High-power fiber lasers are everywhere these days — from cutting and welding in factories to medical surgeries and scientific labs. They're compact, efficient, and deliver excellent beam quality. But there's a catch: if you want to send that powerful laser beam over a long distance (say, to separate the laser source from the workpiece for flexible factory layouts), conventional solid-core fibers hit a wall.

Take a real example: Fujikura's ytterbium-doped fiber laser can transmit 5 kW over just 20 meters. Crank it up to 8 kW, and the distance shrinks to only 3 meters. The problem? Material damage thresholds and nasty nonlinear effects like self‑phase modulation or stimulated scattering.

That's where hollow-core fibers (HCFs) come in.

Instead of traveling through glass, the light zips through an air‑filled or vacuum core. Early experiments already showed that guiding light in air dramatically cuts nonlinearity and raises damage thresholds. Over the past decade, HCF performance has skyrocketed: losses have dropped from >100 dB/km to <1 dB/km in the best anti‑resonant designs — approaching ordinary silica fibers.

So how do they trap light inside an air hole? Solid‑core fibers use total internal reflection, but air has a lower refractive index than glass, so you need a trick. Two main tricks, actually:

1. Photonic bandgap (PBG) – A periodic microstructure in the cladding creates a "bandgap" that prevents light from escaping. The first practical hollow‑core fiber (HC-PBG-PCF) came out of Southampton in the late 1990s. But PBG fibers still have fairly high loss (dB/km scale) and are complex to make.
2. Anti‑resonant guidance – This is the current star. Thin‑walled glass tubes (often nested) act like a Fabry‑Pérot etalon, reflecting light back into the core. Anti‑resonant HCFs (AR-HCFs) offer wider transmission windows, lower loss, and simpler fabrication. Variants include Kagome, Revolver, node‑less, and nested anti‑resonant node‑less fibers (NANFs) – the latter holding the loss record today.

Why are these fibers such a big deal for high‑power and ultrafast lasers? Four key advantages stand out:

· Extremely low nonlinearity – The Kerr effect in air is ~1000× weaker than in glass. That means almost no self‑phase modulation, SBS, or SRS to mess up your beam. Even single‑frequency lasers can be transmitted without parasitic SBS.
· High damage threshold – Light barely touches the glass walls, so surface intensity stays low. You can push over 2 kW without harming the cladding microstructure. Some experiments ran for weeks at >100 mW with zero degradation.
· Broad spectral window – Wavelengths that are impossible for silica fibers (like mid‑IR 2–10 μm or deep UV) work beautifully in hollow‑core fibers. Skylark lasers transmitted >100 mW of UV, again for weeks.
· Low latency – Light in air travels about 31% faster than in glass (refractive index ~1 vs. ~1.45). That's crucial for telecom, but also for precision timing in ultrafast laser systems.

Real‑world results are already impressive.

In 2025, Shi et al. (Nature Communications) demonstrated 2 kW continuous‑wave laser transmission over 2.45 km using an AR-HCF with a record loss of 0.168 dB/km at 1080 nm. The power–distance product was 500× better than previous all‑fiber systems. They even observed Raman scattering inside the silica nested tubes – and managed to suppress it, opening the door to industrial applications like nuclear decommissioning and laser drilling.

Another study showed flexible transmission of mid‑infrared ultrafast pulses (2.8 μm, 100 fs, watt‑level energy) through a 5‑m evacuated hollow‑core PCF. The pulses kept their spatial, spectral, and temporal fidelity – perfect for spectroscopy, surgery, or remote sensing.

Of course, challenges remain.

· Residual loss – Though simulations hit 0.025 dB/km at 1550 nm and experiments reach 0.168 dB/km at 1080 nm, the 0.14 dB/km benchmark of silica isn't yet consistently beaten across all wavelengths.
· End‑face damage – Under high continuous power, the polymer coating and jacket glass (not the microstructure) can degrade. Air ionization inside the core may also limit power scaling.
· Mode purity – Hollow‑core fibers are inherently multimode. Recent designs with quadruple‑truncated dual‑nested structures achieve fundamental mode loss of 0.1 dB/km and high‑order mode loss of 430 dB/km (extinction ratio 5×10⁻⁴) – but maintaining single‑mode operation over long distances is still tricky.
· Fabrication repeatability – Stack‑and‑draw requires sub‑micron precision. Any variation in drawing conditions, pressure, or glass purity affects performance.
· Coupling to solid‑core fibers – Most high‑power laser sources are solid‑core, so efficient coupling demands precise mode‑field matching. Today's systems often use free‑space optics, which limits long‑term stability.

Looking ahead, the future is bright.

Triple‑nested anti‑resonant fibers (TNANFs) have already achieved 0.25 dB/km loss with smaller diameters. All‑fiber transmission (no free‑space coupling) is on the horizon – targeting 10 kW and beyond. Gas‑filled HCFs will continue to generate octave‑spanning supercontinua, VUV femtosecond pulses, and even tabletop attosecond X‑rays. Industrial integration with robot‑mounted flexible fibers is coming. And the market is growing: from $92 million in 2025 to $158 million by 2032 (CAGR 8.1%).