Fiber-based tunable frequency conversion of ultrafast pulses is simple, efficient

While ytterbium-doped fiber (Yb:fiber) lasers produce ultrafast pulses with a center wavelength around 1 μm, slightly longer-wavelength high-energy ultrafast pulses are needed for a number of applications, including scientific (high-harmonic generation) and medical (new forms of optical-coherence tomography) uses. The traditional source of wavelength-tunable ultrafast pulses, the optical parametric amplifier (OPA), frequency-downconverts 1 μm ultrafast pulses and allows continuous tuning from about 1.3 to 4.5 μm, although it cannot reach the important 1.0–1.3 μm range without additional frequency-doubling. In addition, OPAs are large, complex, and expensive.

An international team of researchers has come up with a far-simpler (and, as a result, lower in cost) approach to creating ultrafast pulses in the 1–1.7 μm region: sending 1 μm pump pulses down a long segment of nitrogen-filled hollow-core fiber, causing extreme Raman red-shifting of the light.¹ The approach has the additional benefit of shortening the pulses to about 20 fs from the input pump pulse’s 200 fs. The team members hail from Technische Universität Wien (TUWien; Vienna, Austria), Institut National de la Recherche Scientifique (INRS; Varennes, QC, Canada), the Center for Physical Sciences and Technology (Vilnius, Lithuania), the University of Electronic Science and Technology of China (Chengdu, China), Few-Cycle (Montreal, Quebec, Canada), M. V. Lomonosov Moscow State University (Moscow, Russia), the Russian Quantum Center (Skolkovo, Russia), and Texas A&M University (College Station, TX).

Usually, hollow-core fibers are filled with a monatomic gas such as argon in order to symmetrically broaden the spectrum of the laser and then recompress it into a much-shorter optical pulse. The research team discovered that by using a molecular gas such as nitrogen, spectral broadening was still possible, but in an unexpected asymmetric manner. Once the beam is broadened toward longer infrared (IR) wavelengths, the researchers filter the output spectrum to keep only the band of interest. With this approach, energy is transferred into the near-IR spectral range (with efficiency comparable to that of OPAs) in a pulse three times shorter than the input, without any complex apparatus or additional pulse post-compression system (see figure).

While much of the above R&D was done at INRS by Riccardo Piccoli, Luca Razzari, Roberto Morandotti, and their colleagues, the team of researchers based in Vienna and headed by Andrius Baltuska and Paolo Carpeggiani had a complementary strategy to that of INRS. They also used a nitrogen-filled hollow-core fiber, but rather than filtering the spectrum, they compressed it in time with dispersive mirrors capable of adjusting the phase of the broadened pulse. “In this case, the overall shift in the infrared was less extreme, but the final pulse was much shorter and more intense, perfectly suited to attosecond and strong-field physics,” says Carpeggiani.

Stretched hollow-core fiber

In both cases, the experimental setup consisted of a hollow-core fiber (HCF) placed in holders and stretched, then pumped with 200 fs, 1.03 μm pulses from a Yb-based amplified laser system. The special system to stretch and hold the HCFs is marketed by the startup Few-Cycle (few-cycle.com), a spinoff from INRS. The holders allow HCFs with several-meter lengths to be stretched and held safely, leading to high pulse-compression ratios with up to 80% throughput. The setup at TUWien relied on a 5.5-m-long fiber with 1 mm inner diameter, while the INRS setup had a 6-m-long fiber with a 0.53 mm inner diameter, along with broadband chirped mirrors to compress the pulses.

The Moscow-based team, led by Aleksei Zheltikov, focused on developing a theoretical model to explain these optical phenomena. By combining these three approaches, the researchers were able to fully understand the complex underlying dynamics as well as achieve not only the extreme red shift using nitrogen, but also efficient pulse compression in the IR range.

The research team believes the Raman-shift-based method could very well meet the increasing demand for longer-wavelength ultrafast sources in laser and strong-field applications, starting with less expensive industrial-grade tunable systems based on Yb laser technology. 

REFERENCE

1. P. A. Carpeggiani et al., Optica (2020); https://doi.org/10.1364/optica.397685.