New materials for ultrafast laser technology at 2 µm wavelength

Ultrafast lasers with high peak power are important for current-and-future applications of DOE relevance such as laser-based accelerators and tabletop-scale coherent x-ray sources. However, these applications require lasers with both high peak and high average power for broad impact—and this goal is a longstanding challenge for the laser community. One area of immediate promise is to generate coherent high harmonic soft x-rays in the “water window” spectral region for nanoscale functional imaging and spectroscopy of hard and soft materials. However, this requires high energy, high average power lasers operating in the short-to mid-wave infrared (l > 1.5 µm) with excellent spatial profile and stability. Several approaches have been used for pioneering proof-of-principle demonstrations—but none has emerged as a robust solution for routine applications. Ti:Sapphire laser-driven optical parametric amplifiers can deliver ultrashort pulses in the 1.2-1.8 µm spectral region but are complex, bulky and high maintenance. Fiber lasers at a wavelength of ~2 µm have also been used, but their limited per-pulse energy makes high harmonic up-conversion inefficient. Transition metal chalcogenide lasers operating at ~2.5 µm have been tested, but their poor spectroscopic and thermal properties limit the scalability of this technology. An attractive alternative, that has not as-yet been explored for high peak power femtosecond pulse applications, is to use Holmium-doped laser gain media. These materials can be efficiently pumped by industrial fiber lasers, with exceedingly low thermal loading of the gain medium, resulting in a scalable architecture. We have experimentally investigated the spectroscopic properties of three Ho3+-doped gain media—Ho:CaF2, Ho:YLF(q) and Ho:CALGO and found that they exhibit exceptional properties: a long excited state lifetime ideal for high energy pulses, and a spectral gain bandwidth broad enough to allow for ~200 femtosecond pulses. Numerical simulations using our spectroscopic data show that experimentally relevant performance should be achievable i.e. 10 mJ energy in 200-300 fs pulses at 1 kHz for Ho:CaF2 and Ho:CALGO given appropriately spectrally shaped seed pulses. We also see that Ho:YLF(q) allows for ~600 fs pulses, using a novel orientation of the Ho:YLF crystal and YLF has exceptional thermal properties. In this Phase I STTR project, we propose to develop a fiber based front-end that will deliver >50 nm bandwidth optical pulses at 2 µm wavelength with >100 nJ of energy in a compact and reliable package. Following this fiber front-end, we propose to send these pulses in a free-space pulse shaper that will allow on-demand spectral shaping (phase and amplitude) with low losses. Equipped with high energy, spectrally shapable pulses we will build a regenerative amplifier platform to test and compare the performance of Ho:CaF2, Ho:YLF(q) and Ho:CALGO with the aim to generate 10 mJ, 1 kHz, few hundreds of fs pulses at 2 µm wavelength. These results will guide our Phase II project to reach the 100 mJ, <40 fs duration regime. The team for this project has an excellent record for implementing and commercializing advanced high average power ultrafast lasers. The university team has developed extensive experience in mid-IR ultrafast lasers—for example implementing a robust, hands-free laser operating at 3 µm that has been used to successfully drive high harmonic generation to the carbon K-edge (283 eV) and a 10 mJ, 1 kHz, narrowband (i.e. few ps) Ho:YLF regenerative amplifier. The company has an excellent record for commercializing high-performance ultrafast lasers, with more than 100 high-energy ultrafast laser-amplifier systems in the field, and has also commercialized the coherent x-ray generation technology, with application ranging from nanometrology for semiconductors and ultrahigh resolution bioimaging.