Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 317, 962-967 (1997)

Previous Section Next Section Title Page Table of Contents

2. Catalytic reactions of organometallic chemistry.

The linkage of one or several metal atoms to one or several organic molecules leads to the formation of organometallic complexes. From gas phase mass spectroscopy experiments, Shröder et al. (1991) produced benzene through cyclotrimerization of acetylene via a neutral complex [FORMULA]. First, they ionized the [FORMULA] molecule which led to substitution of CO by [FORMULA]. Secondly, the [FORMULA] (x=2 to 4) complex was neutralized and then reionized. Spontaneous production of benzene occurred only after the neutralization step. Kline et al. (1985) studied the formation of [FORMULA] and [FORMULA] complexes in 15K argon matrices. Formation of these complexes at very low temperature indicate that the activation energy should be very low but no value was published up to now.

Cyclotrimerization of unsaturated hydrocarbons to benzene is also known with various cationic metals: [FORMULA] for [FORMULA] or 5 (Schnabel et al. 1991), [FORMULA] (Buckner et al. 1989) and [FORMULA] (Berg et al. 1994). In these cases, the reaction of ethene [FORMULA] with these metals leads firstly to the formation of the [FORMULA] complex which indicates a spontaneous dehydrogenation of ethene into acetylene upon coordination. Then, an external energy of a few eV is required to activate the triacetylenic complex toward benzene formation. We have considered a generic scheme for the cyclotrimerization of acetylene to benzene, described by the set of chemical reactions given in Fig. 1. In this scheme, M can be a metallic atom or a small metallic aggregate.

[FIGURE] Fig. 1. Scheme for the cyclotrimerization of acetylene to benzene catalysed by the metal M.

Berg et al. (1994) have measured the rate constants of the cyclotrimerization of ethene to benzene. They have found rate constant values in the order of magnitude of the Langevin rates computed for a model described by the collision of an induced dipole with an ion. No constant rate for the reaction with neutral metal and acetylene was measured up to now. Nevertheless, Clary et al. (1994) demonstrated that the Langevin rate for a reaction of neutral reagents can be computed with a model described by interactions between two induced dipoles. This theory leads to the following formula:


where [FORMULA] is the polarizability volume (in Å [FORMULA], [FORMULA] is the ionization potential (in eV) and [FORMULA] is the molecular mass (in g/mol) of the reagent i. Clary et al. (1994) showed that the Langevin rate between neutrals is well verified in various cases: e.g. ground state carbon atoms react with 1-alkenes or with 1-alkynes with this rate. Chemical reactions of our modelling are not very different from that of the previous example where carbon atom has been replaced by a metal atom and 1-alkyne is chosen to be acetylene. In absence of published rate measurements between neutral metals and acetylene in the gas phase, we computed the rates of our set of reactions using formula (1).
One can notice that a small aggregate of n atoms has an abundance of [FORMULA] compared to the atomic state but its cross section is increased by a factor [FORMULA] (in the spherical approximation). Finally, the integrated cross section of clusters containing n metal atoms should be corrected by a factor [FORMULA] compared to the integrated cross section of isolated metal atoms. Then, it is important to notice that the value of the integrated cross section of 100 metal atom clusters is only 1/4 that of the integrated cross section of 1-metal atom. The constant rates being proportional to the cross sections, the colliding efficiency of a given metal should not vary dramatically if M is considered as free atoms or as small aggregates. Our calculations were performed considering M as a 1-metal atom and we discussed the case of small metal aggregates by modifying in consequence the [FORMULA] parameter value of the initial depletion defined in the next section. The values of the parameters [FORMULA] value and [FORMULA], defined in formula (1), do not change significantly from one transition metal to another. Even if the exact nature of the M species can vary, their kinetic parameters can be approximated by those of a given transition metal. In this study, the parameters for M have been deduced from atomic iron because it is the most abundant transition metal according to cosmic abundances. For the intermediate complexes [FORMULA] (x=1, 2 and 3), the polarizability was chosen to be equal to the sum of the polarizabilities of individual components and the ionization potential was chosen identical to that of the naked metal.
We considered that the collision between M and [FORMULA] leads to a [FORMULA]) complex with a Langevin rate but this complex is in an excited state because it must evacuate, at least, the metal-acetylene bonding energy. This energy excess U can be evacuated according to three main channels: i) the molecule radiates U through its vibrational modes, ii) the molecule dissipates U by loosing a fragment, iii) the molecule radiates U by a fluorescence emission. In the case of the reaction of metal PAHs, this problem was treated by Marty et al.(1996a) using the unimolecular rate theory (Forst et al. 1972, 1973). From this theory, Marty (1996b) found that only aggregates containing more than ten metal atoms can coordinate acetylene efficiently (the i process being faster than the ii one). In other cases, [FORMULA]) could be efficiently formed if fluorescence channels exist (the iii process faster than the ii one) but, up to now, no experimental results were reported about the fluorescence of such complexes.
Another problem is to estimate the stability of the intermediate complexes [FORMULA] [FORMULA] and 3) under a given radiation field. No experiments have been performed yet on the photodestruction of the organometallic bond of neutral metal-acetylene complexes in the gas phase. In a first approximation, we postulate that every photon of energy higher than the metal-acetylene bonding energy [FORMULA] can destroy the organometallic bond (i.e. dissociation efficiency=1 if [FORMULA]). This is probably too harsh but it leads to the worst case for the formation of benzene via organometallic cyclotrimerization. In the case where M are large clusters of metals ([FORMULA] atoms), the unimolecular theory predicts that the dissociation energy of complexes must be much larger than the bonding energy El, leading to a better efficiency of benzene formation than that predicted by our modelling.
Kobitchev et al. (1987) computed a bonding energy of 0.8 eV between Fe and C2H2 from ab-initio calculations at the Hartree Fock level. However, they did not take into account the electronic correlation which must lead to a higher value of energy. The bonding energies of metal-acetylene cations were computed by Sodupe & Bauschlisher (1991) for the first row of transition metals. They found values varying between 0.7 and 1.7 eV, depending on the metal. Hettich et al. (1996) have obtained experimentally [FORMULA] and [FORMULA]. However, bonding energies of neutral complexes are expected to be lower. In order to deduce approximately [FORMULA], we can make an analogy using the ratio [FORMULA]. Since Shröder et al.(1991) have found [FORMULA], we have inferred that [FORMULA] was between 0.8 and 1.5eV. In the absence of an experimental photodissociation spectrum of [FORMULA] complexes, the dissociative cross section has been approximated by the geometric section, [FORMULA], when [FORMULA]. Compared to [FORMULA] complexes (Hettich et al. 1986), this approximation should be an overestimation of the real cross section of photoabsorption which leads to an underestimation of the rate of benzene formation in our model. In addition, Schnabel et al. (1991) have shown that metals, which are able to perform the cyclotrimerization of acetylene to benzene, have also a good activity for the dehydrogenation of hydrocarbons. They noticed that the reaction of benzene with [FORMULA] was leading to the formation of benzyne [FORMULA], which is known to be very reactive. It can coordinate to metals, and Bauschlicher et al. (1993) computed that [FORMULA] could be bonded to the triple bond one like a standard acetylenic derivative. According to this, we propose to take into account another catalytic set of reactions, leading to the formation of the triphenylene PAH [FORMULA]), as described in Fig. 2.

[FIGURE] Fig. 2. Scheme for the cyclotrimerization of benzyne ([FORMULA]) and benzene to triphenylene ([FORMULA]) catalysed by the metal M.

Kinetic parameters were computed in the same way as those of the first catalytic process. Other reaction pathways involving cyclotrimerization could be envisaged. However, in order to simplify the discussion of the results of this study, we used only a restricted set of reactions (see Fig. 1 and 2). In fact, the triphenylene production is a first step toward PAH formation. The complete set of parameters for each compound involved in our modelling is defined in Table 1.


Table 1. Set of physical parameters of the reagents involved in the catalytic reactions of the modelling. The parameters associated to M are those of the iron atom.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: July 8, 1998