Elsevier

Fuel Processing Technology

Volume 171, March 2018, Pages 117-123
Fuel Processing Technology

Combination of DOSY and 1D selective gradient TOCSY: Versatile NMR tools for identify the mixtures from glycerol hydrogenolysis reaction

https://doi.org/10.1016/j.fuproc.2017.11.011Get rights and content

Highlights

  • DOSY can pseudo separate the reaction mixture of glycerol hydrogenolysis.

  • 1D selective gradient TOCSY is applied to assist the assignment of signals.

  • Deuterated methanol plays as both solvent and matrix in this analysis.

  • The combined method is a powerful tool for analysis of complex mixtures.

Abstract

Hydrogenolysis is necessary for preparing various value-added chemicals from glycerol. Routine methods like GC and HPLC are not sufficient for analysis the reaction mixtures because the products usually have high boiling or similar polarity, especially when new chromatographic peaks are found which requires other technologies such as NMR. Although they were developed for years, but either DOSY or 1D selective TOCSY NMR has nature limitations. Herein we took their advantages and combined them to analyse glycerol hydrogenolysis mixtures for the first time. A model reaction mixture was first pseudo separated in DOSY spectrum, and 1D selective gradient TOCSY was further applied to assign the overlapped signal which could not be confirmed just by DOSY. A genuine reaction mixture was tried afterwards, and main signals could be assigned well in DOSY diffusion dimension. Notably, n-propanol was found as by-product by 1D selective gradient TOCSY, which is not visible in DOSY spectrum. The experimental results demonstrate that the combination of these two NMR methods could provide a fast, effective, feasible and reliable way for the identification of glycerol hydrogenolysis mixtures and is expected to apply in other research areas.

Introduction

Increasing energy demands, concerns about petroleum supplies and the environmental consequences of fossil fuels have driven the researchers in developing renewable transportation fuels [1]. Biodiesel, usually referred to the fatty acid methyl esters, has become increasingly attractive because of its environmental benefits [2]. High quality biodiesel fuel has been prepared via transesterification of methanol with triglyceride, with glycerol as the by-product in around 10 wt% of the total product [3]. Due to its unique chemical structure, glycerol can be converted into several value-added platform chemicals through oxidation, dehydration, acetalization, carboxylation or hydrogenolysis reactions [4], [5], [6]. Among these reactions, hydrogenolysis of glycerol (Scheme 1) into 1,2-propanediol or 1,3-propanediol has been extensively studied because these diols are important for producing polyester resins, polyurethane resins, liquid detergents, pharmaceuticals etc. [7], [8].

The routine methods for product analysis of glycerol hydrogenolysis involves gas chromatography (GC) and high-performance liquid chromatography (HPLC) [9], [10]. However, the main products obtained from glycerol hydrogenolysis are of similar chemical structure and sometimes even the same molecular weight, which makes the analysis complicated when the hydroxyl group-containing compounds were mixed with the unreacted glycerol in the crude products, it is impossible to run GC analysis without derivatization by silylation [11]. The HLPC method can avoid this tedious derivatization step through adjusting the polarity of mobile phase to achieve separation, but HPLC or HPLC-MS is incapable of providing detailed structural information.

NMR is one of the crucial analytical techniques e.g. for chemistry, biology, pharmacology, where it is used to analyse both pure compounds and diverse mixtures including crude reaction mixtures [12], [13]. For example, the composition of liquid product from glycerol hydrogenolysis was characterized qualitatively by 1H and 13C NMR [14]. By comparison with the 1H and 13C NMR spectra of corresponding genuine samples, 1,2-POD and ethanol were identified as the main products. However, when the hydrogenolysis reactions become more complicated, for instance with low conversion rate and poor selectivity, the application of only 1H or 13C NMR to analyse the resulting mixtures is insufficient.

Diffusion-ordered spectroscopy (DOSY), a pseudo two-dimensional (2D) NMR technique, can separate the signal of each species in NMR tube according to their diffusion coefficients (D) [15], [16]. A two-dimensional spectrum is constructed by DOSY, in which one dimension is the conventional NMR spectrum with chemical shift of the substrate and another dimension is D [17]. DOSY NMR is very useful in acquiring information about composition of complex mixture [18], [19]. However, this technique can be limited by overlapped signals in 1H NMR, which will complicate the signal assignment and lead to inaccurate diffusion coefficient determination [20]. Thus matrix-assisted 1H DOSY technique has been developed and applied to support the separation of overlapped signal in diffusion dimension [21], [22], [23]. Unfortunately, the assisted matrix brings additional disturbing signal and universal use of matrix is limited. Other measures, such as data processing, pulse optimization and 3D–DOSY technique, are also used to resolve the problem of signal stacking [24], [25], [26], [27], [28].

1D selective total correlation spectroscopy can provide the information of the proton which is coupled directly or indirectly with the selected and excited nuclei [29]. This technique can thereby extract a net spectrum of a specific compound from a complicated spectrum, even those with heavily overlapped signal, which demonstrate its potential to simplify and assist the analysis of mixtures with DOSY [30]. However, when the proton-proton coupling constant is approaching 0 or when non-protonated carbon, oxygen, nitrogen atoms are found in the molecular skeleton, the transfer of TOCSY fails which therefore restrict its application in many systems [31]. Moreover, finding an appropriate resonance to excite is crucial for TOCSY and it can be very demanding. Fortunately, DOSY can be used to assign the signals of substrate initially and confirm the peak which should be excited.

The components in glycerol hydrogenolysis mixture have similar structure and molecular weights, but the DOSY NMR method could potentially resolve these components when an appropriate solvent is chosen, due to their different solvation [32]. Considering that some of the components may possess similar chemical shift and hence the assignment of signal thereby complicated, 1D selective TOCSY should be applied. Notably, all components in this hydrogenolysis reaction have a short carbon chain, which are suitable for 1D selective TOCSY detection. In this study, DOSY and 1D selective TOCSY techniques are combined to analyse the mixture of glycerol hydrogenolysis for the first time. A model mixture consisted of five main components will be analysed firstly, and a genuine sample from the glycerol hydrogenolysis reaction will be studied afterwards.

Section snippets

Meterials

Glycerol (analytical grade, 99%), 1,2-propanediol (analytical grade, 99.8%), and n-propanol (analytical grade, 99.8%) were obtained from Sinopharm Chemical Reagent Co., Ltd. 1,3-Propanediol (analytical grade, 98%) and i-propanol (analytical grade, 99.5%) were purchased from Aladdin Reagent Company (Shanghai). Chloroform‑d (CDCl3, 99.8 atom% D), dimethyl sulfoxide‑d6 (DMSO‑d6, 99.8 atom% D) and methanol‑d4 (99.8 atom% D) were supplied by Qingdao Teng Long Microwave Technology Co., Ltd. All

1H DOSY analysis of model sample

A model mixture consisting of glycerol, 1,3-POD, 1,2-POD, NPA and IPA was analysed by the DOSY method to get a pseudo 2D spectrum. Initially, readily available deuterated solvents were chosen, i.e. CDCl3 and DMSO‑d6, to dissolve these components for measurement (1H NMR spectra are displayed in Fig. S1 and S2). The DOSY spectra of model sample mixture are presented in Fig. 1. Unfortunately, these five components are not well separated along the diffusion dimension in both solvents. The model

Conclusion

The combination of DOSY and 1D selective gradient TOCSY techniques is of great use in analysing the glycerol hydrogenolysis mixture. With the optimized DOSY and 1D selective gradient TOCSY condition, five major model compounds (glycerol, 1,2-POD, 1,3-POD, NPA and IPA) involved in the glycerol hydrogenolysis could be identified distinctly. Although these components have similar structure and molecular weight, the DOSY method can help us assign signal initially and confirm the resonance that will

Acknowledgements

This work was the Research Project Supported by Shanxi Scholarship Council of China (2015-123). Yingxiong Wang thanks the Key Research and Development Program of Shanxi Province (International Cooperation) (201703D421041) for financial support.

References (41)

  • Y. Yang et al.

    Silica sol assisted chromatographic NMR spectroscopy for resolution of trans- and cis-isomers

    J. Magn. Reson.

    (2016)
  • D. Sun et al.

    Glycerol hydrogenolysis into useful C3 chemicals

    Appl. Catal. B Environ.

    (2016)
  • P. Lameiras et al.

    Glycerol and glycerol carbonate as ultraviscous solvents for mixture analysis by NMR

    J. Magn. Reson.

    (2011)
  • J. Hill et al.

    Environmental, economic, and energetic costs and benefits of biodiesel and ethanol biofuels

    Proc. Natl. Acad. Sci.

    (2006)
  • M. McCoy

    Glycerin surplus

    Chem. Eng. News

    (2006)
  • C.H.C. Zhou et al.

    Chemoselective catalytic conversion of glycerol as a biorenewable source to valuable commodity chemicals

    Chem. Soc. Rev.

    (2008)
  • M. Pagliaro et al.

    From glycerol to value-added products

    Angew. Chem. Int. Ed.

    (2007)
  • A. Corma et al.

    Chemical routes for the transformation of biomass into chemicals

    Chem. Rev.

    (2007)
  • M. Mittelbach

    Diesel fuel derived from vegetable-oils, V [1]: gas chromatographic determination of free glycerol in transesterified vegetable-oils

    Chromatographia

    (1993)
  • R.C. Breton et al.

    Using NMR to identify and characterize natural products

    Nat. Prod. Rep.

    (2013)
  • Cited by (0)

    View full text