Coherent Manipulation and Decoherence of $S=10$ Single-Molecule Magnets

We report coherent manipulation of $S=10$ ${\mathrm{Fe}}_8$ single-molecule magnets. The temperature dependence of the spin decoherence time $T_2$ measured by high-frequency pulsed electron paramagnetic resonance indicates that strong spin decoherence is dominated by ${\mathrm{Fe}}_8$ spin bath fluctuations. By polarizing the spin bath in ${\mathrm{Fe}}_8$ single-molecule magnets at magnetic field $B=4.6\mathrm{T}$ and temperature $T=1.3\mathrm{K}$, spin decoherence is significantly suppressed and extends the spin decoherence time $T_2$ to as long as 712 ns. A second decoherence source is likely due to fluctuations of the nuclear spin bath. This hints that the spin decoherence time can be further extended via isotopic substitution to smaller nuclear magnetic moments.

Figure 1

(a) Schematic diagram of ${\mathrm{Fe}}_8$ molecule. The ${\mathrm{Fe}}_8$ molecule consists of eight Fe(III) ($S=5/2$) ions which couple to each other to form an $S=10$ ground state. (b) Energy diagram as a function of the magnetic field, ${\mathit{B}}_0$, while the magnetic field is applied along the easy axis. The diagram is calculated using Eq. (1). (c) cw EPR spectrum at 20 K taken by sweeping magnetic fields from 4 T to 10 T. The magnetic field is applied along the easy axis within 10 degrees. Corresponding EPR transitions are indicated by solid lines in Fig. 1b. The spectrum also shows fine structures which are more pronounced in the transitions of high $|m_S|$ values. These were also seen in previous studies [1]. A resonance at 8.5 T is a spurious signal from impurities in the sample holder.

Figure 2

(a) Echo signals and echo decay as a function of $2\tau$. The solid line is a fit using a single exponential. The inset shows echo-detected EPR signals as a function of the magnetic field. (b) Echo decay measured by a stimulated echo sequence. The inset shows a measurement of the echo decay in a short time scale. The long decay curve was fit by a single exponential $e^{-T/T_{\mathrm{long}}}$ as shown in solid lines, while the short decay curve was fit by a double exponential $A{e}^{-T/T_{\mathrm{long}}}+B{e}^{-T/T_{\mathrm{short}}}$ with $T_{\mathrm{long}}=945\mu \mathrm{s}$.

Figure 3

Temperature dependence of $1/T_2$. Data with error bars are shown by square dots, and a simulation of spin bath decoherence is shown by a solid line. The inset shows a plot of $T_2$ vs temperature.