The optical laser, invented by Theodore Maiman in 1960 (Maiman, 1960), is a funda- mental technology in today’s world. Lasers have grown from engaging our collective imagination through works of science fiction, to being an essential part of our day to day lives, facilitating communications, industry, and entertainment. Lasers are now mainstream and commercially available, replacing many existing technologies by enabling superior performance; for example, the once prevalent cassette tape is now obsolete due to CD, DVD, and Blu-Ray technologies – all enabled by optical readout via lasers.
The crucial component of any laser system is the gain medium. The gain medium facilitates the transfer of some form of energy (typically electrical or optical) to co- herent radiation through the transitions of the electronic or molecular states of the host material. Therefore the emission wavelength of a laser is tied to the material properties of the gain medium. When we consider specifically lasers compatible with high peak power operation (for our purposes >10 kW), finding suitable gain media is not trivial. The table below shows a list of popular high-power-compatible gain materials and their emission wavelengths:
Medium Emission wavelength Nd:YAG 1.064 µm Ti:Sapphire 0.7-1.1 µm Ytterbium 1-1.1 µm Erbium 1.5 µm Thulium 1.9-2.1 µm
Table 1.1: Popular high-power gain media and their emission wave- lengths
Considering the emission wavelengths above, it is clear that high power laser tech- nology is segregated to the near-infrared portion of the spectrum, leaving a majority of optical frequencies without applicable sources (notably visible and mid-infrared wavelengths). There are of course many applications requiring lasers in these under- served wavelength regimes: some examples include high power blue-green radiation (4XX nm) for underwater communications and sensing (Akyildiz et al., 2005; Embley et al., 2006; Hanson and Radic, 2008), yellow radiation (589 nm) for sodium guide star applications (Max et al., 1994; Morris, 1994; Avicola et al., 1994; Max, 1997; Feng et al., 2009), mid-infrared radiation (2-14 µm) for explosives detection (Steinfeld and Wormhoudt, 1998; Todd et al., 2002; Caygill et al., 2012), as well as laser sources within the optical transparency windows of biological tissues (1.3 and 1.7 µm) for deep-tissue microscopy (Xu and Wise, 2013). There is accordingly significant interest in developing laser technologies for these regions.
Nonlinear optics (NLO) may provide a potential solution to this problem. A non- linear optical interaction is one in which the response of a material is not linearly proportional to the magnitude of the applied electric field, resulting in a host of dif- ferent effects – including, in some cases, the conversion of the incident light wave to a new frequency (Franken et al., 1961; Armstrong et al., 1962). Unlike a laser gain medium, however, the frequency output from a nonlinear, or parametric, frequency
conversion (PFC) process is not directly related to the nonlinear material; the inter- action only requires that energy and momentum of the interacting light waves are conserved. Therefore an existing high power laser paired with a nonlinear frequency conversion system is an attractive platform for generating high power radiation at wavelengths all across the optical spectrum.
This is the principle at work in an optical parametric oscillator (OPO) (Giord- maine and Miller, 1965) wherein an anisotropic bulk medium facilitates the χ(2)
nonlinearity, leading to interactions involving three photons. Fig. 1·1(a) and (b) schematically demonstrate the two most common χ(2) processes: sum and difference
frequency generation, respectively. By placing the nonlinear medium into a resonant cavity, pump light passes through the crystal and is converted to a new frequency. This frequency can be tuned simply by adjusting the angle of the crystal in order to operate at wavelengths all across the spectrum. Modern OPOs provide hundreds of nanometers of tuning bandwidth, making them a mainstay for applications across all disciplines of optics (Tang et al., 1992; Driscoll et al., 1994; Myers et al., 1995; Coherent, 2010).
While the OPO has become a mature, and useful technology, it is not without drawbacks. Because nonlinear crystals provide only short interaction lengths, OPO systems require the pump to pass through the medium multiple times in order to efficiently convert to a new frequency. This required free space cavity is inevitably composed of many moving parts, thus OPOs typically have large footprints and are notoriously difficult to align and keep aligned. Furthermore, the nonlinear crystal is subject to degradation over time requiring periodic maintenance. As a result, these systems are undesired for rugged, non-laboratory applications.
An ideal system would employ a single-pass (as opposed to resonant) conversion geometry, handle high peak powers, and be capable of alignment-free operation in a
Figure 1·1: Example parametric frequency conversion processes for the χ(2) and χ(3) nonlinear susceptibilities. (a) Difference frequency
generation. (b) Sum frequency generation. (c) Four-wave mixing. (d) Four-photon sum frequency generation.
small-footprint configuration. These are exactly the characteristics that have fueled the development and success of optical fiber lasers (Snitzer, 1961; Koester and Snitzer, 1964). Additionally, the silica glass which comprises most optical fibers is an isotropic medium which has χ(3)nonlinearity. Thus fibers can be used to demonstrate nonlinear frequency conversion processes between four interacting waves. Examples of χ(3)PFC
processes are shown in Fig. 1·1 including four-wave mixing (c) and four-wave sum frequency generation (d). Because fibers guide waves, the interaction lengths for fiber- based nonlinearities can be meter or even kilometer-scale, making them compatible with a single-pass conversion geometry. It follows that a fiber laser paired with a fiber-based PFC mechanism is an ideal means to address the need for high power
lasers at novel wavelengths.