Relaxation processes in optically excited high-Tc films

Physica B 165&166 (1990) 1507-1508 North-Holland RELAXATION PROCESSES IN OPTICALLY EXCITED HIGH-Tc FILMS A.M. KADIN, P.H. BALLENTINE', and W.R. DONALDSON Department of Electrical Engineering and Laboratory for Laser Energetics, University of Rochester, Rochester, New York 14627, USA We have observed the transient electrical response of current-biased YBa2Cu307 films irradiated with 150 ps pulses from a Nd:YAG laser (A=1.06I!m). The magnitude of the response always agrees quantitatively with a simple heating model, as does the speed of the response «2 ns) in optically thin films «=0.2I1m). However, an equally rapid response in some thicker films is faster than thermal diffusion through the film can explain. At low temperatures «20 K), this rapid response provides indirect evidence for nonequilibrium hot-electron transport across the film on the 1 ps timescale. At higher temperatures, a similarly fast response can be achieved only by biasing the film close to the critical current, where rapid current redistribution can account for the results. There has been substantial interest recently in the effects of optical radiation on the electrical properties of high-Tc superconducting films such as YBa2Cu307 (YBCO). The key question is whether the response is due simply to heating the film, or alternatively if a nonequilibrium process must be invoked. Experiments to date have come down on either side of the issue.(1) We have measured the fast response of thin YBCO films to short pulses (150 ps) from a Nd:YAG laser (A=1.06I1m).(2,3) In general, we see a rapid voltage rise (=1 ns), followed by a slow recovery to the superconducting state, as measured using a fast 300 MHz oscilloscope (see Fig. 1). The samples tested included sputtered films (4) from CVC Products, Inc. (Rochester, NY) as well as coevaporated films (5) from J.H. Scofield of Oberlin College, Ohio. For this study, each film was patterned into a crossbar structure, with a central bridge region 2 mm long and 100 - 200l1m wide. The patterning was carried out using the above Nd:YAG laser at higher intensity.(6) In identifying the various active physical processes, it is useful to estimate the relevant characteristic distances and times (see Table I). The two key distances are the film thickness d= 0.1 - 111m, and the optical penetration depth 0 = 0.1I1m. The relation between these two values determines whether the film is optically thin (d<o) or optically thick (d»o). In the simplest model of the optoelectronic response, the incoming photons produce highly excited carriers, which rapidly thermalize with low energy phonons to create local heating in the film. Sufficient 'Present Address: CVC Products, Inc., Rochester, NY 0921-4526/90/$03.50 24,..------ >' 21 18 S l v.I ~6 --, :~ fo9 til >3 o o 40 80 120 160 200 Time (nsec) sputtered YBCO FIGURE 1. Fast response of 0.2 ~m epitaxial film on MgO to 0.5 mJ/cm pulse at T=35 K. Table I: Characteristic Response Times Physical process Electron-phonon time Current redistribution time Optical pulsewidth Electrical risetime Thermal cooling time Time [ps] 1-100 0.01 150 <2000 - >10 4 >10 4 heating drives the superconductor locally into the resistive state. If the sample is optically thin, this heating should be homogeneous, and a voltage response on the time scale of the optical pulsewidth should result. This accounts for the fast voltage rise observed. The much slower recovery of the superconducting state is due to thermal conduction out of the film through the substrate, with a characteristic cooldown time = d 2/D =10 ns for a film with thickness d=0.3I1m and thermal diffusivity 0=0.1 cm 2/s. © 1990 - Elsevier Science Publishers B.V. (North-Holland) A.M. Kadin, P.H. Ballentine, W.R. Donaldson 1508 "hot electrons" can rapidly spread across the thickness of the film. Recent estimates of this relaxation time (7) suggest that 'te-ph ~ 160 psIT [in K], so that using an electron diffusivity ~1 000 cm 2/s and 30 ~...... 20 1 tiD • i;ii ~ • • • • • • • 'te-ph=8 ps (at 20 K) gives a diffusion length ~ 1~m • 0 0.0 O.S 1.0 1.5 2.0 Laser Fluence (mJ/r:m 2 ) FIGURE 2. For the sample in Fig. 1, the dependence of the fast voltage rise on laser fluence/pulse. The dots are experimental values; the solid line results from fitting the de resistance to a uniform heating model. In Fig. 2, the magnitude of the fast voltage rise is plotted against the laser f1uence. This is compared to a simple uniform heating model with no free parameters, in which the film temperature rise is determined, and the measured resistance curve used to calculate the response.(2) This agreement indicates that the response is essentially thermal on the ns scale. For optically thick samples, one expects that substantial heating should be present only at the top surface, tailing off exponentially into the film. Driving the top surface of the film into the resistive state will lead to a very fast current redistribution to the deeper portion of the film that remains superconducting, on a time ~ 02/D em ~ 10 fs, where For small· currents, there should be no Dem = p/~o' voltage until the heat has penetrating through the film, and thermal diffusion is rather slow in this material. For near-critical currents, however, the current redistribution can exceed the critical current in the remainder of the film, thus very rapidly driving the entire film into a resistive state. In this way, by varying the current in the film, we observe a change from a very slow response to a very fast response.(3) In contrast, we see a different trend in optically thick films with low transition temperatures; the response is always rapid, even for small currents, although the magnitude of the response remains thermal. We therefore infer that the heat is distributed through the film more rapidly than allowed by equilibrium thermal diffusion. Our proposed explanation is as follows: the photons are depositing their energy primarily into the electron sub-system, creating an elevated temperature as compared with the low-energy thermal phonons.(7) Before the final thermalization takes place, these nonequilibrium at low T, comparable to d. In summary, we have observed a fast voltage rise across current-biased YBCO films illuminated with a fast optical pulse. These included sputtered and evaporated films, high-quality epitaxial and granular films. In all cases, for optically thin films, the results are fully consistent with uniform heating of the film. For optically thick films, a fast response can be obtained at high temperatures only by biasing the film close to the critical current, in which case rapid current redistribution can account for the observed signal. At lower temperatures, however, a rapid response is achieved even for currents much less than Ic ' suggesting that nonequilibrium electron transport may be responsible for distributing the energy through the film. Finally, by illuminating the film with a much shorter pulse ~ 1 ps, it should be possible to observe a direct nonequilibrium response, consisting of a rapid heating of the electrons, and a comparably fast cooling via local electron-phonon relaxation. ACKNOWLEOOEMENTS This research was supported by NSF DMR8913524 and by the Laser Fusion Feasibility Project at the UR Lab. for Laser Energetics. We would also like to thank J. Allen and R. Rath at CVC Products, Inc. of Rochester, New York, for sputtered films, and Dr. J.H. Scofield of Oberlin College, Ohio, for evaporated films. REFERENCES (1) A.1. Braginski, M.G. Forrester, and J. Talvacchio, Ext. Abstr. of Int. Superconductivity Electronics Conf. (ISEC '89), Tokyo, June 1989, p. 482 (Japan Soc. of Appl. Phys., Tokyo, 1989). (2) W.R. Donaldson, et al., Appl. Phys. Lett. 54 (1989) 2470. (3) P.H. Ballentine, et al., in Proc. Amer. Vac. Soc. ConI. on High Tc Thin Films, Boston, MA, October (4) (5) (6) (7) 1989 (Amer. Inst. of Physics, 1990). A.M. Kadin, P.H. Ballentine, J. Argana, and R.C. Rath, IEEE Trans. Magn. 25 (1989) 950. P. M. Mankiewich, et al., Appl. Phys. Lett. 51 (1987) 1753. P.H. Ballentine, et al., IEEE Trans. Magn. 25 (1989) 950. E.M. Gershenzon, et al., Proc. Int. Cont. High Tc Films & Single Crystals, Ustron Poland, 1989.