Predictions of the ideal gas properties of refrigerant molecules

Generalized Prediction of Enthalpies of Formation Using DLPNO‐CCSD(T) Ab Initio Calculations for Molecules Containing the Elements H, C, N, O, F, S, Cl, Br

“…The investigation in this study covers the following main points: first, the expansion of the method to encompass additional elements and compounds relative to the previous studies, and second, the influence of scaling factors in the context of calculating accurate enthalpies of formation. A further third question arose during the work, namely by which method one should determine the minimum energy structure where multiple conformers exist.…”

“…In this work, we have chosen to use B3LYP with D3BJ [47,48] as well as RI-MP2 for comparison purposes on a subset of the species, as it was used by Paulechka and Kazakov [15] as well as by Demenay et al [32] Despite extensive criticism of B3LYP, [49][50][51][52][53][54][55][56] the functional remains widely used and thus it is valuable to comment upon its performance. BP86 with D3BJ [47,48] is used for its speed, which makes it particularly suitable in cases with finite computational resources.…”

“…However, the method is limited to compounds where reactions with accurately measured or calculated reference species are available, which precludes the calculation of certain compounds. A third method is the use of regression coefficients, which is a modification of the atomization approach. Further, proposed calculation methods used with high‐accuracy computational schemes are HEAT, NEAT, the “method of Fishtik,” or ATOMIC …”

“…In the case of the use of regression coefficients, calculations are carried out for a known set of species and regression parameters are fitted to the elements in the test set to minimize the error in the calculated formation enthalpy. This fitting is intended to compensate for errors in the functional and has been shown to offer good results in the past . This method has been recently modified to use DLPNO‐CCSD(T) for the electronic energy calculation, whereas DFT or RI‐MP2 are used for the geometry optimization and frequency calculation .…”

“…This fitting is intended to compensate for errors in the functional and has been shown to offer good results in the past. [32,41] This method has been recently modified to use DLPNO-CCSD(T) [21,22] for the electronic energy calculation, whereas DFT or RI-MP2 are used for the geometry optimization and frequency calculation. [15] As the authors show, the standard deviation of the error is capable of rivaling the well-established G4 method, [15] while offering a more direct approach to the calculation of formation enthalpies as the weighing of different methods, as used in common compound methods, is no longer required.…”