ion in an nn site, giving aa = 1.424. Repulsive energies are determined with a procedure detailed in reference 15 and applied to CaF2 in reference 13. The result for a lattice F-ion is ER = 2.76 eV. When the same equations are applied to the case of an interstitial F- ion, ER = 2.84 eV. The numbers just determined are listed in table 3, along with the values of EB found by using them in equation (3). Note that the binding energy of a lattice ion (Fs) calculated this way (11.42 eV) agrees quite well with the band-gap energy of CaF2 (12.2 eV [18]). This is as expected and lends credibility to the use of equation (3). Note also that the calculated binding energy of the F-i ion in an nn defect complex (4.37 eV) compares very well with the photon energy of the high-energy peak in the PC data (4.43 eV), which is proposed to be associated with nn defect complexes. Table 3. Quantities calculated for use in evaluating the binding Although the calculated values of EB for the F's and nn Fi ions agree well with the expected values, there remain some points that do not agree as well. For example, when the calculation is repeated for the nnn Fi ion, equation (3) yields EB = 2.13 eV, which is significantly less than the photon energy of the low-energy peak in the PC data (4.05 eV). 5. Nonlinear PC Results Nonlinear PC data measured as described in Section 3 (950-ns-wide laser pulses at 496.5 nm) were obtained from the Optovac sample and BDH samples A and C. The results are shown in figures 8 through 11. Note that the laser was focused inside the Optovac and BDH:A samples (figures 8 and 9) and both inside and outside the BDH:C sample (figures 10 and 11, respectively). Examination of the data in figures 8 through 11 reveals several interesting features. First, at low laser pulse energies, the slopes of all four plots are between 1 and 2, indicating the presence of charge-producing defects that absorb at 496.5 nm. This verifies the statement made in Section 4.1 that there was some very small but finite linear PC at 496.5 nm. A second interesting feature of the data is the nonlinear increase of photocurrent with increasing laser pulse energy in figures 9 through 11. Figure 8 shows no such behavior in the Optovac sample, probably because the pulse energies were simply not high enough to see the nonlinear effects. The nonlinear behavior seen in figures 9 through 11 suggests the onset of multiple-photon-induced PC, which adds to the one-photon linear effect at 496.5 Note that two-photon PC would occur at 248.3 nm, a wavelength at which the linear PC spectra of figures 2 through 6 show significant photoresponse. Three-photon effects at 165.5 nm are also possible; however, since the linear PC was not measured at this wavelength, only speculation to that possibility is suggested. nm. The effects of focusing the laser inside or outside the CaF2 sample are compared in figures 10 and 11. The illuminated volume was much smaller when the laser was focused inside the sample; the photon density was much greater than when the laser was focused outside the sample. Thus, the photocurrent saturation apparent in figure 10 (laser focused outside) is probably due to space charge effects in the larger illuminated volume. No such saturation is apparent in figure 11 (laser focused inside), probably because the photoinduced charge could easily diffuse away from the smaller illuminated volume. Note also that at any given laser pulse energy below the onset of space charge effects, the photoresponse was greater in figure 10 (laser focused outside) than in figure 11 (laser focused inside). The difference can be explained as a greater number of charge carriers generated by the unfocused beam. It is likely that the photon density in the focused beam was great enough to sufficiently deplete the low charge-producing defect population so that many photons were not absorbed. In the unfocused-beam situation with the same number of photons spread over a larger volume, less if any of this bleaching occurred, resulting in more absorbed photons and more photocurrent. Finally, note that the near-vertical slopes of the focused-laser data in figures 9 and 11 at laser pulse energies above 300 mJ suggest the onset of electron avalanche. The samples (BDH:A and C) were slightly colored (blue) at these pulse energies, but not catastrophically damaged, suggesting that nonlinear PC measurements may constitute a nondestructive method of estimating the laser damage threshold of an optical material. In figures 9 and 11, the vertical line represents the asymptote to the experimental data. This asymptote would be the laser damage threshold, i.e., catastropic failure, of the material. The laser damage threshold is approximately 500 mJ for these CaF2 samples. When the pulse width (950 ns) and spot size (estimated to be approximately 50 μm in diameter) of the laser beam are taken into account, the laser damage threshold is estimated to be approximately 27 GW/cm2. 6. Temporal Behavior of Pulse-Induced Photocarriers Figure 12 shows an oscilloscope trace of the voltage induced across the 216-M2 resistor in figure 1 during illumination of BDH sample A by a 400-mJ pulse from the 496.5-nm laser. From the photo, it is evident that the signal decayed with an e-1 time constant TM of approximately 440 μs. The leading edge of the signal in figure 12 is shown on a wider time base in the upper trace of figure 13, along with the response of a fast PIN diode to the laser pulse. The two signals were measured simultaneously, so the photo shows their true temporal relationship. Notice that the photocurrent signal begins simultaneously with the laser pulse and continues to rise until the laser pulse has fallen back to zero. This shows that the laser-induced photocurrent tracks the laser pulse and has a and has a rise time of approximately 1 μs. The 440-μs decay time Tм associated with the falling edge of the signal in figure 12 can be used to infer the recombination lifetime TR of the laser-induced photocarriers. Note that the CaF2 crystal and holder electrodes in figure 1 constitute a capacitor. Thus, the photocurrent can either decay through the 216-M resistor or recombine in the crystal. Actually, both avenues are used, so that the net rate of decay is the sum of the rates associated with both processes. Thus, we can write 1 1 1 TM = TRC + TR (5) where TM, TRC, and TR represent the measured decay time (440 μs), the RC time constant of the 216-M crystal capacitance combination, and the actual recombination lifetime of the photocarriers, respectively. A careful measurement of the crystal capacitance, including everything between the input and the output of the vacuum chamber in figure 1, showed its value to be 16.7 pF. Thus, TRC = 3.6 ms and TR = 0.5 ms. Although it has not been done at this point, it should be possible to associate this recombination lifetime with a specific type of defect and thereby gain additional information that would tend to support or dispute the proposed model. 7. Summary and Conclusions Linear photoconductivity (PC) measurements in CaF2 produced significant photocurrents in the spectral range from 275 to 350 nm. The linear PC results varied with the concentrations of certain trace element impurities and are thought to be due to chargeproducing defects that involve those impurities. A tentative model of the charge-producing defect is proposed. It involves electron removal from interstitial fluorine ions that enter to charge-compensate the trace-element impurities. A Born-model estimate of the electron binding energy of the proposed defect agrees well with the photon energy of the peak PC signal. Nonlinear PC measurements at 496.5 nm revealed a steady transition with increasing laser pulse energy from one to two-photon absorption to the possible onset of electron avalanche. The onset of coloration-not catastrophic damage at the higher laser pulse energies suggests that nonlinear PC can be used as a nondestructive estimator of the laser damage threshold. Measurement of the temporal decay of the photocurrent induced by 400-mJ laser pulses and consideration of the crystal capacitance showed the recombination lifetime of the photocarriers to be 0.5 ms. 8. References [1] Emmony, D. C. Private communication. [2] Lee, R. S.; Merklin, J. F.; Marrs, C. D.; Richter, M. H. "Impurity photoconductivity of y-irradiated LiF," Phys. Status Solidi (b) 103; 605; 1981. [3] Messner, D.; Smakula, A. 1162; 1960. [4] Scouler, W. J.; Smakula, A. 20°C and -190°C," Phys. Rev. "Color centers in alkaline-earth fluorides," Phys. Rev. 120; "Coloration of pure and doped calcium fluoride crystals at 120; 1154; 1960. [5] O'Connor, J. R.; Chen, J. H. "Color centers in alkaline earth fluorides," Phys. Rev. 130; 1790; 1963. [6] Zintl, E.; Udgard, A. Z. Anorg, u. allgem. Chem. 240; 150; 1939. [7] Southgate, P. D. "Anelastic and dielectric loss in yttrium-doped calcium fluoride," J. Phys. Chem. Solids 27; 1623; 1966. [8] Short, J.; Roy, R. "Confirmation of defect character in calcium fluoride-yttrium fluoride crystalline solutions," J. Chem. Phys. 67; 1860; 1963. [9] Stott, J. P.; Crawford, J. H. "Dipolar complexes in calcium fluoride doped with erbium," Phys. Rev. Lett. 26; 384; 1971. [10] Ure, R. W. J. Chem Phys. 26; 1363; 1957. [11] Heist, R. H.; Fong, F. K. "Maxwell-Boltzman distribution of M+3- F- interstitial pairs in fluoride-type lattices," Phys. Rev. B 1; 2970; 1970. [12] den Hartog, H. W. "Defect structure and defect-defect interactions in solid solutions of AlF2 and RbF3 doped with Gd3+ probes," Phys. Rev. B. 27; 20; 1983. [13] Poole, R. T.; Szajman, J.; Leckey, R.G.C.; Jenkin, J. G.; Liesegang, J. "Electronic structure of the alkaline-earth fluorides studied by photoelectron spectroscopy," Phys. Rev. B 12; 5872; 1975. [14] Born, M.; Meyer, J. E. Z. Phys. 75; 1; 1932. [15] Tosi, M. P. "Cohesion of ionic solids in the Born model," in Solid state phys.ics, Vol. 16, F. Seitz and D. Turnbull, eds. New York; Academic Press; 1964. [16] Handbook of chemistry and physics. CRC; Cleveland, OH; 1972. [17] Kittel, C. Introduction to solid state physics, 4th ed. New York; Wiley; 1971. P. 111. [18] Hayes, W., ed. Crystals with the fluorite structure. 1974. P. 13. London; Oxford University Press; |