Activation-Strain Analysis Reveals Unexpected Origin of Fast Reactivity in Heteroaromatic Azadiene Inverse-Electron-Demand Diels–Alder Cycloadditions

Abstract

Heteroaromatic azadienes, especially 1,2,4,5-tetrazines, are extremely reactive partners with alkenes in inverse-electron-demand Diels-Alder reactions. Azadiene cycloaddition reactions are used to construct heterocycles in synthesis and are popular as bioorthogonal reactions. The origin of fast azadiene cycloaddition reactivity is classically attributed to the inverse frontier molecular orbital (FMO) interaction between the azadiene LUMO and alkene HOMO. Here, we use a combination of ab initio, density functional theory, and activation-strain model calculations to analyze physical interactions in heteroaromatic azadiene-alkene cycloaddition transition states. We find that FMO interactions do not control reactivity because, while the inverse FMO interaction becomes more stabilizing, there is a decrease in the forward FMO interaction that is offsetting. Rather, fast cycloadditions are due to a decrease in closed-shell Pauli repulsion between cycloaddition partners. The kinetic-thermodynamic relationship found for these inverse-electron-demand cycloadditions is also due to the trend in closed-shell repulsion in the cycloadducts. Cycloaddition regioselectivity, however, is the result of differences in occupied-unoccupied orbital interactions due to orbital overlap. These results provide a new predictive model and correct physical basis for heteroaromatic azadiene reactivity and regioselectivity with alkene dieneophiles.

Understanding the Reactivity of Endohedral Metallofullerenes: C78versus Sc3N@C78

“…[22] In contrast, as shown in Figure4,t he DE Pauli term is less destabilizing in the [5,6]-pathway for the cycloaddition involvingthe larger fullereneC 78 .Asaconsequence, the computed DDE int is not that significant (DDE int = 3.1 kcal mol À1 ,a tC ···C distance of 2.4 ) for the process involving the latter species, which clearly contributes to the lower computed [6,6]/ [5,6] activation barrierdifference. [41] We then focused on the reaction involving the metallofullerene Sc 3 N@C 78 .T he corresponding reactionp rofiles for the most reactive [6,6]-1 and [6,6]-6 bonds and [5,6]-c bond (see above)a re depicted in Figure 5. On this occasion, the pathway forming the [6,6]-6-adduct through the saddle point [6,6]-6-TS is clearly favoured both kinetically andt hermodynamically.…”

Reactivity and Selectivity of Bowl‐Shaped Polycyclic Aromatic Hydrocarbons: Relationship to C60

“…[42] This finding confirms, once again, that qualitative FMO arguments can not be based only on orbital-energy gaps and that overlap must be taken into account. [43][44][45] …”

Understanding the Reactivity of Ion‐Encapsulated Fullerenes

“…[9,10] Indeed, Li + @C 60 shows greatly enhanced reactivity in photoinduced electron-transfer reductionsw ithe lectron donors compared to hollow C 60 . [10b] Note that, as has repeatedly been reported, [12] the use of frontier molecular orbital( FMO) argumentst or ationalize the reactivity,p articularly in pericyclic reactions,m ay lead to misleading conclusions, as the FMO interactions are exclusively computed at the equilibriumg eometries of the reactants,w hicht herefore ignores interactions occurringi nt he transition-state region or at anyo ther point along the reaction coordinate. [10] Thus, the observed activation barrierf or the DA reactiono fL i + @C 60 and C 6 H 8 (11.0 kcal mol À1 )w as about6kcalmol À1 lower than that for the reaction involving empty C 60 (16.8 kcal mol À1 ).…”

“…[10b] This enhancedD Ar eactivity has been qualitatively attributed to the decrease of the HOMO(diene)-LUMO(fullerene)g ap as ac onsequence of stabilization of the Li + @C 60 LUMO (E LUMO = À3.74and À2.70 eV for Li + @C 60 and C 60 ,r espectively). [10b] Note that, as has repeatedly been reported, [12] the use of frontier molecular orbital( FMO) argumentst or ationalize the reactivity,p articularly in pericyclic reactions,m ay lead to misleading conclusions, as the FMO interactions are exclusively computed at the equilibriumg eometries of the reactants,w hicht herefore ignores interactions occurringi nt he transition-state region or at anyo ther point along the reaction coordinate.…”