A semilog plot of the magnitude of the free induction decay (FID) of an enriched xenon sample, showing the characteristic beat pattern and a signal decay over 5 orders of magnitude. Using a combination of spin exhange optical pumping and phase exchange in a convection cell, we are able to routinely generate 10-12% nuclear polarizations in a enriched solid polycrystalline Xe-129 sample. Capitalizing on our unique ability to produce such extremely large signals, we explore the behavior of the spin-spin interactions in solid xenon in the extremely long-time regime (out many times T2). In this virtually unexplored regime, we study spin correlations and system behavior in an attempt to illuminate long standing questions in solid state theory.
This image shows the long time behavior of decaying spins, all initiated with different spin configurations. Note how the long time behavior of each of the signals tends to the same oscillation frequency and decay constant.
Xe reversibly binds to protein via weak Van der Waals forces resulting in a chemical exchange between solution and protein-bound states. In the fast-exchange limit, one detects a single NMR peak that is chemically shifted as a function of protein concentration. Specific binding occurs at hydrophobic sites with high Xe affinity; non-specific binding refers to all other interactions. BPTI (Bovine Pancreatic Trypsin Inhibitor) is a good model system to study protein dynamics using 129Xe NMR as a probe. It is extensively studied and has many well characterized mutants. Recently, we've studied four different types of BPTI: Wild type, Y23A, Y35G, and F45S. Wild type is believed to not have a specific binding site. Y23A, F45S, and Y35G have manufactured hydrophobic cavities. However, Y35G is known to undergo conformational changes when in solution phase on time scales of microseconds3 and it is not known what fraction of Y35G is folded appropriately to accept Xe.
A home-built flow-through polarizer provides 10 sccm of 129Xe hyperpolarized to ~10%. Hyperpolarized gas boosts signal to noise and, coupled with a high resolution NMR spectrometer, enables us to measure sub-ppm chemical shifts to plus/minue 0.03 accuracy at very low Xe (92-96 microMole) and protein concentrations (0-0.8 mM). To deliver the gas to the protein solution without bubbling, we employ a membrane system suggested by Baume et al. Our adaptation employs 10 inlet and outlet tubes of hydrophobic Celguard© that have ~50 micrometer walls which allow gas to diffuse through without appreciable relaxation.
Data shows a linear dependence at low concentrations, which is consistent with single site fast exchange interaction. The slopes are proportional to binding strength assuming similar chemical shifts for all bound states.
Consistent with a rigid lattice and a manufactured, hydrophobic binding site, the mutants Y23A and F45S demonstrate strong binding relative to wild type and Y35G. Wild type's gentle slope is consistent with the claim that a specific site does not exist. The shallow slope observed for Y35G supports the notion that only a small fraction of solution-phase Y35G is in a conformation such that the binding cavity is accessible to Xe.
The resonance shift of 129Xe will asymptotically approach the shift due only to the site at very high protein concentrations. Since the data appear to still be in the linear regime, we can place upper and lower bounds on the parameters of interest.
We have built a low-pressure flow-through Xe polarizer and characterized its performance by examining both the output Xe129 polarization by NMR and the in situ Rb85 polarization by optically detectd EPR. We have compared our system with a numerical model of expected performance, and discovered interesting behavior.
One can obtain nuclear polarization on the order of unity in Xe129 using spin-exchange optical pumping. The relative difference in nuclear spin populations of these hyperpolarized gases fr exceeds that dervied from thermal distribution and makes it feasible to conduct a wide variety of gas phase magnetic resonance experiments. The physics and applications of HP gases continue to be subjects of intense interest. The polarizer we have built has been employed in studying fundamental physics of NMR decays in solids, as a biosensor to characterize proteins, to characterisze diffusion in nanotubes, and to characterize the porous environments of metal-organic frameworks.
