Methanol, ethanol, acetone, ethyl acetate, dichloromethane, chloroform and n-hexane (analytical grade) used for sample extraction were purchased from Sigma-Aldich, Germany. 2-Azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS), 1,1-diphenyl-2-picrylhydrazyl (DPPH), 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ), and 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox) were all purchased from Sigma-Aldrich Chemical Co., Ltd. (St. Louis, Missouri, USA). Other analytical-grade reagents used in the study were purchased from Tianjin Kemiou Chemical Reagent Co., Ltd. (Tianjin, China). The chemicals used to prepare DESs were purchased from Aladdin Biochemical Technology Co., Ltd. (Guangzhou, China).

Methanol, acetonitrile, acetic acid, and formic acid (Sigma-Aldich, Germany) were HPLC grade and water was purified by a MilliQ system (Milllipore Corporation, Bedford, MA, USA). The standards of taxifolin (98%), luteolin-7-O-glucosid (98%), sinapic acid (98%) and cinnamic acid (98%) used for HPLC analysis were purchased from ChemFaces Biochemical Company, Wuhan. Chlorogenic acid (98%) and p-coumaric acid (98%) standards were purchased from Sigma-Aldich, Germany.

UFLC-DAD was performed on an UFLC Shimadzu system (LC-20AD), detector DAD SPD-M20A, all piloted by Labsolutions software.

Avicennia officinalis L. leaves were obtained from the Mekong Delta region of Ca Mau, Vietnam, and identified at the Can Tho University Department of Biology. To remove dirt and grime, the leaves were washed with water. The moisture content of the dried samples was less than 13%. The raw ingredients were crushed into a fine powder with a size of 2 mm. A. officinalis powder was maintained at room temperature in black glass containers until use.

The aim of the preliminary studies was to choose the ideal conditions (solvent concentration, solid-to-liquid ratio, temperature, and time) for the ultrasonic-assisted extraction of the total phenolic acid content (TPC) and total flavonoid content (TFC) from the extract of A. officinalis leaves. The first step of the experiment was to determine the effect of the solvent on the production of phenolic acids and flavonoids. Leaf powder (1 g) and solvent (15 mL) with varying methanol concentrations (50–100%, v/v) were used for extraction. The extraction flask was submerged in an ultrasonic bath at 40 °C for 15 min. The flask was covered to prevent solvent loss during the extraction. Each filtered extract from the solid-liquid mixture was diluted to a volume of 50 mL using the same extraction solvent after the solid-liquid mixture was filtered. The second step of the experiment determined the solid-to-liquid ratio (1:5 to 1:25). Setting the temperature and proper duration for the ultrasound-assisted extraction was the final step of the experiment. The extraction was conducted between 30 °C and 50 °C, and lasted between 10 and 25 min. The extraction effectiveness was assessed through the sum of the peak regions for TPC and TFC.

RSM was used to optimize the phenolic acid and flavonoid extraction parameters from the A. officinalis leaf extract for the highest antioxidant potential. The methanol concentration (X₁), solid-to-liquid ratio (X₂), temperature (X₃), and duration (X₄) are the independent variables. The levels of independent parameters were set based on early experimental findings. For 20 runs of the five stages of the variable optimization process, a rotatable Box-Behnken design was used. Table 1 shows the ranges and numbers of the independent variables in their coded forms. To represent the TPC (Y₁) and TFC (Y₂) yields as a function of the independent factors, linear models were fitted to the experimental data as follows:

${\mathrm{Y}}_{\mathrm{n}}={\mathrm{a}}_0+\sum _{\mathrm{i}=1}^4{\mathrm{a}}_{\mathrm{i}}{\mathrm{X}}_{\mathrm{i}}+\sum _{\mathrm{i}=1}^4{\mathrm{a}}_{\mathrm{ii}}{\mathrm{X}}_{\mathrm{ii}}+\sum _{\mathrm{i}=1}^4{\mathrm{a}}_{\mathrm{iii}}{\mathrm{X}}_{\mathrm{iii}}+\sum _{\mathrm{i}=\mathrm{j}=1}^4{\mathrm{a}}_{\mathrm{ij}}{\mathrm{x}}_{\mathrm{i}}{\mathrm{x}}_{\mathrm{j}}$

where Yn represents the response variables, a0 is a fixed amount, and ai, aii, and aij represent the linear, quadratic, and interaction coefficients, respectively. The independent variables are Xi and Xj. Three-dimensional surface response plots were produced by altering the two variables within the experimental range and maintaining the third constant at the center point. The coefficients of the response surface equation were calculated using Design Expert 11. A 95% confidence level (p < 0.05) based on the total error criterion was used to test for statistical significance.

A UFLC Shimadzu (LC-20AD) system with a DAD detector, autosampler, injection system, and quaternary pump was used to identify and quantify phenolic acids and flavonoids in the crude methanol extracts. The system was controlled using LabSolution software. A 4.6 × 250 mm Agilent C18 column with 5 µm particle size was used to measure phenolic acid and flavonoid components. The complete loop injection option was used to inject 20-µL samples. The mobile phase was used to elute the column at a constant flow rate of 1 mL/min and consisted of HPLC-grade acetonitrile (solvent A), HPLC-grade methanol (solvent B), and water containing 0.2% ammonium acetate and 0.1% formic acid. The gradient elution program included: 0–17 min, 3% B and 90% C; 32 min, 9% B and 90% C; 41 min, 10% B and 83% C; 48 min, 11% B and 82% C; 55 min, 15% B and 74% C; 72 min, 18% B and 73% C; 86 min, 20% B and 80% C; and 90–95 min, 7% B and 90% C. The DAD detector was set at 280 nm.

Stock solutions of chlorogenic, p-coumaric, sinapic, and cinnamic acids, taxifolin, and luteolin-7-O-glucosid were prepared in methanol at 1000 μg/mL and stored at 4 °C. Then, solutions were diluted in methanol to obtain seven concentrations (from 1.25 to 150 μg/mL). The peak areas and concentrations of each standard were fitted to linear regression and linear regression after square root transformation to select the most suitable regression model.

The validation of the method was performed with three independent series of experiments to analyze the following criteria: selectivity, linearity, limit of detection (LOD), limit of quantification (LOQ), precision, and accuracy according to the Association of Official Analytical Chemists (AOAC) guidelines (Association of Analytical Communities International 2013).

Six crude methanol extracts of A. officinalis leaves collected from different coastal provinces of Vietnam were used. Leaves were harvested from many different A. officinalis trees to obtain 1 kg of dry leaf samples. The methanol leaf extracts were analyzed using validated methods for their concentrations of chlorogenic, p-coumaric, sinapic, and cinnamic acids, taxifolin, and luteolin-7-O-glucosid to analyze variations according to the collection location and determine the best one.

Statistical analyses were performed using Microsoft Excel and GraphPad Prism 8. All measurements were performed in triplicates. Data and figures from the preliminary experiments were generated using Microsoft Excel. Analysis of variance (ANOVA) of the results of the RSM test and the correlations of extraction parameters with TPC and TFC were analyzed using Design Expert 11.

The extraction is very important for the isolation and identification of phenolic acid and flavonoid compounds in plants. In this study, a one-factor-at-a-time approach was used to select the appropriate extraction conditions (solvent concentration, solid-to-liquid ratio, temperature, and time). TPC and TFC were chosen as the indices for the extraction process. The effect of different methanol concentrations (50%, 60%, 70%, 80%, and 100%; v/v) on the TPC and TFC yields was tested at 40 °C for 15 min. The results showed that TPF and TFC depended on methanol concentration (Fig. 1a). The total phenolic acid and flavonoid yields increased with an increase in methanol concentration from 50% to 80% and then decreased when 100% methanol was used. The maximum TPC and TFC yields were 126.47 μg/mL and 75.36 μg/mL, respectively, when methanol concentration was 80%. These results were similar to those previously reported (Liyana-Pathirana et al. 2005; Silva et al. 2007), and might be attributed to the change in solvent polarity with changes in methanol proportion. Furthermore, a high methanol concentration may cause protein denaturation, which prevents the dissolution of polyphenols and affects the extraction rate (Yang et al. 2009). Considering the limitation of the highest methanol concentration in the design, 80% was chosen as the center value for the RSM.

Figure 1. 

Influence of methanol concentration, ratio of solid to liquid, temperature and time ultrasonic on the extration of TPC and TFC from A. officinalis leave extracts.

The effect of the solid-to-liquid ratio was investigated using 80% methanol as the solvent at 40 °C for 15 min. To evaluate the effect of this factor on the extraction yields, we examined different ratios ranging from 1:5 to 1:25 g/mL. Fig. 1b shows that TPC and TFC were significantly affected by the solid-to-liquid ratio. Therefore, a 1:15 ratio was selected as the center point for the RSM.

The effect of the temperature (30 °C, 35 °C, 40 °C, 45 °C, and 50 °C) was investigated using 80% methanol with a solid-to-liquid ratio of 1:15 for 15 min. Fig. 1c demonstrates that TPC and TFC were significantly affected by the temperature. When the temperature increased from 30 to 45 °C, a substantial increase in the TPC and TFC was observed. However, a further increase in temperature induced a decrease in TPC and TFC. The rise in temperature may benefit the extraction of TPC and TFC by decreasing the viscosity coefficient, increasing the diffusion coefficient, and enhancing the solubility of phenolics (Cacace et al. 2003). Simultaneously, heat-sensitive compounds may be degraded at higher temperatures. Therefore, a temperature of 45 °C was fixed as the center point for the RSM.

The effect of the ultrasonic time (5 min, 10 min, 15 min, 20 min and 25 min) was investigated using 80% methanol with a solid-to-liquid ratio of 1:15 at 45 °C. Fig. 1d demonstrates that TPC and TFC were significantly affected by ultrasonic time. When the ultrasonic time increased from 5 to 20 minutes, a substantial increase in the TPC and TFC was observed. However, a further increase in ultrasonic time at 25 minutes induced a decrease in TPC and TFC. Therefore, 20 minutes was fixed as the center point for the RSM.

An experiment was conducted to confirm the reliability of the RSM design under the ideal ultrasound-assisted extraction conditions established through the model. The ideal conditions for TPC extraction were determined using the following parameters: 55.27% methanol concentration, 14:1 (mL/g, v/w) for the liquid-to-solid ratio, 48.8 °C as the temperature, and 9.66 min as the duration of the ultrasound-assisted extraction at optimal conditions. TPC and TFC were 127.841 μg/mL and 46.18 μg/mL, respectively. In contrast, the greatest 162 μg/mL TPC and 75.38 μg/mL TFC were projected with 80% of methanol concentration and a 15:1 (mL/g, v/w) liquid-to-solid ratio at 45 °C for 20 min. When phenolics from Inga edulis were extracted, a similar result was observed (Silva et al. 2007). This could be caused by process flaws, including phenolic breakdown at high temperatures, prolonged extraction times, or use of high solvent concentrations.

Four standards of phenolic acids, including chlorogenic, p-coumaric, sinapic, and cinnamic acids, and two standards, including taxifoli and luteolin-7-O-glucosid were selected to quantify the major phenolic compounds from the methanol extracts of A. officinalis leaves from Vietnam by the validated method we developed.

Selectivity

Selectivity was validated by comparing the retention times and UV spectra of peaks of the crude extract and standards of six analytes at retention times corresponding to the beginning, middle, and end of these peaks. Similar results indicated the selectivity of the method (Fig. 4).

Figure 4. 

HPLC chromatogram of A. officinalis leaves extract. a Mobile phase solvent; b Solute sovent; c Extraction solvent; d Spiked sample; e Mixed standards solution; f Standard addition. (C1: Chlorogenic acid, C2: p-coumaric acid, C3: Sinapic acid, C4: Taxifolin, C5: Luteolin -7-O-glucoside, C6: Cinnamic acid).

The leaves of A. officinalis used in this study were collected from six provinces in Vietnam. Crude methanol extracts were analyzed using the methods developed. Quantification results (Table 5) showed that the concentration of phenolic acids and flavonoids in the methanol extract of the six samples varied from 0.319 to 3.524 mg/g. The highest total content of bioactive components was found in samples collected in the Kien Giang province.

Cinnamic acid is the major phenolic acid component in the leaves of A. officinalis in Vietnam. The highest content of cinnamic acid reached 2.531 mg/g in the crude extract in samples from the Kien Giang province, while no cinnamic acid was detected in the leaf samples collected from the Ben Tre province, and the lowest phenolic acid content was chlorogenic acid, which ranged from 0.005 to 0.102 mg/g.

The leaf samples from the Soc Trang Province had the highest TFC (0.856 mg/g) and contained the highest concentration of luteolin-7-O-glucoside (0.841 mg/g). The sample collected in the Bac Lieu province had no detected luteolin-7-O-glucoside and the content of taxifolin was low (0.018 mg/g).

Huang studied polyphenol-rich Avicennia marina leaf extracts (Huang et al. 2016). The content of bioactive compounds in the water extracts included chlorogenic acid (0.0089 mg/g), p-coumaric acid (0.0058 mg/g), and luteolin (0.0022 mg/g). The A. officinalis polyphenol content in Vietnam was 10 times higher than that of A. marina in Taiwan. To the best of our knowledge, this is the first report on the sinapic acid and taxifolin content in A. officinalis.