NMR is known for its notoriously low sensitivity resulting in weak signal strengths. Different methods are used to get around this problem. These include cross-polarization (INEPT like transfers from high gyromagnetic ratio nuclei), chemically induced dynamic nuclear polarization (CIDNP), polarization transfers from electrons, also called the dynamic nuclear polarization (DNP), spin exchange to xenon from optically pumped alkali metal vapor and the spin-polarization induced nuclear Overhauser effect (SPINOE). In the Pines Lab, we routinely use most of these experimental techniques for achieving hyperpolarization. We are particularly interested in the creation and manipulation of polarization derived from para-hydrogen and hyper-polarized xenon.

Molecular hydrogen can be used to enhance sensitivity via the technique of para-hydrogen induced polarization. Quantum symmetry considerations couple the nuclear spin states of hydrogen to the rotational degrees of freedom, which allows relatively simple means for manipulating the population distributions of the nuclear spin states. By molecular addition onto some molecule of interest the resultant non-equilibrium population distribution gives rise to an increase in the NMR signal intensity (the PASADENA or ALTADENA effects).

Para-hydrogen is nuclear spin hyperpolarized, even at zero magnetic fields or at room temperatures. At the one hand, this directly results in enhanced signal intensities, theoretically 10⁴ to 10⁵ times stronger than what is achievable from a Boltzmann distributed thermal state. The enhancement is state dependent, and at lower fields, can be much larger. For example, at the earth’s field, enhancements up to nine orders of magnitude are achievable.

We have built a modular setup that uses liquid nitrogen to quickly cool gaseous hydrogen and iron(III) oxide to catalyze the interconversion of ortho and para-hydrogen to produce a constant stream of arbitrary concentrations of para-hydrogen, ranging from 25% to 100%. This is achieved by liquid nitrogen and liquid helium cooled cryostats. Flow and pressures of para-hydrogen are under user control. The delivery of the hydrogen to the magnet bore is under complete control of the spectrometer via TTL lines to allow automation by the pulse program. Our circuit also allows for selection between hydrogen and nitrogen flows for effective purging of the system. We can also fill para-hydrogen cylinders, allowing us to have a transportable supply of para-hydrogen.

Recently we demonstrate the creation and observation of para-hydrogen induced polarization in heterogeneous hydrogenation reactions. Wilkinson’s catalyst, RhCl(PPh₃)₃, supported on either modified silica gel or a polymer, was shown to hydrogenate styrene into ethylbenzene and to produce enhanced spin polarizations, observed through NMR, when the reaction was performed with H2 gas enriched in the para spin isomer. Furthermore, gaseous phase para-hydrogenation of propylene to propane with two catalysts: the Wilkinson’s catalyst supported on modified silica gel and Rh(cod)(sulfos) (cod = cycloocta-1,5-diene; sulfos = -O₃S(C₆H₄)CH₂C(CH₂PPh₂) ₃) supported on silica gel, demonstrate heterogeneous catalytic conversion resulting in large spin polarizations. These experiments serve as a direct verification of the mechanism of heterogeneous hydrogenation reactions involving immobilized metal complexes and can be potentially developed into a practical tool for producing catalyst-free fluids with highly polarized nuclear spins for a broad range of hyperpolarized NMR and MRI applications. This is the first demonstration of observing para-hydrogen induced polarization using a heterogeneous (supported) catalytic system. The results are in press with the Journal of American Chemical Society.

The use of heterogeneous catalysis has led us into the possibility of performing hyperpolarized gas phase imaging. Our results have demonstrated enhanced image signal to noise in the images acquired of hyperpolarized propane gas derived from the para-hydrogenation of propylene at elevated temperatures.This result has lead to several important applications, including flow through porous materials, gas-phase reaction kinetics and adsorption studies and MRI in low fields, all using catalyst-free polarized fluids. Our results look extremely promising and are published with the Angewandte Chemie International Edition in the form of a V.I.P. paper as well as a publication in Science. We demonstrate the characterization and direct visualization of gas-phase flow and the density of active catalyst in a packed-bed microreactor, as well as control over the dynamics of the polarized state in space and time to facilitate the study of subsequent reactions. These procedures are suitable for characterizing reactors and reactions in microfluidic devices where low sensitivity of conventional magnetic resonance would otherwise be the limiting factor.

The PASADENA and ALTADENA effects can also be used as the pre-polarization step for alternative detection methodologies. For example, the current optical magnetometer set-up requires some form of magnetization which is subsequently detected by the gradiometer. This magnetization can be derived from the spin order of para-hydrogen. Such an experiment can be carried out at extremely low fields; maybe the earth’s field or even at zero field. In the long run, this will potentially become an important step forward for developing low-cost, mobile, detection-multiplexed optical MRI/NMR systems. At the Pines Lab, work is underway for the first proof-of-principle experiments in this direction. Preliminary results have been obtained that demonstrate that the symmetry of the para-hydrogen can be broken at low magnetic fields, which is one of the main requirements to observe PHIP signals in general. Furthermore we have demonstrated that spectral resolution and sensitivity achieved with optical magnetometry are high enough to resolve antiphase PHIP signals.

Hyperpolarized xenon gas is obtained through contact between xenon atoms and rubidium vapor that has been electronically polarized via optical pumping. The rubidium polarization is achieved by using circularly polarized light to selectively excite one of two possible electronic transitions, which, after relaxation back to the ground state, provides a net electronic polarization of rubidium. This polarization is transferred to xenon nuclei during collisions via the hyper-fine coupling.

The lab has recently finished a homebuilt xenon polarizer which can provide up to 120W of laser emission in order to enable high levels of spin polarization. Research into hyperpolarized xenon applications is currently being applied in the areas of microfluidics, remote detection, and biosensors.

There are wide-ranging applications for hyperpolarized xenon in the fields of biomedical imaging and biochemical spectroscopy, including use as a biosensor. In addition, the xenon polarization can be transferred to other nuclei in solution through an effect referred to as SPINOE¹, first demonstrated by Alex Pines and co-workers. The method relies on the cross-relaxation between hyperpolarized xenon spins and dipolar coupled nuclei in solution. For more details on the SPINOE, read the Pines work, featured as a cover story on Spectroscopy, Volume 14, 1999.

Conventional proton MRI employs relatively large magnetic fields (0.5-7 T) to obtain observable signal at thermal Boltzmann polarization. Although the improved signal-to-noise ratio induced by high field increases both the spatial and temporal resolution of the image, superconducting magnets used by high field MRI systems are expensive, heavy, and not feasible for gas-phase MRI at thermal polarization. However, the polarization of hyperpolarized xenon does not depend on static magnetic fields, and is therefore particularly advantageous for low-field applications. Current research in the Pines lab is focusing on utilizing hyperpolarized xenon gas in conjunction with low-field techniques such as atomic magnetometry and ex situ MRI in order to produce effective low-field gas imaging techniques.

[1] Navon, G. et al., Science 271, 1848-1851 (1996)