Wave Interference

Theory & Devices

 Interference

Interference is the superposition of two or more waves

We take two waves of complex amplitudes 𝑈1 𝑟 and 𝑈2 𝑟 . The superposition is a wave with the same
frequency, and a complex amplitude 𝑈 𝑟 = 𝑈1 + 𝑈2

The intensity of the wave is 𝐼 = 𝑈 2 = 𝑈1 + 𝑈2 2 = 𝑈1 2 + 𝑈2 2 + 𝑈1∗ 𝑈2 + 𝑈1 𝑈2∗

Expressing U as a function of the intensity for the wave components: 𝑈1 = 𝐼1 exp 𝑗𝜑1 and 𝑈2 = 𝐼2 exp 𝑗𝜑2 , we get:
𝐼 = 𝐼1 + 𝐼2 + 2 𝐼1 𝐼2 cos 𝜑 with 𝜑 = 𝜑2 − 𝜑1
If the component waves have the same frequency and a constant phase difference, the interference is coherent.
Otherwise, it is incoherent! Incoherent interference can be viewed as a linear overlap between the different
frequencies, and their induced effect is independent from each other.
𝜑
If 𝐼1 = 𝐼2 = 𝐼0 , then 𝐼 = 2𝐼0 1 + cos 𝜑 = 4𝐼0 cos 2 2 , which ranges from zero to 4𝐼0

Superposition
Phase diagram result
Interferometers

An interferometer is a device that: 1. splits the wave in two; 2. induces an optical path difference between
the two waves; 3. recombines the waves; 4. measures the intensity distribution. The two waves have to be
coherent in order to observe interference
We take two waves having an optical path difference d
𝑈1 = 𝐼0 exp −𝑖𝑘𝑧 𝑈2 = 𝐼0 exp −𝑖𝑘 𝑧 − 𝑑
2𝜋𝑑
Knowing that 𝑘𝜆 = 2𝜋, the phase difference between the two waves is 𝜑 = , and the superposition of the waves is
𝜆

𝑑
𝐼 = 2𝐼0 1 + cos 2𝜋
𝜆
When 𝑑 = 𝑚𝜆, the interference is constructive, and 𝐼 = 4𝐼0 ;
2𝑚+1 𝜆
When 𝑑 = , the interference is destructive, and 𝐼 = 0
2

The average value of intensity is 2𝐼0

The optical path difference can be induced in two ways: geometric

length or velocity modification through the refractive index n
 Interferometers with mirrors

Interferometer layouts: Mach-Zehnder, Michelson, Sagnac

The waves travel in parallel directions; The waves travel in parallel directions; The waves travel in opposite directions;
Any OPD has a phase conversion factor Any OPD has a phase conversion factor Any OPD has a phase conversion factor
of 1. of 2. of 2.
The beam splitters& mirrors have to One beamsplitter is used instead of two; One beamsplitter is used instead of two;
obey layout symmetries; Moving mirrors does not break Moving mirrors breaks symmetry;
Moving a mirror breaks symmetry; symmetry Rotating the layout with an angular
velocity induces a phase difference
2Δ𝑥
𝜆 = 𝑁 , where Δ𝑥 is the induced OPD,
and N is the number of fringes that are
displaced w.r.t. the initial one.
Interferometers with mirrors

For energy conservation purposes: 𝐼0 = 𝐼 + 𝐼′ = 𝑈1 + 𝑈2 + 𝑈1′ + 𝑈2′

Whenever 𝐼 is maximum, 𝐼 ′ has to be minimum. This corresponds to a phase difference 𝜑 = 𝜋 between the output waves
For one output wave, each component transmitted through the beamsplitter once and reflected from it once; For the other
output wave, one component is transmitted twice, and the other component is reflected twice by the beamsplitter. This
𝜋
induces a phase difference 𝜑 = 𝜋. This implies that the phase difference between the reflected and transmitted waves is 2
Diffraction-based interferometers

Before lasers, coherent light was impossible to obtain with two

independent sources!

Young’s experiment: interfering two light cones coming from

the same source!

𝑚 2𝑚+1
Interference condition: 𝑂𝑃𝐷 = 𝜆 for maxima, 𝑂𝑃𝐷 = 𝜆 for minima
2 4
𝑠
Distance between consecutive fringes: 𝑦𝑚 − 𝑦𝑚−1 = 𝑎 𝜆
Diffraction-based interferometers

Wave fronts passing through the aperture slits

Intensity distribution
Diffraction-based interferometers

Far-field interference pattern

Ray and wave model of the double-slit experiment Effect of the spacing The slit modulates the interference
on the fringe thickness pattern according to the spatial Fourier
transform of its geometry
 Newton’s rings

A flat and a spherical surface produce Newton’s rings

Interference through films

𝑚𝜆
𝑂𝑃𝐷 = 2𝑛𝑓 𝑑 cos 𝜃𝑡 cos 𝜃𝑡 = , for maxima
2𝑛𝑓 𝑑
𝑛1 sin 𝜃𝑖 = 𝑛𝑓 sin 𝜃𝑡 (2𝑚+1)𝜆
cos 𝜃𝑡 = 84𝑑 , for minima

Circular wavefronts add up to produce Haidinger rings –

equal thickness & amplitude
Interference through films – fringes of equal inclination

Fringes visible on a small portion Fringes visible on a large Deformation detection: All rays exiting
of the film. Interference pattern portion of the film. the film at a certain angle will be
will reveal local deformations Interference pattern will focused in the same point!
reveal global deformations
Interference through films – fringes of equal thickness

Interference figure for various angles. As the inclination

angle increases, the fringes get narrower and more fringes
appear on the screen. For zero wedge, interference dissapears

𝑂𝑃𝐷 = 𝑥𝛼 𝜆
Maxima 𝑚𝜆 Δ𝑥 =
𝑥𝛼 = 𝑚𝜆𝑓 = 2𝑛𝑓 𝛼
𝑛𝑓
Minima (2𝑚 + 1)𝜆 Film thickness @ various
𝑥𝛼 = 𝑚𝜆𝑓 = 2𝑚+1 𝜆
2𝑛𝑓 maxima is: 𝑑𝑚 = 2 2𝑛𝑓

A film wedge of angle 𝛼 is created

between two flat sheets of glass
Fizeau interferometers
Coherence

Coherence is the property of a wave to convey the same information it has at an initial moment after some time has passed.
Coherence is always a ratio! All waves lose coherence in time due to the fact that the environment induces randomness.
Coherence is a statistical measure! Coherence is both temporal and spatial, due to the double periodicity of the wave.
In interference, coherence can be measured as the effect that a phase difference induces on the interference pattern.
Due to the fact that the two waves have initially no phase difference, the interference pattern will have 𝐼𝑚𝑎𝑥 = 4𝐼 and
𝐼𝑚𝑖𝑛 = 0 for the duration of coherence. When a phase difference is accumulated, the interference pattern becomes
‘dirty’ with not-so-bright maxima, and dim-lit minima. Otherwise said, 𝐼𝑚𝑎𝑥 < 4𝐼 and 𝐼𝑚𝑖𝑛 = 0. For a phase difference
of 𝜋, the maxima and minima have the same intensity 𝐼𝑚𝑎𝑥 = 𝐼𝑚𝑖𝑛 = 2𝐼

Interference pattern for a) two fully The fringe visibility:

correlated waves, b) two partially
correlated waves, c) two completely 𝐼𝑚𝑎𝑥 − 𝐼𝑚𝑖𝑛
𝑉=
uncorrelated waves. In this example, 𝐼𝑚𝑎𝑥 + 𝐼𝑚𝑖𝑛
the initial phase difference is not zero.
Partial coherence

We remind the general interference law for (in intensity): 𝐼 = 𝐼1 + 𝐼2 + 2 𝐼1 𝐼2 cos 𝜑 𝜑 = 𝜑2 − 𝜑1

Due to the fact that the phase difference is induced by processes in time, we can introduce the
complex degree of coherence: 𝛾12 𝜏 , where 𝜏 is the time interval at which we evaluate.

The intensity pattern becomes: 𝐼 = 𝐼1 + 𝐼2 + 2 𝐼1 𝐼2 𝑅𝑒 𝛾12 𝜏

2 𝐼1 𝐼2
The visibility becomes: 𝑉 = 𝛾12 𝜏 for different intensity waves, and 𝑉 = 𝛾12 𝜏 for the same intensities
𝐼1 +𝐼2
Special cases: 𝛾12 𝜏 = 1 – fully coherent; 𝛾12 𝜏 < 1 – partially coherent; 𝛾12 𝜏 = 0 – fully incoherent

Interference patterns for fully and partially correlated waves

Coherence time and coherence length

The autocorrelation function of a wave at a certain time 𝜏 is equal to the Fourier transform of the power spectrum!

This property gives the relation between the coherence time and the bandwidth of the wave: Δ𝜏𝑐𝑜ℎ ⋅ Δ𝑓 ≥ 1. When we
have equality, the wave is Fourier-limited in coherence.

The coherence length is given by: 𝐿𝑐𝑜ℎ = 𝑣𝑝 Δ𝜏𝑐𝑜ℎ .

Spatial coherence

For 2D and 3D waves, spatial inhomogeneities will introduce phase differences, which will lead to a wave front
deformation in the resulting field. This can be viewed as a deformation of the fringe shape.
White light interferometry

Because white light has the broadest spectrum, it will

have the smallest coherence time/length. Events that fit
in that coherence window will produce an interference
pattern, while those who are out of the coherence
widow will produce just background light.

A white light interferometer (WLI) is a modified

Michelson interferometer that is used to determine
surface roughness by checking the variations in the
coherence time which are induced by surface variations.

The two arms are balanced so that the coherence time in

the focal plane is zero. This produces the brightest
interferogram. Any deviation from this position induces
a dimming in the contrast.
Multiple interference

Ray comes in on a thin film of thickness d, having reflection and transmission coefficients r and t on the input face and r’
and t’ on the output face.
𝑟 ′ = −𝑟 The minus indicates a phase shift of 𝜋
𝑡 ′𝑡 = 1 − 𝑟2

OPD between two adjacent rays is: 𝑂𝑃𝐷 = 2𝑛𝑓 𝑑 cos 𝜃𝑡

Multiple interference – scalar interpretation

All reflections except for the first undergo an odd number of

reflections
Phase shift @ reflection is either 𝜋 for 𝜃𝑖 < 𝜃𝑐 or zero for
𝜃𝑖 > 𝜃𝑐 for the TM component
Phase shift @ reflection is zero for the TE component.

Special case: 𝑂𝑃𝐷 = 𝑚𝜆, then the 2nd , 3rd, 4th etc waves will
be in-phase with each other. The first will have its phase
shifted by 𝜋. All beams of orders higher than 1 will add up to
cancel the first order.
1
Special case: 𝑂𝑃𝐷 = 𝑚 + 2 𝜆, the first and second rays
are in-phase, and 𝜙𝑛 = 𝜙𝑛−1 + 𝜋, for each order n. The
resulting amplitude is:
2𝑟
𝐸0𝑟 = 𝐸0 𝑟 + 𝐸0 𝑡𝑟𝑡 ′ 1 − 𝑟 2 + 𝑟 4 − ⋯ = 𝐸0 1+𝑟 2
4𝑟 2 𝐸02
Irradiance: 𝐼 = 1+𝑟 2 2 2
Multiple interference – complex interpretation

The reflected fields will take into account delays:

𝐸1𝑟 = 𝐸0 𝑟 exp 𝑖𝜔𝑡 𝐸2𝑟 = 𝐸0 𝑡𝑟 ′ 𝑡′ exp 𝑖𝜔𝑡 − 𝑖𝛿 𝐸3𝑟 = 𝐸0 𝑡𝑟 ′3 𝑡′ exp 𝑖𝜔𝑡 − 𝑖2𝛿

𝐸𝑁𝑟 = 𝐸0 𝑡𝑟 ′ 2𝑁−3
𝑡′ exp 𝑖𝜔𝑡 − 𝑖 𝑁 − 1 𝛿

Phasor diagrams

No delays
With delay 𝛿 per each reflection

2𝑟 2 1−cos 𝛿
The reflected intensity (to the right of the slab): 𝐼𝑟 = 𝐼𝑖
1+𝑟 4 −2𝑟 2 cos 𝛿
Multiple interference – complex interpretation

The transmitted fields to the left of the slab (with delays included):
2 2(𝑁−1)
𝐸1𝑡 = 𝐸0 𝑡𝑡′ exp 𝑖𝜔𝑡 𝐸2𝑡 = 𝐸0 𝑡𝑡 ′ 𝑟 ′ exp 𝑖𝜔𝑡 − 𝑖𝛿 𝐸𝑁𝑡 = 𝐸0 𝑡𝑡 ′ 𝑟 ′ exp 𝑖𝜔𝑡 − 𝑖 𝑁 − 1 𝛿

2
𝑡𝑡 ′
Irradiance: 𝐼𝑡 = 𝐼𝑖 1+𝑟 4 −2𝑟 2 cos 𝛿

𝛿
We use the identity cos 𝛿 = 1 − 2 sin2 , and obtain:
2

2𝑟 2 2 𝛿
2 sin (2) 1
𝐼𝑟 = 𝐼𝑖 1 − 𝑟 𝐼𝑡 = 𝐼𝑖 2
2𝑟 2 2 𝛿 1+
2𝑟 𝛿
sin2 2
1+ sin 1 − 𝑟2
1 − 𝑟2 2

𝐼𝑡 𝑚𝑎𝑥 = 𝐼𝑖 for 𝛿 = 2𝑚𝜋 𝐼𝑟 𝑚𝑖𝑛 = 0 for 𝛿 = 2𝑚𝜋

For 𝛿 = 2𝑚 + 1 𝜋 we have:

1 − 𝑟2 2 4𝑟 2
𝐼𝑡 𝑚𝑖𝑛 = 𝐼𝑖 𝐼𝑟 𝑚𝑎𝑥 = 𝐼𝑖
1 + 𝑟2 2 1 + 𝑟2 2
Multiple interference – complex interpretation

2𝑟 2
We introduce the coefficient of finesse: 𝐹 = 1−𝑟 2

The expressions for the reflected & transmitted fields become:

𝛿
𝐼𝑟 𝐹 sin2 2 𝐼𝑡 1
= =
𝐼𝑖 1 + 𝐹 sin2 𝛿 𝐼𝑖 1 + 𝐹 sin2 𝛿
2 2 The transmitted field ratio (Airy function)
1
In mathematics, the function 𝐴 𝛿 = 𝛿 is know as the
1+𝐹 sin2 (2)
Airy function

The reflected field ratio (1- Airy function)

Fabry-Perot cavities

The Fabry Perot cavity serves as the basic laser cavity!

Description: The cavity consists of two plane, parallel, highly reflecting surfaces at distance d from each other. The
distance is small (mm to cm) if the cavity is used for interferometry and large if it is used for laser generation. There are
several ways to tune the cavity depending on the application:

If the gap d is modified by moving one of the mirrors, then it is an interferometer

If the gap d is fixed and and the mirrors are adjusted for parallelism using a spacer (quartz), it is an etalon (roughly the
same as an interferometer)

The Fabry Perot etalon

Fabry-Perot cavities

In practice, due to the fact that the mirrors are metallic, part of the field will induce currents in the surface of the mirror, which
will change the relation between reflectance and transmittance. For pure dielectrics, we have:

𝑅 = 𝑟2 𝑇 = 𝑡𝑡′ 𝑅+𝑇 =1

When surface currents appear on the metallic surfaces, this can be interpreted as an absorptance of the EM field energy, and
the relations become:

𝑅 = 𝑟2 𝑇 = 𝑡𝑡′ 𝑅+𝑇+𝐴 = 1

The modified transmittance and reflectance ratios become:

2
𝐼𝑡 𝐴 1
= 1−
𝐼𝑖 1−𝑅 𝛿
1 + 𝐹 sin2 2

The FWHM is 𝛾 = 4 𝐹
The resolution finesse is defined as the power of
separation between adjacent peaks @ FWHM,
𝜋 𝐹
and is given by: ℱ = 2