The study of spin-polarized atomic hydrogen in the three-dimensional gas phase has been extended into a new regime of magnetic field strength. Analysis of decay data for B=10--20 T reveal for the first time the characteristic peak in the three-body dipolar-decay rate constant as predicted by Kagan, Vartanyants, and Shlyapnikov (Zh. Eksp. Teor. Fiz. 81, 1113 (1981) [Sov. Phys.---JETP 54, 590 (1981)]). Beyond 17 T we observe a rapid falloff of the three-body-decay rate constant. We also determine a value for the three-body-exchange rate constant which is in accord with the calculation of Kagan et al. In contrast to the picture which was emerging from previous studies at more moderate fields, the data presented here suggest that existing theories are capable of describing the decay of the system in the very-high-field limit. The new picture which is emerging suggests that Bose-Einstein condensation may be achievable at somewhat higher fields which are currently available at high-magnetic-field facilities.
In view of the failure of the Kagan dipole mechanism to explain the magnetic field dependence of the H+H+H recombination rate in spin-polarized atomic hydrogen, we consider an additional process, the so-called dipole-exchange mechanism. Two simple approaches to estimate its consequences turn out to be promising but the question of the role of different and higher-order processes remains open. We therefore turn to in principle, an exact approach to the three-body recombination, including all possible processes. The first numerical results of the approach are presented.
Three-body recombination in spin-polarized hydrogen has been observed. Doubly polarized hydrogen was compressed by a piston into a thin cell and the temperature and pressure were independently monitored. Measurements were carried out in the temperature range 0.3 to 0.45 K in a magnetic field of 8.0 T at densities up to 1.5\ifmmode\times\else\texttimes\fi{}
1018 ${\mathrm{cm}}^{\ensuremath{-}3}$. The results are in good agreement with the theory of Kagan, Vartanyants and Shlyapnikov.
We have studied relaxation and recombination processes in compressed, doubly polarized atomic hydrogen at temperatures from 0.13 to 0.60 K and magnetic fields from 3 to 9 T. The gas and surface dipole three-body recombination rate constants at a field of 7.6 T are measured to be, respectively,
Lg=8.9(8)\ifmmode\times\else\texttimes\fi{}${10}^{\mathrm{\ensuremath{-}}39}$
cm6 ${\mathrm{s}}^{\mathrm{\ensuremath{-}}1}$ and
Ls=1.2(4)\ifmmode\times\else\texttimes\fi{}${10}^{\mathrm{\ensuremath{-}}24}$
cm4 ${\mathrm{s}}^{\mathrm{\ensuremath{-}}1}$. They decrease slowly with field and
Ls exhibits no significant angular dependence. The three-body recombination rate due to hyperfine mixing has also been measured. Electronic and nuclear relaxation rates have been measured; the b-c electronic relaxation rate constant in the gas is
Gbc=1.03(7)\ifmmode\times\else\texttimes\fi{}${10}^{\mathrm{\ensuremath{-}}15}$
cm3 ${\mathrm{s}}^{\mathrm{\ensuremath{-}}1}$exp(-
Ebc/
kBT). The temperature and field dependence of the nuclear relaxation rate in the gas are observed to be in excellent agreement with recent theoretical calculations. Three-body surface recombination-rate measurements using $^{3}\mathrm{\ensuremath{-}}^{4}$He surfaces indicate that as little as one monolayer of
3He on the liquid
4He surface appreciably decreases the adsorption energy of atomic hydrogen. Densities achieved include 4.5\ifmmode\times\else\texttimes\fi{}
1018 atoms${\mathrm{?}}^{\mathrm{\ensuremath{-}}3}$ at 0.55 K (pure
4He walls), and 1.4\ifmmode\times\else\texttimes\fi{}
1018 atoms${\mathrm{?}}^{\mathrm{\ensuremath{-}}3}$ at 0.19 K (8 at. ? sup 3---He).
The three-body recombination rate in spin-polarized hydrogen has been measured at temperatures from 0.25 to 0.60 K in magnetic fields from 5 to 9 T. In contrast to theoretical expectation, the three-body rate is a weakly decreasing function of field in this region. At 7.6 T the gas and surface three-body rate constants are ${L}_{g}=7.5(3)\ifmmode\times\else\texttimes\fi{}{10}^{\ensuremath{-}39}$
cm6 ${\mathrm{s}}^{\ensuremath{-}1}$ and ${L}_{s}=2.0(6)\ifmmode\times\else\texttimes\fi{}{10}^{\ensuremath{-}24}$
cm4 ${\mathrm{s}}^{\ensuremath{-}1}$. The large value of
Ls can account for effects previously attributed to an anomalously large surface two-body nuclear relaxation rate.