Exited atoms may relax to the ground state by radiative decay, a process which is usually very fast (of order nanoseconds). However, quantum-mechanical selection rules can prevent such rapid decay, in which case these "metastable" states can have lifetimes of order seconds or longer. In this Letter, we determine experimentally the lifetime of the longest-lived neutral atomic state-the first excited state of helium (the 2(3)S1 metastable state)-to the highest accuracy yet measured. We use laser cooling and magnetic trapping to isolate a cloud of metastable helium (He*) atoms from their surrounding environment, and measure the decay rate to the ground 1(1)S0 state via extreme ultraviolet (XUV) photon emission. This is the first measurement using a virtually unperturbed ensemble of isolated helium atoms, and yields a value of 7870(510) seconds, in excellent agreement with the predictions of quantum electrodynamic theory.
We describe an approach to detecting ionizing radiation that combines the special properties of superfluid helium with the sensitivity of quantum optics techniques. Ionization in liquid helium results in the copious production of metastable He2 molecules, which can be detected by laser-induced fluorescence. Each molecule can be probed many times using a cycling transition, resulting in the detection of individual molecules with high signal to noise. This technique could be used to detect neutrinos, weakly interacting massive particles, and ultracold neutrons, and to image superfluid flow in liquid 4He.
We have studied the dynamic properties of the
He(23S) and the ${\mathrm{He}}_{2}(a^{3}\ensuremath{\Sigma}_{u}^{+})$ atomic and molecular metastable states of helium that are produced in electron-bombarded superfluid helium. The steady-state concentration of ${\mathrm{He}}_{2}(a^{3}\ensuremath{\Sigma}_{u}^{+})$ molecules is shown to be limited by collisions between pairs of these molecules. Measurements of the rate coefficient for this bimolecular reaction as a function of temperature and pressure show that the reaction is diffusion limited and that the diffusion coefficient for the metastable molecules in the superfluid is determined by ${\mathrm{He}}_{2}(a^{3}\ensuremath{\Sigma}_{u}^{+})$-roton collisions. The bilinear destruction of metastable molecules has been monitored for times as long as 0.1 sec after the initial population of the state, and this establishes a lower bound of 0.1 sec on the radiative lifetime of the $(v=0) {\mathrm{He}}_{2}(a^{3}\ensuremath{\Sigma}_{u}^{+})$ state in superfluid helium. The lifetime and decay mode of the
He(23S) atoms are distinctly different from that of the corresponding ${\mathrm{He}}_{2}(a^{3}\ensuremath{\Sigma}_{u}^{+})$ molecule. In particular, the characteristic lifetime of the
He(23S) state in superfluid helium is essentially independent of the temperature between 2.08 and 1.4 \ifmmode^\circ\else\textdegree\fi{}K and is measured to be about 15 \ensuremath{\mu}sec. Our experimental observations indicate that the destruction of metastable atoms results in the population of high vibrational levels of the $a^{3}\ensuremath{\Sigma}_{u}^{+}$ molecular state. However, it is shown that the production of $(v=0) {\mathrm{He}}_{2}(a^{3}\ensuremath{\Sigma}_{u}^{+})$ molecules responds immediately to the electron beam (response time 3 \ensuremath{\mu}sec) and thus the formation of this state in the liquid does not depend upon the destruction of metastable atoms. Despite the fact that the lifetimes and decay modes of the metastable
He(23S) atoms and $(v=0) {\mathrm{He}}_{2}(a^{3}\ensuremath{\Sigma}_{u}^{+})$ molecules are not related, the production rates for these states are nearly equal and correspond to about 450 and 500 per incident 160-keV electron, respectively.
Long-lived metastable ${\mathrm{He}}_{2}^{\ensuremath{-}}$ negative ions have been observed for the first time. They were produced from
He2+ by two-step electron capture in a Cs vapor target. The apparent autodetachment decay rates have been measured at several delay times after formation, and indicate the presence of more than one state, with the lifetime of the major component exceeding 100 \ensuremath{\mu}s. The most likely configurations are ${^{4}\ensuremath{\Sigma}_{g}}^{+}$ and/or $^{4}\ensuremath{\Pi}_{g}$ states bound below ${\mathrm{He}}_{2}({^{3}\ensuremath{\Sigma}_{g}}^{+})$.
A time-of-flight technique has been used to measure the radiative lifetime of the
21S0 state of helium. After passing from a cooled source, helium atoms are excited by a pulsed anti-parallel electron beam. The atomic beam then contains not only ground-state helium atoms but also metastable atoms in both the
23S1 and the
21S0 states. The metastable atoms are detected at two positions, 1.9 and 6.7 m from the electron gun, by Auger ejection of an electron from a copper target, the first of which is a 60% transmitting mesh. Ground-state helium atoms cannot cause an electron to be ejected and hence are not detected. The time-of-flight distribution for the metastable
21S0 state is separated from that of the
23S1 state by illuminating the beam with an rf-discharge helium lamp. The
21S0 state is quenched by resonant absorption of a 20 581 - \AA{}A photon, raising the atom to the
21P1 state, which then decays preferentially to the
1S0 ground state; the
23S1 state remains unaffected because it is the ground state for the triplet system. The time-of-flight distribution for the
21S0 atoms is therefore obtained from the difference between the full beam and the quenched beam. A comparison of the number of
21S0 metastables within a given velocity interval at the two detectors determines the number which decay in flight and yields a value for the two-photon radiative lifetime. The value of the singlet lifetime for both
He3 and
He4 is 19.7 \ifmmode\pm\else\textpm\fi{} 1.0 msec, agreeing with the theoretical value of 19.5 msec, but disagreeing with the value 38 \ifmmode\pm\else\textpm\fi{} 8 msec measured by Pearl using a movable detector. The 1.0-msec error is an estimate of the remaining systematic errors in the experiment.