23
Part One · Quantum mechanics

Photoelectric effect and Compton scattering

Shine a dim blue light on a sheet of sodium and electrons leap out at once. Shine a red light a thousand times brighter and not a single electron escapes. Brightness, which is energy per second, does not decide whether electrons come out. Colour does.

In 1905 Einstein proposed that light itself comes in lumps, each carrying energy hfhf. This chapter tests that idea with the photoelectric effect and with X-rays bouncing off electrons, and then asks a sharper question: do those experiments really prove that photons exist? Surprisingly, the photoelectric effect does not. Something else does.

The photoelectric lab

Light falls on a metal plate in a vacuum tube. Choose the metal: , or . Set the light’s wavelength to 450 nm and its brightness to 4.

Brightness changes how many electrons fly out. Wavelength changes how fast. Make the light too red for the metal and nothing comes out, however bright.

Einstein’s equation

Each lump of light gives all its energy to one electron. The electron spends some of it, the work function ϕ\phi, escaping the metal, and keeps the rest as kinetic energy:

Kmax=hf−ϕK_\text{max} = hf - \phi

Below the threshold frequency f0=ϕ/hf_0 = \phi/h, no single lump has enough energy, and piling on more lumps one at a time does not help. Measure KmaxK_\text{max} by the reverse voltage needed to stop the fastest electrons, and plot it against frequency: every metal gives a straight line with the same slope, h/eh/e. Robert Millikan, who expected to disprove Einstein, measured that slope in 1916 and got Planck’s constant to within about one percent. Established

X-rays bouncing off electrons

In 1923 Arthur Compton scattered X-rays off the electrons in graphite and found the scattered X-rays had longer wavelengths, by an amount depending only on the angle. Treat each X-ray as a particle with energy hfhf and momentum h/λh/\lambda, collide it with an electron using the energy-momentum rules of Chapter 17, and you get exactly what he saw:

λ′−λ=hmec (1−cos⁡θ)\lambda' - \lambda = \frac{h}{m_ec}\,(1 - \cos\theta)

The constant h/mec=2.426h/m_ec = 2.426 pm is the electron’s Compton wavelength. Scatter at 90 ° and the wavelength grows by 2.43 pm. Light carries momentum in lumps, just as Einstein said. Established

DeeperWrite the four-momenta pp (photon in), p′p' (photon out), PP (electron at rest), P′P' (electron out). Conservation gives P′=p+P−p′P' = p + P - p'. Square both sides using p2=p′2=0p^2 = p'^2 = 0 and P2=P′2=m2c2P^2 = P'^2 = m^2c^2: 0=p⋅P−p′⋅P−p⋅p′0 = p\cdot P - p'\cdot P - p\cdot p'. With p⋅P=mc E/cp\cdot P = mc\,E/c and p⋅p′=EE′(1−cos⁡θ)/c2p\cdot p' = EE'(1-\cos\theta)/c^2, this rearranges to the Compton formula.

A surprise: this does not prove photons

In 1969 Willis Lamb and Marlan Scully showed that you can explain the photoelectric effect with light as an ordinary classical wave, provided the electrons in the metal are treated with quantum mechanics. The threshold, the linear dependence on frequency and even the absence of a delay all come out. So the photoelectric effect proves that something is quantized, but not necessarily the light. Established

What does prove it: one photon, one click

Send light at a half-silvered mirror with a detector on each output. A classical wave splits, and both detectors can click at the same moment. A single photon cannot split: it is detected at one output or the other, never both. Light from a source that emits one photon at a time should give no coincidences.

In 1986 Philippe Grangier, Gérard Roger and Alain Aspect did this with single photons from calcium atoms and found coincidences far below what any classical wave allows. Earlier, in 1977, Jeff Kimble, Mario Dagenais and Leonard Mandel had seen antibunching: photons from a single atom arriving more evenly spaced than random, impossible for a classical field. and compare the two detectors. Established

How many photons?

A green laser pointer of 5 milliwatts at 532 nm emits Pλ/hc≈1.3×1016P\lambda/hc \approx 1.3\times10^{16} photons every second. Lumps of energy are real, but in everyday light there are so many that the graininess is invisible.

· · ·

Three questions to carry with you.

Why does brighter red light not free electrons from sodium?tap
Each photon has energy hf below the work function, and an electron absorbs one photon at a time.
Why doesn’t the photoelectric effect alone prove photons exist?tap
A semiclassical model, with classical light and quantized electrons, reproduces it (Lamb and Scully, 1969).
Which experiment shows light itself is quantized?tap
Single-photon experiments: no coincidences at a beam splitter (Grangier, Roger and Aspect, 1986) and antibunching (Kimble, Dagenais and Mandel, 1977).

And one to work out. Ultraviolet light of 250 nm hits zinc (ϕ=4.33\phi = 4.33 eV). What is the maximum kinetic energy of the electrons, in eV? Use hc=1240hc = 1240 eV nm.

K =eV

Sources: A. Einstein, Ann. Phys. 17, 132 (1905). R. A. Millikan, Phys. Rev. 7, 355 (1916). A. H. Compton, Phys. Rev. 21, 483 (1923). W. E. Lamb and M. O. Scully, in Polarisation, Matière et Rayonnement (1969). H. J. Kimble, M. Dagenais and L. Mandel, Phys. Rev. Lett. 39, 691 (1977). P. Grangier, G. Roger and A. Aspect, Europhys. Lett. 1, 173 (1986). Work functions from CRC Handbook. Checked by code: λ_C = 2.4263 pm; 1.34×10¹⁶ photons/s.

Next · Chapter 24
Atomic spectra and the Bohr model

Hydrogen sends out a barcode. Decode it and you get the size of the atom.

Turn the page