Multiple linear regression is used to combine a series of experimental measurements and theoretically determined chemical kinetic quantities to derive a predictive methodology to estimate the anti-knock propensity of a series of a series of methyl substituted phenols. The derived cetane number (DCN) is used to determine the anti-knock effect of the additives on a research octane number (RON) 95 gasoline. The use of DCN allows the additive to be tested at one tenth of the scale compared with the analogous, and widely used, RON measurement.<br/>The mechanism of knock inhibition of these additives is proposed to involve the destruction of reactive radical species. Thus, physical quantities representative of this mechanism are determined and used as predictors. The O-H and C-H bond dissociation energies (BDEs) of the hydroxyl and methyl groups are determined computationally. The BDEs are then used to derive rate constants for hydrogen atom abstraction by OH and HO<sub>&lt;span style="font-size:10.8333px"&gt;2&lt;/span&gt;</sub> radicals. It is proposed that a higher rate of abstraction from the OH site yields a more effective anti-knock additive. The Gibbs free energy of formation for each molecule and corresponding radical is calculated and, used to indicate the relative stability of the radical species. Structural parameters, such as the degree of methylation are also determined.<br/>The theoretically determined quantities are combined, using multiple linear regression, as independent variables to generate a numerical model capable of predicting the effect of the additive on the DCN of the gasoline. Through sequential refinement of the model, key molecular properties for effective anti-knock additives are identified.

Development of a predictive methodology for estimating the anti-knock propensity of gasoline additives at the one-gram scale

ACS Fall 2020 Virtual Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.