A sub-ppm upper limit on the cosmological variations of the fine structure constant alpha

TL;DR

The paper tests the cosmological invariance of the fine-structure constant α and the proton-to-electron mass ratio μ by comparing CH and H2O submillimeter absorption lines observed with ALMA toward two high-z quasars. By fitting a common opacity profile and measuring a bulk velocity offset δv between CH and H2O across multiple epochs and sightlines, the authors derive stringent 3σ upper limits on variations: |Δα/α| < 0.55 ppm toward PKS 1830-211 and |Δα/α| < 1.5 ppm toward B 0218+357, assuming independent μ constraints. The results, aided by precise rest frequencies, co-spatial gas tracing, and simultaneous observations that suppress major systematics, provide the deepest single-system constraints to date and offer a radio-based complement to optical studies on α-diversity, including implications for a putative α-dipole anisotropy. Overall, the work demonstrates the power of molecular absorbers in distant galaxies to constrain fundamental-constant variation with unmatched precision and robustness.

Abstract

Absorption spectroscopy toward high-redshift quasars provides strong constraints on the putative variation of fundamental constants of physics on cosmological time scales. The submillimeter ground-state transitions of methylidyne (CH) and water (H2O), both molecules widespread and coeval in the interstellar medium, provide a sensitive test for variations of alpha, the fine structure constant, and mu, the proton-to-electron mass ratio, taking advantage of the unmatched spectral resolution and frequency reliability of radio techniques. We used ALMA simultaneous observations of the two species to constrain any velocity offset between their absorption profiles toward the radio-bright lensed quasars PKS1830-211 (z_abs=0.88582) and B0218+357 (z_abs=0.68466). Our observational setup minimizes instrumental errors and known sources of systematics, such as time variability of the absorption profile and frequency-dependent morphology of the background quasar. The excellent correlation between CH and H2O opacities, the large number of individual narrow velocity components, and the number of independent spectra obtained due to the intrinsic time variability of the absorption profiles ensure that even the chemical segregation bias is minimized. We obtained bulk velocity shifts delta_v = -0.048 pm 0.028 km/s and -0.13 pm 0.14 km/s (1 sigma confidence level) between CH and H2O in the direction of PKS1830-211(NE) and B0218+357(SW), respectively. These values convert into the 3sigma upper limits |Delta_alpha/alpha| < 0.55 ppm and 1.5 ppm, respectively, taking into account the independent upper limits on |Delta_mu/mu| previously obtained for these systems. These constraints on |Delta_alpha/alpha|, at look-back times of about half the present age of the Universe, are two to four times deeper than previous constraints on any other single high-z system.

Figures (3)

• Figure 1: Diagram of the lowest rotational energy levels of CH. The $\Lambda$-doublet and hyperfine levels are not to scale. The rest-frame frequencies of CH transitions observed by ALMA are indicated in red, in GHz, and their relative line strengths are given in parenthesis.
• Figure 2: Spectra of CH and H$_2$O lines toward PKS 1830$-$211 NE and SW images (left, middle, respectively), and toward B 0218$+$357(SW) (right) observed by ALMA at different epochs. The spectra are colored by dates, as indicated in the top-left and top-right boxes. The hyperfine structure of each CH $\Lambda$-doublet is indicated at the bottom of spectra. The spectrum of the H$_2^{18}$O isotopologue, taken on 2014/05/05, is also shown for PKS 1830$-$211(SW) (top-middle box, in light blue). A zoom on the weak $v \sim +170$ km s$^{-1}$ velocity component is shown for PKS 1830$-$211(SW), with the 2016 averaged spectra (in purple) instead of that taken on 2016/03/05.
• Figure 3: Overlay of the opacity profiles of CH (average of the two $\Lambda$-doublets, deconvolved from their respective hyperfine structure) and H$_2$O (scaled by the opacity ratio $\gamma_\tau$), on top of their common fit profile (see Sec. \ref{['sec:gaussfit']}) for PKS 1830$-$211(NE) (top, 2012 data), the isolated velocity component near $v \sim +170$ km s$^{-1}$ toward PKS 1830$-$211(SW) (middle, average of the 2016 data), and B 0218$+$357(SW) (bottom). Note the span of two orders of magnitude between the optical depths of these features.