Laser-Produced Plasmas as Short-Wavelength Incoherent Optical Sources

James F. Young    Department of Electrical and Computer Engineering, Rice University, Houston, Texas

1.1 Introduction

Incoherent sources are useful experimental tools that have long been used throughout the optical spectrum, subject to availability. Recent technological advances have extended the available range into the vacuum ultraviolet and soft x-ray range through the use of laser-produced plasmas. This chapter will concentrate on the description of these new sources. Traditional vacuum ultraviolet sources and appropriate experimental techniques in this spectral range have been discussed in detail previously [1]. Conventional sources in other spectral regions—from the infrared, through the visible, to the ultraviolet quartz absorption edge at about 160 nm—have also been reviewed [2].

Figure 1 shows a typical configuration of a laser-produced plasma. A pulsed high-power laser is focused to a small spot on a metal target. Asmall, hot, radiating plasma or spark is formed that is characterized by very high densities of electrons and ions, very large density gradients, and very high temperatures. The parameters can be comparable to those in stellar interiors. Much of the work on laser-produced plasmas was motivated by the laser inertial confinement fusion program, and therefore concentrated on low-atomic-number targets irradiated at extremely high power densities, perhaps 1015 W cm−2; interest was focused on shock wave formation, ablation rates, and radiation above l keV [3, 4]. Researchers in other fields, however, realized that laser-produced plasmas could be produced at much lower laser fluences, and therefore with smaller, practical lasers, and yet still provide unique sources of short-wavelength radiation [5–18]. Laser-produced plasmas can be used as a point source to illuminate a separate sample, but often, as shown in Fig. 1, a gas of the atoms to be irradiated surrounds the plasma, eliminating collection optics. The combination of a laser-produced plasma and a surrounding gas that can be photoionized also represents a unique pulsed source of hot electrons that can be used to excite optically forbidden transitions [19, 20]. Even a modest pumping laser can produce an effective electron current density of > 105 A-cm−2 with a subnanosecond rise time.

The experimental appeal of laser-produced plasma light sources results from their combination of simplicity and unique characteristics. The physical reality of a typical source is not far different from the schematic in Fig. 1, yet it constitutes a radiation source with a bandwidth and spectral intensity that is difficult to match even at national facilities. The emission from a laser-produced plasma depends on both the laser and the target parameters, but can have a broad smooth spectrum that is often characterized as a black-body radiator with a temperature from 10 eV to several keV. Figure 2 shows a comparison between measured radiant energy spectra from an Au target and calculated black-body radiation spectral profiles. The agreement is reasonably good, but the black-body temperature of a laser-produced plasma, although useful, should be used only qualitatively. The conversion efficiency from input laser energy to total energy radiated by a plasma can vary from a few percent to more than 60%, as shown in Fig. 3 [21]. These data also show the advantages of short-wavelength pump lasers. The time behavior of the plasma temperature and radiation generally follows that of the driving laser pulse, at least down to about 10-pstime scales, providing a unique tool for time-resolved studies [22–25].

Before reviewing the characteristics of black-body radiators and summarizing the physics of laser-heated plasmas, it is helpful to present some typical numbers to place laser-produced plasmas in a physical context. High focused light intensities are required to initiate and sustain a plasma. While densities of 1015 W cm−2 and higher have been used to produce high x-ray yields, typical values for laboratory sources are 1011 to 1013 W cm−2. Thus, for a focal spot of 50 pm diameter, a peak laser power of ~10 MW is required, and pump lasers having a short pulse length must be used to keep the energies practical. The Nd laser fundamental wavelength of 1064 nm plus its harmonics (see Fig. 3) are common choices for the driving radiation because of the commercial availability of the laser. Early fusion work studied CO2 laser-driven plasmas extensively because of the economy and scalability of CO2 lasers, but the long wavelength and generally longer pulse lengths reduce plasma heating efficiency. Typically, a plasma heated by 1064 nm radiation will have a spectral radiance 10 to 30 times greater than one heated by 10.6-μm radiation for the same input energy [26, 27]. This difference is illustrated particularly clearly by the data shown in Fig. 4. Excimer lasers are potentially important sources for laser-produced plasmas, but currently the large beam divergence and long pulse lengths of standard commercial lasers are major problems [28].

Although Nd Q-switched lasers with pulse lengths up to 10 nsec are used to heat plasmas, Nd mode-locked lasers with pulse lengths of about 100 psec are used more often. Minimum required pulse energies are therefore in the range of 1 to 200 mJ. The recent commercial availability of such sources with high repetition rates has made laser-produced plasmas practical experimental light sources. U1-trashort subpicosecond pulse length laser systems based on dye or Tixapphire gain media have been used to produce plasmas with unique characteristics [29–34]. The plasmas produced by very short high-intensity pump pulses are fundamentally different from those produced by more conventional pump sources, which are discussed here. For example, Fig. 5 shows the high-energy spectrum radiated by a plasma heated by a 0.5-TW 120-fsec pulse focused to a density greater than 10l8 W cm−2. Hard x-ray radiation extending beyond 1 MeV was observed. The spectrum was taken through 19 mm of lead to avoid saturating the detector. The physics of plasma formation and heating under such conditions are quite different from that presented here and are still an active area of research.

1.2 Black-Body Radiators

For many experimental applications, for example, absorption spectroscopy, a source with a broad bandwidth and a flux that is reasonably independent of wavelength is required. The meaning of the vague qualifiers "broad" and "reason-ably" vary widely, of course, with the specific application. The prototype incoher ent broadband source is the black-body radiator with a spectral radiance given by Planck's law:

λ,=T=A/λ5expB/λTWcm−2sr−1nm−1,

where the constants A and B are given by

=2hc2=3.97×1020nm5,

and

=chk=1240nmeV,

h is Planck's constant, c is the velocity of light, and k is Boltzmann's constant. The numerical values above and those elsewhere in this chapter assume wavelengths, λ, in nm, and temperatures, T, in units of electron volts, T (eV) = T (K)/11605.

Although the dependence of the black-body spectral radiance on source temperature (Eq. 1) is quite familiar, it is worth repeating that, as the temperature is raised, not only does the peak emission shift to shorter wavelengths, but the emission at every wavelength also increases. Thus, for such sources hotter is always...