OpenWorldLab

Loading interactive view…

Diffraction

Talbot self-images

In 1836 Henry Fox Talbot looked through a magnifier at the light behind a diffraction grating and saw the grating’s lines again, sharp, some way behind it, and again further on. No lens makes these images: the grating’s diffracted waves fall back into step. Lord Rayleigh found the distance in 1881: $2d^2/\lambda$.

What you are looking at

A laser lights a Ronchi ruling, equal clear and opaque bars 0.2 mm apart. The screen stands 126.4 mm behind it, at the Talbot distance.

Zoom the screen’s card in: the bars are back, as sharp as on the ruling itself.

How the bench is set up

HeNe laser with a 1.5 mm waist, a Ronchi ruling of period 0.2 mm, and a screen 126.4 mm beyond it. Select the screen to move it.

The theory

A ruling of period dd sends light into orders at angles sin⁡θn=nλ/d\sin\theta_n = n\lambda/d. In the Fresnel approximation the nnth order falls behind the straight-through light in phase by

ϕn=πn2 λzd2\phi_n = \pi n^2\,\frac{\lambda z}{d^2}

after a distance zz. At zT=2d2/λz_T = 2d^2/\lambda every ϕn\phi_n is a whole number of turns, so the orders add up to the ruling’s own pattern: here zTz_T = 126.4 mm. At zT/2z_T/2 the odd orders are half a turn out, and the image comes back shifted by half a period. A 50% ruling has no even orders, so the strength of the bars’ period goes as cos⁡(2πz/zT)\cos(2\pi z/z_T), which vanishes at zT/4z_T/4: there only detail of half the period is left.

What to measure

Self-image at 2d²/λ, shifted by d/2 at half that

  1. Note where the bars sit at 126.4 mm, then move the screen to 63.2 mm behind the ruling: the bars come back shifted sideways by half a period, 0.1 mm.
  2. Move it to 31.6 mm: the bars’ period halves.

What the model leaves out

  • The ruling is lit by a Gaussian beam 3 mm across, not a plane wave. The orders walk apart by λz/d\lambda z/d, 0.4 mm at zTz_T, so each self-image is a little weaker than the last.
  • The screen shows one plane at a time. The whole pattern behind the ruling, the “Talbot carpet”, needs a side view the lab does not have.

Things to try

  • Change the laser to 532 nm: the self-image moves out to 150.4 mm, since zT∝1/λz_T \propto 1/\lambda.
  • Halve the period to 0.1 mm: the self-image comes four times closer, to 31.6 mm.
  • Move the screen to twice the Talbot distance, 252.8 mm: a second self-image, a little weaker.

References

  • H. F. Talbot (1836), Facts relating to optical science, No. IV, Philosophical Magazine 9, 401.
  • Lord Rayleigh (1881), On copying diffraction-gratings, and on some phenomena connected therewith, Philosophical Magazine 11, 196.

More experiments, and how the lab works, are in the laser lab’s overview.