Ultra-high laser frequency stabilization using spectral holes in a cryogenically cooled crystal as a frequency reference
Final Report Abstract
Our most important results are: We successfully developed an apparatus for SH spectroscopy, consisting of a reliable narrowlinewidth laser system at a wavelength of 580 nm, a compact cryogenic setup with a fiber connection and precision temperature stabilization, optical link to our metrology laboratory, and sophisticated computer control of the complete experiment. - A detailed investigation of persistent spectral holes in Eu3+:Y2SiO5 crystals for use as a frequency reference, which included an accurate measurement of the temperature-induced frequency shift of the SHs, as well as sensitivity to magnetic field and mechanical vibrations. - Demonstration of the longest SH lifetime (50 days) and smallest long-term frequency drift < 2.3×10^-19 s^-1 fractionally, to the best of our knowledge. - Development of a new approach to SH spectroscopy based on a multi-tone laser. - Locking the laser to an ensemble of SHs for many hours without significant degradation of its short-term frequency stability, as evaluated from the measurement data at the beginning and at the end of the measurement. Achieved a laser frequency instability of about 2×10^-15 for an integration time of 1 s by locking to spectral holes. Result is limited by the optical reference used. While working on the project, we investigated several Eu3+:Y2SiO5 crystals with different Eu dopant concentration, and optimized the spectroscopic conditions to achieve the narrowest and, as well, the strongest SHs. We tried various methods for generating a multi-tone signal, including driving the AOM with short (ns) RF pulses in combination with phase modulation by means of an EOM, and a set of 12 phase locked DDS synthesizers with slightly different frequencies to control the AOM. Intensive experimental work was carried out to improve the photodetector. We performed a detailed characterization of the sensitivity and noise properties of various photodetectors, including Hamamatsu solid-state photomultipliers (MPPC), home-made ultra-low noise photodetectors with PIN diodes, and developed our own variable-gain amplifier for MPPC chips. We established broadband stabilization of the laser power using an AOM and characterized the noise properties of the laser in the probe modulation range (3 kHz). For frequency stabilization on the SHs, we compared the different methods including the Pound-Drever-Hall method with modulation sidebands at a high frequency of about 100 kHz, side-of-fringe lock and lock-in detection of frequency-modulated light. We note that the instability observed in absence of cryostat vibrations, is not far away from the best results, achieved at NIST. Our group continues the experimental work on the subject even after the DFG funding ran out, with own funding. In order to achieve frequency stability below 1×10^-15 a cryostat with much lower acceleration level at the location of the crystal. If this were available, then a further improvement of the signal-to-noise ratio in detection, by using more tones, should significantly improve the frequency stability. Crucial steps in this direction have been undertaken: design of a cryogenic passive vibration isolation stage, identification of and proposal submission for a low-vibration-noise cryostat.
Publications
- Phys. Rev. A 98, 062516 (2018)
René Oswald, Michael G. Hansen, Eugen Wiens, Alexander Yu. Nevsky, and Stephan Schiller
(See online at https://doi.org/10.1103/PhysRevA.98.062516)
