ETD

• basi azotate
• composite methods
• geometrie
• geometries
• metodi compositi
• nucleobases
• PCS
• tautomerism
• tautomerismo

Tautomerism is a ubiquitous phenomenon and represents a major challenge in modern chemistry. The relationship between tautomers depends strongly on both intramolecular and environmental stereoelectronic effects. To unravel the complexity of these effects, accurate quantum chemical methods capable of addressing such subtle phenomena are necessary. In this context, preliminary gas-phase studies allow for the disentanglement of intramolecular effects from environmental ones. However, employing standard high-level quantum chemical methods is often unsuitable, as their high computational cost makes the characterization of large biochemical molecules inaccessible.
Over the years, various approaches known as "composite schemes" have been developed to overcome these computational limitations, achieving an optimal balance between accuracy and computational cost. Composite schemes typically operate under the energy additivity approximation, which assumes that the total energy can be separated into independent contributions which can be computed at cheaper levels of theory, depending on the magnitude of their contribution. The Pisa Composite Scheme (PCS) emerges as an accurate and cost-efficient framework capable of predicting reliable energies, geometries, and (ro-)vibrational properties. In the PCS framework, the energy is partitioned into a frozen-core contribution and a core-valence correlation correction, both computed using MP2 and subsequently refined by applying a CCSD(T)-F12 correction with a smaller basis set. PCS methods are identified by an integer following the "PCS" acronym, which indicates the highest angular momentum of the basis set employed for the low-level valence contribution. Furthermore, different variants of the PCS exist and are identified by specific prefixes; for instance, "H-" denotes a hybrid DFT variant which makes use of the B3LYP-D3(BJ) functional, "D-" identifies a double-hybrid DFT variant employing revDSD-PBEP86-D4, and "B-" indicates that empirical core-valence corrections are considered.
In this study, 39 nucleobase tautomers are investigated in the gas phase. To achieve energy precision within the sub-kJ/mol threshold, the PCS3 variant is employed, while PCS2 is used to determine molecular geometries with spectroscopic accuracy. Furthermore, to enable a meaningful comparison between experimental and computational data, the inclusion of vibrational properties is fundamental; in this study harmonic vibrational frequencies are computed at the DPCS3 level (a double-hybrid DFT variant) and anharmonic corrections are obtained via HPCS2 (a hybrid DFT variant). As a cost-effective alternative for molecular geometries, the semi-empirical BDPCS3 method is also evaluated. The study focuses on geometrical parameters and relative enthalpies at 0 K between tautomers.
Regarding the geometrical parameters, the tautomers can be divided into two distinct sets: those with available microwave (MW) spectra and those with no experimental data. For the first set, rotational constants obtained at the PCS2 and BDPCS3 levels, including HPCS2 vibrational corrections (hereafter denoted as PCS2//HPCS2 and BDPCS3//HPCS2, respectively), are compared with MW data. More generally, in the "//" notation, the method on the right is the one used to optimize the molecular geometries, while the method on the left denotes the one used to evaluate vibrational corrections.
Additionally, the PCS2 molecular structures of thymine, uracil, and 2-thiouracil are compared with semi-experimental (SE) structures available in the literature. For the second set, BDPCS3//HPCS2 rotational parameters are compared with the PCS2//HPCS2 results to assess consistency.
For energetics, PCS3 electronic energies are calculated on PCS2 geometries (hereafter denoted PCS3/PCS2), while Zero-Point Energies (ZPE) are computed at the DPCS3 level with HPCS2 anharmonic corrections. These results are evaluated against MP2, DFT, and, where available, experimental or high-level computations (such as HEAT or W1) data.
The PCS2//HPCS2 approach demonstrates spectroscopic accuracy, with a relative error of approximately 0.1% compared to the corresponding experimental ground-state rotational constants. Notably, BDPCS3//HPCS2 achieves similar precision with a dramatic decrease in computational cost. PCS2 geometries agree with SE structures within 2 mÅ for bond lengths and 0.2° for bond angles. For tautomers not yet detected experimentally, BDPCS3//HPCS2 shows results consistent with PCS2//HPCS2. Regarding relative enthalpies, PCS3 results are in excellent agreement with experimental values available for pyridones and purines, as well as with W1 and HEAT references for other tautomers, whereas standard MP2 or DFT methods tend to underestimate or overestimate these gaps. The study also shows that the tautomers lowest in energy are consistently those detected experimentally.
In conclusion, the PCS protocol, with the inclusion of vibrational corrections, provides benchmark-quality results at an affordable computational cost. This protocol can be extended to larger, biologically significant systems using more efficient methods such as BDPCS3 or FPCS3 (which includes the Frozen Natural Orbital approximation). Furthermore, the PCS protocol appears to offer accurate results when integrated with the ONIOM method for the study of complex environments.