The contents of this thesis include several experiments and a theory, the unifying theme of which is resonant recombination of atomic hydrogen (H) and deuterium (D) at low temperatures. Measurements of the interaction of H and D with the liquid helium (l-⁴He) wall coating used in low-temperature experiments are also included. Atomic hydrogen and deuterium were studied by ESR in a field of 41 kG at temperatures below 1 K, in a second generation of experiments on an existing apparatus. The binding energy of H on l-⁴He was measured, by direct observation of the physisorbed atoms, to be E[sub B]= 1.00(5) K. Extensive meaurements of recombination rate constants were made. Resonant recombination of H via the (v, J) = (14,4) level of H₂ was observed and identified. Atomic deuterium was also observed to recombine in a resonant process. At low temperatures, resonant recombination is possible due to the predissociation of weakly bound molecular levels by the intra-atomic hyperfine interaction. The theory of resonant recombination of H and D due to hyperfine predissociation is developed in detail. In order to keep the development self-contained, it is presented as appendix D of this thesis. The theory is combined with our ESR data in order to predict the magnetic-field dependence of resonant recombination of D. It is shown that observation of the threshold magnetic fields for resonant recombination will permit accurate measurement of the dissociation energy of the (v, J) = (21,0) and perhaps (21,1) levels of the D₂ molecule. An apparatus was constructed to study atomic deuterium by hyperfine magnetic resonance at temperatures near 1 K. The D atoms were produced from D₂ of 99.65% isotopic purity by an RF discharge in a bulb coated with a film of superfluid l-⁴He. Magnetic resonance in a field of 39 G on the longitudinal β-δ transition at its minimum frequency of 309 MHz was used to observe the atoms. The experiment was designed with the capability of measuring the recombination rate as a function of magnetic field for fields up to 30 kG by moving the sample between the 309 MHz resonator and a superconducting solenoid. A number of preliminary experiments were made using this apparatus. It was discovered that the D atom density decays away exponentially in time. The decay rate was measured as a function of temperature and found to follow an Arrhenius (thermally-activated) form, with an activation energy of about 15 K. No second-order decay was seen, permitting an upper bound to be placed on the D-D recombination rate. The exponential decay of the D atom density was due to the penetration by the atoms of the l-⁴He film coating their container. Our measurement of the temperature dependence of the density decay rate gives the energy required to dissolve a D atom in l-⁴He, Es[sub s] ≃ 14 K. This is the first experimental determination of this quantity. The longitudinal relaxation time T₁ for the β-δ transition was found to be too short to be due to D-D spin-exchange. It was conjectured that the sample was contaminated with substantial amounts of H, spin-exchange in H-D collisions causing the short T₁. Measurements in an existing 1420 MHz apparatus showed directly, by magnetic resonance on the hydrogen zero-field transition, that H atoms were, in fact, present. By varying the H concentration, we were able to show that the decay of the D density by was not due to recombination of H with D. Several attempts to substantially reduce the H impurity have failed. The penetration of the l-⁴He film is undoubtedly one reason why the D is so susceptible to contamination by H, since the film is permeable to D but not to H. For use in interpreting these experiments, the effects of spin-exchange relaxation on hyperfine resonance in mixtures of H and D were calculated.
Spin-polarized atomic hydrogen (H↓) gas at low temperatures (<0.5 K) has been studied using Electron Spin Resonance (ESR) techniques which allow direct measurements of the densities of H atoms in each of the two lower hyperfine states (a and b). It has been demonstrated that ESR can be made to yield accurate and detailed information about the decay of H↓ densities in time. Crucial to the analysis of the data is the ability to confine the sample of H↓, gas inside the liquid ⁴He coated microwave cavity where the magnetic field and temperature are well known. Measurements of K[sub aa] and K[sub ab], the two surface recombination rate constants, and their ratio γ=K[sub aa]/K[sub ab] have been made and from these results the binding energy of H on liquid ⁴He has been extracted. Also reported are the results of the first detailed study of the one-body surface relaxation rate which we find is due to microscopic magnetic impurities in the cell walls. Theoretical calculations of the two-body dipole-dipole relaxation rate of H↓ in the gas and on the surface are also presented.
A calculation is presented of the relaxation rate ${{T}_{1}}^{\ensuremath{-}1}$ between the lowest two hyperfine states of a dilute gas of H atoms in high magnetic field at low temperatures. Dipole-dipole interactions dominate ${{T}_{1}}^{\ensuremath{-}1}$, leading to long relaxation times at low temperatures (${{T}_{1}}^{\ensuremath{-}1}\ensuremath{\approx}{10}^{\ensuremath{-}2}{n}_{\mathrm{H}}{T}^{\frac{1}{2}}$ ${\mathrm{sec}}^{\ensuremath{-}1}$ at $H\ensuremath{\approx}100$ kG). The recombination rate due to wall collisions is much greater than ${{T}_{1}}^{\ensuremath{-}1}$ and can lead to a depletion of the lowest hyperfine state from thermal equilibrium and an effective recombination rate equal to ${{T}_{1}}^{\ensuremath{-}1}$.
In sufficiently large magnetic fields, the weakly bound levels of H2 acquire finite lifetimes owing to hyperfine predissociation and become manifest as scattering resonances in spin-polarized atomic hydrogen (H↓). In this work we investigate in detail the contribution to the reaction kinetics in H↓ due to the resonance, denoted [Formula: see text], arising from the predissociation of the (v,J) = (14,4) level of H2. To characterize the formation and disintegration of [Formula: see text], we calculate the predissociation as a function of both the applied magnetic field and the dissociation energy, D, of the level. Recombination occurs when [Formula: see text] is stabilized in collisions that cause transitions to lower molecular levels. For low H↓, densities, collisions with atoms of the saturated 4He vapour dominate; we calculate the cross section in the distorted wave Born approximation and also use a coupled-channels approach. At higher densities, stabilization occurs in collisions with H↓ as well; the effect of proton exchange between the atom and the molecule is derived. Previous experiments, analyzed neglecting proton exchange, are reanalyzed including this; the dissociation energy D is not as well determined by the data as was previously thought. Prospects for future experiments are discussed.
Spin-polarized, atomic hydrogen at temperatures below 1K shows promise of becoming the first physically realizable example of a degenerate, weakly interacting Bose gas. Because of the hyperfine structure of the hydrogen atom the Bose condensed state is expected to exhibit novel magnetic properties such as spin wave excitations of the superfluid and possibly a coherent precessing, spontaneous magnetization. This paper discusses a variety of theoretical problems which arise in the study of atomic hydrogen gas. Theories are presented for spin relaxation of the hyperfine levels and for the recombination of H-atoms into H2 molecules. The problem of the binding of H atoms to a surface of superfluid 4He is reviewed together with the theory of surface superfluidity in the hydrogen film proposed by Edwards and Mantz. Finally we consider the problem of superfluidity in the bulk atomic hydrogen gas. We describe the theory of Siggia and Ruckenstein of double Bose condensation into the two lowest hyperfine states and speculate about the possible observable consequences of this phenomenon.
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).