PNNL

Abstract

The heats of formation of carbonate, bicarbonate and bicarbonate/hydroxide metal complexes and hydrated versions these complexes of Mg2+, Ca2+, Fe2+, and Cd2+ are predicted from atomization energies using correlated molecular orbital theory at the CCSD(T) level extrapolated to the complete basis set limit following the Feller-Peterson-Dixon (FPD) approach. Using the calculated gas phase values and the available experimental solid state values, the cohesive energies of selected minerals were calculated. The gas phase decomposition energies into MO, CO2 and H2O follow the order Mg ~ Ca > Cd ~ Fe, and correlate with the hardness of the metal +2 ions. Gas phase hydration energies show that the order is Mg > Fe > Ca ~ Cd. There are a number of bulk hydrated Mg and Ca complexes that occur as minerals but there are few if any for Fe and Cd, suggesting that a number of factors are important in determining the stability of the bulk mineral hydrates. The FPD heats of formation were used to benchmark a range of density functional theory exchange-correlation functionals, including those commonly used in solid state mineral calculations. None of the functionals provided chemical accuracy agreement (± 1 kcal/mol) with the FPD results The best functionals are ?B97X and ?B97X-D with mean average unsigned errors of 10 kcal/mol. The worst functionals PW91, BP86, and PBE with mean average unsigned errors of 32 to 36 kcal/mol. Much of the functional error in the heats of formation is due to an incorrect treatment of the O-O bond dissociation energy in O2 which affects the heat of formation of the O atom in the gas phase needed to obtain heats of formation from atomization energies.