Dissociative recombination of cold and its interstellar implications

0 Departments of Chemistry and Astronomy, University of Illinois at Urbana-Champaign , 600 South Mathews Avenue, Urbana, IL 61801 , USA HC3 plays a key role in interstellar chemistry as the initiator of ion-molecule chemistry. The amount of HC3 observed in dense interstellar clouds is consistent with expectations, but the large abundance of HC3 seen in diffuse clouds is not easily explained by simple chemical models. A crucial parameter in predicting the abundance of HC3 in diffuse clouds is the rate constant for dissociative recombination (DR) with electrons. The value of this constant has been very controversial, because different experimental techniques have yielded very different results, perhaps owing to varying degrees of rotational and vibrational excitation of the HC3 ions. If the value of this rate constant under interstellar conditions were much lower than usually assumed, the large HC3 abundance could be easily explained. In an attempt to pin down this crucial rate constant, we have performed DR measurements at the CRYRING ion storage ring in Stockholm, using a supersonic expansion ion source to produce rotationally cold HC3 ions. These measurements suggest that the DR rate constant in diffuse clouds is not much lower than usually assumed and that the abundant HC3 must be due to either a low electron fraction or a high ionization rate. 1. Introduction (a ) H C3 in dense clouds Interstellar clouds are typically classified into dense molecular clouds and diffuse clouds (Snow & McCall 2006). Dense molecular clouds have typical number densities of 104106 cmK3 and temperatures of w2030 K. In these clouds, almost all hydrogen atoms are in the form of H2, and almost all carbon atoms are in the form of CO. HC3 is produced by cosmic ray ionization of H2 to form HC2, followed by the fast ionneutral reaction HC2C H2 / HC3C H. The cosmic ray ionization is the rate-limiting step and it proceeds with a rate of zn(H2), where z is usually assumed to be w3!10K17 sK1. HC3 is destroyed by H C4H2+ C C3H3+ H2 C3H+ C+ CO CH5+ H2 CH3+ H2 CH2+ H2 CH+ C H + 3 CO H2+ H2 Figure 2. After being produced from H2 by cosmic ray ionization, HC3 initiates a network of ionneutral reactions that is responsible for the production of a wide variety of molecules. chemical reactions with atoms and molecules other than H2, dominantly by CO note about this result. First, it is independent of the cloud density n(H2), because the ratio of H2/CO is constant (all H atoms are in the form of H2, all C atoms are in the form of CO). Second, the HC3 number density is very small, approximately one part per billion of the H2 number density. The infrared observations (figure 3) directly yield the column density N(H3 )w15!1014 cmK2 in dense clouds (McCall et al. 1999). Since the column C density is the integral of the number density along the line of sight, the absorbing y t isn 0.98 e t n i e v i lta 1.01 e r path length can be directly inferred as L wN HC3=nHC3, because nHC3 is a constant. This analysis yields path lengths of the order of 1 pc and resulting average number densities hn(H2)iwN(H2)/L of w105 cmK3, in agreement with our expectations for dense clouds. This agreement confirms the overall picture of HC3 chemistry in dense clouds and suggests that the adopted values of z and k CO are probably correct. (b ) H C3 in diffuse clouds Diffuse interstellar clouds have typical number densities of w101103 cmK3 and higher temperatures of w50100 K owing to the influence of starlight. The starlight also photodissociates H2, leading to a mixture of H and H2, and photoionizes C to CC, producing abundant electrons. While HC3 is formed in the same way as in dense clouds, it is now dominantly destroyed by electrons, with a rate of kenHC3ne. Once again making the assumption of steady state, we can C solve for nH3 Substituting nzwH33 !w101K0K177scKm1K,3k.eZ5!10K7 cm3 sK1 and n(H2)/n(e)w2400, we find that C This is approximately three orders of magnitude lower than in dense clouds and is also independent of the cloud density to the extent that n(H2)/n(e) is a constant. y t i sn1.01 e t n i e v ta1.00 i l e r Downloaded from http://rsta.royalsocietypublishing.org/ on November 16, 2014 Cold H3C and its interstellar implications 2. Dissociative recombination of cold HD3 The rate constant ke can be measured in the laboratory, so one might think that its value should not be suspect. However, as reviewed by Larsson (2000), Plasil et al. (2002) and Oka (2003), laboratory values of ke have varied by as much as four orders of magnitude over the past 30 years. To make matters worse, theoretical calculations had yielded (until the work of Kokoouline & Greene (2003)) unreasonably low values of ke, so they did not help in discriminating among the different experiments. Near the turn of this century (e.g. Larsson et al. 2003), it became widely recognized that the different experimental results were likely to be due to the lack of experimental control over the rotational and vibrational levels populated in the various plasmas used for the measurements. In the interstellar medium, almost all of the HC3 ions are in the lowest two rovibrational levels, and none of the experiments that had been conducted were really applicable to this situation. A major advance in controlling the state distribution of molecular ions for DR measurements was the introduction of the ion storage ring method. At a storage ring such as the CRYRING facility in Stockholm, ions are produced in an external plasma source, extracted and mass selected, accelerated and then injected and stored in a ring consisting of a series of bending magnets. Ions can be stored as long as several tens of seconds, so that all ions will relax to their vibrational ground states by spontaneous emission. The DR measurement is made by overlapping a nearly mono-energetic electron beam with the ion beam and counting the number of neutral fragments (H and H2) that fly out of the ring. In addition to the benefit of the vibrational cooling, the storage ring technique has the advantage that it is also conceptually simple and does not require extensive modelling, in contrast to afterglow experiments. Furthermore, the electronion impact energy can be precisely controlled (by varying the velocity of the electron beam), thus enabling detailed measurements of the DR crosssection. However, a major limitation of this technique has been that hot plasma sources have traditionally been used to produce the ions. For a symmetric ion such as H3 , there is very little rotational cooling in the ring (as HC3 has only C forbidden rotational transitions), and the rotational temperature of the ions remains high despite the vibrational cooling. In order to study rotationally cold HC3 ions, we developed a new ion source that utilizes a supersonic expansion. In this source, high-pressure gas is pulsed through a small nozzle into the vacuum of the rings endstation; as the gas rushes into v (...truncated)