Development of a HPLC-UV Method for the Separation and Quantification of Hesperidin, Neohesperidin, Neohesperidin Dihydrochalcone and Hesperetin

Abstract An analysis method was developed for the separation and quantification of hesperidin, neohesperidin, neohesperidin dihydrochalcone and hesperetin by using HPLC-UV. Single factor experiments and Box-Behnken Designs were used to optimize separation of four flavonoids, in which a gradient elution method was adopted with 99% acetonitrile and 0.1% formic acid as mobile phases at a flow rate of 0.9 mL/min. A C18 column was used with a column temperature of 35 °C. LODs and LOQs were below 0.84 µg/mL and 2.84 µg/mL, respectively. Linearity with good correlation coefficients (r > 0.99, n = 5) was attained, recovery rate of four flavonoids ranged from 88% to 130%, the RSD indicating results precision for analyzing hesperidin, neohesperidin, neohesperidin dihydrochalcone and hesperetin ranged from 1.2% to 4.6%. Finally, the present method could be successfully applied to identify and quantify hesperidin, neohesperidin and hesperetin in Fructus Aurantii Immaturus and Pericarpium Citri Reticulatae. Graphical Abstract


Introduction
Flavonoids exist widely in plants. Studies have found that there are more than 8000 kinds of flavonoids in plants (Bae et al. 2012). Flavonoids mainly exist in plants in the form of glycosides. Most types of flavonoids are composed of flavonoid glycosides and monosaccharides (Guo et al. 2019).
Hesperidin (HES) is isolated from citrus peel by French scientist Labreton for the first time, and it abundantly exists in citrus fruits (Rajasekar 2021;Zareiyan and Khajehsharifi 2021). HES is composed of a hesperetin aglycone and a rutinose. Neohesperidin (NH) is a natural flavonoid biomolecule composed of neohesperidos and hesperetin moiety. NH is the isomer of HES and is a precursor of neohesperidin dihydrochalcone and neodiosmin (Sammani et al. 2017). Neohesperidin dihydrochalcone (NHDC) is an active flavonoid, which is produced by the hydrogenation of NH (Bozo glan et al. 2014). Hesperetin (HTIN) is a natural flavonoid compound, it widely exists in fruits, vegetables, such as oranges (Scholz et al. 2007) ( Figure S1).
High performance liquid chromatography (HPLC) is an analytical technique that can be used to separate and identify biomass in mixed solutions. Citrus juices contain HES, NH and HTIN (Di Donna et al. 2013). NHDC as a new sweetener can also be added to beverages. Wang et al. established a UHPLC method for the separation of naringin, HES, NH, and HTIN from Fructus aurantii decoction (Wang et al. 2018). As far as we know, there is no literature on the separation of HES, NH, NHDC and HTIN by HPLC method, though the four substances are usually mixed among their transformations.

Optimization of method
The four flavonoids typical HPLC-UV chromatogram was shown in Figure S2. As a result, the separation of four flavonoids occurred within 30 min and the peak shapes of all analytes were sharp, symmetrical, and easily distinguishable. However, the separation degree of HES was less than 1.5.
In HPLC, pH, ratio of organic components, column temperature and flow rate are common factors to improve separation degree and shorten analysis time (F. Houdiere et al. 1997). As shown in Table S2, 70% acetonitrile had the highest separation degree, but there was a certain loss of four flavonoids and longer retention time. It was due to the poor peak symmetry and long retention time of flavonoids in methanol chromatographic system ( Sat ınsk y et al. 2013). Therefore, 99% acetonitrile was chosen as the optimal mobile phase B because of a good peak area, a good separation degree and analyze time.
0.1% formic acid was a widely used liquid mobile phase additive to separate flavonoids (Kuppusamy et al. 2018). The formic acid content increased from 0.05% to 0.15%, the separation degree increased from 1.17 to 1.29. Results in Table S3 showed that the formic acid content had little difference in the area and analysis time. Considering the bad influence of higher formic acid content on the column, 0.1% formic acid was chosen as a mobile phase A. As shown in Table S4, 1 mL/min had the highest separation degree, reaching 1.27. Temperature can affect the analysis speed and separation effect by affecting the mass transfer rate and solute diffusion, and the conventional silicon stationary phase has strict requirements on temperature (Dolan 2002, Han et al. 2015. Results of the influence of temperature on the separation degree were shown in Table S5, the separation degree increased with the increase of temperature. When the temperature rose from 25 C to 40 C, the separation degree increased from 0.99 to 1.56. Obviously, temperature had a significant effect on separation degree. In addition, considering that excessive temperature would reduce the service life of the column, therefore, 35 C was selected as the separation temperature.

Response surface optimization
The BBD (Box-Behnken Designs) design was shown in Table S6. Analysis of variance (ANOVA) is a widely used method for evaluating the quality of simulation models (Yang et al. 2010). R 2 was 0.8319 and CV was 7.26 showing the regression model had low dispersion, high accuracy and reproducibility. As shown in Table S7, the pvalue of regression model and lack of fit were significant and not significant, respectively, indicating this model can fit the experimental data well (Bezerra et al. 2008). Figure S3-S5 were 3D models of various factors on the separation degree. The 3D plots and contour plots directly reflected the relationship between separation degree and various factors. The separation degree under optimum conditions with the flow rate of 0.9 mL/min, formic acid content of 0.1%, and column temperature of 35 C was 1.3, unlike the 16th RSM experiment with a special separation degree of 1.60. Possible reasons may be that, in the actual experimental process, due to the limited temperature control ability of the column temperature box, it was impossible to select a higher temperature as a zero level of temperature factor for the RSM experiment, resulting in no convergence on the response surface related to temperature. In the future, if the instrument conditions are possible, we would do more work by making a better selection of zero level of temperature factor to achieve a better separation degree with good convergence.

Method validation
The linearity equation, LODs (Limit of detections), LOQs (Limit of quantifications) and correlation coefficients of the typical standard curves of the four substances were shown in Table S8, all standard curves showed good linearity with correlation coefficients (r) above 0.99. The LODs of 0.84, 0.69, 0.67 and 0.4 mg/mL for HES, NH, NHDC and HTIN were attained, respectively, and the LOQs of HES, NH, NHDC and HTIN were 2.84, 2.04, 2.22 and 1.34 mg/mL, respectively. The recoveries of HES, NH, NHDC and HTIN ranged from 88% to 130%, with an RSD value varying from 0.9 to 4.4% (Table  S9). As shown in Table S10, the separation method gave good repeatability, the RSD of HES, NH, NHDC and HTIN were 1.3%, 1.8%, 4.6% and 1.2% with average mass fraction 92.7%, 119.6%, 107.9% and 128.4%, respectively.

Method application
All the analyses were carried out under the optimal conditions. As shown in Figure S6, NH and HTIN in Fructus Aurantii Immaturus powder were separated successfully. As shown in Figure S7, HES and NH in Pericarpium Citri Reticulatae powder could be detected and separated using the developed method. However, HTIN was not detected under the optimal condition, which might be due to differences in citrus varieties and planting environment (Zheng et al. 2020). These results indicated that the new method could be applied in the analysis of flavonoids in plants.

Conclusions
In this paper, we have established a new HPLC-UV method to analyze HES, NH, NHDC and HTIN simultaneously by continuously optimizing the volume of acetonitrile, formic acid content, flow rate, and column temperature. The newly established HPLC-UV method could be applied to the separation of flavonoids from Fructus Aurantii Immaturus and Pericarpium Citri Reticulatae. Compared with Wang's study (Wang, et al. 2018), our method is the first to be applied to the separation of four substances, and can be successfully applied to the identification of related substances in the extracts of Fructus Aurantii Immaturus and Pericarpium Citri Reticulatae, which shows the potential of this method in the separation and identification of flavonoids in food and beverage, and the higher recoveries of HES and NH by the new method indicate that our work is meaningful and the analytical method is more accurate.