Gas-phase spin relaxation ofXe129

Abstract

We have completed an extensive study of $^{129}\mathrm{Xe}$ longitudinal spin relaxation in the gas phase, involving both intrinsic and extrinsic mechanisms. The dominant intrinsic relaxation is mediated by the formation of persistent ${\mathrm{Xe}}_{2}$ van der Waals dimers. The dependence of this relaxation on applied magnetic field yields the relative contributions of the spin-rotation and chemical-shift-anisotropy interactions; the former dominates at magnetic fields below a few tesla. This relaxation also shows an inverse quadratic dependence on temperature $T$; the maximum low-field intrinsic relaxation for pure xenon at room temperature (measured here to be $4.6\phantom{\rule{0.3em}{0ex}}\mathrm{h}$, in agreement with previous work) increases by $\ensuremath{\approx}60%$ for $T=100\phantom{\rule{0.2em}{0ex}}\ifmmode^\circ\else\textdegree\fi{}\mathrm{C}$. The dominant extrinsic relaxation is mediated by collisions with the walls of the glass container. Wall relaxation was studied in silicone-coated alkali-metal-free cells, which showed long (many hours or more) and robust relaxation times, even at the low magnetic fields typical for spin-exchange optical pumping $(\ensuremath{\approx}3\phantom{\rule{0.3em}{0ex}}\mathrm{mT})$. The further suppression of wall relaxation for magnetic fields above a few tesla is consistent with the interaction of $^{129}\mathrm{Xe}$ with paramagnetic spins on or inside the surface coating. At $14.1\phantom{\rule{0.3em}{0ex}}\mathrm{T}$ and sufficiently low xenon density, we measured a relaxation time ${T}_{1}=99\phantom{\rule{0.3em}{0ex}}\mathrm{h}$, with an inferred wall-relaxation time of $174\phantom{\rule{0.3em}{0ex}}\mathrm{h}$. A prototype large storage cell ($12\phantom{\rule{0.3em}{0ex}}\mathrm{cm}$ diameter) was constructed to take advantage of the apparent increase in wall-relaxation time for cells with a smaller surface-to-volume ratio. The measured relaxation time in this cell at $3\phantom{\rule{0.3em}{0ex}}\mathrm{mT}$ and $100\phantom{\rule{0.2em}{0ex}}\ifmmode^\circ\else\textdegree\fi{}\mathrm{C}$ was $5.75\phantom{\rule{0.3em}{0ex}}\mathrm{h}$. Such a cell (or one even larger) could be used to store many liters of hyperpolarized $^{129}\mathrm{Xe}$ produced by a flow-through polarizer and accumulator for up to three times longer than currently implemented schemes involving freezing xenon in liquid nitrogen.