Interference, Coherence and Diffraction
CSIR NET optics: Young's fringes, thin films, Michelson and Fabry–Pérot, temporal and spatial coherence, single slit, gratings, Airy disc, Fresnel zones.
- 19 min
On this page · 15 sections
Prerequisites
- EM Waves in Dielectrics and Conductors: Reflection, Refraction, Fresnel and Polarization: intensity \( \propto \lvert E\rvert^2 \), phase change \( \pi \) on external reflection, Stokes relations.
- Fourier transforms: a Fraunhofer pattern is the Fourier transform of the aperture, and a spectrum is the Fourier transform of the temporal coherence function.
- Small-angle approximations and sums of geometric series of complex exponentials.
The Hook: Two Sodium Lines in a Michelson Interferometer
Two-Beam Interference
Two waves of the same frequency with intensities \( I_1 \), \( I_2 \) and phase difference \( \delta \) superpose to give
$$\boxed{I = I_1 + I_2 + 2\sqrt{I_1I_2}\,\lvert\gamma_{12}\rvert\cos\delta}$$where \( \lvert\gamma_{12}\rvert \) is the degree of coherence (1 for fully coherent beams, 0 for incoherent). The visibility is
$$V = \frac{I_{\max} - I_{\min}}{I_{\max} + I_{\min}} = \frac{2\sqrt{I_1I_2}}{I_1 + I_2}\lvert\gamma_{12}\rvert$$Equal intensities and full coherence give \( V = 1 \). Unequal beams reduce it: \( I_1 = 4I_2 \) gives \( V = 0.8 \) even for perfect coherence. Interference also needs the polarisations to overlap: orthogonally polarised beams do not interfere.
Phase difference and optical path difference \( \Delta \) are related by \( \delta = 2\pi\Delta/\lambda_0 \), with \( \Delta = \sum n_il_i \) plus \( \lambda_0/2 \) for each external reflection (low to high index).
Division of Wavefront
Young’s double slit
Slits separated by \( d \), screen at distance \( D \gg d \). The path difference at height \( y \) is \( d\sin\theta \approx dy/D \).
$$\text{bright: } y_m = \frac{m\lambda D}{d}, \qquad \text{dark: } y = \left(m + \tfrac{1}{2}\right)\frac{\lambda D}{d}, \qquad \boxed{\beta = \frac{\lambda D}{d}}$$- Angular fringe width \( \lambda/d \), independent of \( D \).
- Immersing the apparatus in a medium of index \( \mu \) divides \( \beta \) by \( \mu \).
- A thin plate of thickness \( t \) and index \( n \) over one slit adds \( (n - 1)t \) to that path; the whole pattern shifts toward the covered slit by \( (n - 1)tD/d \), that is by \( (n - 1)t/\lambda \) fringes.
- With white light the central fringe is white (zero path difference for all \( \lambda \)), flanked by coloured fringes that wash out after a few orders.
Lloyd’s mirror (one source and its reflection at grazing incidence) has a dark fringe where the path difference is zero, direct evidence of the \( \pi \) phase change on reflection. The Fresnel biprism makes two virtual sources from one and gives Young-type fringes with \( d \) set by the prism angle.
Division of Amplitude: Thin Films
A film of index \( n \) and thickness \( t \), viewed in reflection at refraction angle \( r \) inside the film, gives an optical path difference \( 2nt\cos r \) between the two reflected beams. Add \( \lambda_0/2 \) if exactly one of the two reflections is external.
For a film in air (soap bubble, oil on water is different: check the indices), one reflection is external:
$$\text{reflected bright: } 2nt\cos r = \left(m + \tfrac{1}{2}\right)\lambda_0, \qquad \text{reflected dark: } 2nt\cos r = m\lambda_0$$The transmitted pattern is complementary. A very thin film (\( t \to 0 \)) looks dark in reflection: the soap film just before it bursts.
Wedges and Newton’s rings
- Air wedge of small angle \( \alpha \): straight fringes of spacing \( \lambda/2\alpha \) (in a medium of index \( n \), \( \lambda/2n\alpha \)).
- Newton’s rings between a lens of radius \( R \) and a flat plate, viewed in reflection: the air gap at radius \( r \) is \( r^2/2R \). The centre is dark, and dark rings satisfy $$r_m^2 = m\lambda R \quad\text{(air)}, \qquad r_m^2 = \frac{m\lambda R}{\mu}\quad\text{(liquid of index } \mu\text{)}$$ Radii grow as \( \sqrt m \), so rings crowd together outward. The standard measurement uses \( D_{m+p}^2 - D_m^2 = 4p\lambda R \), which removes any error from dust at the contact point.
Antireflection coatings
A single layer of index \( n_c \) on a substrate \( n_s \) (with \( 1 < n_c < n_s \)) has two external reflections, so no extra \( \pi \) appears. Destructive interference for reflection needs \( 2n_ct = \lambda_0/2 \), and equal reflected amplitudes need \( n_c = \sqrt{n_s} \):
$$\boxed{t = \frac{\lambda_0}{4n_c}, \qquad n_c = \sqrt{n_s}}$$For glass with \( n_s = 1.5 \) the ideal index is 1.22; real coatings use a slightly higher index and accept a small residual reflection.
Interferometers
Michelson
A beam splitter sends light to two mirrors; with one mirror displaced by \( d \) from the equal-path position, the path difference is \( 2d \) and the circular fringes satisfy \( 2d\cos\theta = m\lambda \). Moving a mirror by \( \Delta d \) sweeps
$$N = \frac{2\Delta d}{\lambda}$$fringes past the centre. A gas cell of length \( L \) in one arm, evacuated or filled, shifts \( N = 2L(n - 1)/\lambda \) fringes, which measures refractive indices of gases.
With a two-line source of separation \( \Delta\lambda \) and equal line strengths, the visibility varies as \( \lvert\cos(\pi\Delta\,\Delta\lambda/\lambda^2)\rvert \) with path difference \( \Delta \). It vanishes periodically, with period \( \lambda^2/\Delta\lambda \) in path difference, which is a mirror movement of
$$\Delta d = \frac{\lambda^2}{2\Delta\lambda}$$For the sodium doublet (\( \lambda \approx 589.3 \) nm, \( \Delta\lambda \approx 0.6 \) nm) this is about 0.29 mm: the observation in the hook.
Fabry–Pérot
Two parallel surfaces of reflectance \( R \) separated by \( d \) (index \( n \)) produce multiple-beam interference. The transmitted intensity is the Airy function
$$\frac{I_t}{I_{\max}} = \frac{1}{1 + F\sin^2(\delta/2)}, \qquad F = \frac{4R}{(1 - R)^2}, \qquad \delta = \frac{4\pi nd\cos\theta}{\lambda_0}$$| Quantity | Formula | For \( R = 0.9 \) |
|---|---|---|
| Coefficient of finesse | \( F = 4R/(1 - R)^2 \) | 360 |
| Finesse | \( \mathcal F = \pi\sqrt R/(1 - R) = \tfrac{\pi}{2}\sqrt F \) | about 30 |
| Free spectral range | \( \Delta\nu_{\text{FSR}} = c/2nd \) | 15 GHz for \( nd = 1 \) cm |
| Resolving power | \( \lambda/\delta\lambda \approx m\mathcal F \) | very large: \( m \) is huge |
High reflectance makes the transmission peaks very sharp: multiple-beam fringes are far narrower than the \( \cos^2 \) fringes of two-beam instruments. A laser cavity is a Fabry–Pérot resonator.
Coherence
Real sources are not perfectly monochromatic or point-like. Coherence measures how well the phase of a field at one place and time predicts it at another.
Temporal coherence
A source of bandwidth \( \Delta\nu \) (or \( \Delta\lambda \)) has a coherence time and length
$$\boxed{\tau_c \approx \frac{1}{\Delta\nu}, \qquad l_c = c\tau_c \approx \frac{\lambda^2}{\Delta\lambda}}$$Interference fringes are visible only when the path difference is less than about \( l_c \). The number of visible fringes is about \( l_c/\lambda = \lambda/\Delta\lambda \).
| Source (illustrative) | \( \Delta\lambda \) at 550 nm | \( l_c \) |
|---|---|---|
| White light (whole visible band) | about 300 nm | about 1 μm: only a few fringes |
| White light through a 10 nm filter | 10 nm | about 30 μm |
| Narrow line source, \( \Delta\lambda = 0.01 \) nm | 0.01 nm | about 3 cm |
The Wiener–Khinchin theorem makes this exact: the power spectrum is the Fourier transform of the self-coherence function \( \Gamma(\tau) = \langle E^*(t)E(t + \tau)\rangle \). A Michelson interferometer measures \( \lvert\gamma(\tau)\rvert \) as visibility against mirror position, and Fourier-transform spectroscopy inverts it to the spectrum. A Lorentzian line gives an exponentially decaying \( \lvert\gamma(\tau)\rvert \); a doublet gives a beating visibility.
Spatial coherence
An extended source illuminating two pinholes produces fringes only if the pinholes are close enough. For a source of angular size \( \theta_s \) seen from the pinholes:
$$\text{coherence width } \approx \frac{\lambda}{\theta_s}$$The van Cittert–Zernike theorem gives the exact result: the complex degree of coherence across the observation plane is the normalised Fourier transform of the source intensity. For a uniform slit source of width \( b \) at distance \( L \), visibility first vanishes at \( d = \lambda L/b \). For a uniform circular disc of angular diameter \( \theta \), it first vanishes at
$$d = \frac{1.22\lambda}{\theta}$$which is how the Michelson stellar interferometer measured stellar diameters: increase the mirror separation until the fringes disappear, then read off \( \theta \).
Diffraction
Diffraction is the Huygens–Fresnel superposition of secondary wavelets from every point of an aperture. The Fresnel number \( N_F = a^2/L\lambda \) (aperture size \( a \), distance \( L \)) decides the regime: \( N_F \ll 1 \) is Fraunhofer (far field, plane wavefronts, a lens focal plane), \( N_F \gtrsim 1 \) is Fresnel (near field). The Fraunhofer amplitude is the Fourier transform of the aperture function.
Single slit
For a slit of width \( a \):
$$\boxed{I(\theta) = I_0\left(\frac{\sin\beta}{\beta}\right)^2, \qquad \beta = \frac{\pi a\sin\theta}{\lambda}}$$- Minima at \( a\sin\theta = m\lambda \), \( m = \pm1, \pm2, \ldots \)
- Central maximum twice as wide as the others: angular full width \( 2\lambda/a \), linear width \( 2\lambda f/a \) in the focal plane of a lens of focal length \( f \).
- Secondary maxima near \( \beta \approx 1.43\pi, 2.46\pi, \ldots \) (not exactly halfway), with relative intensities about 4.7%, 1.6%, …
- Narrowing the slit widens the pattern: the uncertainty principle in optical form.
Double slit and missing orders
Two slits of width \( a \) separated by \( d \):
$$I = I_0\cos^2\gamma\left(\frac{\sin\beta}{\beta}\right)^2, \qquad \gamma = \frac{\pi d\sin\theta}{\lambda}$$Interference fringes (\( \cos^2\gamma \)) sit under the single-slit envelope. An interference maximum at \( d\sin\theta = m\lambda \) that coincides with a diffraction minimum \( a\sin\theta = p\lambda \) is missing: orders \( m = p\,(d/a) \). When \( d/a \) is an integer, the central envelope holds \( 2(d/a) - 1 \) bright fringes.
Diffraction grating
\( N \) slits of spacing \( d \):
$$I = I_0\left(\frac{\sin\beta}{\beta}\right)^2\left(\frac{\sin N\gamma}{\sin\gamma}\right)^2$$| Property | Result |
|---|---|
| Principal maxima | \( d\sin\theta = m\lambda \); intensity \( \propto N^2 \) |
| Between principal maxima | \( N - 1 \) minima, \( N - 2 \) weak secondary maxima |
| Angular half-width of a principal maximum | \( \lambda/(Nd\cos\theta) \) |
| Highest order | \( m < d/\lambda \) |
| Angular dispersion | \( d\theta/d\lambda = m/(d\cos\theta) \) |
| Resolving power (Rayleigh) | \( \boxed{\lambda/\delta\lambda = mN} \) |
Orders overlap when \( m\lambda_1 = (m + 1)\lambda_2 \): the second-order 600 nm line falls at the same angle as the third-order 400 nm line. The same \( d\sin\theta = m\lambda \) condition, applied to planes of atoms with the path difference \( 2d\sin\theta \), is Bragg’s law for X-ray diffraction.
Circular aperture and resolution
A circular aperture of diameter \( D \) gives the Airy pattern, with the first dark ring at
$$\boxed{\sin\theta_1 = 1.22\frac{\lambda}{D}}$$About 84% of the light falls inside the central (Airy) disc. Rayleigh’s criterion: two point sources are just resolved when the central maximum of one falls on the first minimum of the other, so the angular resolution of a telescope is \( 1.22\lambda/D \). For \( D = 10 \) cm at 550 nm, that is \( 6.7\times10^{-6} \) rad, about 1.4 arcseconds. A microscope resolves a minimum separation of about \( 0.61\lambda/\text{NA} \).
Fresnel zones
Divide a plane wavefront at distance \( L \) from an observation point into zones whose distances to the point increase by \( \lambda/2 \). Zone radii are
$$r_m = \sqrt{m\lambda L}$$and all zones have nearly equal area. Successive zones contribute with alternating signs, and the total unobstructed amplitude is half that of the first zone. Consequences:
- A circular hole exposing only the first zone gives four times the unobstructed intensity on axis; exposing two zones gives nearly zero.
- A zone plate (alternate zones blocked) focuses like a lens with focal length \( f = r_1^2/\lambda \), and has further foci at \( f/3, f/5, \ldots \)
- An opaque disc has a bright spot at the centre of its shadow (the Poisson–Arago spot), a prediction once offered as a reductio against the wave theory and then observed.
Babinet’s principle: complementary screens (a slit and a strip of the same width) give identical Fraunhofer patterns everywhere except at the central point, because their fields add to the unobstructed field.
Memory Tricks
Don’t Confuse
| This | vs | That |
|---|---|---|
| Single slit: \( a\sin\theta = m\lambda \) gives minima | Grating or double slit: \( d\sin\theta = m\lambda \) gives maxima | |
| Temporal coherence: bandwidth, limits path difference | Spatial coherence: source size, limits slit separation | |
| Newton’s rings in reflection: dark centre | In transmission: bright centre | |
| Lloyd’s mirror: zero-path fringe dark | Young: zero-path fringe bright | |
| Fraunhofer: far field, Fourier transform | Fresnel: near field, zones | |
| Mirror moves \( \Delta d \) | Path changes \( 2\Delta d \) in a Michelson | |
| Fringe visibility drop from unequal beams | Visibility drop from partial coherence |
Common Mistakes
Practice
All questions below are original practice questions written for this page. They are not from any CSIR NET paper.
Level 1: Foundation
- Write the fringe width in Young’s experiment.
- Estimate the coherence length of light of mean wavelength 550 nm passed through a filter of bandwidth 10 nm.
- State Rayleigh’s angular resolution for a circular aperture of diameter \( D \).
Answers
- \( \beta = \lambda D/d \).
- About 30 μm: \( \lambda^2/\Delta\lambda = (550)^2/10 = 30\,250 \) nm.
- \( \theta_{\min} = 1.22\lambda/D \).
Level 2: Part B Style
Practice Q1. In a Young’s experiment with \( \lambda = 600 \) nm, \( d = 0.5 \) mm and \( D = 1 \) m, one slit is covered by a transparent sheet of index 1.5, and the central fringe moves by 5 fringe widths. The fringe width and the sheet thickness are:
(a) 1.2 mm, 6 μm (b) 1.2 mm, 3 μm (c) 0.6 mm, 6 μm (d) 2.4 mm, 12 μm
Answer: (a)
\( \beta = \lambda D/d = 600\times10^{-9}\times1/0.5\times10^{-3} = 1.2 \) mm. A shift of 5 fringes means \( (n - 1)t = 5\lambda \), so \( t = 5\times600\ \text{nm}/0.5 = 6\ \mu \)m. Option (b) takes the extra path as \( t \) itself instead of \( (n - 1)t \).
Practice Q2. A grating is to resolve two lines at 589.0 nm and 589.6 nm in the first order. The minimum number of illuminated lines is about:
(a) 98 (b) 491 (c) 982 (d) 1964
Answer: (c) 982
\( N \geq \dfrac{\lambda}{m\,\delta\lambda} = \dfrac{589.3}{1\times0.6} \approx 982 \). Option (b) is the second-order answer.
Practice Q3. A non-reflecting single-layer coating is designed for a substrate of index 1.96 at 560 nm. Its index and minimum thickness are:
(a) 1.40, 100 nm (b) 1.40, 200 nm (c) 1.96, 71 nm (d) 1.98, 140 nm
Answer: (a) 1.40, 100 nm
\( n_c = \sqrt{1.96} = 1.40 \); \( t = \lambda_0/4n_c = 560/(4\times1.40) = 100 \) nm. Option (b) is a half-wave layer, which reflects as if uncoated.
Practice Q4. One mirror of a Michelson interferometer is moved by 0.1 mm using light of wavelength 500 nm. The number of fringes that cross the field is:
(a) 100 (b) 200 (c) 400 (d) 800
Answer: (c) 400
\( N = 2\Delta d/\lambda = 2\times10^{-4}/5\times10^{-7} = 400 \). Option (b) forgets the round trip.
Level 3: Part C Style
Practice Q5. A Michelson interferometer is illuminated by a lamp emitting two lines of equal strength at 589.0 nm and 589.6 nm. As one mirror moves away from the equal-path position, the fringe visibility is recorded. Which statements are correct?
- At equal path lengths the visibility is maximum.
- The visibility falls to zero first when the mirror has moved about 0.145 mm.
- Successive visibility minima are separated by about 0.29 mm of mirror travel.
- The data alone cannot distinguish a doublet from a single line of width 0.6 nm.
(a) 1, 2 and 3 only (b) 1 and 3 only (c) 2 and 4 only (d) 1, 2, 3 and 4
Answer: (a) 1, 2 and 3 only
Visibility \( \propto \lvert\cos(\pi\Delta\,\delta\lambda/\lambda^2)\rvert \) with path difference \( \Delta = 2d \).
Statement 1: \( \Delta = 0 \) gives \( \lvert\cos0\rvert = 1 \). True.
Statement 2: first zero when \( \Delta = \lambda^2/2\delta\lambda \), mirror travel \( d = \lambda^2/4\delta\lambda = (589.3)^2/2.4\ \text{nm} = 0.145 \) mm. True.
Statement 3: zeros recur every \( \Delta = \lambda^2/\delta\lambda \), mirror travel \( \lambda^2/2\delta\lambda = 0.289 \) mm. True.
Statement 4: a single broad line gives a visibility that decays and stays low (it is the Fourier transform of a single smooth line), while a doublet gives periodic revivals. The visibility curve distinguishes them. False.
Practice Q6. A double slit has slit width 0.1 mm and centre-to-centre separation 0.4 mm, illuminated normally by 500 nm light, with the pattern observed in the focal plane of a lens of focal length 1 m. Consider:
- The interference orders \( m = \pm4, \pm8, \ldots \) are missing.
- Seven bright fringes lie within the central diffraction maximum.
- The fringe spacing in the focal plane is 1.25 mm.
- The full width of the central diffraction maximum is 5 mm.
Which are correct?
(a) 1 and 2 only (b) 1, 2 and 3 only (c) 1, 2, 3 and 4 (d) 3 and 4 only
Answer: (b) 1, 2 and 3 only
Statement 1: \( d/a = 4 \), so orders \( 4, 8, \ldots \) coincide with diffraction minima. True.
Statement 2: the central envelope holds \( m = 0, \pm1, \pm2, \pm3 \): 7 fringes. True.
Statement 3: spacing \( \lambda f/d = 500\times10^{-9}\times1/(4\times10^{-4}) = 1.25 \) mm. True.
Statement 4: the full width of the central maximum is \( 2\lambda f/a = 2\times5\times10^{-7}\times1/10^{-4} = 10 \) mm, not 5 mm (5 mm is the half-width). False. Consistency check: the envelope spans \( \pm4 \) fringe spacings, \( 8\times1.25 = 10 \) mm.
Practice Q7. A zone plate has first-zone radius 0.5 mm. For 500 nm light, which statements are correct?
- Its principal focal length is 0.5 m.
- It also focuses at 0.25 m.
- It also focuses at about 0.167 m.
- Opening only the first zone of a circular hole at the same distance would give four times the unobstructed on-axis intensity.
(a) 1 and 3 only (b) 1, 3 and 4 only (c) 1, 2 and 4 only (d) 1, 2, 3 and 4
Answer: (b) 1, 3 and 4 only
\( f = r_1^2/\lambda = (5\times10^{-4})^2/5\times10^{-7} = 0.5 \) m. True.
Secondary foci are at \( f/3, f/5, \ldots \) (odd fractions): \( 0.167 \) m yes, \( 0.25 \) m (\( f/2 \)) no. Statement 2 false, statement 3 true.
The first zone alone gives twice the unobstructed amplitude, so four times the intensity. True.
Quick Revision Table
| Result | Formula |
|---|---|
| Two-beam | \( I = I_1 + I_2 + 2\sqrt{I_1I_2}\lvert\gamma\rvert\cos\delta \) |
| Visibility | \( 2\sqrt{I_1I_2}\lvert\gamma\rvert/(I_1 + I_2) \) |
| Young | \( \beta = \lambda D/d \); plate shift \( (n - 1)tD/d \) |
| Thin film (reflection, film in air) | bright \( 2nt\cos r = (m + \frac{1}{2})\lambda \) |
| Newton’s rings (dark, reflected) | \( r_m^2 = m\lambda R/\mu \) |
| AR coating | \( n_c = \sqrt{n_s} \), \( t = \lambda/4n_c \) |
| Michelson | \( N = 2\Delta d/\lambda \); doublet period \( \lambda^2/2\Delta\lambda \) in mirror travel |
| Fabry–Pérot | \( F = 4R/(1 - R)^2 \), \( \mathcal F = \pi\sqrt R/(1 - R) \), FSR \( c/2nd \) |
| Coherence | \( l_c \approx \lambda^2/\Delta\lambda \); width \( \approx \lambda/\theta_s \); disc \( 1.22\lambda/\theta \) |
| Single slit | \( I_0(\sin\beta/\beta)^2 \); minima \( a\sin\theta = m\lambda \) |
| Grating | \( d\sin\theta = m\lambda \); \( \lambda/\delta\lambda = mN \) |
| Airy / Rayleigh | \( 1.22\lambda/D \) |
| Fresnel zones | \( r_m = \sqrt{m\lambda L} \); zone plate \( f = r_1^2/\lambda \) |
Key Takeaways
What’s Next
- Charged Particles in Static Uniform Fields: the last official line of the unit.
- EM Waves, Reflection, Refraction and Polarization for the reflection phases and Stokes relations used here.
- Electromagnetic Theory: Quick Revision.
- The optical uncertainty relation (narrow slit, wide pattern) is the same mathematics as in Wave–Particle Duality and the Schrödinger Equation.