Optical Resonator

CHAPTER 1---- Optical Resonator
Chapter 1
Optical Resonator
Fundamentals of Photonics 24/3/11 1

What is an optical resonator?

An optical resonator, the optical counterpart of an electronic resonant
circuit, confines and stores light at certain resonance frequencies. It may be
viewed as an optical transmission system incorporating feedback; light
circulates or is repeatedly reflected within the system, without escaping.
(a) (b)
(c) (d)

Contents:
• 1.1 Brief review of matrix optics
• 1.2 Planar Mirror Resonators
– Resonator Modes
– The Resonator as a Spectrum Analyzer
– Two- and Three-Dimensional Resonators
• 1.3 Spherical-Mirror Resonators
– Ray confinement
• 1.4 Gaussian waves and its characteristics
– The Gaussian beam
– Transmission through optical components
– Gaussian Modes
– Resonance Frequencies
– Hermite-Gaussian Modes
– Finite Apertures and Diffraction Loss

1.1 Brief review of Matrix optics

Light propagation in a optical system, can use a matrix M,
whose elements are A, B, C, D, characterizes the optical system
Completely ( known as the ray-transfer matrix.)
matrix to describe the
rays transmission in the optical components.
One can use two

y2  Ay1  B1
 2  Cy1  D1
parameters:
y: the high
: the angle above z axis
y2  Ay1  B1  y2   A B   y1 
   C D   
 2  Cy1  D1  2   1

y2  y1  d  tg1
2  2  1
1
y2 For the paraxial rays tg  
y1
 y2  1 d   y1 
   0 1   
 2   1
y2,2 y2  y1 y1
2 y1,1 2 
1  2  y1  1 R
 R
Along z upward angle is positive,
and downward is negative

Free-Space Propagation Refraction at a Planar Boundary
1 0
1 d 
M  M  n1 
 0
0 1   n2 
Refraction at a Spherical Boundary Transmission Through a Thin Lens
 1 0  1 0
M   (n2  n1 ) n1  M  1 
  1
 n2 R n2   f 
Reflection from a Planar Mirror Reflection from a Spherical Mirror
 1 0
1 0  M  2 
M    1
0 1   R 

A Set of Parallel Transparent Plates
 di 
M  1 n 
 i

0 1 
Matrices of Cascaded Optical Components
M  M 1M 2 ....M N

Periodic Optical Systems

The reflection of light between two parallel mirrors forming an optical
resonator is a periodic optical system is a cascade of identical unit system.
Difference Equation for the Ray Position
A periodic system is composed of a cascade of identical unit systems

(stages), each with a ray-transfer matrix (A, B, C, D). A ray enters the system
with initial position y, and slope 8,. To determine the position and slope (y,,,,
0,) of the ray at the exit of the mth stage, we apply the ABCD matrix m times,
 ym   A
m
B   y0  ym 1  Aym  B m
   C D   0   m 1  Cym  D m
 m 

ym 1  Aym  B m
 m 1  Cym  D m
From these equation, we have
ym 1  Aym
m 
B
y  Aym 1
 m 1  m  2
B
And then:
ym  2  2bym 1  F 2 ym linear differential equations
A D 2
where b , F  AD  BC  det  M 
2

If we assumed: y m  y0 h m
So that, we have h 2  2bh  F 2  0 h  b  i F 2  b2
If we defined  
  cos 1 b F
We have b  F cos  , F  b  F sin 

2 2
then h  F (cos   i sin  )  Fe  i ym  y0 F m e  im
A general solution may be constructed from the two solutions with positive
and negative signs by forming their linear combination. The sum of the two
exponential functions can always be written as a harmonic (circular) function
ym  y0 F m sin(m  0 )  ymax F m sin(m  0 )

If F=1, then ym  ymax sin( m  0 )
Condition for a Harmonic Trajectory: if ym be harmonic, the cos-1b must

be real, We have condition
A D
b 1 or
2
1
The bound b  1 therefore provides a condition of stability (boundedness) of

the ray trajectory
If, instead, |b| > 1,  is then imaginary and the solution is a

hyperbolic function (cosh or sinh), which increases without bound.
A harmonic solution ensures that y, is bounded for all m, with a
maximum value of ymax. The bound |b|< 1 therefore provides a
condition of stability (boundedness) of the ray trajectory.

Condition for a Periodic Trajectory
Unstable b>1
Stable and periodic
Stable nonperiodic
The harmonic function is periodic in m, if it is possible to find an integer

s such that ym+s = ym, for all m. The smallest such integer is the period.
The necessary and sufficient condition for a periodic trajectory is:

s = 2q, where q is an integer

EXERCISE : A Periodic Set of Pairs of Different Lenses. Examine the trajectories of

paraxial rays through a periodic system composed of a set of lenses with alternating focal
lengths f1 and f2 as shown in Fig. Show that the ray trajectory is bounded (stable) if
d d
0  (1  )(1  ) 1
2 f1 2 f2
 d d2 
 1 0  1 0  1  2d  
 1 d   1 d    f1 f1
M  1 
 
 1 
 1  0 1    1  0 1   d 1 1 d d d 
 f 2   f1       (1  )(1  ) 
 f1 f 2 f1 f 2 f2 f1 f2 

Home works
4 X 4 Ray-Transfer Matrix for Skewed Rays. Matrix methods may be
generalized to describe skewed paraxial rays in circularly symmetric
systems, and to astigmatic (non-circularly symmetric) systems. A ray
crossing the plane z = 0 is generally characterized by four variables-the
coordinates (x, y) of its position in the plane, and the angles (e,, ey) that its
projections in the x-z and y-z planes make with the z axis. The emerging ray
is also characterized by four variables linearly related to the initial four
variables. The optical system may then be characterized completely, within
the paraxial approximation, by a 4 X 4 matrix.
(a) Determine the 4 x 4 ray-transfer matrix of a distance

d in free space.
(b) Determine the 4 X 4 ray-transfer matrix of a thin
cylindrical lens with focal length f oriented in the y
direction. The cylindrical lens has focal length f for rays
in the y-z plane, and no focusing power for rays in the x-
z plane.

1.2 Planar Mirror Resonators
This simple one-dimensional

resonator is known as a
Fabry-Perot etalon.
A. Resonator Modes
Resonator Modes as Standing Waves
A monochromatic wave of frequency v has a wavefunction as


u (r , t )  Re{U (r ) exp( j 2 vt )}
Represents the transverse component of electric field.
The complex amplitude U(r) satisfies the Helmholtz equation;
Where k =2v/c called wavenumber, c speed of light in the medium

the modes of a resonator must be the solution of Helmholtz equation

with the boundary conditions:
 z  0
U (r )  0 
z  d
So that the general solution is standing wave:


U (r )  A sin kz
d
With boundary condition, we have kd  q
c
F 
2d
q c
kq  q  q , q  1, 2,...,

d 2d
 q  q 1
Resonance frequencies c
F 
2d

c
The resonance wavelength is: q   2d
q q
The length of the resonator, d = q q /2, is an integer number of half

wavelength
Attention: c  c0 / n Where n is the refractive index in the resonator
Resonator Modes as Traveling Waves

A mode of the resonator: is a self-reproducing wave, i.e., a wave that reproduces
itself after a single round trip , The phase shift imparted by a single round trip of
propagation (a distance 2d) must therefore be a multiple of 2.
4 n 4
  k 2d  d d  q 2 q= 1,2,3,…
0 c

Density of Modes (1D)
The density of modes M(v), which is the number of

modes per unit frequency per unit length of the resonator, is
4
M ( )  For 1D resonator
c
The number of modes in a resonator of length d within the frequency

interval v is:
4
d 
c
This represents the number of degrees of freedom for the optical waves
existing in the resonator, i.e., the number of independent ways in which
these waves may be arranged.

Losses and Resonance Spectral Width

The magnitude ratio of two consecutive phasors is the round-trip
amplitude attenuation factor r introduced by the two mirror reflections and
by absorption in the medium. Thus: Mirror 1 Mirror 2
4 nd U3
i
U1  hU 0   e U 0   e  i 
U 0   e  i 2 kdU 0
U2
So that, the sum of the sequential reflective light with field of U1
U U0
U  U 0  U1  U 2  U 3  ...  U 0 (1  h  h 2  h3  ...)  0
(1  h)
2
2 U0 I0 I0
IU   
 i 2 (1    2 cos  )
 
(1   ) 2  4 sin 2  
2
1  e
 2 
I max I0
finally, we have I , I max 
1  (2F /  ) 2 sin 2 ( / 2) (1   ) 2
 1/ 2
F  Finesse of the resonator
1 

The resonance spectral peak has a full width of half maximum (FWHM):
c F
   
4 d F
4 d I max I max

Due to   We have I  I 
c 1  (2F /  ) 2 sin 2 ( /  F ) min 1  (2F /  ) 2
where  F  c 2d    q  q F , q  1, 2,...,
c
F 
2d
F
 
F

Spectral response of Fabry-Perot Resonator

The intensity I is a periodic function of  with period 2. The dependence of I
on , which is the spectral response of the resonator, has a similar periodic
behavior since = 4d/c is proportional to . This resonance profile:
I max
I
1  (2 F /  ) 2 sin 2 ( / F )
The maximum I = Imax, is achieved at the     q , q  1, 2,...,

q F
resonance frequencies
I max
whereas the minimum value I min 
1  (2 F /  ) 2
c F
The FWHM of the resonance peak is    
4 d F

Sources of Resonator Loss

• Absorption and scattering loss during the
round trip: exp (-2asd)
• Imperfect reflectance of the mirror: R1, R2
Defineding that
  R1 R2 exp(2 s d )
2  2  exp(2 r d )
1 1
r  s  ln
2d R1 R2
we get: 1 1
r  s  ln   s   m1   m 2
ar is an effective overall 2d R1 R2
distributed-loss coefficient,
which is used generally in the
system design and analysis 1 1 1 1
 m1  ln  m1  ln
2d R1 2d R1

If the reflectance of the mirrors is very R1  R2  R  1

high, approach to 1, so that
1  R1 1  R2 1 R
The above formula can approximate as  m1     m2 
2d 2d 2d
1 1 1 R
r  s  ln   s   m1   m 2 r  s 
2d R1 R2 d
The finesse F can be expressed as a function of the effective loss coefficient ar,
 exp( r d / 2)
F 
1  exp( r d )

Because ard<<1, so that exp(-ard)=1-ard, we have: F 
r d
The finesse is inversely proportional to the loss factor rd

Photon Lifetime of Resonator

The relationship between the resonance linewidth and the resonator loss
may be viewed as a manifestation of the time-frequency uncertainty relation.
Form the linewidth of the resonator, we have
c / 2d c
   r
 /  r d 2
Because ar is the loss per unit length, car is the 1
loss per unit time, so that we can Defining the p 
characteristic decay time as the resonator lifetime or c r
photon lifetime
The resonance line broadening is seen to be 1

governed by the decay of optical energy arising  
2 p
from resonator losses

The Quality Factor Q

The quality factor Q is often used to characterize electrical resonance circuits
and microwave resonators, for optical resonators, the Q factor may be determined
by percentage of that stored energy to the loss energy per cycle:
2 ( storedenergy )
Q Large Q factors are associated with low-loss resonators
energylosspercycle
For a resonator of loss at the rate cr (per unit time), which is equivalent to the rate
cr /0 (per cycle), so that
0
Q  2  1 
 (c r /  0 )     c 2 r Q

The quality factor is related to the resonator lifetime (photon lifetime)

  1
c r

1
Q  2 0 p
2
p
0
The quality factor is related to the finesse of the resonator by Q  F
F

• In summary, three parameters are convenient for

characterizing the losses in an optical resonator:
– the finesse F
– the loss coefficient r (cm-1),
– photon lifetime p = 1/cr, (seconds).
• In addition, the quality factor Q can also be used for this

purpose

B. The Resonator as a Spectrum Analyzer

Transmission of a plane wave across a planar-mirror resonator (Fabry-Perot etalon)
It
T ( ) 
I
t1 r1 r2 t2 Tmax
T ( ) 
1  (2F /  ) 2 sin 2 ( /  F )
Where:
2
U2 t
Tmax  , t  t1t 2 ,    1 2
(1   ) 2
U1  1/2
F 
1 
U0 The change of the length of the cavity
Mirror 1 Mirror 2 will change the resonance frequency
qc  q d
 q   2 d  
2d d

C. Two- and Three-Dimensional Resonators

• Two-Dimensional Resonators
q y q 2 2
ky  , k z  z , q y  1, 2,..., qz  1, 2,..., k 2  k y2  k z2  ( )
• Mode ddensityd c
Determine an approximate expression for the number of modes in a

two-dimensional resonator with frequencies lying between 0 and , 4
M ( ) 
assuming that 2/c >> d, i.e. d >>/2, and allowing for two c2
orthogonal polarizations per mode number.

Three-Dimensional Resonators
Physical space resonator Wave vector space
q x q y q 2 2
kx  , ky  , k z  z , qx , q y , qz  1, 2,..., k 2  k x2  k y2  k z2  ( )
d d d c
Mode density
The number of modes lying in the frequency interval
8 2
between 0 and v corresponds to the number of
M ( )  3 points lying in the volume of the positive octant of a
c sphere of radius k in the k diagram

1.2 Optical resonators and stable condition

A. Ray Confinement of spherical resonators
The rule of the sign: concave mirror (R < 0), convex (R > 0). The
planar-mirror resonator is R1 = R2=∞
The matrix-optics methods introduced which are valid only for paraxial rays, are
used to study the trajectories of rays as they travel inside the resonator

C. Stable condition of the resonator

For paraxial rays, where all angles are small, the relation between
(ym+1, m+1) and (ym, m) is linear and can be written in the matrix form
R1 R2  ym 1   A B   ym 
y1
-1    C D   
 m 1    m
A B 1 0  1 d   1 0  1 d 
z C D    2 1   0 1   R22 1   0 1 
   R1
y2
2
0 reflection reflection
y0 from a from a
mirror of mirror of
d radius R1 radius R2
Attention here: we just take general case propagation a distance

spherical so doesn’t take the sign d through free space

A  1  2d
R2
det M  Ad  BC  1  F 2
B  2d (1  d )
R2
ym  ymax sin(m  0 )
C 2  2  4d
R1 R2 R1 R2  d  d 
b  ( A  D) / 2  2 1   1    1
D  2d  ( 2d  1)( 2d  1)  R1   R2 
R1 R1 R1
It the way is harmonic, we need cos-1b must be real, that is
 d  d 
b  1 b  ( A  D) / 2  2 1   1   1  1
 R1   R2 
for g1=1+d/R1; g2=1+d/R2

 d  d 
0  1   1   1 0  g1 g 2  1
 R1   R2 

resonator is in conditionally stable, there will be:
 d  d 
0  1   1    1
 R1   R2 
0  g1 g 2  1
In summary, the confinement condition for paraxial rays in a spherical-

mirror resonator, constructed of mirrors of radii R1,R2 seperated by a
distance d, is 0≤g1g2≤1, where g1=1+d/R1 and g2=1+d/R2
For the concave R is negative, for the convex R is positive

Stable and unstable resonators

a. Planar
g2
(R1= R2=∞)
Non stable b. Symmetrical confocal
e
d (R1= R2=-d)
1 a
stable c. Symmetrical concentric
b g1 (R1= R2=-d/2)
-1 0 1
d. confocal/planar
c Non stable
(R1= -d,R2=∞)
l
e. concave/convex
na rica
et
rs
m
to
m
Sy
(R1<0,R2>0)
so
re
d/(-R) = 0, 1, and 2, corresponding to planar, confocal, and concentric resonators

The stable properties of optical resonators
Crystal state resonators a. Planar

(R1= R2=∞)
g1 g 2  0 or g1 g 2  1
b. Symmetrical confocal
(R1= R2=-d)
Stable
c. Symmetrical concentric
unstable (R1= R2=-d/2)

Unstable resonators Unstable cavity corresponds to the high loss

g1 g 2  0 or g1 g 2  1
a. Biconvex resonator
d
b. plan-convex resonator
c. Some cases in plan-concave resonator
When R2<d, unstable

R1
d. Some cases in concave-convex resonator
d
When R1<d and R1+R2=R1-|R2|>d
e. Some cases in biconcave resonator

g1 g 2  (1  d / R1 )(1  d / R2 )  0
R1  R2  d
g1 g 2  (1  d / R1 )(1  d / R2 )  1

1.2 Gaussian waves and its characteristics

A. Gaussian beam
Helmholtz equation
 2U  k 2U  0
Normally, a plan wave (in z direction) will be
U  U 0 exp{i (k  r )}  U 0 exp( ikz )
When amplitude is not constant the wave is
U  A( x, y , z ) exp( ikz )
An axis symmetric wave in the amplitude
U  A(r ) exp( ikz ) z
2
frequency   2 Wave vector k


Paraxial Helmholtz equation

Substitute the U into the Helmholtz equation we have:
A 2 2
 A  i 2k
2
T 0 where 
2
T  2 2
z x y
 A1 2
One simple solution is spherical wave ： A(r )  exp( jk )  2  x 2  y 2
z 2z
The other solution is Gaussian wave ：
 W0 2 2
U (r )  A0 exp[ 2 ]exp[ikz  ik  i ( z )]
W ( z) W ( z) 2 R( z )
z 2 1/ 2 z
where W ( z )  W0 [1  ( ) ]  ( z )  tan 1 z0 is Rayleigh range
z0 z0
z0 2  z0 1/2 1 1 
  i
R ( z )  z[1  ( ) ] W0  ( )  W ( z ) z 0  W (0) q( z ) R( z )  W 2 ( z )
z 
q parameter

Electric field of Gaussian wave propagates in z dirextion
A0 ( x 2  y 2 ) x2  y 2
E ( x, y , z )  exp[ 2
]  exp[ ik (  z )  i ( z )]
W ( z) W ( z) 2 R( z )
z 2 1/ 2
Beam width at z W ( z )  W0 [1  ( ) ]
z0
Waist width W0  W (0)
 W02 2 z0 2  W02
Radii of wave front at z R ( z )  z[1  ( ) ]  z[1  ( ) ], z0 
z z 
z 1 z
 Phase factor  ( z )  arctan  tg
 W02
z0

Gaussian beam at z=0

A0 r2
E ( x, y,0)  exp[ 2 ] where r 2  x 2  y 2 E A0
W0 W0 W0
z 2 1/ 2
Beam width: W ( z )  W0 [1  ( ) ] will be minimum A0
z0
eW0

    W 2 2 
 
wave front: lim R ( z )  lim  z 1   0
   
 z   
  
z 0 z 0
 
-W0 W0
at z=0, the wave front of Gaussian beam is a plan

surface, but the electric field is Gaussian form
W0 is the waist half width

B. The characteristics of Gaussian beam
Beam
radius
Gaussian beam is a axis symmetrical wave, at z=0

phase is plan and the intensity is Gaussian form,
at the other z, it is Gaussian spherical wave.

Intensity of Gaussian beam

W0 2 2
Intensity of Gaussian beam I ( , z)  I0[ ] exp[ 2 ]
W ( z) W ( z)
z=0 z=z0 z=2z0

y y y
x x x
I 1
I 1 I 1
I0 I0 I0
0 0 0
-1 0 1
-1 0 1
 W0  -1 0 1
W0 
W0
The normalized beam intensity as a function of the radial distance at different axial distances

On the beam axis ( = 0) the intensity

W0 2 I0
I (0, z )  I 0 [ ] 
W ( z) z
1  ( )2
z0
I
I0
1
1
0.5
 zo
0
0 zo z
The normalized beam intensity I/I0
at points on the beam axis (p=0) as a function of z

Power of the Gaussian beam

The power of Gaussian beam is calculated by the integration
of the optical intensity over a transverse plane
1
P I 0 W02
2
So that we can express the intensity of the beam by the power
2P 2 2
I ( , z)  exp[  2 ]
W 2 ( z) W ( z)
The ratio of the power carried within a circle of radius r. in the

transverse plane at position z to the total power is
1 0 2  02
P 0
I (  , z )2 d   1  exp[  2 ]
W ( z)

z 2 1/ 2 W
Beam Radius W ( z )  W0 [1  ( ) ] W ( z )  0 z  0 z
z0 z0
W(z)
Beam
2W0
waist
W0
 z
-z0 z0
The beam radius W(z) has its minimum value W0 at the waist (z=0)
reaches 2W0 at z=±z0 and increases linearly with z for large z.
Beam Divergence
dW ( z ) 2z  2W02 2
1
 
2  2  [( )  z2] 2 0 
dz W0   W0

The characteristics of divergence angle
z  0, 2  0
 W02
z , 2  2 /  W0

2 2W ( z )
z  , 2  or 2  lim
 W0 x  z
 W02
Let’s define f=z0 as the confocal parameter of Gaussian beam f  z0 

The physical means of f ： the half distance between two section of width
f z2
2 (1  2 )
2W ( z )  f 
2  lim  lim 2
z  z z  z f

Depth of Focus
Since the beam has its minimum width at z = 0, it achieves its best focus at the
plane z = 0. In either direction, the beam gradually grows “out of focus.” The
axial distance within which the beam radius lies within a factor 20.5 of its
minimum value (i.e., its area lies within a factor of 2 of its minimum) is known
as the depth of focus or confocal parameter
 
0 z 2 W02
2 z0 

2z
The depth of focus of a Gaussian beam.

Phase of Gaussian beam

The phase of the Gaussian beam is,
z0 2
k 2
R( z )  z[1  ( ) ]
 (  , z )  kz   ( z )  z
2 R( z )
On the beam axis (p = 0) the phase
 (0, z )  kz   ( z )
kz Phase of plan wave
 ( z) an excess delay of the wavefront in comparison with a

plane wave or a spherical wave
The excess delay is –/2 at z=-∞, and /2 at z= ∞
The total accumulated excess retardation as the wave travels from z

= -∞ to z =∞is. This phenomenon is known as the Guoy effect.

Wavefront
W02 2 f2
R ( z )  z[1  ( ) ] z
z z
Confocal field and its equal phase front

Parameters Required to Characterize a

Gaussian Beam
How many parameters are required to describe a plane wave, a spherical
wave, and a Gaussian beam?
 The plane wave is completely specified by its complex amplitude and

direction.
 The spherical wave is specified by its amplitude and the location of its origin.
 The Gaussian beam is characterized by more parameters- its peak
amplitude [the parameter A, its direction (the beam axis), the location of its
waist, and one additional parameter: the waist radius W0 or the Rayleigh
range zo,
 q-parameter q(z) is sufficient for characterizing a Gaussian beam of known
peak amplitude and beam axis
If the complex number q(z) = z + iz0, is known
the real part of q(z) z is the beam waist place
the imaginary parts of q(z) z0 is the Rayleigh range

Parameters required to describe a Gaussian beam

The Gaussian beam is characterized by : its peak amplitude [the
parameter A, its direction (the beam axis), the location of its waist, and
one additional parameter: the waist radius W0 or the Rayleigh range zo
q parameter is an sufficient factor to characteristic a Gaussian beam

of known peak amplitude and beam axis
1 1  1 1
 i  
q( z ) R( z ) W 2 ( z ) q ( z ) z  iz0
If the complex number q(z) = z + iz0, is known, the distance z to the beam
waist and the Rayleigh range z0. are readily identified as the real and
imaginary parts of q(z).

C. TRANSMISSION THROUGH OPTICAL

COMPONENTS
a). Transmission Through a Thin Lens
2 2 2 1 1 1 1 1 1
kz  k   k  kz  k     
2R 2f 2R ' R' R f R R' f
Notes:
R is positive since the wavefront of the incident beam is diverging and
R’ is negative since the wavefront of the transmitted beam is
converging.
In the thin lens transform, we have
W W '
1 1 1
 
W0 , z1 , f
R' R f
If we know W0, Z1, f, we can get  W 2 2 1

W  W [1  (
'2 2
) ]
0
R '
 R ' 2 1
 z '  R '[1  ( ) ]
W 2
The minus sign is due to the waist lies to the right of the lens.

W R'
W0 '  z ' 
[1  ( W 2 /  R ') 2 ]1/ 2 1  ( R '/ W 2 ) 2
because R  z[1  ( z0 / z ) 2 ] and W  W0 [1  ( z / z0 ) 2 ]1/2
Waist radius W0 '  MW0

The beam waist is
Waist location ( z ' f )  M ( z  f )
2
magnified by M, the
beam depth of focus is
Depth of focus 2 z0'  M 2 (2 z0 ) magnified by M2, and the
angular divergence is
2 0 minified by the factor M.
Divergence angle 2 0' 
， M
Mr z0 f
magnification M where r  Mr 
(1  r 2 )1/ 2 z f z f

Limit of Ray Optics

Consider the limiting case in which (z - f) >>zo, so that the lens is well
outside the depth of focus of the incident beam, The beam may then
be approximated by a spherical wave, thus
z0
r 0 and M  Mr
z f
z z’
W0 '  MW0
1 1 1
 
z' z f
f
M  Mr 
2W0 2W0’ z f
The magnification factor Mr is that based on ray optics. provides that M <
Mr, the maximum magnification attainable is the ray-optics magnification Mr.

b). Beam Shaping

Beam Focusing If a lens is placed at the waist of a Gaussian beam, so z=0, then
W0
W0 ' 
[1  ( z0 / f ) 2 ]1/ 2
f
z'
1  ( f / z0 ) 2
If the depth of focus of the incident beam 2z0, is much longer than the
focal length f of the lens, then W0’= ( f/zo)Wo. Using z0 =W02/, we obtain

W0 ' 
 W0
f  0 f z' f
The transmitted beam is then focused at the lens’ focal plane as

would be expected for parallel rays incident on a lens. This occurs
because the incident Gaussian beam is well approximated by a plane
wave at its waist. The spot size expected from ray optics is zero

Focus of Gaussian beam W' 

2
0
W02
z W02 2
(1  )  ( )
f f
For given f, W '02 changes as
•when z1  f W '02 decreases as z decreases

z1  0 W0 ' reaches minimum, and M<1, for f>0, it is focal effect
•when z1  f , W0 ' increases as z increases
•when z1  f the bigger z, smaller f, better focus

 W02
•when z f W '0 reaches maximum, when

 f , it will be
• focus

In laser scanning, laser printing, and laser fusion, it is desirable to generate the
smallest possible spot size, this may be achieved by use of the shortest
possible wavelength, the thickest incident beam, and the shortest focal length.
Since the lens should intercept the incident beam, its diameter D must be at
least 2W0. Assuming that D = 2Wo, the diameter of the focused spot is given by
4 f
2W0 '   F# F# 
 D
where F# is the F-number of the lens. A microscope objective with small

F-number is often used.

Beam collimate
locations of the waists of the incident and transmitted beams, z and z’ are
z' z / f 1
1 
f ( z / f  1) 2  ( z0 / f ) 2
The beam is collimated by making the location of the new waist z’ as

distant as possible from the lens. This is achieved by using the
smallest ratio z0/f (short depth of focus and long focal length).

Beam expanding
A Gaussian beam is expanded and collimated using two

lenses of focal lengths f1 and f2,
Assuming that f1<< z and z - f1>> z0, determine the optimal

distance d between the lenses such that the distance z’ to the
waist of the final beam is as large as possible.
overall magnification M = W0’/Wo

C). Reflection from a Spherical Mirror

Reflection of a Gaussian beam of curvature R1 from a mirror of curvature R:
1 1 2
W2  W1   f = -R/2.
R2 R1 R
R > 0 for convex mirrors and R < 0 for concave mirrors,

R R1   R1   R
 If the mirror is planar, i.e., R =∞, then R2= R1, so that the mirror reverses the
direction of the beam without altering its curvature
 If R1= ∞, i.e., the beam waist lies on the mirror, then R 2= R/2. If the mirror is
concave (R < 0), R2 < 0, so that the reflected beam acquires a negative
curvature and the wavefronts converge. The mirror then focuses the beam to a
smaller spot size.
 If R1= -R, i.e., the incident beam has the same curvature as the mirror, then R 2=
R. The wavefronts of both the incident and reflected waves coincide with the
mirror and the wave retraces its path. This is expected since the wavefront
normals are also normal to the mirror, so that the mirror reflects the wave back
onto itself. the mirror is concave (R < 0); the incident wave is diverging (R 1 > 0)
and the reflected wave is converging (R2< 0).

d). Transmission Through an Arbitrary Optical

System
An optical system is completely characterized by the matrix M of

elements (A, B, C, D) ray-transfer matrix relating the position and
inclination of the transmitted ray to those of the incident ray
The q-parameters, q1 and q2, of the incident and transmitted

Gaussian beams at the input and output planes of a par-axial optical
system described by the (A, B, C, D) matrix are related by

ABCD law
The q-parameters, q1 and q2, of the incident and transmitted
Gaussian beams at the input and output planes of a par-axial optical
system described by the (A, B, C, D) matrix are related by
Aq1  B
q2 
Cq1  D
Because the q parameter identifies the width W and curvature R of the

Gaussian beam, this simple law, called the ABCD law
Invariance of the ABCD Law to Cascading

If the ABCD law is applicable to each of two optical systems with
matrices Mi =(Ai, Bi, Ci, Di), i = 1,2,…, it must also apply to a system
comprising their cascade (a system with matrix M = M1M2).

C. HERMITE - GAUSSIAN BEAMS

The self-reproducing waves exist in the resonator, and resonating inside of
spherical mirrors, plan mirror or some other form paraboloidal wavefront
mirror, are called the modes of the resonator
Hermite - Gaussian Beam Complex Amplitude
W0 2x 2y x2  y2
U l ,m ( x, y, z )  Al ,m [ ]Gl [ ]Gm [ ]  exp[  jkz  jk  j (l  m  1) ( z )]
W ( z) W ( z) W ( z) 2 R( z )
u 2
where Gl (u )  H l (u ) exp( ), l  0,1, 2,...,
2
is known as the Hermite-Gaussian function of order l, and Al,m is a constant
Hermite-Gaussian beam of order (I, m). The Hermite-Gaussian

beam of order (0, 0) is the Gaussian beam.

H0(u) = 1, the Hermite-Gaussian function of order O, the Gaussian function.

G1(u) = 2u exp( -u2/2) is an odd function,
G2(u) = (4u2 - 2) exp( -u2/2) is even,
G3(u) = (8u3 - 12u)exp( -u2/2) is odd,

Intensity Distribution
The optical intensity of the (I, m) Hermite-Gaussian beam is
2 W0 2 2 2 x 2 2 y
I l ,m ( x, y, z )  Al ,m [ ] Gl [ ]Gm [ ]
W ( z) W ( z) W ( z)

C. Gaussian Modes
• Gaussian beams are modes of the spherical-mirror resonator;
Gaussian beams provide solutions of the Helmholtz equation under
the boundary conditions imposed by the spherical-mirror resonator
Beam
radius
a Gaussian beam is a circularly symmetric wave whose energy is confined

about its axis (the z axis) and whose wavefront normals are paraxial rays

2( x 2  y 2 )  i  k  z  x  y   tg 1 z 
2 2
2
Gaussian beam intensity: I  I  W0  e
  
  
W 2 (z)   2 R  z0 
0   e
W ( z ) 
The Rayleigh range z0  W02

z0 

where z0 is the distance called Rayleigh range, at which the beam
wavefronts are most curved or we usually called confocal
prrameter z
Beam width W ( z )  W0 [1  ( ) 2 ]1/ 2
z0
minimum value W0 at the beam waist (z = 0).
z02
The radius of curvature R( z )  z 
z
  z 2   z z0  z0 2
R  R( z )  z 1      z0     z 
  z0    z0 z  z
 z0
Beam waist W0 


Gaussian Mode of a Symmetrical Spherical-

Mirror Resonator
z2  z1  d
R1 d R2 z02
R1  z1 
z1 2
z0
 R2  z2 
z2
 d ( R2  d )
z1  , z2  z1  d
R2  R1  2d
z1 0 z2 z z 2   d ( R1  d )( R2  d )( R2  R1  2d )
0
( R2  R1  2d )
2

the beam radii at the mirrors
zi 2 1/ 2
Wi  W0 [1  ( ) ] , i  1, 2.
z0
An imaginary value of z0 signifies that the Gaussian beam is in fact a

paraboloidal wave, which is an unconfined solution, so that z0 must be
real. it is not difficult to show that the condition z02 > 0 is equivalent to
d d
0  (1  )(1  )  1
R1 R2

Gaussian Mode of a Symmetrical Spherical-

Mirror Resonator
Symmetrical resonators with concave mirrors that is R1 = R2= -/RI
so that z1 = -d/2, z2 = d/2. Thus the beam center lies at the center
d R
z0  (2  1)1/ 2
2 d
d R
W 2
(2  1)1/ 2
2
0
d
d / 
W12  W22 
{(d / R )[2  (d / R )]}1/ 2
The confinement condition becomes

d
0 2
R

Given a resonator of fixed mirror separation d, we now examine the

effect of increasing mirror curvature (increasing d/lRI) on the beam
radius at the waist W0, and at the mirrors Wl = W2.
As d/lRI increases, W0 decreases

until it vanishes for the concentric
resonator (d/lR| = 2); at this point
W1 = W 2 = ∞
The radius of the beam at the
mirrors has its minimum value, WI =
W2= (d/)1/2, when d/lRI = 1
d  d 1/ 2
z0  W0  ( ) W1  W2  2W0
2 2

C. Resonance Frequencies
k ( x2  y 2 )
The phase of a Gaussian beam,  ( x, y, z )  kz  tg ( z
1
z0
)
2 R( z )
At the locations of the mirrors z1 and z2 on the optical axia (x2+y2=0), we have,
 z 
 (0, z2 )   (0, z1 )  k ( z2  z1 )  [ ( z2 )   ( z1 )]  kd   where  ( z )  tg 1  
 z0 
As the traveling wave completes a round trip between the two

mirrors, therefore, its phase changes by
2kz  2
For the resonance, the phase must be in condition 2kz  2  2q , q  1, 2,3...
If we consider the plane wave resonance frequency
 k  2 and  F  c 2d
c
We have  q  q F  F


Spherical-Mirror Resonator Resonance Frequencies (Gaussian Modes)


 q  q F  F

1. The frequency spacing of adjacent modes is VF = c/2d, which is the same
result as that obtained for the planar-mirror resonator.
2. For spherical-mirror resonators, this frequency spacing is independent of
the curvatures of the mirrors.
3. The second term in the fomula, which does depend on the mirror
curvatures, simply represents a displacement of all resonance frequencies.

D. Hermite - Gaussian Modes

The resolution for Helmholtz equation An entire family of solutions, the
Hermite-Gaussian family, exists. Although a Hermite-Gaussian beam of order
(I, m) has the same wavefronts as a Gaussian beam, its amplitude distribution
differs . It follows that the entire family of Hermite-Gaussian beams represents
modes of the spherical-mirror resonator
W0 2x 2y x2  y2
U l ,m ( x, y, z )  Al ,m [ ]Gl [ ]Gm [ ]  exp[  jkz  jk  j (l  m  1) ( z )]
W ( z) W ( z) W ( z) 2 R( z )
 (0, z )  kz  (l  m  1) ( z ) 2kd  2(l  m  1)  2 q, q  0, 1, 2,...,
Spherical mirror resonator Resonance Frequencies

(Hermite -Gaussian Modes)

 l ,m ,q  q F  (l  m  1) F


longitudinal or axial modes: different q and same indices (l, m) the

intensity will be the same
transverse modes: The indices (I, m) label different means different

spatial intensity dependences

 l ,m ,q  q F  (l  m  1) F

Longitudinal modes corresponding to a given transverse mode (I, m) have resonance

frequencies spaced by vF = c/2d, i.e., vI,m,q – vI’,m’,q = vF.
Transverse modes, for which the sum of the indices l+ m is the same, have the same
resonance frequencies.
Two transverse modes (I, m), (I’, m’) corresponding mode q frequencies spaced

 l ,m,q  l ',m ',q  [(l  m)  (l ' m ')] F


*E. Finite Apertures and Diffraction Loss

Since the resonator mirrors are of finite extent, a portion of the optical power escapes
from the resonator on each pass. An estimate of the power loss may be determined by
calculating the fractional power of the beam that is not intercepted by the mirror. That is
the finite apertures effect and this effect will cause diffraction loss.
For example:
If the Gaussian beam with radius W and the mirror is circular with radius a and a= 2W, each time
there is a small fraction, exp( - 2a2/ W2) = 3.35 x10-4, of the beam power escapes on each pass.
Higher-order transverse modes suffer greater losses since they have greater spatial extent in
the transverse plane.
 In the resonator, the mirror transmission and any aperture limitation will
induce loss
 The aperture induce loss is due to diffraction loss, and the loss depend
mainly on the diameters of laser beam, the aperture place and its diameter
 We can used Fresnel number N to represent the relation between the size
of light beam and the aperture, and use N to represent the loss of
resonator.

Diffraction loss
The Fresnel number NF
a2 a2 a2
NF   
d  2 z  W 2
Attention: the W here is the beam width in the mirror, a is the dia. of mirror
Physical meaning ： the ratio of the accepting angle (a/d) (form

one mirror to the other of the resonator )to diffractive angle of
the beam (/a) .
The higher Fresnel number corresponds to a smaller loss

N is the maximum number of trip that light will propagate in side resonator
without escape.
1/N represent each round trip the ratio of diffraction loss to the total energy
a12 a22
Symmetric confocal resonator   NF
 W1  W2
2 2
For general stable concave mirror resonator, the Fresnel number for two
mirrors are:
1
a12 a12 g1
N F1   [ (1  g1 g 2 )] 2
 W1 2
d  g2
1
a22 a22 g 2
NF 2   [ (1  g g )] 2
 W22 d  g1 1 2

Home work 2
• The light from a Nd:YAG laser at wavelength 1.06 mm is a Gaussian beam of 1 W optical power and
beam divergence 2q0= 1 mrad. Determine the beam waist radius, the depth of focus, the maximum
intensity, and the intensity on the beam axis at a distance z = 100 cm from the beam waist.
• Beam Focusing. An argon-ion laser produces a Gaussian beam of wavelength l = 488 nm and waist
radius w0 = 0.5 mm. Design a single-lens optical system for focusing the light to a spot of diameter
100 pm. What is the shortest focal-length lens that may be used?
• Spot Size. A Gaussian beam of Rayleigh range z0 = 50 cm and wavelength l=488nm is converted
into a Gaussian beam of waist radius W0’ using a lens of focal length f = 5 cm at a distance z from
its waist. Write a computer program to plot W0’ as a function of z. Verify that in the limit z - f >>z0 ,
the relations (as follows) hold; and in the limit z << z0 holds.
• Beam Refraction. A Gaussian beam is incident from air (n = 1) into a medium with a planar
boundary and refractive index n = 1.5. The beam axis is normal to the boundary and the beam waist
lies at the boundary. Sketch the transmitted beam. If the angular divergence of the beam in air is 1
mrad, what is the angular divergence in the medium?
W0 '  MW0 M  M r  f
z f
W0
W0 ' 
[1  ( z0 / f ) 2 ]1/ 2

 Resonance Frequencies of a Resonator with an Etalon. (a) Determine the

spacing between adjacent resonance frequencies in a resonator
constructed of two parallel planar mirrors separated by a distance d = 15
cm in air (n = 1).(b) A transparent plate of thickness d1= 2.5 cm and
refractive index n = 1.5 is placed inside the resonator and is tilted slightly to
prevent light reflected from the plate from reaching the mirrors. Determine
the spacing between the resonance frequencies of the resonator.
 Mirrorless Resonators. Semiconductor lasers are often fabricated from
crystals whose surfaces are cleaved along crystal planes. These surfaces
act as reflectors and therefore serve as the resonator mirrors. Consider a
crystal with refractive index n = 3.6 placed in air (n = 1). The light reflects
between two parallel surfaces separated by the distance d = 0.2 mm.
Determine the spacing between resonance frequencies vF, the overall
distributed loss coefficient ar, the finesse F, and the spectral width dv.
Assume that the loss coefficient (as= 1 cm-1).

b). Parameter q transform for thin lens

If put the original point at the waist of beam, the parameter q will be
1 1  The real part is concern with wave

 i
q( z ) R( z )  W 2 ( z ) front R(z), the imaginary part is
corresponding to beam radii w(z)
For q0 is the value at R(0)→∞

i W02
q0  q (0) 

Finally we have ：
W02
q( z )  i  z  q0  z


Parameter q in the thin lens transform

W1  W2 1 1 1 i W02
  q01  q (0) 
1 1 1 q' q f 
 
R' R f
q1  q01  z
q1  q  z1
1 1 1
 
q2 q1 f q2  q ' z2
1 1 1
 
q1 q2 f
1 1 1
 
q ' z2 q  z1 f

continue
z2 zz Real part
(1  )q  ( z1  z2  1 2 )
f f
q'  z z f  z1 W012
q z (1  1 )q02  (1  2 )q01  2
( )  (1  1 )
f f f f f  z 2 W02
 W012
q  q01  i Imaginary part

 W022 q01q02 z1 z 2  2W012W022 z1
q  q02  i
'   z1  z 2   z  z (1  )
2 f
1 2
 f f f
W012 W
W  2
W0 ' 
02
z1 2 W01 2 [1  ( W 2 /  R ') 2 ]1/ 2
(1  ) ( )
f f
z
(1  1 )
f
z 2  f [1  ]
z1 2 W01 2
(1  )  ( )
f f

Special cases
• when z1→∞, then z2=F when incident beam waist is
in the far infinite the outlet beam waist is located at
the focal plan
• when z1=f, then z2=F, the incident beam waist is in

the object focal plan, the outlet beam waist will be in
the imaging focal plan
  W012   z1 
2
1 1 1
• when   f   1  f  , then z  z  f
  1 2

Beam waist radius W0 '  MW0
Waist location ( z ' f )  M 2 ( z  f )
，
Depth of focus 2 z0'  M 2 (2 z0 )
2 0
Divergence 2 0' 
M
Mr z0 f
Magnification M r  Mr 
(1  r 2 )1/ 2 z f z f

CHAPTER 2
PHOTONS AND ATOMS

Max Planck Albert Einstein

Nobel prize 1918 Nobel prize 1921

1. Atoms, Molecules, and Solids

• Atomic Physics’ starting point: the old Bohr model
Schrodinger equation

 2   (r , t )
   (r , t )  V (r , t ) (r , t )  j 
2
2m t
Time independent Schrodinger equation
 2 2 
   (r )  V (r ) (r )  E (r )
2m

-on discrete stationary states

-radiative transitions 
quantum jumps between levels
-The old Bohr model 
-energy levels
For molecular oscillation:

1
Eq  (q  )
2

Vibrational and Rotational Energy Levels of Molecules

e e
v N2 CO2 v
0.4 050 0.4
200
040
0.3 q=1 001 0.3
Energy
9.6um 030
10.6um
0.2 100 0.2
020
0.1 010 0.1
q=0 000
0 0
Asymmetric Symmetric Bending
stretch stretch
Figure Lowest vibrational energy levels of the N2 and CO2 molecules (the
zero of energy is chosen at q=0). The transitions marked by arrows represent
energy exchanges corresponding to photons of wavelengths 10.6um and
9.6um, as indicated. These transitions are used in CO2 lasers.
Electron Energy Levels of Atoms and Molecules

e e
v He Ne v
21 1s2s 1S0 2p55s 3.39um 2p54 21
p
20 1s2s 3S1 2p54 20
632.8nm
s
Energy
19 19
18 2p53p 18
1 2p53s 1
7 7
1 1
6 6
Odd Even
parity parity
Figure 12.1-3 Some energy levels of He and Ne atoms. The
He transitions marked by arrows correspond to photons of
wavelengths 3.39m and 632.8nm, as indicated. These
transitions are used in He-Ne lasers.
Electron Energy Levels in Solids

Isolated atoms and molecules exhibit discrete energy levels, as shown in
Figs.4.3-5 to 4.3-8. For solids, however, the atoms, ions, or molecules in close
proximity to each other and cannot therefore be considered as simple collections
of isolated atoms; rather, they must be treated as a many-body system.
Vacuum
level
3p Eg E
g
3s
2
Energy
2
s
1
s Isolate Semi- Insulator
Metal
d conduct
Figure 12.1-5 Broadening of the discrete energy levels of an isolatedor
atom atom into bands for solis-state materials.

Quantum-well Structure
Multi-layers of different semiconductor materials
Conduction
band
Energy
GaAs AlGaAs
Valence
band
0 20 4 6 8 10 12
Distance
0 0 0 0 0
nm
Figure 12.1-8 Quantized energies in a single-crystal AlGaAs/GaAs
multiquantum-well structure. The well widths can be arbitrary (as
shown) or periodic.

2. Occupation of energy levels in thermal

equilibrium
Boltzmann distribution
P ( Em )  exp( Em / k BT ), m  1, 2,...,

Em Em
E3 E3
E2 E2
E1 E1
P(Em)
Energy levels Occupation
N2 E2  E1
 exp( )
N1 k BT

N2 g2 E2  E1
 exp( )
N1 g1 k BT

Fermi-Dirac Distribution
Pauli exclusion principle

E
Boltzman
n
P(Em)
Ef
Fermi-
Dirac
f(E)
0 1/2 1
Figure 12.1-11 The Fermi-Dirac distribution f(E) is well approxiamated by the

Boltzmann distribution P(Em) when E>>Ef.
1
f (E) 
exp[( E  E f ) / k BT )]  1

3.Interactions of photons with atoms

• Semi-classical view of atom excitations
e
Ze
Atom in ground
state
e
Ze
Atom in excited
state

Spontaneous Emission
Figure 4.4-1 Spontaneous

h emission of a photon into the
mode of frequency by an
atomic transition from energy
level 2 to level 1. The photon
点击查看 flash 动 energy h
画
c
psp   ( ) transition cross-section.
V

Decay of the number of the excited atoms

N(t)
N(0)
1/Psp t
Figure 12.2-2 Spontaneous emission into a single mode

causes the number of excited atoms to decrease
exponentially with time constant 1/Psp
N (t )  N (0) exp( psp t )

Absorption
Figure 12.2-3 Absorption of a

h photon h leads to an upwoard
transition of the atom from energy
level 1 to energy level 2.
点击查看 flash 动
画
c
Pab  n  ( )
V

Stimulated Emission
Energy
Super-excited
state
excited
state
c
h
Pst   ( )
V
h
c
h Pst  n  ( )
V
Wi  Pab  Pst
ground
点击查看 flash 动画 state
When a photon enters, it “knocks” an electron

from the inverted population down to the ground
state, thus creating a new photon. This
amplification process is called stimulated
emission. 10
Fundamentals of Photonics 24/3/11
Lineshape Function
Transition Strength

S    ( )d

Lineshape function
 ( )  Sg ( )

 
g ( )d  1

Figure The transition cross section and the lineshape function

g()

Total Spontaneous Emission into All Modes
8 S
Psp  M ( 0 )cS 
 2

1/ tsp  Psp  M ( 0 )cS
1
Psp 
tsp
2
S
8 tsp

• Relation between the Transition Cross Section

and the Spontaneous Lifetime
2
 (v )  g (v )
8 tsp
2
 0   (v0 )  g (v0 )
8 tsp

Stimulated Emission and Absorption

• Transitions Induced by Monochromatic Light
I V
 n 
h c
Wi   ( )
• Transitions Induced by Broadband Light
  ( )V c
Wi   [  ( )]d
0 h V

 ( 0 )V   ( 0 )
Wi  c   ( )d  cS
h 0 0 h 0
3
Wi   ( 0 ) (4.4-17)
8 htsp

Einstein Coefficients
Psp  A
Wi  B  ( 0 )

Psp  A 1
A
tsp
Wi  B  ( 0 )
3
1 B
Psp  8 htsp
tsp

Lineshape Broadening
h 1 1 h 1
Life-time broadening E  E1  E2  (  )
2  1  2 2 
1 1 1
Caused by the time
  (  )
2  1  2
uncertainty of the
 / 2
occupation of the g ( ) 
(  0 ) 2  ( / 2) 2
energy level
2 1
0 
2 2 tsp 
2
0 
2
Collision Broadening
Figure 12.2-8 A sinewave interrupted at the rate fcol by random phase

jumps has a Lorentzian spectrum of width △=fcol/

Inhomogeneous Broadening
• Doppler effect
4 1
Direction
of observation
Figure 12.2-10 The radiated frequency is dependent on the direction of

atomic motion relative to the direction of observation. Radiation from
atom 1 has higher frequency than that from atoms 3 and 4. Radiation
from atom 2 has lower frequency.

0
c
g ( )
g ( )
P(v)

g (  0 )
c
0 0  
Velocity v
Figure 12.2-11 The velocity distribution and average lineshape

function of a Doppler-broadened atomic system.
 V
g ( )   g (v  v0 ) p (V )dV
 c

3. Thermal light
Thermal Equilibrium Between Photons and Atoms
1
Psp  A A
tsp
Wi  B  ( 0 ) Wi  Psp  Pst
dN 2 N 2 nN1 nN 2
  
dt tsp tsp tsp

N2 E2  E1 h
 exp( )  exp( )
N1 k BT k BT
The average number of photons in a mode of frequency 
1
n
exp(h / k BT )  1

Black-body Radiation
h
E
exp(h / k BT )  1
8 h 3 1
 ( ) 
c 3 exp(h / k BT )  1

E
kT
0
kT kT kT v
10
10h h h
Figure 12.3-2 Semilogarithmic plot of the average energy E of

an electromagnetic mode in thermal equilibrium at temperature
T as a function of the mode frequency At T=300K, KBT/h =
6.25THz, which corrsponds to a wavelength of 48um.
Spontaneous emissionl
返回

Absorption
返回

Stimulated Emission
返回

Chapter 4
Laser

LASERS
In 1958 Arthur Schawlow, together with Charles Townes, showed how to

extend the principle of the maser to the optical region. He shared the 1981
Nobel Prize with Nicolaas Bloembergen. Maiman demonstrated the first
successful operation of the ruby laser in 1960.

LASERS
an oscillator is an amplifier with positive feedback
Two conditions for an oscillation:

1. Gain greater than loss: net gain
2. Phase shift in a round trip is a multiple of 2π

Stable condition: gain = loss

Gain
Loss
0 Power
Steady-state
power
If the initical amplifier gain is greater than the loss, oscillation may initiate.
The amplifier then satuates whereupon its gain decreases. A steady-state
condition is reached when the gain just equals the loss.
An oscillator comprises:
◆ An amplifier with a gain-saturation mechanism
◆ A feedback system
◆ A frequency-selection mechanism
◆ An output coupling scheme

Light amplifier with positive feedback
Pout  gPin
Pin Gain medium (e.g. 3-
level system w
population inversion)
When the gain exceeds the roundtrip losses,

the system goes into oscillation
+
 g
+

LASERS
Amplified once Initial photon
画
Gain medium
Reflected (e.g. 3-level
system w
population
Amplified Output
inversion)
twice
Reflected
Amplified Partially
Again reflecting
Mirror
Light
Amplification through
Stimualted
Emission
Radiation

Mirror
Active medium
Partially Laser
d transmitting output
mirror
A laser consists of an optical amplifier (employing

an active medium) placed within an optical
resonator. The output is extracted through a
partially transmitting mirror.

Optical amplification and feedback

★ Gain medium
The laser amplifier is a distributed-gain device
characterized by its gain coefficient
2
 0 ( )  N 0 ( )  N 0 g ( ) 5.1-43 Small signal
8 tsp Gain Coefficient
 0 ( )
 ( )  5.1-42 Saturated
1   / s ( ) Gain Coefficient

  0
 ( )   ( ) Phase-shift Coefficient
 (Lorentzian Lineshape)
Figure 5.1-5 Spectral dependence of the gain and phase-shift

coefficients for an optical amplifier with Lorentzian lineshape function

Optical Feedback-Optical Resonator

Feedback and Loss: The optical resonator
A Fabry-Perot resonator, comprising two mirrors separated by a distance d, contains

the medium (refractive index n). Travel through the medium introduces a phase shift
per unit length equal to the wavenumber
2
k
c
In traveling a round trip through a resonator of length d, the photon-flux density is
reduced by the factor R1R2exp(-2sd). The overall loss in one round trip can therefore
be described by a total effective distributed loss coefficient r, where
exp(2 r d )  R1 R2 exp(2 s d )

Loss coefficient  r   s   m1   m 2
1 1
 m1  ln
2d R1
1 1
 m2  ln
2d R2
1 1
 m   m1   m 2  ln
2d R1 R2
1
Photon lifetime p 
rc
r represents the total loss of energy (or number of photons) per

unit length, arc represents the loss of photons per second

 q  q F , q  1, 2,...,
F
  , F  c / 2d
F

F  2 p F
r d
c
F 
2d
Resonator
response 
 q 1 q  q 1 
Resonator modes are separated by the frequency
 F  c / 2d and have linewidths    F / F  1/ 2 p .

Conditions for laser oscillation

Gain condition: Laser threshold
 0 ( )   r Threshold Gain
Condition
N0  Nt
r 1
where Nt  or Nt 
 ( ) c p ( )
8 tsp 1
Nt  2 Threshold Population
 c  p g ( ) Difference

For a Lorentzian lineshape function, as g ( 0 )  2 / 
2 2 tsp
Nt  2
 c p
If the transition is limited by lifetime broadening with a decay time t sp
2 2 r
Nt  2  2
 c p 
As a numerical example, if m, p=1 ns, and the refractive index n=1, we
obtain Nt=2.1×107 cm-3

Conditions for laser oscillation(2)

Phase condition: Laser Frequencies
2kd  2 ( ) d  2 q, q  1, 2……
Frequency Pulling
c   0
  ( )   q
2 
c   0
or   q   ( )
2 
   q'   q
c  q  0
  q 
'
 ( q )
2 
q


   q  ( q  0 )
'
Laser Frequencies

q

The laser oscillation frequencies fall near the cold-

resonator modes; they are pulled slightly toward the
atomic resonance central frequency 

Characteristics of the laser output

Internal Photon-Flux Density
Gain Clamping
 0 ( ) /[1   / s ( )]   r

Laser
turn-on
 0 ( ) Time Steady
state
r Loss coefficient
Gain coefficient
s ( ) s ( )
0 10s ( )
10
Photon-flux density
Determination of the steady-state laser photon-flux density 

The smaller the loss, the greater the value of 

Steady Photon Density

 0 ( )
  s ( )[  1],  0 ( )   r
r
  0,  0 ( )   r
Since  0 ( )  N 0 ( ) and  r  N t ( )
N0
  s ( )(  1), N 0  N t Steady-State Laser
Nt Internal Photon-Flux
Density
  0, N 0  N t

Steady-state values of the population difference N, and the laser internal photon-flux
density , as functions of N0 (the population difference in the absence of radiation; N 0,
increases with the pumping rate R).
T
Output photon-flux density 0 
2
h T 
Optical Intensity of Laser Output I0 
2

Optimization of the output photon-flux density

1 1 1
From  m1  ln   ln(1  T )
2d R1 2d
1
We obtain  r   s   m 2  ln(1  T )
2d
1 g0
0  sT [  1], g0  2 0 ( )d , L  2( s   m 2 ) d
2 L  ln(1  T )
When T 1
,
use the approximation ln(1  T )  T
Then Top  ( g 0 L)1/ 2  L


Spectral Distribution
Determined both by the atomic lineshape and by the
resonant modes
B
M Number of Possible
F Laser Modes
Linewidth   ?
Schawlow-Townes limit

 0 (
) Gain

r Loss
B
  
F
Resonator modes
  ……
   
Allowed modes
Figure 5.3-3 (a) Laser oscillation can occur only at frequencies for which
the gain coefficient is greater than the loss coefficient (stippled region).
(b) Oscillation can occur only within of the resonator modal
frequencies (which are represented as lines for simplicity of illustration ).

Homogeneously Broadened Medium

 0 ( )
 ( )  点击查看 flash 动
1   i 1 j / s ( j )
M
画

Inhomogeneously Broadened Medium

 0 ( )  ( ) 点击查看 flash 动
画
 s
c
r 2d

Frequency 
 q 1  q  q 1 
(a) (b)
Figure 5.3-6 (a) Laser oscillation occurs in an inhomogeneously broadened medium by each mode
independently burning a hole in the overrall spectral gain profile. (b) Spectrum of a typical
inhomogeneously broadened multimode gas laser.

Hole burning in a Doppler-broadened medium
画

Hole burning in a Doppler-broadened medium

画

Spatial distribution and polarization

Spatial distribution
x,y
Laser
intensity
Spherical Spherical
mirror mirror
The laser output for the (0,0) tansverse mode of a spherical-

mirror resonator takes the form of a Gaussian beam.

 0,0
 1,1
TEM0,0
11
00
B11
B00

(1,1) modes
(0,0) modes
Laser TEM1,1
output

Figure 5.3-8 The gains and losses for two transverse modes, say (0,0) and (1,1), usually
differ because of their different spatial distributions.

Two Issues: Polarization, Unstable Resonators

Mode Selection
Selection of
1. Laser Line
2. Transverse Mode
3. Polarization
4. Longitudinal Mode

High Output mirror

reflectance
mirror
Laser
Active medium output
Prism Aperture
Unwanted
line
Figure 5.3-9 A paticular atomic line may be selected by

the use of a prism placed inside the resonator. A
transverse mode may be selected by means of a
spatial aperture of carefully chosen shaped and size.

Brewster Active Brewster

window midum window Polarized
laser output
B
High Output
reflectance mirror
mirror
Figure 5.3-10 The use of Brewster windows in a gas laser provides a

linearly polarized laser beam. Light polarized in the plane of incidence
(the TM wave) is transmitted without reflection loss through a window
placed at the Brewster angle. The orthogonally polarized (TE) mode
suffers reflection loss and therefore does not oscillate.

Etalon
Selection of Longitudinal Mode
Active midum
High reflectance d1 Output
mirror mirror
d
Resonator loss
c/2d
Resonator mdoes
Etalon mdoes
c/2d1
Laser output
Figure 5.3-11 Longitudianl mode selection by the use of an intracavity etalon. Oscillation
occurs at frequencies where a mode of the resonator coincides with an etalon mode; both
must, of course, lie within the spectral window where the gain of the medium exceeds the loss.

Multiple Mirror Resonators
(a)
(b)
(c)
Figure 5.3-12 Longitudinal mode selection by use of (a) two

coupled resonators (one passive and one active); (b) two
coupled active resonators; (c) a coupled resonator-
interferometer.

Characteristics of Common Lasers
Solid State Lasers: Ruby, Nd3+:YAG, Nd3+:Silica, Er3+:Fiber, Yb3+:Fiber
Gas Lasers: He-Ne, Ar+; CO2, CO, KF;
Liquid Lasers: Dye
Plasma X-Ray Lasers
Free Electron Lasers


Pulsed Lasers
Method of pulsing lasers External Modulator or Internal Modulator?
Modulator Modulator
Average
CW power
power
t t
(a) (b)
Figure 5.4-1 Comparison of pulsed laser outputs achievable
with (a) an external modulator, and (b) an internal modulator
1. Gain switching 2. Q-Switching

3. Cavity Dumping 4.Mode Locking

Gain Switching

Q- Switching
Loss
Modulator
Gain
t
Laser
output
t t
Figure 5.4-3 Q-switching.

Cavity Dumping
Figure 5.4-4 Cavity dumping.

Mode locking
• Laser modes coupling together
• Lock their phases to each other
Rate equation for the photon-number density

dn n
   NWi
dt p
p photon lifetime
Wi   ( )  cn ( )
From  ( )  1/ c p N t Probability density for induced

absorption/emission
We have dn n N n
  Photon-Number Rate
dt  p Nt  p Equation

Rate equation for the Population Difference

For a three level system
dN 2 N
 R  2  Wi ( N 2  N1 )
dt tsp
Note N1  ( N a  N ) / 2, N 2  ( N a  N ) / 2, N  N 2  N1
dN N 0 N
Then    2Wi N
dt tsp tsp
Where the small signal population difference N 0  2 Rtsp  N a
Substituting Wi  n / N t p
dN N 0 N N n Population-difference rate
We have   2
dt tsp tsp Nt  p equation (Three-level system)

Situation for the gain switching

Situation for the Q-switching

画

Dynamics of a Q-Switching process
Dynamics of the Q-switching process

Note the time relationship between the photon density and the
population inversion variations!

Determination of the peak power, energy, width

and shape of the optical pulse
dn N n
 (  1)
dt Nt p
dN N n
 2
dt Nt  p
dn 1 N t
Dividing  (  1)
dN 2 N
1 1
n  N t ln( N )  N  cons tan t
2 2

1 N 1
n  N t ln  ( N  Ni )
2 Ni 2
1 c
Power P0  h A0  h cTAn  h T Vn
2 2d
Peak pulse power 1 N N N
np  N i (1  t ln t  t )
2 Ni Ni Ni
c
Pp  h T Vn p
2d
When Ni>>Nt
1 1 c
It is clear np  Ni So Pp  h T VN i
2 2 2d
The larger the initial population inversion,
the higher the Q-switched pulse peak power.

c. Pulse energy:
c tf c Nf dt
E  h T V  n(t )dt h T V  n(t ) dN
2d ti 2d N i dN
1 c Ni dN
E  h T VN t p 
2 2d Nf N
1 c N
E  h T VN t p ln i
2 2d Nf
The final population difference Nf
Ni Ni  N f
ln 
Nf Nt
1 c
E  h T V p ( Ni  N f )
2 2d

d. Pulse width:
A rough estimation of the pulse width is the ratio

of the pulse energy to the peak pulse power.
Ni / Nt  N f / Nt
 pulse   p
N i / N t  ln( N i / N t )  1
When Ni>>Nth and Ni>>Nf
 pulse   p The shorter the photon life time,

the shorter the Q-switched pulses.


Techniques for Q-switching

1. Mechanical rotating mirror method:
Q-switching principle: rotating the cavity mirror results in

the cavity losses high and low, so the Q-switching is obtained.
Advantages: simple, inexpensive.
Disadvantages: very slow, mechanical vibrations.

2. Electro-optic Q-switching
Disadvantages:
complicate and expensive
Advantages:
very fast and stable.
Pockels effect: applying electrical field in a uniaxial crystal results in

additional birefringence, which changes the polarization of light when
passing through it.
Q-switching principle: placing an electro-optic crystal between crossed

polarizers comprises a Pockels switch. Turning on and off the electrical
field results in high and low cavity losses.

Electro-optic Q-switch operated at (a) quarter-wave and (b) half-wave retardation voltage

3. Acousto-optic Q-switching

Diffracted light
Incident light
 
Sound Transmitted light
Piezoelectric transducer RF
Bragg scattering: due to existence of the acoustic wave, light

changes its propagation direction.
Q-switching principle: through switching on and off of the acoustic
wave the cavity losses is modulated.
Advantages: works even for long wavelength lasers.
Disadvantages: low modulation depth and slow.

4. Saturable absorber Q-switching

What’s a saturable absorber?
0 Absorption coefficient of the material is

 
I reversely proportional to the light intensity.
1
Is Is: saturation intensity.
Saturable absorber Q-switching:
Insertion a saturable absorber in the

laser cavity, the Q-switching will be
automatically obtained.

General characteristics of laser Q-switching

• Pulsed laser output:
– Pulse duration – related to the photon lifetime.
– Pulse energy - related to the upper level lifetime.
• Laser operation mode:
– Single or multi-longitudinal modes.
• Active verses passive Q-switching methods:
– Passive: simple, economic, pulse jitter and intensity fluctuations.
– Active: stable pulse energy and repetition, expensive.
• Comparison with chopped laser beams:
– Energy concentration in time axis.
• Function of gain medium
– Energy storage

Laser mode-locking
Aims:
1. Familiarize with the principle of laser mode-locking.
2. Familiarize with different techniques of achieving laser
Mode-locking.
Outlines:
1. Principle of laser mode-locking.
2. Methods of laser mode-locking.
3. Active mode-locking.
4. Passive mode-locking.
5. Transform-limited pulses.

Principle of laser mode-locking

1. Lasing in inhomogeneously broadened lasers:
i) Laser gain and spectral hole-burning.
ii) Cavity longitudinal mode frequencies.
iii) Multi-longitudinal mode operation.

2. Laser multimode operation:

Single mode lasers: E t   E0 cos 0t   t 
N
Multimode lasers: E t    Ei cos i t   i t 

i 1
c
Mode-frequency separations: ~
nL
Phase relation between modes: Random and independent!
Total laser intensity fluctuates with time !
The mean intensity of a

multimode laser remains
constant, however, its
instant intensity varies with
time.

3. Effect of mode-locking:
(i) Supposing that the phases of all modes are locked together:
 i t    0  0
(ii) Supposing that all modes have the same amplitude:
purely for the convenience of the
Ei  E0 mathematical analysis
(iii) Under the above two conditions, the total electric field
of the multimode laser is:
N i i t 
E t   Re  Ei e  where
 i 1 
 N  1 c
 i   0  i    c  c 
 2  L

0 is the frequency of the central mode, N is the number

of modes in the laser, c is the mode frequency separation.
i is the frequency of the i-th mode.
Calculating the summation yields:
  c t 
sin N  Note this is the optical
 2 
E t   E 0 cos  0 t field of the total laser
   c 
t
sin  Emission !
 2 
The optical filed can be thought to consist of a carrier wave of

frequency 0 that amplitude modulated by the function
sin Nx 
AN x  
sin x 

4. Characteristics of the mode-locked lasers:

The intensity of the laser field is:
  c t 
sin 2  N 
 2 
I t   E02
  c t 
sin 2  
 2 
The output of a mode-locked laser consists of a series of pulses.

The time separation between two pulses is determined by RT and
the pulse width of each pulse is tp.

5. Properties of mode-locked pulses:

i) The pulse separation RT:
  c t 
Sin 2  0 c t  2
 2 
2 2L
 RT   The round-trip time of the cavity!
 c c
ii) The peak power: Considering sin    when  is small,
E 2 0  N 2 E 02
N times of the average power. N: number of modes.
The more the modes the higher the peak power of the
Mode-locked pulses.

iii)The individual pulse width:

  c t  2
sin N 0 t p 
 2  N  c
N
 a 2 1
 c
t p   a: bandwidth of
 a  a the gain profile.
 RT
Narrower as N increases. t p 
N
The mode locked pulse width is reversely proportional to

the gain band width, so the broader the gain profile, the shorter
are the mode locked pulses.

Techniques of laser mode-locking

Active mode-locking:
Actively modulating the gain or loss of a laser cavity in a periodic way,
usually at the cavity repetition frequency c/2nL to achieve mode-locking.
Amplitude modulation:
A modulator with a transmission function of
   2t  
T  1   1  cos  

    RT  
is inserted in the laser cavity to modulate the light. Where  is the
modulation strength and  < 0.5. Under the influence of the modulation
phases of the lasing modes become synchronized and as a consequence
become mode-locked.
Operation mechanism of the technique:

Time domain analysis:

Consider the extreme case where a shutter is inside the cavity, and
the opens only for a short time every second. Is the cavity round
trip time. In this case only a pulse with pulse width narrower than
the opening time can survive in the cavity, all the CW type of
operation will be blocked by the shutter. To have a pulse moving
in the cavity the phase of all lasing modes must be synchronized.
The laser modes will arrange themselves to realize such a state.
It is a natural competition and the fittest will survive.
Shutter losses

Frequency domain analysis:

Amplitude modulation Sidebands generation
E m t    m sin  m t   m   m   0 1   cos t 
Electrical filed of each mode Amplitude of each mode is modulated
   
E m t    0 sin m t   m   sin m   t   m   sin m   t   m 
 2 2 
Sidebands are generated by the modulation
m m- m m+
Without modulation After amplitude modulation

In the case of a multimode laser
As all modes are modulated by the same frequency, the

sidebands of one mode will drive its adjacent modes, and
as a consequence, all modes will oscillate with locked phase.
From both the time domain and the frequency domain analysis it
is easy to understand why the modulation frequency must be
exactly the cavity longitudinal mode separation frequency.

Passive mode-locking:
Inserting an appropriately selected saturable absorber inside the laser
cavity. Through the mutual interaction between light, saturable absorber
and gain medium to automatically achieve mode locking.
A typical passive mode locking laser configuration:
laser medium
saturable
absorber
Mechanism of the mode-locking:
i) Interaction between saturable absorber and laser gain:
Survival takes all!
ii)Balance between the pulse shortening and pulse broadening:
Final pulse width.

Transform limited pulses

Gaussian pulses:
In our analysis we have assumed
that En=E0
The real gain line has a Gaussian

profile,which results in that the lasing
mode amplitudes have a Gaussian
distribution.
A Gaussian gain line shape function a Gaussian

mode-locked pulse intensity variation, namely a Gaussian pulse.
 1   
2

   2    t   expi t 
E t   E0     exp   
   
 2 ln 2    0
1 0

   2ln 2 2   


Intensity of the pulse:

2   
2

     
I t     exp     Gaussian intensity profile!
  0     
1
  2ln 2 2
 2 ln 2
 
Transform limited pulses:
If the product of pulse width and spectral bandwidth of a Gaussian

pulse equals 0.441, then the Gaussian pulse is called a transform
limited pulse as in this case the pulse width is purely determined
by the Fourier transformation of the pulse spectral distribution.
t p  L  0.441 For transform limited

Gaussian pulses!
tp: pulse width, L: spectral bandwidth

激光振荡器
返回
均匀加宽激光器模式竞争
返回
非均匀加宽激光器模式竞争返回
多普勒加宽增益饱和返回
兰姆凹陷返回
脉冲泵浦调 Q
返回

Optical Resonator

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Optical Resonator

Uploaded by

Copyright:

Available Formats

CHAPTER 1---- Optical Resonator

Fundamentals of Photonics 24/3/11 1

What is an optical resonator?

Fundamentals of Photonics 24/3/11 2

Fundamentals of Photonics 24/3/11 3

1.1 Brief review of Matrix optics

One can use two

: the angle above z axis

Fundamentals of Photonics 24/3/11 4

Fundamentals of Photonics 24/3/11 5

Free-Space Propagation Refraction at a Planar Boundary

Refraction at a Spherical Boundary Transmission Through a Thin Lens

Reflection from a Planar Mirror Reflection from a Spherical Mirror

Fundamentals of Photonics 24/3/11 6

A Set of Parallel Transparent Plates

Matrices of Cascaded Optical Components

Fundamentals of Photonics 24/3/11 7

Periodic Optical Systems

Difference Equation for the Ray Position

A periodic system is composed of a cascade of identical unit systems

Fundamentals of Photonics 24/3/11 8

ym  2  2bym 1  F 2 ym linear differential equations

Fundamentals of Photonics 24/3/11 9

So that, we have h 2  2bh  F 2  0 h  b  i F 2  b2

We have b  F cos  , F  b  F sin 

then h  F (cos   i sin  )  Fe  i ym  y0 F m e  im

ym  y0 F m sin(m  0 )  ymax F m sin(m  0 )

Fundamentals of Photonics 24/3/11 10

If F=1, then ym  ymax sin( m  0 )

Condition for a Harmonic Trajectory: if ym be harmonic, the cos-1b must

The bound b  1 therefore provides a condition of stability (boundedness) of

If, instead, |b| > 1,  is then imaginary and the solution is a

Fundamentals of Photonics 24/3/11 11

Condition for a Periodic Trajectory

Stable and periodic

The harmonic function is periodic in m, if it is possible to find an integer

The necessary and sufficient condition for a periodic trajectory is:

Fundamentals of Photonics 24/3/11 12

EXERCISE : A Periodic Set of Pairs of Different Lenses. Examine the trajectories of

Fundamentals of Photonics 24/3/11 13

(a) Determine the 4 x 4 ray-transfer matrix of a distance

Fundamentals of Photonics 24/3/11 14

1.2 Planar Mirror Resonators

This simple one-dimensional

A monochromatic wave of frequency v has a wavefunction as

Fundamentals of Photonics 24/3/11 15

the modes of a resonator must be the solution of Helmholtz equation

So that the general solution is standing wave:

Fundamentals of Photonics 24/3/11 16

The length of the resonator, d = q q /2, is an integer number of half

Resonator Modes as Traveling Waves

Fundamentals of Photonics 24/3/11 17

Density of Modes (1D)

The density of modes M(v), which is the number of

The number of modes in a resonator of length d within the frequency

Fundamentals of Photonics 24/3/11 18

Losses and Resonance Spectral Width

So that, the sum of the sequential reflective light with field of U1

Fundamentals of Photonics 24/3/11 19

4 d I max I max

Fundamentals of Photonics 24/3/11 20

Spectral response of Fabry-Perot Resonator

The maximum I = Imax, is achieved at the     q , q  1, 2,...,