Vibrational autodetachment of sulfur hexafluoride anions at its long-lifetime limit

We have investigated the autodetachment of electrons from rovibrationally hot SF${}_6{}^{-}$ anions using a cryogenic ion-beam trap. Extremely low residual gas densities of ${10}^4$ cm${}^{-3}$ provided undisturbed observation of the neutralization rates due to vibrational autodetachment (VAD) over almost five orders of magnitude and over times up to 100 ms. We successfully explain our experimental decay curves using statistical rate theory combined with electron attachment data and vibrational frequencies calculated for a $C_{4v}$-distorted SF${}_6{}^{-}$. The unprecedented sensitivity of the experiment to the decay constants at the VAD threshold allows us to infer from the data the adiabatic electron affinity of SF${}_6$ to be $(0.91\pm 0.07)$ eV and to confirm the recently predicted $C_{4v}$ symmetry of SF${}_6{}^{-}$.

Figure 1

Measured neutralization rate $R\left(t\right)$ of SF${}_6{}^{-}$ as a function of the storage time $t$ (dots). Previous storage ring data [9], rescaled in amplitude to match the present data, are shown for comparison (dashed line). The solid line is a fit of the present data for $t\le 5$ ms by a power law ($\propto t^n$) with $n=-1.4$. Upper inset: Neutralization rate for long storage times (symbols) and exponential fit (solid curve). The dotted line represents the mean dark count rate of the MCP. Lower inset: Schematic ground state potential energy curves of SF${}_6{}^{-}$ and SF${}_6$ as a function of a single S-F-bond distance and decay paths of excited SF${}_6{}^{-}$ ions (VAD: vibrational autodetachment, Diss: dissociation into ${\mathrm{SF}}_5{}^{-}+F$, and Rad: radiative cooling).

Figure 2

(a) VAD rate coefficients $k(E,J)$ from Eq. (2) for different EA, as functions of $E_v=E-E_r^{-}\left(J\right)$; solid lines for $J=0$ and dotted lines for $J=150$. (b) Vibrational energy distribution $f_v(E_v,t)$ of the stored SF${}_6{}^{-}$ ions for various storage times (labeled) and strongly different initial vibrational temperatures $T_v$ ($J=0$). The normalization was set to make the heights of the 1 ms curves match at the energy of the first vibrational level in SF${}_6$ ($E_v^0=43$ meV). (c) VAD decay curves $R_{\mathrm{VAD}}\left(t\right)$ from Eq. (3) with $\mathrm{EA}=0.90$ eV and different $T_v$, $J$, and $k_{\mathrm{rad}}$.

Figure 3

Background-subtracted VAD decay rates (symbols with statistical error bars) observed at four different SF${}_6$ source pressures (highest to lowest pressures for top to bottom curves) and $V_c=800$ V. Solid curves are fits of Eq. (3) using $\mathrm{EA}=0.90$ eV and $k_{\mathrm{rad}}=25$ s${}^{-1}$ and yielding $T_v$ and $T_r$ as shown in Fig. 4. Inset: Semilogarithmic plot of $R\left(t\right)$ for times $t>10$ ms.

Figure 4

(a) Global reduced ${\chi }^2$/DOF from the fits of all ten VAD decay curves over a grid (dots) of globally set EA and $k_{\mathrm{rad}}$ values, with the colors and contours obtained by interpolation. (b) Initial temperatures derived for vibrations (circles) and rotations (squares) from fits of the measured VAD decay curves with Eq. (3), setting $\mathrm{EA}=0.90$ eV and $k_{\mathrm{rad}}=25$ s${}^{-1}$. Full symbols: $V_c=800$ V; open symbols: $V_c=2000$ V.