2020MNRAS.497.3591S

We present high-quality Atacama Large Millimeter/submillimeter Array (ALMA) Band 8 observations of the [C I] ³P₁-³P₀ line and 609-µm dust continuum emission towards the nearby luminous infrared galaxy (LIRG) IRAS F18293-3413, as well as matched resolution (300-pc scale) Band 3 CO J = 1-0 data, which allow us to assess the use of the [C I] ³P₁-³P₀ line as a total gas mass estimator. We find that the [C I] line basically traces structures detected in CO (and dust) and a mean (median) [C I]/CO luminosity (L^′^_ [C I]_/L^′^_ CO_) ratio of 0.17 (0.16) with a scatter of 0.04. However, a pixel-by-pixel comparison revealed that there is a radial L^′^_ [C I]_/L^′^_ CO_ gradient and a superlinear L^′^_ CO_ versus L^′^_ [C I]_ relation (slope = 1.54 ± 0.02) at this spatial scale, which can be explained by radial excitation and/or line opacity gradients. Based on the molecular gas masses converted from the dust continuum emission, we found that the CO-to-H₂ and [C I]-to-H₂ conversion factors are relatively flat across the molecular gas disc with a median value of 3.5^+1.9_–1.3 and 20.7^+9.2_–4.9 M_☉ (K km s^–1 pc²)^–1, respectively. A non-LTE calculation yields that typical molecular gas properties seen in nearby (U)LIRGs (n_ H₂ = 10^3–4 cm^–3, T_kin ∼ 50 K, and X_ C I_ = (0.8-2.3) x 10^–5) can naturally reproduce the derived [C I]-to-H₂ conversion factor. However, we caution that a careful treatment of the physical gas properties is required in order to measure H₂ gas mass distributions in galaxies using a single [C I] line. Otherwise, a single [C I] line is not a good molecular gas estimator in a spatially resolved manner.