Constraints on Spin-Dependent Short-Range Interaction between Nucleons

We search for a spin-dependent $P$- and $T$-violating nucleon-nucleon interaction mediated by light pseudoscalar bosons such as axions or axionlike particles. We employ an ultrasensitive low-field magnetometer based on the detection of free precession of colocated ${}^3\mathrm{He}$ and ${}^{129}\mathrm{Xe}$ nuclear spins using SQUIDs as low-noise magnetic flux detectors. The precession frequency shift in the presence of an unpolarized mass was measured to determine the coupling of pseudoscalar particles to the spin of the bound neutron. For boson masses between 2 and $500\mu \mathrm{eV}$ (force ranges between $3\times {10}^{-4}\mathrm{m}$ and ${10}^{-1}\mathrm{m}$) we improved the laboratory upper bounds by up to 4 orders of magnitude.

Figure 1

Sketch of experimental setup. The lower plane SQUIDs in module $D$, $E$, and $I$ marked as (●) are used to detect the ${}^3\mathrm{He}/{}^{129}\mathrm{Xe}$ free spin precession. The center of the cylindrical spin sample cell ($D=60\mathrm{mm}$, ${\oslash }_D=58\mathrm{mm}$) has an average distance of $\overline{z}=66\mathrm{mm}$ to the sensors. The relative position of the cell in the projection onto the ($x$, $y$) plane is indicated by the gray square. SQUIDs in module $S$ marked as (■) are used for the gradiometric sensor arrangements. The unpolarized mass (cylindrical BGO crystal: $d=70\mathrm{mm}$, ${\oslash }_d=60\mathrm{mm}$) can be moved along the $x$ axis ($B$-field axis) to “close” ($\Delta x_c=2.2\mathrm{mm}$) and “distant” ($\Delta x_d=170\mathrm{mm}$) positions and vice versa (see text). This is accomplished by a piston driven glass tube with the BGO fixed at its cell-facing side. The two measuring arrangements “left” ($\mathcal{L}$) and “right” ($\Re$) are shown.

Figure 2

(a) Extracted frequency shifts ${\overline{\Delta \nu }}_{\mathrm{sp}}$ (with correlated $1\sigma$ error) of the 10 measurement runs. The triangles specify the $\Re$, the circles the $\mathcal{L}$ arrangement of the BGO crystal. Full symbols indicate the $c\to d$ sequence, hollow symbols the opposite case ($d\to c$). (b) Results ${\overline{\Delta \nu }}_{\mathrm{check}}$ obtained from the LV data using the same fit model [Eq. (6)]. Since no mass was moved, we expect no shift in the spin precession frequency. The rightmost symbols in both plots (stars) indicate the respective weighted means ($1\sigma$ error).

Figure 3

The experimental 95% confidence upper limit on $|g_s^N{g}_p^n|$ plotted versus $\lambda$, the range of the Yukawa force with $\lambda =\hslash /(m_{a}c)$. The axion window is indicated by the light gray area. (1) result of 28, (2) result of 29, (3) result of 30, (4) result of 31, (5) result of 32, (6) result of 33, (7) result of 34, (8) this experiment ($\Delta x=2.2\mathrm{mm}$) [35], (9) expected results for $\Delta x\approx 0\mathrm{mm}$ using the same data set demonstrates the gain in measurement sensitivity for $\lambda <{10}^{-3}\mathrm{m}$. See 36 for bounds on the pseudoscalar short-range force between polarized electrons and unpolarized nucleons. Raffelt [37] points out that much tighter constraints on $|g_s^N{g}_p^n|$ can be inferred by combining constraints on $g_s$ from stellar energy-loss arguments and $g_p$ from searches for anomalous monopole-monopole forces.