Search

ConspectusWe review the Best Theory + Reliable High-Resolution Experiment (BTRHE) strategy for obtaining highly accurate molecular rovibrational line lists with InfraRed (IR) intensities. The need for highly accurate molecular rovibrational line lists is twofold: (a) assignment of the many rovibrational lines for common stable molecules especially those that exhibit a large amplitude motion, such as NH

The azirinyl cation is an aromatic cyclic molecule that is isoelectronic with cyclopropenylidene, c-C3H2, and c-C3H3+. Cyclopropenylidene has been shown to be ubiquitous, existing in many different astrophysical environments. Given the similar chemistry between C and N, and the relative abundances between C and N in astrophysical environments, it is expected that there should be aromatic ringed molecules that incorporate N in the ring, but as yet, no such molecule has been identified. To address this issue, the present study uses high levels of electronic structure theory to compute a highly accurate quartic force field (QFF) for the azirinyl cation and its two lowest lying isomers, the cyanomethyl and isocyanomethyl cations. The theoretical approach uses the singles and doubles coupled-cluster method that includes a perturbative correction for connected triple excitations, CCSD(T), together with extrapolation to the one-particle basis set limit and corrections for scalar relativity and core-correlation. The QFF is then used in a second-order vibrational perturbation theory analysis (VPT2) to compute the fundamental vibrational frequencies and rovibrational spectroscopic constants for all three C2NH2+ isomers. The reliability of the VPT2 vibrational frequencies is tested by comparison to vibrational configuration interaction (VCI) calculations, and excellent agreement is found between the two approaches. Fundamental vibrational frequencies and rovibrational spectroscopic constants for all singly substituted 13C, 15N, and D isotopologues are also reported. It is expected that the highly accurate spectroscopic data reported herein will be useful in the identification of these cations in high-resolution experimental or astronomical observations.

The semi-empirical molecular rovibrational IR line lists, such as ExoMol, TheoReTs, and Ames, combine the experimental accuracy and theoretical power to reach better than 0.1 cm−1 accuracy for line positions and better than 80–90% agreement for line intensities. The quality of these existing semi-empirical IR lists allows further improvements of intensity and line positions for those unobserved minor isotopologues. This paper presents our new BTRHE (Best Theory + Reliable High-resolution Experiment) strategy implementation. For line intensity, the isotopologue consistency and the patterns of mass dependence in the Ames-296 K SO2 and CO2 IR lists are quantitatively presented along the mass-inverse coordinates. The consistency and patterns are better than those in existing experimental data. The methodology proposed here can be used to identify inconsistencies, outliers, and mistakes in intensities, and help improve Effective Dipole Model (EDM) and molecular IR databases. We call for an experimental study on the 50006 and 60007 bands of CO2 628. For line position predictions, a simple approach combining the variational IR line lists with Effective Hamiltonian (EH) model may refine the effective rotational constants A0/B0/C0 and quartic centrifugal distortion constants of minor isotopologues. The prediction accuracy may be improved by two orders of magnitude, i.e. reaching 0–5 MHz prediction accuracy in the range of J

The quality of Ames-296K SO2 Infrared (IR) line list intensities is first validated by quantitative exploration of several dipole moment surfaces (DMSs) and partition sum convergence. The DMSs are computed with several of Dunning's correlation-consistent basis sets and their vibrational dependence are compared to the empirical model derived from Stark effect experiments reported by D. Patel, D. Margolese, and T.R. Dyke [J. Chem. Phys. 70, 2740 (1979)]. The effective dipole deviations from the DMS adopted in the Ames IR lists is 0.2–0.4% for vibrational states up to 3ν3. The vibrational dependence of the dipole moment is also in good agreement, except for nν1. Partition sum convergence at 296 K is confirmed by new calculations with rotational quantum number J up to 150 and upper state E’ up to 8000 cm−1. The isotopologue consistency of the Ames IR line lists is superior relative to the regular Effective Hamiltonian (EH) models and Effective Dipole Moment (EDM) models. The ν1 + ν2 and ν2 + ν3 intensity consistency check reveals the recently reported experimental intensities need significant improvement or re-analysis. After the accuracy, convergence, and isotopologue consistency have been confirmed, the theoretical Ames-296K intensities are combined with the experimental line positions or EH models that experimental spectroscopists published after 2009. Three high-resolution IR line sets are reported for the 32/33/34S16O2, 32S18O2 and 16O32S18O isotopologues: (1) the “New Lines Sets” include experimentally measured line positions; (2) the “Expanded Line Sets” include possible transitions among new rovibrational levels assigned in experiments and ground state (GS) levels predicted by reliable EH models; (3) the “Ames + MARVEL Sets” include possible transitions among all those levels reported in a recent MARVEL analysis. [Tobias et al, JQSRT 208, 152 (2018)]. Compared to the limited data in High-resolution TRANsmission molecular absorption database (HITRAN), these line sets have significantly improved the data coverage up to 4000 cm−1. Some missing bands can be traced to the unpublished experimental data. The isotopologue consistency of these line sets will help identify the uncertainties and defects in the experimental EH and EDM models. These line sets are good candidates for the next HITRAN update, if line shape parameters are available. The line sets can be downloaded from supplementary files or from the Ames Molecular Spectroscopic Database at http://huang.seti.org .

A new 626-based Ames-2 PES refinement and Ames-2016 line lists for 13 CO2 isotopologues are reported. A consistent σRMS = ±0.02 cm−1 is established for hundreds of isotopologue band origins using the Ames-2 PES. Ames-2016 line lists are computed at 296 K, 1000 K and 4000 K using the Ames-2 PES and the same DMS-N2 dipole surface used previously, with J up to 150, E′ up to 24,000 cm−1 or 18,000 cm−1 and appropriate intensity cutoffs. The lists are compared to the CDSD-296, CDSD-4000 databases, UCL line lists, and a few recent highly accurate CO2 intensity measurements. Both agreements and discrepancies are discussed. Compared to the old Ames CO2 lists, the Ames-2016 line lists have line position deviations reduced by 50% or more, which consequently leads to more reliable intensities. The line shape parameters in the Ames-2016 line lists are predicted using the newly assigned conventional vibrational polyad quantum numbers for rovibrational levels below 12,000 cm−1 so the quality of the line shape parameters is similar to that of CDSD or HITRAN. This study further proves that a semi-empirically refined PES (Ames-1 and Ames-2) coupled with a high quality ab initio DMS (DMS-N2 and UCL) may generate IR predictions with consistent accuracy and is thus helpful in the analysis of laboratory spectra and simulations of various isotopologues. The Ames-2016 lists based on DMS-N2 have reached the ∼1% intensity prediction accuracy level for the recent 626 30013-00001 and 20013-00001 bands, but further quantification and improvements require sub-percent or sub-half-percent accurate experimental intensities. The inter-isotopologue consistency of the intensity prediction accuracies should have reached better than 1–3% for regular bands not affected by resonances. Since the Effective Dipole Models (EDM) in CDSD and HITRAN have 1–20% or even larger uncertainties, we show that the Ames lists can provide better alternative IR data for many hard-to-determine isotopologue bands. Comparison at 4000 K suggests that the Ames-4000 K 12C16O2 line list is reliable and consistent within the current cutoffs of J ≤ 150 and E′ ≤ 24,000 cm−1, but intensity contributions involving higher energy levels should not be omitted and future computations need to be converged up to at least 32,000 cm−1 or higher. The remaining issues are discussed regarding the source of energy level discrepancies, intensity underestimations by ∼50% for some weak bands, etc. and also future work.

The simple aromatic hydrocarbon, cyclopropenylidene (c-C3H2), is a known, naturally-occurring molecule. The question remains as to whether its isoelectronic, cyclic, fellow aromatics of c-N3+, c-CNN, HCNN+, and c-CNC− are as well. Each of these are exciting objects for observation of Titan, and the rotational constants and vibrational frequencies produced here will allow for remote sensing of Titan's atmosphere or other astrophysical or terrestrial sources. None of these four aromatic species are vibrationally strong absorbers/emitters, but the two ions, HCNN+ and c-CNC−, have dipole moments of greater than 3 D and 1 D, respectively, making them good targets for rotational spectroscopic observation. Each of these molecules is shown here to exhibit its own, unique vibrational properties, but the general trends put the vibrational behavior for corresponding fundamental modes within close ranges of one another, even producing nearly the same heavy atom, symmetric stretching frequencies for HCNN+ and c-C3H2 at 1600 cm−1. The c-N3+ cation is confirmed to be fairly unstable and has almost no intensity in its ν2 fundamental. Hence, it will likely remain difficult to characterize experimentally.

Using the latest published, empirically refined potential energy surface (PES) Ames-1 and purely ab initio CCSD(T)/aug-cc-pV(Q+d)Z dipole moment surface (DMS), we have computed infrared line lists for five symmetric isotopologues of sulfur dioxide: 32S16O2 (626), 33S16O2 (636), 34S16O2 (646), 36S16O2 (666), and 32S18O2 (828). The line lists are based on J = 0–80 rovibrational variational calculations with E′ ⩽ 8000 cm−1. The 34S16O2 and 33S16O2 line lists are compared to the experiment-based models in the HIgh-resolution TRANsmission molecular absorption database (HITRAN2012, http://www.cfa.harvard.edu/hitran/ ) and the Cologne Database for Molecular Spectroscopy, CDMS ( http://www.astro.uni-koeln.de/cdms/ ). The accuracy for computed 646 band origins is similar to what has been reported for the main isotopologue, i.e. 0.01–0.03 cm−1 for bands up to 5500 cm−1. For rovibrational transitions, the 646 line position and intensity deviation patterns are much simpler and more self-consistent than those of the main isotopologue 626. The discrepancies are mainly found for higher Ka/J transitions. 626 and 646 exhibit comparable line position and intensity agreement for lower Ka/J transitions. The line position deviations for the 636 purely rotational band are parallel to those of 626 and 646, while its line intensity deviations do not show branching patterns as we found in the 626 and 646 cases. Predictions for the other minor isotopologues are expected to exhibit similar accuracy. These line lists are accurate enough to provide alternatives for missing bands of 626 and the minor isotopologues. It may significantly facilitate the laboratory spectroscopic measurement and analysis, as well as to identify these isotopologues in various astrophysical environments.

New semi-empirical Infrared (IR) line lists for 12C16O2, Ames-296 K and Ames-1000 K, have been computed using a newly updated ab initio CCSD(T)/aug-cc-pVQZ dipole moment surface (denoted DMS-N2) and an empirically refined potential energy surface (Ames-1). J=0–150 rovibrational levels are computed up to 30,000 cm−1, and related transitions are cut off at 1E−42 cm molecule−1 (296 K) and 1E−36 cm molecule−1 (1000 K). These are the first line lists available to cover reliably the energy region as high as ~20,000 cm−1. Recent experimental data at 1.1 μm has confirmed the predicted intensities for the 50013-00001 and 50014-00001 band transitions have better than 90% agreement. Comparisons are made against the Wattson 750 K line list and HITEMP/HITRAN at 300 K, 500 K, 725 K, 1000 K, 1500 K, 2000 K and 3000 K. The temperature dependence and accuracy of the new Ames-296 K/1000 K line lists are investigated and we claim both line lists are capable of providing reliable opacities up to 18,000–23,000 cm−1, while the highest applicable wavenumber range drops as T rises. We suggest caution is used for T>1000 K simulations. Comparison to recent experiments at 1000 K, 1550 K and 1773 K shows that the Ames-1000 K line list and HITEMP perform similarly in the 3200–3800 cm−1 and 4600–5200 cm−1 ranges. In the 2000–2100 cm−1 range, Ames-1000 K yields better agreement relative to experiment. Existing problems and possible future solutions in the new Ames-296 K/1000 K line lists for line positions and intensities are also discussed.

Quartic force fields (QFFs) have been shown to be one of the most effective ways to efficiently compute vibrational frequencies for small molecules. In this letter we discuss how the simple-internal or bond-length bond-angle (BLBA) coordinates can be transformed into Morse-cosine (-sine) coordinates which produce potential energy surfaces from QFFs that possess proper limiting behavior and can describe the vibrational (or rovibrational) energy levels of an arbitrary molecular system to 5 cm −1 or better compared to experiment. We investigate parameter scaling in the Morse coordinate, symmetry considerations, and examples of transformed QFFs making use of the MULTIMODE, TROVE, and VTET variational vibrational methods.

Besides the ν(1) O-H stretching mode at 3435 cm(-1) for HOCS(+), the fundamental vibrational frequencies for this cation and its HSCO(+) isomer have not been determined experimentally. Because these systems are analogues to HOCO(+), a detected interstellar molecule, and are believed to play an important role in reactions of OCS, which has also been detected in the interstellar medium, these cations are of importance to interstellar chemistry and reaction surface studies. This work provides the fundamental vibrational frequencies and spectroscopic constants computed with vibrational perturbation theory (VPT) at second order and the vibrational configuration interaction (VCI) method conjoined with the most accurate quartic force field (QFF) applied to date for these systems. Our computations match experiment to better than 2 cm(-1) for the known O-H stretch. Additionally, there is strong agreement in the prediction of the fundamentals across methods and choices of QFFs. The consistency in the computations and the correspondence for the known mode should give accurate reference data for the rovibrational spectra of these cations and their singly substituted isotopologues for D, (18)O, and (34)S.

The effect of approximating the three- and four-virtual molecular orbital integrals in single and double coupled-cluster theory including a perturbational correction for connected triple excitations [CCSD(T)] is investigated for the calculation of higher-order properties, specifically the calculation of a molecular quartic force field and spectroscopic constants. The approach was proposed previously, but investigated for only second- and lower-order properties. It is shown that the conclusions reached previously are essentially unchanged on moving to higher-order properties. That is, approximating the selected integrals has essentially no effect on the accuracy of CCSD(T) calculations, and the error due to approximating integrals is much smaller than the residual error due to one-particle basis set deficiencies. The advantage of this approach is that it significantly reduces the amount of data needed to perform CCSD(T) calculations, thereby reducing computational requirements associated with input/output ...

We report variational calculations of rovibrational energies of CH 4 using the code MULTIMODE and an ab initio force field of Schwenke and Partridge. The systematic convergence of the energies with respect to the level of mode coupling is presented. Converged vibrational energies calculated using the five-mode representation of the potential for zero total angular momentum are compared with previous, benchmark calculations based on Radau coordinates using this force field for zero total angular momentum and for J = 1. Very good agreement with the previous benchmark calculations is found.

Extensions to the code MULTIMODE to obtain rovibrational wave functions and properties are described. An application of these new capabilities is made to a calculation of the Franck–Condon factors for photoionization of CF3 to CF . These calculations make use of a new, full-dimensional ab initio potential energy surface, which is also described here.

This review focuses on the calculation of rovibrational energies of polyatomic molecules using the code MULTIMODE. This code, which uses normal coordinates and a hierarchical n-mode representation of the potential, aims to be applicable to a wide class of molecules and molecular complexes. The theoretical and computational methods used in this code are described, followed by a review of selected applications. These applications illustrate various features of the code and also point out some limitations of the current version of the code. The review concludes with some ideas about possible future directions in this area of research.

A new potential energy surface, based on high quality ab initio electronic structure calculations, is presented for the hydronium ion (H3O+). The new potential surface is used in rigorous calculations of vibrational energies of H3O+, D3O+, H2DO+, and HD2O+. Comparison with experiment shows significant improvement over our previous calculations using an earlier potential [X. Huang, S. C. Carter, and J. M. Bowman, J. Phys. Chem. B 106, 8182 (2002)]. Vibrational calculations are also presented with a new version of the code MULTIMODE. In this version the maximum number of coupled modes in the potential in any grouping of modes is increased from four (the previous maximum) to five. The importance of five-mode terms in the potential is demonstrated for several vibrational states in H3O+ and H2DO+. Also, in the new version of MULTIMODE the number of coupled modes in the Coriolis term can be varied independently from the number of coupled modes in the potential. Rovibrational calculations for J=1 are also presen...

Protonated acetylene, C2H3(+), is among the simplest carbocations. Comprehensive experimental or highly accurate computational spectroscopic data is lacking for this system due to its inherent complexities. Utilizing state-of-the-art quartic force fields (QFFs), the spectroscopic constants and fundamental vibrational frequencies are provided in this work for the nonclassical, bridged, cyclic global minimum. The rotational constants match experiment to better than 0.1%, and the computed ν2 antisymmetric HCCH stretch is less than 3.0 cm(–1) different from experiment. Hence, the rovibrational spectroscopic data provided herein for c-C2H3(+) and its deuterated isotopologues enrich the chemical understanding of this system. Unfortunately, the same rovibrational spectroscopic data is not as trustworthy for the classical, linear form of protonated acetylene due to the shallow well in which it resides on the potential energy surface. However, spectroscopic data are provided for this isomer in the Supporting Information to enhance future studies.

In this work, computational procedures are employed to compute the rotational and rovibrational spectra and line lists for H2O, CO2, and SO2. Building on the established use of quartic force fields, MP2 and CCSD(T) Dipole Moment Surfaces (DMSs) are computed for each system of study in order to produce line intensities as well as the transition energies. The computed results exhibit a clear correlation to reference data available in the HITRAN database. Additionally, even though CCSD(T) DMSs produce more accurate intensities as compared to experiment, the use of MP2 DMSs results in reliable line lists that are still comparable to experiment. The use of the less computationally costly MP2 method is beneficial in the study of larger systems where use of CCSD(T) would be more costly.

Through established, highly-accurate ab initio quartic force fields (QFFs), a complete set of fundamental vibrational frequencies, rotational constants, and rovibrational coupling and centrifugal distortion constants have been determined for both the cyclic 1(sup 1) 1A' and bent 2(sup 1)A' DCCN, H(C13)CCN, HC(C-13)N, and HCC(N-15) isotopologues of HCCN. Spectroscopic constants are computed for all isotopologues using second-order vibrational perturbation theory (VPT2), and the fundamental vibrational frequencies are computed with VPT2 and vibrational configuration interaction (VCI) theory. Agreement between VPT2 and VCI results is quite good with the fundamental vibrational frequencies of the bent isomer isotopologues in accord to within a 0.1 to 3.2 / cm range. Similar accuracies are present for the cyclic isomer isotopologues. The data generated here serve as a reference for astronomical observations of these closed-shell, highly-dipolar molecules using new, high-resolution telescopes and as reference for laboratory studies where isotopic labeling may lead to elucidation of the formation mechanism for the known interstellar molecule: X 3A0 HCCN.

An isotopic-independent, highly accurate potential energy surface (PES) has been determined for CO(2) by refining a purely ab initio PES with selected, purely experimentally determined rovibrational energy levels. The purely ab initio PES is denoted Ames-0, while the refined PES is denoted Ames-1. Detailed tests are performed to demonstrate the spectroscopic accuracy of the Ames-1 PES. It is shown that Ames-1 yields σ(rms) (root-mean-squares error) = 0.0156 cm(-1) for 6873 J = 0-117 (12)C(16)O(2) experimental energy levels, even though less than 500 (12)C(16)O(2) energy levels were included in the refinement procedure. It is also demonstrated that, without any additional refinement, Ames-1 yields very good agreement for isotopologues. Specifically, for the (12)C(16)O(2) and (13)C(16)O(2) isotopologues, spectroscopic constants G(v) computed from Ames-1 are within ±0.01 and 0.02 cm(-1) of reliable experimentally derived values, while for the (16)O(12)C(18)O, (16)O(12)C(17)O, (16)O(13)C(18)O, (16)O(13)C(17)O, (12)C(18)O(2), (17)O(12)C(18)O, (12)C(17)O(2), (13)C(18)O(2), (13)C(17)O(2), (17)O(13)C(18)O, and (14)C(16)O(2) isotopologues, the differences are between ±0.10 and 0.15 cm(-1). To our knowledge, this is the first time a polyatomic PES has been refined using such high J values, and this has led to new challenges in the refinement procedure. An initial high quality, purely ab initio dipole moment surface (DMS) is constructed and used to generate a 296 K line list. For most bands, experimental IR intensities are well reproduced for (12)C(16)O(2) using Ames-1 and the DMS. For more than 80% of the bands, the experimental intensities are reproduced with σ(rms)(ΔI) < 20% or σ(rms)(ΔI∕δ(obs)) < 5. A few exceptions are analyzed and discussed. Directions for future improvements are discussed, though it is concluded that the current Ames-1 and the DMS should be useful in analyzing and assigning high-resolution laboratory or astronomical spectra.

The anion chemistry of the interstellar medium (ISM) has almost exclusively been limited to linear hydrocarbons and cyanocarbons. Of the hydrocarbons, only the even n C n H– chains have been detected in the ISM, and lines hypothesized to originate with b-C3H– have been conclusively linked to the corresponding cation, as originally claimed. However, no reason has yet been provided as to why other anions cannot form, and the cyclic form of C3H– is actually the lowest-energy isomer on the anion's potential energy surface. As such, this work provides the necessary rovibrational reference data for the potential detection of this anion in the ISM or related laboratory experiments. Improvements over previously calculated rovibrational spectroscopic constants are contained herein along with graphical depictions of the pure rotational spectra at 100 K, 40 K, 20 K, and 2.7 K.

A purely ab initio potential energy surface (PES) was refined with selected (32)S(16)O2 HITRAN data. Compared to HITRAN, the root-mean-squares error (RMS) error for all J=0-80 rovibrational energy levels computed on the refined PES (denoted Ames-1) is 0.013 cm(exp -1). Combined with a CCSD(T)/aug-cc-pV(Q+d)Z dipole moment surface (DMS), an infrared (IR) line list (denoted Ames-296K) has been computed at 296K and covers up to 8,000 cm(exp -1). Compared to the HITRAN and CDMS databases, the intensity agreement for most vibrational bands is better than 85-90%. Our predictions for (34)S(16)O2 band origins, higher energy (32)S(16)O2 band origins and missing (32)S(16)O2 IR bands have been verified by most recent experiments and available HITRAN data. We conclude that the Ames-1 PES is able to predict (32/34)S(16)O2 band origins below 5500 cm(exp -1) with 0.01-0.03 cm(exp -1) uncertainties, and the Ames-296K line list provides continuous, reliable and accurate IR simulations. The Ka-dependence of both line position and line intensity errors is discussed. The line list will greatly facilitate SO2 IR spectral experimental analysis, as well as elimination of SO2 lines in high-resolution astronomical observations.

We report a CCSD(T)/cc-pCV5Z quartic force field (QFF) and a semi-global CCSD(T)-F12b/aug-cc-pVTZ potential energy surface (PES) for singlet, cyclic C4. Vibrational fundamentals, combinations, and overtones are obtained using vibrational second-order perturbation theory (VPT2) and the vibrational configuration-interaction (VCI) approach. Agreement is within 10 cm(-1) between the VCI calculated fundamentals on the QFF and PES using the MULTIMODE (MM) program, and VPT2 and VCI results agree for the fundamentals. The agreement between VPT2-QFF and MM-QFF results is also good for the C4 combinations and overtones. The J = 1 and J = 2 rovibrational energies are reported from both VCI (MM) on the PES and VPT2 on the QFF calculations. The spectroscopic constants of (12)C4 and two C2v-symmetry, single (13)C-substituted isotopologues are presented, which may help identification of cyclic C4 in future experimental analyses or astronomical observations.

The interstellar presence of protonated nitrous oxide has been suspected for some time. Using established high-accuracy quantum chemical techniques, spectroscopic constants and fundamental vibrational frequencies are provided for the lower energy O-protonated isomer of this cation and its deuterated isotopologue. The vibrationally-averaged B0 and C0 rotational constants are within 6 MHz of their experimental values and the D(J) quartic distortion constants agree with experiment to within 3%. The known gas phase O-H stretch of NNOH(+) is 3330.91 cm(-1), and the vibrational configuration interaction computed result is 3330.9 cm(-1). Other spectroscopic constants are also provided, as are the rest of the fundamental vibrational frequencies for NNOH(+) and its deuterated isotopologue. This high-accuracy data should serve to better inform future observational or experimental studies of the rovibrational bands of protonated nitrous oxide in the interstellar medium and the laboratory.

Several aspects of ammonia rovibrational spectra have been investigated using the new HSL-2 potential energy surface that includes an approximate correction for nonadiabatic effects. The unprecedented accuracy of rovibrational energy levels and transition energies computed using HSL-2 was demonstrated in Part I of this study. For (14)NH(3), new assignments for a few ν(3) + ν(4) band transitions and energy levels are suggested, and discrepancies between computed and HITRAN energy levels in the 2ν(4) band are analyzed (2ν(4) is the most difficult band below 5000 cm(-1)). New assignments are suggested for existing or missing 2ν(4) levels. Several new vibrational bands are identified from existing, unassigned HITRAN data, including 2ν(2) + ν(4), (ν(3) + ν(4)) -A(')∕A("), ν(1) + 2ν(2), and 2ν(2) + 2ν(4). The strong mixing between the 2ν(4) and 2ν(2) + ν(4) bands is carefully examined and found to be the source of the difficulties in the experimental modeling of 2ν(4). Discussion is presented for preliminary J = 10 results, where the overall root-mean-square error is estimated to be less than 0.039 cm(-1). The analysis of the 4ν(2) band demonstrates both the reliability and the accuracy of predictions from HSL-2. The full list of computed J = 0 band origins (with assignments) and the inversion splittings up to 7000-8000 cm(-1) above the zero-point energy are presented. J = 0-2 levels are reported for those bands below 5100 cm(-1) that are missing from the HITRAN database. For (15)NH(3), excellent agreement is found for the available ν(2) and ν(3) + ν(4)(E) transition energies, but significant deficiencies are shown for HITRAN levels and several corrections are suggested. The (15)N isotopic effects are presented for the J = 0-6 levels of 13 HITRAN bands. For (14)ND(3), we reproduce the pure rotational inversion spectra line frequencies with an accuracy similar to that for (14)NH(3). However, it is not possible to reproduce simultaneously all four pairs of inversion-split vibrational fundamentals to better than 0.05 cm(-1) uncertainty. It is suggested that a reanalysis of some suspicious (14)ND(3) fundamental bands is required. The analyses presented here and in Part I show that rovibrational energy levels and transition frequencies computed with HSL-2 (with nonadiabatic corrections) remain highly accurate well beyond the experimental data used in the refinement procedure. Calculations using HSL-2 are capable of revealing many deficiencies in experimental analyses of ammonia spectra and provide reliable predictions with similar accuracy. It is expected that the results of this study will be useful in the future interpretation of high-resolution spectra from laboratory experiments or from astronomical observations. The present work represents a very significant advance in the state of our knowledge of the spectroscopy of ammonia and its isotopologues.

In this work, we build upon our previous work on the theoretical spectroscopy of ammonia, NH(3). Compared to our 2008 study, we include more physics in our rovibrational calculations and more experimental data in the refinement procedure, and these enable us to produce a potential energy surface (PES) of unprecedented accuracy. We call this the HSL-2 PES. The additional physics we include is a second-order correction for the breakdown of the Born-Oppenheimer approximation, and we find it to be critical for improved results. By including experimental data for higher rotational levels in the refinement procedure, we were able to greatly reduce our systematic errors for the rotational dependence of our predictions. These additions together lead to a significantly improved total angular momentum (J) dependence in our computed rovibrational energies. The root-mean-square error between our predictions using the HSL-2 PES and the reliable energy levels from the HITRAN database for J = 0-6 and J = 7∕8 for (14)NH(3) is only 0.015 cm(-1) and 0.020∕0.023 cm(-1), respectively. The root-mean-square errors for the characteristic inversion splittings are approximately 1∕3 smaller than those for energy levels. The root-mean-square error for the 6002 J = 0-8 transition energies is 0.020 cm(-1). Overall, for J = 0-8, the spectroscopic data computed with HSL-2 is roughly an order of magnitude more accurate relative to our previous best ammonia PES (denoted HSL-1). These impressive numbers are eclipsed only by the root-mean-square error between our predictions for purely rotational transition energies of (15)NH(3) and the highly accurate Cologne database (CDMS): 0.00034 cm(-1) (10 MHz), in other words, 2 orders of magnitude smaller. In addition, we identify a deficiency in the (15)NH(3) energy levels determined from a model of the experimental data.

A series of high-quality, purely ab initio, quartic force fields (QFFs), computed using a procedure we recently proposed, is reported for NH2− and CCH−. The singles and doubles coupled-cluster method with a perturbational estimate of the effects of connected triple excitations, denoted CCSD(T), was used with TZ, QZ, and 5Z quality basis sets and was combined with extrapolation to the one-particle basis-set limit, core-correlation effects, scalar relativistic effects, and higher-order correlation effects to yield accurate QFFs. A “best-guess” reference geometry was determined at the CCSD(T)/5Z level of theory. Analytical transformation removes nonzero gradients to facilitate a second-order perturbation theory spectroscopic analysis. The QFF is transformed into Morse/cosine coordinates in order to perform exact vibrational configuration interaction computations. Equilibrium structures, vibrational frequencies, rotational constants, and selected spectroscopic constants are reported in comparison with experim...

A global potential energy surface (PES) that includes short and long range terms has been determined for the NH(3) molecule. The singles and doubles coupled-cluster method that includes a perturbational estimate of connected triple excitations and the internally contracted averaged coupled-pair functional electronic structure methods have been used in conjunction with very large correlation-consistent basis sets, including diffuse functions. Extrapolation to the one-particle basis set limit was performed and core correlation and scalar relativistic contributions were included directly, while the diagonal Born-Oppenheimer correction was added. Our best purely ab initio PES, denoted "mixed," is constructed from two PESs which differ in whether the ic-ACPF higher-order correlation correction was added or not. Rovibrational transition energies computed from the mixed PES agree well with experiment and the best previous theoretical studies, but most importantly the quality does not deteriorate even up to 10 300 cm(-1) above the zero-point energy (ZPE). The mixed PES was improved further by empirical refinement using the most reliable J=0-2 rovibrational transitions in the HITRAN 2004 database. Agreement between high-resolution experiment and rovibrational transition energies computed from our refined PES for J=0-6 is excellent. Indeed, the root mean square (rms) error for 13 HITRAN 2004 bands for J=0-2 is 0.023 cm(-1) and that for each band is always