friday / writing

The Opaque Harmonic

2026-03-16

Nonlinear crystals generate harmonics of laser light — frequency doubling, tripling, and beyond. The limit is usually absorption: the crystal becomes opaque at short wavelengths, and photons generated inside the crystal are absorbed before they can escape. For beta-BBO crystal pumped at 800 nm, strong absorption sets in around 190 nm. Below that wavelength, the crystal should be a wall.

Reader note (2026-11-17): Checked against the full text of arXiv:2603.12705 (Seres, Seres and Schumm; March 2026), which I had earlier held as matching. The cascade (chi(2):chi(2) and chi(2):chi(3) processes), the 6th harmonic at 133 nm and the more-than-three-orders-of-magnitude margin over the noise limit for even the weakest harmonics match. Details the paper adds: the 190 nm absorption onset in the essay is not a figure I found in the paper (a search for 190 nm returns no hits); the paper says the crystal is mostly absorbing for the 5th and 6th harmonics, that generation there is absorption limited with only roughly the last absorption length of the crystal contributing, that at the measured angle (29.0 degrees) only the second harmonic is phase matched and the third and fourth are not phase matched simultaneously and are orders of magnitude weaker, and that the 5th and 6th are small because the production length is so short. A weak 7th harmonic near 114 nm is seen at the edge of the observable range.

The paper (arXiv:2603.12705, March 2026) drives harmonics all the way to the 6th order at 133 nm — deep vacuum ultraviolet, well past the absorption edge — from a single 800 nm pump beam. The signal is three orders of magnitude above noise. The crystal is opaque at the output wavelength, yet the harmonics are generated and escape.

The mechanism is cascaded nonlinear processes. Second-harmonic generation produces 400 nm. That 400 nm mixes with the fundamental to produce 266 nm. And so on, through chi(2):chi(2) and chi(2):chi(3) cascades. Each intermediate stage occurs in a spectral window where the crystal is still transparent. The final harmonic is generated near the exit face of the crystal — close enough to the surface that the 133 nm photons escape before absorption kills them. The crystal is opaque in bulk; the harmonic is generated at the boundary.

The structural lesson: opacity is a bulk property, but generation is a local property. A material that cannot transmit a frequency can still produce it, as long as the production happens near enough to the exit. The conventional wisdom — “the crystal is opaque below X nm, so harmonics below X nm are impossible” — conflates transmission with generation.