Here is our laser system.

Our laser system is a little special, because we have a bunch of nontrivial constraints:

  • Our muons arrive at random times, and we have to trigger the laser on the (random) muon arrival. The average rate is about 500/sec.
  • Then the muon stops and makes (hopefully) an atom in the 2S state. The 2S state lives only 1 µs, so the laser light has to hit the atom at most about 1 µs,after the laser was triggered.
  • The required wavelength is between 5.5 and 6.0 µm, and the laser has to be quickly (within minutes) tunable over a (few ) hundred GHz.
  • We need about 0.2 mJ pulse energy at µm.

What we came up with is this:

Let's start on the top right:

  • The cw titanium sapphire laser (TiSa) lases around 708 nm. It is locked to a stable Fabry-Perot (FP) reference cavity with a free spectral range (FSR) of about 1.5 GHz. This cw light is injection-seeding a pulsed Ti:Sa oscillator:

Now we turn our attention to the work horse of the laser system: The cw-pumped Ytterbium-YAG (Yb:YAG) disk laser in the top left part of the picture:

  • The Yb:YAG disk laser consists of two parallel MOPA-systems (mater-oscillator, power amplifier). All 4 laser disks (2x osci, 2x ampli) are at all times pumped with a total of about 1.5 kW of cw light from two fiber-coupled pump diodes.
    The two Yb:YAG oscillators are lasing "a bit", just above threshold. When a muon is detected, we close the oscillator cavities (by pulsing a Pockels cell) and we have a quick pulse buildup: About 150 ns after the trigger we can extract up to 10 mJ of light per oscillator.
  • This light is amplified in two fancy multi-pass Yb:YAG disk amplifiers, giving a total of about 85 mJ of pulsed light (20 ns pulse length) at 1030 nm wavelength.
  • In order to be able to use the pulsed light for pumping a Ti:Sa laser we generate about 55 mJ of green light at 515 nm in three SHG (second harmonic generation) stages using BBO crystals.

Now to the central part of the figure above:

  • A small fraction of 7 mJ of the green light is used to pump the Ti:Sa crystal of the pulsed Ti:Sa oscillator. This leads to rapid amplification of the cw Ti:sa light which is already circulating inside the pulsed laser cavity (due to the injection seeding).
    We obtain about 1.5 mJ of red pulses, with a duration of about 5 ns.
  • This pulse is furher amplified in a Ti:Sa amplifier in bow-tie configuration. This Ti:Sa crystal is pumped with the majority of the 515 nm light.
  • Finally, the red pulses (15 mJ, 4-5 ns) are sent into a multi-pass Raman cell filled with 15 bar of hydrogen gas. Here, three sequential Stokes shifts convert the 708 nm light (via 1.0 µm and 1.7 µm) to the desired 6 µm wavelength. This works by virtual excitation of the H2 molecules, and subsequent de-excitation to the 1st excited vibrational level of H2. Hence, 4155 cm1 of energy are removed from the photons in each step:

The 6 µm light is finally separated from the other wavelengths in a CaF2 prisma and by Ge plates (which are transparent for 6 µm light). Then we send the light over to the muon beam (in an evacuated tube, to avoid light absorption in the water (humidity!) in air).

And now a bunch of pictures:

The disk laser. Right box: both oscillators, left bx: both amplifiers.

Detail of the oscillator.

Disk amplifiers.

And the mirror wall needed to guide the light multiple ties onto the amplfier disk.

Converting the 1030 nm light of the Yb:YAG disk laser to the 515 nm.

Pulsed Ti:Sa oscillator (left) and amplifier (right). The Raman cell can be seen in the back.

Left side of the pulsed Ti:Sa amplifier. You can see the green pump beam, and the red laser as it is sent through the crystal (on the right, outside the picture) multiple times. This picture was taken through a filter to reduce the intense green light, but no image manipulation was done!

The evacuated tube we use to get the 6 µm light from the laser hut to the muon beam.

And finally: The target!

Muons arrive from the left (through the rings), and are stopped in the target. Laser light enters the target (bottom, from the left), bounces off the off-axis paraboloid and enters the target cavity (long horizontal bars) through a tiny hole.

The 10 square-shaped things are large-area avalanche photo diodes (LAAPDs). We use these to detect x-rays from muoninc hydrogen (in particular the ones emitted after laser-excitation 2S->2P).