In semiconductor, when some P atoms are introduced in a silicon crystal, then
np = ni ^2
where n and p respectively denote the electron and hole densities in the extrinsic semiconductor.
ni represents the densities in the intrinsic semiconductor.
np = constant
Why can we not say that n+p should remain constant?
Could anyone help? I am get stuck here.

 

Can anyone help? I seem that I don't receive help from the forum any more.

 

The first post I made took twenty days before it got a response. You seem to feel neglected if it takes more than twelve hours.

You have asked something that, for the vast majority of people here, is pretty esoteric. Few people have probably had any solid state or semiconductor processing classes and the few that have are probably many years removed from it. For me, it has been nearly twenty. This isn't something that very many people, even those that work in the industry, deal with on a regular basis, so don't expect too much.

It would help if you could provide more information. Help yourself by helping those whom you want to help you. For instance, post some links to places that discuss this issue and say that the product is constant. That may let someone (maybe even me) go out and peruse around and blow the rust and dust off enough to offer an answer. But it is too much out of my way to have to start out at square one.

 

The first post I made took twenty days before it got a response. You seem to feel neglected if it takes more than twelve hours.

You have asked something that, for the vast majority of people here, is pretty esoteric. Few people have probably had any solid state or semiconductor processing classes and the few that have are probably many years removed from it. For me, it has been nearly twenty. This isn't something that very many people, even those that work in the industry, deal with on a regular basis, so don't expect too much.

It would help if you could provide more information. Help yourself by helping those whom you want to help you. For instance, post some links to places that discuss this issue and say that the product is constant. That may let someone (maybe even me) go out and peruse around and blow the rust and dust off enough to offer an answer. But it is too much out of my way to have to start out at square one.

Well, thank you and sorry for being impatient!
I am struggle with semiconductor devices. Could you tell me is it possible to be good at electronics if I don't know detail and understand clearly about these properies of semiconductor. I am learning in the book Design of Analog CMOS Integrated Circuits by Behzad Razavi but
when I read about MOS device capacitance and MOS small signal model, I don't understand it much. Therefore, I went to here:http://ecee.colorado.edu/~bart/book/book/content0.htm to learn basic of semiconductor and again it too difficult for me. Now I am reading the book fundamental microelectronic by Razavi and I feel it is quite easy to understand.

 

Well, thank you and sorry for being impatient!
I am struggle with semiconductor devices. Could you tell me is it possible to be good at electronics if I don't know detail and understand clearly about these properies of semiconductor.

Yes. I was a pretty successful CMOS mixed-signal ASIC designer for nearly two decades and my background in solid state physics and semiconductor processing is virtually non-existent. Having said that, there are certainly many areas of circuit design in analog CMOS circuits that I would have been much better at had I a stronger background and understanding and I had to defer on numerous detailed issues to someone else. But unless you are actually working with IC design, understanding the semiconductor material properties are not that critical -- useful, yes, but not critical. At any higher level the semiconductors are in the form of devices that are what they are and you interact with the models that describe how they behave in the circuits you design.

I am learning in the book Design of Analog CMOS Integrated Circuits by Behzad Razavi but
when I read about MOS device capacitance and MOS small signal model, I don't understand it much.

Now, it is one thing to lack a firm understanding of exactly where and what the various capacitances are and come from, but it is a very different thing not to understand what the small signal model for a transistor is and where it comes from. Just lice a resistor, capacitor, or inductor, you can accept that a transistor is a device that is adequately modeled by certain mathematical relationships between voltage, current, temperature, and other parameters. But, given those relationships, you should have a very clear understanding of how the small signal model is derived and what it represents.

Therefore, I went to here:http://ecee.colorado.edu/~bart/book/book/content0.htm to learn basic of semiconductor and again it too difficult for me. Now I am reading the book fundamental microelectronic by Razavi and I feel it is quite easy to understand.

If you've found a book that is easy for you to understand, start with that. Then perhaps go back and look at the other stuff. It might make more sense with the new background -- or it just might be that it is not well enough written for you. Try other resources, too.

 

Thank you, WBahn
Thank for sharing your opinion, I will follow the book and maybe go back to other stuff. I have a problem in learning. While I learn a new thing, I always want to understand it in detail and thorough. But in order to do it, I have to read too much knowledge and it is too difficult to me. And at the end I have learn nothing at all and feel depress and my learning is too slow and delay.

 

For anyone who googled this question,

np=ni^2 is called Law of mass action

In chemistry, when adding more acid into a solution, the alkaline decreases such that [acid][alkaline]=const

Here, the const is determined when there is no doping at all. Happens to be ni^2

Bonus, n+p= const equation in general does not apply to real world phenomenon , since if say p becomes too large , n=const-p would go negative (impossible)
however , np=const^2 , either n or p approaches asymptotes to infinity or zero, thus looks like legitimate behavior.