KEVIN C. GROSS,1 PAUL G. SEYBOLD,1 CHRISTOPHERM. HADAD2 1Department of Chemistry, Wright State University, Dayton, Ohio Department of Chemistry, The Ohio State University, Columbus, Ohio International Journal of Quantum Chemistry, Vol. 90, 445–458 (2002) © 2002 Wiley Periodicals, Inc.

Introduction Any definition of atomic charge in a molecule is immensely useful in its ability to correlate with the experimental properties such as pKa of a compound. A good correlation between aqueous acidity and easily calculated gas-phase properties is especially convenient if the atomic charge method is practical for rapid evaluation. For substituted anilines and phenols, the charges on the functional group heavy atom (e.g., N in – NH 2 or the O - of the OH) and on the acidic hydrogens(e.g., in –NH 3 + or OH) can serve as good regression parameters in determination of the pKa of the compound.

I ntroduction In the phenol series, the –OH group pKa varies with substitution, increasing with electron-donating group (EDGs) and decreasing with electron-withdrawing groups (EWGs). Three population analysis were examined for their ability to generate charges that correlate with the experimental pKa’s for a series a para and meta monosubstituted anilines and phenols. the Mulliken charge Q(H+) Mul, the natural population charge Q(H+) NPA electrostatic-potential Q(H+) ESP Geometry optimizations were carried out at the 6-31G(D) level for each compound.

Definitions Mulliken Analysis Divides overlap population equally between the two atoms of a bond and thus has advantage of simplicity. However, its results vary with the basis set employed inducing sometimes unnatural and ineffective results. Natural Population Analysis (NPA) The analysis of the electron density distribution in a molecular system is based on the orthonormal natural atomic orbitals. Natural populations, ni(A) are the occupancies of the natural atomic orbitals which rigorously satisfy the Pauli exclusion principle: 0 < ni(A) <2. Electrostatic Potential Analysis (ESP) Assigns point charges to the atomic centers in order to best reproduce the electrostatic potential.

Aniliniums m-Aminom- Methoxym- Nitro p-Nitrop- Aminop- Methoxy

Phenols m- Aminom- Methoxym- Nitro p-Amino p- Methoxyp- Nitro

4-tert-amylphenol 4-phenylphenol 4-ethylphenol4-hydroxybenzaldehyde Compounds used to verify the effectiveness of estimating pKa using different atomic charge methods.

Plots of the averaged amoni-group hydrogen pKa vs. charge using several charge definitions for a series of monosubstituted aniliniums: the Mulliken charge Q(H+) Mul, the natural population charge Q(H+) NPA, and the electrostatic-potential Q(H+) ESP. Results

Plots of the phenolic hydrogen pKa vs. charge using several charge definitions for a series of monosubstituted phenols: Q(H+) Mul, Q(H+) NPA, and Q(H+) ESP. and Q(H+) ESP. Results

Plots of the phenolic hydrogen pKa vs. charge using several charge definitions for a series of the para monosubstituted phenols as indicated on the graphs: Q(H+) Mul, Q(H+) NPA, and Q(H+) ESP. Results

These plots indicate the effectiveness of determining pKa values through different atomic charge methods by using the actual experimental values (known) and the training set of data points. Results Actual experimental values Training set of data points (phenols)

Conclusion NPA and Mulliken charges are observed to be quite useful and effective measures in correlating with the observed pKa’s of the known compound. ESP produced less effective results for the succesful correlation with the observed pKa’s as indicated by the regression results. Atomic charges methods are very useful for rapid evaluation and determination of the pKa for anilines and phenols as observed experimentally.

Thank You