Fe57NMR and relaxation by strong collision in the tunneling regime in the molecular nanomagnet Fe8

Abstract

$^{57}\mathrm{Fe}$ NMR measurements have been performed in single crystal and oriented powder of enriched $^{57}\mathrm{Fe}8$ molecular cluster in the temperature range $0.05--1.7\phantom{\rule{0.3em}{0ex}}\mathrm{K}$ in zero external field and with small perturbing field up to $1\phantom{\rule{0.3em}{0ex}}\mathrm{T}$ for both transverse and longitudinal orientation of $H$ with respect to the anisotropy axis. The $^{57}\mathrm{Fe}$ NMR spectrum is analyzed in terms of a dominant contribution due to the hyperfine interaction arising from core polarization. The measured temperature dependence of the resonance frequency is explained well by calculating the local average magnetic moment of the ${\mathrm{Fe}}^{3+}$ ion with a simple model which incorporates the effects of thermal average in the low lying energy states. Nuclear spin-lattice relaxation rate $(1∕{T}_{1})$ and spin-spin relaxation rate $(1∕{T}_{2})$ were investigated via temperature and field dependences. The obtained results are analyzed in terms of both intrawell thermal fluctuations of the hyperfine fields due to spin-phonon interaction, and interwell fluctuations due to phonon assisted quantum tunneling of the magnetization. It is argued that in zero external field and at low $T$ the $^{57}\mathrm{Fe}$ and the proton $1∕{T}_{1}$ is dominated by a strong collision relaxation mechanism due to the fact that phonon assisted tunneling transitions generate a sudden reversal of the local quantization field at the nuclear site. The data could be explained satisfactorily by assuming that the $^{57}\mathrm{Fe}$ $1∕{T}_{1}$ measures directly the effective tunneling rate. However, in order to fit the data we had to assume a larger in-plane anisotropy than previously reported, resulting in a bigger tunneling splitting in zero field. A comparison with published data of $^{55}\mathrm{Mn}$ in Mn12 indicates that a strong collision relaxation mechanism may apply also in Mn12. Finally the $H$ and $T$ dependence of $^{57}\mathrm{Fe}$ $1∕{T}_{2}$ is well explained simply in terms of thermal fluctuations of the magnetization without any tunneling contribution. At very low $T$ the $1∕{T}_{2}$ approaches a limiting value which can be explained in terms of the dipolar interaction between proton and $^{57}\mathrm{Fe}$ nuclei in the quasistatic regime.