Polysaccharides and polyphenol in dried Morinda citrifolia fruit tea after different processing conditions: Optimization analysis using response surface methodology

Overview of experimental program

Wash the collected M. citrifolia fruit with distilled water, slice and dry them, grind them into powder, extract with water, and determine the content of PSC and PPN with an ultraviolet spectrophotometer. Figure 1 describes the experimented program.

Collection of fruit samples

The M. citrifolia fruit was collected from plantations located Yuanjiang, Yunnan Province, China (25° 3″41.92″N, 102° 45″20.39″E). M. citrifolia was introduced by the research team from Hawaii in the United States in 2007 and planted in Yuanjiang plantations. M. citrifolia is a tropical plant, and its optimal growth temperature range is 20 to 30 °C (Gupta et al., 2020). Yuanjiang has a typical dry and hot valley climate, with an average annual temperature of 24.3 °C (Fei et al., 2017; Zhao et al., 2014). Collect M. citrifolia fruit in the white maturity stage and select fruit with the same maturity and size. Immediately after harvesting, immediately wash it carefully with distilled water to remove dust and other contaminants. Wipe off the residual moisture on the surface with absorbent paper for later use.

Chemical and solvents

The chemicals used in this study, such as Folin-Ciocalteu reagent, gallic acid and glucose standard, were purchased in Sigma-Aldrich. All other solvents such as sulfuric acid, phenol and sodium carbonate were of analytical grade.

Sample preparation and extraction

The M. citrifolia fruit was slice up of 0.5–0.8 mm and dried at 60 °C until the weight was constant, and grind it into a powder, pass through an 80-mesh sieve. The powder samples were put into a sealed bag and placed in a dryer at 60 °C for the later use. 0.5 g of M. citrifolia fruit powder was weighed, and explore the effect of different ratio of water/M. citrifolia fruit powder (70, 80 and 90 mL/g), extraction temperature (PSC: 40, 45 and 50 °C; PPN: 85, 90 and 95 °C) and extraction time (5, 10 and 15 min) on the extraction yield of PSC and PPN. The extraction solution was centrifuged at 3880 g for 10 min. The supernatant was used for the determination of PSC and PPN.

Determination of PSC contents

The total PSC content of M. citrifolia extract was determined according to the phenol–sulfuric acid method (Qu et al., 2016) with modification using glucose as the standard. Each 0.1 mL sample added 0.9 mL distilled water and one mL of 5% phenol after mixed with 3.5 mL concentrated sulfuric acid rapidly. The maximal absorption was at 489 nm and the absorbance was measured after standing for 15 min.

The yield of PSC (%) was calculated in the same way as that of Qu et al. (2016).

Determination of PPN contents

The total PPN content in the M. citrifolia extract was determined by using Folin–Ciocalteu method (Ab Rahman et al., 2019) with modification using gallic acid as the standard. Each 0.5 mL sample added 0.5 mL distilled water and one mL of Folin–Ciocalteu reagent, stood for 5 min. Then two mL of 10% sodium carbonate solution was added and stood about 1 h at room temperature in darkness. Finally, the maximal absorption at 762 nm was monitored.

The yield of PPN (%) was calculated in the same way as that of (Qu et al., 2016).

Experimental design and data analysis

To determine the optimized aqueous extraction parameters of PSC and PPN, a Box-Behnken design (BBD) with three levels (−1, 0 and 1) was applied. Based on the single-factor results, the main scope of extraction parameters was identified. Table 1 showed the range and center points of the three operational parameters.

All data in this study were calculated as the average of experiments in triplicate. Response surface methodology (RSM) carried out the model and the experimental design and data analysis were implemented using Design-Expert software 8.0.6. To assess the statistical significance and the reliability of the model, response surface analysis and variance analysis (ANOVA) were performed. The statistically regarded was as significant if P < 0.05.

Model fitting analysis

The optimization of aqueous extraction parameter for high PSC extracts and PPN content from dried M. citrifolia fruit slice tea was conducted using RSM. The yield of PSC and PPN was considered a dependent variable for BBD. The experimental conditions of 15 runs were done respectively, with predicted results from PSC change interval of 13.84–17.02% and PPN change interval of 7.4−8.9% (Table 2). The regression equation of the predicted response Y for PSC and PPN yield was obtained as follows:

(1)${\displaystyle Y_{\mathrm{PSC}}\left(\text{\%}\right)=-\mathrm{89}.\mathrm{52375}+\mathrm{0}.\mathrm{63058}{\mathrm{X}}_1+\mathrm{3}.\mathrm{49650}{\mathrm{X}}_2+\mathrm{0}.\mathrm{45367}{\mathrm{X}}_3-\mathrm{0}.\mathrm{00165}{\mathrm{X}}_1{\mathrm{X}}_2-\mathrm{0}.\mathrm{00525}{\mathrm{X}}_1{\mathrm{X}}_3+\mathrm{0}.\mathrm{0225}{\mathrm{X}}_2{\mathrm{X}}_3-\mathrm{0}.{\mathrm{0032208}}_3{\mathrm{X}}_1^2-\mathrm{0}.\mathrm{039683}{\mathrm{X}}_2^2-\mathrm{0}.\mathrm{048883}{\mathrm{X}}_3^2}$ (2)${\displaystyle Y_{\mathrm{PPN}}\left(\text{\%}\right)=-\mathrm{24}.\mathrm{35}+\mathrm{0}.\mathrm{29417}{\mathrm{X}}_1^{\prime }+\mathrm{0}.\mathrm{35750}{\mathrm{X}}_2^{\prime }+\mathrm{0}.\mathrm{25083}{\mathrm{X}}_3^{\prime }-\mathrm{0}.\mathrm{0005}{\mathrm{X}}_1^{\prime }{\mathrm{X}}_2^{\prime }-\mathrm{0}.\mathrm{0035}{\mathrm{X}}_1^{\prime }{\mathrm{X}}_3^{\prime }+\mathrm{0}.\mathrm{004}{\mathrm{X}}_2^{\prime }{\mathrm{X}}_3^{\prime }-\mathrm{0}.\mathrm{00116667}{\mathrm{X}}_1^{\prime 2}-\mathrm{0}.\mathrm{00166667}{\mathrm{X}}_2^{\prime 2}-\mathrm{0}.\mathrm{015667}{\mathrm{X}}_3^{\prime 2}}$

The ANOVA results determined the reliability of the model of PSC and PPN in predicting the optimum conditions and exactly displaying the real interrelationship between the selected parameters (Table 3).

The regression coefficients were showed in Table 3. For PSC, X₃, X₁², X₂², X₃², X₁X₃, X₂X₃were significant (P < 0.001). For PPN, X′₁, X′₂, X′₁X′₃, X′₃² was extremely significant (P < 0.001). In all the significant terms, the most significant effect on PSC and PPN were X₃² and X′₂, respectively.

Response surface analysis of PSC

The relationship between the response of PSC extraction condition was directly reflected by Fig. 2. At 40∼45 °C, PSC yield with the increase of extraction temperature, while the further addition of extraction temperature caused to decrease in the PSC yield (Fig. 2A). Moreover, under the extraction temperature of 45 °C and the ratio of water/M. citrifolia fruit powder of 80 mL/g, the extraction value could reach 17.02%. PSC yield was gradually increased by extension of extraction time but reduced at higher from 11 min extraction times. Moreover, up to 16.84% of PSC yield was obtained at an extraction time of 10 min and the ratio of water/M. citrifolia fruit powder of 80 mL/g (Fig. 2B).The extraction yield of PSC increases first and then decreases with the increase of extraction time and temperature (Fig. 2C). When the extraction time and temperature was 10 min and 45 °C, respectively, the maximum PSN yield was achieved.

Response surface analysis of PPN

The extraction yield of PPN increased with the extraction temperature and the ratio of water/M. citrifolia fruit powder (Fig. 3A). The optimal region was supposed to be in the ratio of water/M. citrifolia fruit powder above 80 mL/g, a temperature higher than 92 °C which could lead to the extraction yield of PPN over 8.8%. Furthermore, Fig. 3B illustrated that the ratio of water/M. citrifolia fruit powder was positively correlated with the extraction yield of PPN. An increase in the ratio of water/M. citrifolia fruit powder tend to enhanced the PPN yield. Moreover, the extraction time showed quadratic effects on the PPN extraction response. From 5 to 10 min, PPN yield showed an upward trend, and decline slightly over 10 min. Overall, the PPN showed the maximum yield when the ratio of water/M. citrifolia fruit powder was 90 mL/g for 10 min. The optimal region temperature and time of PPN extraction were about 91−95 °C and 8-15 min, respectively, and peaked at PPN yield when 95 °C and 11 min (Fig. 3C).

Optimization conditions and model effectiveness

The optimal conditions for the PSC and PPN extraction were selected based on the constructed RSM models. The predicted yield of 16.98% could be obtained under the ratio of water/M. citrifolia fruit powder of 77.27 mL/g, the temperature of 45.56 °C and time of 10.98 min for PSC. The predicted yield of 8.91% could be obtained under the ratio of water/M. citrifolia fruit powder of 90 mL/g, the temperature of 95 °C and time of 10.09 min for PPN. To verify the effectiveness of models in predicting the yields of PSC and PPN, the conditions were modified based on the actual operation situation and experimental verification on optimized conditions. In the adjusted conditions, the practical yield of PSC and PPN was 17.00 ± 0.08 mg/g and 8.93 ±  0.04 mg/g, respectively (Table 4), which were approximately the predicted results, confirming that the response models were sufficient and precise for the prediction of PSC and PPN extraction.