First Constraint on Cosmological Variation of the Proton-to-Electron Mass Ratio from Two Independent Telescopes

A high signal-to-noise spectrum covering the largest number of hydrogen lines (90 ${\mathrm{H}}_2$ lines and 6 HD lines) in a high-redshift object was analyzed from an observation along the sight line to the bright quasar source J2123-005 with the Ultraviolet and Visual Echelle Spectrograph on the European Southern Observatory Very Large Telescope (Paranal, Chile). This delivers a constraint on a possible variation of the proton-to-electron mass ratio of $\Delta \mu /\mu =(8.5\pm {3.6}_{\mathrm{stat}}\pm {2.2}_{\mathrm{syst}})\times {10}^{-6}$ at redshift $z_{\mathrm{abs}}=2.059$, which agrees well with a recently published result on the same system observed at the Keck telescope yielding $\Delta \mu /\mu =(5.6\pm {5.5}_{\mathrm{stat}}\pm {2.9}_{\mathrm{syst}})\times {10}^{-6}$. Both analyses used the same robust absorption line fitting procedures with detailed consideration of systematic errors.

Figure 1

Spectrum of J2123-005 observed with UVES-VLT clearly showing several ${\mathrm{H}}_2$ transitions, each comprising two distinct velocity features. Left: Region 3097–3106 Å with a relatively low signal-to-noise ratio. Right: Region 3251–3263 Å toward longer wavelength at a higher signal-to-noise ratio, also containing a line of HD. The upper trace shows residuals in the fit regions. The tick marks above the spectrum refer to the velocity components of the ${\mathrm{H}}_2$ and HD lines (green), modeled Lyman-$\alpha$ forest lines (blue), and Fe ii lines (red). ${\mathrm{H}}_2$ spectral line identifications are shown at the bottom.

Figure 2

Composite residual spectrum composed of the 25 cleanest ${\mathrm{H}}_2$ lines for fits to 2-, 3-, and 5-velocity components. Upper panel: Residuals. Lower panel: The spectrum as simulated from the final parameters resulting in the 2- and 5-VC models. The tick marks indicate the fitted positions in 2- and 5-VC models; in the 5-VC model, two VCs lie at almost the same velocity and appear as a single tick mark on this plot.

Figure 3

Results from various fits. (a) Resulting value for $\Delta \mu /\mu$ for progressively increasing the number of velocity components; there are two results for 3-VC fits, obtained by putting the 3rd component in both distinct velocity features, which are horizontally offset for clarity; (b) the ${\chi }_{\nu }^2$ value; (c) result on $\Delta \mu /\mu$ from fits to 7 subclasses of spectral lines.

Figure 4

Current constraints on $\Delta \mu /\mu$ on a cosmological time scale. Black points for $z<1$ refer to ${\mathrm{NH}}_3$ values from objects B0218 [25, 26] and PKS1830 [27]; blue diamonds refer to Ref. 12; and the green triangle to Ref. 10, i.e., the HIRES-Keck analysis of J2123. The gray circle refers to a recent reanalysis of Q0347 [22]. The red square is the present result of our UVES-VLT analysis of J2123. All uncertainties are at $1\sigma$. Independent constraints on the same object have been offset in redshift for clarity.