Studying Infall in Infrared Dark Clouds with Multiple HCO${}^{+}$ Transitions

Mass accumulation by inward gravitational motions is a basic step in models of star formation (e.g. 1969larson; 1987anglada; 2001bonnell; 2018motte). Infall motions often are revealed in observations with moderate optical depth molecular lines, which show blue-shifted self-absorption dips at frequencies where optically thin lines peak, produced by the temperature gradients of dense cores and the infalling gas (e.g. 1986walker; 2003wu). Multiple transitions of many molecules such as CS, CO, H${}_{2}$CO, HCO${}^{+}$, HCN, CH${}_{3}$CN, and C${}_{3}$H${}_{2}$ have been investigated to search for infall signatures in various environments in low mass star forming regions (e.g. 1995myers; 1997mardones; 1998tafalla; 2016keown), as well as massive star forming regions (e.g. 1998zhang; 2003wu; 2005fuller; 2008velusamy; 2010barnes; 2016qin).

The comparisons of different tracers including multiple transitions of the same tracer have been made through both observations and simulations, to explore which tracers are more efficient to reveal infall signatures in what kind of sources. Higher-$J$ transitions, such like $J\,$=3–2 and $J\,$=4–3, of HCN and HCO${}^{+}$ are considered as good tracers for dense clusters based on their infall asymmetries in lines generated in numerical simulations (2014chira). Observationally, HCN $J\,$=3–2 has been considered as a very good infall tracer in a dense clump sample associated with H${}_{2}$O masers (2003wu). The $J\,$=1–0 transition of HCN was found to have the strongest infall signatures (2010stahleryen) in simulations although the line shapes depend on the viewing angle (2012smith). Compared to HCN, HCO${}^{+}$ changes less drastically in abundances in chemical models at temperatures ranging from 10 K to 40 K (2012vasyunina), which is the temperature range of dark clouds. Thus, HCO${}^{+}$ is potentially better for comparing infall in different star forming environments as it is less affected by chemistry. Among HCO${}^{+}$, the lowest transition, HCO${}^{+}$ $J\,$=1–0, showed the most blue-asymmetric profiles towards a sample of 77 candidate high mass protostellar objects (HMPOs) (2005fuller), which is consistent with the results of numerical simulations that HCO${}^{+}$ $J\,$=1–0 is a better indicator of collapse in high mass star formation (2013smith) where the gas is not dense enough to excite the higher-$J$ transitions.

Being cold and dense, infrared Dark Clouds (IRDCs) are believed to represent the initial conditions of massive star formation and the formation of the associated stellar clusters (e.g. 2000carey; 2006rathborne; 2009perettofuller; 2018motte). Some pilot HCO${}^{+}$ observations towards two massive IRDCs have revealed that the clumps/cores in there are undergoing rapid collapse (2013peretto; 2018contreras). Considering IRDCs are at an earlier stage than HMPOs or UCHII massive clumps, it raises the question of what molecular or ion line tracers can best trace infall in IRDCs? Specifically, will lower-$J$ transitions with smaller critical density (like HCO${}^{+}$$J\,$=1–0) reveal infall signature better, or higher-$J$ transitions are needed to probe collapse in these IRDCs? In addition, it is important to understand the properties of infall in the IRDC phase compared with infall in more evolved regions.

Recently, in a HCO${}^{+}$$J\,$=1–0 mapping survey towards a sample of 27 IRDCs (Peretto et al. in preparation), a significant fraction presented infall signatures in their line profiles. In this paper, we report our follow-up study on a sub-sample of these IRDCs, all of which show infall signatures in HCO${}^{+}$$J\,$=1–0 in at least one clump in the IRDCs. We mapped 11 IRDCs with HCO${}^{+}$$J\,$=4–3, and made single-point observations with HCO${}^{+}$$J\,$=3–2, towards 23 HCO${}^{+}$$J\,$=4–3 peak positions, using James Clerk Maxwell Telescope (JCMT). The aims of the study are to use multiple HCO${}^{+}$ transitions to fit and constrain models of infall motion, to understand infall properties in IRDCs, and to investigate if higher-$J$ transitions of HCO${}^{+}$ can also probe infall and reveal infall properties well in these HCO${}^{+}$$J\,$=1–0 selected IRDCs .

The organization of the rest of the paper are the follow: We describe source selection and observations in sect. 2. The observational results and and data analysis are presented in sect. LABEL:sect:results. We discuss infall tracers and infall properties of IRDCs in sect. LABEL:sect:discussion, and summarise our conclusions in sect. LABEL:sect:concl.

The sources in this work are a sub-sample of an IRAM 30 m HCO${}^{+}$ $J\,$=1–0 survey towards 27 IRDCs (Peretto et al. in preparation). The parent sample was selected to span a range of geometries and a factor of $\sim$100 in mass range from $\sim$200 to $\sim$2$\times$10${}^{4}$M${}_{\odot}$, to cover a representative sample of IRDCs. From the parent sample we selected 11 IRDCs who show blue asymmetry indicative of infall in HCO${}^{+}$ $J\,$=1–0 in at least one clump. We made follow-up higher-$J$ HCO${}^{+}$ observations and we fit multiple transitions with models, in order to study infall properties in these candidate infalling IRDCs. The basic information of the sources are listed in Table 1.

All the sources have been mapped in HCO${}^{+}$ $J\,$=1–0 with IRAM 30 m telescope (half power beam width (HPBW): 29”; $\eta_{mb}$=0.75). Since N${}_{2}$H${}^{+}$ $J\,$=1–0 has been found in good agreement with H${}^{13}$CO${}^{+}$ $J\,$=1–0 in both velocity and line width (2005fuller), we use N${}_{2}$H${}^{+}$ $J\,$=1–0, which was observed simultaneously with HCO${}^{+}$ $J\,$=1–0, to locate the central velocity of the cloud and to identify infall signature for all the HCO${}^{+}$ transitions. The IRAM 30 m observations and data reduction of HCO${}^{+}$ $J\,$=1–0 and N${}_{2}$H${}^{+}$$J\,$=1–0 of the 27 IRDCs will be presented in a separated paper (Peretto et al. in preparation).

Eleven IRDCs were mapped in HCO${}^{+}$$J\,$=4–3 with Heterodyne Array Receiver Programme (HARP) ¹¹1https://www.eaobservatory.org/jcmt/instrumentation/heterodyne/harp/. (harp) (HPBW:: 14”; $\eta_{mb}$=0.63) on the 15 m James Clerk Maxwell Telescope (JCMT)²²2The JCMT is operated by the EAO on behalf of NAOJ; ASIAA; KASI; CAMS as well as the National Key R&D Program of China (No. 2017YFA0402700). Additional funding support is provided by the STFC and participating universities in the UK and Canada. on Mauna Kea, Hawaii during August and September 2016 (Project M16BP081), and August 2017 (Project M17BP087). The Auto-Correlation Spectral Imaging System (ACSIS) spectrometer was used. Maps of 180” by 180” (except one source SDC18.624, whose map size is 240”$\times$180”) were made with RASTER scans with T${}_{sys}$ ranging from 364 K to 613 K (with an average of 443 K), under weather Band 3 and Band 4 with 0.08 $<\tau_{225\,GHz}<$ 0.20.

Single-point observations were carried out in HCO${}^{+}$$J\,$=3–2 towards the HCO${}^{+}$$J\,$=4–3 peak positions with JCMT RxA3m frontend and ACSIS backend ³³3https://www.eaobservatory.org/jcmt/instrumentation/heterodyne/rxa/ (HPBW: 20”; $\eta_{mb}$=0.57) from April to June 2018 under Project M18AP073. No H${}^{13}$CO${}^{+}$$J\,$=3–2 observations were available due to a low LO current at 260 GHz. The molecule line frequencies are listed in Table 2. Each position was observed in position-switching mode (GRID) for an integration time of 300 s in weather Band 4 (0.12 $<\tau_{225\,GHz}<$ 0.2). An off-position (+600”, +600”) (J2000) from the observed position was also observed. The system temperatures ranged from 778 K to 1356 K, with an average of 988 K. For both the HARP and RxA3m observations, ACSIS was configured to cover 250 MHz wide windows, 8192 channels in each window, resulting to a velocity resolution of $\sim$ 0.02 km s${}^{-1}$ for HARP and $\sim$ 0.03 km s${}^{-1}$ for RxA3m. All data the HARP and RxA3 observations were smoothed with a Gaussian kernel which has the FWHM equals 5.9 channels, resulting a velocity resolution of 0.15 km s${}^{-1}$ and 0.20 km s${}^{-1}$ for HARP and RxA3, respectively. The telescope pointing was checked before observing a new source and was checked every 1-1.5 hours, by observing one or more calibration sources in CO(2–1) at 234.591 GHz and CO(3–2) at 350.862 GHz for RxA3m and HARP, respectively. The uncertainty in the flux calibration is estimated to be about 10%.

The HARP and RxA3m data reduction were undertaken using the Starlink (starlink) software package SMURF, KAPPA, and GAIA. Each integration was first visually checked. The data were converted to spectral cubes and baselines were subtracted before been written out as FITS format files using standard Starlink routines (pipeline).

All data were converted from the antenna temperature scale T${}_{A}^{*}$ to main-beam brightness temperature T${}_{mb}$ using T${}_{mb}$ = T${}_{A}^{*}$/$\eta_{mb}$, where main beam efficiencies are listed in Table 2, as well as the noise levels and the velocity resolutions.