GAS-DYNAMIC EFFECTS IN THE LASER-PULSE SPUTTERING OF ALN - IS THERE EVIDENCE FOR PHASE EXPLOSION

The overall light intensity (fluorescence) of the sputtered atoms, ion s, and molecules has been measured for polycrystalline AlN which was b ombarded with 248 nm laser pulses in the presence of a background pres sure of N-2. AlN is unusual in that, in spite of a similar to 6 eV ban d gap, it is easily rendered highly absorptive of 248, 308, or 694 nm laser pulses. In fact, since it is well established in other work that similar to 1.5 J/cm(2) (308 nm) brings AIN to the melting temperature , similar to 3050 K, we will assume that the fluence used here(similar to 20 J/cm(2)) was more than enough to bring the target surface first to a temperature sufficient for normal vaporization but finally to th e vicinity of the thermodynamic critical temperature, T-tc. As a resul t a significant quantity of particles can be assumed to have been expe lled by phase explosion. The tentativeness in the argument rests in th e problem that some part of the incident fluence beyond similar to 1.5 J/cm(2) will have been consumed in laser-plume interaction. Neverthel ess there is evidence in work by Pedraza et al. [A.J. Pedraza, J-Y. Zh ang, H. Esrom, Mater. Res. Sec. Symp. Proc. 285 (1993) 209] that both AIN and Al respond linearly to the fluence up to at least 6 J/cm(2). I t was found that the assumed phase-exploded particles decelerated rapi dly, possibly due to their encounter with the normally vaporized parti cles, or possibly due to an electric field arising from positive charg ing of the target surface. The fluorescence maximum (which can be safe ly assumed to be also a density maximum) was then nearly stationary, a situation which characterized the lowest background pressures of N-2 (less than or equal to 3.5 Pa). At higher pressures (greater than or e qual to 3.5 Pa) a second fluorescence maximum appeared nearer the cont act front and was found to move. Following the suggestion of Horwitz ( 1) we take this feature as being an artifact of electrons near the con tact front diffusing (or scattering) backwards and causing fluorescenc e which is unrelated to the particle density. From the velocity of the contact front one obtains explicit information on the mean kinetic en ergies (E-4) of the particles in the plume (1.5-2 eV), Another estimat e of E-4 follows from the initial expansion observed from 0-200 ns (1. 5-3 eV). Such energies suggest, independently of the fact that the flu ence was high (similar to 20 J/cm(2)), that a temperature near T-tc wa s reached and that phase explosion may have occurred. We finally note that, however tentative is the claim for phase explosion, it is certai n that a close relative of phase explosion, due to subsurface heating, was not involved. This is because the numerical demonstrations of sub surface heating have been flawed. (C) 1998 Elsevier Science B.V. All r ights reserved.