Transition states have been obtained by fixing the length of the bond to be broken or
formed and relaxing the geometry. We stretch the bond in steps of 0.1 Å until a
maximum in energy, i.e. the bond length for the transition state, is found. For hydrogen
shifts these calculations must include the Ag(100) surface explicitly, as the transition may
be surface mediated, e.g. the [1,3]-hydrogen shift on Int2. As we are first searching for
the fixed bond length which describes the transition state, rather than its energy, we may
employ a simplified 6×6×1 slab model for the Ag(100) surface. On the other hand, when
forming C bonds, the surface plays less of a role in determining the bond length which
describes the transition state. To model the first cyclization barrier, we have used gas
phase calculations for the reaction pathway from reactant 1 to product 10. These methods
allow us to quickly obtain the required bond length for the transition state, which we then
employ within a 6×6×3 slab model to obtain accurate transition state energies. The
transition state energies as a function of bond length are shown in Fig. S3 for the (A)
[1,3]-hydrogen shift, (B) [1,2]-hydrogen shift, and (C) first cyclization step along with
schematics of the initial, transition, and final states. The initial cyclizations along the
pathways towards products 2 or 3 are expected to be energetically similar, as concluded
from the comparable product distribution after annealing to different temperatures. The
calculated transition state energy for the initial cyclization is furthermore in good
agreement with estimations from the 90 °C threshold temperature observed
experimentally, in combination with an Arrhenius function and typical attempt
frequencies on the order of 1013 Hz.
Transition states have been obtained by fixing the length of the bond to be broken orformed and relaxing the geometry. We stretch the bond in steps of 0.1 Å until amaximum in energy, i.e. the bond length for the transition state, is found. For hydrogenshifts these calculations must include the Ag(100) surface explicitly, as the transition maybe surface mediated, e.g. the [1,3]-hydrogen shift on Int2. As we are first searching forthe fixed bond length which describes the transition state, rather than its energy, we mayemploy a simplified 6×6×1 slab model for the Ag(100) surface. On the other hand, whenforming C bonds, the surface plays less of a role in determining the bond length whichdescribes the transition state. To model the first cyclization barrier, we have used gasphase calculations for the reaction pathway from reactant 1 to product 10. These methodsallow us to quickly obtain the required bond length for the transition state, which we thenemploy within a 6×6×3 slab model to obtain accurate transition state energies. Thetransition state energies as a function of bond length are shown in Fig. S3 for the (A)[1,3]-hydrogen shift, (B) [1,2]-hydrogen shift, and (C) first cyclization step along withschematics of the initial, transition, and final states. The initial cyclizations along thepathways towards products 2 or 3 are expected to be energetically similar, as concludedfrom the comparable product distribution after annealing to different temperatures. Thecalculated transition state energy for the initial cyclization is furthermore in goodagreement with estimations from the 90 °C threshold temperature observedexperimentally, in combination with an Arrhenius function and typical attemptfrequencies on the order of 1013 Hz.
การแปล กรุณารอสักครู่..
