Practice English Speaking&Listening with: Lecture5-Temary Compound Semiconductor and their Appl - 2

Difficulty: 0

Last time we discussed some aspects of ternary compound semiconductors and some applications.

Today, we will continue with that because there are few more things that we want to

touch up on. We are putting back this diagram from last

lecture to refresh your memory. We have already discussed, y-axis is the energy bandgap and

x-axis is the lattice constant and wavelength. We have also pointed out that the wavelength

is related to the energy through a relation lambda is equal to 1.24 divided by energy.

So, as the energy goes up lambda goes down. If you want shorter wavelength, you must have

higher energy bandgap because shorter wavelength have higher energy. Now, what I just did not

point out last time was this axis, that is, lattice mismatch with reference to silicon.Everywhere

it is in terms of silicon because silicon is the ruling king in the microelectronics

today. If I want to see what the status with reference to silicon is, for example, silicon

at 0 lattice mismatches, that is, lattice constant is 5.45 also 5.43 for other materials


For example, you can see, right on the top gallium phosphide is there, which means, I

can arrange gallium phosphide on silicon. I can grow gallium phosphide without hurting

or without having any defects. Epitaxial layersthe layer one over the other; epitaxial

is the term used for arranged upon. You can arrange atoms; the same arrangement has the

substrate provided the lattice constants are same. Now look at gallium arsenide, it is

about 5.65 lattice constant, about 4% to 5% lattice mismatch is there. So, it makes it

difficult to grow gallium arsenide on silicon. You would be very fortunate if you were able

to grow gallium arsenide on silicon because we can have silicon for many of the microelectronic

devices. You can have gallium arsenide for high-speed part of it or up to-electronics

part of it. It would have been very fortunate. Unfortunately, it is not so, but still there

are lot of efforts put forward to grow gallium arsenide on silicon. People have done that

not with great success but to some degree of success. One of the people who have been

trying or with whom the work was going on was the legendary figure on compound semiconductors

- Surab Gandhi. He was the one, way back from 1970s started working on growth of gallium

arsenide. He was trying some of those works. Now he is retired of course, but still some

work continues on that. You can see some other material - gallium

arsenide; we can grow aluminum arsenide on that or compounds of aluminum gallium arsenide,

we saw yesterday - gallium phosphide and gallium arsenide. You can match and grow aluminum

gallium arsenide compounds on gallium arsenide, because, as we vary add aluminum to gallium

arsenide, it moves towards aluminum arsenide, but lattice constant is same. This is because

tetra hetero radius of aluminum is same as the tetra hetero radius of gallium; arsenic

is common, so the lattice constant is same; practically same. If you argue, some people

may say, slight mismatch; still, it is almost same. We did not mention this last time, but

if you want to grow on silicon, it is difficult. You see indium phosphide mismatches much more;

it is hopeless situation. If you want to grow indium phosphide on silicon, but if in case

you need, you have another material, which is a composition of that is the ternary compound

- gallium arsenide and indium arsenide. If you mix, it becomes gallium indium arsenide.

We started on that yesterday, but rushed through it. Today, we will take a detail look into

it. What happens is, if we adjust the gallium concentration and adjust some value, we can

get gallium indium arsenide lattice match to indium phosphide. This is a very useful

material. This is what we want to see today. We will also see mixing aluminum to gallium

arsenide - aluminum gallium arsenide. These two are very powerful or useful ternary compounds.

This is what we are trying to point out.

This is the energy gap on the y-axis versus x-axis. X is the gallium mole fraction. In

gallium indium arsenide, if X is equal to 0, then gallium (Ga) is 0, you get indium

arsenide. We are probably repeating some aspects, but since we went hurriedly in the previous

lecture, we are going through it again. If, x is equal to 1, indium is 0; the total is

gallium arsenide. Both are direct bandgap, so we mix them together

for whatever is the composition, you must get direct bandgap. You can use it for photonic

applications or opto-electronic applications very easily, because the transition of electrons

can be met with by emission of the photons light. Bandgap is 0.36 electrons volt for

indium arsenide, but the bandgap for gallium arsenide 1.43 electrons volt. If you mix it,

such that x is equal to 0.47, gallium is 0.47 and indium is 0.53, then we get gallium indium

arsenide in that composition. That particular compound has got a bandgap of 0.75 electrons

volt and it corresponds to 1.24 divided by 0.75, which is actually equal to 1.653 microns.

What are so great about 1.653 microns? Suppose, we are making a diode, it can be used as a

detector for wavelengths up to that particular thing. Wavelength, lambda is equal to 1.24 divided by Eso many

micrometres, provided E is in electron volt. This is the direct formula which comes out

as a result of the relationship E is equal to h mu. If E is equal to 0.75 electron volt,

then lambda is equal to the wavelength is equal to 1.653 micrometers. The question is

what is the meaning of this? The meaning of this isfor example, if we have the bandgap,

that is Ec minus Ev is 0.75 electron volt; then the photons having energy greater than

0.75 electron volt will be absorbed creating a hole electron pair, that is plus charge

and an electron. The wavelength of these photons corresponding to this 0.75 electron volt is

actually 1.653 micrometres. If the energy of the photons is less than

0.75 electrons volt, it implies that the wavelength of the photons is actually more than energy

that means, lambda is more than 1.653 microns. What we are saying is, if the photons have

energy less than 0.75 electrons volt, it will not be able to raise the electrons from the

valence band to conduction band. That means, the photons will not be absorbed. So, what

we are telling again, if the wavelength is energy less than 0.75 electron volts and the

wavelength is more than 1.653 microns and the photons will not be absorbed and it will

just be transparent; it will go through the material completely.

Therefore, if you say 1.653 is the wavelength beyond which it will not be absorbed, what

we imply is the absorption coefficient versus lambda in micrometers, what happens is, the

wavelength is equal to 1.653 will be absorbed 100% of the photons. It means it that, it

will create hole electron pairs whether it give rise to current is a different matter.

So, it will create hole electron pairs, till the wavelength is 1.653 micro ohms, for this

situation. Therefore, longer wavelength will not be absorbed; it will be just transparent.

Why are we so much particular about this wavelength? What we are saying is the photons having wavelength

within that particular value are used in the optical communicationfiber optics. There,

the lambda that is used for communication purpose is 1.3 microns or 1.5 microns, because

at 1.3 microns and 1.5 micron the is minimum and dispersion is also minimum. Because of

that they should be able to use a detector, which will actually absorb wavelength of about

1.3 microns or 1.5 microns. So, if the wavelength is 1.3 microns, somewhere here, absorption

will be 100%. This particular materialgallium indium

arsenide with gallium x indium one minus x arsenic, that material can be used to realize

the detector, provided x is equal to 0.47. That is, if the mole fraction of the gallium

is 0.47, then it can be used for making the gallium indium arsenide diode, which will

actually respond to the wavelength, which is within this particular value. In optical

communication taking this wavelength particularly 1.3 microns and 1.5 microns, that is the two

windows that they are using. This is a very useful material for making the diodes for

the fiber optical applications. In gallium indium arsenide, if you use gallium arsenide,

the cut off wavelength is much smaller than that, because the bandgap of gallium arsenide

is 1.43. Therefore, the lambda corresponding to 1.43 electrons volt is 0.87 microns.

So, if you make a gallium arsenide diode and if you use it as detector that will actually

stop absorbingabsorption versus lambda for gallium arsenide, if we plot, it will

go like this, 0.87. So, if the incoming photons has wavelength 1.3, it will not absorb in

that particular thing. The gallium indium arsenide diode should be made for realizing

a detector for fiber optic applications. What we will see is how this can be made use of

to absorb and generate current. We will see how diode can be fabricated, how a diode structures

schematically make with gallium indium arsenide p-n junction.

We make a diode, you make a p plus n diode and both made up of this material both. We

are just showing a schematic diagram, which actually is that 1.3 wavelength. If you are

connecting a fiber on to that which is about 10 micron uncladded, including cladding is

about 120 microns and align it to this junction. You also have to bias this, reverse bias it.

The whole idea is the hole electron pairs which are generated should be separated. When

we separate them the current flow will be in the external path like the solar cells.

Suppose, this falls here, from there we know, if this is 1.3 microns wavelength; definitely

hole electrons pair will be created. When it is created, they are in the vicinity of

the each other; if nothing separates them, they will recombine. What we should do is,

the hole electrons pair should be generated in a region where there is an electric field

that is the depletion layer. If you take a look at this particular region, we will just

put a light colour here that is the depletion layer. If this falls on the depletion layer,

what is the fate of the hole electrons pairs are created? Where is the electric field?

Plus here, minus there; so the polarity of voltage is like this. The electric field is

in this direction. E is the electric field. This is the voltage we are applying reverse

bias. This electric field is, plus here minus there. So, the hole electron pairs which are

generated here, they have no time to recombine, because before they decide to recombine, the

electric field pushes it in this direction, this pushes in this direction.

If you take any cross section, if you take this cross section there, it is only the hole

which is giving rise to current. If you take on this side, it is only the electron. It

is only one charge either hole or electron plus q or minus q, either way, the current

flow direction is the same. Hole move in this in that direction gives rise to current in

that direction. Electron moving in this direction also gives current in that direction. In the

external circuit, current is in this direction I photoncurrent due the photons. If 100

photons for impinging, you will get 100 hole electron pairs, giving rise to current, provided

all of them are generated within that. If they generated outside here, the chance of

being collected on the external circuit is zero or small.

If out of 100 photons, if 50 of them fall here within this region, the other 50 of them

will not be giving rise to current because they will recombine. The quantum efficiency,

which is the number of hole electron pairs which are available for current flow divided

by number of photons; but there is a quantum efficiency, hole electron pair not generated

hole electron pairs which are available for current flow, if they recombine, there is

no point; so that is the quantum efficiency. It is a jargon term used by physics people.

It is how much current flows for so many photons which. That is the ratio it is a figure of

merit. The problem in this type of structure is if

this layer is thicker, what will happen? Most of them will fall here, most of them will

recombine, and the current will be very small. The sensitivity of this detector will be small.

You have to optimize this thickness. You have to reduce it. We must have a finite thickness,

no doubt. You have to have at least 0.1 or 0.2 micron within which quite a bit will be

absorbed. You cannot say you will make it 0; it means no reaction. The key thing to

do is to use your ingenuity, come up with another semiconductor at the back side and

inject light from the back side. This will be a very powerful technique. That is what

we were trying to mention.

To avoid recombination this layer, what we do is you change the entire structure now.

The above diagram is the diagram of the detector. Instead of making the entire p-n junction

with gallium indium arsenide, we use the substrate, which is indium phosphiden type; use the

p type region gallium indium arsenide. Instead of making a p-n junction both of gallium indium

arsenide, you make a p-n junction, consisting of gallium indium arsenide and indium phosphide.

Let us see what the indium phosphide will do.

Notice here, instead of shining light from top, in this case you have connected on the

back side the fiberthe blue things is the fiber; you have connected that on back

side, light is connected to that. We have shown it small, because the fiber is just

about 9 to 10 microns uncladded but when we clad them it may be 120 microns. What happens

to the light, which is coming here on the back side? The junction now is here; this

is a p-type and this we marked it as new layer. What is new layer? New layer is n layer, very

lightly doped. In the previous lecture, we wrote it as intrinsic deliberately, because

we did not want to confuse the issue at that time. It is actually very lightly doped region.

If you have very lightly doped region, what happens? This is P plus; this is N layer.

The depletion layer will be occupying completely this layer. Take this N layer doping, something

like 10 to the power of 17; does not matter and take this as 10 to the power of 19 or

so; this is very lightly doped. If we reverse bias that, make this region negative and this

region positive. Here, we have got from bottom side indium phosphide and on the top side,

it is the p layer, we have P plus and a new layer and the new layer is very lightly doped.

That layer can be a micron. Now, what happens when a wave length comes into this? Let us


In indium phosphide, what is the bandgap? It is 1.35 electrons volt. What is the cut

off wavelength? Lambda corresponding to that is 1.24 divided by 1.35 micrometer, which

is actually equal to 0.92 micrometer. What does it mean? If we have lambda versus absorption,

up to the wavelength of 0.92, energy is greater than thatsmaller wave length larger energy.

Therefore, it will go like that. At lambda equal to 0.92 microns beyond that point, if

we have light falling wavelength, it will not absorb. It is transparent for wavelength

greater than 0.92 microns. Now, this wavelength is of no interest for us. If you use indium

phosphide, on this side and if a light of whatever wavelength comes in, until the wavelength

is about 0.92, it will be absorb. It is because the energy is greater than 1.25 electron volts;

it will be absorbed by, on this layer. If it is thick layer of 200 micron, if it is

absorbed there, it is lost by recombination. If you do not care about this, in fact this

is very good, because some other light coming, it is suppressed there. So, up to 0.92 microns,

this is absorbed. Beyond that point, if a wavelength is more than that, indium phosphide

is transparent. It means, this layer is empty layer, as if the layer is absent. When the

wave length is more than 0.92 microns, the light goes through all the way and picks this

lightly doped new layer. The lightly doped new layer is depleted. Why it is depleted?

Doping very low concentration depletes layer of 1 micron very easily. The entire red colour

portion is depleted. The electric field is actually in this direction, from bottom to

top. Wavelengths, which are greater than 0.29 micron, will regenerate this. Once it hits

it, it generates hole electron pairs because that can be absorbed. This is for indium phosphide.

For indium arsenide, we saw what it was like; we will put it like this that is the cut off.

This is 1.65 microns; it is the cut off point for the indium phosphide; gallium 0.47 indium

0.53 arsenide. Now, can you tell, if we have bottom layer consisting of indium phosphide

and the top layer consisting of gallium indium arsenide; if the light comes in from bottom

up to this wavelength, it will not reach that. It is not reaching that because it absorbs

right here. However, beyond that 0.92, it will be absorbed here. You actually have a

window current, so whatever is generated here will give rise to current, though it is electric

field. So, till that point, till lambda equal to I photon for this diode, but we are sure

a detector will be till lambda equal to 0.9, up to that there is no current. We have put

it out as 1.65; up to 0.92, it is absorbed by indium phosphide, no current. Beyond that

point, it starts coming up like this; gradually goes up there steeply and becomes like that.

The I photon, in fact this will be about 70% quantum efficiency. QE is the quantum efficiency

current generated due to one photon, hundred photons are generated in that layer. Some

of them escape, because they are just generated near the edge, so you get about 70%. That

is, 70 hole electron pairs are created and all of them gives rise to current.

This is actually a window. If we have a window, right from our 0.92 to 1.6 microns, where

you will get the currentit will be sensitive to that. This is very powerful technique for

using device or used in fiber optic application, where you need some air filling that region.

If you want 1.3 or 1.5, this can be used. In fact, we are happy to say that, in the

EE department in the microelectronics laboratory, a device has been made like that with gallium

indium arsenide on indium phosphide. You need to actually do lot of passivantion work etc.,

which may come up, when we discuss in the later stage. In this, when we talk about little

bit of technology, that sought of device has been made.

Including connecting this fiber on the backside, you etch on the back side so that the fiber

is connected on that. In fact, we were trying to get one device today, but somehow were

not able to, because they handed over it to the company in Hyderabad. We just made prototype,

just few of them and handed them over. They were very happy. The signal output that you

get there, the current through the diode that will be quiet large, but still if we want

amplification, we can put amplifier. Therefore, what people like to do is, the detector and

the ftp amplifier can be put together. That is something, which we are looking into now.

The idea of putting the indium phosphide layer on the substrate is what is the benefit you

get? You do not have to worry about this particular junction type, because you choose this about

1 micron; when it comes from this end right up to this, almost everything is absorbed.

If it is not absorbed, little bit loss will be there. That is a small loss you will get.

That is the key; you do not have to worry about the thickness of the junction there.

This presence of indium phosphide layer, automatically decides, the moment the wavelength is more

than 0.92, it is available for hole electron pairs. This is very powerful technique.

This is all we have said today. We have written those things about the bandgap, lambda 0.92

micron; photons with energy greater than 0.75, which will be absorbed by the high layer;

that is lambda in this region, below that it will be absorbed, we will get that thing.

Without the slide, I have just mentioned that with the help of the board there.

This is the plot, where lambda micron 0.9 and that is about 70% quantum efficiency.

The quantum efficiency definition is put there. Photon current divided by q; I photoncurrent

that you get on the external circuit divided by q gives you number of hole electron pairs.

Please note it is pair. Because the hole electrons are separated. Therefore, it is only one particle,

which gives rise to current in any cross section. So, that is why you have got, the current

divided by q gives number of pairs, which are made availabledivided by the optical

power divided by h mu, that gives you number of photons. Total power divided by total energy

or power divided by the energy of one particle gives you number of photons. Therefore, that

is the quantum efficiency of a detector. In a LED, it will be other way. You talk of how

much current flows and how many photons are emitted. That is called as quantum efficiency.

So quantum efficiency means, whether you talk of the emission or deduction.

This is the summary of very popular detector in fiber optic communication, because you

are looking for detectors in that particular wavelength. The window, where it is about

1.3 microns wavelength, it acts as an excellent window and absorbs the photons in the range

0.92 to 1.6 microns. That is about that particular compound which we have been discussing. We

will go back to this original slide here.

This is the one which you know, when there was lot of work, picking up on compound semiconductor,

whichever conference you go, this slide was projected and at least 5 minutes was spent

on that. We have spent actually more than 5 minutes because we want to get it a greater

look into the thing rather than splashing it and taking it off. We have first set this

material yesterday gallium arsenide phosphide for light emitting diodes. We have seen this

portion, where you have gallium indium arsenide, lattice match to indium phosphide. Therefore,

we can grow gallium indium arsenide by epitaxial technique arranging atoms of gallium indium

arsenide on indium phosphide, it is perfectly matched. You can make the detector. Such wafers

are available commercially. If you want, you can buy. p plus made up of gallium indium

arsenide on indium phosphide, we can buy that. In fact, when we made the device, we bought

that and etched and made all contacts etc. Now, the other one that is very popular, the

first of those compounds, which we were looking into, was gallium arsenide lattice match to

aluminum arsenide. You can see, gallium arsenide is direct bandgap material. If you keep on

adding aluminum at that point onwards, from that arrow point, you see the dotted line,

saying that it is indirect bandgap. Therefore, you have got direct bandgap switching over

to indirect bandgap beyond certain point. The nice thing is, any composition of aluminum

gallium arsenide is lattice matched with gallium arsenide because the worst case is gallium

is zero that is matched. Therefore, any other combination is matching with the thing. Mismatches

are practically zero. Now, let us see that compound consisting of this and bandgap we

can vary from 1.43 right up to 0.26, but changing from direct to indirect.

The next slide, if you see is the aluminum gallium arsenideAlGaAs. This is another

term which we can use, because it takes lot of time saying aluminum gallium arsenide.

We will say AlGaAs, do not stop with the alga, it is AlGaAs. Now, the red colour or if you

take gallium arsenide, it has a direct bandgap at 1.43 electrons volt. It has the indirect

bandgap at 1.8 electrons volt. We hope you are able to recall that if you see the energy band diagram.

putting those to all just to recapitulate your memory, electron volt one point eight

electron, that is the energy versus momentum. That is the conduction band and valence band.

So, that is what we projected there1.43, where the red colour is the direct bandgap

and the green is 1.8. If you take gallium arsenide, that is x is equal to 0 gives gallium

arsenide; x equal to 1 gives you aluminum arsenide. You have got 3.02 electrons volt

there. This

is gallium arsenide. Aluminum arsenide has got a direct bandgap of 3.02 electrons volt.

It has also an indirect bandgap; we

just did not put it down there, which is actually equal to 2.168 electron volt. This point here

shows 2.168 electrons volt. As we pointed out yesterday, when you vary, where the change,

what we are looking for now is, aluminum gallium(1 minus x) AsAlGaAs. This quantity is actually

x equal to 0; this quantity is, x equals to 1. When you vary x is equal to 0 to 1, this

which is 1.43 keeps on increasing to 3.02 that is what is plotted there. The indirect

bandgap, which is 1.8 electron volts, keeps on changing, coming to 2.168 electron volts.

This is linearly varying. At x is equal to 0.3, both of them become equal. That is, from

x is equal to 0 that is from gallium arsenide to gallium aluminum arsenide, with x is equal

to 0.3, up to that point, that is indirect; this is direct. The material in this region,

this is smaller than that, so that is direct band here. Here, this is smaller than this,

which is indirect band here. So, if you grow, as you keep on increasing x, this becomes

wider and this also becomes wider. they finally become almost equal. So, up to this point

on red, that is a direct band here is smaller than the indirect band here, which means,

it is a direct bandgap material here, till x is equal to 0.3.

So, if we say, GaAs, Eg is 1.43 electron volt, direct bandgap. X is equal to 0.3, which is

actually aluminum 0.3 mole factions, gallium 0.7, arsenideAlGaAs, for this Eg is equal

to 1.93 electron volts. This is direct. Up to that point 1.43 to 1.93 electron volt,

you get a direct bandgap semiconductor. If you want to choose, you can choose for opto-electronic

applications. Beyond, that point, if you increase the aluminum mole faction, more than 0.3,

the bandgap will go up from 1.93 right up to 2.168 electron volt. It will go right up

to that point. The whole thing is green colour here. Green is indirect; red is direct. So,

this will become an indirect semiconductor. This is actually a very popular semiconductor,

which has been tried out for years together. Professor Herbert Kroemer got Nobel Prize

related to this hetero-junction etc. This is just to illustrate, you can get the bandgap

varying. You can engineer the bandgap. You can make from direct to indirect. You can

do all those things.

What are the different applications? Why are we looking at this aluminum gallium arsenide?

If you can vary the bandgap, what is the consequence of that? Where is the more powerful application?

One of the application is of course is in opto-electrical application. For a direct

bandgap up to x equal to 0.3, you can use it for opto-electronic application. Mostly

this material is used for x is equal 0.3, because beyond x is equal to 0.3, the aluminum

concentration becomes more and more and the material becomes unstable. We are not touching

that point now. Up to the 0.3, if people are using it, it is a manageable material. You

can not only do bandgap engineering, but also manage the material. Growing and see the whether

the material is stable or not.

Therefore, up to 0.3 is direct bandgapvery stable material and a perfect lattice match

on to gallium arsenide. The vertical line you saw. It means, you can grow aluminum gallium

arsenide layers on gallium arsenide. That means, you can make hetero-junctions using

AlGaAs on GaAs. If you remember, in indium phosphide, if you want to make hetero-junction,

what material you will use? Gallium indium arsenide, because the lattice match of indium

phosphide is with gallium indium arsenide. That is where the hetero-junction is made.

One application you saw. Here you can use AlGaAs. The nice thing about this is, the

bandgap of AlGaAs is more than that of gallium arsenide 1.43 there. That is right up to 1.93.

Therefore, you have no worry about leakage current. Bandgap is larger in the case of

AlGaAs. So, with a perfect lattice match these hetero-structures

are very popular for different applications. One of them is for laser. In laser, it is

a p-n junction again. You make a p-type material with aluminum gallium arsenide and n-type

region with gallium arsenide, like this. Not exactly n-type material with gallium arsenide,

it is like this. It is a very popular structure with laser. is also diode only. The diode

is so powerful that it will haunt everybody, in microelectronic, optoelectronic, and everywhere.

So, you have got, this material you can use as gallium arsenide p-type; this material

you can use as aluminum gallium arsenide within x is equal to 0.3 p-types; this material you

can use as aluminum gallium arsenide n-type. Now, what happens is, this material is a wider

bandgap material, this material is a wider bandgap material, and this is like this. This

is Ec and Ev. We are just giving you an idea of this. Here, there is a notch, if you forward

bias this device, pump in current, then of course it will bend. You will have the electrons

injected from here; they will be trapped there. They can be forced on to recombine. Therefore,

you have got a notch, in which it will capture those electrons and those electron injected

from n-region to p-region, forward bias p-n junction will inject electrons. That will

be trapped here. It will not go to the p-region; it will be available. Once it is available

there, it will lose its energy to this side there and you will get the laser wavelength

corresponding to gallium arsenide. That is 0.87. This is one of the applications hetero

junction, aluminum gallium arsenide, and gallium arsenide. Another application on which we

are going to spend lot of time is high electron mobility transistor.

The high electron mobility transistor (HEMT) - we are just showing the application that

is all. This is p gallium arsenide, on the top of that you will have AlGaAs. You will

have contact made here; please do not worry, if you do not understand the whole thing,

because we will have lot of discussions on this. Source drain, this is aluminum gallium

arsenide and you will have a gate region put here, which will actually create a channel

below this. In fact, you will have n plus. What we are trying to point out is, in a high

electron mobility transistor, aluminum gallium arsenide there, gallium arsenide here, the

aluminum gallium arsenide will be heavily doped and gallium arsenide will be very lightly

doped. In the case of MOSFET, if you recall, this is a oxide layer and this region will

be inverted. In this case, this whole region will be depleted, because it is made very

thin, and instead of oxide, you will have this layer.

This n plus layer will donate electrons into this region. So, the electrons are here, when

you invert, when you supply this, into this region. Those electrons are in the very lightly

doped region. In a MOSFET, if you recall, it will go to smaller and smaller scaling,

doping in that region goes up and mobility goes down. Here, you do not touch the doping

on this layer. Increase the doping here which supplies electrons on to that. So, the electrons

are in a region, which are not scattered by the doping very lightly doped. What you get

is, the electrons in this region do not suffer from ionised impurity scattering, no doubt,

they will experience lattice scattering.

If they experience lattice scattering, what is the consequences? Mobility versus temperature

was like that, if you remember. But, we cut this out, because this is and this is, lattice

scattering. We will have only one curve; this is removed. So, we can go to lower temperatures.

We can go to 100,000 centimeter square per volt second mobility. That is a powerful technique.

The other application which we can mention is a hetero-junction bipolar transistor. If

you make a n-p-n transistor; the n-region made up of aluminum gallium arsenide and p-region

made up of gallium arsenide; and again collector made up to gallium arsenide or AlGaAs, whatever

it is, it does not matter. This is HEMThigh electron mobility transistor.

We will have at least few lectures discussing on this. This is a very important device for

high speed operations. We will also see its circuit applications.

HBThetero-junction bipolar transistor n-type aluminum gallium arsenide; p-type gallium

arsenide; n-type either gallium arsenide or aluminum gallium arsenide. This is emitter,

base, and collector. This junction is hetero-junction. In fact, what we have done is, we have replaced

the emitter instead of making that gallium arsenide. We have made a wider bandgap material

AlGaAs. We are just giving you the result; the consequence of that isthe beta of

the transistor will be very high and it will allow you to reduce the doping here. If you

get high beta, you do not have to make this n plus. It will allow you to reduce the doping.

If you reduce the doping, depletion layer will be wider and the device will become faster.

I will discuss this later in full detail.

To sum up today, compound semiconductor based on gallium arsenide and aluminum arsenide,

indium arsenide and indium phosphide etc., have opened up several avenues and devices

for microelectronics and optoelectronics. With that, we will stop today. We will take

a closer look into the crystal structure etc., for these materials in the next lecture.

The Description of Lecture5-Temary Compound Semiconductor and their Appl - 2