Testing grain-surface chemistry in hot core regions

Many complex organic molecules have been detected in star-forming regions like CH3OCH3 and CH3CH2CN. Studying their chemistry is important, since these species may eventually be incorporated into circumstellar disks and thus become part of the material from which future planetary systems are made. Two scenarios for their formation have been proposed: grain-surface formation and high-temperature gas-phase reactions (van Dishoeck & Hogerheijde 1999). In the former case, grain-surface reactions during the cold pre-stellar and protostellar phases are thought to lead to the formation of various hydrogenated molecules, which subsequently evaporate into the gas when the young star heats its surroundings. In the latter scenario, high-temperature gas-phase reactions involving evaporated molecules (primarily CH3OH) produce complex organic species (Charnley et al. 1992). The resulting molecules from these two schemes are also known as “first generation” and “second generation” species, respectively. Currently, it is very difficult to distinguish between these two scenarios since the rates of even some of the most basic surface reactions are not known. Accordingly, we have started a combined laboratory and observational program to test the scenarios. The basic scheme for grain surface chemistry has been outlined by Tielens & Hagen (1982) and Tielens & Charnley (1997). The first step is hydrogenation of the main species that accrete from the gas onto the grains, i.e., C, N, O and CO, leading to CH4, NH3, H2O, H2CO and CH3OH. These molecules can react further in the ice with C, H, N, and O creating several more complicated species. In the warm regions close to the protostars the grain mantles can evaporate, returning many of the “first generation” molecules into the gas phase. Deep JCMT searches for a set of molecules have been performed for 1 low mass and 7 high mass young stellar objects (YSO’s) (e.g. Figure 1). The focus is on relatively nearby sources with narrow line widths to avoid confusion. The molecules studied are HCO, H2CO, CH3OH, CH2CO, CH3CHO, C2H5OH, HNCO and NH2CHO. HNCO, CH3OH, H2CO and C2H5OH are detected for most or all sources, whereas NH2CHO, CH2CO and HCOOH are detected for some and HCO and CH3CHO are not detected at all. For HCO the reason is likely due to its low intrinsic line strength, however CH3CHO has previously been detected by Ikeda et al. (2002). The rotational temperatures found by Ikeda et al. are only 20-30 K and thus much lower than the lowest rotational transition of 71 K that can be probed with the JCMT. Initial correlations between the abundances with respect to CO of these species have been studied, and a more detailed analysis using radiative transfer models will follow. The correlation between the abundances relative to CO are found to be strong for CH3OH, H2CO, C2H5OH, and HNCO (see for example Figure 1 for HNCO and C2H5OH), whereas no clear trends are yet found for the abundances of HCOOH, NH2CHO, and CH2CO. The strong trends found for four of these molecules indicate that they are either present in the same environment or share a common formation mechanism. The resulting branching ratios can be compared with various models. For example, pure gas-phase models by Lee et al. (1996) predict CH2CO/CH3CHO/CH3CH2OH ratios of ∼ 1 : 10 : 10, whereas grain-surface chemistry predicts ratios much closer to unity. Our inferred branching ratios are found to be much closer to the latter values, indicating that grain-chemistry plays an important role.