Life in the fast lane – developing the world’s fastest laser
In the early 1990s, JILA Fellows Margaret Murnane and Henry Kapteyn played a key role in revolutionizing femtosecond lasers, by uncovering the beautiful physics of how lasers operate at the fundamental limits of pulse duration and stability.Their group at Washington State University was responsible for understanding how the ultrashort-pulse–mode-locked titanium-sapphire (Ti:S) laser can generate very short (sub-10 femtoseconds) pulses. In a series of papers, they showed in experiment and theory, that as the laser was re-designed to generate shorter and shorter pulses, the laser also became more stable. This laser is now a standard fixture in thousands of laboratories around the world for applications in AMO science, chemistry, biology, plasma science, materials science, nanoscience, medicine and industry.
As graduate students, Margaret and Henry were already designing and building ultrashort-pulse lasers and even using them to produce short (picosecond = 10-12 sec) bursts of incoherent X-rays i.e. the equivalent of an x-ray light bulb. After graduation the two scientists set up a joint laboratory at Washington State University. Their efforts took off in 1990 when Wilson Sibbett of Scotland’s St. Andrews University reported that a laser employing crystals of Titanium-doped sapphire could spontaneously generate pulses as short as 60 fs. Sibbett’s discovery was a big surprise in the ultrafast research community because theorists had predicted that it would be impossible to use a crystal material to make a short light pulse – all past work had used dye lasers to generate very short pulses of light. But Sibbett’s discovery showed that past understanding was not correct. Moreover, crystal material in theory had the ability to generate shorter pulses because Titanium sapphire emits a broader range of colors than any other laser material – so the race was on to understand how to generate even shorter pulses, to understand the limits of femtosecond lasers.
Henry and Margaret and their students soon realized that if all the colors generated by the Titanium sapphire crystal traveled with the same net speed in the laser, then it should be possible for the first time to generate pulses less than 10 fs directly from a laser. Usually however, different colors travel with different speeds in materials – a phenomenon called dispersion (which is why a prism separates white light into it’s various colors). So Henry and Margaret and their students redesigned the Titanium sapphire laser so that all the colors could march in step - coherently adding together light waves of many different colors. Constructive interference between the different coherent waves then creates a very short laser pulse at a specific point in time (or position within the laser). The challenge was figuring out how to keep the red colors from traveling faster through the laser crystal than the blue colors, a phenomenon that tears apart the light pulse. However, by using short laser crystals and prisms of the right material, the light waves covering a broad range of colors can travel at the same net speed in the laser.
Why do Titanium-sapphire lasers spontaneously generate short pulses? The mechanism took a few years to identify and turned out to be a very beautiful example of the new laser science discovered in the 1990s. When a few waves of different color add together to form a spike of light, the crystal properties change. Instead of behaving like a window, the crystal starts to behave like a lens – focusing the light beam as a result. Essentially, the crystal’s refractive index increases, allowing it to act like a lens. This focusing gets stronger and stronger as the pulse gets shorter, because the light spike gets larger. Fortunately, the laser can be adjusted to favor this strong focusing (called self-focusing), by also focusing the power source for the laser. Thus, the laser favors the strong self-focusing regime. It is even more stable when very short pulses are generated by the laser. This was another big surprise – usually lasers become more unstable and sensitive when they generate very short pulses.
By 1993, Henry and Margaret had invented a self-mode-locked Ti:Sapphire laser that generated 11-fs pulses, and some as short as 8 fs. Their secret laser design was simple and robust. The trick was to use a short (< 5 mm) highly doped Ti:Sapphire crystal and a pair of fused-silica prisms in the laser. Although the laser crystal caused the different colors (light waves) to travel at different speeds, the two prisms could perfectly compensate for this.
Once they had created the fastest laser in the world, Margaret and Henry were inundated with requests from other scientists for how to duplicate their laser. Henry prepared a document with lots of good information about the design, where to buy the parts, and made it readily available to researchers willing to take the time to build their own laser. However, some scientists did not have access to machine shops to build the laser themselves. So in 1994, Henry and Margaret also formed a company, KMLabs, to manufacture and sell their Ti:S mode-locked laser to researchers in academia and industry who preferred a simpler turnkey approach to getting a new 10fs laser.
Laying the Groundwork for a Tabletop X-Ray Laser
With their new ultrafast laser in hand, Margaret and Henry turned their attention to generating very short bursts of X-rays – to capture the dizzying dance of electrons in materials and molecules. Here again, a new discovery allowed them to make a huge step forward in generating laser-like x-ray beams using a simple tabletop femtosecond laser.
One year after the laser was first demonstrated in 1960, one of the first new phenomena observed was that by focusing a laser beam onto a material, it was possible to combine two laser photons together to generate a shorter wavelength photon. This ability was first demonstrated by Franken, who combined two Ruby laser red photons to generate a single blue photon. This phenomenon is called nonlinear optics. Just like when you pluck a guitar or violin string very hard, and you hear higher frequencies or harmonics or the sound waves, a laser beam can pluck the electrons in material very hard, to generate new colors, also called harmonics of the light beam. However, for decades only a small number of harmonics were observed.
This situation completely changed with the discovery of high harmonic generation (HHG) in 1987 by the Rhodes group in Chicago. They focused a femtosecond laser beam into a gas, and unexpectedly saw a very large number of odd harmonics of the fundamental driving laser - corresponding to adding together an odd number of photons (i.e. 3rd, 5th, 7th, up to 17th).
Henry and Margaret and their students realized that this approach would work even better if they used very short femtosecond laser pulses. They quickly showed that they could generate high-harmonic X-rays with wavelength as short as 3 nm (or 30 Å) and pulse durations even shorter than the laser pulses used to create them – in the sub-femtosecond or attosecond regime (1 as = 1—18 sec). In high-harmonic generation, a high intensity laser pulse plucks electrons out of atoms (or ions) such as argon. The electron then gains energy from the laser, and when the high-speed electrons crash back into the atoms, the atoms emit short bursts of X-rays.
Henry and Margaret and their students worked hard to understand how these x-ray bursts were generated, so they could make the high harmonic generation process as efficient as possible, to make bright beams of laser-like x-rays, especially at high photon energies needed in order to visualize fast motions in the nanoworld. Like their work on femtosecond lasers, in order to make high harmonic generation efficient, they had to ensure that the laser and x-ray waves move in sync throughout the gas, so that the x-rays add coherently to make a bright laser-like (coherent and phase matched) beam. This succeeded beyond their expectations. Recent work by Tenio Popmintchev and Ming-Chang Chen in their group, in collaboration with the Baltuska group in Vienna, has shown that they can efficiently generate laser-like beams of x-rays that span all the way from the ultraviolet to the kilo-electron-volt region, corresponding to wavelengths below 8 Å. This efficiently combines >5000 laser photons — a record for nonlinear optics.
The key improvement was to propagate the visible laser pulse through a waveguide (hollow capillary tube) filled with argon or another noble gas. The waveguide makes it possible to carefully control the phase velocity of the laser beam and eliminated beam divergence, allowing much improved phase matching. The waveguide also made it possible to easily adjust the gas pressure so that the laser light and the X-ray light traveled at the same speed, which was also necessary for phase matching.
Murnane and Kapteyn predicted this ability would lead to many new experiments in X-ray science, a prediction they would go on to realize in the years to come.