Laser system for precision spectroscopy of the ground state hyperfine splitting in muonic hydrogen

Abstract

This thesis describes the laser system for the spectroscopy of the 1S HFS (1SF=1 1/2 -1SF=01/2 ) in muonic hydrogen (μH), pursued by the CREMA collaboration with its current experiment HyperMu. Muonic hydrogen is an atomic system formed by a muon (μ−), a negatively charged unstable lepton with a lifetime of ∼2.2 μs, and a proton (p), much like the hydrogen atom (eH). As the muon is ∼ 207 times more massive than the electron, the wavefunction of the muon has a larger overlap with that of the nucleus, making its bound-state energy levels characteristically perturbed by nuclear structure effects. A measurement of the 1S HFS to about 1ppm in μH allows the determination of the nuclear structure contribution to about 200 ppm. In a second step from this contribution, the Zemach radius and the polarizability contribution can be deduced, assuming polarizability from theory or the Zemach radius from ep scattering of eH. The experiment is to be conducted at the Paul Scherrer Institute (PSI), Switzerland, where a high-intensity continuous muon beam is available. The muons from the ΠE5 beamline are stopped within a H2 target (22K temperature, 0.5 bar pressure) to form μH. After thermalization, these atoms are excited by the laser pulses that undergo multiple reflections in a toroidal enhancement cell, placed in the H2 target, to increase the 1SF=0 1/2 - 1SF=1 1/2 transition probability. The lifetime of the 1SF=1 1/2 state is longer than that of the muons, which makes the fluorescence from the deexcitation not a viable indicator of laser excitation. An indirect detection scheme is designed based on diffusion of μH through the H2 and the production of X-rays at its arrival at the target walls. The laser-excited μH undergoing a collision with H2 molecules acquires a kinetic energy of 0.1 eV and diffuses efficiently to the target walls coated with gold. The transfer of μ− from the μH to the Au atom creates μAu in an excited state. The deexcitation of μAu generates characteristic X-rays that are detected by X-ray detectors to indicate laser excitation. The resonance curve of the HFS transition is obtained by plotting the number of μAu events versus the frequency of the laser. Simulations indicate that the laser pulses for the spectroscopy must have an energy of 3 mJ with 100MHz frequency bandwidth at the predicted transition wavelength of 6.8 μm. Moreover, as the muons arrive stochastically within the target, the laser system must be stochastically triggerable with an average repetition rate of >100 s−1. The muon lifetime constrains the maximum pulse build-up time to be ∼1 μs. Generating such high-energy pulses in the mid-infrared in such a short time and with adequate frequency control is technologically challenging. We pursue a two-stage design that begins with the generation of NIR 1030nm pulses, followed by a second stage that downconverts the 1030nm pulses into the required 6.8μm pulses via nonlinear difference frequency generation (DFG). Pulses of 1030nm of 50 ns duration and energy 30 mJ are generated in a thin-disk oscillator by the method of cavity dumping within <1 μs. These pulses are amplified to the energy of 300 mJ by a thin-disk multipass amplifier (TDA). These 1030nm pulses are used for pumping two optical parametric oscillators (OPO) operating at 2.1μm and 3.1μm wavelengths. The 2.1 μm-OPO converts the 1030nm pulse into pulses of wavelength 2.1μm and 1.9μm while the 3.1 μm-OPO converts the 1030nm pulse into pulses of wavelength 3.1μm and 1.5 μm. In a similar process, the 2.1μm and 3.1μm pulses are amplified by 1030nm pulses to energies of 25 mJ and 3 mJ, respectively, in their respective optical parametric amplifiers. The 6.8μm pulses are eventually obtained by difference frequency generation of the 2.1μm pulse and the 3.1μm pulse. The TDO and the two OPOs are injection-seeded and Pound-Drever-Hall (PDH) stabilized to ensure single-frequency operation. While the wavelength of the 1030nm and 2.1μm pulses are fixed, the wavelength of the 3.1μm pulses can be varied by changing the frequency of the seed-laser. This allows the 6.8μm pulses to be scanned across the search range of 6798nm - 6785nm of the 1S HFS in μH. This thesis deals with the development of the 2.1 μm-OPO, the 2.1 μm-OPA, the 6.8 μm- DFG stage and the frequency calibration of the 6.8μm pulses. Chapter 1 of the thesis elaborates on the motivation of the experiment along with a summary of the theoretical efforts parallel to the experiment. Details of the experimental scheme are given, focusing on the aspects that constrain the requirements of the laser system. Chapter 2 of the thesis discusses the layout of the laser system under development, designed to satisfy these requirements. A brief review of laser physics and nonlinear optics, as well as the current status of the laser system, is provided. Chapter 3 compiles the results on the 2.1μm-OPO and the implications on its variablefinesse cavity layout. The effect of injection-seeding and PDH stabilization of the cavity on the energy and stability is studied. Generation of 2.1μm pulses of energy 1 mJ of average beam qualityM2 ∼ 1.12 with 5 mJ of input 1030nm pulse energy is demonstrated. Chapter 4 describes the amplification of the 2.1μm pulses of 1 mJ energy by the 2.1 μm- OPA to ∼5 mJ while providing a beam with average M2 of 1.56, for input 1030nm pulse energy of 25 mJ from the TDO. Preliminary tests of the 2.1 μm-OPA with 1030nm pulses of energy 95 mJ are shown to amplify the 2.1μm pulses to 22 mJ. Chapter 5 focuses on the 6.8 μm-DFG stage that converts the 2.1μm pulses into 6.8μm pulses. 300 μJ of 6.8μm is generated from 6 mJ of 2.1μm and 300 μJ of 3.1 μm-seed pulses. The dependence of the nonlinear process on the orientation and temperature of the crystal, as well as the frequency of the 6.8μm beam, is studied for optimum frequency and energy control during the μH spectroscopy campaign. Chapter 6 reports on the frequency calibration of the 6.8μm pulses. By absorption spectroscopy in a H2O vapour cell, three resonances of H2O are studied for various pressures between the range 0.1 mbar and 7 mbar. From a fit to the measured transition, the centroid position and linewidth of the transitions are obtained. The obtained centroid positions were all systematically deviating by 50MHz from the HITRAN value, indicating the need to recheck the frequency calibration. From the linewidth, the upper limit of the laser bandwidth was determined to be 110 MHz. This fulfils the minimum requirement for the spectroscopy in μH. In this thesis, we have demonstrated for the first time our ability to produce 6.8μm pulses with the necessary frequency control. Energy scaling of the laser system will be needed to reach 3 mJ energy at the 6.8μm wavelength, but the observed efficiency is promising.