A Coupled Dynamical and Chemical Model of Starless Cores of Magnetized Molecular Clouds. II. Chemical Differentiation
Dense cores of molecular clouds are the basic units of isolated low-mass star formation. They have been observed extensively in various molecule lines and dust continuum with the aim of revealing their chemical and dynamical state. In a previous paper, we formulated a coupled dynamical and chemical model for data interpretation and carried out an initial investigation focusing on the effects of a magnetic field on the core dynamics and chemistry. Here, we update our chemical network and the
... network and the treatment of magnetic field-matter coupling and explore the effects of changing various parameters, including the initial gas-phase metal abundances, adsorption energies, cosmic-ray ionization rate, sticking probability onto dust grains, cloud mass, as well as magnetic field strength. The model results are compared with the velocity field and column density distributions of CO, CS, CCS, NH 3 , N 2 H + , and HCO + inferred observationally for the well-studied starless core L1544. We find that, in agreement with previous work, models with the so-called high metal abundances produce excessive CS and CCS by more than 2 orders of magnitude. Models of magnetized clouds with " low metal " and " mixed metal " (with a strong initial depletion of sulphur) abundances can fit the available data on L1544 reasonably well, with the low-metal model fitting somewhat better the chemical data (except for CS) and the mixed-metal model the velocity field. Taking into account of a newly recalculated rate for the neutral-neutral reaction S + CCH ! CCS + H increases the abundance of CCS substantially, leading to a better agreement with observation for the mixed-metal model. We considered two sets of adsorption energies, compiled respectively by Aikawa et al. and Hasegawa & Herbst. Our results favor the former over the latter. For our standard models, we adopted a cosmic ionization rate of 1:3 Â 10 À17 s À1 and a sticking probability of 0.3. Increasing their values does not improve the model fits. Somewhat surprisingly, removing the magnetic support of the cloud leads to relatively modest changes in the peak column densities of the species except for CS. However, the spatial distributions of CS and CCS become more centrally concentrated than observed in L1544, and the infall speed is too large to be acceptable. This illustrates the need for both chemical and dynamical data to provide the tightest possible model constraints. A generic feature of our coupled dynamical and chemical model is that NH 3 and, to a lesser extent, N 2 H + are concentrated in the slowly contracting, central plateau region of the growing core, whereas CS and CCS are most abundant in the lower density envelope surrounding the plateau, which has a faster infall motion. The chemical differentiation offers an exciting possibility of directly probing the velocity field of core evolution leading to star formation.