Krypton-83 Relaxometery by Remote Detection

For the first time, a hyperpolarized (hp) noble gas with a nuclear electric quadrupole moment is available for high field nuclear magnetic resonance (NMR) spectroscopy and MR imaging. Hp 83Kr (I = 9/2) is generated by spin exchange optical pumping and separated from the rubidium vapor used in the pumping process. Signal enhancements of more than three orders of magnitude compared to the thermal equilibrium 83Kr signal at 9.4 T magnetic field strength are obtained. The spin-lattice relaxation of 83Kr is caused by quadrupolar couplings during the brief adsorption periods of the krypton atoms on the surrounding container walls and significantly limits the currently obtained spin polarization. Measurements in macroscopic glass containers and in desiccated canine lung tissue at field strengths between 0.05 T and 3 T using remotely detected hp 83Kr NMR spectroscopy reveal that the longitudinal relaxation dramatically accelerates as the magnetic field strength decreases.

Apparatus and procedures used for production and detection of hp 83Kr. a, Experimental setup used for continuous flow experiment (valve A & D permanently open and B & C permanently closed, peristaltic pump operating) and stopped flow experiments (valve D permanently closed) for remotely detected relaxometry. b, Diagram for stopped flow experiments for remotely detected relaxometry with movable probe head indicating the position of 83Kr within the magnetic field, gas valve status, and r.f. pulse sequence used to obtain field depended T1 times. c, Diagram for stopped flow experiments for remotely detected relaxometry with gas storage cell (through valve C) indicating the position of the 83Kr within the magnetic field, gas valve status, and r.f. pulse sequence alternatively used to obtain remotely detected T1 times at 0.05 T.

Enhancement of the 83Kr NMR signal by spin exchange optical pumping. a, Hp 83Kr NMR spectrum recorded every 10 s under continuous flow optical pumping conditions. The magnetic (B) field at the optical pumping cell is switched on, twice inverted in direction and finally switched off at the times indicated. The very small signal at the beginning of the experiments arises from optical pumping in the arbitrarily aligned stray field from the superconducting NMR magnet. The effect on the NMR spectrum is time delayed because the Kr flows at a rate of 150 cm3/s through approximately 3.5 m of PFA tubing into the detection cell within the high field region. The maximum enhancement factor is 27 times the thermal equilibrium signal at 9.4T b, Hp 83Kr NMR signal obtained from a stopped flow experiment with an enhancement factor of 1200 times the thermal signal at 9.4 T. The flow is stopped for 15 min and subsequently the hp krypton is released into a pre-evacuated detection cell via PFA tubing. The inset shows the spectrum from the continuous flow experiment of Figure 1a set to scale.

Enhancement over thermal signal of hp 83Kr NMR as a function of polarization time. Data are collected under two different laser powers (60 W – open squares and 30 W – open circles). The curves are produced by fitting Eq. 1 using one pre-exponential fitting parameter (i.e. maximum enhancement factor) and the exponential fitting parameter . The highest signal enhancement is about half that shown in Figure 2b because a slightly different experimental setup with much longer transfer tubing is used.

Field dependent longitudinal relaxation of hp 83Kr. a, Longitudinal relaxation decay curves. Data are collected a six different magnetic field strengths (3.0 T – closed squares, 1.5 T – closed triangles, 1.0 T – closed circles, 0.5 T – open circles, 0.25 T – open triangles, and 0.15 T – open squares). The reported intensities are obtained from the r.f. pulse sequence shown in Figure 1b, and the curves are mono-exponential fittings. The reported field strengths are at the center of the 40 mm sample region. b, T1 as a function of field strength in gas-phase (closed circles) and desiccated canine lung tissue (closed squares). All values, except the highest field strength (9.4 T) values, are calculated from mono-exponential fitting of longitudinal relaxation decay curves (Figure 4a). Horizontal bars represent the lower and upper limits of the field strength along Bz within the sample region (see Figure 1a). Lines in Figure 4b are intended to guide the eye.

Figure 5 | Remotely detected longitudinal relaxation decay curves at 0.05 T. Data produced by the procedure described in Figure 1c are collected at two storage cell temperatures (433 K – open circles and 297 K – open triangles). The curves are produced by mono-exponential fitting these data resulting in T1 = 85 s at 297 K and T1 = 220 s at 433 K.

A hyperpolarized quadrupolar noble gas (hp 83Kr) free from paramagnetic and highly reactive rubidium vapor has been produced for the first time. Signal enhancements up to 1200 times that of thermally polarized 83Kr are obtained. Further improvements may come from the use of laser line width narrowing external cavity devices 48, increased laser power 38, and isotopic enrichment. Improved 83Kr optical pumping is also to be expected in light of the relaxation studies presented in this work that will lead to improved optically pumping and gas transfer system designs. A strongly field dependent T1 relaxation caused by quadrupolar interactions on surfaces is observed by remotely detected NMR relaxometry with hp 83Kr. Additionally, the T1 of hp 83Kr is shown to increase with increasing surface temperature due to reduced surface adsorption times. The longitudinal relaxation times are also prolonged by the presence of small amounts of water. Neither of these results can be explained by relaxation due to diffusion through inhomogeneous magnetic fields. The results are encouraging for further increasing the hp 83Kr signal intensity and suggest that in vivo hp 83Kr MRI at 1.5 T may be feasible.