White light supercontinuum: Power struggle

Marie Freebody, marie.freebody@photonics.com

Researchers are pioneering a practical way of generating high-power white light suitable for applications including spectroscopy, microscopy and optical coherence tomography. Producing high-power light across a broad spectrum, also known as a supercontinuum, is not easy. It requires high-power pumping to trigger significant nonlinear effects (usually within a nonlinear medium such as an optical fiber). In this process, incident light can be converted into supercontinuous white light.

Today’s pump source of choice is the pulsed laser. With typical peak powers of up to several tens of kilowatts, it provides more than enough power to excite most nonlinear effects. On the other hand, the average power of a compact and low-cost continuous wave (CW) laser is much lower (between 10 and 100 W), falling far short of meeting the requirements of the job. Now, Alexandre Kudlinski and colleagues at the University of Lille have managed to compensate for this power shortfall, generating a CW-pumped supercontinuum spanning the elusive visible region.

This photograph is of the illuminated fiber spool during CW supercontinuum experiments. The red light is generated first and progressively turns into yellow as the fiber length increases. Finally, all visible wavelengths are generated, leading to white light collimated with a lens at the fiber output.

The group’s white light source extends from 470 to more than 1750 nm, with almost 10 W of average power for a pump of just 45 W. Kudlinski believes that this alternative approach opens up possibilities that could benefit many applications, including flow cytometry, endoscopy, optical coherence tomography and fluorescence microscopy.

The white-light-output spot, collimated and dispersed with a prism, shows the whole visible spectrum generated in the fiber.

“Most commercial confocal microscopes are generally made up of several visible CW lasers to excite fluorophores in various biological samples,” he explained. “Many interesting fluorophores are unusable because of the lack of suitable laser sources to excite them. Our CW supercontinuum source provides the required wavelengths with sufficient spectral power density for efficient excitation of new, useful fluorophores.”

In this experimental setup of CW supercontinuum generation, the illuminated fiber spool and fiber end are delivering high-power white light through a collimating lens. Images courtesy of Université Lille 1.

What’s more, a CW supercontinuum is much easier to use than its pulsed counterparts. There is no need to synchronize the detection setup with the source, as opposed to when using a pulsed supercontinuum where expensive and cumbersome synchronization elements are required.

The researchers found that achieving their goal involved a careful balancing act between increasing the nonlinear response of the optical fiber and maintaining optimized dispersion. They opted for a photonic crystal fiber with a germanium dioxide-doped core to promote nonlinear behavior, and they controlled the resulting dispersion using a cleverly tapered fiber design.

Shown are scanning electron microscope images of the tapered photonic crystal fiber output face (outer diameter of 85 µm). The black regions are airholes and the gray regions, silica. The germanium dioxide-doped core appears slightly lighter at the center of the structure.

“It is well-known that germanium dioxide enhances both the Kerr and Raman nonlinear responses of silica glass, which is, of course, beneficial for supercontinuum generation,” Kudlinski said. “On the other hand, germanium dioxide modifies the dispersion properties, making it generally unsuitable for efficient supercontinuum generation. The key to our work was to adjust the dispersion by transversally controlling the size of the airholes within the microstructured fiber and by longitudinally decreasing the zero-dispersion wavelength along the length of the fiber.”

In the next stage of the research, Kudlinski aims to reach even shorter wavelengths (down to 350 nm), which is particularly interesting for many applications, including fluorescence microscopy. He also plans to test the source by integrating it into the laboratory’s confocal microscopes.