American Chemical Society
jp8b10146_si_002.avi (1.11 MB)

Exploratory Direct Dynamics Simulations of 3O2 Reaction with Graphene at High Temperatures

Download (1.11 MB)
posted on 2018-11-27, 00:00 authored by Seenivasan Hariharan, Moumita Majumder, Ross Edel, Tim Grabnic, S. J. Sibener, William L. Hase
Direct chemical dynamics simulations at high temperatures of reaction between 3O2 and graphene containing varied number of defects were performed using the VENUS-MOPAC code. Graphene was modeled using (5a,6z)-periacene, a poly aromatic hydrocarbon with 5 and 6 benzene rings in the armchair and zigzag directions, respectively. Up to six defects were introduced by removing carbon atoms from the basal plane. Usage of the PM7/unrestricted Hartree–Fock (UHF) method, for the simulations, was validated by benchmarking singlet-triplet gaps of n-acenes and (5a,nz) periacenes with high-level theoretical calculations. PM7/UHF calculations showed that graphene with different number of vacancies has different ground electronic states. Dynamics simulations were performed for two 3O2 collision energies Ei of 0.4 and 0.7 eV, with the incident angle normal to the graphene plane at 1375 K. Collisions on graphene with one, two, three, and four vacancies (1C-, 2C-, 3C-, and 4C-vacant graphene) showed no reactive trajectories, mainly due to the nonavailability of reactive sites resulting from nascent site deactivation, a dynamical phenomenon. On the other hand, 3O2 dissociative chemisorption was observed for collisions on four- (with a different morphology), five- and six-vacant graphene (4C-2-, 5C- and 6C-vacant graphene). A strong morphology dependence was observed for the reaction conditions. On all reactive surfaces, larger reaction probabilities were observed for collisions at Ei = 0.7 eV. This is in agreement with the nucleation time measured by supersonic molecular beam experiments wherein about 2.5 times longer nucleation time for O2 impinging at 0.4 eV compared with 0.7 eV was observed. Reactivity at both collision energies, viz., 0.4 and 0.7 eV, showed the following trend: 5C- < 6C- < 4C-vacant graphene. Formation of carboxyl/semiquinone (CO)- and ether (−C–O–C−)-type dissociation products was observed on all reactive surfaces, whereas a higher probability of formation of the ether (−C–O–C−) group was found on 4C-vacant graphene on which dangling carbon atoms are present in close proximity. However, no gaseous CO/CO2 formation was observed on any of the graphene vacancies even for simulations that were run up to 10 ps. This is apparently the result of the absence of excess oxygen atoms that can aid the formation of larger groups, the precursors for CO/CO2 formation. Although the results of this study do not provide a conclusive understanding of the mechanism of graphene/graphite oxidation, this work serves as an initial study attempting to understand the 3O2 dissociative chemisorption dynamical mechanism on defective-graphene/graphite surfaces at high temperatures.