Physics · my PhD and postdoc lab
Making light, pulsing it, catching it
Two-photon experiments need two things: a laser that squeezes its light into femtosecond pulses, and a detector that can count the faint glow that comes back. Open the laser, compare continuous and pulsed light, and try both detectors.
The physics behind one photon or two and my research on singlet oxygen.
Part of Work · What I've done
These are the instruments I spent my lab years with. In my PhD it was an 800 nm femtosecond pump laser, kept mode-locked with an optical isolator and checked with an autocorrelator. In my postdoc on Light4Lungs it was photodiodes and photomultiplier tubes, measuring singlet oxygen inside bacteria.
The laser turns green light into near-infrared pulses about 100 fs long. Bunching the light like this makes two-photon excitation about 100,000 times stronger at the same average power. To catch the faint glow that comes back, a photomultiplier turns each photon into about a million electrons.
Terms used here
- Ti:sapphire
- A sapphire crystal doped with titanium. It absorbs green light and amplifies near-infrared light from about 700 to 1000 nm.
- Cavity
- The mirrors light bounces between inside a laser. The output coupler lets a few percent out on each round trip.
- Modes
- The colours that fit a whole number of waves into the cavity, spaced 80 MHz apart here.
- Mode-locking
- Making all the modes keep the same phase, so they add up into one short pulse per round trip.
- Kerr lens
- Very intense light changes the crystal's refractive index and focuses itself. This favours pulses over continuous light.
- Femtosecond (fs)
- A millionth of a billionth of a second. In 100 fs light travels 30 µm, a third of a hair's width.
- Quantum efficiency (QE)
- The share of photons a detector turns into electrons.
- Dynode
- An electrode inside a photomultiplier. Each electron that hits it knocks out several more.
- Signal-to-noise ratio (SNR)
- How far a signal stands above the noise. Here, 3 or more counts as seen.
01 · The laser
Ti:sapphire, Kerr-lens mode-locked
Inside a Ti:sapphire laser
Step through how green light becomes a train of femtosecond pulses. The light is slowed down about 200 million times so you can follow it.
Why locking the modes makes pulses
The cavity only holds colours that fit a whole number of waves between its mirrors. Add them with random phases and you get a flickering continuous beam. Lock their phases and they all peak at the same instant: one pulse per round trip, shorter and brighter the more modes join in. Drag across the plot to bring them into step yourself.
02 · Continuous or pulsed
same average power
Same power, very different punch
Two-photon excitation needs two photons at the same molecule at the same instant. A continuous laser spreads its photons evenly; a pulsed laser bunches them. The average power is the same, so one-photon excitation is too. Hold the continuous lane to store its photons, then let go and send them as one bunch.
Pulses of 100 fs at 80 MHz: the light is on for one 125,000th of the time. Peak power = average ÷ (rate × length). Two-photon excitation follows the average of intensity squared: 1/(rate × length) times more for square pulses, 0.588 times that for the sech² shape of a mode-locked laser (Xu and Webb, 1996).
03 · Catching light
photodiode or photomultiplier
One electron, or a million
A photodiode turns a photon into one electron, then an amplifier makes it big enough to read and adds its own noise. A photomultiplier turns the photon into an electron and multiplies it a million times before any amplifier sees it. Dim the light and see which one can still tell it is on, or tap inside a detector to send it a single photon.
The shaded band is when the light is on. A detector sees the light when the band stands out from the noise around it.
Counts per millisecond. Photodiode: QE 80% (silicon at 500 nm, InGaAs at 1270 nm), amplifier noise 5,000 electrons rms. Photomultiplier: QE 25% at 500 nm and about 2% at 1270 nm, gain 4¹⁰ ≈ 1,000,000, excess noise factor 1.2, dark counts 50 per second (visible) and 100,000 per second (cooled infrared tube), counting linearly up to 10⁷ photons per second. Typical values for each type of detector.
Sources
- Ti:sapphire as a laser material: P. F. Moulton, J. Opt. Soc. Am. B 3, 125 (1986).
- Kerr-lens (self) mode-locking: D. E. Spence, P. N. Kean and W. Sibbett, Opt. Lett. 16, 42 (1991).
- Two-photon excitation with pulsed lasers and the sech² factor: C. Xu and W. W. Webb, J. Opt. Soc. Am. B 13, 481 (1996).
- Photomultiplier gain, noise and dark counts: Hamamatsu, Photomultiplier Tubes: Basics and Applications.