Lab

How the instruments work

Interactive physics from my lab work, for anyone curious about the instruments behind my research. Everything runs in your browser. The numbers come from the real systems or from published values, and each drawing says which.

Photoelectron-photoion coincidence

One molecule, one electron, one ion

A molecule absorbs ultraviolet light and loses an electron; what is left is an ion. PEPICO measures both pieces from the same molecule. The electron's energy tells you how it was bound, and the ion's flight time tells you its mass. To be sure the two belong together, the experiment keeps the rate low, so most laser shots ionise nothing, and almost every event is one electron and one ion from the same molecule; the few false pairs that remain are subtracted statistically.

An 800 nm pulse comes in over two mirrors into the optical parametric amplifier (OPA) and leaves as 309 nm ultraviolet, which a lens focuses inside the vacuum chamber, at 10−10 mbar. When a shot ionises a molecule, the electron spirals up the magnetic bottle: a strong field at the focus turns electrons flying off in every direction into one beam towards the detector, and it lands in under a microsecond. The ion waits for the extraction field, is pulled down the time-of-flight tube, and arrives about 4 µs later.

Invisible light is drawn by convention: 800 nm deep red, 309 nm violet. Time is slowed: the light's trip of a few nanoseconds takes two seconds, and after the focus one second here is one microsecond, the same for the electron and the ion. Most shots ionise nothing on purpose: with more than about one ionisation every two shots, electrons get paired with the wrong ions. Schematic, not to scale; the button in the corner pauses it.

Flight tube, fields and timing follow the spectrometer's documentation (J. P. Müller, PhD thesis, FU Berlin, 2013). Magnetic bottle: P. Kruit and F. H. Read, J. Phys. E 16, 313 (1983). Count-rate limit: V. Stert et al., Eur. Phys. J. D 5, 97 (1999).

Chirped pulse amplification

How a femtosecond pulse survives its own power

The ultraviolet light for my experiments started as pulses of a few millijoules, only about 50 fs long. You cannot amplify a femtosecond pulse straight to that energy. Its peak power would distort the pulse and could damage the crystal doing the amplifying. Chirped pulse amplification gets round this: stretch the pulse, amplify it while it is long, then compress it again. Donna Strickland and Gérard Mourou published the idea in 1985 and shared the 2018 Nobel Prize in Physics for it.

800 nm pulse: near-infrared and invisible, drawn deep redits colours spread apart, 770 to 830 nm: false colour527 nm pump: true colourlayout schematic, not to scale

  1. An oscillator sends out 80 million pulses a second near 800 nm, each carrying 8.6 nJ. Their 60 nm wide spectrum could support pulses as short as 16 fs.

  2. Gratings fan the colours out and send the red end along a shorter path. The pulse leaves about 200 ps long, red end first. Its peak power drops more than ten thousand times.

  3. The long pulse passes again and again through a titanium-sapphire crystal pumped with green light at 527 nm. Each pass adds energy, up to about 6 mJ. Spread over 200 ps, that is still only tens of megawatts.

  4. A second grating pair sends blue along the shorter path, so the tail catches up with the head. The pulse is back to 50 to 55 fs and its peak power jumps to about 80 GW. Not 16 fs: amplification narrowed the spectrum to about 30 nm.

I operated a 1 kHz titanium-sapphire system at Uni Kassel from 2023 to 2026. It ran this chain, with two amplifier stages where the drawing shows one. Every millisecond it turned 8.6 nJ from the oscillator into 4.2 to 4.5 mJ in 50 to 55 fs. That is about half a million times the energy of the seed.

The 800 nm pulse is near-infrared and invisible, so I draw it deep red. Inside the stretcher and compressor its colours are drawn apart in false colour, blue for 770 nm through red for 830 nm. The real fan at 1200 lines per mm is only about 5 degrees wide; the drawing widens it. The 527 nm pump is drawn in its true green.

Grating density and the 200 ps stretch are typical values for kilohertz titanium-sapphire amplifiers (Rudd et al., Opt. Lett. 18, 2044 (1993); Fu et al., Opt. Lett. 22, 712 (1997)). The energies, the spectra and the output duration come from the Kassel system; the 16 fs is the limit its 60 nm spectrum sets. Peak powers assume Gaussian pulses. The real pulse holds about six optical cycles at 16 fs and some 75,000 once stretched. Method: D. Strickland and G. Mourou, Opt. Commun. 56, 219 (1985).