Gas-Grain Chemistry: Problems in Chemical Models

Gas-Grain Chemistry: Problems in Chemical Models

M. S. El-Nawawy1 and M. Y. Amin2

1 Astronomy Department, Faculty of Science, Cairo University, Cairo, Egypt.

E mail:

2 King Fesal University, Physics Department, Faculty of Science, Saudi Arabia.

Abstract

In this article the gas-grain chemistry is studied. Its importance in the chemical evolution of diffuse and dense interstellar clouds are discussed. In order to investigate the role of grain chemistry, the chemical evolution is studied in many objects. There are many challenges appeared in the models and in the chemical processes included. The gas-grain chemistry is tested in contracting and non-contracting models. The results show some challenges which are outlined here. The results in the contracting models seems to be more fitting with observations than the non-contracting models. The models including gas-grain chemistry succeeded in explaining the chemistry in diffuse clouds and some of the observed species in dense clouds such as TMC-1.

In this review we discuss the impact of inclusion of grain chemistry in addition to the gas phase chemistry. Particular attention is given to the variations happened in the abundances of the gas phase species in this case. We have also outlined the experimental efforts given to the problem of H2 formation on grain surface and the methods used to study gas-grain chemistry. Some other points related to grain chemistry is also discussed; such as the formation of H2O and CO2 on grain surface and the structure of the grain mantel under different conditions.

Introduction

The interstellar grains play an important role in the chemical evolution of the interstellar clouds. This role is partially recognized since the pioneer work on H2 formation on grain surface by Hollenbach and Salpeter(1970 & 1971) and Watson and Salpeter(1972 a and b). The astronomer start to revisit this point again recently through experimental and theoretical work. Pirronello et al.(1997a and b) and Pirronello et al.(1999) reported the results of experiments to measure the recombination rate of hydrogen in ultra-high-vacuum (UHV) chamber by irradiating the sample with two beams of H and D atoms and monitoring the HD production rate. Two different substrates have been used: a natural olivine( a polycrystalline silicate containing Mg2Sio4 and Fe2SiO4) slab and an amorphous carbon sample. The substrate temperatures during hydrogen irradiation were in the range between 5 and 15 K. It is found that H and D atoms adsorbes on the surface form molecules only above 9 K in the case of olivine and above 14 K in the case of amorphous carbon. This indicates that tunneling alone does not provide enough mobility to H adatoms to enable recombination, and thermal activation is required. The measurements also give lower values for the recombination efficiency than previously estimated. These results are extended in other papers ( e.g. Katz et al. 1999, Biham et al., 2001 and Biham and Lipshtat, 2002) to formulate the discrete distribution of atomic hydrogen . These efforts tried to investigate the rate of H2 formation on grain surface under different conditions particularly in the case of limited abundance of atomic hydrogen.

Charnley(1998) presented the stochastic treatment of the chemical evolution of dense molecular clouds. His algorithm is based on the Monte Carlo method. Green et al.(2001) give a more detailed studies for a stochastic approach to grain surface chemical kinetics for small systems of H, O and H, N, O. Modifications of the rate equations for reactions on grain surface are given by Caselli et al.(1998) and Stantcheva et al.(2001). More studies on the modeling of surface chemistry on interstellar grains are given by Herbst and Shematovich(2003). Lipshtat and Biham(2003) introduced a set of moment equations for the analysis of the formation of molecular hydrogen and other chemical reactions on dust grain surfaces in the interstellar medium. These equations are derived from the master equations that describe these processes.

Many of the chemical processes that occur on grain surface have been deeply studied in many recent papers. The energy distribution in the H2 formation process on the surface of dust grains is investigated theoretically by Takahashi et al.(1999) and the chemical effects of H2 formation in excited states have been studied by Garrod et al.(2003) and they found that the presence of the excited H2 has only marginal effects on the chemistry of interstellar clouds. Formation of carbon dioxide by surface reactions on ices has been investigated by Roser et. al. (2001) experimentally in conditions close to those encountered in the interstellar medium. The interaction of carbon monoxide(CO) with vapor-deposited water(H2O) ices has been studied by Collings et al.(2003a). The adsorption and desorption of CO on and from amorphous water ice at interstellar cool temperatures has been studied by Collings et al.(2003a) and Collings et al.(2003b). The chemical desorption of an adsorbed CO molecule in the vicinity of H2-forming sites on interstellar grains in cold dense clouds is investigated theoretically by Takahashi and Williams(2000). A nice review of the chemical structure of the grain mantel under different conditions is given by Allamandola et al.(1999). The formation of H2O either in gas phase or on grain surface is also studied by O’Neil and Williams(1999). The thermal desorption of water ice in the interstellar medium is studied by Fraser et al.(2001).Measurements of conversion rates of CO to CO2 in UV-induced reaction of D2O(H2O)/CO amorphous ice is given by Watanabe and Kouchi(2002). In this article we review most of the recent studies concerning the chemical processes on grain surface.

There are many studies on grain chemistry have been carried out using both pseudo time dependent and contraction models ( e.g. Hasegawa et al.,1992; Ruffle and Herbst, 2000; El-Nawawy, 2000 a and b; and El-Daly et al., 2002). Discussion of the results in our models both in pseudo-time dependent and contraction models are outlined.

Theoretical modeling

We studied the chemical evolution in theoretical models, using both pseudo- time dependent ( non-contracting) and contracting models. The chemical network contains gas-phase and gas- grain reactions. Details of the chemical and dynamical structure in our models are explained in many of our papers ( e.g. El-Nawawy et al. 1997, El-Nawawy 2000 a and b, and El-Nawawy 2003 ). In the models of contraction we solved the hydrodynamic equations in one dimension and the method of solution is outlined by El-Nawawy et al.(1997) and Amin and El-Nawawy(1997). The cloud is divided into shells and the coordinates are allowed to move with contraction such that we can follow the deep center of the cloud. The chemistry is studied only in the central core for simplicity. Considering the chemistry in other shells needs a huge amount of computation. The study of one model of contraction until reaching to the density of TMC-1 requires about 10 CPU hours.

The temperature changes during contraction, in terms of the density n and the visual extinction Av, according to the following relation (Tarafdar et al. 1985);

(2.21)

Where the visual extinction Av is calculated by integration of the column densities over the cloud shells. The initial temperature is assumed to be 100 K in all the models. A restriction is placed on the temperature not to fall below 10 K.

We have included the following gas-grain chemical processes in our computational models:

1-  Accretion of neutral species (144 reactions) on grain surface

2-  Surface reactions ( 249 reactions)

3-  Cosmic ray desorption ( 144 reactions)

4-  Cosmic ray induced photodesorption ( 144 reactions)

5-  Desorption induced by the energy of molecular hydrogen formation (144 reactions)

6-  Accretion of positive ions ( 108 reactions) on grain surface. This type of reaction is tested in our studies only in pseudo-time dependent models.

The network used in the present work contains 4680 reactions of gas- phase and gas-grain chemistries, involving 572 species. The sum of gas-grain reactions is 933. UMIST data base are used for gas-phase reactions, while the grain chemistry rates of reactions are collected from Hasegawa et al.(1992) and Hasegawa and Herbst(1993).

The results shown in Table(1) is for gas phase (model 1) and gas-grain (model 2) under the same conditions. The results are compared with the observations taken for the dense molecular cloud TMC-1 by Hiraahara et al. (1992), Ohishi et al.(1992), Ohishi and Kaifu(1998) and Dickens et al.(2000). The detailed results of the models discussed here and many others will appear in a forthcoming paper. We only focus in the present paper on the main results that merge from our studies. The comparison of the two models show that some of the species increase in the gas-grain model than given by the gas-phase model, such as: OCS, H2S, C3N, NH3, HCOOH, CH3OH and many others. Clearly these molecules increase as a result of grain chemistry. But from the other side the models of gas phase is more successful in explaining the observations and this may refer to many problems in the rain chemistry, particularly the rate equations needs to be more studied experimentally and theoretically. These two models are contracting models and their results are in agreement with the observations better than the pseudo time dependent models. This is because in the contracting models the initial physical parameters are reasonable and the chemistry grow with the physical parameters as the contraction proceeds.

Comparison between a pseudo time dependent model and a contracting model is given in Table(2) . Model(3) is a pseudo time dependent model, while model 2 is a contracting model. In general the results are in comparison with the observations are better in model(2) than given by model(3). The electron density and the most dominant ions and their variations with density in model 1 are shown in Figures 1 and 2. These figures confirm the decrease of electron and ion density as the neutral density increases with contraction. The electron density in model 2 is also shown in Figure 1. It is clear that the electron density decreases more in the gas-grain model than given by the pure gas-phase model. This results is also confirmed by Bergin and Langer (1997). This result needs to be tested again in models containing more details of the grain chemistry than used in these models. For example how the freezing of ions can affect this result? The models of contraction carried out in our calculations do not include the freezing of ions on grain surface.

The importance of grain chemistry has also been studied in ζ Ophiuchi as a diffuse cloud ( El-Nawawy, 2003). The chemistry is studied in ζ Ophiuchi using pseudo time dependent models. The results also confirm the role of grain chemistry in understanding the abundances observed for many species such as NH, OH , CO, H2O and many other species. A new results on B68 as one of the dense molecular clouds given by Amin and El-Nawawy(2004) confirm the role of grain chemistry in predicting the observations of many species.

Table (1) The fractional abundances(relative to total hydrogen) of species observed

in TMC-1.

Species / Observations / Model 1 /

Model 2

C2 / 2.5(-8) / 4.5(-8) / 5.2(-8)
CH / 1.0(-8) / 2.1(-8) / 1.0(-8)
CN / 1.5(-8) / 3.2(-7) / 3.8(-7)
CO / 4.0(-5) / 5.1(-5) / 5.9(-5)
NO / 1.5(-8) / 1.7(-8) / 3.4(-8)
OH / 1.5(-7) / 2.5(-8) / 2.6(-8)
SO / 2.5(-9) / 6.9(-10) / 1.8(-9)
C2H / 2.5(-8) / 4.9(-8) / 2.5(-8)
CS / 5.0(-9) / 5.9(-8) / 2.3(-8)
C2S / 4.0(-9) / 5.1(-9) / 3.6(-9)
C3S / 5.0(-10) / 1.4(-9) / 1.1(-9)
OCS / 1.0(-9) / 5.8(-10) / 1.1(-9)
H2S / <1.0(-9) / 3.6(-11) / 1.1(-10)
C2O / 3.0(-10) / 1.0(-10) / 4.7(-11)
HCN / 1.0(-8) / 7.3(-8) / 9.7(-8)
HNC / 1.0(-8) / 3.1(-8) / 5.9(-8)
SO2 / <5.0(-10) / 2.5(-10) / 1.7(-11)
C3H / 2.5(-10) / 6.9(-9) / 6.4(-9)
C3N / 5.0(-10) / 5.9(-9) / 1.2(-8)
C3O / 5.0(-11) / 3.4(-12) / 1.6(-11)
H2CO / 1.0(-8) / 7.8(-9) / 7.0(-9)
H2CS / 1.5(-9) / 3.6(-10) / 3.3(-10)
NH3 / 1.0(-8) / 5.1(-9) / 1.1(-8)
C2H2N / 2.5(-9) / 4.7(-9) / 6.8(-9)
CH2CO / 5.0(-10) / 1.0(-9) / 1.6(-9)
C3H2 / 5.0(-9) / 5.1(-9) / 7.3(-9)
C4H / 8.0(-7) / 1.5(-8) / 1.8(-8)
HCOOH / 1.0(-10) / 2.6(-11) / 2.3(-10)
HC3N / 3.0(-9) / 1.2(-8) / 3.2(-8)
CH3CN / 5.0(-10) / 1.5(-9) / 1.9(-9)
C4H2 / 4.0(-10) / 3.5(-10) / 5.5(-10)
CH3OH / 1.0(-9) / 7.6(-11) / 3.1(-10)
CH3CHO / 3.0(-10) / 5.3(-14) / 2.2(-13)
H3C3N / 1.0(-10) / 1.0(-11) / 4.5(-11)
C3H4 / 3.0(-9) / 6.7(-11) / 2.0(-10)
CH3C3N / 2.5(-10) / 2.5(-11) / 1.0(-10)
C5H / 1.5(-10) / 3.6(-10) / 2.7(-10)
CH3C4H / 1.0(-10) / 2.7(-10) / 1.4(-9)
HC5N / 1.5(-9) / 3.0(-10) / 7.4(-10)
C6H / 5.0(-11) / 1.3(-10) / 1.4(-10)
HCO+ / 4.0(-9) / 5.9(-9) / 7.3(-9)
HCS+ / 3.0(-10) / 3.6(-10) / 2.0(-10)
N2H+ / 2.5(-10) / 7.1(-12) / 1.2(-11)
H2CN+ / 1.0(-9) / 5.5(-10) / 9.9(-10)

Figure 1. The electron fractional abundance as a function of

density for pure gas-phase and gas-grain models.

Figure 2. The dominant ions in producing electrons as function of density

for the pure gas-phase chemical-dynamical model, Model 1.

Table (2) The fractional abundances(relative to total hydrogen) of species observed in TMC-1.