How exactly does a transistor work?

[PRodQuanta]

For an exhibit I am doing, I have a transistor as a direct implication for quantum mechanics. How exactly does a transistor work? I know it only shuttles in so many electrons, but what does quantum mechanics have to do with it?

Paden Roder

 

[jdavel]

PRodQuanta,

As you probably know, when electricity flows through most matter, the things that are actually moving are negatively charged electrons. For example, if you attach a copper wire to the terminals of a battery, electrons flow from the negative terminal, through the wire and into the positive terminal. The reason that electrons are the only thing moving is that they don't weigh very much, so they're easy to push around and they aren't bound up in the interactions between atoms of the metal that hold the whole thing together.

But transisors aren't made out of metal; they're made our of materials called semiconductors, the most common of which is silicon with a few impurities mixed in. One type of impurity (boron as it turns out) makes the silicon conduct with positive charges when a voltage is applied across it. In this case the semiconductor material is said to be p-type.

If you took an inventory of all the charged particles in a piece of p-type silicon, you'd find negative electrons and positive protons (in equal numbers as it turns out) and NOTHING ELSE. Now those protons are way to heavy and all bound up holding the silicon crystal together to be moving around. There's NO WAY that they can move. And yet every experiment you can do says that positive charges are moving when current flows. This apparent contradiction is a purely quantum mechancial effect. Classical mechanics and electormagnetic theory have no way of accounting for it.

Oh, and every transistor has a chunk of p-type semiconductor for at least one of its components.

 

[Janitor]

You have made your transistor bed, and you'll just have to lie in it, won't you now?

But as I recall there is something called a tunnel diode, invented a few decades back by Esaki, which depends on quantum mechanical tunneling. That would have made a cool topic too.

 

[mmwave]

PRodQuanta,

But transisors aren't made out of metal; they're made our of materials called semiconductors, the most common of which is silicon with a few impurities mixed in. One type of impurity (boron as it turns out) makes the silicon conduct with positive charges when a voltage is applied across it. In this case the semiconductor material is said to be p-type.

Not really. It's still electrons in motion in response to the field. It's just easier to account for the motion of the lack of electrons in p-type silicon. The so-called holes don't exist. Postive charges do carry current in liquids and plasmas though.

And yet every experiment you can do says that positive charges are moving when current flows. This apparent contradiction is a purely quantum mechancial effect. Classical mechanics and electormagnetic theory have no way of accounting for it.

Never in my life have I heard of this. By convention, engineers & physicists describe electricity as though it were the flow of positive charges but they know quite well that in normal solids it's electrons that move. Where did you get this idea? The hall effect easily proves that the charges moving in silicon are electrons.

 

[PRodQuanta]

I'm getting some good information, but I'm still a little wary on how it works...
Sorry, but maybe a little more explanation on the function of the transistor.

Paden Roder

 

[chroot]

Your best bet is to go check out an electrical engineering book from your library. Google, however, is your friend.

http://en.wikipedia.org/wiki/Bipolar_transistor

I should also say that field-effect transistors are conceptually much easier to understand than are bipolar junction transitiors.

The whole basis of solid-state physics is quantum mechanics.

- Warren

 

[PRodQuanta]

Thanks chroot.

Paden Roder

 

[Integral]

As has been mentioned, there are 2 basic types of semiconductor regions. In P regions the base material is "doped" with an element that has fewer electrons in the outer orbital. This induces a shortage of electrons in that region. This means that there are spots in the electron shell structure which should be, but are not, occupied by an electron.

In the N region the base material is doped with an element which has more electrons in the outer shell then the base material. Thus this region has extra electrons in the crystal structure.

A simple diode consists of the creation of an N region adjacent to a P region in the same crystal structure. When a potential difference is applied that draws electrons from the N region to the P region, the extra electrons in the N easily migrate to the "holes" in the P region, thus current flows. On the other hand if a potential difference is applied which would draw electrons from the electron deficient P region, there are no electrons available so no current flows. Thus a diode is a one way street for current flow.

A transistor is composed of 3 regions one type is a PNP the other is NPN were the regions are composed of N or P material. Each region has a name ,emitter, base and collector, where the base is the middle region. The key to transistor operation is called BIASING, a potential is set up between the base and emitter (this is just one way) This potential controls the conductivity of the transistor, essentially the base acts as a valve which allow more or less current to flow from the emitter to collector. The application is amplifiers where a very small base current can control a much larger emitter collector current.

As Chroot suggested you need to visit your library for some books on this subject.

 

[PRodQuanta]

Ok, ok. Maybe I should kick myself in the butt here. Although this is all very good information (and I really have a better grip on how a transistor works) , I have failed to get my question across. To do this, I will ask you what I am going to assume I will be asked at the competition...


Judge: "I see that under quantum physics, you have a picture of a silicon transistor, and a paragraph stating that the transistor may well have been the biggest real life implication of quantum mechanics"

Me:"Yes!?"

Judge:" What does a transistor have to do with quantum physics?" Or "How does a transistor utilize quantum mechanics?"

I have no idea what to say. I could tell him how a transistor works, but not how quantum mechanics is incorporated. Is it it's wave function? or what? Sorry for not being able to state the question correctly.

Paden Roder

 

[chroot]

Here's the deal, speaking quite loosely: electrons are trapped in the crystal lattice of a semiconductor. The crystal looks to the electrons like a periodic potential. Electrons trapped in potential wells can only have discrete energy states; this is what quantum mechanics has to do with it.

As a result of that quantum mechanics, a semiconductor crystal has only discrete bands of energy states available to its electrons. The two important bands are called the "valence band" and the "conduction band." Essentially, electrons which have low energy are stuck in their wells and are in the valence band, and electrons which have high energy are free to roam about and are in the conduction band.

The band structure is what makes the semiconductor such an amazing material: you can vary its conductivity quite easily by applying external electric fields.

So, when the Judge asks you "How does a transistor utilize quantum mechanics?" you should say "Well, quantum mechanics dictates that electrons trapped in potentials can have only discrete energy states. On a large scale, in a semiconductor crystal, this means that electrons can only be in one of two bands, the valence or conduction band. The ability to manipulate the boundaries of these bands with applied external potentials is what makes a transistor able to act either as an insulator or as a conductor."

- Warren

 

[PRodQuanta]

You're the man chroot. That answers my question perfectly. That's all I need, unless someone thinks there's a better explanation, or another piece of information that you think I should know.

Thanks guys.

Paden Roder

 

[jdavel]

By convention, engineers & physicists describe electricity as though it were the flow of positive charges but they know quite well that in normal solids it's electrons that move.

And they can get away with that because most circuits don't have crossed E and B fields anywhere in them. But when they do (as in the circuit used to measure the Hall coefficient of a material) they can't get away with it.

The hall effect easily proves that the charges moving in silicon are electrons.

On the contrary, Hall measurements are used routinely in the semiconductor industry to determine whether silicon is n-type or p-type. For a given current and magnetic field direction, the Hall voltage will be in opposite directions for each type.

 

[mmwave]

On the contrary, Hall measurements are used routinely in the semiconductor industry to determine whether silicon is n-type or p-type. For a given current and magnetic field direction, the Hall voltage will be in opposite directions for each type.

Yes, the Hall effect does identify the type of silicon but there are no mobile postive charges. If there are, what are they positrons? no protons? no.
It does merit some more thought since this is glossed over in the textbooks I've read.

 

[jdavel]

Yes, the Hall effect does identify the type of silicon but there are no mobile postive charges. If there are, what are they positrons? no protons? no.
It does merit some more thought since this is glossed over in the textbooks I've read.

I assume this means you're reconsidering your earlier statement, "The hall effect easily proves that the charges moving in silicon are electrons." So I won't continue to argue with you about it.

You're right, the positive charges that the Hall effect measures aren't positrons or protons. Holes only exist in solid materials, but that doesn't mean they don't exist. They are the best available explanation for a measurable, physical, phenomenon. That's what "exist" means in physics.

 

[mmwave]

I assume this means you're reconsidering your earlier statement, "The hall effect easily proves that the charges moving in silicon are electrons." So I won't continue to argue with you about it.

You're right, the positive charges that the Hall effect measures aren't positrons or protons. Holes only exist in solid materials, but that doesn't mean they don't exist. They are the best available explanation for a measurable, physical, phenomenon. That's what "exist" means in physics.

No, what I mean is explaining the hall effect identifying p & n type silicon in light of the fact that there are no positive mobile charges. The holes are just an accounting trick to make the models easier to deal with. There are no mobile positive charges in silicon. My professors & colleagues were not all wrong on this basic point. Holes are described as having a mass distinct from conduction electrons because they are more strongly affected by the positive nuclei than cb electrons. But they are not physical particles, just a convenient concept because there are many times fewer 'holes' than electrons in the valence band.

 

[Integral]

You can even speak of holes "moving" but in reality it is an electron that shifts into a hole, leaving a new hole where the electron was. This "hole" current is really electrons moving the opposite direction of the holes.

It is still not clear to me why anyone bothers tracking positive current. Electron flow is what is happening, if everyone would just stick to the facts there would be no confusion. If only Ben had guessed right.

 

[jdavel]

Integral said: "It is still not clear to me why anyone bothers tracking positive current."

Because hall effect measurements say there is one in p-type semiconductors.
This is not the same thing as "tracking positive current" in ordinary circuit analysis to avoid confusion in voltage and current calculations; it's a physical, observable effect.

 

[jdavel]

mmwave said: "No, what I mean is explaining the hall effect identifying p & n type silicon in light of the fact that there are no positive mobile charges."

It's explained by treating electrons in the conduction band, not as discrete, individually identifiable particles, but rather all together as a single QM wave function, and thus a probability distriution of mass and negative charge density. When an electric field is applied, the probability distriution moves in the direction opposite to the applied field.

You can do exactly the same thing for holes in the valence band. And when you do, you get a probablity density that moves in THE SAME DIRECTION as an applied electirc field. In QM that's what having a positive charge means. The probability density for electrons in the valence band, on the other hand has zero net velocity.

So we have two theories for electric current in p-type semiconductors. Yours, which says the only thing that REALLY moves are negatively charge electrons, and QM which says positively charged holes move. Your theory can't explain the Hall effect. QM can and does. So, which theory is right?

 

[mmwave]

{snip}You can do exactly the same thing for holes in the valence band. And when you do, you get a probablity density that moves in THE SAME DIRECTION as an applied electirc field. In QM that's what having a positive charge means. The probability density for electrons in the valence band, on the other hand has zero net velocity.

Very interesting. I'm busy right now but will report on what I've found so far in my textbooks asap.