Search

The energetics of the gas-phase S(sub N)2 reactions Y(super -) + CH3X right arrow CH3Y + X(super -) (X,Y = F,Cl, Br), were studied using the W1 and W2 ab initio computational thermochemistry methods. The nonidentity S(sub N)2 reactions and F/F identity reaction possess transition state structures below the reactants energy while Cl/Cl and Br/Br transition structures are above the reactants energy.

A new benchmark study has been performed for six isomers of C20 and four isomers of C24 using explicitly correlated methods, together with coupled cluster theory with large basis sets and DFT with advanced functionals. The relative energy trends obtained are extremely sensitive to the methods used. Combining our best CCSD(T)-MP2 difference with our best MP2 basis set limit, the dehydrocorannulene bowl is found to be the most stable for C20, followed by the cage at about 8 kcal/mol, and the ring at about 46 kcal/mol. For C24, the D3d cage is found to be the most stable isomer, followed at only a few kilocalories per mole by dehydrocoronene, and at larger separations by then octahedral cage and the ring, respectively. This makes C24 the smallest classical fullerene. The estimated residual basis set error of the estimated CCSD(T) basis set limit is conservatively expected to be ±1 kcal/mol. In general, DFT exhibits large errors for relative energies with RMSD values in the 8-34 kcal/mol range. However, among the DFT functionals, the DSD-PBEP86-D3BJ double hybrid comes close to our best ab initio results, while the ωB97X-V range-separated hybrid is in semiquantitative agreement. [ABSTRACT FROM AUTHOR]

Inearlier work [Gruzman, D.; Karton, A.; Martin, J. M. L.J. Phys. Chem. A2009, 113, 11974], we showed that conformer energiesin alkanes (and other systems) are highly dispersion-driven and thatuncorrected DFT functionals fail badly at reproducing them, whilesimple empirical dispersion corrections tend to overcorrect. To gaingreater insight into the nature of the phenomenon, we have mappedthe torsional surface of n-pentane to 10-degree resolutionat the CCSD(T)-F12 level near the basis set limit. The data obtainedhave been decomposed by order of perturbation theory, excitation level,and same-spin vs opposite-spin character. A large number of approximateelectronic structure methods have been considered, as well as severalempirical dispersion corrections. Our chief conclusions are as follows:(a) the effect of dispersion is dominated by same-spin correlation(or triplet-pair correlation, from a different perspective); (b) singlet-paircorrelation is important for the surface, but qualitatively very dissimilarto the dispersion component; (c) single and double excitations beyondthird order are essentially unimportant for this surface; (d) connectedtriple excitations do play a role but are statistically very similarto the MP2 singlet-pair correlation; (e) the form of the damping functionis crucial for good performance of empirical dispersion corrections;(f) at least in the lower-energy regions, SCS-MP2 and especially MP2.5perform very well; (g) novel spin-component scaled double hybrid functionalssuch as DSD-PBEP86-D2 acquit themselves very well for this problem. [ABSTRACT FROM AUTHOR]

Basis set extrapolations are typically rationalized either from analytical arguments involving the partial-wave or principal expansions of the correlation energy in helium-like systems or from fitting extrapolation parameters to reference energetics for a small(ish) training set. Seeking to avoid both, we explore a third alternative: extracting extrapolation parameters from the requirement that the BSSE (basis set superposition error) should vanish at the complete basis set limit. We find this to be a viable approach provided that the underlying basis sets are not too small and reasonably well balanced. For basis sets not augmented by diffuse functions, BSSE minimization and energy fitting yield quite similar parameters.

We report an update and enhancement of the ACONFL (conformer energies of large alkanes [ J. Phys. Chem. A 2022, 126 , 3521-3535]) dataset. For the ACONF12 ( n -dodecane) subset, we report basis set limit canonical coupled-cluster with singles, doubles, and perturbative triples [i.e., CCSD(T)] reference data obtained from the MP2-F12/cc-pV{T,Q}Z-F12 extrapolation, [CCSD(F12*)-MP2-F12]/aug-cc-pVTZ-F12, and a (T) correction from conventional CCSD(T)/aug-cc-pV{D,T}Z calculations. Then, we explored the performance of a variety of single and composite localized-orbital CCSD(T) approximations, ultimately finding an affordable localized natural orbital CCSD(T) [LNO-CCSD(T)]-based post-MP2 correction that agrees to 0.006 kcal/mol mean absolute deviation with the revised canonical reference data. In tandem with canonical MP2-F12 complete basis set extrapolation, this was then used to re-evaluate the ACONF16 and ACONF20 subsets for n- hexadecane and n- icosane, respectively. Combining those with the revised canonical reference data for the dodecane conformers (i.e., ACONF12 subset), a revised ACONFL set was obtained. It was then used to assess the performance of different localized-orbital coupled-cluster approaches, such as pair natural orbital localized CCSD(T) [PNO-LCCSD(T)] as implemented in MOLPRO, DLPNO-CCSD(T 0 ) and DLPNO-CCSD(T 1 ) as implemented in ORCA, and LNO-CCSD(T) as implemented in MRCC, at their respective "Normal", "Tight", "vTight", and "vvTight" accuracy settings. For a given accuracy threshold and basis set, DLPNO-CCSD(T 1 ) and DLPNO-CCSD(T 0 ) perform comparably. With "VeryTightPNO" cutoffs, explicitly correlated DLPNO-CCSD(T 1 )-F12/VDZ-F12 is the best pick among all the DLPNO-based methods tested. To isolate basis set incompleteness from localized-orbital-related truncation errors (domain, LNOs), we have also compared the localized coupled-cluster approaches with canonical DF-CCSD(T)/aug-cc-pVTZ for the ACONF12 set. We found that gradually tightening the cutoffs improves the performance of LNO-CCSD(T), and using a composite scheme such as vTight + 0.50[vTight - Tight] improves things further. For DLPNO-CCSD(T 1 ), "TightPNO" and "VeryTightPNO" offer a statistically similar accuracy, which gets slightly better when T CutPNO is extrapolated to the complete PNO space limit. Similar to Brauer et al.'s [ Phys. Chem. Chem. Phys. 2016, 18 (31), 20905-20925] previous report for the S66x8 noncovalent interactions, the dispersion-corrected direct random phase approximation (dRPA)-based double hybrids perform remarkably well for the ACONFL set. While the revised reference data do not affect any conclusions on the less accurate methods, they may upend orderings for more accurate methods with error statistics on the same order as the difference between reference datasets.

We have explored the use of range separation as a possible avenue for further improvement on our revDSD minimally empirical double hybrid functionals. Such ωDSD functionals encompass the XYG3 type of double hybrid ( i.e. , x DSD) as a special case for ω → 0. As in our previous studies, the large and chemically diverse GMTKN55 benchmark suite was used for evaluation. Especially when using the D4 rather than D3BJ dispersion model, x DSD has a slight performance advantage in WTMAD2. As in previous studies, PBEP86 is the winning combination for the semilocal parts. x DSD n -PBEP86-D4 marginally outperforms the previous "best in class" ωB97M(2) Berkeley double hybrid but without range separation and using fewer than half the number of empirical parameters. Range separation turns out to offer only marginal further improvements on GMTKN55 itself. While ωB97M(2) still yields better performance for small-molecule thermochemistry, this is compensated in WTMAD2 by the superior performance of the new functionals for conformer equilibria. Results for two external test sets with pronounced static correlation effects may indicate that range-separated double hybrids are more resilient to such effects.

For revDSD double hybrids, the Görling-Levy second-order perturbation theory component is an Achilles' heel when applied to systems with significant near-degeneracy ("static") correlation. We have explored its replacement by the direct random phase approximation (dRPA), inspired by the SCS-dRPA75 functional of Kállay and co-workers. The addition to the final energy of both a D4 empirical dispersion correction and of a semilocal correlation component lead to significant improvements, with DSD-PBEdRPA 75 -D4 approaching the performance of revDSD-PBEP86-D4 and the Berkeley ωB97M(2). This form appears to be fairly insensitive to the choice of the semilocal functional but does exhibit stronger basis set sensitivity than the PT2-based double hybrids (due to much larger prefactors for the nonlocal correlation). As an alternative, we explored adding an MP3-like correction term (in a medium-sized basis set) to a range-separated ωDSD-PBEP86-D4 double hybrid and found it to have significantly lower WTMAD2 (weighted mean absolute deviation) for the large and chemically diverse GMTKN55 benchmark suite; the added computational cost can be mitigated through density fitting techniques.

Expanded porphyrins provide a versatile route to molecular switching devices due to their ability to shift between several π-conjugation topologies encoding distinct properties. DFT remains the workhorse for modeling such extended macrocycles, when taking into account their size and huge conformational flexibility. Nevertheless, the stability of Hückel and Möbius conformers depends on a complex interplay of different factors, such as hydrogen bonding, π···π stacking, steric effects, ring strain, and electron delocalization. As a consequence, the selection of an exchange-correlation functional for describing the energy profile of topological switches is very difficult. For these reasons, we have examined the performance of a variety of wave function methods and density functionals for describing the thermochemistry and kinetics of topology interconversions across a wide range of macrocycles. Especially for hexa- and heptaphyrins, the Möbius structures have a stronger degree of static correlation than the Hückel and twisted-Hückel structures, and as a result the relative energies of singly twisted structures are a challenging test for electronic structure methods. Comparison of limited orbital space full CI calculations with CCSD(T) calculations within the same active spaces shows that post-CCSD(T) correlation contributions to relative energies are very minor. At the same time, relative energies are weakly sensitive to further basis set expansion, as proven by the minor energy differences between the extrapolated MP2/CBS energies estimated from cc-pV{T,Q}Z, diffuse-augmented heavy-aug-cc-pV{T,Q}Z and explicitly correlated MP2-F12/cc-pVDZ-F12 calculations. Hence, our CCSD(T) reference values are reasonably well-converged in both 1-particle and n -particle spaces. While conventional MP2 and MP3 yield very poor results, SCS-MP2 and particularly SOS-MP2 and SCS-MP3 agree to better than 1 kcal mol -1 with the CCSD(T) relative energies. Regarding DFT methods, the range-separated double hybrids, such as ωB97M(2) and B2GP-PLYP, outperform other functionals with RMSDs of 0.6 and 0.8 kcal mol -1 , respectively. While the original DSD-PBEP86 double hybrid performs fairly poorly for these extended π-systems, the errors drop down to 1.9 kcal mol -1 for the revised revDOD-PBEP86-NL, which eliminates the same-spin correlation energy. Minnesota meta-GGA functionals with high fractions of exact exchange (M06-2X and M08-HX) also perform reasonably well, outperforming more robust and significantly less empirically parametrized functionals like SCAN0-D3.

A new benchmark study has been performed for six isomers of C20 and four isomers of C24 using explicitly correlated methods, together with coupled cluster theory with large basis sets and DFT with advanced functionals. The relative energy trends obtained are extremely sensitive to the methods used. Combining our best CCSD(T)-MP2 difference with our best MP2 basis set limit, the dehydrocorannulene bowl is found to be the most stable for C20, followed by the cage at about 8 kcal/mol, and the ring at about 46 kcal/mol. For C24, the D3d cage is found to be the most stable isomer, followed at only a few kilocalories per mole by dehydrocoronene, and at larger separations by then octahedral cage and the ring, respectively. This makes C24 the smallest classical fullerene. The estimated residual basis set error of the estimated CCSD(T) basis set limit is conservatively expected to be ±1 kcal/mol. In general, DFT exhibits large errors for relative energies with RMSD values in the 8-34 kcal/mol range. However, among the DFT functionals, the DSD-PBEP86-D3BJ double hybrid comes close to our best ab initio results, while the ωB97X-V range-separated hybrid is in semiquantitative agreement.

We have obtained uniform frequency scaling factors λ(harm) (for harmonic frequencies), λ(fund) (for fundamentals), and λ(ZPVE) (for zero-point vibrational energies (ZPVEs)) for the Weigend-Ahlrichs and other selected basis sets for MP2, SCS-MP2, and a variety of DFT functionals including double hybrids. For selected levels of theory, we have also obtained scaling factors for true anharmonic fundamentals and ZPVEs obtained from quartic force fields. For harmonic frequencies, the double hybrids B2PLYP, B2GP-PLYP, and DSD-PBEP86 clearly yield the best performance at RMSD = 10-12 cm(-1) for def2-TZVP and larger basis sets, compared to 5 cm(-1) at the CCSD(T) basis set limit. For ZPVEs, again, the double hybrids are the best performers, reaching root-mean-square deviations (RMSDs) as low as 0.05 kcal/mol, but even mainstream functionals like B3LYP can get down to 0.10 kcal/mol. Explicitly anharmonic ZPVEs only are marginally more accurate. For fundamentals, however, simple uniform scaling is clearly inadequate.

The alkali halides sodium fluoride, sodium bromide, and sodium iodide exist in the gas phase as both monomer and dimer species. A reanalysis of gas electron diffraction (GED) data collected earlier has been undertaken for each of these molecules using the EXPRESS method to yield experimental equilibrium structures. EXPRESS allows amplitudes of vibration to be estimated and correction terms to be applied to each pair of atoms in the refinement model. These quantities are calculated from the ab initio potential-energy surfaces corresponding to the vibrational modes of the monomer and dimer. Because they include many of the effects associated with large-amplitude modes of vibration and anharmonicity, we have been able to determine highly accurate experimental structures. These results are found to be in good agreement with those from high-level core-valence ab initio calculations and are substantially more precise than those obtained in previous structural studies.

Reference quality conformational energies have been obtained for the 52 unique conformers of melatonin by means of explicitly correlated ab initio methods as well as the ccCA composite method. These data have then been used to evaluate more approximate methods, including a variety of density functionals both on their own and paired with various empirical dispersion corrections. Owing to the presence of internal contacts of the C-H···O and C-H···N variety, basis set convergence is much slower than for alkane conformers, for example, and basis sets of aug-cc-pVQZ or def2-QZVP quality seem to be required to obtain firm estimates of the basis set limit. Not just HF, but also many DFT functionals, will transpose the two lowest conformers unless empirical dispersion corrections are added. Somewhat surprisingly, many DFT functionals reproduce the reference data to fairly high accuracy when combined with the D3BJ empirical dispersion correction or the "nonlocal" Vydrov-Van Voorhis dispersion model. The two best performers including dispersion corrections are the double hybrids DSD-PBEP86-D3BJ and B2GP-PLYP-D; if no such correction is permitted, then M06-2X puts in the best performance. Of lower-cost ab initio-like models, MP2.5 yields the best performance, followed by SCS-MP2.

The dynamics of O((3)P) + CO(2) collisions at hyperthermal energies were investigated experimentally and theoretically. Crossed-molecular-beams experiments at = 98.8 kcal mol(-1) were performed with isotopically labeled (12)C(18)O(2) to distinguish products of nonreactive scattering from those of reactive scattering. The following product channels were observed: elastic and inelastic scattering ((16)O((3)P) + (12)C(18)O(2)), isotope exchange ((18)O + (16)O(12)C(18)O), and oxygen-atom abstraction ((18)O(16)O + (12)C(18)O). Stationary points on the two lowest triplet potential energy surfaces of the O((3)P) + CO(2) system were characterized at the CCSD(T)/aug-cc-pVTZ level of theory and by means of W4 theory, which represents an approximation to the relativistic basis set limit, full-configuration-interaction (FCI) energy. The calculations predict a planar CO(3)(C(2v), (3)A'') intermediate that lies 16.3 kcal mol(-1) (W4 FCI excluding zero point energy) above reactants and is approached by a C(2v) transition state with energy 24.08 kcal mol(-1). Quasi-classical trajectory (QCT) calculations with collision energies in the range 23-150 kcal mol(-1) were performed at the B3LYP/6-311G(d) and BMK/6-311G(d) levels. Both reactive channels observed in the experiment were predicted by these calculations. In the isotope exchange reaction, the experimental center-of-mass (c.m.) angular distribution, T(θ(c.m.)), of the (16)O(12)C(18)O products peaked along the initial CO(2) direction (backward relative to the direction of the reagent O atoms), with a smaller isotropic component. The product translational energy distribution, P(E(T)), had a relatively low average of = 35 kcal mol(-1), indicating that the (16)O(12)C(18)O products were formed with substantial internal energy. The QCT calculations give c.m. P(E(T)) and T(θ(c.m.)) distributions and a relative product yield that agree qualitatively with the experimental results, and the trajectories indicate that exchange occurs through a short-lived CO(3)* intermediate. A low yield for the abstraction reaction was seen in both the experiment and the theory. Experimentally, a fast and weak (16)O(18)O product signal from an abstraction reaction was observed, which could only be detected in the forward direction. A small number of QCT trajectories leading to abstraction were observed to occur primarily via a transient CO(3) intermediate, albeit only at high collision energies (149 kcal mol(-1)). The oxygen isotope exchange mechanism for CO(2) in collisions with ground state O atoms is a newly discovered pathway through which oxygen isotopes may be cycled in the upper atmosphere, where O((3)P) atoms with hyperthermal translational energies can be generated by photodissociation of O(3) and O(2).

The thermochemistry of linear and branched alkanes with up to eight carbons has been reexamined by means of W4, W3.2lite and W1h theories. "Quasi-W4" atomization energies have been obtained via isodesmic and hypohomodesmotic reactions. Our best atomization energies at 0 K (in kcal/mol) are 1220.04 for n-butane, 1497.01 for n-pentane, 1774.15 for n-hexane, 2051.17 for n-heptane, 2328.30 for n-octane, 1221.73 for isobutane, 1498.27 for isopentane, 1501.01 for neopentane, 1775.22 for isohexane, 1774.61 for 3-methylpentane, 1775.67 for diisopropyl, 1777.27 for neohexane, 2052.43 for isoheptane, 2054.41 for neoheptane, 2330.67 for isooctane, and 2330.81 for hexamethylethane. Our best estimates for DeltaH(f,298K)(o) are -30.00 for n-butane, -34.84 for n-pentane, -39.84 for n-hexane, -44.74 for n-heptane, -49.71 for n-octane, -32.01 for isobutane, -36.49 for isopentane, -39.69 for neopentane, -41.42 for isohexane, -40.72 for 3-methylpentane, -42.08 for diisopropyl, -43.77 for neohexane, -46.43 for isoheptane, -48.84 for neoheptane, -53.29 for isooctane, and -53.68 for hexamethylethane. These are in excellent agreement (typically better than 1 kJ/mol) with the experimental heats of formation at 298 K obtained from the CCCBDB and/or NIST Chemistry WebBook databases. However, at 0 K, a large discrepancy between theory and experiment (1.1 kcal/mol) is observed for only neopentane. This deviation is mainly due to the erroneous heat content function for neopentane used in calculating the 0 K CCCBDB value. The thermochemistry of these systems, especially that of the larger alkanes, is an extremely difficult test for density functional methods. A posteriori corrections for dispersion are essential. Particularly for the atomization energies, the B2GP-PLYP and B2K-PLYP double hybrids and the PW6B95 hybrid meta-GGA clearly outperform other DFT functionals.

Hydrogen bonds are of utmost importance in both chemistry and biology. As the applicability of density functional theory and ab initio methods extends to ever larger systems and to liquids, an accurate description of such interactions is desirable. However, reference data are often lacking, and ab initio calculations are only possible and done in very small basis sets. Here, we present high level [CCSD(T)] ab initio reference calculations at the basis set limit on a large set of hydrogen-bonded systems and assess the accuracy of second-order perturbation theory (MP2). The possibilities of using basis set extrapolations for geometries and dissociation energies are discussed as well as the results of R12 methods and density functional and local correlation methods.

Benchmark total atomization energies (TAE0 values) were obtained, by means of our recent W4 theory [Karton, A.; Rabinowitz, E.; Martin, J. M. L.; Ruscic, B. J. Chem. Phys. 2006, 125, 144108], for the molecules Be2, BeF2, BeCl2, BH, BF, BH3, BHF2, B2H6, BF3, AlF, AlF3, AlCl3, SiH4, Si2H6, and SiF4. We were then able to deduce "semi-experimental" heats of formation for the elements beryllium, boron, aluminum, and silicon by combining the calculated TAE0 values with experimental heats of formation obtained from reactions that do not involve the species Be(g), B(g), Al(g), and Si(g). The elemental heats of formation are fundamental thermochemical quantities that are required whenever a molecular heat of formation has to be derived from a calculated binding energy. Our recommended DeltaH degrees f,0 [A(g)] values are Be 76.4+/-0.6 kcal/mol, B 135.1+/-0.2 kcal/mol, Al 80.2+/-0.4 kcal/mol, and Si 107.2+/-0.2 kcal/mol. (The corresponding values at 298.15 K are 77.4, 136.3, 80.8, and 108.2 kcal/mol, respectively.) The Be value is identical to the CODATA recommendation (but with half of the uncertainty), while the B, Al, and Si values represent substantial revisions from established earlier reference data. The revised B and Si values are in agreement with earlier semi-ab initio derivations but carry much smaller uncertainties.

We examine the applicability of density functional theory (DFT) to the polarizability of Cn- (n = 3-9) cluster anions. This was achieved by comparing DFT calculations using two different exchange-correlation functionals (the non-empirical local density approximation, LDA, and the semiempirical hybrid functional B97-1) to quantum chemical calculations using the coupled cluster method in the CCSD(T) "gold standard" approximation. We find that, unless the extra electron is not bound at all by DFT, both LDA and B97-1 agree with the CCSD(T) calculation to within 5-10%, allowing for a meaningful qualitative and semiquantitative analysis. Furthermore, the polarizability is found to increase monotonically with chain size, consistent with the trend inferred from electron detachment experiments.

Rozen's epoxidation reagent, CH(3)CN.HOF, and a prototype epoxidation reaction employing it, have been subjected to an extensive ab initio and density functional study. Its anharmonic force field reveals a very strong red shift for the OH stretch and a strong blue shift for the HOF bend, in semiquantitative agreement with experiment. The very strong hydrogen bond (8.20 kcal/mol at the W1 level) not only serves to stabilize the reactant but also considerably lowers the barrier height for epoxidation of ethylene. Moreover, the reaction byproduct HF is found to act autocatalytically. The OH moiety acquires HO(+) character in the transition state. Our W1 benchmark data for the reaction profile allow the performance of various DFT functionals to be assessed. In general, "kinetics" functionals overestimate barrier heights, the BMK functional less so than the others. The B1B95 and TPSS33B95 meta-GGA functionals both perform very well, whereas general-purpose hybrid GGAs underestimate barrier heights. The simple PBE0 functional does reasonably well.

The performance of a wide variety of DFT exchange-correlation functionals for a number of late-transition-metal reaction profiles has been considered. Benchmark ab-initio reference data for the prototype reactions Pd + H2, Pd + CH4, Pd + C2H6 (both C-C and C-H activation), and Pd + CH3Cl are presented, while ab-initio data of lesser quality were obtained for the catalytic hydrogenation of acetone and for the low-oxidation-state and high-oxidation-state mechanisms of the Heck reaction. "Kinetics" functionals such as mPW1K, PWB6K, BB1K, and BMK clearly perform more poorly for late-transition-metal reactions than for main-group reactions, as well as compared to general-purpose functionals. There is no single "best functional" for late-transition-metal reactions, but rather a cluster of several functionals (PBE0, B1B95, PW6B95, and TPSS25B95) that perform about equally well; if main-group thermochemical performance is additionally considered, then B1B95 and PW6B95 emerge as the best performers. TPSS25B95 and TPSS33B95 offer attractive performance compromises if weak interactions and main-group barrier heights, respectively, are also important. In the ab-initio calculations, basis set superposition errors (BSSE) can be greatly reduced by ensuring that the metal spd shell has sufficient radial flexibility in the high-exponent range. Optimal HF percentages in hybrid functionals depend on the class of systems considered, increasing from anions to neutrals to cations to main-group barrier heights; transition-metal barrier heights represent an intermediate situation. The use of meta-GGA correlation functionals appears to be quite beneficial.

A variety of theoretical procedures, including the high-level ab initio methods G3, G3[CC](dir,full), and W2C//ACQ, have been used to predict the structures and heats of formation of several small calcium-containing molecules (CaH, CaH2, CaO, CaOH, Ca(OH)2, CaF, CaF2, CaS, CaCl, and CaCl2). B3-LYP and CCSD(T) with both the (aug-)cc-pWCVQZ and (aug-)cc-pWCVQ+dZ basis sets are found to give molecular geometries that agree well with the experimental results. The CCSD(T)(riv)/(aug-)cc-pWCVQ+dZ results are found to be the most accurate, with a mean absolute deviation from experiment of just 0.008 angstroms. Zero-point vibrational energies (ZPVEs) and thermochemical corrections are found to be relatively insensitive to the level of theory, except in the case of molecules with highly anharmonic calcium-centered bending modes (CaH2, Ca(OH)2, CaF2, CaCl2), where special procedures need to be employed in order to obtain satisfactory results. Several potential improvements to the W2C method were investigated, most of which do not produce significant changes in the heats of formation. It was observed, however, that for CaO and CaS the scalar relativistic corrections are unexpectedly large and highly basis set dependent. In these cases, Douglas-Kroll CCSD(T)/(aug-)cc-pWCV5Z calculations appear to give a converged result. The G3[CC](dir,full) and best W2C-type heats of formation are both found generally to agree well with experimental values recommended in recent critical compendia. However, in some cases (CaO, Ca(OH)2, and CaF2), they differ from one another by more than their predicted error margins. The available experimental data are not sufficiently precise to distinguish definitively between the two sets of results although, in general, when discrepancies exist the W2C heats of formation are lower in energy and tend to be in better agreement with experiment. In the case of CaO, the W2C heat of formation (20.7 kJ mol(-1)) is approximately 20 kJ mol(-1) lower than the G3[CC](dir, full) result and most of the experimental data. Extensive investigation of possible refinements of the W2C method has failed to reveal any weaknesses that could account for this discrepancy. We therefore believe that the heat of formation of CaO is likely to lie closer to the more recent direct experimental determination of 27 kJ mol(-1) than to the value of approximately 40 kJ mol(-1) recommended in recent thermochemical reviews.