pbqff: Push-Button Quartic Force Fields

From the Automated Calculation of Potential Energy Surfaces to Accurate Infrared Spectra

Advances in the development of quantum chemical methods and progress in multicore architectures in computer science made the simulation of infrared spectra of isolated molecules competitive with respect to established experimental methods. Although it is mainly the multidimensional potential energy surface that controls the accuracy of these calculations, the subsequent vibrational structure calculations need to be carefully converged in order to yield accurate results. As both aspects need to be considered in a balanced way, we focus on approaches for molecules of up to 12-15 atoms with respect to both parts, which have been automated to some extent so that they can be employed in routine applications. Alternatives to machine learning will be discussed, which appear to be attractive, as long as local regions of the potential energy surface are sufficient. The automatization of these methods is still in its infancy, and the generalization to molecules with large amplitude motions or molecular clusters is far from trivial, but many systems relevant for astrophysical studies are already in reach.

Performance of EOM-CCSD(T)(a)*-Based Quartic Force Fields in Computing Fundamental, Anharmonic Vibrational Frequencies of Molecular Electronically Excited States with Application to the Ã1A″ State of :CCH2 (Vinylidene)

Highly accurate anharmonic vibrational frequencies of electronically excited states are not as easily computed as their ground electronic state counterparts, but recently developed approximate triple excited state methods may be changing that. One emerging excited state method is equation of motion coupled cluster theory at the singles and doubles level with perturbative triples computed via the (a)* formalism, EOMEE-CCSD(T)(a)*. One of the most employed means for the ready computation of vibrational anharmonic frequencies for ground electronic states is second-order vibrational perturbation theory (VPT2), a theory based on quartic force fields (QFFs),fourth-order Taylor series expansions of the potential portion of the internuclear Watson Hamiltonian. The combination of these two is herein benchmarked for its performance for use as a means of computing rovibrational spectra of electronically excited states. Specifically, the EOMEE-CCSD(T)(a)* approach employing a complete basis set extrapolation along with core electron inclusion and relativity (the so-called "CcCR" approach) defining the QFF produces anharmonic fundamental vibrational frequencies within 2.83%, on the average, of reported gas-phase experimentally assigned values for the test set including the A ~ 1 A ″ states of HCF, HCCl, HSiF, HNO, and HPO. However, some states have exceptional accuracy in the fundamentals, most notably for ν2 of A ~ 1 A ″ HCCl in which the CcCR QFF value is within 1.8 cm-1 at 927.9 cm-1 (or 0.2%) of the experiment. Additionally, this approach produces rotational constants to, on the absolute average, within 0.41% of available experimental data, showcasing notable accuracy in the computation of rovibronic spectral data. Furthermore, utilizing a hybrid approach composed of harmonic CcCR force constants along with a set of simple EOMEE-CCSD(T)(a)*/aug-cc-pVQZ QFF cubic and quartic force constants is faster than using pure CcCR and better represents those modes that suffer from numerical instability in the anharmonic portion of the QFF, implying that this so-called "CcCR + QZ" QFF approach may be the best for future applications. Finally, complete, rovibrational spectral data are provided for A ~ 1 A 2 :CCH2, a molecule of potential astrochemical interest, in order to aid in its potential future experimental rovibronic characterization.

Quantum Chemistry and Astrochemistry: A Match Made in the Heavens

Quantum chemistry can uniquely answer astrochemical questions that no other technique can provide. Computations can be parallelized, automated, and left to run continuously providing exceptional molecular throughput that cannot be done through experimentation. Additionally, the granularity of the individual computations that are required of potential energy surfaces, reaction mechanism pathways, or other quantum chemically derived observables produces a unique mosaic that make up the larger whole. These pieces can be dissected for their individual contributions or evaluated in an ad hoc fashion for each of their roles in generating the larger whole. No other scientific approach is capable of reporting such fine-grained insights. Quantum chemistry also works from a bottom-up approach in providing properties directly from the desired molecule instead of a top-down perspective as required of experiment where molecules have to be linked to observed phenomena. Furthermore, modern quantum chemistry is well within the range of "chemical accuracy" and is approaching "spectroscopic accuracy." As such, the seemingly difficult questions asked by astrochemistry that would not be asked initially for any other application require quantum chemical reference data. While the results of quantum chemical computations are needed to interpret astrochemical observation, modeling, or laboratory experimentation, such hard questions, regardless of the original need to answer them, produce unique solutions. While questions in astrochemistry often require novel developments in and implementations of quantum chemistry as outlined herein, the applications of these solutions will stretch beyond astrochemistry and may yet impact fields much closer to Earth.

A Vision for the Future of Astrochemistry in the Interstellar Medium by 2050

By 2050, many, but not nearly all, unattributed astronomical spectral features will be conclusively linked to molecular carriers (as opposed to nearly none today in the visible and IR); amino acids will have been observed remotely beyond our solar system; the largest observatories ever constructed on the surface of the Earth or launched beyond it will be operational; high-throughput computation either from brute force or machine learning will provide unprecedented amounts of reference spectral and chemical reaction data; and the chemical fingerprints of the universe delivered by those of us who call ourselves astrochemists will provide astrophysicists with unprecedented resolution for determining how the stars evolve, planets form, and molecules that lead to life originate. Astrochemistry is a relatively young field, but with the entire universe as its playground, the discipline promises to persist as long as telescopic observations are made that require reference data and complementary chemical modeling. While the recent commissionings of the James Webb Space Telescope and Atacama Large Millimeter Array are ushering in the second "golden age" of astrochemistry (with the first being the radio telescopic boom period of the 1970s), this current period of discovery should facilitate unprecedented advances within the next 25 years. Astrochemistry forces the asking of hard questions beyond the physical conditions of our "pale blue dot", and such questions require creative solutions that are influential beyond astrophysics. By 2050, more creative solutions will have been provided, but even more will be needed to answer the continuing question of our astrochemical ignorance.

DFT + F12 QFFs for Cost-Effective Rovibrational Spectral Data Predictions of Ground and Excited Electronic States

The quest for faster computation of anharmonic vibrational frequencies of both ground and excited electronic states has led to combining coupled cluster theory harmonic force constants with density functional theory cubic and quartic force constants for defining a quartic force field (QFF) utilized in conjunction with vibrational perturbation theory at second order (VPT2). This work shows that explicitly correlated coupled cluster theory at the singles, doubles, and perturbative triples levels [CCSD(T)-F12] provides accurate anharmonic vibrational frequencies and rotational constants when conjoined with any of B3LYP, CAM-B3LYP, BHandHLYP, PBE0, and ωB97XD for roughly one-quarter of the computational time of the CCSD(T)-F12 QFF alone for our test set. As the number of atoms in the molecule increases, however, the anharmonic terms become a greater portion of the QFF, and the cost comparison improves with HOCO+ and formic acid, requiring less than 15 and 10% of the time, respectively. In electronically excited states, PBE0 produces more consistently accurate results. Additionally, as the size of the molecule and, in turn, QFF increase, the cost savings for utilizing such a hybrid approach for both ground- and excited-state computations grows. As such, these methods are promising for predicting accurate rovibrational spectral properties for electronically excited states. In cases where well-behaved potentials for a small selection of targeted excited states are needed, such an approach should reduce the computational cost compared to that of methods requiring semiglobal potential surfaces or variational treatments of the rovibronic Hamiltonian. Such applications include spectral characterization of comets, exoplanets, or any situation in which gas phase molecules are being excited by UV-vis radiation.