Modeling the intrusion of molecules into graphite: Origin and shape of the barriers.

We performed density functional theory calculations to explore the energetic and geometric aspects of the permeation of H2, BeH x , OH x (x = 1, 2) and CH y (y = 1-4) through the central hexagon of coronene. Coronene serves as a cluster model for extended graphene which can be regarded as the first layer of a graphite (0 0 0 1) surface. We compare the energy barriers encountered by these molecular projectiles with the ones that are obtained for atomic H, Be, C and O. The barriers are substantially lower if projectiles possess free valences that can bind to the carbon entity. Furthermore, for some of the species fragmentation is observed. Implications with respect to plasma-surface interaction are discussed.

Introduction

Plasma-wall interactions (PWI) are the basis of plasma coating [1] and of the use of microplasmas for medical applications [2]. In thermo-nuclear research [3,4] they must be understood in terms of material degradation. PWI comprises how atoms and molecules and their ions in the plasma interact with surface materials in contact with them. Important aspects of PWI are how projectiles from the plasma or a hot gas in front of the wall can penetrate through the surface. In case of molecular projectiles, eventual fragmentation of the molecules is an issue either. The interaction of atoms and molecules with carbon materials is an especially important issue in thermo-nuclear research. Carbon fiber composites (CFC) are possibly used in the ITER (International Tokamak Experimental Reactor) device as partial coverage of the first wall [5]. There is vast experience with this material from earlier fusion experiments [6,7]. The main drawback of CFC is its high reactivity with respect to hydrogen [6]. It is expected that the latter will eventually lead to a substantial amount of tritium retained in the wall material as well as to the formation of hydrocarbon impurities polluting the fusion plasma and re-depositing at other parts of the wall [6]. The use of beryllium at the main wall of ITER [5] is expected to lead to the formation of beryllium hydride molecules forming another source of impurities. A rest concentration of oxygen will have the same effect leading to the formation of hydroxyl radicals and water molecules. Furthermore, recombination of hydrogen atoms leads to the formation of hydrogen molecules. Hence, the rather low temperatures of about 2–3 eV in the divertor region due to the realization of partial detachment [8,9] will give rise to a complex mixture of hydride molecules interacting with the CFC surface. For this reason, we want to explore some aspects of the interaction of such molecules with carbon sheets (as a model for the graphite structures of which CFC consists) in this work.

The interaction of molecular projectiles with fusion-relevant wall materials has been subject to a variety of experimental studies conducted in the last years [10-20]. Especially the groups of Hermann and of Märk conducted several mass-spectrometric experiments with respect to molecule/surface interactions [12,13,16-20]. They let deuterated hydrocarbon ions interact with stainless steel [10,19], carbon [11-15,17-20], beryllium [18] and tungsten [16,18]. All of these surfaces are of relevance in thermo-nuclear fusion research. The impact energies range from a few eV to hundreds of eV. Analyzed were the fragmentation processes of the incident molecules, sputtering of the material and effects due to heating of the surface. However, interpretation of experimental results is complicated due to technical difficulties such as contamination of the surfaces in course of their fabrication and unknown details of the surface structure which hinders the interpretation of the microscopic processes governing the investigated systems.

In contrast, simulations and (numerical) modeling can work on the atomistic level. Thus, PWI has also attracted a lot of interest from the theoretical side. Especially the interaction of hydrogen (and/or isotopes) with carbon-based materials has been studied extensively employing both force-field approaches [21-34] as well as quantum-chemical methods [35-43]. Also the interaction of molecules with fusion-relevant surfaces has been explored to some extent, mostly by employing force-field based simulations [28,30,34,44,45].

This study is of the quantum-chemical type. In particular, we use density functional theory (DFT) to explore the energetics of permeation of the small hydrides H, BeH, OH and CH with x = 1,2 and y = 1–4 through the center of coronene (C24H12). The latter molecule serves as a model of an extended graphene sheet. We are particularly interested in the effect of the saturation of free valences via the addition of hydrogen atoms on the energy barrier associated with the permeation process. The processes studied are far from chemical equilibrium and thus are difficult to describe quantitatively accurate. Hence, we rather aim for a qualitative exploration. Nevertheless, we will show that the trends explored indicate substantial differences in the energy barriers for atomic and molecular projectiles. This might eventually lead to a better understanding of some types of PWI.

The work is structured as follows. In Section 2, we summarize our method. In particular, in Section 2.1 we describe the cluster model used to represent the graphite (0 0 0 1) surface. In Section 2.2, we describe the quantum chemical methods employed. In Section 2.3, we describe how the energy barriers are obtained and how they can be decomposed further in order to aid their interpretation. In Section 3, we discuss our results. We begin with the discussion of the differences between the energy barriers for permeation of H and H2 in Section 3.1. This is followed by analogous discussions concerning the projectiles BeH, CH and OH with x = 0, 1, 2 and y = 0–4 in Sections 3.2–3.4, respectively. In Section 4, we discuss implications of our results in terms of conditions eventually met in future nuclear fusion applications. In Section 5, we summarize our conclusions.

Method

Cluster model

In order to model extended graphene sheets (serving per se as a model for a graphite (0 0 0 1) surface) we use a cluster approach. In particular, we use the polycyclic aromatic hydrocarbon (PAH) coronene (C24H12) as model. The molecule is depicted in Fig. 1(a). In an earlier theoretical work, it has been shown that although coronene appears rather small it is large enough to yield essential trends governing the interaction of atoms with extended graphene [43]. In this work, we adopt the strategy of this earlier investigation, but now we move molecules rather than atoms through the center of the central hexagon of coronene. By calculating the energy at each step we determine the barrier for permeation at this site. Except for the hydrogen atoms at the edge of coronene all atomic positions are relaxed corresponding to the adiabatic interaction regime, see reference [43]. This corresponds to a velocity of the projectiles which is small enough for the carbon atoms constituting the graphite surface to adapt instantly. This appears as a reasonable assumption given the low temperature of about 2–3 eV in the divertor of a nuclear fusion scenario as envisaged for ITER. This was modeled by moving the approaching molecule along the C6 axis of coronene (the z-axis), thus performing a relaxed potential energy surface scan of the system. The z-coordinates of the coronene hydrogen atoms are fixed at z = 0 and the z-coordinate of one atom of the approaching molecule is changed in steps of Δz = −0.2 Å and is also fixed while the remaining z-coordinates and all x and y coordinates of all other atoms are allowed to relax. The atoms of the approaching molecules with fixed z-coordinates have been (a) one hydrogen atom in H2, (b) the beryllium atom in BeH and BeH2, (c) the carbon atom in CH (y = 1–4) and (d) the oxygen atom in OH and H2O. A schematic depiction of this adiabatic permeation is shown in Fig. 1(b) for H2.

Fig. 1

(a) Coronene used as cluster model for extended graphene and (b) schematic depiction of the procedure used to obtain the energy barrier for permeation of H2 through coronene. H and C atoms are depicted as light blue and yellow spheres, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

DFT calculations

In order to obtain the energy profiles and intermediate geometries for permeation as described in the foregoing section we use unrestricted DFT calculations using the B3LYP [46] and PBE0 [47] functionals. Both are hybrid functionals and include a mixture of Hartree–Fock exchange and DFT exchange-correlation. The widely used B3LYP functional has been constructed semi-empirically by fitting its three parameters to experimental data. The parameter-free PBE0 functional is based on the fulfillment of a number of physical constraints. The independent and complimentary design of these functionals makes it worth to compare their results. As basis sets 6-31G [48] and 3-21G [49] have been chosen in conjunction with B3LYP and PBE0, respectively. Although they are small basis sets, it has been shown in earlier work [43] that basic trends are well preserved in them, compared with larger ones. The main reason for the choice of these functionals and basis sets is the comparability with a former work on the interaction of bare atoms with cluster models for the graphite (0 0 0 1) surface [43]. To our knowledge, it has never been investigated which combination of functional and small basis set would be the best for barriers. A comparison of thermochemical data obtained from a large number of density functionals combined with small basis sets has found functionals without HF-exchange to perform best [50]. On the other side, it is known that with basis sets like 6-31G* barriers can be calculated more accurately if more HF-exchange than 20 or 25% (PBE0 and B3LYP, respectively) is incorporated like in BH&HLYP [51,52]. For future studies on larger systems it might well be beneficial to check if functionals without HF-exchange perform equally good or better with the specific basis set for barriers and for other high-energy geometries like small interatomic distances or large distortions.

Using the mentioned model chemistries, the geometries of the target molecule coronene as well as the molecular projectiles have been optimized and then used as input for the calculation of the energy barriers. Another important issue is the choice of the spin configuration of the total system. It has been reported [43] that adiabatic barriers are lowest if the total system exhibits the lowest possible spin multiplicity. Our interest is also to find the lowest barriers and the spin configurations have been chosen accordingly. In case of C, O, CH and CH2 the lowest spin configurations, i.e. singlet, singlet, doublet and singlet, respectively, correspond to excited states (rather than the ground state configurations, i.e. triplet, triplet, quartet and triplet, respectively) of the projectiles for large distances between them and the target molecule. At a temperature equivalent to 2–3 eV, these excited spin states are accessible via collision induced electronic transitions. These states might therefore be at least partially occupied. For most of the other species like H, H2, Be, BeH2, OH2 and CH4, electronically excited states are not accessible at such temperatures. They may be more relevant for the remaining radicals BeH, OH and CH3. Electronic excitations in the surface model coronene are also not of interest in this temperature range due to the HOMO-LUMO gap of about 9 eV [53]. For the static calculations of the present work we restricted ourselves to the singlet and triplet cases mentioned. A more thorough analysis would require more demanding theoretical methods. All calculations were performed using the Gaussian 09 suite of programs [54].

Energy decomposition

The planar coronene becomes distorted when the projectile (either an atom or a molecule) approaches. In most cases, it bends away from the projectile, ‘outward’ of the original molecular plane, in order to reduce the repulsion between the electrons. The energy barrier Eb is thenwhere Etot(M − C) is the total energy of the (relaxed) system consisting of the projectile and the (deformed) coronene molecule, Etot(C) is the energy of the isolated coronene molecule in its equilibrium geometry and Etot(M) is the energy of the isolated projectile. Moreover, Eb can be decomposed into the energy required for the deformation, Edef > 0, of the coronene molecule and the rest energy containing contributions from temporary bonding and the deformation of the projectile molecule, ΔE, i.e.

The deformation energy Edef was calculated by using the deformed coronene geometries at each scan step. Subtraction of this energy from Eb then yields ΔE. The latter quantity has proven to be helpful for the interpretation of the individual processes as it accounts for eventually attractive interactions between the projectile and coronene, see particularly Sections 3.2–3.4.

Results and discussion

Energy barriers for H and H2

The energy barriers for adiabatic permeation of H and H2 through the center of coronene are depicted in the top panel of Fig. 2. First, we note that both model chemistries employed, i.e. PBE0/3-21G (black lines in the top panel of Fig. 2) and B3LYP/6-31G (red lines in the top panel of Fig. 2) yield very similar results. Second, we note that the energy barrier for H2 as projectile (dashed lines) is substantially higher than for the projectile H (solid lines), more specifically, it is about twice as high. This is a consequence of the different electronic configuration of these two projectile species. H is an open-shell doublet and H2 is a closed-shell singlet. It can be assumed that the latter projectile is thus chemically much more inert and this is nicely seen also in the bottom panel of Fig. 2. There, the natural charges as obtained from a natural bond orbital analysis [55] are plotted against the position of the projectile. In the central region of the barrier, H donates a significant fraction of its electronic density to the coronene molecule. It acquires thus a positive charge of about 0.6 e, see the black line (augmented with black circles) in the bottom panel of Fig. 2. In contrast, in H2 such a donation does not take place (dashed blue line augmented with squares in Fig. 2). By decomposing the total natural charge on H2 into individual contributions from the two hydrogen atoms (dashed red and green lines augmented with diamonds and triangles, respectively, in Fig. 2), we find that this is the case for each of the two atoms either rather than a cancelation between them. In summary, permeation of H is facilitated via temporary bonding, but this is not the case for H2 which has a much higher barrier.

Fig. 2

Energy barriers obtained for the permeation of H and H2 through coronene (top panel). Natural charges as obtained using PBE0/3-21G (bottom panel). (For interpretation of the references to color in the text, the reader is referred to the web version of the article.)

Energy barriers for the Be, BeH and BeH2

In Fig. 3, we depict the energy barriers for permeation of (a) Be, (b) BeH and (c) BeH2 through the center of coronene (black lines augmented with circles). For simplicity, we plot only the barriers obtained at the PBE0/3-21G level of theory (solid lines) as the curves obtained at the B3LYP/6-31G level of theory yield generally very similar results. In Fig. 3(b), however, we show also the data obtained with B3LYP/6-31G as for the PBE0/3-21G calculations beyond z = 1.2 Å no SCF-convergence could be achieved.

Fig. 3

Energy barriers obtained for the permeation of (a) Be, (b) BeH and (c) BeH2 through coronene. The total barriers are decomposed into the deformation energy Edef accounting for geometrical changes in the coronene molecule and the rest energy ΔE according to Eq. (2). The latter contribution is denoted simply as “difference” in the figure. (For interpretation of the references to color in the text, the reader is referred to the web version of the article.)

Again, the total energy barrier was decomposed into the contributions Edef (red lines augmented with squares) and ΔE (blue lines augmented with diamonds and denoted simply as “difference” in the figure) according to Eq. (2). This helps significantly to interpret the results. In contrast, a natural bond orbital analysis as done for the projectiles H and H2 in Section 3.1 turns out to be rather complicated and difficult to interpret in this case.

For the permeation of the atomic projectile Be through coronene, depicted in Fig. 3(a), we note that both the deformation energy Edef and the rest energy ΔE are positive in the central region of the barrier. The local minima in front of the barrier at z = −0.6 Å and after the barrier at z = 1.0 Å indicate at least a slight adaptation of the electronic configuration of the atomic projectile facilitating permeation. Nevertheless, the 2s subshell of Be would need energy to be promoted into a hybrid orbital before bonding and so both contributions carry a positive sign.

The situation is different for BeH for which the barrier and the two contributions to it are plotted in Fig. 3(b). Here, slightly attractive forces beyond z = −1.8 Å become apparent. They are much more pronounced for PBE0/3-21G but it is well known that B3LYP has difficulties with the reproduction of small attractive energies. These contributions lead to a somewhat lower energy barrier, compared to the atomic case. Furthermore, permeation (accompanied by flipping of the sheet) occurs later compared to the atomic case, at z = 1.2 Å. Upon permeation of Be, the H atom is stripped off BeH leading to isolated Be and H atoms in the outgoing channel. Due to this fragmentation process both the total energy as well as the rest energy ΔE are positive by the amount of binding energy of BeH (at this level of theory) on the right side of the barrier in Fig. 3(b). In contrast, the deformation energy Edef goes smoothly back to zero indicating that the coronene molecule remains undamaged.

In Fig. 3(c), the energy barrier for BeH2 is depicted. BeH2 is a closed-shell singlet species and as in the atomic case we observe no attractive contributions from ΔE as well as no local minimum in front of the barrier in Fig. 3(c). Both contributions to the total barrier are strictly positive over the entire range. However, in analogy to BeH, both H atoms are stripped off the BeH2 molecule. The local minimum after the barrier corresponds to the one observed for the atomic case. After the barrier, a positive total energy remains which is identical to the fragmentation energy required to dissociate BeH2 into Be and H2.

Energy barriers for the hydrocarbon projectiles CH, x = 0, …, 4

For the analysis of the results concerning (a) C, (b) CH, (c) CH2, (d) CH3 and (e) CH4 we adopt the same strategy, as for the BeH species presented in Section 3.2. The energy barriers and contributions Edef and ΔE are depicted in Fig. 4.

Fig. 4

Energy barriers obtained for the permeation of (a) C, (b) CH, (c) CH2, (d) CH3 and (e) CH4 through coronene. The total barriers are decomposed into the deformation energy Edef accounting for geometrical changes in the coronene molecule and the rest energy ΔE according to Eq. (2). The latter contribution is denoted simply as “difference” in the figure. H and C atoms are depicted as light blue and yellow spheres, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

For atomic C a strong attractive contribution accompanied by a strong repulsive deformation energy in the central region of the permeation barrier can be seen in Fig. 4(a). We find that in this case permeation is facilitated by temporary bonding which is reflected also in the intermediate geometry of the total system shown on the right of Fig. 4(a). The formation of three C—C bonds upon approach of the projectile is in line with the amount of negative energy of about −10 eV contained in ΔE. The resulting distortion of the coronene molecule, however, significantly overcompensates this gain in energy due to temporary bonding.

For CH, we observe similarities with the atomic case, occurring at a somewhat closer distance to the sheet: Energy is gained due to temporary bonding of CH to three adjacent C atoms of the target molecule. This is again accompanied by a strong distortion of the coronene molecule as shown in Fig. 4(b) to the right. After permeation of the C atom from CH the H atom is stripped off as was observed already for the BeH species, see Section 3.2. However, when the C atom departs from the target molecule, one of the C atoms of coronene remains attached such that a rather high deformation is remaining over the rest of the scan range. However, it can be expected that this situation becomes more and more unfavorable as the C atom departs further such that for larger distances a partially hydrogenated coronene and an isolated C atom will remain.

For the projectile CH2 temporary bonding to three C atoms of the coronene molecule is not favorable. This is also nicely reflected in Fig. 4(c). During permeation the central hexagon of coronene is cut in two pieces. The permeating CH2 molecule forms a temporary pentagon with four of the C atoms, while the other two C atoms exhibit a stronger bonding than in the original situation.

For CH3 temporary bonding can facilitate permeation only via the formation of one C-C bond as can be seen in the region beyond z = 1 Å in Fig. 4(d). Beyond z = 1.8 Å fragmentation of into CH2 and H takes place. For larger z values no convergence could be achieved, but it is expected that for the remaining part of the scan range the coronene molecule would flip back and one would remain with C24H13 and CH2. In general, if the geometry and thus the wavefunction changes substantially and abruptly, it can be difficult to achieve convergence in the SCF procedure.

For the projectile CH4, we observe a relatively early fragmentation into CH2 and H2 upon approaching coronene, see Fig. 4(e). After this fragmentation the permeation is facilitated by the formation of two temporary C—C bonds as in the case of CH2discussed above but in a somewhat more symmetrical manner.

Energy barriers for the projectiles OH, x = 0,1,2

For the chosen nomenclature and representation of our results see Section 3.2. The energy barriers and contributions Edef and ΔE for the projectiles (a) O, (b) OH and (c) OH2 are depicted in Fig. 5.

Fig. 5

Energy barriers obtained for permeation of (a) O, (b) OH and (c) OH2 through coronene. The total barriers are decomposed into the deformation energy Edef accounting for geometrical changes in the coronene molecule and the rest energy ΔE according to Eq. (2). The latter contribution is denoted simply as “difference” in the figure. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

The atomic projectile O exhibits significantly negative ΔE contributions in the central region of the barrier, see Fig. 5(a). Similar to the case of CH2 discussed in Section 3.3, this is accompanied by temporary bonding between O and two C atoms of the central hexagonal ring in coronene as well as enhanced deformation of the target molecule.

For OH, we observe a jump in the ΔE contribution at larger values than for the atomic projectile O, see Fig. 5(b). This stronger repulsion for smaller values of z is also familiar from the so far discussed hydrogenated species. Unfortunately, we could not calculate the data for values of z larger than 1.4 Å. However, our results at least indicate that for these values the O atom will have passed the coronene molecule while the H atom is stripped off as it was observed already for BeH and CH.

For water as projectile, i.e. OH2, we note for z < 1.2 Å also a strongly enhanced repulsion, see Fig. 5(c). This can be expected from the closed-shell configuration of the water molecule. Beyond z = 1.2 Å, however, we observe a significant attractive contribution from ΔE indicating temporary bonding in the region between 1.4 and 2.0 Å. The mechanism underlying this feature is depicted in Fig. 6. At z = 1.2 Å the coronene is strongly distorted and repelled by the water molecule. At z = 1.4 Å, however, one of the H atoms (light blue sphere in Fig. 6) of the water molecule is donated to one of the carbon atoms (yellow spheres in Fig. 6) of the central hexagon of coronene. At z = 1.6 Å both of the hydrogen atoms have been donated and O (red sphere in Fig. 6) binds to two C atoms. This situation remains also for z = 1.8 Å. At z = 2.0 Å the departing O atom re-captures the two H atoms and a water molecule is formed again.

Fig. 6

Permeation mechanism of water through the center of coronene. H, C and O atoms are depicted as light blue, yellow and red spheres, respectively. Values for z are given in Å. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

Discussion

We note for the total energy barriers for permeation of the different species through coronene, discussed in Sections 3.3 and 3.4, that these barriers generally exhibit various local maxima. One of those is usually located in front of the region in which temporary bonding as discussed above occurs. In the cases of CH2, CH3 and OH2, these maxima are also the global maxima, whereas in all other cases they are at least of similar height with the global ones. Hence, projectiles with energies well below these pre-barrier maxima will be repelled way before they could be subject to fragmentation. In a nuclear fusion scenario corresponding to partial detachment with a temperature of a few eV in the divertor region, most of the projectiles will have such low energies. Their dynamics is then governed by these pre-barriers.

For the reason given above, we collect the values of the energy barriers for the range of positions of the projectile species BeH, CH and OH corresponding to z = 0.0–1.0 Å in Fig. 7. These results can be interpreted as follows. From H and H2, (not included in Fig. 7), we have already observed that the permeation barrier depends strongly on the electronic configuration. For atomic H temporary bonding is possible when approaching the coronene molecule, but not for H2 due to its closed-shell nature. The same holds for the more complex molecules. As noted in Section 3.2, Be exhibits a closed subshell, while BeH exhibits a valence electron allowing temporary bonding, and BeH2 corresponds to a closed-shell species. This is reflected in the pre-barriers in Fig. 7. In particular, the barriers are smaller for BeH than for Be, and larger for BeH2 than for BeH, see the left panel in Fig. 7. The exception for z = 1.0 Å in Fig. 7 is due to the different locations of the barrier maxima for the different species. At this z value, Be and BeH2 have already passed the coronene molecule, whereas this happens later for BeH.

Fig. 7

Comparison of barrier heights at different positions of the impact reference atom in dependence of the number of hydrogen atoms in the BeH, CH and OH species considered in this work. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

For the CHy species, we note that both C and CH allow temporary bonding to three C atoms of the coronene molecule. However, for the CH molecule the repulsion between the projectile and the target molecule can be expected to be bit larger due to higher electron density due to its C—H bond. Attaching two/three H atoms to C, which results in a CH2/3 molecule reduces the number of possible temporary bonds to two/one in contrast to the foregoing cases. The expected order of permeation barriers is thus Eb(CH3) > Eb(CH2) > Eb(CH) > Eb(C). This is indeed the case as can be seen from the central panel in Fig. 7. The early fragmentation of CH4 into CH2 and H2 upon approaching the coronene surface could lead to the expectation that it would resemble CH2. However, the cost for breaking two C—H bonds is not fully compensated by the formation of the H2 molecule and thus the resulting barrier is between the ones for CH2 and CH3 (Fig. 7).

For the last species considered in this work, OH with x = 0, 1, 2, we observe that the barriers are the higher the more of the free valences of the projectile are saturated. In particular, the atomic projectile O yields two free valences, whereas the hydroxyl radical OH yields one, and the water molecule, OH2 exhibits a closed-shell configuration. Inspection of Fig. 7 shows that the permeation barriers for these projectiles yield the thus expected order, i.e. Eb(OH2) > Eb(OH) > Eb(O).

Concerning the relevance of the work to fusion setups like ITER, which will exhibit a complex materials mixture [5,6], the observation of large differences in permeation capability of species in a chemical equilibrium with each other might eventually have implications for the use of carbon in the divertor region. Assumed that low temperatures accompanied by high gas densities and high recombination rates are present in this region, a significant fraction of the impurities in front of the wall will consist of neutral molecules, rather than atomized or ionized species [56]. Those molecular species will then experience a stronger repulsion upon approaching the surface due to their limited number of free valences, as our results indicate. Hence, the situation in such a scenario might be substantially different from exploratory experiments in which materials are exclusively bombarded by energetic ions. In especially, we expect that effects like material degradation, tritium retention and impurity production will be rather sensitive to operation parameters such as the degree of detachment realized in the reactor.

Conclusion

We calculated energy barriers for the adiabatic permeation of H, H2, BeH as well as CH and OH with x = 0, 1, 2 and y = 0–4 through the central hexagon of coronene. We find that the permeation barrier can be lowered by temporary bonding depending strongly on the electronic configuration of the projectile species. In many cases, we observe energetically driven fragmentation processes, generally occurring already upon close approach to the target molecule. The dynamics of projectiles with energies of a few eV is governed by barriers that are found to be lowered in cases of temporary bonding which is in turn enabled by free valences in the projectiles. For comparable systems, the strength of the bonding is related to the number of free valences. In particular, some barriers are found to be as much as twice as high for fully hydrogenated species as for the corresponding bare atoms. It is argued that therefore caution must be exercised when experiments using ion beams are compared to the scenario in plasmas of a few eV ion temperature.