Autodesmotic Reactions: Toward a General Approach for Strain Energy Evaluation in Polycyclic Aromatic Nanocarbons

Accurate quantification of strain energy (SE) is essential for rational design and synthesis of aromatic nanocarbons, as intrinsic strain strongly affects their structure, stability, reactivity, and properties. Existing approaches often suffer from arbitrariness in reference selection, limited applicability, or high computational cost. In particular, the balance of π-energy between the target molecule and its strain-free references has long been overlooked, which is critical for π-conjugated systems. Here, we introduce “autodesmotic reactions” as a general and efficient strategy for reliable SE evaluation in polycyclic aromatic nanocarbons. An autodesmotic reaction transforms a strained molecule into a uniquely defined single-molecule reference. This reference resides within a virtual strain-free chemical space derived from physically motivated models trained on planar benzenoid polycyclic aromatic hydrocarbons (PAHs). Crucially, preservation of molecular topology ensures both unique reference choice and proper π-energy balance. Benchmarking against asymptotic and StrainViz models across diverse carbon nanobelts (CNBs) validates the accuracy and robustness of the approach. It is broadly applicable to both macrocyclic and nonmacrocyclic systems while requiring only a single quantum chemical calculation per molecule. Applications to more challenging systems, including helicene-containing and saddle-shaped CNBs, helicenes, bowl-shaped PAHs, carbon nanotubes, and fullerenes, demonstrate high computational efficiency and provide valuable insights into strain–structure–property relationships. Overall, autodesmotic reactions establish a general, conceptually transparent, and efficient framework for SE analysis in π-conjugated nanocarbons. The accuracy and generalizability of the current method can be further enhanced through rapidly advancing machine-learning techniques.