The down-conversion system for the muonic hydrogen hyperfine splitting experiment at PSI

Abstract

This thesis focuses on the development of the laser system for the HyperMu experiment, conducted by the CREMA collaboration at the Paul Scherrer Institute (PSI), Switzerland. The experiment aims to measure the ground-state hyperfine splitting (HFS) in muonic hy- drogen (μp) with 1 ppm precision using pulsed laser spectroscopy. This accuracy allows for a precise extraction of the proton structure contributions, including the Zemach radius and the proton polarizability. In the HFS experiment, a muon beam stops in a cryogenic H2 gas target, forming μp atoms. These atoms are excited from the (1S, F = 0) state to the (1S, F = 1) state by a laser pulse at 6.8 μm wavelength. A subsequent inelastic collision with an H2 molecule de- excites the μp atom back to the (1S, F = 0) state, transferring the de-excitation energy into kinetic energy. The additional kinetic energy causes the μp atoms to diffuse to gold-coated walls, where they form μAu atoms in an excited state. X-rays from μAu de-excitation are detected as a signature of a successful transition. The resonance is obtained by recording the number of μAu de-excitations versus laser frequency. The primary challenge of the HyperMu experiment lies in developing a laser system ca- pable of delivering pulses with a few mJ of energy at the target wavelength of 6.8 μm, corresponding to the hyperfine splitting transition energy. This pulse energy is dictated by the weak transition matrix element of the HFS. The pulses need to be single-frequency with a narrow bandwidth below 100 MHz due to the transition linewidth and have a pulse length of 50 ns. A tunability range of 50 GHz is required to account for the uncertainty in the theoretical prediction. Additionally, the 6.8 μm pulses should exhibit good beam quality to ensure efficient incoupling into an enhancement cavity surrounding the cryogenic muon stopping volume. This thesis centers on the nonlinear frequency down-conversion stages to generate mid- infrared light at 6.8 μm, starting from a 1030 nm thin-disk laser. The system uses two parallel branches: optical parametric oscillators (OPOs) followed by optical parametric amplifiers (OPAs). Their outputs at 2.1 μm and 3.1 μm, respectively, are combined in a differencefrequency generation (DFG) stage to produce pulses with the required properties for HFS measurement. We realized a novel OPO cavity layout with variable reflectivity achieved through the use of polarization elements. This design allowed control of the outcoupling of the OPO cavity, enabling the optimization of nonlinear conversion efficiency. A total of 1.4 mJ pulse energy at 3.1 μm with 44% conversion efficiency was achieved. Single-frequency operation was achieved by locking the OPO cavity to the injection seeding laser via a modified Pound-Drever-Hall lock with an “infinite” capture range. Injection seeding at 1532 nm was validated through optical heterodyne measurements, showing a frequency chirp below 10 MHz. Temperature tuning of the OPO provided wavelength control over a range of 10 nm. An OPA boosted the OPO pulse energy to 3.4 mJ at 3.1 μm with a pulse length of 50 ns. Long-term pulse energy RMS stability of 1.5% was demonstrated, with a beam quality of M2 = 1.3. Methane absorption spectroscopy confirmed 20 GHz tunability and a laser bandwidth below 100 MHz, meeting experimental requirements. The search for the resonance is critical to the experiment, requiring a precise theoretical prediction for the HFS transition. We provided an updated theoretical prediction for the HFS energy shift, incorporating new values from chiral perturbation theory and the data-driven approach for the proton structure contribution.