Constraints on Changes in Fundamental Constants from a Cosmologically Distant OH Absorber or Emitter

We have detected the four 18 cm OH lines from the $z\sim 0.765$ gravitational lens toward PMN J0134–0931. The 1612 and 1720 MHz lines are in conjugate absorption and emission, providing a laboratory to test the evolution of fundamental constants over a large lookback time. We compare the HI and OH main line absorption redshifts of the different components in the $z\sim 0.765$ absorber and the $z\sim 0.685$ lens toward $\mathrm{B}0218+357$ to place stringent constraints on changes in $F\equiv g_p[{\alpha }^2/\mu {]}^{1.57}$. We obtain $[\Delta F/F]=(0.44\pm {0.36}^{\mathrm{stat}}\pm {1.0}^{\mathrm{syst}})\times {10}^{-5}$, consistent with no evolution over the redshift range $0<z\lesssim 0.7$. The measurements have a $2\sigma$ sensitivity of $[\Delta \alpha /\alpha ]<6.7\times {10}^{-6}$ or $[\Delta \mu /\mu ]<1.4\times {10}^{-5}$ to fractional changes in $\alpha$ and $\mu$ over a period of $\sim 6.5\mathrm{G}\mathrm{yr}$, half the age of the Universe. These are among the most sensitive constraints on changes in $\mu$.

Figure 1

GBT HI 21 cm [(a), $\sim 5\mathrm{km}/\mathrm{s}$ resolution)] and OH mainline [(b), $\sim 14.5\mathrm{km}/\mathrm{s}$ resolution] spectra toward PMN J0134–0931, with negative optical depth $(-{10}^3\times \tau )$ plotted against barycentric frequency, in MHz. The solid line shows the three-Gaussian fit to each spectrum. The vertical lines in each figure indicate the locations of the three components (marked $A$, $B$, and $C$), with the dashed and dotted lines in (b) showing the 1667 and 1665 components, respectively.

Figure 2

OH satellite line spectra ($\sim 5\mathrm{km}/\mathrm{s}$ resolution), with negative optical depth ($-{10}^3\times \tau$) plotted against barycentric velocity, in $\mathrm{km}/\mathrm{s}$, relative to $z=0.76355$, the redshift of component $C$. (a) 1720 MHz transition, redshifted to $\sim 975.6\mathrm{MHz}$; (b) 1612 MHz transition, redshifted to $\sim 912.2\mathrm{MHz}$; (c) sum of 1612 and 1720 MHz spectra; this spectrum is consistent with noise. Note that the feature at $\sim -30\mathrm{km}/\mathrm{s}$ is not statistically significant. The satellite lines peak $\sim 10\mathrm{km}/\mathrm{s}$ blueward of component $C$ (the dashed vertical line), which has the lowest redshift of the three spectral components.