Glass-wool study of laser-induced spin currents en route to hyperpolarized Cs salt

The nuclear spin polarization of optically pumped Cs atoms flows to the surface of Cs hydride in a vapor cell. A fine glass wool lightly coated with the salt helps greatly increase the surface area in contact with the pumped atoms and enhance the spin polarization of the salt nuclei. Even though the glass wool randomly scatters the pump light, the atomic vapor can be polarized with unpolarized light in a magnetic field. The measured enhancement in the salt enables study of the polarizations of light and atomic nuclei very near the salt surface.

Figure 1

Schema for optical pumping of Cs atoms and Cs NMR of salt. A permanent magnet produces the field 0.56 T, and the homogeneity is 15 ppm for a cylindrical cell of inner diameter 8 mm. The cell is heated with a copper tube and an electric heater and irradiated with $\pi$ or $\sigma$ light passing through a half-wave ($\lambda /2$) plate.

Figure 2

(a) Absorption cross section and (b) nuclear spin current calculated for each light polarization at the $D_2$ line, 0.56 T, and 1.5 kPa ${\mathrm{N}}_2$. We assumed uniform pumping at 1 W/cm${}^2$ and the boundary condition, $\langle S_z\rangle =\langle I_z\rangle =0$, at the sidewall ($r=4$ mm) of the cylindrical cell. The spin-exchange interaction between Cs atoms was neglected.

Figure 3

Cs NMR signal optically enhanced by $\sigma$ pumping. CsD salt was formed on glass wool and filled with Cs metal and 3.0 kPa ${\mathrm{N}}_2$ gas. The positive (red, enhancement $61.5$) and negative (orange, $-79.8$) signals were recorded without averaging, respectively, for the laser frequencies 351 720.0 and 351 731.5 GHz. Thermal signal scaled by the right-hand vertical axis (dotted green curve) was averaged a hundred times with the repetition interval of 10 min.

Figure 4

Laser-frequency dependence of the NMR area enhanced by $\sigma$ pumping at 1.5 kPa ${\mathrm{N}}_2$. The vertical axis is in arbitrary units but can be compared with the other plots in this paper. Radio-frequency pulses of the tipping angle $\pi$/3 were applied at the repetition interval 100 s. (a) CsH at the sidewall of two cylindrical cells ($ˆ$, $▵$) with no glass wool. The signal was averaged 70 times for each point. (b) CsH and CsD formed on the glass fibers. The apparent density of the glass wool was ($▴$) dense, ($$) medium, and ($•$) sparse, judged visually.

Figure 5

Pressure dependence of NMR enhancement for $\sigma$ pumping in the glass-wool cells. (a) CsH filled with ${\mathrm{N}}_2$ gas at ($•$) 1.5 kPa, ($$) 4.5 kPa, and ($▴$) 7.5 kPa. (b) CsD in vacuum ($•$, $90\hspace{0.5em}{}^{ˆ}$C; $$, $85\hspace{0.5em}{}^{ˆ}$C) and CsH with the original pressure 4.0 kPa of ${\mathrm{H}}_2$ gas ($◂$).

Figure 6

(a) Light-polarization dependence of enhancement for CsD filled with 1.5 kPa and 3.0 kPa ${\mathrm{N}}_2$ gas. NMR area was measured with ($•$, $▴$) $\sigma$ light and ($$, $▾$) $\pi$ light. (b) Nuclear spin current simulated for $\sigma$ (dotted blue) and mixed (dashed red, solid orange) polarizations in a cylindrical cell with no glass wool. The relative power of each polarization is shown as a percentage. The current by polarization mixing is not simply a sum of currents by each component of light polarization.

Figure 7

Light transmission through a test cell with ($▴$, right vertical scale) and without ($•$, left vertical scale) the glass wool. Horizontal axis is the rotation angle of the second polarizer. The inset shows the optics arrangement.