Laser: Fundamentals and Applications

Prof. Manabendra Chandra

Department of Chemistry
Indian Institute of Technology, Kanpur

Lecture – 06
Population inversion, 2-level system and 3-level system

Hello and welcome. Today is the first day of this week, and I hope so far you found these
initial lessons useful. So, what we were looking at the last class was the conditions to have
laser actions. And what we did? We used the Principle of Detailed Balance to equate the
total rate of absorption and total rate of emission. Thereby we boiled down to the condition
where Einstein’s formalism matches with the Planck’s one, and from there we got two
useful relations. We found, the rate of stimulated emission is equal to the rate of absorption
and we also saw why it is difficult to achieve a laser at higher frequencies. We ultimately
came to the conditions at which we can or we cannot get a laser action.

(Refer Slide Time: 01:38)

N2
So, what we found out that, if >1, only then I can have a condition where stimulated
N1

emission dominates, and I can achieve laser action.

Suppose for time being let us assume that I can achieve this condition. So, at this condition
what do I see? What I will see can be shown graphically, if I plot the intensity of light as
 N
a function of frequency. So, when I have the condition, which is equal to N2 =1, I have no
1

net gain. So, whatever be the scale of intensity, that is flat.

Now, our concern is at a particular frequency. So, let me is put this as 12 as we have been
N
using. So, here I do not have any kind of amplification. Now the condition when I have N2
1

>1, then what I will have? I am putting some number of photons and I am getting number
of photons which is much higher. So, when I give x input and get n number of times of x
as output, then I say I have gain. Here also I am giving photon and I am not getting anything
N
out of it when I have N2 =1. So, my income and expenditure both are same. No gain. So,
1

N
if I come to this particular condition, N2 >1, then what I will say? I have much more photon
1

than I have already used. So, therefore, at that particular frequency what I am using (12),
I will see a net gain. So, this is not any other type of gain, but it is called optical gain at
this particular frequency, 12.

N
Now, mathematically we got it. But in reality, how we can get this condition N2 >1?
1

If I can get that condition, I can make a laser, not much trouble! So, now, what is preventing
me to have that condition? So, if I can do it mathematically, why cannot I actually do it?
Simply, if I look at the equilibrium population distribution, and if we again look at the
−h
N
Boltzmann statistics, then what do I have, this is equal to this, N2 = 𝑒 𝑘𝑇 , right?
1

(Refer Slide Time: 05:27)

So, normally what I can see that, N2 is always very very very less than N1, but I want N2
be much higher than N1. N2 should be greater than N1 in order to have laser action.

So, suppose I have again 2-state models, and I have several atoms. Suppose I have taken
total 8 atoms. Now in the step one, I come with some photons. So, there are many photons
I am sending in. Two of these atoms are going from 1 to 2. So, they make a transition here.
So, I essentially have two atoms at state 2, and the same ones are non-existing in state 1.
So, after sending these photons, I get this condition of population distribution.

Now, if I can achieve a condition where half of the atoms are in this state 1 and half of the
atoms are in the excited state i.e. state 2, then that will be something like this. So, I have 4
atoms in state 1, 4 in state 2. This is a saturated condition. Initially all the atoms were in
ground state and very few in the excited state. That was my equilibrium condition and that
was governed by Boltzmann distribution. Then by putting light, more and more light, I
somehow achieved a condition which is like this where I have equal population in both the
states. When I get this condition, this is known as saturation.

Now, in the equilibrium condition, I had, N1>>N2. Then, after sending photons I got still
N1>N2. At the saturation condition, I got, N1=N2. But what I want? I want N1 to be less
than N2, so N1<N2, for laser action, right? So, I want to actually invert the population! I
want to invert the populations so that now my state 2 contains more number of atoms than
state 1. And this is known as population inversion. So, this condition N2 greater than N1 is
essentially a population inversion condition. Ok?

So how can I get this population inversion? Now here we will describe in detail that
whether or not we can find population inversion in this system, which is essentially a 2-
level system. So, can we? So, the question is, can we get population inversion in a 2-level
system? We will try to answer that. Obviously here, we will take the same 2-level model,
their rates and all the details for answering this question.
(Refer Slide Time: 10:42)

By the way, I think I have missed out one thing here. We were talking about these A21,
B12, B21 - these rate constants. They are known as Einstein’s coefficients. I do not
remember if I have mentioned this one earlier, but this A and Bs, they are known as
Einstein coefficients. Many time you will come across these in the books where these
coefficients will be written as Einstein’s A coefficient, Einstein’s B coefficient, and so on.
So, you should definitely know this one. If I have not mentioned it earlier, you know now.
All right. So, let us go back to our discussion.

(Refer Slide Time: 11:43)

So, we are still considering this 2-level system. Let me call the states simply as 1 and 2
without bothering about the bracket notation. And, N1, N2 are their populations. Now in
order to answer the question that can we have a population-inversion in 2-level system, we
will deal with the population of state 2, which is N2. We have to see how we can make N2
greater than N1. That is our aim.

So, let us see what is the change in population of state 2. So, from the basic knowledge of
kinetics we can easily get that. So, how do I write that? I write, dN2/dt. Now this N2 is time
dependent. When I am shining light, then as a function of time, N2 changes. How it is
changing? So, how this state 2 is populated? It is populated by absorption process. Correct?
And, how it is depopulated? By the emission process. Correct?

So, the absorption processes will come with positive sign, and the emission processes,
emission rates will come with negative sign. So, if I write this one, what do I have?

dN2
= B12N1- A21N2-B21N2 (Refer Slide Time: 11:43)
dt

So, first absorption, B12N1, (I am not writing that (h12) anymore for simplicity) minus
spontaneous emission A21N2 minus stimulated emission B21N2. So, from previous
discussion we already know these rates. We also know that the absorption rate constant is
equal to the rate constant of stimulated emission. So, I have this condition B12 equals to
B21. If you remember this from the previous class, you have already seen this. So, if I take
this B12 equals to B21 equal to be something as B, and suppose I write A21 to be just A.
Then what I get? I get,

dN2
= BN1- AN2-BN2 (Refer Slide Time: 11:43)
dt

dN2
So, what I have is, = B(N1-N2)- AN2 (Refer Slide Time: 11:43)
dt

Now, this is a differential equation and we need to solve it. So, what are the conditions
that we have in our hand so that I can solve this one? At time t = 0, N2 = 0. We all agreed.
Right? Time t=0 means when there was no photon. We started giving the photon at time t.
When there was no photon, there was no population of the state 2. So, N2=0, and if there
were total N number of molecules, and if I call it like Ntotal, then essentially at that time
t=0, my N1 equals to Ntotal. So, what I have, I can get this expression.
 𝐵N𝑇𝑜𝑡𝑎𝑙
N2 = A+2B

Now, what do I have for N2? Here is A. A is the rate constant for spontaneous emission,
right? So, Spontaneous emission is always there. It does not require any photon or
anything. In vacuum if you keep it, it will emit because there is a lifetime of the state. It
will come back to the ground state by emitting photons and that is what is spontaneous
emission. So, no matter what, spontaneous emission is always there. So, the rate constant
A is nonzero. So, therefore, I can say, A > 0 always. This condition is always there. So,
now I have here this BNTotal divided by this factor (A + 2B), where A > 0. Let me write
it over here.

(Refer Slide Time: 18:40)

So, from the above expression I get,

N2 𝐵
=
N𝑇𝑜𝑡𝑎𝑙 A + 2B

Now, when A is very very less, it comes to something like

N2 N2 𝐵 1 1
𝑜𝑟 = 2B = or < 2
N𝑇𝑜𝑡𝑎𝑙 N1 +N2 2
When A is very small, the value will be around half. The maximum value is half here. If
A is not too small, the value will be less than half.

Now does it tell you something? We drew this 2 level system and we started exciting it
and we got to a condition, when we had a saturated condition, right? So, we had a condition
where we had 4 atoms in state 1 and 4 atoms in state 2 and we called the condition as
saturated. So, here if you have equal population of two states. Essentially N2 by total
number of atoms is half that is maximum possible. I cannot get more than that. So,
essentially when N2 divided by Ntotal is equal to half or N2 divided by N1+N2 is equal to
half, that means, N1 actually equals to N2, which is essentially this condition.

N2
So, because A is always positive; essentially what I have is N is always less than half.
𝑇𝑜𝑡𝑎𝑙

So, forget about going to N2 > N1. I have N2<< N1 all the time. So, N2 > N1 , this is not
possible for a 2-level system. In other word, population inversion is not possible for a 2-
level system. This is an important relation to remember.

So, I know that if I have to get a laser I must fulfill this condition N2 > N1 or I must have
a population inversion. And, after seeing everything I figured out now that I cannot achieve
a population inversion in a 2-level system. So, I have answered the question that I have
asked few minutes back. Now then question is, if not 2-level, whether a 3-level system can
give rise to population inversion? Let us look at it.

(Refer Slide Time: 23:25)

So, question is, 2-level system cannot. Now is a 3-level system capable of having
population inversion? So, if I draw three level system, just for clarity I am drawing in this
fashion, this has nothing to do the actual energy level picture. So, I have state 1, I have
state 2, and I have state 3 (Refer Slide Time: 23:25).

Now, what are the possibilities that we have here? We can have absorption as well as
emission from these two states 1 and 2. I can have spontaneous emission, I can have
stimulated emission, and if I just consider these 2 states without considering state 3, it is
fairly a 2-level system. So, I know everything about it what I should know. I have already
learnt it. Now between these 2 levels what I can have? I am going sequentially. So, I can
hit this system with light, then the atom goes to state 2. So, at that condition either it can
come down here or it can come down state 3, all right? So, the single headed arrow means
there is a spontaneous emission and from here, from state 3, I can have spontaneous
emission. I can also have stimulated emission.

Now, whether the stimulated emission will dominate over spontaneous emission will tell
me whether I can have a laser using three level system. So, let me answer you first that
whether this three level system can act as a laser, can it give population inversion? Answer
is yes, it can. Under certain conditions. So, what are those conditions? That is this level 3
which is lying below level 2 must have a lifetime. So, if I write 3 as the lifetime of state
3, then 3 should be much greater than 2. So, in other word the lifetime of the state 2
should be much shorter than that of state 3. This kind of state, this state 3, which is long-
lived, is known as metastable state. So, my state 3 is a metastable state. And also this state
3 should not be a dark state. That means, it should be able to emit radiatively. So, at this
condition, I can actually achieve population inversion, how? That we will discuss in the
following class, till then goodbye.

Thank you very much.