2015
DOI: 10.1007/s00894-015-2787-x
|View full text |Cite
|
Sign up to set email alerts
|

Critical test of some computational methods for prediction of NMR 1H and 13C chemical shifts

Abstract: Performance of 18 DFT functionals (B1B95, B3LYP, B3PW91, B97D, BHandHLYP, BMK, CAM-B3LYP, HSEh1PBE, M06-L, mPW1PW91, O3LYP, OLYP, OPBE, PBE1PBE, tHCTHhyb, TPSSh, wB97xD, VSXC) in combinations with six basis sets (cc-pVDZ, aug-cc-pVDZ, cc-pVTZ, aug-cc-pVTZ, IGLO-II, and IGLO-III) and three methods for calculating magnetic shieldings (GIAO, CSGT, IGAIM) was tested for predicting (1)H and (13)C chemical shifts for 25 organic compounds, for altogether 86 H and 88 C atoms. Proton shifts varied between 1.03 ppm to 1… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

4
44
0

Year Published

2017
2017
2023
2023

Publication Types

Select...
4
1

Relationship

1
4

Authors

Journals

citations
Cited by 38 publications
(48 citation statements)
references
References 86 publications
4
44
0
Order By: Relevance
“…In fact, this heavily parameterized DF was designed for thermochemical kinetic studies and not for NMR calculations. Highly unreliable results with BMK were previously reported in prediction of 13 C NMR chemical shifts . In turn, a mid‐ranked position of M06‐L, with the poorest a , b , RMSD, and MAD data, coincides well with the poor performance of this DF in predicting 2–4 J HH .…”
Section: Resultssupporting
confidence: 85%
See 4 more Smart Citations
“…In fact, this heavily parameterized DF was designed for thermochemical kinetic studies and not for NMR calculations. Highly unreliable results with BMK were previously reported in prediction of 13 C NMR chemical shifts . In turn, a mid‐ranked position of M06‐L, with the poorest a , b , RMSD, and MAD data, coincides well with the poor performance of this DF in predicting 2–4 J HH .…”
Section: Resultssupporting
confidence: 85%
“…The greater value of the Pearson correlation coefficient r or its square r 2 was considered as a very simple indication of better adjustment of the compared J data (vide infra). In addition, the four standard statistical metrics for errors, namely, the root‐mean‐square deviation (RMSD), mean absolute deviation (MAD=(1/ n )∑ i |JicalcdJiobsd |, where n is the total number of J CH values), corrected RMSD, that is, CRMSD=[(1/ n )∑ i (JicorrJiobsd ) 2 ] 1/2 , and corrected MAD (CMAD=(1/ n )∑ i |JicorrJiobsd |), were used as much more reliable estimates of the uncertainties of calculation results. Moreover, maximum positive or negative deviations between the Jicalcd (or Jicorr ) and Jiobsd data, that is, max(+/−), were evaluated.…”
Section: Computational Detailssupporting
confidence: 84%
See 3 more Smart Citations