Fiber-Based System Compresses Mid-Infrared Laser Pulses at 80 Watts Instead of Kilowatts

Published in IEEE Journal of Quantum Electronics. DOI: 10.1109/JQE.2025.3638679. Research by SASTRA Deemed University, Thanjavur, India.

Ultrashort mid-infrared laser pulses are workhorses across molecular spectroscopy, nonlinear microscopy, and biomedical imaging. Generating them has traditionally required complex, power-hungry systems: high pump powers, elaborate optical setups, and the kind of precise alignment that keeps lasers confined to specialized laboratories. Taking these capabilities into clinical or field settings has been impractical.

A team at SASTRA Deemed University in Thanjavur, India, has now demonstrated a compact, fiber-based alternative that drops the required input power from the kilowatt range to just 80 watts while producing clean, high-contrast pulses. Their results, published in the IEEE Journal of Quantum Electronics, mark the first demonstration of sub-200 femtosecond pulse compression in the mid-infrared using a holmium-doped ZBLAN photonic crystal fiber.

From kilowatts to 80 watts

The system compresses 5-picosecond input pulses down to 187 femtoseconds -- a compression factor of 26.7 -- with pedestal energy as low as 0.63%. Pedestal energy is the fraction of pulse energy that leaks into unwanted temporal sidelobes rather than staying concentrated in the main pulse. Low pedestal energy means clean, well-defined pulses suitable for precision measurements.

Lead author G. Sornambigai explained that by combining rare-earth-enabled gain and the nonlinear pulse shaping mechanism of a nonlinear optical loop mirror (NOLM) configuration, the team reduced the required input power by more than an order of magnitude. The power reduction is not just about energy efficiency. Lower power means less heat, less risk of fiber damage, and longer operational lifetimes -- all factors that matter for real-world deployment outside controlled laboratory environments.

How the fiber does the work

The key innovation is integrating three functions into a single fiber-based architecture. A carefully engineered tapered fiber geometry enables self-similar pulse evolution -- a process where the pulse maintains its shape as it propagates, ensuring efficient compression without distortion. Holmium doping provides optical gain near 2.86 micrometers, amplifying the signal and compensating for losses during propagation. And the NOLM configuration provides the nonlinear pulse shaping needed to compress the pulse temporally.

Together, these components suppress the unwanted temporal pedestals that plague other compression schemes, prevent fiber damage under high-intensity operation, and minimize alignment complexity. The entire system operates within a single fiber, eliminating the free-space optical elements -- mirrors, lenses, beam splitters -- that make conventional systems bulky and alignment-sensitive.

Co-author R. Vasantha Jayakantha Raja noted that the architecture delivers clean, high-contrast pulses well suited for mid-IR spectroscopy and nonlinear imaging applications.

Mid-infrared: the molecular fingerprint region

The mid-infrared spectral region, roughly 2 to 20 micrometers in wavelength, is sometimes called the molecular fingerprint region because many molecules have strong, distinctive absorption features in this range. This makes mid-IR light essential for identifying chemical compounds in settings from pharmaceutical quality control to environmental monitoring to medical diagnostics.

Ultrashort pulses in this region are particularly valuable because they enable time-resolved measurements -- capturing molecular dynamics on femtosecond timescales -- and nonlinear optical techniques that require high peak intensities concentrated in brief bursts. The challenge has been generating these pulses without the bulk and expense of traditional laser systems.

The SASTRA team's fiber-based approach addresses this by trading the complexity of free-space optics for the simplicity and robustness of fiber systems. Fibers are inherently alignment-free -- the light propagates through the core without needing external mirrors or adjustments -- which makes them more suitable for deployment in clinical settings, industrial environments, or field stations.

Performance in context

The 187-femtosecond pulse duration achieved is well within the ultrafast regime needed for most spectroscopic and imaging applications. The compression factor of 26.7 is competitive with existing free-space systems, and the pedestal energy of 0.63% is low enough to avoid the artifacts that degrade measurement quality in applications requiring high temporal contrast.

The researchers describe this as the first Ho:ZBLAN-based NOLM system to produce sub-200 femtosecond pulses in the mid-infrared, marking a milestone in compact ultrafast mid-IR source development.

What remains to be demonstrated

The results are based on numerical modeling and simulations, not experimental measurements from a built device. Self-similar pulse modeling played a key role in optimizing the design, and the researchers conducted system-level analysis to ensure reliable compression. But the transition from simulation to a working fiber system involves practical challenges -- splice losses, fiber defects, thermal management -- that could affect real-world performance.

The ZBLAN glass used in the fiber is fragile compared to silica fibers and more difficult to handle in manufacturing. Scaling production of holmium-doped ZBLAN photonic crystal fibers to commercially viable quantities remains an engineering challenge.

The system has been designed for a single wavelength near 2.86 micrometers. Extending the approach to other mid-IR wavelengths would require different rare-earth dopants and potentially different fiber designs. Tunability, a feature valued in spectroscopic applications, has not been demonstrated.

Still, the principle is sound: by combining gain, nonlinear shaping, and careful fiber engineering in a single integrated architecture, the power requirements for ultrashort mid-IR pulse generation drop dramatically. If the simulation results hold up experimentally, this approach could bring ultrafast mid-infrared capabilities to laboratories and clinics that cannot currently access them.