Both electron thermal conductivity and thermal exchange with the lattice can cool an electron distribution initially heated on a metallic surface with an ultrashort laser pulse. The interplay between the two processes allows the electron-lattice coupling parameter to be determined. We report measurements of optical damage to molybdenum and copper. Damage caused by pulses have a duration TL ^ 1 nsec can be understood only with a two-temperature model of metals.PACS numbers: 72.15.Eb, 78.47.+p A reliable value of the electron-phonon coupling constant g in metals ^ has wide implications, perhaps even in the field of superconductivity.^ Recent investigations^'^ have deduced or inferred experimental values of g. However, a large uncertainty remains because of the difficulty in firmly relating observations to theory. The experimental approach has been to diagnose the electron temperature either by photon-assisted electron emission^ or by the change in reflection^'^ (or transmission^) of visible light near the rf-state resonance. These processes are all limited to relatively weak excitation, the former because of space-charge effects and the latter because of nonlinear saturation processes. ^ Moreover, in the case of thick samples, the initial reduction of the electron temperature that is measured at the surface is mainly due to the fast electron diff*usion process,^ and unless correctly treated leads to an incorrect value of g. Thin samples, on the other hand, can have anomalously fast electron cooling due to electron-impurity and electron-surface scattering.This paper reports a new method of determining g. It relies on competition between two fast processes: electron-lattice energy exchange and electron thermal conduction. The former cools electrons heated with an ultrashort 10-/im pulse (penetration depth -200 A) by transferring their heat to the lattice. The latter removes energy from the surface.The paper has a second purpose. A theory is developed to describe the pulse-duration dependence of optical damage to metals, and experimental results are presented in support of the theory.The heat transport inside the metal can be described with the following one-dimensional, two-temperature model ^:Ce-l-Te-^^K^Te-g(Te-Ti) +A(x,t) (l) at ox ox andQ^Ti=g(Te-Ti),(2) ot where x is the direction perpendicular to the surface.The electron heat capacity Ce is given by Ce =CeTe, Ce being a constant,^ and K is the heat conductivity. According to Sommerfeld's model,^ KozTe/v, where v = Vee + Vei and Vee and Vei are the electron-electron and electron-phonon collision frequencies, respectively. Te and Ti are, respectively, the electron and lattice temperatures and d is the lattice heat capacity. A{xj) is a source term and g is the electron-phonon coupling constant.As a guide, we will derive some approximate scaling laws for the response of a metal to an ultrashort pulse. These scaling laws have two essential roles. First, they allow quantities that are necessary to understand the following measurements to be introduced. Second, they...

Supercontinua extending from the ultraviolet to the infrared are observed from high-pressure (1-40 atm) Ar, Kr, Xe, H2, N2, or CO2 illuminated with 2-psec or 70-fsec, 0.6-/xm pulses with an energy ;$500 /xJ. The blue spectral component is shown to display a nearly universal behavior for all gases and pulse durations. Although the maximum intensity of the focused, femtosecond pulse in an evacuated cell was -10 B W/cm 2 , continuum generation was only observed with the femtosecond pulse when the threshold for self-focusing was exceeded. PACS numbers: 42.65.Jx, Discovered in 1970, supercontinuum generation has now been demonstrated in a wide variety of solids and liquids. Self-phase modulation, four-wave mixing, and plasma production (for the blue spectral component) are the processes most commonly invoked to explain continuum generation. 2 Recent plane-wave theoretical models based on cubic nonlinearities 2 ' 3 describe continuum generation in terms of a parameter Q :as rf2E 2 z/cr, where 172 is the nonlinear refractive index, E is the incident laser field amplitude, z is the interaction length, T is the pulse duration, and c is the speed of light. For small Q, the continuum bandwidth is given by Aco -Qco. A value of Q ^ 1 is required for the generation of a broad (8v~~ v) supercontinuum. 2,3 Experimentally, self-focusing (the spatial analog of self-phase modulation) is difficult to avoid. Self-focusing limits the validity of current rj2 continuum theories to e<(4/r)(r ?2 P/cW) 1/2 , (1) where P is the laser power. Equation (1) is obtained by the requirement that the interaction length be less than the self-focusing length for a Gaussian beam and by the assumption of only full-beam self-focusing. Mierofilamentation lowers Q even further. 4 Although no continuum-generation experiments have clearly satisfied the inequality in Eq. (1), the parameter Q is used as a guide to estimate the expected spectral broadening. Based on this parameter, gases are not likely to be sources of supercontinua. For example, xenon, a highly nonlinear gas, has a nonresonant 772 "^x 10~~2 6 m 2 /V 2 at atmospheric pressure 5 (r] 2 of water is 1.5x 10~2 2 m 2 /V 2 ). If we consider the maximum field that can be applied to Xe in the absence of breakdown 6 (/ -10 13 W/cm 2 ) then, even with a 100-fsee (2 psec) pulse, an interaction length of 2=6 xlO 3 cm (1.2xl0 5 cm) is required for

No abstract

Static and optical field enhancement in metallic nanotips studied by two-photon photoemission microscopy and spectroscopy excited by picosecond laser pulses Appl. Phys. Lett. 94, 093508 (2009); 10.1063/1.3095480Experimental studies of generation and propagation of high frequency acoustic waves in various solid materials using ultraviolet picosecond laser pulses

No abstract