This is the first of two parts that discuss this extensive study of femtosecond lasers. The second part will be published in next month's issue.
A German project on femtosecond-laser technology and applications—established in 1999 and presented to the public in September 2003—is a joint effort by companies and research institutes, and is supported by the German Federal Ministry for Education and Research, which provided about €27 million (US$31.98 million), with industry donating another €15 million (US$17.77 million). The goal was to explore the potential of femtosecond lasers in manufacturing and medicine. Results are highly encouraging and further public funding is expected to be invested into a broad application field termed "femtonics" by the representative of the main sponsor.
Divided into packages
The project was divided into six so-called packages. The first, called Primus, pursued materials processing with ultrashort laser pulses and was a collaboration between five institutes and nine companies in different working groups. The predominant applications were microdrilling, microstructuring, and diamond cutting. Different laser setups were investigated, with the aim of using them in production. Oscillator-amplifier systems with both Yb:YAG and Nd:YVO4 as amplifying media were realized in disk, rod, and fiber design. For example, fiber systems with chirped-pulse amplification had, in one case, an average power of 76 W at 400-fs pulse length and repetition rate of 75 MHz, and in another, a high pulse energy of 50 µJ at a 120-kHz repetition rate.
The characterization of optical components and laser beams (CHOCLAB) became an important issue in the project. Many existing prescriptions for standardization were also found to hold for the characterization of ultrashort pulse lasers and components. Some had to be revised, however; for example, current wavefront analyzers do not suffice for beam characterization.
For metal drilling, subpicosecond pulses were found to give rise to nonlinear effects that lowered the quality of microscopic holes, unless an additional air stream was applied. Longer nanosecond pulses resulted in deposits due to melting, an effect avoided by applying trepanned drilling with picosecond pulses and keeping the direction of polarization perpendicular to the bore-wall surface. For fuel-injection nozzles, drilling of 100-µm-diameter holes of about 1-mm length was achieved within 200 seconds with a quality markedly superior (no deposits by melting) to that obtainable by nanosecond pulses, although processing speed is still too low for economical application.
In cooperation with the automotive industry, surface structuring in metal was investigated to create small depressions in the surface of cylinder walls that can store lubricants; a significant reduction in friction was found. In another application, image-pattern microdepressions could be directly written—with no aftertreatment necessary—on the surface of printing rolls. Good printing results were the result, although economic break-even would require greater than a 100-kHz repetition rate, which is not yet available.
The project also investigated the microstructuring of diamond surfaces to sharpen cutting edges of milling cutters. Using 120-fs pulses and adjusting the direction of polarization perpendicular to the forward feed, promising results were obtained for cut depths up to 300 µm. The results of Primus were encouraging for industrial introduction of many of these tools.
Coherence radar
One of the packages was femtosecond measurement technology (Fesmet), a working group consisting of six companies and four research institutes. This was focused on coherence radar, a technique that interferes light scattered from a spot on the probe surface with that from a spot of a microrough reference surface, both mounted in the legs of a Michelson interferometer. A CCD array is placed in the image plane of the recombined beam; each pixel collects light from homologous areas on the probe and the reference surface. Probe and reference surfaces are illuminated by a laser with broad spectral bandwidth. As the reference plane travels axially, an interferogram is recorded for each pixel. Whenever the difference of optical paths is around zero, white-light interference occurs; in this way, an object can be height-scanned parallel in all pixels, with the height resolution being larger for a broader spectral width. Lateral resolution is determined by the CCD pixel size.
A main goal of Fesmet was to develop broadband sources based on short-pulse laser technology. The straightforward approach was to optimize a Ti:sapphire laser with a saturable absorber for Q-switching the cavity, producing bandwidths of up to nearly 50 nm. In a second step, light was coupled into a microstructured optical fiber, leading to a supercontinuum extending from 700 to 1600 nm. A complete measurement system, including software, was able to scan objects of 45 mm in height with a resolution down to 10 nm.
First applications were in quality control, aerospace, and dentistry. In aerospace, for example, the shape of turbine buckets is extremely important for the efficiency of the turbine. Using coherence radar, the shape of those buckets could be precisely compared to the computer-aided design data. In dentistry, the goal is to measure the shape of casts in order to generate data for replacement parts such as crowns, bridges, and inlays.
Information on the German femtosecond-laser project can be found at www.fgsw.de/fst.