Adsorption mechanism of triterpenoid saponins in reversed-phase liquid chromatography and hydrophilic interaction liquid chromatography: Mogroside V as test substance

https://doi.org/10.1016/j.chroma.2020.461010Get rights and content

Highlights

  • Study of retention behavior and adsorption model of mogroside V in RPLC and HILIC.

  • Strong dependency of peak shape of mogroside V from ACN ratio and temperature in RP.

  • Langmuir model to describe adsorption of mogroside V in MeOHsingle bondH2O of RPLC and HILIC.

  • Moreau or BET models simulated adsorption of mogroside V in ACNsingle bondH2O of RPLC or HILIC.

  • Three methods for purification of mogroside V (>98%) from S. grosvenorii extract.

Abstract

In this paper, adsorption mechanism of triterpenoid saponins in reversed-phase liquid chromatography (RPLC) and hydrophilic interaction liquid chromatography (HILIC) was proposed based on the study of the retention behavior of mogroside V as test substance. The change of peak shape of mogroside V and its influencing factors was first investigated. As the increase of sample loading, a tailing peak of mogroside V was observed in MeOHsingle bondH2O of both two modes. It was the fronting peak in ACNsingle bondH2O of HILIC while there was a transition from fronting peak to tailing peak in ACNsingle bondH2O of RPLC that was largely affected by column temperature and ACN concentration. The adsorption isotherm of mogroside V in ACNsingle bondH2O of RPLC was fitted by Moreau model, where a monolayer adsorption with large inter-molecular interaction was formed on the C18 surface. While in ACNsingle bondH2O of HILIC, the adsorption of mogroside V was in accordance with BET model, showing multilayer adsorption behavior. In MeOHsingle bondH2O of both HILIC and RPLC, there was always monolayer adsorption, which was fitted by Langmuir model. At last, by choosing the suitable chromatographic mode, controlling the key factors such as the solvent concentration and column temperature, and predicting the broadening trend of peak, three methods were screened out, namely, C18 column with 22% ACN (30 °C), Click XIon column with 90% MeOH or 70% ACN, to get mogroside V of purity greater than 98% from Siraitia grosvenorii extract. Among them, the RPLC method of 22% ACN that showed the highest loading sample per hour (1.92%) and the lowest solvent consumption emerged as the best approach.

Introduction

Because of the hydrophobic aglycones and the hydrophilic sugar chains in the structure of saponins, reversed-phase liquid chromatography (RPLC) and hydrophilic interaction liquid chromatography (HILIC) are the most commonly used high performance liquid chromatography (HPLC) methods in saponin analysis. RPLC mostly uses C18 column to separate saponins, and the difference of each method is mainly in the detector, i.e. ultraviolet detector (UV) [1, 2], evaporative light scattering detector (ELSD) [3, 4] and mass spectrometry (MS) [5, 6]. In a variety of HILIC applications, researchers often choose different columns to separate saponin samples. For example, a silica column was selected to separate the saponin of Gypsophila paniculata [7] and a polyacrylamide-based silica stationary phase was used to separate saponins of Paris polyphylla [8]. Haller et al. used the silica column to isolate three saponins, including two new derivatives of oleanane triterpenoidal saponins [9]. Guo et al. compared the retention behavior of ginsenoside Rc on a Click XIon zwitterion stationary phase using MeOHsingle bondH2O or ACNsingle bondH2O as mobile phases, revealing the effect of organic solvent on the selectivity of ginsenoside separation [10]. In addition, by making full use of the difference of retention mechanism between RPLC and HILIC, two-dimensional liquid chromatography (2D-LC) was established to separate saponins and it was further combined with MS for structural characterization [11], [12], [13].

Although there have been many reports on the separation of saponins by HPLC, the study on the chromatographic adsorption mechanism of saponins is rare. Because the inside of chromatography can not be observed directly, it is feasible to use adsorption isotherm to describe chromatographic separation [14, 15]. Based on the measured isotherm data, the isotherm model is established, usually including Langmuir [16], Moreau [17], BET [18], Quadratic models [19], etc. According to the nature of the model and the meaning of the parameters, the retention mechanism of the sample can be inferred, so as to accurately predict the chromatographic curve and finally improve the success rate of the separation. In RPLC, Georges Guiochon et al. found that methanol absorbed monolayer and acetonitrile absorbed multilayer could be formed on the surface of C18, which affected the adsorption mechanism of compounds. For example, in MeOHsingle bondH2O mobile phase, the adsorption isotherms of phenol on Kromasil C18 and Discovery C18 were fitted with bi-Langmuir and tri-Langmuir models respectively, while in ACNsingle bondH2O mobile phase, the adsorption isotherm was better to use the Langmuir combined with BET models to fit [20]. The effect of column temperature on adsorption isotherm was also examined by the same group [21]. The results showed the reduced adsorption constant and monolayer saturation capacity on the low- and high-energy sites with the increase of temperature, but the increased interaction between absorbents. The application of adsorption isotherm in HILIC is not as extensive as that in RPLC. Researchers paid more attention to the change of water-rich layer [22, 23]. In addition, the effect of water content in mobile phase on adsorption behavior of solute was also investigated [24].

Mogroside V is typical triterpenoid saponin containing five glycosyl groups in its structure (Fig. 1) and its sweetness is about 350 times that of sucrose. It is widely used in food, pharmaceutical and health products, which is an ideal sugar substitute for diabetic, obese and hypertensive patients. HPLC is a common method for the separation of mogroside V and its quantitative experiments are mostly carried out in RPLC mode [25, 26].

In this paper, the work was carried out from two aspects of theoretical research and practical application, to explore the adsorption mechanism of triterpenoid saponins (mogroside V as test substance) in RPLC and HILIC. First of all, chromatographic parameters were investigated including the stationary phase (carbon content, pore size), column temperature, the type and concentration of organic solvents, additives (acid and buffer salt) to put forward the key factors affecting the peak shape of mogroside V. Then, the adsorption isotherm of mogroside V under the key factors was determined to establish the adsorption model. According to the model fitting results and the structural characteristics of mogroside V, the adsorption mechanism in RPLC and HILIC was discussed. Finally, based on the optimization of the key parameters and the prediction of the change trend of peak shape, the preparative-scale separation was carried out to obtain high-purity mogroside V from Siraitia grosvenorii extract.

Section snippets

Experimental instrument and equipment

Waters Alliance HPLC (USA), including 2695 quaternary gradient pump, 2489 UV–Vis detector, autosampler and column oven system; empower 3 chromatographic workstation is used for data acquisition and processing; Waters automatic purification system (USA), including 2545 binary gradient pump, 2767 fraction collector, 2489 UV–Vis detector and autosampler, data acquisition and processing are controlled by software masslynx 4.1. XP105DR analytical balance (Mettler Toledo, Switzerland), electronic

The influence of chromatographic conditions on the peak shape of mogroside V in RPLC

Acetonitrile (ACN) and methanol (MeOH) are the common organic solvents in the RPLC. Firstly, mogroside V was separated using different composition of ACNsingle bondH2O or MeOHsingle bondH2O as mobile phase. Obvious fronting peaks were formed in the ACNsingle bondH2O mobile phase at low sample loading, while a tailing trend was gradually showing with the increase of sample amount (Fig. 2A). In the MeOHsingle bondH2O mobile phase, it always showed a tailing peak no matter the change of sample amount (Fig. 2B). Then, the influence of six

Conclusion

In this paper, the key factors affecting the peak shape of mogroside V were found out, i.e. the type of organic solvent (MeOH and ACN), column temperature and ACN concentration. In both RPLC and HILIC, the peak shape of mogroside V in MeOHsingle bondH2O was always a tailing peak and it formed a monolayer adsorption on the surface of the stationary phase, which can be fitted by Langmuir model. Considering that the peak shape did not change with the chromatographic parameters such as column temperature, pH,

CRediT authorship contribution statement

Qing Fu: Conceptualization, Methodology, Formal analysis, Investigation, Writing - original draft. Lin Ni: Data curation, Software. Dasen Jiang: Software. Yanxiong Ke: Methodology, Formal analysis, Writing - review & editing. Yu Jin: Supervision, Writing - review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (28)

View full text