Approximate treatment of higher excitations in coupled-cluster theory

The possibilities for the approximate treatment of higher excitations in coupled-cluster (CC) theory are discussed. Potential routes for the generalization of corresponding approximations to lower-level CC methods are analyzed for higher excitations. A general string-based algorithm is presented for the evaluation of the special contractions appearing in the equations specific to those approximate CC models. It is demonstrated that several iterative and noniterative approximations to higher excitations can be efficiently implemented with the aid of our algorithm and that the coding effort is mostly reduced to the generation of the corresponding formulas. The performance of the proposed and implemented methods for total energies is assessed with special regard to quadruple and pentuple excitations. The applicability of our approach is illustrated by benchmark calculations for the butadiene molecule. Our results demonstrate that the proposed algorithm enables us to consider the effect of quadruple excitations for molecular systems consisting of up to 10-12 atoms.     

Analytical gradient evaluation in coupled-cluster theory

First derivatives of the coupled-cluster doubles (CCD) energy have been implemented for the first time. Formulae are presented for the Hartree-Fock reference case. These equations are illustrated by applying CCD gradients to the geometry optimization and harmonic vibrational frequency determination of H₂O and the ¹A₁ and the ¹A₁ state of CH₂.     

Three-electron excitation in open-shell coupled-cluster theory

The three-electron excitation operator T₃ is included approximately in the coupled-cluster method with single and double excitations for open shells (CCSD). The resulting CCSD + T scheme incorporates all energy diagrams up to and including third order in the perturbation. All ten valence-shell energy differences of O and its ions were calculated. They show significant effects of T₃ and are within 0.1 eV of the basis set limit.     

Error analysis and improvements of coupled-cluster theory

Summary The error in the energy of the traditional coupled-cluster (TCC) approach and of several variants is analyzed in terms of the error of the cluster operatorS. A key feature of this analysis is that TCC can be based on an energy functional (asymmetric inS andS ) that is made stationary with respect to variation ofS . The error of TCC scales with the particle numbern, but it is not quadratic in . An improved coupled-cluster method (ICC) is presented that is the next step in a hierarchy from TCC to an exact variational theory. An alternative hierarchy is possible that leads to the extended coupled-cluster (ECC) method of Arponen. Variational (VCC) and unitary (UCC) coupled cluster theories and their stationary conditions and errors are analyzed along similar lines and practicable VCC or UCC approaches are presented. An infinite summation of certain terms in the VCC expectation value is shown to lead to a coupled-pair functional of the type proposed by Ahlrichs. The various CC schemes discussed here are compared on the CC-D, CC-SD and CC-SDT levels and beyond this. Special aspects referring to properties are also discussed.     

Closed-shell coupled-cluster theory with spin-orbit coupling

A two-component closed-shell coupled-cluster (CC) approach using relativistic effective core potentials with spin-orbit coupling included in the post-Hartree-Fock treatment is proposed and implemented at the CC singles and doubles (CCSD) level as well as at the CCSD level augmented by a perturbative treatment of triple excitations [CCSD(T)]. The latter invokes as an additional approximation the neglect of the occupied-occupied and virtual-virtual blocks of the spin-orbit coupling matrix in order to avoid the iterative N(7) steps in the treatment of triple excitations. The computational effort of the implemented two-component CC methods is about 10-15 times that of its corresponding nonrelativistic counterpart, which needs to be compared to the by a factor of 32 higher cost for fully relativistic schemes and schemes with spin-orbit coupling included already at the Hartree-Fock self-consistent field (HF-SCF) level. This substantial computational saving is due to the use of real molecular orbitals and real two-electron integrals. Results on 5p-, 6p-, and 7p-block element compounds show that the bond lengths and harmonic frequencies obtained with the present two-component CCSD method agree well with those computed with the CCSD approach including spin-orbit coupling at the HF-SCF level even for the 7p-block element compounds. As for the CCSD(T) approach, high accuracy for 5p- and 6p-block element compounds is retained. However, the difference in bond lengths and harmonic frequencies becomes somewhat more pronounced for the 7p-block element compounds.     

Local correlation potentials from Brueckner coupled-cluster theory

Local correlation potentials have been obtained from the nonlocal Brueckner coupled-cluster correlation potentials for the rare-gas atoms He, Ne, and Ar and the CO molecule. It is shown that the local correlation potential can mainly be expressed as a sum of two components: a "pure" correlation part and a relaxation contribution. While the total correlation potentials show an oscillating behavior near the nuclei, indicating the atomic shell structure, their components decrease rather monotonously, with a step structure in case of Ne and Ar. By looking at the determinantal overlap and one-electron properties it has been found that the orbitals obtained from these local potentials form a determinant which very well corresponds with the Brueckner determinant. Thus the previously found closeness between the Hartree-Fock determinant and the exchange-only Kohn-Sham determinant [Della Sala and Gorling, J. Chem. Phys. 115, 5718 (2001)] is confirmed also for the correlated case.     

Multireference coupled-cluster theory: the easy way

The multi-ionization equation-of-motion coupled-cluster (CC) method is developed for multireference (MR) problems. It is operationally single reference, depending upon a formal matrix diagonalization step to define the coefficients in the wavefunction in an unbiased way that allows for important MR character. The method is illustrated for the autoisomerization of cyclobutadiene, which has a very large multireference effect and compared to other MR-CC results. The newly implemented methods are also used to obtain the vertical double ionization (DI) potentials of several small molecules (H(2)O, CO, C(2)H(2), C(2)H(4)). Also, the performance of the new methods is analyzed by plotting the potential energy curve for twisted ethylene as a function of a dihedral angle between two methylenes. Evaluation of the total molecular energy via MR-DI-CC calculations makes it possible to avoid an unphysical cusp.     

Computation of expectation values from vibrational coupled-cluster at the two-mode coupling level

In this work we show how the vibrational coupled-cluster method at the two-mode coupling level can be used to calculate zero-point vibrational averages of properties. A technique is presented, where any expectation value can be calculated using a single set of Lagrangian multipliers computed solving iteratively a single linear set of equations. Sample calculations are presented which show that the resulting algorithm scales only with the third power of the number of modes, therefore making large systems accessible. Moreover, we present applications to water, pyrrole, and para-nitroaniline.     

Symmetry exploitation in closed-shell coupled-cluster theory with spin-orbit coupling

In the present work, we report exploitation of spatial symmetry in calculations of ground state energy and analytic first derivatives of closed-shell molecules based on our previously developed coupled-cluster (CC) approach with spin-orbit coupling. Both time-reversal symmetry and spatial symmetry for D(2h) and its subgroups are exploited in the implementation. The symmetry of a certain spin case for the amplitude, intermediate, or density matrix is determined by the symmetry of the corresponding spin functions and the direct product decomposition method is employed in computations involving these quantities. The reduction in computational effort achieved through the use of spatial symmetry is larger than the order of the molecular single point group. Symmetry exploitation renders application of the CC approaches with spin-orbit coupling to larger closed-shell molecules containing heavy elements with high accuracy.     

Anharmonic vibrational spectroscopy and investigation of intramolecular mode couplings in adenine

Vibrational frequencies for the nucleobase adenine are calculated by the vibrational self-consistent field (VSCF) and correlation corrected vibrational self-consistent field (CC-VSCF) methods using Hartree-Fock (HF), density functional theory (DFT) and second order Møller-Plesset (MP2) theories. A large number of potential energy surface (PES) points were computed in the anharmonic calculations corresponding to each method. The quartic force field (QFF) approximation was used to generate the full grid of points for the VSCF solver. We have implemented our new procedure for computing the mode-mode coupling integrals in the 2-mode coupling representations of the quartic force field (2MR-QFF) for prediction of coupling magnitudes. Calculations were performed using the 6-31G(d,p) basis set. Comparison of the calculated ab initio anharmonic spectra with Ar matrix experimental data of adenine reported in the literature reveals that, the CC-VSCF (DFT) wavenumbers show the best agreement. The experimental geometric parameters of adenine are compared with the theoretically optimized molecular structural parameters. These are found to be in good agreement. Vibrational assignments are based on the calculated potential energy distribution (PED) values.     

Calculation of nonadiabatic couplings in density-functional theory

This paper proposes methods for calculating the derivative couplings between adiabatic states in density-functional theory (DFT) and compares them with each other and with multiconfigurational self-consistent field calculations. They are shown to be accurate and, as expected, the costs of their calculation scale more favorably with system size than post-Hartree-Fock calculations. The proposed methods are based on single-particle excitations and the associated Slater transition-state densities to overcome the problem of the unavailability of multielectron states in DFT which precludes a straightforward calculation of the matrix elements of the nuclear gradient operator. An iterative scheme employing linear-response theory was found to offer the best trade-off between accuracy and efficiency. The algorithms presented here have been implemented for doublet-doublet excitations within a plane-wave-basis and pseudopotential framework but are easily generalizable to other excitations and basis sets. Owing to their fundamental importance in cases where the Born-Oppenheimer separation of motions is not valid, these derivative couplings can facilitate, for example, the treatment of nonadiabatic charge transfers, of electron-phonon couplings, and of radiationless electronic transitions in DFT.     

Application of the coupled-cluster theory to atomic frequency-dependent polarizabilities

Summary A time-dependent coupled-cluster approach may be employed to describe dynamic processes of many-electron systems. Atomic properties, such as the frequency-dependent polarizability, can be treated as a response of the system described by the coupled-cluster expansion to an external radiation field. The major difficulty in the realization of such a formalism is to deal with dynamic pair functions. The procedure reported here is to simplify the full set of single- and pair-excitation expansion equations to a subset of equations which includes polarization and relaxation effects to all orders and is solved by using a complete set of discrete basis functions. Calculations of excitation energies and frequency-dependent electric dipole polarizabilities for helium are presented. Application of the procedure to calculate photoionization cross sections is discussed.     

Spin-projected coupled-cluster theory with single and double excitations

Coupled-cluster (CC) theory including single (S) and double (D) excitations and carried out with a spin-unrestricted Hartree–Fock (UHF) reference wave function is free from S + 1 spin contamination as can be confirmed by an analysis of the expectation value of the spin operator, Ŝ ². Contamination by the S + 2 contaminant can be projected out by an approximate procedure (APCCSD) with a projection operator, P^, represented by the product of the spin annihilation operators ? _{s+ 1} and ?_s+2. The computational cost of such a projection scales with O(M ⁶) (M is the number of basis functions). The APCCSD energy obtained after annihilation of the S + 2 contaminant can be improved by adding triple (T) excitations in a perturbative way, thus leading to APCCSD(T) energies. For the 17 examples studied, the deviation of the UHF-CCSD(T) energies from the corresponding full configuaration interaction values is reduced from 4.0 to 2.3 mhartree on the average as a result of annihilating the S + 2 contaminant in an approximate way. In the case of single-bond cleavage, APCSSD leads to a significant improvement of the energy in the region where the bonding electrons recouple from a closed shell to an open shell singlet electron pair. Received: 13 April 2000 / Accepted: 12 July 2000 / Published online: 24 October 2000     

A coupled-cluster study of the structure and vibrational spectra of pyrazole and imidazole

Harmonic force fields were calculated at the corresponding optimized geometries for pyrazole and imidazole at the HF, B3LYP, MP2, CCSD and CCSD(T) levels using the 6-31G* basis set and at the HF and B3LYP levels using the cc-pVTZ basis set. The agreement between the calculated and experimental geometries by the CCSD and CCSD(T) methods was generally similar to that obtained with the B3LYP and MP2 methods. The force fields were scaled using one-scale-factor (1SF), 3SF and 7SF scaling schemes. The scale factors were varied with respect to the experimental frequencies. Using 7SF scaling, the root-mean-square (RMS) deviation of the calculated frequencies from the experimental frequencies by the HF, B3LYP, MP2, CCSD and CCSD(T) methods and the 6-31G* basis set was 16, 7, 13, 11 and 11 cm(-1), respectively. This shows that the B3LYP method is preferred for force field calculations over the perturbative MP2, CCSD and CCSD(T) methods. Using 1SF scaling, the CCSD(T) scale factor was 0.931, the highest among the five methods used but close to that obtained with the B3LYP method and the cc-pVTZ basis set with lower RMS deviation.     

Accurate computational thermochemistry from explicitly correlated coupled-cluster theory

Explicitly correlated coupled-cluster theory has developed into a valuable computational tool for the calculation of electronic energies close to the limit of a complete basis set of atomic orbitals. In particular at the level of coupled-cluster theory with single and double excitations (CCSD), the space of double excitations is quickly extended towards a complete basis when Slater-type geminals are added to the wave function expansion. The purpose of the present article is to demonstrate the accuracy and efficiency that can be obtained in computational thermochemistry by a CCSD model that uses such Slater-type geminals. This model is denoted as CCSD(F12), where the acronym F12 highlights the fact that the Slater-type geminals are functions f(r ₁₂) of the interelectronic distances r ₁₂ in the system. The performance of explicitly correlated CCSD(F12) coupled-cluster theory is demonstrated by computing the atomization energies of 73 molecules (containing H, C, N, O, and F) with an estimated root-mean-square deviation from the values compiled in the Active Thermochemical Tables of σ = 0.10 kJ/mol per valence electron. To reach this accuracy, not only the frozen-core CCSD basis-set limit but also high-order excitations (connected triple and quadruple excitations), core–valence correlation effects, anharmonic vibrational zero-point energies, and scalar and spin–orbit relativistic effects must be taken into account.     

Analytical energy gradients for a unitary coupled-cluster theory

Our recently developed and tested unitary multiconfigurational coupled-cluster electronic wavefunction method is extended to permit, for the first time, the analytical evaluation of energy derivatives. The unitary nature of this method admits a variational energy functional whose stationary nature plays a key role in simplifying our derivation. Explicit expressions are given for the gradient (first energy derivative) for both the full unitary coupled cluster and its coupled electron pair approximation (CEPA).     

Fermi resonance in CO2: a combined electronic coupled-cluster and vibrational configuration-interaction prediction

The authors present a first-principles prediction of the energies of the eight lowest-lying anharmonic vibrational states of CO(2), including the fundamental symmetric stretching mode and the first overtone of the fundamental bending mode, which undergo a strong coupling known as Fermi resonance. They employ coupled-cluster singles, doubles, and (perturbative) triples [CCSD(T) and CCSDT] in conjunction with a range of Gaussian basis sets (up to cc-pV5Z, aug-cc-pVQZ, and aug-cc-pCVTZ) to calculate the potential energy surfaces (PESs) of the molecule, with the errors arising from the finite basis-set sizes eliminated by extrapolation. The resulting vibrational many-body problem is solved by the vibrational self-consistent-field and vibrational configuration-interaction (VCI) methods with the PESs represented by a fourth-order Taylor expansion or by numerical values on a Gauss-Hermite quadrature grid. With the VCI, the best theoretical estimates of the anharmonic energy levels agree excellently with experimental values within 3.5 cm(-1) (the mean absolute deviation). The theoretical (experimental) anharmonic frequencies of the Fermi doublet are 1288.9 (1285.4) and 1389.3 (1388.2) cm(-1).     

Multiple basis sets in calculations of triples corrections in coupled-cluster theory

Multiple basis sets are used in calculations of perturbational corrections for triples replacements in the framework of single-reference coupled-cluster theory. We investigate a computational procedure, where the triples correction is calculated from a reduced space of virtual orbitals, while the full space is employed for the coupled-cluster singles-and-doubles model. The reduced space is either constructed from a prescribed unitary transformation of the virtual orbitals (for example into natural orbitals) with subsequent truncation, or from a reduced set of atomic basis functions. After the selection of a reduced space of virtual orbitals, the singles and doubles amplitudes obtained from a calculation in the full space are projected onto the reduced space, the remaining set of virtual orbitals is brought into canonical form by diagonalizing the representation of the Fock operator in the reduced space, and the triples corrections are evaluated as usual. The case studies include the determination of the spectroscopic constants of N₂, F₂, and CO, the geometry of O₃, the electric dipole moment of CO, the static dipole polarizability of F⁻, and the Ne⋯Ne interatomic potential. Received: 28 December 1996 / Accepted: 8 April 1997     

Towards a spin-adapted coupled-cluster theory for high-spin open-shell states

A spin-adapted coupled-cluster (SA-CC) scheme based on the additional consideration of spin constraints is proposed for the quantum chemical treatment of high-spin open-shell cases. Its computational feasibility is demonstrated via a pilot implementation within the singles and doubles approximation. Test calculations indicate that the suggested SA-CC scheme provides results of similar accuracy as the more traditional schemes without spin adaptation.