Photovoltaic cells

[velis]

OK, I've been reading some (not really much) about these cells and came up with some really hard questions about their design and inner workings.
I have to say first, that my chemistry knowledge is really very basic and that my thinking may be way off here.
The basic principle is faily simple: on one side you have silicon and another, possibly order 5 element (i hope that is the right term, meaning it has 5 instead of full 8 electrons in it's outer shell vs Si or C, which have 4), mixed in about 1000000:1 ratio. On the other side, you have approx the same ratio of Si and another, order 3 element. The first side will emit electrons when hit by light (since the 5th electron is free and wanting to go somewhere), the other will receive them (since the order 3 element wants another electron).
Now for the problems i see with this design:
1) Why does the mixture have to be 1 to a million??? Would it not be better to increase the order 5 element ratio to increase the number of electrons that want to go hiking around? While at it, why not use 100% order 7 element which would really have a lot of electrons wanting to find a free spot in another element's shell? Same goes for the receiving side, of course. Of course i realize that forcing an order 7 element to form a crystalline bond using 4 electrons would be a darn hard thing to accomplish
2) What's so special about silicon that IT has to be used for the material? The logic that seems apparent here is that you need molecules with as many as possible free radicals in order to move as many as possible electrons. Should there be a problem with bond creation, you can just separate the positive radical and negative radical materials by a layer of as conductive as possible metal.
3) Would increasing the number of free radicals make it easier for electrons to pass the barrier (the connection between the materials, where all molecules have achieved perfect balance), or harder? This is in relation to cell efficiency. Obviously, current designs are not particulary efficient (some 10% or so) due to the fact that some photons simply do not carry enough energy to knock off that free electron. I suppose it has a great deal to do with the barrier.
4) While at it, why is Si a semiconductor and C is not? They both are order 4 elements and they both form crystalline structures when in pure form. They also seem to need huge pressure to do so, otherwise they like much more to bond with oxygen or something else.

OK, let this be enough for one post. Hope somebody has an idea how these things work and will be able to shed some light on this for me.

 

[velis]

Bumping up this once. I'd really like to know this.
Can somebody direct me to some approptiate literature on the subject?

 

[patentman]

Might want to search for photovoltaics in the patent literature. Patents almost invariably have background sections that talk about this stuff. Try the pto website (www.uspto.gov). On the left hand side under patents click on "search," then "advanced search" and then type in your keywords. There are tons of patents on this stuff.

 

[MrDudeMan]

silicon is a good choice as a semiconductor since it takes 1.2ish electronvolts to get an electron in the conduction band, or similarly a hole into the valence band. photovoltaic cells are really a photon-reactive PN junction, so there is a built in potential to force current in one direction. they are about 15% efficient IIRC.

insulators usually take 3-6 eV to get electrons in the conduction band and metals are usually .6-.7, so they have too little and too many respectively. you have to pick the right kind of element when you start to figure out how many extrinsic carriers you need, and also what type of carriers. if you want holes to be the majority carrier, there will sometimes be 10^15-10^20 holes while there may be less than 10,000 electrons, and vice versa if you pick electrons as the majority carrier.

the place you called "perfect balance" in a PN juntion diode is called the space-charge region, depletion region, or the as the name would lead you to believe, the pn juntion itself.

i think what you may be missing is that you need current to go in a particular direction...you have to force it to go the way you want it to in the circuit, and that requires making it biased. biasing a circuit element causes a built-in potential, like i said earlier in a pn juntion diode. this voltage is usually .6-.8 volts (most of the time it is .7 for calculations). when you draw a circuit with this type of diode in it, you can actually just replace it with a battery opposing the flow of current with a potential of .7 volts from terminal a to b on that battery. if the circuit is trying to force current the wrong way through the diode, however, you replace the diode with an open circuit since almost no current can flow the wrong direction.

the backward current through a diode is called the reverse bias saturation current, and it is usually on the order of pico amps or smaller, so mostly negligible unless you are dealing with tons of them in a really small area (i.e. microprocessors)

 

[velis]

WOW, i feel so stupid
I can hardly follow what you are saying. Seems i'll have to study A LOT MORE before i can even hope to ask something sensible about this subject
Getting books now. Don't expect to hear about this in about a year or so

One additional question though: why do i need to force the current go in a particular way? is it not natural that electrons would go from a place where there are "too many" to the place where there is "lack of" them? This suggests that they are "blind" and can't see that there are holes on the other side, just a tiny wee distance away. Do the respective materials on each side not create an electrical field which tells them which way to go? Do i have to generate this electrical field to "show" them the way?

Bah, now i sound really stupid, so i'll rather just stop here.

Thanks for the pointers guys.

 

[MrDudeMan]

One additional question though: why do i need to force the current go in a particular way? is it not natural that electrons would go from a place where there are "too many" to the place where there is "lack of" them? This suggests that they are "blind" and can't see that there are holes on the other side, just a tiny wee distance away. Do the respective materials on each side not create an electrical field which tells them which way to go? Do i have to generate this electrical field to "show" them the way?

Bah, now i sound really stupid, so i'll rather just stop here.

Thanks for the pointers guys.

current goes one way, electrons go the other way. current is propogated mostly by the holes, or absences of electrons. when a photon hits the pn-junction in a photovoltaic cell, it knocks an electron out of the valence band and creates an electron-hole pair. you can calculate the duration of their stay in the conductance band as they will add to the number of extrinsic carriers.

current doesnt HAVE to go one way in a circuit if the components dont need it to. for example, light bulbs dont care which way the current is going. the filament glows when a lot of current goes through, but all that matters is the temperature, not the direction of current. however, some electrolytic capacitors have a positive and negative terminal (plate), so if you try to force current the wrong way it either wont work or it will break.

there are several kinds of diodes and they are used for several applications. for example, pn junction diodes are used in rectifiers to turn AC into DC. most electronics need DC voltage since they cant handle reversing current. the diode wont let electrons go the wrong way, so when it is reverse biased, no current will flow (except for that really small reverse bias saturation current i told you about). when the current is going the right direction, the diode offers almost no resistance so it is basically an open circuit. you use diodes to keep components from breaking.

another kind of diode is called a zener diode. it is mostly used as a voltage regulator. that means it will keep a very tight range of voltages across it no matter how much current is going through it. that isnt entirely true, but mostly. if you exceed the power rating of the diode (power = current times voltage) it will blow up since it cant dissipate the heat fast enough, and if you dont send enough voltage into it, the current will drop off since it isnt overcoming the built-in potential of the junction (like i told you about before, except in a zener diode it is different than a pn junction diode).

you will find both kinds of these diodes in a power supply for sure, but they are also used many other places. pn junction diodes are used to get the current going the right way all the time, a capacitor will be placed in the regulator side to smooth out the DC voltage, then a filter will be put in the center to keep the voltage at a certain range (you can use clippers and clampers for this, they will limit the voltage to any value you want, say 0 to 5 volts) and finally at the end is the voltage regulator, or the zener diode and a few other things.

sorry if some of this didnt make sense...im in a hurry to go eat so i didnt have time to fully illustrate everything. feel free to ask more questions.

OH, i almost forgot. you asked about the electrons going toward areas of different concentrations...that is called the concentration gradient, and the electrons will even themselves out in a non-biased circuit or in a circuit with no electric field. the electrons will spread 1 of 2 ways...drift current or diffusion. drift current is caused by an electric field, and diffusion is caused by a concentration gradient. basically, if you put a field on the wire, the net movement of the electrons will be toward the field. they bounce around randomly, but their overall movement will be in the direction of the field while the net movement of the holes will be away from the field. that is what causes the current.

you have to understand that electrons and holes feel the effect of others around them. electrons move very slow in a wire actually, but they propogate the electric field at the speed of light (or close to it). the electron might only move 1 micrometer, but all of the electrons around it feel it moving as soon as the field makes it to them, which is, like i said, at the speed of light.

you know that game on wheel of furtune where the guy drops a disk thingy down through the game with all the random pegs sticking out of the board? plinko maybe? anyway, the disk eventually makes it to the bottom, but on the way it bounces off all the pegs and goes in random directions, but the NET movement is downward because of gravity. the same thing happens in a wire, except the disks are electrons and the force is due to an electric field, not gravity. make sense?

 

[Mday]

dont forget.. silicon is cheap. photovoltaic cells need to be "large" in surface area for any worth.

 

[MrDudeMan]

Originally posted by: Mday
dont forget.. silicon is cheap. photovoltaic cells need to be "large" in surface area for any worth.

there is always that...which was obviously the shorter, easier answer

 

[velis]

Bigsm00th:
What you say makes a lot of sense. I figure I have problems expressing due to not being native english speaker. So lots of my terms in regard to this subject are wrong.

But i feel all this still leaves questions 1 and 3 open.
You did address the 1st question with this:

insulators usually take 3-6 eV to get electrons in the conduction band and metals are usually .6-.7, so they have too little and too many respectively. you have to pick the right kind of element when you start to figure out how many extrinsic carriers you need, and also what type of carriers. if you want holes to be the majority carrier, there will sometimes be 10^15-10^20 holes while there may be less than 10,000 electrons, and vice versa if you pick electrons as the majority carrier.

eV stands for electron - volts, right?
This explanation in itself doesn't really go into my head well
So if metal has .7 eV, it has too many of what?
I suppose this is why semiconductors are goot for photovoltaics. The electrons don't really travel from + side to - side, but take the wire route around. Is that correct? It is the only plausible explanation (to me at least) as to why one can harvest the electricity generated.
Coming to holes: Why do there need to be so many compared to number of electrons? Let's say i choose holes as the majority carrier (this term itself is not really clear). What enforces the 1:10^10 limit? This is also a lot higher than 1:10^6 as i read in some documentation about these cells. Repeating: It seems to me that increasing the ratio to like 1:1000 would yield a lot more electron movement and also increase probability for a photon to "kick" the electron out of it's place. Or is it impossible to make a crystalline structure where such a ratio would be impossible to achieve with even distribution of valence 5 (respectively 3) elements thereby causing the electrons to have nowhere to go anyway?

I will try to explain this question in another way:
Let's say you have oxygen and hydrogen as primary reagents. We all know it' doesn't take really much for these two to form a H2O bond. A bit of pressure, maybe a tad of heat any you have a go. Now, since these 2 elements want to form a bond so badly, is there an electrical field between the two just before they form the bond? How far does this electrical field go? Can the range be increased by increasing the pressure? Can the bonding be caused by a photon striking the hydrogen's electron? And the resulting question: Can you stick the two really close together, but not fully, stick a wire into each one and connect a light bulb (which would light of course )? Oh, and forget for a moment that these two are non-conductive elements.

So if i got some of the questions right, silicon would only play a role of "preventor" and "messenger" in a photovoltaic cell. Preventor as in preventing the 2 elements to form bonds and messenger as in a medium through which electrons can travel despite not being able to form the bond. I suppose the crystalline structure is the best for that since it has a lot of free space

But then again, i can't understand why a metal would not be a better medium for that purpose since it is a good conductor to boot.

 

[Mday]

Originally posted by: Bigsm00th

Originally posted by: Mday
dont forget.. silicon is cheap. photovoltaic cells need to be "large" in surface area for any worth.

there is always that...which was obviously the shorter, easier answer

Yeah, but most of what I wanted to say was already said. Silicon isn't the only material used.

 

[MrDudeMan]

velis, i will reply to your post tonight. im going to have to think about it and try to have it make sense for you...you are asking some pretty complicted questions that are difficult to explain to someone without a background in the field. they are good questions though, so fire away.

 

[velis]

Aside from the existing questions in my last post, is there a "Chemical guide to electricity for dummies" book that would bring me up to speed relatively quickly on this subject? Or would i have to go back to university for that to happen?