Light: Wave Model
How diffraction, interference and polarisation settled the argument between Newton's particles and Huygens' waves, and how the wave model turns fringe spacings, grating angles and polariser angles into numbers.
70 min · Module 7: The Nature of Light
Assumes you have read The Electromagnetic Spectrum.
The full teaching. Start here.
The short version
Two models. Newton said light is a stream of particles; Huygens said it is a wave. Both explained reflection and refraction, but Newton's refraction needed light to go faster in water and Huygens' needed it slower. Foucault found it slower (1850).
Diffraction is light spreading through a gap or around an edge. It is noticeable only when the gap is not much wider than the wavelength, so for light the slit must be a fraction of a millimetre. Narrower gaps and longer wavelengths spread more.
Interference. Light from two slits overlaps to give bright fringes where the path difference is a whole number of wavelengths and dark fringes at half numbers:
A diffraction grating has thousands of slits: the maxima are at the same angles, but far sharper and brighter.
Polarisation. A polariser passes only the electric field along its axis. Unpolarised light is halved; after that,
Only a transverse wave can be polarised, which ruled out particles and Huygens' longitudinal waves and matched Maxwell's electromagnetic wave.
Two models of light
By the late 1600s there were two explanations of what light is.
Isaac Newton proposed that light is a stream of tiny particles, which he called corpuscles, fired out from a source in straight lines.
- Straight line travel and sharp shadows: particles travel in straight lines unless something acts on them.
- Reflection: particles bounce elastically off a smooth surface, so the angle of incidence equals the angle of reflection.
- Refraction: as a particle reaches a denser medium such as water, the medium attracts it, increasing its speed at right angles to the surface. Its path bends towards the normal, and it travels faster in water.
- Partial reflection, some light reflecting from glass and some passing through, needed an extra idea: particles arrive in "fits" of easy reflection and easy transmission.
- Colour: particles of different masses, with red light the heaviest.
Christiaan Huygens proposed that light is a wave. Every point on a wavefront acts as a source of small secondary wavelets, and the new wavefront is the surface touching all of them. He pictured the wave as longitudinal, a compression travelling through an invisible medium, like sound.
- Reflection: wavelets from each point of the mirror build up a reflected wavefront obeying the law of reflection.
- Refraction: if the wavelets travel more slowly in the denser medium, the wavefront swings round and bends towards the normal.
- Partial reflection is natural for waves, which always partly reflect at a boundary.
- Diffraction: wavelets from the edge of an opening spread into the region behind it.
Newton's reputation, and the fact that light's diffraction is too small to notice in everyday life, kept the particle model dominant for over a century. The contest was eventually decided by experiments where the two models made different predictions.
| Observation | Newton's particles | Huygens' waves |
|---|---|---|
| Straight line travel, sharp shadows | Yes | Yes, if is very small |
| Reflection | Yes, elastic collisions | Yes, wavelets |
| Refraction | Yes, but predicts faster in water | Yes, predicts slower in water |
| Partial reflection | Only with the extra idea of fits | Yes |
| Diffraction | No | Yes |
| Interference (Young, 1801) | No | Yes |
| Speed in water (Foucault, 1850) | Wrong: measured slower | Right |
| Polarisation | Only by giving particles "sides" | No, as a longitudinal wave |
Foucault settles refraction
In 1850 Léon Foucault used a rapidly rotating mirror to compare the time light takes to travel through a tube of water and through the same length of air. The beam returning through water was deflected further, meaning it took longer: light is slower in water. That is what Huygens' wave model predicts and the opposite of Newton's. The Electromagnetic Spectrum lesson describes how Foucault later used the same apparatus to measure .
Why the two models disagree about water
AnswersNewton and Huygens both explained refraction towards the normal. How can they predict opposite speeds?
In Newton's model, the water pulls particles towards it as they cross the surface. That pull adds to the part of the velocity at right angles to the surface and leaves the part along the surface unchanged. The particle turns towards the normal because it is going faster into the water.
In Huygens' model, one end of a wavefront enters the water first. If the wavelets there travel more slowly, that end falls behind the part still in air, and the whole wavefront pivots. Its direction of travel turns towards the normal because the wave is going slower in the water.
The bending is the same, so refraction alone could not decide between them. The speed in water could.
Worked example4 marks
Judging Newton and Huygens
Analyse how the observations of refraction, diffraction and polarisation supported or challenged the models of light proposed by Newton and Huygens.
Refraction
Both models explain it, so the evidence has to be the speed they each need.
Both models explain the bending of light towards the normal in water. Newton's requires light to travel faster in water, Huygens' slower. Foucault's measurement in 1850 showed light is slower in water, supporting Huygens and contradicting Newton.
Diffraction
Only one model predicts it, which makes it strong evidence.
Huygens' wavelets spread into the region behind a gap, so the wave model predicts diffraction. Particles travelling in straight lines do not. Young's interference fringes (1801) and the bright Arago spot at the centre of a circular shadow (1819) are diffraction effects, supporting the wave model.
Polarisation
An analyse answer should show where the preferred model fails too.
Light passing a polariser behaves differently depending on its orientation. Huygens' longitudinal wave oscillates along its direction of travel, so it has no orientation to select, and Newton's particles needed "sides" added purely to fit. Neither model explained it. It needed light to be a transverse wave.
Answer
Refraction and diffraction both supported Huygens' wave model over Newton's particle model, most decisively through Foucault's measurement that light is slower in water. Polarisation challenged both, and was only explained when the wave model was changed to a transverse wave.
Is that answer sensible?
Each paragraph names an observation, what each model said about it, and which model the evidence supports. That is the structure markers look for in analyse and evaluate answers.
Common mistake
Writing that Newton's model could not explain refraction.
Why it happens: Assuming the model that lost explained nothing.
Newton's model did explain refraction. What failed was its prediction that light is faster in water. Saying so is more accurate and earns more marks, because it shows why the evidence decided between the models.
CheckpointAnswer before reading on.
Both Newton's particle model and Huygens' wave model could explain why light bends towards the normal as it enters water. Why did Foucault's 1850 experiment decide between them?
Hint 1What did each model say about the speed of light in water?
Hint 2Newton explained the bending by water attracting the particles.
Hint 3Which way did the measurement come out?
CheckpointAnswer before reading on.
Huygens modelled light as a longitudinal wave, like sound. Which property of light could this model not explain?
Hint 1Which of these depends on the direction in which the wave oscillates?
Hint 2A longitudinal wave oscillates along its direction of travel.
Hint 3Can a sound wave be blocked by turning a filter?
Exam question
Harder · about 10 min
6 marks
Evaluate the models of light proposed by Newton and Huygens, with reference to experimental evidence.
Hint 1List the observations: straight line travel, reflection, refraction, partial reflection, diffraction, interference, polarisation.
Hint 2For each, which model explained it? Where did they make different predictions that could be tested?
Hint 3Evaluate means a judgement. Which model did the evidence support, and what did neither model get right?
Where students lose marks on this one
Describing both models without a judgement.
Why it happens: Treating evaluate as describe.
Say which model the evidence supports, why, and what it still could not explain.
Saying Newton's model could not explain refraction.
Why it happens: Assuming the losing model explained nothing.
It did explain refraction, but with a prediction about speed that turned out wrong. That is a stronger point to make.
Written for this site.
Diffraction
Diffraction is the spreading of a wave as it passes through a gap or around the edge of an obstacle. Every point in the gap acts as a source of wavelets, as Huygens said, and they spread out behind it.
How much a wave spreads depends on the gap compared with the wavelength. A gap many wavelengths wide lets the wave through almost straight, with only its edges spreading. A gap about one wavelength wide turns it into an almost circular wave spreading in every direction. Sound, with wavelengths around a metre, spreads through a doorway. Light, with a wavelength of about nm, passes through the same doorway two million wavelengths wide and casts a sharp shadow, which is why light seems to travel only in straight lines.
Diffraction is also what makes Young's experiment possible: each slit spreads light into the region behind it, so the light from two slits overlaps. Young's pattern comes first below, treating each slit as a point that spreads light evenly. The pattern from one slit of real width follows, and then the two are combined into the pattern a real double slit gives.
Interference: Young's double slit
In 1801 Thomas Young shone light through two narrow slits close together. Particles would give two bright lines behind the slits. Instead the screen showed a row of evenly spaced bright and dark fringes.
Each slit diffracts the light, so the two spreading waves overlap. Because both slits are lit by the same wavefront, the waves leave them in step: they are coherent. At any point on the screen, what you see depends on the path difference, how much further the light from one slit has travelled than the light from the other.
A plane wave arrives from the left at a barrier with two narrow slits. Each slit diffracts the wave into circular wavefronts, and where they overlap they interfere. Shaded bands are crests. On the screen at the right, the time averaged brightness shows bright fringes where the waves always arrive in step and dark fringes where they always cancel.
- Brightness on the screen
The simulation shows the waves from each slit in shades of colour. Where crest meets crest the colour is strong, and those lines of reinforcement run out to the bright fringes on the screen. Between them the waves cancel and the screen stays dark. Widen the slit separation and more lines of reinforcement appear, closer together; lengthen the wavelength and they spread apart.
Path difference in the double slit
1Two coherent sources
The two slits are lit by the same wavefront, so they act as two sources oscillating exactly in step. That is what makes a steady pattern possible. Two separate lamps change phase at random millions of times a second, so their fringes would shift too fast to see.
2Path difference
For a distant screen the two rays to P are nearly parallel. Drop a perpendicular from the upper slit to the lower ray: the short leg left over is the extra distance the lower ray travels. From the small triangle, it is .
3Bright fringe
At a bright fringe the path difference is a whole number of wavelengths, so crest meets crest and the waves reinforce: constructive interference. , with at the centre, then , , outwards. P, at , is two wavelengths further from the lower slit.
4Dark fringe
At a dark fringe the path difference is a whole number of wavelengths plus a half, . Crest meets trough and the waves cancel: destructive interference. Both slits light this spot, yet it is dark. Particles cannot do that; waves can.
Bright fringes from two slits or a gratingon the NESA formulae sheet
- Symbols
- distance between neighbouring slits (for a grating, 1 divided by the lines per metre)m
- angle from the straight through direction to the bright fringe°
- order of the fringe: 0 at the centre, then 1, 2, 3 on each side1
- wavelength of the lightm
- Valid when
- Coherent light of one wavelength passing through two slits or a grating, with the screen far away compared with the slit spacing, so rays from neighbouring slits to one point are nearly parallel.
- Not valid when
- Finding the dark bands of a single slit: that uses the slit width, a sin θ = mλ, and gives minima rather than maxima. It also gives no answer when mλ is larger than d, because no order exists there.
- Rearranged
- for \lambda: for \theta: for m_{\max}:
- Where it turns up
- Finding a wavelength from the angle of a bright fringe
- Finding the angle of each order from a grating
- Counting how many orders a grating produces
- Where marks go missing
- Using lines per millimetre as d instead of converting to the spacing in metres
- Using the angle between the two first order fringes, which is 2θ
- Leaving the calculator in radians
The fringes are usually only millimetres apart on a screen metres away, so the angles are tiny and there is a simpler form in terms of positions on the screen.
Derivation
Fringe spacing
- Starts from
- the condition for bright fringes, using small angles
- Ends at
- Holds only if
- The screen distance L is much larger than the slit separation d
- The fringes are close to the centre, so θ is small
Fringe spacing
3 steps
- 1
Position on the screen
Convert the angle into a distance we can measure with a ruler.
The th bright fringe is at a distance from the centre, with .
- 2
Use the small angle
For small angles, the sine and the tangent are the same.
, so becomes
- 3
Spacing between neighbours
Neighbouring fringes differ by one in m.
The fringes are evenly spaced, as Young observed.
Fringe spacing for a double slitnot on the formulae sheet, learn it
- Symbols
- distance between neighbouring bright fringes on the screenm
- distance from the slits to the screenm
- wavelength of the lightm
- distance between the slitsm
- Valid when
- Small angles, when the screen is much further away than the fringes are wide, so sin θ ≈ tan θ = y/L. This is the usual double slit arrangement, with fringes a few millimetres apart on a screen metres away.
- Not valid when
- Large angles, such as the orders of a diffraction grating, which can be tens of degrees from the centre. Use d sin θ = mλ and then y = L tan θ instead.
- Rearranged
- for \lambda: for d:
- Where it turns up
- Finding the wavelength of a laser from measured fringe spacing
- Predicting how the pattern changes when the slits, screen or colour change
- Where marks go missing
- Mixing millimetres and metres in the same calculation
- Measuring from the centre to the fifth fringe and treating it as one spacing
- Applying it to a grating at large angles
- Brightness on the screen
- Fringe spacing, Δy = Lλ/d
- 4.80 mm
- Angle to the first bright fringe, sin θ = λ/d
- 0.138°
- Path difference at the first dark fringe
- λ/2 = 300 nm
- Bright fringes across the 60 mm shown
- 13
Every slider is a normal range input, so the arrow keys move it one step and Home and End jump to the extremes.
This graph is the ideal double slit pattern: it treats each slit as so narrow that it sends light equally in every direction, so every fringe is equally bright and the fringes carry on evenly across the screen. That is the model describes, and the one HSC questions use. Real slits have a width, and the fringes of a real double slit fade away from the centre. The reason is single slit diffraction, which comes next.
Worked example3 marks
Measuring a wavelength with a double slit
In an investigation, light from a lamp with a yellow filter passes through two slits mm apart onto a screen m away. The distance across six bright fringes, from the centre of the first to the centre of the sixth, is mm. Calculate the wavelength of the light.
The fringe spacing
Measuring across several fringes and dividing reduces the uncertainty of each.
Six fringes have five spacings between them:
Rearrange for the wavelength
Every length in metres.
Answer
m, about nm.
Is that answer sensible?
nm is yellow orange, as expected for a yellow filter, and within a few nanometres of the sodium lines at nm. Two significant figures in the answer matches the slit separation.
CheckpointAnswer before reading on.
Light of wavelength nm passes through two slits. At point P on the screen, the light from one slit has travelled nm further than the light from the other. What is seen at P?
Hint 1Write the path difference as a number of wavelengths.
Hint 2Whole numbers of wavelengths arrive in step; what about half numbers?
Hint 3
CheckpointAnswer before reading on.
A laser shines on two slits mm apart, and the fringes are viewed on a screen m away. A student measures mm from the centre of the first bright fringe to the centre of the eleventh. Calculate the wavelength of the laser light, in nanometres.
Give it to 2 significant figures.
Hint 1How many fringe spacings are there between the first and eleventh bright fringes?
Hint 2Use with every length in metres.
Hint 3
Exam question
Standard · about 4 min
3 marks
Explain how the pattern produced in Young's double slit experiment supports the wave model of light.
Hint 1What does the pattern look like, and what would particles produce?
Hint 2Why are some places dark even though both slits are open?
Hint 3Relate bright and dark fringes to the path difference.
Where students lose marks on this one
Saying the dark fringes are shadows of the slits.
Why it happens: Thinking in terms of rays.
The dark fringes are lit by both slits; the two waves cancel there. Covering one slit makes those places brighter.
Written for this site.
Single slit diffraction
One slit of width , with nothing else, also gives a pattern of bright and dark bands. The graph below is the single slit diffraction pattern: a broad, bright central maximum, twice the width of the others, with much fainter secondary maxima beside it.
- Brightness
- Colour
- orange
- Slit width ÷ wavelength
- 13.3
- First dark band, sin θ = λ/a
- ±4.3°
- Central bright band
- 8.6° wide
Every slider is a normal range input, so the arrow keys move it one step and Home and End jump to the extremes.
A single slit gives a broad central bright band with much fainter bands beside it. Light from different parts of the slit arrives at the screen with different path lengths. At the angle where the top half of the slit and the bottom half are exactly out of step, they cancel, giving the first dark band:
Use the graph to check the two trends: halve the slit width and the central band doubles in width; change red to violet and it shrinks.
Worked example3 marks
Predicting how far light spreads
A red laser of wavelength nm shines through a slit mm wide onto a screen m away. Estimate the width of the central bright band, then predict how it changes if the slit is narrowed to mm, and if a green laser of nm is used with the original slit.
The angle to the first dark band
The central band runs from the dark band on one side to the one on the other.
The angle is small, so on the screen the first dark band is mm from the centre, and the central band is about mm wide.
Narrower slit
The spreading depends on λ divided by a.
Halving doubles , so the central band doubles to about mm. The light spreads more because the gap is closer to the wavelength.
Shorter wavelength
The same ratio, changed by the wavelength instead.
, so the band shrinks to mm.
Answer
The central bright band is about mm wide. Narrowing the slit to mm doubles it to about mm; green light makes it narrower, about mm.
Is that answer sensible?
The trend is the whole of the qualitative dot point: narrower gap or longer wavelength, more spreading. A slit mm wide is about wavelengths, which is why the spreading is small but still visible across a room.
In the investigation, shine a laser through slits of different widths, or through a gap you can adjust, onto a wall a few metres away in a darkened room. Also try the diffraction of light around a thin wire or a hair, which gives the same pattern as a slit of the same width. Never look along the beam or at its reflection from a shiny surface.
Common mistake
Saying that a wider slit gives a wider central band.
Why it happens: Particle thinking: a wider opening lets through a wider beam.
For waves, the opposite holds. A wider gap lets the wave through with less spreading, so the central band gets narrower. The band is widest when the gap is about one wavelength.
CheckpointAnswer before reading on.
Red laser light passes through a single narrow slit onto a distant screen. Which change makes the central bright band on the screen wider?
Hint 1Diffraction is the spreading of a wave after it passes through a gap.
Hint 2How does the amount of spreading depend on the gap compared with the wavelength?
Hint 3The first dark band is where .
CheckpointAnswer before reading on.
You can hear someone talking in the next room through an open doorway, even though you cannot see them. A student says this proves light is not a wave. What is the best reply?
Hint 1Both sound and light diffract. What decides how much?
Hint 2Compare each wavelength with the width of a doorway, about 1 m.
Hint 3Sound in speech is about 1 m long; visible light is about m.
Exam question
Standard · about 4 min
3 marks
Outline an investigation to show qualitatively how the diffraction of light depends on the width of a slit and on the wavelength of the light. Include what you would expect to observe.
Hint 1What light source gives a single, bright wavelength?
Hint 2Which variable do you change first, and what do you keep the same?
Hint 3What happens to the width of the central band as each variable changes?
Where students lose marks on this one
Using a white light globe without a filter.
Why it happens: Any light source seems fine for a qualitative investigation.
Each wavelength spreads a different amount, so white light gives blurred, coloured bands. A laser gives one wavelength and a clear pattern.
Written for this site.
Fraunhofer diffraction: the real double slit pattern
A real double slit is two single slits side by side. Each slit spreads its light according to the single slit pattern, and the light from the two slits then interferes according to the double slit pattern. The brightness at any angle is the product of the two: the evenly spaced double slit fringes, scaled by the single slit pattern of either slit. This combined pattern, seen on a distant screen or at the focus of a lens, is called Fraunhofer diffraction.
- Two slit interference alone
- Single slit diffraction envelope
- Fraunhofer pattern: the product
- Fringe spacing, Δy = Lλ/d
- 4.80 mm
- Angle to the first bright fringe
- 0.138°
- Path difference at the first dark fringe
- λ/2 = 300 nm
- Envelope minimum, a sin θ = λ
- 24.0 mm from the centre
- Bright fringes in the central envelope
- 9
- Missing orders
- m = ±5, ±10 and so on, since d/a = 5
Every slider is a normal range input, so the arrow keys move it one step and Home and End jump to the extremes.
Read the graph in three layers. The faint curve is the ideal two slit interference, every fringe equally bright. The dashed curve is the single slit envelope, set by the slit width . The shaded curve, the product, is what appears on the screen: the fringes keep the spacing but their brightness follows the envelope. Make the slits narrower and the envelope widens, letting more fringes through; move them further apart and more fringes fit inside it. Where a fringe lands on a dark band of the envelope, it vanishes: a missing order.
Common mistake
Using the envelope's dark bands to find the fringe spacing.
Why it happens: Both patterns appear together, so the two sets of dark bands are easy to confuse.
The closely spaced fringes come from the slit separation , through . The broad envelope comes from the slit width , through for its first dark band. The slits are always further apart than they are wide, so the fringes are always narrower than the envelope.
Diffraction gratings
A diffraction grating is a sheet with thousands of equally spaced slits, usually given as lines per millimetre. Every pair of neighbouring slits obeys the same condition as the double slit, so the bright maxima are at exactly the same angles:
The difference is what happens between the maxima. With two slits, moving slightly off a maximum leaves the waves only slightly out of step, so the fringes are broad. With a thousand slits, the same small move puts the first slit and the five hundredth slit exactly out of step, and so on in pairs, so the light cancels almost everywhere except very near the exact angles. The maxima become narrow, bright and widely separated, which is why gratings rather than double slits are used to measure wavelengths.
- Brightness
- Line spacing d
- 3.33 × 10⁻⁶ m
- Order 1, sin θ = 1λ/d
- 10.4°
- Order 2, sin θ = 2λ/d
- 21.1°
- Order 3, sin θ = 3λ/d
- 32.7°
- Highest order, m < d/λ
- 5
- Half width of each maximum
- about 1.72°
Every slider is a normal range input, so the arrow keys move it one step and Home and End jump to the extremes.
Increase the number of slits on the graph and watch the maxima sharpen without moving. Increase the lines per millimetre and the orders spread further apart, until the higher orders disappear: cannot exceed , so the highest order is the largest whole number with .
With white light, the central maximum is white, because every wavelength has zero path difference there. Every other order is spread into a spectrum, with violet nearest the centre and red furthest out, since the angle increases with wavelength. A prism does the opposite.
Worked example4 marks
A grating and its orders
A helium neon laser, nm, shines on a grating with lines per millimetre. Find the angles of the first and second order maxima, the position of the first order on a screen m away, and the highest order produced.
The line spacing
d is the gap between neighbouring lines, in metres.
The angles
d sin θ = mλ for m = 1 and m = 2.
Position on the screen
At grating angles the small angle formula is not reliable, so use the tangent.
The highest order
sin θ cannot be more than 1.
so the highest order is , at , or .
Answer
The first order is at , m from the centre, and the second at . Orders up to appear on each side, maxima in all.
Is that answer sensible?
The second order angle is roughly, but not exactly, twice the first: it is that doubles, not . A double slit would put its fringes millimetres apart; the grating's tiny spacing spreads them tens of centimetres, which is what makes precise wavelength measurements possible.
Common mistake
Substituting the lines per millimetre directly as d.
Why it happens: The number of lines is given, not their spacing.
is the distance from one line to the next. lines per millimetre means lines share m, so m. A quick check: for a grating is between about one and ten micrometres.
CheckpointAnswer before reading on.
A green laser of wavelength nm shines on a diffraction grating with lines per millimetre. Calculate the angle of the second order maximum from the central maximum, in degrees.
Give it to 3 significant figures.
Hint 1The line spacing is one millimetre divided by the number of lines.
Hint 2 m.
Hint 3 with .
CheckpointAnswer before reading on.
Red light of wavelength nm shines on a grating with lines per millimetre. How many bright maxima appear in total, counting the central one, on a screen wide enough to catch every order?
Hint 1 can never be more than .
Hint 2So . Find first.
Hint 3Count the orders on both sides, then add the central maximum.
CheckpointAnswer before reading on.
A double slit is replaced by a grating with the same slit spacing but thousands of slits. What happens to the bright maxima?
Hint 1The maxima are where . Has changed?
Hint 2With many slits, what happens a little away from an exact maximum?
Hint 3Try it on the grating graph by increasing the number of slits.
CheckpointAnswer before reading on.
White light passes through a diffraction grating. In the first order spectrum, which colour is found furthest from the central maximum?
Hint 1The first order is at .
Hint 2Which colour has the longest wavelength?
Hint 3A prism does the opposite, but for a different reason.
The Arago spot
In 1818 Augustin Fresnel entered a wave theory of diffraction in a competition of the French Academy of Sciences. One judge, Siméon Poisson, a supporter of the particle model, worked through Fresnel's equations and found what he thought was an absurd prediction: in the shadow of a small circular disc, waves diffracting round the edge all travel the same distance to the centre, so they should arrive in step and make a bright spot in the middle of the shadow.
François Arago did the experiment in 1819 and found the spot. A prediction meant to discredit the wave model became some of its most convincing evidence, and Fresnel won the prize.
Polarisation
Interference and diffraction showed that light is a wave, but not which kind. Sound, a longitudinal wave, diffracts and interferes too. Polarisation settled the question.
A polariser, such as a sheet of Polaroid, is made of long chain molecules lined up in one direction. Electrons can move freely along the chains, so the component of the light's electric field along the chains drives currents and is absorbed. The component at right angles to the chains passes through. That direction is the polariser's transmission axis.
Ordinary light from a lamp or the Sun is unpolarised: its electric field oscillates in every direction across the beam, changing at random. A polariser passes only the component along its axis, so the light leaving it is plane polarised, oscillating in one plane only, and it has half the original intensity.
3D model
Unpolarised light from a lamp passes a polariser, which lets through only the electric field component along its transmission axis. A second polariser, the analyser, turned through an angle θ, passes only the component along its own axis, so the transmitted intensity falls as cos²θ.
Unpolarised light from a lamp passes through a polariser, which passes only the electric field along its axis, and then an analyser that can be turned through an angle θ. The transmitted beam dims as θ increases and disappears at 90°. A third polariser can be added between them at 45°.
Malus' law
Turn a second polariser, the analyser, so its axis is at an angle to the plane of polarisation. The field arriving has amplitude . Only its component along the analyser's axis passes:
The intensity of a wave is proportional to the square of its amplitude, so
Malus' lawon the NESA formulae sheet
- Symbols
- intensity transmitted by the analyserW m⁻²
- intensity of the plane polarised light reaching the analyserW m⁻²
- angle between the plane of polarisation and the transmission axis of the analyser°
- Valid when
- Light that is already plane polarised, arriving at an ideal polariser. With several polarisers, apply it once at each, using the angle between neighbouring axes.
- Not valid when
- Unpolarised light reaching the first polariser. There is no single angle, and an ideal polariser passes exactly half of the intensity.
- Rearranged
- for \theta: for I_{\max}:
- Where it turns up
- Intensity through two or three polarisers
- Finding the angle that transmits a given fraction of light
- Explaining why crossed polarisers block light
- Where marks go missing
- Forgetting to square the cosine
- Applying cos²θ to unpolarised light instead of halving it
- Using the angle from the first polariser for every later polariser
- I = Imax cos²θ
- After the first polariser, Imax
- 400 W m⁻²
- cos θ
- 0.866
- cos²θ
- 0.750
- Leaving the analyser, I
- 300 W m⁻²
Every slider is a normal range input, so the arrow keys move it one step and Home and End jump to the extremes.
At all of the polarised light passes; at , crossed polarisers, none does. Halfway, at , , so half gets through. The same average, , is why a single polariser passes half of unpolarised light whatever its angle.
Worked example4 marks
Three polarisers
Unpolarised light of intensity passes through three polarisers. The first has a vertical axis, the second is at to the vertical, and the third is horizontal. Calculate the intensity after each polariser, and compare with what happens if the middle one is removed.
First polariser
Unpolarised light has no single angle, so it is halved.
Second polariser
Malus' law with the angle between the first and second axes.
Third polariser
The light is now polarised at 30°, so the angle to the horizontal axis is 60°.
Without the middle polariser
The first and last are crossed.
Answer
The intensity falls to , then , then . Without the middle polariser no light passes. Adding a polariser between crossed polarisers lets light through.
Is that answer sensible?
Each polariser does not just block light; it re-polarises it along its own axis, so each angle is measured from the polariser before it. Set the 3D model to a middle polariser at and the analyser at to see the light return.
To investigate Malus' law, fix a polariser in front of a lamp, place an analyser on a protractor after it, and measure the transmitted intensity with a light sensor as the analyser is turned in steps of . Subtract the reading with the lamp off, to remove background light. Plotting against should give a straight line through the origin with gradient .
Other ways light is polarised
- Reflection: light reflected from water, glass or a road is partly polarised parallel to the surface, and completely polarised at one angle of incidence, Brewster's angle. Polarising sunglasses have vertical axes to block this glare.
- Scattering: sunlight scattered by the air is partly polarised, most strongly at from the Sun. Bees navigate by it, and photographers use polarising filters to darken the sky.
- Stress in plastics rotates the plane of polarisation by different amounts for different colours, so a strained ruler between crossed polarisers shows coloured bands. Geologists identify minerals in thin rock slices the same way.
Why polarisation matters
Only a transverse wave can be polarised
AnswersWhy does blocking light by turning a filter prove anything about what light is?
A longitudinal wave, like sound, oscillates along its direction of travel. Viewed along the beam it looks the same from every side, so there is no direction for a filter to select, and turning the filter could not change anything.
Light does change: turn a polariser and the light through it dims and disappears. So light must oscillate in some particular direction across its path. It is a transverse wave.
That ruled out Huygens' longitudinal waves, as well as Newton's particles, which could be made to fit only by giving them "sides" with no other reason. Young and Fresnel reworked the wave model as a transverse wave between 1817 and 1821. Maxwell's theory then explained what oscillates: the electric and magnetic fields, at right angles to the direction of travel. Malus' law is exactly what a transverse field predicts.
Say which kind of wave
When asked about the significance of polarisation, "it shows light is a wave" earns little. Interference and diffraction already showed that. The mark earning point is that polarisation shows light is a transverse wave, because only transverse waves have a direction of oscillation across the beam, ruling out the longitudinal model and matching Maxwell's electromagnetic waves.
Common mistake
Using cos²θ at the first polariser when unpolarised light arrives.
Why it happens: Applying the formula to every polariser in a row.
Malus' law needs light that is already polarised at a known angle to the axis. Unpolarised light has no such angle, and an ideal polariser always passes exactly half of it. Halve first, then use at each polariser after that.
CheckpointAnswer before reading on.
Unpolarised light from a lamp falls on a single ideal polariser whose axis is at to the vertical. What fraction of the intensity passes through?
Hint 1In unpolarised light, which direction does the electric field point?
Hint 2Does the angle of the polariser to the vertical matter, then?
Hint 3The average value of over all angles is .
CheckpointAnswer before reading on.
Unpolarised light of intensity passes through a polariser and then an analyser whose axis is at to the polariser's. Calculate the intensity leaving the analyser, in .
Give it to 3 significant figures.
Hint 1What fraction of unpolarised light passes the first polariser?
Hint 2That is for the analyser.
Hint 3
CheckpointAnswer before reading on.
Two polarisers are crossed, with axes at and , so no light passes. A third polariser with its axis at is placed between them. If unpolarised light of intensity enters the first polariser, what intensity leaves the last, in ?
Give it to 2 significant figures.
Hint 1The first polariser halves the unpolarised light.
Hint 2Apply Malus' law at the middle polariser, using the angle between the first and middle axes.
Hint 3Then again at the last, using the angle between the middle and last axes.
CheckpointAnswer before reading on.
Plane polarised light falls on an analyser. At what angle between the plane of polarisation and the analyser's axis is of the intensity transmitted? Give your answer in degrees.
Give it to 3 significant figures.
Hint 1
Hint 2Take the square root first.
Hint 3
Evaluatetypically 6 to 8 marks
- Demands
- Judge worth against criteria, weighing strengths against weaknesses, and conclude.
- Shape
- A position stated early, defended with evidence, with the counter-case acknowledged and answered.
- Loses marks
- Listing advantages and disadvantages with no weighting, so nothing is actually evaluated.
Through a marker’s eyes
3 marks
Explain why the polarisation of light was significant in developing a model for light. (3 marks)
The attempt
1 out of 3
Polarisation is when light only vibrates in one plane. It showed that light is a wave and not particles, which supports Huygens.
What the marker sees
The definition is worth a mark. But Huygens' model could not explain polarisation, so it does not support him, and saying light is a wave was already shown by interference. The response misses the key point: polarisation shows the wave is transverse.
The same answer, fixed
3 out of 3
A polariser transmits light depending on its orientation, and two crossed polarisers block light completely, so light must oscillate in a particular direction across its path. Only a transverse wave can do this; a longitudinal wave looks the same from every side.
This ruled out Huygens' longitudinal wave model and Newton's particle model, neither of which could explain polarisation, and led to light being modelled as a transverse wave. Maxwell's theory later explained it as transverse oscillating electric and magnetic fields, and Malus' law, , follows from that transverse field.
Exam question
Harder · about 4 min
3 marks
Evaluate the significance of polarisation in developing a model of light.
Hint 1What does polarisation show that interference and diffraction do not?
Hint 2Could Huygens' longitudinal wave be polarised? Could Newton's particles?
Hint 3Which later theory predicts exactly this kind of wave?
Where students lose marks on this one
Saying polarisation proves light is a wave, and stopping there.
Why it happens: Treating all wave evidence as the same.
The point of polarisation is the type of wave: it must be transverse.
Written for this site.
Exam question
Harder · about 6 min
4 marks
Describe an investigation to test Malus' law, including how you would analyse the results to show that they support it.
Hint 1You need polarised light, an analyser you can turn to known angles, and a way to measure intensity.
Hint 2What graph would give a straight line if ?
Hint 3What would you control, and how would you deal with background light?
Where students lose marks on this one
Plotting $I$ against $\theta$ and saying it is a straight line.
Why it happens: Expecting every relationship to be linear in the variable measured.
Intensity depends on , so plot that on the horizontal axis to get a straight line.
Written for this site.
Why the single slit has dark bands
Split a slit of width into a top half and a bottom half. For each point in the top half there is a point exactly below it in the bottom half. At the angle where the path difference between these partners is , every pair cancels, and the screen is dark. That path difference is , so the first dark band is at . Split the slit into four, six or more strips and the same argument gives further dark bands at . Note the reversal: the same form of equation gives dark bands for one slit of width and bright fringes for slits a distance apart.
Missing orders
Because the double slit pattern sits inside a single slit envelope, a bright fringe that falls exactly on a dark band of the envelope is missing. That happens when is a whole number: with slits mm apart and mm wide, every fifth fringe is missing, and nine bright fringes fit in the central envelope. Try it on the Fraunhofer graph: set the separation to a whole number of slit widths and watch those orders disappear.
Resolving power
A grating's maxima sharpen as the number of slits grows: each one's width in angle is roughly . Two wavelengths can be told apart when their maxima are further apart than that, which works out as . The two sodium lines at and nm need of about : lines in first order, or in second. The gratings in astronomical spectrographs have tens of thousands of lines, which is how they measure the tiny Doppler shifts in the Electromagnetic Spectrum lesson.
Brewster's angle
When light reflects from glass or water at the angle where the reflected and refracted rays are at to each other, the reflected light is completely polarised parallel to the surface. That angle satisfies , where is the refractive index: about for water and for glass. Lasers use windows set at Brewster's angle to let one polarisation through without any reflection loss.