grains. There are two basic surface reaction mechanisms: the Langmuir-Hinshelwood
mechanism and the Eley-Rideal mechanism [2, 3]. In the case of the former one, both
H atoms are bound to the surface and at least one of them must be mobile. The
problem with this mechanism is that it can only be effective in a certain temperature
range (e.g. in the diffuse interstellar medium) because at very low temperatures the
diffusion rate of H atoms in the ice is too low, whereas at higher temperatures the H
atoms cannot stay on the surface long enough for an efficient H2formation. In the case
of the Eley-Rideal mechanism one H atom is bound on the surface and then a gaseous
H atom collides and reacts with this atom. This mechanism requires a relatively high
concentration of H atoms. Therefore, considering simple physisorption, it cannot be
considered as a main interstellar H2formation mechanism either, especially above 20 K
where the residence time of the H atoms is too short.
There are several complex solid-phase reaction mechanisms that might explain the
formation rate of interstellar H2[1–3]. The residence time of H atoms near the surface
can be increased by capturing them in the pores of amorphous carbon or silicate grains.
Another possibility is that H atoms are bound on carbon grains at a low temperature,
and a sudden increase of the temperature (for example due to a supernova explosion)
can lead to explosive recombination of H atoms in a runaway event. A third mechanism
involves the chemisorption of H atoms on the surface of interstellar grains.
In the case of the latter mechanism, a molecule can first chemisorb an H atom
which in the second step is abstracted by another H atom, resulting in the formation
of an H2molecule and the reformation of the original catalyst molecule. Therefore,
this process can be considered as a catalytic cycle. Being a cycle, the mechanism can
also start with abstraction of a H atom, followed by an addition, resulting in the
original molecule. Starting from a closed-shell molecule regardless to the first step
being abstraction or addition, former studies revealed that this first step of the cycle
usually has an activation energy [4–7]. The second step is a reaction that includes a
free radical and a H atom, and it is generally a barrierless process. Consequently, the
rate-limiting process is the first step, which can take place via H-atom tunneling at
low temperatures.
It was suggested on the basis of astronomical observations by Habart and co-workers
that the high number of polycyclic aromatic hydrocarbons (PAHs) present in the
photon-dominated regions (PDRs) may be related to the rapid formation of molecular
hydrogen in these regions. Therefore, PAHs might act as catalysts in the H2forma-
tion in PDRs [15–18]. The potential catalytic role of various PAHs has already been
investigated both experimentally and theoretically [8–12].
Based on computations, the H-atom-addition reaction of benzene [8, 9] and pyrene
[10] have a reaction barrier that is permeable for H atoms by tunneling. Moreover, the
reaction rates of these reactions are non-negligible at low temperatures in contrast to
those of the H-atom-addition reactions of graphene or graphite. In addition, experi-
ments have also proved that H atoms and H2molecules can react with benzene and
small PAHs, forming so-called superhydrogenated species [19–22]. Another experimen-
tal evidence was provided by Menella and coworkers [11], showing that coronene can
react with deuterium atoms in a D-atom-addition reaction and the resulting species
is able to react with another deuterium atom, resulting in the formation of HD or D2
and the reformation of coronene [11]. Therefore, this experiment demonstrated that
neutral PAH molecules can act as catalysts in the formation of interstellar H2.
Computations and experiments have also revealed that the carbon atoms at dif-
ferent positions in PAHs have different reactivity, and there is a specific order of
sequences of hydrogenation of PAHs [26, 27]. It was shown that among the hydro-
2