Resolving the discrepancy between theoretically predicted and experimentally measured 129Xe nuclear spin polarization

By bringing spins out of thermal equilibrium, hyperpolarization techniques can enhance the nuclear magnetic resonance (NMR) signal by several orders of magnitude. Hyperpolarization makes the nuclear polarization independent of the magnetic field strength, enabling studies at lower magnetic field strengths or the detection of low-density nuclei. All hyperpolarization methods require a source of spin order that can be transferred to the nuclei of interest. Some methods such as dynamic nuclear polarization use the intrinsically higher spin order of electrons obtained under high field and low temperature conditions and transfer it to nuclei using microwave radiation. Others use the nuclear spin order of molecules like parahydrogen. Here I focus on the method of polarizing 129Xe called spin-exchange optical pumping that uses photons as its primary source of spin order. Xenon is an ideal NMR probe due to its inert nature, high solubility, and large electron cloud that makes it exceedingly sensitive to its physical and chemical environment. Hyperpolarized 129Xe is widely used as a magnetic resonance contrast agent for probing lung ventilation function. This technique recently gained approval by the United States Food and Drug Administration for clinical use. The large polarization achievable with current SEOP techniques makes micromolar concentrations of xenon MR-visible and recently enabled detection of gas exchange in the lungs, kidney and brain perfusion, and detection of thermogenic activity and perfusion in brown adipose tissue. The feasibility of arising biomedical imaging applications of hyperpolarized 129Xe depends on the polarization levels achieved with spin-exchange optical pumping. The optimization of spin-exchange optical pumping experimental conditions relies on modeling. However, the spin-exchange optical pumping model used for the past three decades has fallen short of accurately predicting experimental 129Xe polarization values. Specifically, theoretical models of spin-exchange optical pumping have overestimated experimental polarization values, leading to a search for additional 129Xe depolarization mechanisms, without much avail. Indeed, the proposed depolarization mechanisms were either unable to explain the discrepancy between theoretically predicted and experimentally measured 129Xe polarization values or used ad hoc parameters in the model to reproduce experimental values. The first part of my work seeks to address the discrepancy between theoretically predicted and experimentally obtained 129Xe polarization values. I show that an overestimation of the Rb vapor density, along with an underestimation of the spin-exchange cross section, led to this long-standing discrepancy. The second part of my work is focused on enabling optimization of 129Xe polarization values. I present a new setup to measure Rb polarization values directly during optical pumping. I then describe modifications to a current large-scale polarizer and provide standard operating procedures for filling optical cells. Finally, I describe how to transfer spin order from 129Xe to other nuclei.