Cascading photoinduced, elastic, and thermal switching of spin states triggered by a femtosecond laser pulse in an Fe(III) molecular crystal

In this paper we put on solid ground the optical time-resolved study of spin-state photoinduced switching in an Fe(III) compound whose structural dynamics have been recently unveiled. We provide the experimental evidence of complex dynamics, occurring in succession, and spanning ten decades in time. We show that in addition to the ultrafast photoinduced molecular switching occurring on the subpicosecond timescale, there exists two additional switching processes significantly affecting the fate of the macroscopic material: one driven by elastic interactions as volume expansion occurs on a tens of nanoseconds timescale, and another one associated with heat diffusion and thermal switching taking place on a tens of microseconds timescale.

Figure 1

(a) Fraction of molecules in high spin state during thermal crossover, as extracted from magnetic susceptibility measurements (from Ref. 19). (b) Temperature dependence of heat capacity measured with DSC on a bulk sample. Sketched with arrows: temperature increase $\Delta$$T_1$ and $\Delta$$T_2$ induced at $T_1$ and $T_2$ with the same laser energy and the resulting thermal conversion of HS fraction, $\Delta$$X_{\mathit{HS}1}$${}^{\mu }$${}^{\mathrm{s}}$ > $\Delta$$X_{\mathit{HS}2}$${}^{\mu }$${}^{\mathrm{s}}$.

Figure 2

The timing set-up and the optical layout of the time-resolved pump-probe experiment. The thin lines show schematically the connections between different instruments synchronized to the trigger signals delivered by the delay pulse generator. The electronic trigger signals of different frequencies and phases are drawn for clarity. The delay pulse generator sets the desired timing pattern for the four lasers: amplification lasers (1) and (2) and the regenerative amplifiers RGA(1) and RGA(2). Any pump-probe delay, $t$($j$, $i$), being a multiple number of 13 ns (inverse of $f_{\mathit{RF}}$ $=$ 76 MHz) is thereby generated. Shorter pump-probe delays, from 100 fs to 1 ns, are attained by adjusting the optical path difference ($\Delta$$PL$) with a mechanical translation stage. The thick lines represent the laser beams as they propagate from the sources to the sample. The laser beams are polarized parallel to the crystalline axis $a$. The sample is cooled with nitrogen flow ($N_2$). The transmitted laser light is detected with two balanced photodiodes. The phase and intensity of the differential signals are recorded with a lock-in amplifier.

Figure 3

(a) Differential intensity of transmitted probe light ($\Delta$$I$/$I$) measured on a single crystal of [(TPA)Fe(TCC)]PF${}_6$ at 200 K by femtosecond excitation at 800 nm and 4 $\mu$J/pulse with probe at 600 nm, obtained with 110 fs time resolution inside 1 ns time window (open circles) and obtained by electronically gating two synchronized femtosecond lasers (full circles). (b), (c) Time-resolved x-ray diffraction experimental data taken from Ref. 18: lattice parameter $a$ (b) and variation of Debye-Waller factor $B$ between $t$ < 0 and $t$ > 0 (c). Solid lines are drawn to guide the eye.

Figure 4

Time-resolved differential intensity of light transmitted by the crystal ($\Delta$$I$/$I$) from −50 ns to 1 ms at different temperatures; pump and probe as described in Fig. 3a. Symbols are experimental data points. Solid lines are drawn to guide the eye.

Figure 5

(a) Temperature dependence of the laser switched molecular fraction, $\Delta$$X_{\mathit{HS}}$${}^{fs}$ (dots), compared to $X_{\mathit{LS}}$ (solid line). (b) Temperature dependence of the molecular fraction switched during microsecond step, $\Delta$$X_{\mathit{HS}}$${}^{\mu }$${}^{\mathrm{s}}$ (dots), compared to the model of thermal conversion due to laser heating (solid lines). The model calculations were performed at three different values of temperature increase, $\Delta$$T$ $=$ 10 K, 15 K, 20 K. The corresponding calculated $\Delta$$X_{\mathit{HS}}$${}^{\mu }$${}^{\mathrm{s}}$ at 180 K are pointed to with arrows.