Binding energies and bonding nature of MX(CO)(PH3)2(C60) (M = Rh or Ir; X = H or Cl): Theoretical study

IrH(CO)(PH₃)₂(C₆₀), IrCl(CO)(PH₃)₂(C₆₀), and RhH(CO)(PH₃)₂(C₆₀) were theoretically investigated with DFT and MP2 to MP4(SDQ) methods.  Because the DFT method considerably underestimates the binding energy compared to the MP2 method, their binding energies were evaluated by the ONIOM(MP4(SDQ):UFF) method.  The binding energy decreases in the order IrH(CO)(PH₃)₂(C₆₀) (59.4) > RhH(CO)(PH₃)₂(C₆₀) (48.2) > Pt(PH₃)₂(C₆₀) (47.2) > IrCl(CO)(PH₃)₂(C₆₀) (43.0), where in parentheses are the binding energy (in kcal/mol) calculated with the ONIOM(MP4(SDQ):UFF) method and that of Pt(PH₃)₂(C₆₀) was calculated with the same method and the same basis sets in our previous work.  This decreasing order is interpreted in terms of the d_π orbital energy, the d orbital expansion, the presence of the empty d_σ orbital, and the distortion energy of the metal fragment induced by the complexation; for instance, the d_π orbital is at higher energy and more expands in IrH(CO)(PH₃)₂ than in the Rh analogue, which leads to the larger binding energy of IrH(CO)(PH₃)₂(C₆₀) than that of the Rh analogue. IrCl(CO)(PH₃)₂ is less favorable than IrH(CO)(PH₃)₂ because of the lower energy of d_π orbital.  Although the π-back donation is stronger in IrCl(CO)(PH₃)₂(C₆₀) than in RhH(CO)(PH₃)₂(C₆₀), the binding energy of IrCl(CO)(PH₃)₂(C₆₀) is smaller than that of RhH(CO)(PH₃)₂(C₆₀) due to the larger distortion energy of the IrCl-(CO)(PH₃)₂ moiety.  Although the d_π orbital of Pt(PH₃)₂ is at higher energy than that of IrH-(CO)(PH₃)₂, the binding energy of IrH(CO)(PH₃)₂(C₆₀) is larger than that of Pt(PH₃)₂(C₆₀) because the distortion energy is large and the d_σ orbital is doubly occupied in Pt(PH₃)₂(C₆₀).  It is also noted that these binding energies are much larger than those of the ethylene analogues like those of the Pt(0) complexes, which is reasonably interpreted in terms that the LUMO of C₆₀ is at much lower energy than those of ethylene.     

Cation-π interactions of curved polycyclic systems: M (M=Li and Na) ion complexation with buckybowls

B3LYP/6-311+G** calculations on alkali metal ion (Li⁺ and Na⁺) complexation with corannulene and sumanene indicate stronger binding compared to [5]-radialene or benzene. The dependence of binding to the convex and concave site is marginal, albeit the preference was consistent for convex binding in the range of 1-4 kcal/mol. The bowl-to-bowl inversion barriers are only marginally affected, below 2 kcal/mol, by metal ion complexation.     

Adsorption of NH(3) and H(2)O in acidic chabazite. Comparison of ONIOM approach with periodic calculations

The adsorption of NH(3) and H(2)O in acidic chabazite has been studied with the B3LYP method within the cluster approach (5T, 48T clusters) and the periodic approach adopting a Si/Al = 11/1 chabazite and a basis set of polarized double-zeta quality. The 5T cluster has been treated fully ab initio at the B3LYP level whereas the 48T cluster has been treated with the ONIOM2 scheme using B3LYP as the high level of theory and the MNDO, AM1, and HF/3-21G methods as low levels of theory. Periodic calculations show that the adsorption of NH(3) in acidic chabazite takes place through an ion pair (NH(4)(+)-CHA(-)) structure, the computed adsorption energy being -32 kcal/mol. The adsorption of H(2)O leads to a hydrogen bonded (H(2)O-HCHA) complex with the computed adsorption energy of -20 kcal/mol. All ONIOM combinations provide similar structures to those obtained with periodic calculations. Adsorption energies, however, are sensitive to the low level used, especially for NH(3). The ONIOM B3LYP:HF/3-21G method is the one that provides more satisfactory results. Present results show that, for larger zeolites, the ONIOM scheme can be successfully applied while drastically reducing the cost of a fully ab initio treatment.     

Metallophilic attractions between d8-d10 heterometallic compounds trans-[Pt(PH3)2(CN)2] and M(PH3)2+ (M=Ag or Cu): Ab initio study

The weak metal-metal interactions of Pt(II)-Ag(I)/Cu(I) have been investigated by ab initio method at MP2 level through the model complexes [trans-Pt(PH3)2(CN)2-M(PH3)2+] (M=Ag,Cu). The calculated interaction energy of 12.9 and 11.5 kcal mol(-1) for [trans-Pt(PH3)2(CN)2-Ag(PH3)2+] and [trans-Pt(PH3)2(CN)2-Cu(PH3)2+] respectively, are in the middle of the van der Waals force and the strong hydrogen bond. The estimated equilibrium separations between Pt and M, r(eq)(Pt-M) (3.32 A for M=Ag and 3.23 A for M=Cu), lie within the region expected for weak metal-metal interaction. The electronic dispersive contributions dominate the weak interaction.     

Ab initio and DFT benchmark study for nucleophilic substitution at carbon (SN2@C) and silicon (SN2@Si)

To obtain a set of consistent benchmark potential energy surfaces (PES) for the two archetypal nucleophilic substitution reactions of the chloride anion at carbon in chloromethane (S(N)2@C) and at silicon in chlorosilane (S(N)2@Si), we have explored these PESes using a hierarchical series of ab initio methods [HF, MP2, MP4SDQ, CCSD, CCSD(T)] in combination with a hierarchical series of six Gaussian-type basis sets, up to g polarization. Relative energies of stationary points are converged to within 0.01 to 0.56 kcal/mol as a function of the basis-set size. Our best estimate, at CCSD(T)/aug-cc-pVQZ, for the relative energies of the [Cl(-), CH(3)Cl] reactant complex, the [Cl-CH(3)-Cl](-) transition state and the stable [Cl-SiH(3)-Cl](-) transition complex is -10.42, +2.52, and -27.10 kcal/mol, respectively. Furthermore, we have investigated the performance for these reactions of four popular density functionals, namely, BP86, BLYP, B3LYP, and OLYP, in combination with a large doubly polarized Slater-type basis set of triple-zeta quality (TZ2P). Best overall agreement with our CCSD(T)/aug-cc-pVQZ benchmark is obtained with OLYP and B3LYP. However, OLYP performs better for the S(N)2@C overall and central barriers, which it underestimates by 2.65 and 4.05 kcal/mol, respectively. The other DFT approaches underestimate these barriers by some 4.8 (B3LYP) to 9.0 kcal/mol (BLYP).     

Computational investigation of adsorption of molecular hydrogen on lithium-doped corannulene

Density functional theory and classical molecular dynamics simulations are used to investigate the prospect of lithium-doped corannulene as adsorbent material for H(2) gas. Potential energy surface scans at the level of B3LYP/6-311G(d,p) show an enhanced interaction of molecular hydrogen with lithium-atom-doped corannulene complexes with respect to that found in undoped corannulene. MP2(FC)/6-31G(d,p) optimizations of 4H(2)-(Li(2)-C(20)H(10)) yield H(2) binding energies of -1.48 kcal/mol for the H(2)-Li interaction and -0.92 kcal/mol for the H(2)-C interaction, whereas values of -0.94 and -0.83 kcal/mol were reported (J. Phys. Chem. B 2006, 110, 7688-7694) for physisorption of H(2) on the concave and the convex side of corannulene using MP2(full)/6-31G(d), respectively. Classical molecular dynamics simulations predict hydrogen uptakes in Li-doped corannulene assemblies that are significantly enhanced with respect to that found in undoped molecules, and the hydrogen uptake ability is dependent on the concentration of lithium dopant. For the Li(6)-C(20)H(10) complex, a hydrogen uptake of 4.58 wt % at 300 K and 230 bar is obtained when the adsorbent molecules are arranged in stack configurations separated by 6.5 A, and with interlayer distances of 10 A, hydrogen uptake reaches 6.5 wt % at 300 K and 215 bar.     

Theoretical study of Pt(PR3)(2)(AlCl3) (R = H, Me, Ph, or Cy) including an unsupported bond between transition metal and non-transition metal elements: geometry, bond strength, and prediction

The molecular structure and the binding energy of Pt(PR(3))(2)(AlCl(3)) (R = H, Me, Ph, or Cy) were investigated by DFT, MP2 to MP4(SDTQ), and CCSD(T) methods. The optimized structure of Pt(PCy(3))(2)(AlCl(3)) (Cy = cyclohexyl) by the DFT method with M06-2X and LC-BLYP functionals agrees well with the experimental one. The MP4(SDTQ) and CCSD(T) methods present similar binding energies (BE) of Pt(PH(3))(2)(AlCl(3)), indicating that these methods provide reliable BE value. The DFT(M06-2X)-calculated BE value is close to the MP4(SDTQ) and CCSD(T)-calculated values, while the other functionals present BE values considerably different from the MP4(SDTQ) and CCSD(T)-calculated values. All computational methods employed here indicate that the BE values of Pt(PMe(3))(2)(AlCl(3)) and Pt(PPh(3))(2)(AlCl(3)) are considerably larger than those of the ethylene analogues. The coordinate bond of AlCl(3) with Pt(PR(3))(2) is characterized to be the σ charge transfer (CT) from Pt to AlCl(3). This complex has a T-shaped structure unlike the well-known Y-shaped structure of Pt(PMe(3))(2)(C(2)H(4)), although both are three-coordinate Pt(0) complex. This T-shaped structure results from important participation of the Pt d(σ) orbital in the σ-CT; because the Pt d(σ) orbital energy becomes lower as the P-Pt-P angle decreases, the T-shaped structure is more favorable for the σ-CT than is the Y-shaped structure. [Co(alcn)(2)(AlCl(3))](-) (alcn = acetylacetoneiminate) is theoretically predicted here as a good candidate for the metal complex, which has an unsupported M-Al bond because its binding energy is calculated to be much larger than that of Pt(PCy(3))(2)(AlCl(3)).     

Dissecting the concave–convex π‐π interaction in corannulene and sumanene dimers: SAPT(DFT) analysis and performance of DFT dispersion‐corrected methods

The characteristics of the concave–convex π‐π interactions are evaluated in 32 buckybowl dimers formed by corannulene, sumanene, and two substituted sumanenes (with S and CO groups), using symmetry‐adapted perturbation theory [SAPT(DFT)] and density functional theory (DFT). According to our results, the main stabilizing contribution is dispersion, followed by electrostatics. Regarding the ability of DFT methods to reproduce the results obtained with the most expensive and rigorous methods, TPSS‐D seems to be the best option overall, although its results slightly tend to underestimate the interaction energies and to overestimate the equilibrium distances. The other two tested DFT‐D methods, B97‐D2 and B3LYP‐D, supply rather reasonable results as well. M06‐2X, although it is a good option from a geometrical point of view, leads to too weak interactions, with differences with respect to the reference values amounting to about 4 kcal/mol (25% of the total interaction energy). © 2017 Wiley Periodicals, Inc.     

The computation of the potential energy surface for H2 + OH → H2O + H using ab initio and density functional theory methods

The reaction energy profile for H₂ + OH → H + H₂O was computed using HF, MP2, MP4, QCISD, G1, G2, and G2MP2 ab initio methods. In addition, the B3LYP, B3P86, B3PW91, BLYP, BP291, and SVWN density functional theory (DFT) methods were also used. All the ab initio methods, with the exception of the G series, produced much higher activation barriers and heats of reaction than the experimental values. On the other hand, the DFT methods produced negative forward and reverse barriers which were too low, with the exception of the hybrid DFT methods. The G2 ab initio method generated energies which deviated from the experimental values by ∼ 1 kcal/mol and therefore should be considered a very accurate computational method. The hybrid DFT methods produced positive forward reaction barriers with energies that were 2–4 kcal/mol lower than the experimental values. The geometries of the transition state and energies computed by the ab initio and DFT methods were compared. These results suggest that, in the hybrid exchange functional, the portion of the Slater exchange term should be increased. This may be the reason why the computed energies were too low. © 1997 John Wiley & Sons, Inc. Int J Quant Chem 62: 639–644, 1997     

Ab initio and density function theory computational studies of the CH4 + H → CH3 + H2 reaction

The activation barrier for the CH₄ + H → CH₃ + H₂ reaction was evaluated with traditional ab initio and Density Functional Theory (DFT) methods. None of the applied ab initio and DFT methods was able to reproduce the experimental activation barrier of 11.0-12.0 kcal/mol. All ab initio methods (HF, MP2, MP3, MP4, QCISD, QCISD(T), G1, G2, and G2MP2) overestimated the activation energy. The best results were obtained with the G2 and G2MP2 ab initio computational approaches. The zero-point corrected energy was 14.4 kcal mol⁻¹. Some of the exchange DFT methods (HFB) computed energies which were similar to the highly accurate ab initio methods, while the B3LYP hybrid DFT methods underestimated the activation barrier by 3 kcal mol⁻¹. Gradient-corrected DFT methods underestimated the barrier even more. The gradient-corrected DFT method that incorporated the PW91 correlational functional even generated a negative reaction barrier. The suitability of some computational methods for accurately predicting the potential energy surface for this hydrogen radical abstraction reaction was discussed.     

Hetero‐π‐systems from 2 + 2 cycloreversion,part 2.1Ab initio thermochemical study of heterocyclobutanes 2 + 2 cycloreversion to form heteroethenes H2C=X (X=NH,O, SiH2, PH,S)

Ab initio and DFT thermochemical study of diradical mechanism of 2 + 2 cycloreversion of parent heterocyclobutanes and 1,3‐diheterocyclobutanes, cyclo‐(CH₂CH₂CH₂X), and cyclo‐(CH₂XCH₂X), where X = NH, O, SiH₂, PH, S, was undertaken by calculating closed‐shell singlet molecules at three levels of theory: MP4/6‐311G(d)//MP2/6‐31G(d)+ZPE, MP4/6‐311G(d,p)//MP2/6‐31G (d,p)+ZPE, and B3LYP/6‐311+G(d,p)+ZPE. The enthalpies of 2 + 2 cycloreversion decrease on going from group 14 to group 16 elements, being substantially higher for the second row elements. Normally endothermic 2 + 2 cycloreversion is predicted to be exothermic for 1,3‐diazetidine and 1,3‐dioxtane. Strain energies of the four‐membered rings were calculated via the appropriate homodesmic reactions. The enthalpies of ring opening via the every possible one‐bond homolysis that results in the formation of the corresponding 1,4‐diradical were found by subtracting the strain energies from the central bond dissociation energies of the heterobutanes CH₃CH₂—CH₂XH, CH₃CH₂—XCH₃, and HXCH₂—XCH₃. The latter energies were determined via the enthalpies of the appropriate dehydrocondensation reactions, using C—H and X—H bond energies in CH₃XH calculated at G2 level of theory. Except 1,3‐disiletane, in which ring‐opening enthalpy attains 69.7 kcal/mol, the enthalpies of the most economical ring openings do not exceed 60.7 kcal/mol. The 1,4‐diradical decomposition enthalpies found as differences between 2 + 2 cycloreversion and ring‐opening enthalpies were negative, the least exothermicity was calculated for ⋅ CH₂SiH₂CH₂CH₂. The only exception was 1,3‐disiletane, which being diradical, CH₂SiH₂CH₂SiH₂, decomposed endothermically. Since decomposition of the diradical containing two silicon atoms required extra energy, raising the enthalpy of the overall reaction to 78.9 kcal/mol, 1,3‐disiletane was predicted to be highly resisting to 2 + 2 cycloreversion. © 2007 Wiley Periodicals, Inc. Heteroatom Chem 18:704–720, 2007; Published online in Wiley InterScience ( www.interscience.wiley.com ). DOI 10.1002/hc.20377     

Electronic structure and chain-length effects in diplatinum polyynediyl complexes trans,trans-[(X)(R3P)2Pt(C triple bond C)(n)Pt(PR3)2(X)]: a computational investigation

Structure and bonding in the title complexes are studied using model compounds trans,trans-[(C6H5)(H3P)2Pt(C triple bond C)(n)Pt(PH3)2(C6H5)] (PtCxPt; x = 2n = 4-26) at the B3LYP/LACVP* level of density functional theory. Conformations in which the platinum square planes are parallel are very slightly more stable than those in which they are perpendicular (DeltaE = 0.12 kcal mol(-1) for PtC8Pt). As the carbon-chain length increases, progressively longer C triple bond C triple bonds and shorter triple bond C-C triple bond single bonds are found. Whereas the triple bonds in HCxH become longer (and the single bonds shorter) as the interior of the chain is approached, the PtC triple bond C triple bonds in PtCxPt are longer than the neighboring triple bond. Also, the Pt-C bonds are shorter at longer chain lengths, but not the H-C bonds. Accordingly, natural bond orbital charge distributions show that the platinum atoms become more positively charged, and the carbon chain more negatively charged, as the chain is lengthened. Furthermore, the negative charge is localized at the two terminal C triple bond C atoms, elongating this triple bond. Charge decomposition analyses show no significant d-pi* backbonding. The HOMOs of PtCxPt can be viewed as antibonding combinations of the highest occupied pi orbital of the sp-carbon chain and filled in-plane platinum d orbitals. The platinum character is roughly proportional to the Pt/Cx/Pt composition (e.g., x = 4, 31 %; x = 20, 6 %). The HOMO and LUMO energies monotonically decrease with chain length, the latter somewhat more rapidly so that the HOMO-LUMO gap also decreases. In contrast, the HOMO energies of HCxH increase with chain length; the origin of this dichotomy is analyzed. The electronic spectra of PtC4Pt to PtC10Pt are simulated. These consist of two pi-pi* bands that redshift with increasing chain length and are closely paralleled by real systems. A finite HOMO-LUMO gap is predicted for PtCinfinityPt. The structures of PtCxPt are not strictly linear (average bond angles 179.7 degrees -178.8 degrees ), and the carbon chains give low-frequency fundamental vibrations (x = 4, 146 cm(-1); x = 26, 4 cm(-1)). When the bond angles in PtC12Pt are constrained to 174 degrees in a bow conformation, similar to a crystal structure, the energy increase is only 2 kcal mol(-1). The above conclusions should extrapolate to (C triple bond C)(n) systems with other metal endgroups.     

Chemisorption of (CHx and C2Hy) hydrocarbons on Pt(111) clusters and surfaces from DFT studies

We used the B3LYP flavor of density functional theory (DFT) to study the chemisorption of all CH(x) and C(2)H(y) intermediates on the Pt(111) surface. The surface was modeled with the 35 atom Pt(14.13.8) cluster, which was found to be reliable for describing all adsorption sites. We find that these hydrocarbons all bind covalently (sigma-bonds) to the surface, in agreement with the studies by Kua and Goddard on small Pt clusters. In nearly every case the structure of the adsorbed hydrocarbon achieves a saturated configuration in which each C is almost tetrahedral with the missing H atoms replaced by covalent bonds to the surface Pt atoms. Thus, (Pt(3))CH prefers a mu(3) hollow site (fcc), (Pt(2))CH(2) prefers a mu(2) bridge site, and PtCH(3) prefers mu(1) on-top sites. Vinyl leads to (Pt(2))CH-CH(2)(Pt), which prefers a mu(3) hollow site (fcc). The only exceptions to this model are ethynyl (CCH), which binds as (Pt(2))C=CH(Pt), retaining a CC pi-bond while binding at a mu(3) hollow site (fcc), and HCCH, which binds as (Pt)HC=CH(Pt), retaining a pi bond that coordinates to a third atom of a mu(3) hollow site (fcc) to form an off center structure. These structures are in good agreement with available experimental data. For all species we calculated heats of formation (DeltaH(f)) to be used for considering various reaction pathways on Pt(111). For conditions of low coverage, the most strongly bound CH(x) species is methylidyne (CH, BE = 146.61 kcal/mol), and ethylidyne (CCH(3), BE = 134.83 kcal/mol) among the C(2)H(y) molecules. We find that the net bond energy is nearly proportional to the number of C-Pt bonds (48.80 kcal/mol per bond) with the average bond energy decreasing slightly with the number of C ligands.     

Fluxionality of hydrogen ligands in Fe(H)2(H2)(PEtPh2)3

Extensive computational investigations along with additional quasielastic neutron scattering data were used to obtain a consistent picture of the extensive fluxionality of hydride and dihydrogen ligands in Fe(H)(2)(H(2))(PEtPh(2))(3) over a wide range of temperatures from 1.5 to 320 K. We were able to identify three different regimes in the dynamical processes based on activation energies obtained from line spectral broadening. The rotational tunneling lines (coherent exchange of the two hydrogens of the H(2) ligand) are broadened with increasing temperature by incoherent exchange up to about 80 K at which point they merge into a quasielastic spectrum from 100 K to about 225 K. The effective activation energies for the two regions are 0.14 and 0.1 kcal mol(-1), respectively. A third dynamical process with a higher activation energy of 0.44 kcal mol(-1) dominates above 225 K, which we attribute to a quantum dynamical exchange of dihydrogen and hydride ligands. Our detailed density functional theory (DFT) structural calculations involving the three functionals (B3LYP, TPSS, and wB97XD) provide a good account of the experimental structure and rotational barriers when only the hydrogen ligands are relaxed. Full relaxation of the "gas-phase" molecule, however, appears to occur to a greater degree than what is possible in the crystal structure. The classical dihydrogen-hydride exchange path involves a cis-dihydrogen and tetrahydride structure with energies of 6.49 and 7.38 kcal mol(-1), respectively. Experimental observation of this process with much lower energies would seem to suggest involvement of translational tunneling in addition to the rotational tunneling. Dynamics of this type may be presumed to be important in hydrogen spillover from metal particles, and therefore need to be elucidated in an effort to utilize this phenomenon.     

Structure and bonding of transition metal-boryl compounds. Theoretical study of [(PH3)2(CO)ClOs-BR2] and [(PH3)2(CO)2ClOs-BR2] (BR2 = BH2, BF2, B(OH)2, B(OCH=CHO), Bcat)

Quantum chemical DFT calculations using the B3LYP functionals have been carried out for the electronically unsaturated 16 VE five-coordinate osmium boryl-complexes [(PH3)2(CO)ClOs-BR2] and the 18 VE six-coordinate complexes [(PH3)2(CO)2ClOs-BR2] with BR2 = BH2, BF2, B(OH)2, B(OHC=CHO), and Bcat (cat = catecholate O2C6H4). The bonding situation of the Os-BR2 bond was analyzed with the help of the NBO partitioning scheme. The Os-B bond dissociation energies of the 16 VE complexes are very high, and they do not change very much for the different boryl ligands. The 18 VE complexes have only slightly lower bond energies than the 16 VE species. The Os-B bond in both classes of compounds is provided by a covalent sigma-bond which is polarized toward osmium and by strong charge attraction. Os-->B pi-donation is not important for the Os-B binding interactions, except for the Os-BH2 complexes. The stability of the boryl complexes [Os]-BR2 comes mainly from B<--R pi-donation, which is clearly higher than the Os-->B pi-donation. The intraligand charge distribution of the BR2 group changes little when the Os-B bond is formed, except for BH2. The CO ligand in [(PH3)2(CO)2ClOs-BR2] which is trans to BR2 has a relatively weak bond to the osmium atom.     

Density functional theory calculations for endohedral complexes of non-pi C60H60 cage with small guest molecules

Hybrid density-functional theory (B3LYP) calculations were carried out to determine the structures and energies of endohedral complexes of non-pi C(60)H(60) with H(2), CO, and LiH. It was demonstrated that the endohedral complexes of C(60)H(60) with the above three guest molecules are more stable than the corresponding complexes with C(60). Furthermore, the interaction between C(60)H(60) and the inside H(2) or CO is negligible, but the formation of the LiH-C(60)H(60) complex is exothermic with a stabilization energy of -6.0 kcal/mol. While the bond lengths of H(2) and CO changed a little when placed inside the cages, that of the LiH molecule increased and decreased inside C(60)H(60) and C(60), respectively.     

Development of a 3-body:many-body integrated fragmentation method for weakly bound clusters and application to water clusters (H2O)(n = 3-10, 16, 17)

A 3-body:many-body integrated quantum mechanical (QM) fragmentation method for non-covalent clusters is introduced within the ONIOM formalism. The technique captures all 1-, 2-, and 3-body interactions with a high-level electronic structure method, while a less demanding low-level method is employed to recover 4-body and higher-order interactions. When systematically applied to 40 low-lying (H(2)O)(n) isomers ranging in size from n = 3 to 10, the CCSD(T):MP2 3-body:many-body fragmentation scheme deviates from the full CCSD(T) interaction energy by no more than 0.07 kcal mol(-1) (or <0.01 kcal mol(-1) per water). The errors for this QM:QM method increase only slightly for various low-lying isomers of (H(2)O)(16) and (H(2)O)(17) (always within 0.13 kcal mol(-1) of the recently reported canonical CCSD(T)/aug-cc-pVTZ energies). The 3-body:many-body CCSD(T):MP2 procedure is also very efficient because the CCSD(T) computations only need to be performed on subsets of the cluster containing 1, 2, or 3 monomers, which in the current context means the largest CCSD(T) calculations are for 3 water molecules, regardless of the cluster size.