Theoretical prediction of 13C NMR spectrum of mixed triglycerides by mean of GIAO calculations to improve vegetable oils analysis

https://doi.org/10.1016/j.chemphyslip.2020.104973Get rights and content

Highlights

  • A specific theoretical NMR prediction method is proposed for triglycerides present in vegetable oils.

  • An excellent fitting between experimental and calculated chemical shift after a careful molecular modelling was achieved.

  • This novel approach can open up the way to several studies in NMR food analyses driven by quantum calculations.

Abstract

This pioneering study based on GIAO-DFT methods is aimed to the best prediction of 13C Nuclear Magnetic Resonances (NMR) arising from triglycerides (and also glycerols), known to be the main component of vegetable oils. Provided that fatty esters bound to the glycerol moiety are not affected by the other esterification chains, and slightly affected by their own esterification position (2- internal, or 1/3- external), eight natural molecules are first optimized despite the challenging presence of many non-hydrogen atoms and the large conformational freedom. This preliminary study sheds light on the total chemical shift prediction concerning five fatty esters (Oleic, Palmitic, Linoleic, Stearic and Linolenic) either present in internal or external positions (ten fragments in total); these results display a very good matching to the experimental profile recorded for several vegetable oils chosen as natural mixtures of glycerides. In order to further improve the theoretical to experimental matching, ten simplified triglycerides with the mentioned fatty esters in the two different esterification positions, and flanked by acetyl esters, were studied and optimized. Beyond the best matching reached so far, we notice that the theoretical rationalisation of the overcrowding in the 28.7-29 ppm spectral region in unable to decode the necessary resolution, nonetheless the same theoretical prediction can still drive the appropriate assignments (as for the fifth and sixth carbon attribution of every chain) even against actual misleading reports.

Introduction

NMR spectroscopy is undoubtedly one of the most powerful chemical tool available for structural elucidation of compounds inside natural mixtures since the last decades of the past century. However, complicated systems often generate challenging chemical shift attribution.(Schlegel et al., 2002; Porco et al., 2006; Saielli and Bagno, 2009) Recent developments of computational quantum chemistry methods have highly improved the confidence in structural assignment. This marriage between experimental and theoretically predicted NMR spectra spread over for successful characterization and elucidation of new synthetic and natural chemical compounds. (Aliev et al., 2009) Nowadays, there are many calculation software allowing (even non-experts in theoretical chemistry) to perform reliable predictions on chemical shifts and coupling constants. Orca (Neese, 2012), Gamess (Schmidt et al., 1993) Dalton (Aidas et al., 2014), NWChem (Valiev et al., 2010), ADF (Velde et al., 2001), ReSpect (Malkin et al., 2012) and Gaussian (Frisch et al., 2016) (that was used in this work) are just a few of a long list. Commonly, the most investigated nuclei in routine calculations are 1H (Lomas, 2014) and 13C (Tulyabaev and Khalilov, 2011), however recent research in the field has accomplished excellent quality results even on heteronuclei such as 15N (Samultsev et al., 2014) and 31P (Fedorov et al., 2014). The theoretical aim consists in calculating the molecular second-order magnetic response properties, such as nuclear magnetic shielding tensor. In these terms the most commonly used approaches are: the IGLO (Kutzelnigg et al., 1991) (Individual Gauge for Localized Orbital), the IGAIM (Keith and Bader, 1992) (Individual Gauges for Atoms In Molecules), the CSGT(Keith and Bader, 1993) (Continuous Set of Gauge Transformation) and the GIAO (Wolinski et al., 1990) (Gauge Including Atomic Orbital). The last approach turned out to be the most popular because it is more precise in the convergence of the shielding tensor with respect to the size of the basis set (Yasiltepe et al., 2018). Frequently, calculations of NMR chemical shift involve a single point energy on a fixed conformation, it is clear that NMR experimental spectra depends on several conformations, from the computational point of view this feature is improved by performing a Boltzmann process with respect to the energies of all conformers.(Willoughby et al., 2014) The most popular quantum mechanical methods for NMR simulation are: HF(Roothaan, 1951) (Hartree Fock ab-initio method), higher-level post Hartree Fock method including electron correlation such as MP2 (Møller and Plesset, 1934) (Moller-Plesset theory), DFT(Hohenberg and Kohn, 1964) (density functional theory). DFT methods have become well known in the last years thanks to the possibility of being able to access a wide range of functional forms and the excellent compromise between accuracy and CPU time demand. The physical observable provided by calculation programs is the isotropic shielding tensor which has to be converted in to the experimentally observable chemical shift (c.s.). It can be accomplished by mean of two methods: 1) Using a reference compound (Lodewyk et al., 2012) (usually TMS, that is the experimental calibration standard for 1H and 13C NMR) and modelling both reference and target compound with the same method, the equation for the conversion is the following; δt = σr + δr - σt (where δt and δr are the chemical shifts of the target molecule and of the reference, respectively, while σt and σr are the corresponding isotropic magnetic shielding values). This is the simpler but less accurate method. 2) Carrying out a linear regression, (Pierens, 2014) in this method many isotropic shielding values belonging to several reference compounds are plotted against their experimental chemical shifts. The intercept and slope of the best fitting are used to transform isotropic magnetic shielding values into a chemical shifts according to the following equation:δ = (σ – intercept) / (-slope)

In this paper we reported a computational studies based on a slight modification of the latter method in order to investigate the chemical shift related to the main components of vegetable oil mixture, whose single compound assignment by NMR spectroscopy is quite complex. In this context, according to our knowledge no evidence concerning the combination between theoretical calculation and NMR vegetable oil mixture investigation were reported. This mixture is mainly formed by triglycerides whose ester derivations are fatty acyl residues, such as stearic, oleic, palmitic, linolenic, linoleic etc., these spectra are particularly difficult in interpretation because of the long alkyl and alkenyl chains whose chemical shifts (both 1H and 13C) are extremely close to each other in localized areas of the spectrum. Because of this challenging system, we could not notice any reference concerning theoretical approaches for NMR studies about triglycerides. This is why we focused on 13C chemical shift that turns out to be more selective even though less sensitive than 1H in the study of mixed triglycerides.

Section snippets

Results and discussion

Glycerol esters with three fatty acyl residues are generally called tri-acyl-glycerols or triglicerides (TAG); these are the main components of vegetable oils. Usually, also mono- (MAG) and di- (DAG) esterified glycerols are remarkable compounds within the same natural products. Because of the worthy nutritional and cosmetic use of vegetable oils, it is important to provide smart analytical methods for chemical quantification covering most of the components (product labelling). In this context,

NMR experimental procedures

Deuterated chloroform with a small amount of Tetra-Methyl-Sylane (TMS), used as internal reference, was purchased at reagent grade from Cambridge Isotope Laboratories (CIL) Inc. Briefly, all the CDCl3 solutions were kept homologous by stirring 122 μL of oil and 478 μL of deuterated chloroform (samples of 13.5% of oil weight). Tubes were immediately sealed to prevent the solvent evaporation, which would affect the sample concentration and, in turn, the chemical shift of many signals, especially

Conclusion

In the present work, relatively inexpensive basis set and quantum model were used for geometry optimizations and related 13C NMR calculations concerning the main components of vegetable oils. As it is demonstrated that ester fatty chains of triglycerides do not affect to each other, some modified triglycerides were used for calculation leading to a both reduction of computational times and in the case of unsaturated chains, to a significant improvement in the matching between theoretical and

Declaration of Competing Interest

All the authors declare no competing interests.

References (43)

  • K. Aidas et al.

    The Dalton quantum chemistry program system

    WIREs Comput. Mol. Sci.

    (2014)
  • A.D. Becke

    A new mixing of Hartree–Fock and local density‐functional theories

    J. Chem. Phys.

    (1993)
  • S.V. Fedorov et al.

    Towards the versatile DFT and MP2 computational schemes for 31P NMR chemical shifts taking into account relativistic corrections

    Magn. Reson. Chem.

    (2014)
  • M.J. Frisch et al.

    Gaussian 16, Revision A.03

    (2016)
  • P. Hohenberg et al.

    Inhomogeneous Electron

    Gas Phys. Rev.

    (1964)
  • J. Klikarova et al.

    The Phenolic Fraction of Italian Extra Virgin Olive Oils: Elucidation Through Combined Liquid Chromatography and NMR Approaches

    Food Anal Meth.

    (2019)
  • W. Kutzelnigg et al.
  • R. Liezel et al.

    A novel approach to the rapid assignment of 13C NMR spectra of major components of vegetable oils such as avocado, mango kernel and macadamia nut oils

    Magn. Reson. Chem.

    (2009)
  • M.W. Lodewyk et al.

    Computational prediction of 1H and 13C chemical shifts: a useful tool for natural product, mechanistic, and synthetic organic chemistry

    Chem. Rev.

    (2012)
  • J.S. Lomas

    1H NMR spectra of alcohols in hydrogen bonding solvents: DFT/GIAO calculations of chemical shifts

    Magn. Reson. Chem.

    (2014)
  • V.G. Malkin et al.

    ReSpect program

    (2012)
  • Cited by (11)

    View all citing articles on Scopus
    1

    Both authors equally contributed to the paper.

    View full text