Consequences of (129)Xe-(1)H cross relaxation in aqueous solutions

We have investigated the transfer of polarization from (129)Xe to solute protons in aqueous solutions to determine the feasibility of using hyperpolarized xenon to enhance (1)H sensitivity in aqueous systems at or near room temperatures. Several solutes, each of different molecular weight, were dissolved in deuterium oxide and although large xenon polarizations were created, no significant proton signal enhancement was detected in l-tyrosine, alpha-cyclodextrin, beta-cyclodextrin, apomyoglobin, or myoglobin. Solute-induced enhancement of the (129)Xe spin-lattice relaxation rate was observed and depended on the size and structure of the solute molecule. The significant increase of the apparent spin-lattice relaxation rate of the solution phase (129)Xe by alpha-cyclodextrin and apomyoglobin indicates efficient cross relaxation. The slow relaxation of xenon in beta-cyclodextrin and l-tyrosine indicates weak coupling and inefficient cross relaxation. Despite the apparent cross-relaxation effects, all attempts to detect the proton enhancement directly were unsuccessful. Spin-lattice relaxation rates were also measured for Boltzmann (129)Xe in myoglobin. The cross-relaxation rates were determined from changes in (129)Xe relaxation rates in the alpha-cyclodextrin and myoglobin solutions. These cross-relaxation rates were then used to model (1)H signal gains for a range of (129)Xe to (1)H spin population ratios. These models suggest that in spite of very large (129)Xe polarizations, the (1)H gains will be less than 10% and often substantially smaller. In particular, dramatic (1)H signal enhancements in lung tissue signals are unlikely.