Introduction

By bringing spins out of thermal equilibrium, hyperpolarization techniques can enhance the NMR signal by several orders of magnitude. Hyperpolarization methods are particularly advantageous at low magnetic field strengths, where thermal polarization may fail to bring the NMR signal above noise levels. Here we demonstrate the ability to directly transfer spin polarization from hyperpolarized (HP) ¹²⁹Xe spins to thermally polarized ¹H via the nuclear Overhauser effect (SPINOE). While previous works have used the freeze-thaw method to enhance the polarization transfer between hyperpolarized ¹²⁹Xe spins and thermally polarized ¹H spins^1–3, here we simply bubble hyperpolarized ¹²⁹Xe gas in solution. We show that after a single bubbling of HP ¹²⁹Xe gas in solution, the proton signal can be increased by more than 100-fold. At low field, where thermal polarization may fail to bring the NMR signal above noise level, simultaneous detection of ¹²⁹Xe and ¹H resonances enables one to watch the full spin dynamics of the polarization transfer in real time.

Methods

All measurements were performed on a lab-built NMR spectrometer operating at 2.1 mT using a dual-tuned volume coil. Before each acquisition, HP ¹²⁹Xe gas was bubbled into solution in a 5 mm NMR tube for 6 seconds. After a set delay, τ, the spectrometer acquisition was triggered to simultaneously excite and detect ¹²⁹Xe and ¹H resonances (6 s bubble – τ – 90°(¹²⁹Xe/¹H) – acquire). Because at 2.1 mT thermal polarization is on the order of 10^-9, no presaturation pulses were needed to destroy the thermal magnetization of ¹H. After each acquisition, remaining ¹H/¹²⁹Xe magnetizations were destroyed using a train of RF pulses. Measurements were repeated for 2 ml samples of toluene, pentane, and cyclohexane, which have a measured ¹H T₁ of 5.6, 3.8, and 1.9 s, respectively, at room temperature. Measurements were also repeated with an extended bubbling time of 18 s and at lower temperatures.

Results and Discussion

Figure 1 shows the experimental setup used for these experiments, while Figure 2 shows spectra acquired from a sample of pentane before and after bubbling HP ¹²⁹Xe gas in solution. In addition to the expected dissolved-phase xenon resonance, we observed also a ¹H resonance well above noise level. Spectra acquired after a single dissolution of hyperpolarized xenon gas show that the transfer of polarization from the long-lived dissolved-phase hyperpolarized ¹²⁹Xe spins in solution to the fast relaxing ¹H spins is a continuous process that leads to an observable ¹H signal enhancement up to four minutes after a single delivery of HP ¹²⁹Xe gas (Figure 3). Signal averaging over this time can effectively increase the observable ¹H signal (Figure 4), which appears to relax with a T₁ that qualitatively agrees with what is predicted by the Solomon relaxation theory^4,5:

$$S_{1H}(t)\propto e^{-R_{Xe}t}-e^{-R_{H}t},$$
where,
$$R_{Xe(H)} = \frac{1}{T_1^{Xe(H)}}\pm\frac{\sigma_{XeH}\sigma_{HXe}}{1/T_1^{Xe}-1/T_1^{H}},$$
and σ is the cross-relaxation rate between the two spins. This implies the maximum signal enhancement can be obtained at a time,

$$t_m=\frac{1}{R_H-R_{Xe}}\ln{\frac{R_H}{R_{Xe}}}.$$

Experiments performed at lower temperatures show a decrease in the value of t_m, which could be attributed to a decrease in T₁ as expected for small molecules. A concurrent increase in the overall polarization enhancement may be attributed to the known increase in xenon solubility at low temperatures, leading to an enhancement in the cross-relaxation rate (Figures 3 and 5).

Conclusions

We demonstrate that transfer of spin order from hyperpolarized ¹²⁹Xe gas to thermally polarized ¹H spins can be obtained from simply bubbling ¹²⁹Xe in a solution containing ¹H spins. The transfer of polarization is a continuous process that lasts for the entire lifetime of the dissolved-phase ¹²⁹Xe hyperpolarized state, leading to observed enhancement of the ¹H polarization up to several minutes after xenon dissolution.

Acknowledgements

This work was supported by the National Institute of Biomedical Imaging and Bioengineering grant R21EB031319 and by the National Institutes of Diabetes and Digestive and Kidney Diseases grants R01DK108231and R01DK12306.