Ultrashort pulse (USP), or “ultrafast,” lasers emit extremely brief pulses of light, generally with duration of a picosecond (10-12 seconds) or less. The pulses are characterized by a high optical intensity that induces nonlinear interactions in various materials, including air.

One remarkable aspect in which USP lasers differ from traditional “long pulse” or continuous wave (CW) lasers is in their mechanism of ablation. A CW laser uses a process of linear excitation, generating substantial heat during the ablation process. The generated heat can vaporize the target, but it also creates widespread collateral damage since the heat generated can transfer in an uncontrolled fashion to the area surrounding the target. This can lead to melting, material reflow, or tissue charring.

In contrast, USP laser pulses deposit their energy in a time interval too short for significant thermal diffusion. Since the USP interaction avoids strong electron-to-phonon coupling, material removal is mediated by ionization (plasma formation) and Coulombic explosion. The ablation event occurs without leaving behind any heat or collateral damage. The benefits of athermal USP laser ablation in biomedical, material science, and micromachining applications have been investigated both theoretically and experimentally in research laboratories around the world with systems using high energy and low repetition rate optical ultrashort pulse trains. Real-world applications have been scarce due to the historical lack of robust, affordable and flexible laser sources with meaningful energy and average power specifications. More recent commercially available systems have leveraged new fiber-based technology to bring higher peak and average power to the market place.

Optical Fiber

For many applications, such as advanced biomedicine, it is critical to deliver the ultrashort laser pulses using a fiber optic path for precise, safe and minimally invasive ablation of soft or hard tissue. Historically, delivering high peak power ultrashort pulses with an optical fiber has been challenging due to optical nonlinear distortion, beam quality degradation, and dielectric damage of the fiber facets. For USP lasers, conventional systems still require hard-optic beam delivery (lenses and mirrors), either built into gantry robots or by means of an articulated arm. These solutions are expensive, bulky, and prone to frequent repair and maintenance. Hollow core plastic Bragg fiber USP beam delivery could decrease the cost of these machines and improve the flexibility by simplifying the system design. An optical fiber suitable for USP laser delivery should satisfy the following:

  1. Low transmission losses: the pulses must be delivered to the process site, usually a few meters to a few tens of meters, without significant power loss. Less than 1 dB/m transmission loss is required for most applications.
  2. High damage threshold: the fiber facets should not be damaged, or ablated, by the ultrashort pulses with peak power > 10 MW and pulse energy >10μJ.
  3. Low nonlinearities: the high pulse quality of the input pulses should not be distorted enroute to the target.
  4. Low chromatic dispersion: the ultra-short pulse duration should not be increased while traversing the fiber.
  5. Near diffraction-free beam quality: the output ultrashort pulses must be focusable to a tight spot on the target for various applications. The beam propagation parameter, or M2 value, should be less than 1.3.

Bragg Fiber

The hollow core plastic Bragg fiber is a promising candidate for USP laser beam delivery. As shown in Figure 1, the Bragg fiber consists of a hollow core, a bi-layer structure with alternating layers of high- and low-index materials with strong refractive index contrast and a plastic cladding layer for mechanical protection. The high- and low-index stack forms a “perfect mirror” on the inner surface of the hollow core, akin to multi-layer dielectric mirrors. The structure opens up a full photonic band-gap within the fiber that prevents the light in a specific wavelength band from leaking out during propagation1. Similar fiber has been fabricated and demonstrated commercially for continuous wave CO2 laser delivery2.

Figure 1. Geometrical structure of a hollow core Bragg delivery fiber

In conventional fibers, transmission losses are limited by bulk absorption of the glass material, waveguide leakage, and scattering. The Bragg fiber has a hollow core in which 99.9% of the light is localized. The resulting core material absorption loss is dramatically lower than that of a solid core fiber or bi-layer materials. The degree of confinement of the beam in the core is proportional to the number of layers of the bi-layer structure. By increasing the number of layers, typically up to 30, the waveguide leakage can be significantly reduced. Also by using purified materials and industrial preform and fiber fabrication processes, the scattering loss of the Bragg fiber can be kept at a low level. The theoretically predicted transmission loss of the Bragg fiber is approximately 0.2 dB/m (4.5%) and the measured value shows approximately 0.9dB/m (18.5%). By improving the fabrication processes, the transmission loss is expected to be further reduced.

Figure 2. Autocorrelation traces for a 270 fs pulse before and after propagating through 1m of hollow core Bragg fiber at 1552.5 nm

In small diameter solid core glass fibers, like those used for telecommunications, high energy USP laser pulses are significantly distorted by nonlinear optical effects such as self-phase modulation (SPM). In a hollow core fiber, most of the light propagates through the air in the core, and only the low intensity perimeter of the beam interacts with the bi-layer materials. Air has a nonlinear refractive index about three orders of magnitude lower than glass, so the hollow Bragg fiber design largely prevents any optical nonlinearities. In initial experiments, no significant spectral or pulse distortions were observed for the propagation of a 1ps pulse inside a 1m fiber sample with input energy of 4μJ. Within the fundamental band-gap of the Bragg fiber, chromatic dispersion is predicted and has been measured to be at a level that does not increase the duration of the pulses (Figure 2). We estimate the chromatic dispersion of the fiber to be less than 1 ps/nm/km. In addition, because less than 0.1% of the light overlaps with the bi-layer structure, the damage threshold of the hollow core Bragg fiber is dramatically higher than that of solid core delivery fibers. As proof, a USP laser beam with pulse energy of 5μJ, peak power of 5 MW, and average power of 2.5W has been coupled into the Bragg fiber with- out damage.

In order to preserve high beam quality, the Bragg fiber is designed for HE11 single mode — closest to Gaussian mode — operation by tailoring the bi-layer structure. The simulation and test show good mode discrimination even with a hollow core diameter of 100 μm — a size that helps to make launching light into the fiber much easier. The fiber supports single mode propagation for reasonable bending radii due to a high loss differential between the fundamental and higher order modes. The measured beam quality of a fabricated fiber sample shows M2 < 1.3 for the output beam (nearly diffraction limited).

In short, the single HE11 mode guidance, low chromatic dispersion and low nonlinearities ensure both spatially and temporally distortion-free transmission of high peak power USP pulses.

This article was written by Mike Mielke, Vice President of Engineering; Tim Booth, Vice President of Project Management; and Xiang Peng, Senior Research Scientist; Raydiance, Inc. (Petaluma, CA). For more information, contact Mr. Mielke at This email address is being protected from spambots. You need JavaScript enabled to view it., Mr. Booth at This email address is being protected from spambots. You need JavaScript enabled to view it., or Mr. Peng at This email address is being protected from spambots. You need JavaScript enabled to view it., or visit http://info.hotims.com/22916-201 .


  1. Yoel Fink, et al. (Nov. 27, 1998). Science, 282, pp. 1679-1682.
  2. Gregor Dellemann, et al. (Jun. 2003). Photonics Spectra.