16.5 Diffraction and Polarisation
Light bends round the edges of a narrow slit and spreads into a broad central band flanked by faint fringes, with minima where $a\sin\theta=n\lambda$. Light is also a transverse wave whose vibrations can be sorted by a Polaroid ($I=I_0\cos^2\theta$) or by reflection at Brewster's angle, $\tan\theta_B=n$.
Look at a distant street lamp through the narrow gap between two fingers held almost together: the lamp smears out sideways into a streak crossed by dark lines, and the streak gets wider as you squeeze the gap narrower. Light, which we usually picture travelling in straight lines, is bending round the edges of the gap. This is diffraction, and it is one more sign that light is a wave. It also sets a limit on everything we see: no lens, eye or camera can produce a perfectly sharp point image, because the light passing through its aperture spreads out.
Now look at the same lamp's reflection in a wet road through a pair of Polaroid sunglasses and tilt your head: the glare brightens and fades. Light reflected from the road is polarised — its electric field vibrates mostly in one direction — and the sunglasses are designed to block that direction. Polarisation is possible only because light is a transverse wave. In this section we work out the single-slit diffraction pattern and the width of its central maximum, compare diffraction with interference, and then study plane-polarised light, Polaroids, Brewster's law and some of the many uses of polarised light.
Diffraction at a single slit
Every wave spreads out when it passes through an opening or round an obstacle. How much it spreads depends on the size of the opening compared with the wavelength. Sound waves (wavelengths of about a metre) bend easily round doors and buildings, so you hear people you cannot see. Light has a wavelength of about $5\times10^{-7}\ \text{m}$, so its spreading is noticeable only for openings or obstacles less than about a millimetre across — or when we look very carefully at the edges of shadows.
Diffraction is the spreading of a wave into the region behind an obstacle or an aperture — the "geometrical shadow" — together with the pattern of bright and dark bands this produces. It is appreciable when the size of the aperture or obstacle is comparable with the wavelength. In Fraunhofer diffraction, the case we study, the incident wavefronts are plane and the pattern is observed at a great distance (or in the focal plane of a lens), so the rays from the aperture to any point of the screen are effectively parallel.
Setting up. A plane wavefront of monochromatic light falls normally on a long narrow slit of width $a$. By Huygens' principle every point of the wavefront in the slit sends out secondary wavelets, all in phase as they leave the slit. A screen at a distance $D\gg a$ collects the light; consider the wavelets that travel at angle $\theta$ to the normal and meet at a point $P$ of the screen.
The centre. Straight ahead ($\theta=0$) all the wavelets travel equal distances and arrive in phase: the central point is bright — the central maximum.
The first minimum. At angle $\theta$ the wavelet from one edge of the slit travels farther than the wavelet from the other edge by $a\sin\theta$. Suppose this edge-to-edge path difference is exactly $\lambda$. Divide the slit into two halves. Each wavelet in the upper half has a partner in the lower half exactly $a/2$ below it, and the partner's path differs by $\tfrac{a}{2}\sin\theta=\tfrac{\lambda}{2}$: the two arrive in antiphase and cancel. Pair by pair, the whole slit cancels and $P$ is dark. So the first minimum is at $a\sin\theta=\lambda$.
Further minima. If $a\sin\theta=2\lambda$, divide the slit into four quarters; the first two quarters cancel each other as above, and so do the last two: dark again. In general the screen is dark whenever $a\sin\theta$ is a whole number of wavelengths.
Secondary maxima. If $a\sin\theta=\tfrac32\lambda$, divide the slit into three thirds; two of them cancel, and the third is left over and gives some light. So there are weaker maxima roughly halfway between the minima, near $a\sin\theta=\left(n+\tfrac12\right)\lambda$. Since only a fraction of the slit (one third, one fifth, …) is left uncancelled, and that fraction's own wavelets are partly out of step, the secondary maxima are much fainter than the central one: the first is only about $4.5\%$ of the central intensity, the second about $1.6\%$.
For monochromatic light of wavelength $\lambda$ falling normally on a slit of width $a$ (Fraunhofer conditions):
- minima (dark) at $\ a\sin\theta=n\lambda,\quad n=\pm1,\pm2,\pm3,\dots$ (but not $n=0$);
- the central maximum at $\theta=0$, and much weaker secondary maxima approximately at $a\sin\theta=\left(n+\tfrac12\right)\lambda$, $n=\pm1,\pm2,\dots$, with intensities falling rapidly away from the centre.
For small angles, $\sin\theta\approx\theta$ and the $n$th minimum lies at $y_n=\dfrac{n\lambda D}{a}$ on a screen at distance $D$.
The central maximum extends between the first minima on either side, $\theta=\pm\lambda/a$ (small angles). So its
$$\text{angular width}=\frac{2\lambda}{a},\qquad \text{linear width on a screen at distance }D:\quad w=\frac{2\lambda D}{a}$$(with a lens of focal length $f$ focusing the pattern, replace $D$ by $f$). Each secondary maximum is half as wide, $\lambda D/a$. The narrower the slit, the wider the central maximum; for $a\gg\lambda$ the spreading is negligible and ray optics works.
Two further points are worth remembering. Narrowing the slit widens the pattern but also lets less light through, so the central maximum becomes broader and dimmer. And red light ($\lambda$ larger) spreads more than blue: with white light the central maximum is white with reddish edges, and the secondary maxima are coloured. If the whole arrangement is immersed in a liquid of refractive index $n$, the wavelength becomes $\lambda/n$ and the pattern shrinks by the factor $n$.
In a single-slit diffraction experiment the slit is made narrower, everything else unchanged. The central maximum becomes
Diffraction compared with interference
Diffraction and interference are not two different kinds of physics: both are the superposition of coherent waves. The difference is in what is superposed. In Young's experiment we add the waves from two narrow sources; in single-slit diffraction we add the wavelets from a continuous range of points across one wavefront. That difference shows up clearly in the patterns.
| Feature | Interference (double slit) | Diffraction (single slit) |
|---|---|---|
| What is superposed | waves from two separate coherent sources | wavelets from all points of one wavefront |
| Condition for a dark fringe | $d\sin\theta=\left(n-\tfrac12\right)\lambda$ | $a\sin\theta=n\lambda$, $n\ne0$ |
| Fringe widths | all equal, $\lambda D/d$ | central $2\lambda D/a$, others $\lambda D/a$ |
| Intensity of bright fringes | all (nearly) equal | falls rapidly: $100\%$, $4.5\%$, $1.6\%$, … |
| Dark fringes | perfectly dark for equal slits | perfectly dark, but the maxima between them are faint |
Note the trap in the second row: the condition $a\sin\theta=n\lambda$ gives dark fringes in diffraction, while $d\sin\theta=n\lambda$ gives bright fringes in interference. The same-looking formula has opposite meanings, because in diffraction a path difference $\lambda$ across the slit makes the two halves cancel.
Real double slits. In any real Young's experiment each slit has a width $a$, so the light from each slit is itself diffracted. The interference fringes (spacing $\lambda D/d$) are therefore seen only inside the diffraction pattern of a single slit, and their brightness follows its outline — the diffraction envelope. Since $d\gt a$, many interference fringes fit inside the central diffraction maximum:
$$\text{number of fringe widths in the central maximum}=\frac{2\lambda D/a}{\lambda D/d}=\frac{2d}{a}.$$If $d/a$ is a whole number $m$, the interference maximum of order $m$ falls exactly on the first diffraction minimum and is missing, so the central maximum contains $2m-1$ bright fringes.
In Interactive 1 set $\lambda=600\ \text{nm}$, slit width $a=0.20\ \text{mm}$ and $D=2.0\ \text{m}$ (one slit): the central maximum is $12.0\ \text{mm}$ wide. Predict its width when you halve the slit width to $a=0.10\ \text{mm}$.
In a single-slit pattern, the condition $a\sin\theta=\lambda$ gives
Polarisation and Polaroids
Light is an electromagnetic wave: an oscillating electric field $\mathbf E$ and magnetic field $\mathbf B$, perpendicular to each other and both perpendicular to the direction of travel. A wave on a string can be made to vibrate up and down, or side to side, or in any direction across the string — and a vertical slot that the string passes through will let the up-and-down vibrations through but stop the side-to-side ones. Light behaves the same way; the direction of vibration of its electric field is called its direction of polarisation. A longitudinal wave such as sound vibrates along its direction of travel, so it has no such choice of direction and cannot be polarised. The fact that light can be polarised is direct evidence that light waves are transverse.
In unpolarised (natural) light — from the Sun, a flame or a bulb — the electric field vibrates in all directions perpendicular to the direction of travel, the direction changing randomly and rapidly; on average there is no preferred direction. In plane-polarised (linearly polarised) light the electric field vibrates along one fixed direction perpendicular to the direction of travel. The plane containing that direction and the direction of travel is the plane of vibration.
Unpolarised light can be treated as two equal, incoherent, plane-polarised components at right angles, each carrying half the intensity. In diagrams, vibrations in the plane of the page are drawn as short double arrows (↕) along the ray, and vibrations perpendicular to the page as dots (•).
A Polaroid is a thin plastic sheet containing long-chain molecules all aligned in one direction (made conducting with iodine). The component of the electric field along the molecules drives currents in them and is absorbed; the component perpendicular to them passes. That direction is the sheet's pass axis (transmission axis). A Polaroid used to produce plane-polarised light is called a polariser; one used to examine light is called an analyser.
Unpolarised light through one Polaroid. Each instantaneous vibration direction makes some angle with the pass axis; averaged over all the random directions, exactly half the intensity passes. The light that emerges is plane-polarised along the pass axis, and turning the Polaroid about the beam does not change the transmitted intensity: $I=I_0/2$.
Polarised light through a Polaroid. Now let plane-polarised light of amplitude $E_0$ meet a Polaroid whose pass axis makes an angle $\theta$ with the vibration direction. Only the component $E_0\cos\theta$ along the pass axis gets through, and intensity goes as the square of amplitude.
When plane-polarised light of intensity $I_0$ falls on a Polaroid (analyser) whose pass axis makes an angle $\theta$ with the direction of vibration, the transmitted intensity is
$$I=I_0\cos^2\theta ,$$and the transmitted light is polarised along the pass axis. Two Polaroids with parallel axes ($\theta=0$) pass all the light from the first; crossed Polaroids ($\theta=90^\circ$) pass none. Malus's law applies to plane-polarised light only; for unpolarised light falling on a single Polaroid the transmitted intensity is $I_0/2$ at every orientation.
Rotating the analyser through a full turn in a polarised beam gives two positions of maximum and two of complete darkness (extinction). That is how we test light for polarisation: unpolarised light shows no change; plane-polarised light goes completely dark twice per turn; partially polarised light (a mixture) brightens and dims but never goes completely dark.
A surprise: crossed Polaroids let no light through, but slip a third Polaroid between them at $45^\circ$, and light appears. The middle sheet re-polarises the light along its own axis, which is no longer at right angles to the last sheet. For a middle sheet at angle $\theta$ to the first, $I=\dfrac{I_0}{2}\cos^2\theta\,\cos^2(90^\circ-\theta)=\dfrac{I_0}{8}\sin^2 2\theta$, largest ($I_0/8$) at $\theta=45^\circ$.
- If the incoming light is unpolarised, the first Polaroid passes $I_0/2$, whatever its orientation.
- After that, the light is polarised along the last sheet's pass axis. At each later sheet multiply by $\cos^2$ of the angle between its axis and the previous sheet's axis (not the first sheet's).
- Any sheet at $90^\circ$ to the one before it stops everything that follows.
- Ignore absorption in the sheets unless the problem gives it; a real Polaroid passes a little less than these ideal values.
Unpolarised light of intensity $I_0$ passes through two Polaroids whose pass axes are at $45^\circ$ to each other. The emerging intensity is
Brewster's law and the uses of polarised light
Polaroids are not the only way to polarise light. When unpolarised light is reflected from a transparent surface such as glass or water, the two components are not reflected equally. The component vibrating perpendicular to the plane of incidence (parallel to the surface — the dots) is reflected more strongly than the component vibrating in the plane of incidence (the arrows). So reflected light is partly polarised — and at one particular angle it is completely polarised.
Why. The reflected wave is radiated by the electrons in the surface layer of the glass, set vibrating by the refracted wave's electric field. A vibrating charge radiates nothing along its own line of vibration. The "arrow" vibrations in the glass are perpendicular to the refracted ray. If the reflected ray happens to be perpendicular to the refracted ray, it lies exactly along those vibrations, and they cannot send any light into it. Only the "dot" component is reflected: the reflected light is plane-polarised.
When unpolarised light falls on the surface of a transparent medium of refractive index $n$ (from air) at the polarising angle or Brewster angle $\theta_B$ given by
$$\tan\theta_B=n ,$$the reflected light is completely plane-polarised, vibrating perpendicular to the plane of incidence (parallel to the reflecting surface). At this angle the reflected and refracted rays are perpendicular: $\theta_B+r=90^\circ$. The refracted light is only partially polarised. For light going from a medium of index $n_1$ to one of index $n_2$, $\tan\theta_B=n_2/n_1$.
Proof. With $\theta_B+r=90^\circ$, Snell's law gives $n=\dfrac{\sin\theta_B}{\sin r}=\dfrac{\sin\theta_B}{\sin(90^\circ-\theta_B)}=\dfrac{\sin\theta_B}{\cos\theta_B}=\tan\theta_B$. For crown glass ($n=1.5$), $\theta_B=56.3^\circ$; for water ($n=1.33$), $\theta_B=53.1^\circ$. Only a small fraction (about $15\%$ for glass) of the "dot" component is reflected at each surface, so a single reflection gives polarised light that is fairly weak; a stack of many glass plates at the Brewster angle gives a strongly polarised reflected beam, and the transmitted beam becomes more and more polarised too.
Uses of plane-polarised light and Polaroids.
- Sunglasses and camera filters. Glare from roads, water and car bonnets is reflected light near the Brewster angle, polarised horizontally. Polaroid sunglasses have vertical pass axes and cut the glare while letting through half of the unpolarised light from everything else; photographers rotate a polarising filter to remove reflections from water and glass and darken a blue sky.
- Liquid-crystal displays (LCDs). Each pixel of a calculator, watch, phone or monitor screen sits between crossed polarisers; an applied voltage changes how a liquid-crystal layer turns the plane of polarisation, switching the pixel between dark and bright.
- 3-D cinema. The two images for the two eyes are projected with different polarisations, and the viewer's polarising glasses send each image to the correct eye.
- Stress analysis (photoelasticity). A plastic model of a hook, gear or bridge seen between crossed Polaroids shows coloured bands where it is strained, revealing where a real part would be most stressed.
- Chemistry and industry. Solutions such as sugar rotate the plane of polarisation by an amount proportional to their concentration; a polarimeter (two Polaroids with the sample between) measures sugar content in the food industry and in medical tests.
- Controlling light. Two Polaroids, one rotated, form an adjustable "dimmer", used in some windows and goggles; polarised light is also used in microscopes to examine crystals and minerals.
In Interactive 2 set the refractive index to $1.50$. At what angle of incidence does the reflected light become completely plane-polarised, and what is then the angle between the reflected and refracted rays?
1. Minima, not maxima. In single-slit diffraction $a\sin\theta=n\lambda$ ($n\ne0$) gives the dark fringes; the secondary maxima are near $(n+\tfrac12)\lambda$. And $a$ is the width of the slit.
2. Width of the central maximum. It runs from the first minimum on one side to the first on the other: $2\lambda D/a$ (angular $2\lambda/a$) — twice the width of the other maxima. Using $\lambda D/a$ halves the answer.
3. The first Polaroid halves unpolarised light. $I_0/2$ at any orientation; $\cos^2\theta$ applies only from the second sheet on, with $\theta$ measured from the previous sheet's axis.
4. Brewster's angle. $\tan\theta_B=n$ (not $\sin$); reflected and refracted rays are at $90^\circ$; the reflected light is completely polarised, perpendicular to the plane of incidence, while the refracted light is only partly polarised. For light inside a medium use $\tan\theta_B=n_2/n_1$.
5. Angles from the surface. "The Sun's elevation", "the angle with the surface" and "the glancing angle" are measured from the surface, not from the normal: subtract from $90^\circ$ first.
Light falls from air onto a glass plate at the polarising angle. Which statement is correct?
Exercises
Treat all Polaroids as ideal (no absorption of the transmitted component) and take the refractive index of air as $1$ unless told otherwise. In diffraction problems $a$ is the slit width and $D$ the slit-to-screen distance; use small-angle approximations where the angles are small.
- 1 Light of wavelength $500\ \text{nm}$ falls on a slit $0.10\ \text{mm}$ wide. Find the angular position of the first minimum, in milliradians.
- 2 For part 1, find the width of the central maximum on a screen $1.5\ \text{m}$ away.
- 3 A central maximum $6.0\ \text{mm}$ wide is seen on a screen $2.0\ \text{m}$ from a slit lit by light of $600\ \text{nm}$. Find the slit width.
- 4 The first minimum for light of $650\ \text{nm}$ is at $30^\circ$. Find the slit width in $\mu\text{m}$.
- 5 For light of $600\ \text{nm}$, a slit $0.20\ \text{mm}$ wide and a screen $1.0\ \text{m}$ away, find the distance between the first and the third minima on the same side of the centre.
- 6 Find the approximate angular position of the first secondary maximum for light of $500\ \text{nm}$ and a slit $0.25\ \text{mm}$ wide, in milliradians.
- 7 A slit is slowly made narrower. Describe what happens to the width and the brightness of the central maximum. What happens if it is made very wide?
- 8 The central maximum is $12\ \text{mm}$ wide in air. Find its width when the whole apparatus is placed in water ($n=\tfrac43$).
- 1 State three differences between the fringes of a double slit and those of a single slit.
- 2 Explain why we hear sound round a corner easily but cannot see round it.
- 3 A double slit has slit width $0.080\ \text{mm}$ and slit separation $0.40\ \text{mm}$. How many interference fringe widths fit inside the central diffraction maximum?
- 4 For the double slit of part 3 with light of $600\ \text{nm}$ and a screen $1.0\ \text{m}$ away, find the interference fringe width.
- 5 For part 3, which interference orders are missing, and how many bright fringes lie within the central diffraction maximum?
- 6 For what slit width is the central maximum on a screen $1.0\ \text{m}$ away exactly as wide as the slit itself, for light of $500\ \text{nm}$?
- 7 The angular width of the central maximum is $0.12^\circ$ for light of $600\ \text{nm}$. Find the slit width.
- 1 Why does the polarisation of light show that light is a transverse wave? Can sound waves in air be polarised?
- 2 Unpolarised light of intensity $40\ \text{W/m}^2$ falls on a Polaroid. Find the transmitted intensity.
- 3 The light of part 2 then meets a second Polaroid whose axis is at $60^\circ$ to the first. Find the intensity that emerges.
- 4 At what angle between the axes of two Polaroids does $\tfrac38$ of the incident unpolarised light pass through both?
- 5 Plane-polarised light passes through an analyser. Through what angle (from the position of maximum transmission) must the analyser be turned to reduce the intensity to $10\%$ of its maximum?
- 6 Unpolarised light of intensity $64\ \text{W/m}^2$ falls on two crossed Polaroids with a third placed between them at $45^\circ$ to each. Find the emerging intensity.
- 7 In part 6 the middle sheet is turned to $30^\circ$ from the first. Find the emerging intensity.
- 8 The middle sheet of part 6 is rotated steadily. How many times per revolution of that sheet is the emerging light at its brightest? How would you use a Polaroid to tell unpolarised, plane-polarised and partially polarised light apart?
- 1 Find the Brewster angle for glass of refractive index $1.5$.
- 2 Find the Brewster angle for water ($n=1.33$), and the angle of refraction at that incidence.
- 3 The polarising angle for a transparent medium is $60^\circ$. Find its refractive index and the angle of refraction.
- 4 Light travelling in water ($n=\tfrac43$) is reflected from a submerged glass plate ($n=1.5$). Find the angle of incidence for which the reflected light is completely polarised.
- 5 The critical angle of a medium is $37^\circ$ (take $\sin37^\circ=0.6$). Find its refractive index and its polarising angle.
- 6 For a medium of index $1.6$, find the angle of refraction when light is incident at the Brewster angle.
- 7 Explain why Polaroid sunglasses cut the glare from a wet road. Why do anglers wear them?
- 8 Light reflected from a glass surface at the Brewster angle is viewed through a Polaroid. Through what angle must the Polaroid be turned from its brightest position to reduce the intensity to one quarter?
E1. JEE Main · MCQ Light of wavelength $500\ \text{nm}$ falls normally on a single slit, and the central maximum of the diffraction pattern on a screen $1.0\ \text{m}$ away is $5.0\ \text{mm}$ wide. The width of the slit is (A) $0.10\ \text{mm}$ (B) $0.20\ \text{mm}$ (C) $0.40\ \text{mm}$ (D) $0.05\ \text{mm}$
E2. JEE Main · Numerical Unpolarised light of intensity $320\ \text{W/m}^2$ passes through three ideal Polaroids in turn. The pass axis of the second is at $30^\circ$ to that of the first, and the third is crossed with the first. Find the intensity of the emerging light, in $\text{W/m}^2$. Give the answer to the nearest integer.
E3. JEE Main · MCQ Unpolarised light falls from air on a transparent medium at the polarising angle, and the refracted ray makes $30^\circ$ with the normal. The refractive index of the medium is (A) $1.5$ (B) $\sqrt2$ (C) $1/\sqrt3$ (D) $\sqrt3$
E4. JEE Main · Numerical A single slit is lit by light containing two wavelengths. The third minimum of the wavelength $\lambda$ falls at the same angle as the second minimum of light of wavelength $600\ \text{nm}$. Find $\lambda$ in nm. Give the answer to the nearest integer.
E5. JEE Main · Numerical In a double-slit experiment each slit is $0.20\ \text{mm}$ wide and the slit centres are $1.0\ \text{mm}$ apart. How many bright interference fringes are observed within the central maximum of the single-slit diffraction envelope? Give the answer to the nearest integer.
E6. JEE Main · MCQ The critical angle for light going from a certain glass into air is $45^\circ$. The polarising angle for light falling from air onto this glass is (A) $45^\circ$ (B) $60^\circ$ (C) $\tan^{-1}\sqrt2$ (D) $\tan^{-1}(1/\sqrt2)$
Research question. Can you measure the refractive index of glass, water or plastic using nothing but polarisation — by finding the angle at which reflected light vanishes through a Polaroid?
Why it matters. Polarisation measurements are everywhere in modern technology and research. Ellipsometers in semiconductor factories in Taiwan, Korea and the USA use the polarisation of reflected light to measure films only a few nanometres thick on silicon wafers. Satellites measure the polarisation of sunlight reflected by clouds and aerosols to monitor air pollution and climate. Engineers view plastic models between crossed Polaroids to find stress concentrations before building aircraft parts, and pharmacies check sugar and drug solutions with polarimeters.
What to do.
- Place a flat piece of glass (a clean picture-frame glass laid on black paper, so that only the top surface reflects) on a table under a lamp in an otherwise dark room.
- Look at the reflection of the lamp through one lens of Polaroid sunglasses (or a Polaroid sheet) held with its pass axis in the plane of incidence (vertical).
- Move your eye (or the lamp) to change the angle of reflection until the reflection is darkest. Mark the line of sight and measure the angle from the vertical with a protractor and thread. Repeat five times and average: this is $\theta_B$.
- Compute $n=\tan\theta_B$ and compare with the expected value ($\approx1.5$ for glass). Repeat with a shallow dish of water ($\approx1.33$) and with a sheet of clear acrylic.
- Rotate the Polaroid by $90^\circ$ at the same position and describe what you see.
Tools: Polaroid sheet or sunglasses, a small lamp or LED torch, clean flat glass, black paper, dish of water, acrylic sheet, protractor, thread, ruler, clamp stand.
Hand in: A table of your measured angles with their mean and spread for each material, the calculated refractive indices with uncertainties (how does an error of $1^\circ$ in $\theta_B$ affect $n$?), and a short explanation, with a ray diagram, of why the reflection disappears.
Going further. Look at a smartphone or laptop screen through the Polaroid and rotate it: what does this tell you about LCD light? Then place a clear plastic ruler or a transparent food box between the screen and your Polaroid and look for coloured stress patterns — where were the most stressed regions?
🧪 Self-check quiz
Q1. In a single-slit diffraction experiment the slit width is doubled. The angular width of the central maximum
Q2. The phenomenon that shows that light waves are transverse is
Q3. Light is incident on water ($n=1.33$) at the Brewster angle. The angle between the reflected and the refracted rays is
Q4. Unpolarised light of intensity $I_0$ passes through two Polaroids whose axes make $60^\circ$ with each other. The emerging intensity is
Q5. Compared with the fringes of Young's double slit, the fringes of a single slit