While conventional coupled-cluster and equation-of-motion methods allow one to
tackle electronic structure ranging from well-behaved closed-shell molecules to
various open-shell and electronically excited species,
729
Annu. Rev. Phys. Chem.
(2008),
59,
pp. 433.
Link
meta-stable electronic states, called resonances, present a difficult case
for theory. By using complex scaling, complex basis functions (CBFs),
and complex absorbing potential (CAP)
techniques, Q-Chem extends these powerful methods to describe auto-ionizing
states, such as transient anions, highly excited electronic states, and
core-ionized species
156
J. Chem. Phys.
(2013),
138,
pp. 124106.
Link
,
624
J. Phys. Chem. Lett.
(2014),
5,
pp. 310.
Link
,
623
Annu. Rev. Phys. Chem.
(2017),
68,
pp. 525.
Link
(CBF extensions of CC and EOM-CC calculations
are described in Sections 4.9.5 and 8.7). In addition,
users can employ stabilization techniques using charged sphere and scaled
atomic charges options.
741
J. Chem. Phys.
(2012),
136,
pp. 244109.
Link
These methods are only available within
CCMAN2. The complex CC/EOM code is engaged by COMPLEX_CCMAN;
the specific parameters should be specified in the $complex_ccman section.
COMPLEX_CCMAN
COMPLEX_CCMAN
Requests complex-scaled or CAP-augmented CC/EOM calculations.
TYPE:
LOGICAL
DEFAULT:
FALSE
OPTIONS:
TRUE
Engage complex CC/EOM code.
RECOMMENDATION:
Not available in CCMAN. Need to specify CAP strength or complex-scaling parameter
in $complex_ccman section.
The $complex_ccman section is used to specify the details of the complex-scaled/CAP calculations, as illustrated below. If user specifies CS_THETA, complex scaling calculation is performed.
$complex_ccman CS_THETA 10 Complex-scaling parameter theta=0.01, r->r exp(-i*theta) CS_ALPHA 10 Real part of the scaling parameter alpha=0.01, ! r->alpha r exp(-itheta) $end
Alternatively, for CAP calculations, the CAP parameters need to be specified.
$complex_ccman CAP_ETA 1000 CAP strength in 10-5 a.u. (0.01) CAP_X 2760 CAP onset along X in 10^-3 bohr (2.76 bohr) CAP_Y 2760 CAP onset along Y in 10^-3 bohr (2.76 bohr) CAP_Z 4880 CAP onset along Z in 10^-3 bohr (4.88 bohr) CAP_TYPE 1 Use cuboid cap (CAP_TYPE=0/2 will use spherical/Voronoi CAP) $end
One can also add real absorbing potential by using CAP_RE_ETA; it follows the same format as CAP_ETA. For example, this setup would add purely real absorbing potential with =0.01:
$complex_ccman CAP_ETA 0000 CAP strength in 10-5 a.u. (0.00) CAP_RE_ETA 1000 real CAP strength in 10-5 a.u. (0.01) CAP_X 2760 CAP onset along X in 10^-3 bohr (2.76 bohr) CAP_Y 2760 CAP onset along Y in 10^-3 bohr (2.76 bohr) CAP_Z 4880 CAP onset along Z in 10^-3 bohr (4.88 bohr) CAP_TYPE 1 Use cuboid cap (CAP_TYPE=0/2 will use spherical/Voronoi CAP) $end
The CAP_TYPE field specifies the type of the CAP. The current
options are: spherical CAP (CAP_TYPE = 0), cuboid CAP
(CAP_TYPE = 1), and smooth Voronoi
1283
J. Chem. Theory Comput.
(2015),
11,
pp. 4627.
Link
CAP
(CAP_TYPE = 2). In the calculations with a Voronoi CAP, the onset is
specified by the CAP_X variable.
CS_THETA is specified in radian 10. CS_ALPHA, CAP_X/Y/Z are specified in a.u. 10, i.e., CS_THETA = 10 means ; CAP_ETA is specified in units of . The CAP is calculated by numerical integration and the default grid is (). For testing the accuracy of numerical integration, the numerical overlap matrix is calculated and compared to the analytical one. If the performance of the default grid is poor, the grid type can be changed using the keyword XC_GRID (see Section 5.5 for further details). When CAP calculations are performed, CC_EOM_PROP = 1 by default; this is necessary for calculating first-order perturbative correction.
EOM-CC with CBFs (see Section 4.9.5) is deployed by setting COMPLEX_CCMAN = TRUE and COMPLEX_EXPONENTS = TRUE. As with mean-field calculations the complex basis must be specified as described in Section 8.7.
Advanced users may find the following options useful. Several ways of
conducing complex calculations are possible, i.e., complex scaling/CAPs can be
either engaged at all levels (HF, CCSD, EOM), or not. When applied at post
Hartree-Fock level, CAP can either be added to all blocks of the Fock matrix or
restricted to the virtual-virtual block only. The latter approach, known as
projected CAP,
1197
J. Chem. Phys.
(2002),
117,
pp. 5511.
Link
improves the stability of the calculation
results with respect to CAP onset by reducing a CAP-induced perturbation on the
target states through the occupied orbital space.
This type of CAP projection is currently implemented only for EE/EA calculations and is invoked by setting PROJ_CAP keyword in the $complex_ccman section as follows. PROJ_CAP = 1 deploys CAP/EOM-CCSD with projected CAP added at the CCSD and EOM steps and PROJ_CAP = 2 deploys CAP/EOM-CCSD/MP2/MP2T with projected CAP added at the EOM step. The latter implies that -amplitudes (Sec. 7.9.3) are obtained from a real-valued calculation (for zero CAP strength) and can be reused to generate complex eigenvalue trajectories by specifying ETA_STEP and NSTEPS parameters in $complex_ccman.
PROJ_CAP = 3 deploys another form of CAP projection
1284
Int. J. Quantum Chem.
(2001),
82,
pp. 218.
Link
in which the CAP Hamiltonian
is projected onto the subspace spanned by a set of pre-computed EOM eigenvectors. By default,
the excited-state eigenvectors are obtained from a single real-valued calculation, and the CAP matrix
represented in the state basis is printed in the output for each irreducible representation.
This functionality is available for all EOM-CC models for which transition properties
between EOM target states are available. To generate eigenvalue trajectories, CAP_ETA
should be set to a non-zero CAP strength, and subsequent points are specified using
the ETA_STEP and NSTEPS parameters in $complex_ccman. Trajectories
are written to a separate output file for each irreducible representation. Additionally, first-order
perturbative corrections can be obtained by setting PROJ_PROP = 1. Note that
when PROJ_PROP = 1, the initial set of real eigenvectors are obtained using the
complex valued code at zero CAP strength. As such, first-order perturbative corrections
are only available for complex EOM-CC models. The complete set of one-particle state and transition OPDMs between each pair of states (which is all that is required for CAP projection) will be exported to checkpoint file for further analysis when GUI = 2 and PROJ_CAP = 3.
By default, if COMPLEX_CCMAN is specified, the EOM calculations are conducted using complex code. Other parameters are set up as follows:
$complex_ccman CS_HF = true CS_CCSD = true $end
Alternatively, the user can disable complex HF. These options are experimental and should only be used by advanced users. For CBF-EOM-CC, only CS_HF = TRUE and CS_CCSD = TRUE is implemented.
Non-iterative triples corrections are available for all complex scaled and CAP-augmented CC/EOM-CC models and requested in analogy to regular CC/EOM-CC (see Section 7.9.27 for details).
To take account of the impact of the environment on electronic resonances CAP-EOM-CCSD and CBF-EOM-CCSD can be combined with projection-based embedding,
1043
Mol. Phys.
(2021),
119,
pp. e1943029.
Link
as described in Sections 7.9.3.1 and 11.6.
Molecular properties and transition moments can be requested for complex scaled or
CAP-augmented CC/EOM-CC calculations in analogy to regular CC/EOM-CC (see Section 7.9.21 for details). Natural orbitals and
natural transition orbitals can be computed and the exciton wave-functions can
be analyzed, similarly to real-valued EOM-CCSD (same keywords are used to
invoke the analysis). Analytic gradients are available for complex CC/EOM-CC only for cuboid CAPs (CAP_TYPE = 1) introduced at the HF level
(CS_HF = TRUE), as described in Ref.
108
J. Chem. Phys.
(2017),
146,
pp. 031101.
Link
. The
frozen-core approximation is disabled for CAP-CC/EOM-CC gradient
calculations. Geometry optimization can be requested in the same way as in
regular CC/EOM-CC (see Section 7.9.21 for details).
To reduce the memory requirements, the RI-approximation can be invoked to decompose
the electron-repulsion integrals. The code has the same structure as for real-valued
RI-EOM-CCSD
366
J. Chem. Phys.
(2013),
139,
pp. 134105.
Link
but uses a complex-valued RI
decomposition.
1397
J. Chem. Phys.
(2019),
151,
pp. 111101.
Link
The RI-approximation is available for CCSD
and EOM-IP/EA/EE/SF-CCSD in combination with complex-scaling, complex basis functions
and complex absorbing potentials.
185
J. Phys. Chem. A
(2025),
129,
pp. 4797.
Link
It can be invoked by setting AUX_BASIS to
your favorite auxiliary basis set (e.g. RIMP2-aug-cc-pVTZ). For calculations
with complex basis functions, COMPLEX_AUX_BASIS needs to specified as well,
which is usually the same as AUX_BASIS. When using user-defined auxiliary
basis sets, AUX_BASIS is specified in $aux_basis while
COMPLEX_AUX_BASIS is specified in $aux_zbasis. Complex-valued RI integrals
are currently only available for energies and unrelaxed one-electron properties. It
is also possible to use the RI-approximation for the determination of partial Auger
decay widths by decomposing the CCSD energy or by the ACP method (see sections
7.9.11 and 7.9.12).
The RI-EOM-CC2 method is also available for the description of metastable states through electron attachment (EA) in combination with complex absorbing potentials (CAPs) and complex basis functions (CBFs). Two implementations are available in the CCMAN2 and LIBGMBPT modules, which differ in their algorithmic efficiency and supported features.
The CCMAN2 implementation follows a code structure analogous to that of RI-EOM-CCSD, and can be invoked by specifying method = CCMAN2 in the input. However, this implementation does not employ the most efficient algorithm, as the doubles amplitudes are not evaluated on-the-fly. Consequently, the memory requirements remain at , same as EOM-CCSD.
The LIBGMBPT implementation is based on a more efficient algorithm, reducing the memory requirements to , making it the preferred choice for larger systems. Spin-scaled variants of EA-CC2, including spin-component-scaling (SCS) and scaled opposite-spin (SOS), are also available exclusively in this module and can be invoked in the same manner as in MP2 (see section 6.6.5).
For this implementation, the mixed Davidson/DIIS solver is not employed as in the real-valued EOM-CC2 (see section 7.9.4). Instead, a modified Davidson algorithm is introduced for the solution of the CC2 pseudoeigenvalue equations for electron attachment.
Several input features are worth noting. The user can target attachment to a specific state by specifying the
EOM_USER_GUESS keyword along with the $eom_user_guess section, in the same manner as in EOM-CC (see section 7.9.16). The orbital indexing in LIBGMBPT starts from the first occupied orbital, in contrast to the first unoccupied orbital as in CCMAN2, and the frozen core orbitals must be subtracted from the total orbital count. In addition, sequential calculation of multiple roots with a user-defined guess can be requested by setting $eom_user_guess = 2.
Finally, point group symmetry is not yet supported in this implementation. Further details on the implementation and its performance on temporary anions can be found in Ref.
1382
J. Chem. Phys.
(2025),
163,
pp. 244103.
Link
.
The LIBGMBPT implementation can be invoked by setting CCMAN2 = –1, the default currently is CCMAN2.