The CTOCD-DZ approach is competitive with other methods when the dimension of the basis set is not too small, both for magnetizabilities and shieldings [60]. However, for small basis sets the results can be very unreliable. A good basis set for CTOCD-DZ nuclear magnetic shielding calculations, both at the SCF and the correlated level, is the aug-cc-pCVTZ-CTOCD-uc basis set (Ref. [60]), derived from the aug-cc-pVTZ basis set. This basis set is included in the DALTON basis set library.
In SCF calculations, the convergence toward the HF limit is slower than when employing London Orbitals since in the CTOCD-DZ expressions for the magnetic properties (nuclear magnetic shieldings and magnetizabilities) the diamagnetic terms also depend on the first order perturbed density matrix.
On the other hand, the CTOCD scheme does not only fulfill the requirement of translational invariance of the calculated magnetic properties but also guarantees current-charge conservation which is not the case for methods using London Orbitals. In the case of molecules with a vanishing electric dipole moment, CTOCD magnetic susceptibilities are origin independent and the continuity equation is automatically satisfied.
At the present it is possible to obtain CTOCD-DZ magnetizabilities and nuclear magnetic shieldings with SCF, MCSCF, MP2 (using SOPPA) and CCSD (using SOPPA(CCSD)) wave functions via the ABACUS program. Is also possible to compute these properties using the RESPONSE program and using various CC wave functions. Finally is possible to use the quadratic response functions to compute hypermagnetizabilities and shieldings polarizabilities for SCF, MCSCF and CC wave functions.
Using **RESPONSE, one has to be sure that the calculations of the diamagnetic and paramagnetic contributions are both carried out with the gauge origin set at the same positions, since only the full property and not the diamagnetic or paramagnetic contributions are gauge origin independent.
For calculations of nuclear magnetic shieldings (and shielding polarizabilities) using symmetry, the gauge origin has to be placed at the center of mass, otherwise DALTON will give wrong results! This is the default choice of gauge origin. Nuclear magnetic shielding calculations with the gauge origin set at the respective atoms can only be carried without symmetry and by setting the gauge origin on that atom in the **INTEGRAL section.