Entrance examination ENS SIS 2013

Physics
Michelson interferometer

This problem is composed of three parts : a general introduction about interferences created
by two coherent point sources in a peculiar configuration (part 1), the use of a Michelson interfe-
rometer for spectroscopy (part 2) and the conception of a wavelength meter (part 3).

Generally, a light point source located in S emits, in a pulsed manner, light wavetrains which are
supposed to have the same angular velocity ω. In scalar representation, monochromatic light wave
is then characterized, at point M and time t, by the electric field E(M, t) = a cos [ωt − ϕ(M, t)],
where a is the constant field amplitude and ϕ(M, t) the dephasing, at point M and time t, with
respect to the reference point S. The complex quantity associated with E(M, t) is : E(M, t) =
a exp [i (ωt − ϕ(M, t))], where i2 = −1.

The wavetrain model assumes that the phase at the point source ϕS remains constant during
time intervals of constant duration τC between which the phase value changes randomly. The wave
emitted during this time interval called the coherence time is a "wavetrain". The wavetrain is thus
limited in time and propagates in vacuum at speed c = 3 × 108 m.s−1 . The coherence time is the
average travelling duration of the wavetrains at a given point. The wave phase at the source ϕS
takes a new random value at each new wavetrain.

In the following, the light intensity I(M ) is measured by a detector located at M which is
sensitive to the time average value of E 2 (M, t). The intensity is conventionally defined (except for
a multiplicative constant) by :

I(M ) = ⟨E(M, t)E ∗ (M, t)⟩ = ⟨|E(M, t)|2 ⟩.

1 Preamble : interference pattern created by two coherent

monochromatic point sources
Two coherent point sources S1 and S2 emit in vacuum two monochromatic waves of same
wavelength λ0 , same amplitude a0 and in phase at their respective origins S1 and S2 . These
sources, apart from a distance b, are symmetrical with respect to C (see Figure 1).
An observation screen (E) is placed perpendicularly to the line connecting S1 and S2 at a
distance D from the point C between the two sources. The perpendicular line to the screen passing
through C defines the system axis and this axis intersects the screen in B. The sources are in the
plane (BXZ). The observation point M in the coordinate system (BXY Z) is defined by ρ = BM ,
with D ≫ b and D ≫ ρ.
The intensity observed on the screen hiding one of the two sources is denoted I0 = a20 .
1. How do we realize such sources ?
2. Give the intensity I(M ) at point M as a function of the path difference δ(M ) = (S1 M ) −
(S2 M ), λ0 and I0 .
−−→ −−→
3. Express δ(M ) as a function of b and the angle θ = (CB, CM ).

1
 S1
b
C

S2

q
D

r M
B
X
(E)

Y Z

Figure 1 –

4. Deduce the intensity I(M ) at point M as a function of θ, b, λ0 and I0 ; then as a function of

ρ, D, b, λ0 and I0 .
5. What is the shape of the observed interference fringes ?
6. Define the interference order p(M ) at point M . Is it increasing or decreasing from the center
B?

2 The Michelson interferometer : a tool for spectroscopy

Figure 2 schematizes the Michelson interferometer principle. The mirrors are aligned so that
circular fringes are observed on screen (E) which is located in the focal plane of a convergent lens
(L) ("thin film" configuration of the Michelson interferometer). The lens focal distance is denoted
f ′ and its axis Oz passes trough the screen at B (OB = D).

The interferometer is composed of :

– a beam splitter (SP ) which reflects half of the light ; the origin O of the coordinate system
(Oxyz) is centered on the beam splitter which makes a fixed angle of π/4 with the axis Ox
and Oz. The dephasing introduced by the beam splitter is compensated by a compensating
plate (not shown in the figure) paralleled to the beam splitter.
– two plane mirrors (M1 ) and (M2 ), perpendicular to the figure plane, of fixed orientations
with respect to the beam splitter (SP ) with a corresponding angle π/4. Mirror (M1 ) can
be translated in the Oz direction whereas mirror (M2 ) remains fixed. The distance between
mirror (M2 ) and origin O is denoted L0 .

Only waves which have been reflected once and only once on the beam splitter will be considered.
Starting from the configuration where the image of (M2 ) with respect to the beam splitter (SP )
coincides with (M1 ), mirror (M1 ) is translated by length e which is counted positively if the mirror
moves away from the beam splitter.

2
 z

(M1)
e Image of (M2) by the
beam splitter (SP)

(SP) (M2)

LS O L0
S x

(L) O’

q f’ Focal plane
(E) B M of lens (L)
r

Figure 2 – Scheme of a Michelson interferometer in the "thin film" configuration.

2.1 Circular fringes of equal inclination

The monochromatic point source S of wavelength λ0 is placed at a finite distance LS = SO
from the beam splitter. The optical system composed of (SP ), (M1 ) and (M2 ) gives two images S1
and S2 of source S : S1 corresponds to the beams hitting (M1 ) and S2 to the beams hitting (M2 ).
The intensity observed on the screen hiding one of the two sources is denoted I0 .
1. By symmetry arguments, place on a scheme the secondary sources S1 and S2 , images of source
S, and specify their coordinates in the coordinate reference (Oxz). Deduce the expression of
−−−→
S1 S2 as a function of e.
2. The distance between points M and B on screen (E) is denoted ρ = BM . Taking into account
the condition ρ ≪ f ′ , express the path difference δ(M ) (defined here as a positive parameter)
−−→ −−−→
as a function of e and angle θ = (O′ B, O′ M ) ; then as a function of ρ, e, and f ′ . Establish,
as a function of e, the path difference ∆ = δ(B) at B for ρ = 0.
3. Express the intensity I(M ) at point M as a function of ρ, e, f ′ , λ0 and I0 . Deduce the
interference pattern projected on (E).
4. The center B of the circular fringes corresponds to a maximum of intensity. What is the
interference order p0 , supposed to be an integer, at the center of the circular fringes ? Give
the expression of the radius ρk of the k st bright circular fringe counted from the center as a
function of f ′ , p0 and its interference order pk .
5. Express ρk as a function of k and ρ1 , radius of the first circular fringe counted from the
center. Comment the interference pattern.
6. Describe (and justify) the fringes evolution when the thin film thickness value e is progres-
sively increased. Do the circular fringes seem to "enter" or "leave" the center ? Is there an
increasing or decreasing number of circular fringes visible on the screen ?

3
2.2 Interferograms : case of a monochromatic light
Mirror (M1 ) can move from e = 0 to e = Lmax (Lmax > 0). A point detector is placed at center
B of the interference pattern. This detector gives an electrical signal proportional to the detected
intensity and dependent on the path difference ∆. Increasing e induces a variation of the optical
path at B from ∆ = 0 to ∆ = ∆max and, consequently, a scrolling of the circular fringes. One calls
interferogram I(∆) the recording of the intensity evolution I as a function of ∆.

The interferometer is illuminated by a monochromatic point source of wavelength λ0 . Express

the intensity I(∆) and show on a graph the interferogram I(∆) as a function of ∆ with its charac-
teristic parameters.

2.3 Interferograms : case of a non-monochromatic light

2.3.1 Sodium doublet
We put at S a sodium lamp emitting with same intensity two monochromatic waves of wave-
lengths λ1 and λ2 close to the average wavelength λ0 = 21 (λ1 + λ2 ), with δλ = (λ2 − λ1 ) ≪ λ0 .
1. Establish the intensity I(∆) and show that its expression differs from the previous intensity
by a factor γ(∆) called degree of temporal coherence. Express this factor.
2. Define and calculate the interference fringes contrast C(∆).
3. Show on a graph I(∆) as a function of ∆ with its characteristic parameters. What is the
beating period ∆0 ? Express this period as a function of λ0 and δλ.
4. Experimentally, we measure a period ∆0 = 0, 58 mm. Knowing that the sodium doublet
is centered on λ0 = 589 nm, calculate δλ. Conclude on the interest of such interferometric
setup.

2.3.2 Source of Gaussian spectral profile

Now, we put at S a quasi-monochromatic source, characterized by a Gaussian spectral profile :
( )
dI I0 (σ − σ0 )2
= g(σ) = √ exp − (1)
dσ W π W2
where σ = λ1 , and σ0 and W ≪ σ0 are positive constants. In the following, we assume that g can
be extended to the negative values of σ.
1. Express the intensity I(∆) as a function of ∆ and the problem parameters. We will introduce
the function ge(x), Fourier transform of g(σ), defined as :
∫ ∞
ge(x) = g(σ) exp (2iπσx) dσ (2)
−∞
2. The following result is given :
∫ ∞ ( 2)
u √ ( )
exp − 2 exp (2iπux) du = a π exp −π 2 a2 x2 (3)
−∞ a
Calculate I(∆) and give the expression of the temporal coherence degree γ(∆).
3. Show on a graph the shape of I(∆) as a function of ∆, with its characteristic parameters,
knowing that δσ ≪ σ0 .
4. We admit that the interference contrast is clear enough until ∆C , path difference correspon-
ding to half of the maximum contrast, ie while |∆| ≤ ∆C . Express ∆C as a function of W ,
then as a function of λ0 and δλ.
5. ∆C is called coherence length. Explain the physical meaning of ∆C and comment the inter-
ference condition |∆| ≤ ∆C .
6. The Michelson interferometer is illuminated by a low pressure lamp of wavelength λ0 =
600 nm and line width δλ = 10−2 nm. Evaluate its coherence length. Answer to the same
question for a laser of wavelength λ0 = 600 nm and line width δλ = 10−6 nm. Comment.

4
2.4 Spectral analysis of interferograms
The interferogram spectral analysis relies on the calculation of the Fourier transform I(ω) of
intensity I(∆), defined by :
∫ ∆max ( )
ω∆
I(ω) = I(∆) cos d∆
0 c

1. Calculate I(ω) in case of an ideal monochromatic source of angular velocity ω0 .

2. Show on a graph the shape of I(ω) specifying :
– the amplitudes as a function of I0 and ∆max ;
– the base widths δωbase (see Figure 3) as a function of ∆max and c.

dwbase

Figure 3 – Base width for function sinc(x).

3. What happens to I(ω) when ∆max becomes very large ?

4. Without further calculations, deduce the shape of I(ω) for a source emitting two waves of
angular velocities ω1 and ω2 , close to the average angular velocity ω0 , and of same intensity
(ω2 > ω1 ).
In spectroscopy, a spectrometer is used to distinguish two radiation lines of angular velocities ω1
and ω2 close to each other, which depends on the resolution of the system. The common criterion to
evaluate the order of magnitude of a spectrometer resolution is the Rayleigh criterion : the smallest
measurable difference is obtained when two peaks of two different radiations are separated by half
of the base width, ie when the main maximum of one peak corresponds to the first cancelation of
the other.
5. Estimate the smallest spectral difference ∆ωR = ω2 − ω1 observable by the system as a
function of c and ∆max .
6. The spectrometer resolution R is defined by R = ∆ωω0
R
. Show that R is fixed by the number of
intensity maxima Nmax recorded by the detector during the finite course of the interferometer.
Comment.

3 Double Michelson interferometer : the wavelength meter

A wavelength meter (see Figure 4) allows a direct measurement of the wavelength of a stabi-
lized laser. It appears as a double Michelson interferometer which compares the wavelength of an
unknown stabilized laser with the known wavelength of a reference laser. The wavelength meter
requires only one beam splitter (SP ), two identical corner cube reflectors and one mirror (M ). All
the reflection angles are equal to π/4.

"Corner cubes" are reflectors which have the property of reflecting light in the same direction
as the incident light. They are made of glass of index n = 1.5 and each of the three opposite angles
are equal to 90◦ with a better than arcsecond accuracy. A light beam hitting one of the three faces

5
of the cube is reflected successively three times and therefore shifts slightly in position to stand
alongside its incident direction.
The corner cube 2 is mobile and moves in a vacuum chamber. It is heavy enough to make
negligible friction during translation. It is attached to a pulley of a stepper motor through a wire
and is guided in a stainless steel tube. The designers have tried to get as close as the free fall.

The coherence lengths of the reference laser (Helium-Neon laser emitting at 632.8 nm) and
the stabilized laser of unknown wavelength (CO2 laser) are of the order of 300 m and 30 km,
respectively.

vacuum
Mobile corner
g cube reflector 2

(SP)
S1 Fixed corner
Laser 1 A cube reflector 1
(L)
D1
B
f’
(L)
S2 f’
Laser 2
D2
(M)

Figure 4 – Scheme of the wavelength meter principle.

1. The beam coming from Laser 1 hits the beam splitter at A and the central circular fringe
of the interference pattern is detected by photodiode (D1 ). Likewise, the beam coming from
Laser 2 hits the beam splitter at B and the central circular fringe of the interference pattern
is detected by photodiode (D2 ). Compare the path differences for Laser 1 and Laser 2 at the
two interference pattern centers (D1 ) and (D2 ), respectively.
2. Laser 1 is the reference laser of wavelength λ1 = 632.8 nm. Laser 2 is a stabilized CO2 laser
of wavelength λ2 to be measured. During the fall of corner cube 2, a counter connected to
the two photodiodes evaluates to p1 = 3160556 the number of sparklings detected by (D1 ),
and meanwhile, p2 = 188679 sparklings are detected by (D2 ). Evaluate λ2 .
3. Calculate the fall hight e of the mobile corner cube. Compare to the lasers coherence lengths
and comment.
4. Determine the duration of the supposed free fall of the corner cube. Comment.
5. The fringes count is done with a one fringe accuracy. Considering that the reference wave-
length λ1 is known without uncertainty, give the relative uncertainty on the evaluation of λ2 .
Comment.