is often possible to obtain pulse energies, durations, bandwidths, and repetition rates that are optimum for probing particular processes.

Currently, the most common laser system for producing tunable radiation in the range of a few microns to 100 nm is a visible dye or tunable solid-state laser using associated nonlinear techniques to reach shorter and longer wavelengths. The pump source for dye lasers is often a frequency doubled or tripled Nd:YAG laser that is either Q-switched or mode locked. Typically, nonlinear techniques—such as frequency doubling and mixing, stimulated Raman scattering, and four-wave mixing—provide wavelengths beyond the fundamental range (400 nm to 800 nm) of a pulsed dye laser. This is a well-developed, proven commercial technology that is currently undergoing potentially revolutionary advances. Nonlinear crystals are likely to be central to developments in this area, but the limited resources of the small businesses that usually produce them slow their development into optimized optical elements.

The development of titanium-doped sapphire lasers into practical devices has changed the production of ultrafast light pulses with durations from a few picoseconds to less than 10 femtoseconds. These solid-state devices, which produce tunable fundamental radiation in the range of 700 nm to 1100 nm, are excellent for mode-locking to create the short pulses that are efficient for producing other wavelengths by nonlinear processes. Other tunable solid-state lasers such as those based on forsterite or alexandrite, although less developed commercially, also provide tunable fundamental radiation, and, in all cases, nonlinear doubling and mixing techniques extend the wavelength range far beyond the fundamental region.

The other seminal development that is just beginning to change the landscape for commercial pulsed lasers is the introduction of optical parametric oscillators (OPOs) for nanosecond-duration pulses and optical parametric amplifiers (OPAs) for shorter pulses. These systems use b-barium borate (BBO), potassium trihydrogen phosphate (KTP), lithium niobate (LiNbO3), and new nonlinear crystals such as AgGaSe2. For example, several commercial manufacturers now offer OPOs that span the range from 440 nm to 2 µm with 10-ns, 20- to 100-mJ pulses of light and produce 1- to 10-mJ pulses in the range of 2 to 3.5 µm. Others promise an OPA that can generate 1-ps, tunable infrared pulses at a kilohertz repetition rate at wavelengths between 1 and 5 µm. Nonlinear frequency difference and mixing schemes should routinely provide radiation with wavelengths as short as 200 nm and as long as 5 µm from these devices. Many laboratory devices operate within a factor of a few of the Fourier transform limit of bandwidth determined by the pulse duration. A 100-fs ultrafast laser pulse has a transform-limited bandwidth of 150 cm−1, corresponding to a fractional bandwidth of Δλ/λ = 0.01 at 15,000 cm−1. Sophisticated seeding technologies can produce longer-duration (nanosecond) pulses that are near the transform-limited bandwidth as well, but, even without these measures, bandwidths are routinely less than 0.05 cm−1, corresponding to a fractional bandwidth of Δλ/λ ≤ 3 × 10−6 .The good spatial properties, large pulse energies, and small bandwidths of conventional laboratory lasers make them very bright sources in the regions where they operate.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement