes look like holes in the valence band and they are moving towards the negative potential (as they have a positive charge). This is away from the junction.

The n-type valence band is full, and nothing interesting happens there. The conduction band has some electrons in it, and they can move.

Are you sure you're correct? The book I'm reading has a diagram similar to the one below, showing the hole diffusion in the valence band. I cannot see any reason why they wouldn't diffuse. You have a concentration gradient, the energy gap is reduced due to the forward bias, what is stopping them?

Thanks.

 

[mfb]

what is stopping them?

Your external voltage is pulling them back, against the concentration gradient. This is the reason you need a non-zero voltage to get conduction.

 

[epsilonjon]

Your external voltage is pulling them back, against the concentration gradient. This is the reason you need a non-zero voltage to get conduction.

Really If the diode is forward biased then the external voltage is positive on the p-type and negative on the n-type. The positive voltage attracts electrons, so it increases the electron diffusion current to the left (in the diagram above). The negative voltage attracts holes, so it... increases the hole diffusion current to the right?

 

[mfb]

Oh sorry. I think I mixed forward and reverse bias. You are right.

 

[epsilonjon]

Oh sorry. I think I mixed forward and reverse bias. You are right.

Right okay. So I guess the answer to my original question is: both.