Lecture 8

Download as pdf or txt
Download as pdf or txt
You are on page 1of 19

Fall 2024

ECE 455

Black Body Radiation


Einstein’s View and Stimulated Emission
Line Shape
Let’s Start With Some Basics

• Materials can both absorb and emit light


• There are lots of ways to have this
happen
• We’ll get to those a bit later

• Absorbing light, for example, puts the


system into a state of higher energy
• “Excited state”
• These states can be ‘discrete’ or
continuum-like
• We will use the strength of the absorption, competition between two types of emission,
and the shape of the distributions of these states (i.e., the lineshape) to understand gain
• But first, let’s try to understand the origins of modern laser analysis
Rayleigh-Jeans Emission

• We all know that when you heat an object


(sufficiently, of course), it can start to
glow…starting from ‘red hot’ and marching
towards ‘white hot’

• How do we go about deriving the mathematics


describing this process? If we can do this, maybe
we can harness it…
Rayleigh-Jeans Emission

So, how did people approach this problem?

• First, the assumption was to select a rectangular resonator with a small hole in it
• A 3-D box is selected since you can define a very broad range of associated
modes

• We are looking to find the power spectral ρ(ν) dν


density ρ(ν)

• This has units of Watts per Hertz

• This describes the probability of a photon


of frequency ν can be produced (note the
z
distinction between photon energy hν)
Rayleigh-Jeans Emission

• So, let’s find the frequencies associated with this box

• First, we can define an arbitrary plane wave inside the box that has
a k-vector associated with it
kx2 ky2 kz2

• Where does this come from? It’s simply the FSR


applied to all 3 dimensions, for example
Note we can have both TE
and TM modes so a factor
of 2 will show up later
Rayleigh-Jeans Emission

• If we assume a = b = d, and substitute ν = ω/2π we obtain

This is the equation of a sphere with radius R

• Only those modes within the sphere are allowed


- the volume of the sphere is 4πR3/3
• As the mode spacing gets smaller (wavelength is much
smaller than the resonator), we just
map out the sphere, so the total number of
modes is the volume of the sphere with “adjustments”
m, p, q > 0 (1/8 of a sphere)
TE+TM
Rayleigh-Jeans Emission

• The power spectral density per unit volume is just the derivative of this with
respect to ν, divided by the volume of the material (a3), and so we obtain
Using the chain rule
If dn/dν is approximately zero, if not n3 → n2ng where ng = n + ν(dn/dν)

• This is a number density (number per unit volume per hertz) of states or
modes
• If we assume each mode is driven by the same thermal energy, each mode
has energy kT giving

• This can’t be right since it blows up to infinity with increasing ν (the


“ultraviolet catastrophe”)
Planck’s Correction

• Rather than giving each mode an energy of kT, maybe we are


dealing with photons with energy hν
• The probability of having those photons follows Boltzmann
statistics where the expected energy at a frequency
is the number of photons times that photon energy
where n is the number of photons produced at frequency ν
• We can add up the total probability of having n = 1, 2, 3, 4, etc.
photons at frequency ν
Planck’s Correction and Black Body Radiation

Verdeyen brings the ng back here


Replace kT
• And we obtain this expression With This

• Giving the power per unit volume per unit Hz that is produced by
the system
• Total energy is the integral of this equation over dν and volume
• Since the function is bound (goes to zero at small and large ν)
the power no longer blows up to infinity
• This expression was validated experimentally and will serve as the
basis for the introduction of the concept of stimulated emission
Einstein’s View

• Rather than assuming a resonator full of modes, let’s instead assume that we have two energy levels
associated with an emitter (again this can be a lot more complicated, and it usually is!)

• There are two processes easy to understand: absorption and spontaneous emission
• First, we can absorb (b) a photon of energy hν to excite the system to a higher energy
• It can also naturally decay (exponentially) via spontaneous emission (a) of a photon
• Einstein further said: let there be a third process by which one photon can lead to the emission of
another via stimulated emission (c)
Einstein’s View
N2 = Number per unit volume in excited state
N1 + N2 is a constant = “Number
per volume of possible emitters”
N1 = Number per unit volume in ground state

• Let’s further take these process to be time dependent, and are governed by the
following differential equations and coefficients A & B
Spontaneous Stimulated
Absorption (b)
Emission (a) Emission (c)

ρ(ν) is the
intensity of light
• And that they all take place at the same time present in units
of Watts/Hz/m3
Einstein’s View

• Now, let’s assume that we are in thermal equilibrium, such that the
time derivatives are zero (we are in the “steady state”)
Rewrite
0

• Next let’s make the assumption that we still follow Boltzmann


statistics, and introduce the concept of a degeneracy, or the
number of ‘configurations’ where energy hν can be achieved
Solve for ρ(ν)
Einstein’s View

• Comparing

• Which are equivalent iff


Lineshape Function and Cross-Sections

• The Einstein coefficients are not practical to use


• We would prefer to use a different set of coefficients or parameters to describe
the energy distribution so that we can use units of power (Watts) or intensity
(W/m2), rather than Watts/Hz/volume
• Let us first introduce the concept of the lineshape
• You can think of this as moving away from a delta-like energy level diagram to
one where there somehow is a distribution g(ν)
• We’ll talk more later about where these come from
Lineshape Function and Cross Sections

• ρ(ν) is power spectral density (W/Hz) per unit volume

• If we know we are propagating in a cavity along the optical axis, we can


convert the volume into an area by multiplying by a Length L (= vgxΔt)

• We can turn this into a flow (flux, really) if we know that we traversed a
length L in a time Δt

This is true for all flows, such as


Watts/m2 Per Unit Time
Watts/Hz/m3 c/ng fluids in pipes, rivers, EM waves, etc.
Lineshape Function and Cross Sections

• Let’s continue and change our rate equation for the two-level system

• By substituting , , and

• And rearranging to obtain

We will lump these together into an “cross-section” σ(ν) which has units of m2
Lineshape Function and Cross Sections

• Caution must be exercised here


• Absorption is relative to N1 population, and stimulated emission to
N2 population
• In many laser systems, the “degeneracies” are not easy to identify,
and the emission and absorption cross-sections differ. In that case
and more generally

• And we can say that This has units of m-1 and will
become a gain coefficient γ
• So, “g2/g1” is also a function of ν in many systems
Gain = eγz
Example: Erbium Doped Fiber

Er3+

emission
NEXT TIME

MORE ON GAIN

You might also like