This article investigates how molecular topology governs stability in polycyclic aromatic hydrocarbons (PAHs) using a Z₁-based composite descriptor, C(G) = log₁₀W + log₁₀Z₁ + log₁₀R + log₁₀MDP(G,2). Indices were computed on 2D, hydrogen-suppressed heavy-atom graphs for nine PAHs (naphthalene → coronene) and paired with a normalized stability score derived from quantum-chemical energies (total, per-atom, per-ring). A single-descriptor model, S = a + bCG, captured the baseline trend (R² = 0.245). A parsimonious multi-descriptor model combining CG, E_atom, and fusion-topology dummies improved fit (R² = 0.729). The best performance was delivered by an enhanced model, S = α + β₁C(G) + β₂E_atom + β₃E ring + β₄( C(G)× E_atom), achieving R² = 0.962, adjusted R² = 0.924, and Q² = 0.855 from LOOCV (leave-one-out cross-validation) with n = 9. Descriptor-level “bifurcation” diagnostics on C_G reveal interpretable regime shifts—anthracene → phenanthrene (angularization), phenanthrene → pyrene (compact pericondensation), and pyrene → tetracene (acene lengthening), highlighting sensitivity to fusion topology beyond ring count. Overall, the Z₁-based composite is compact and interpretable; when minimally augmented with energetic terms, it provides accurate small-sample prediction and clear structural insight, and it is readily extensible to larger PAH sets and heteroatom variants.

How to Cite this paper?