Introduction

Bioactive compounds like polyphenols and flavonoids derived from plants have been reported to possess antioxidant potentials associated with the prevention of several human diseases like cancer, diabetes, inflammation, cardiovascular diseases, and aging [1, 2]. Antioxidants are chemical compounds that possess the ability to scavenge free radicals and/or reduce their production rate and lipid peroxidation in human bodies [3]. Thus, these bioactive compounds are being extracted from plant matrix for food and pharmaceutical industries utilization. In addition, World Health Organization had reported that traditional medicine will continue to play an important role in health care system since 80% population of the third world countries relies on the use of traditional medication [4]. Although, medicinal plants are prone to extinction, hence, the bioactive phytochemicals needs to be identified and isolated for use in therapeutic drugs.

Bitter leaf, scientifically known to be V. amygdalina is an important medicinal plant mostly found in Africa and Asia. The leaves of this plant are popularly consumed as food and medicine. It has been used to treat different ailments like inflammation, diabetes, malaria, oxidative stress diseases, cancer, infertility and hypoglycaemia without any record of toxicity [5,6,7,8]. These pharmacological activities had been associated with the presence of polyphenols, flavonoids, alkaloids, saponins, steroidal glycosides, sesquiterpenes lactones, tannins, and terpenoids in the leaf extracts [8, 9]. More so, the greenish powders are being produced by Medical Traditional Healer Association in Rukararwe, Uganda for consumption as a tea for patient suffering from malaria [10].

In general, the conventional methods including Soxhlet, maceration and decoction used in the extraction of phenolic compounds responsible for the antioxidant property from plant matrix are very time-consuming and require large quantities of solvents [11, 12]. Therefore, the use of microwave-assisted extraction technique has been found to be more effective and environmental friendly in obtaining higher and quality yields of phenolic compounds from plant matrix using a lesser solvent in a shorter time of extraction [12, 13]. The mechanism at which microwave-assisted extraction (MAE) works is different from other types of extraction methods because the extraction occurs as a result of changes in the cell structure caused by electromagnetic waves [11]. Thus, involves a synergistic combination of mass and heat transfers working in the same direction. In contrary, the mass transfer in conventional methods occurs from inside to outside of the substrates and heat transfer occurs from the outside to inside of substrate [13]. The efficiency of MAE is usually influenced by variables like irradiation time, microwave power level, feed-to-solvent ratio, and solvent concentration, therefore optimization of the process variables are inevitable. Response surface methodology (RSM) is an optimization tool that involves the compilation of mathematical and statistical techniques for the evaluation of several experimental parameters interactions. The main advantage of this approach over single factor experiment is that it involves the evaluation of multiple variables and their interactions on the independent variables with a minimal number of experimental runs [14, 15].

Thus, this study evaluate the optimization of the MAE extraction conditions (irradiation time, microwave power level, feed-to-solvent ratio, and ethanol concentration) in obtaining the optimal yields of phenolics, flavonoids and antioxidant activities from V. amygdalina leaf ethanolic extract using RSM, and to further identify the polyphenols and flavonoids which are responsible for its antioxidant activities using liquid chromatography mass spectrometry quadrupole time of flight (LCMS QTOF).

Materials and methods

Plant sample and extract preparation

Vernonia amygdalina fresh leaves were procured from a garden located in Gambang, Malaysia. The leaves were manually removed from the stalks, washed and shade-dried until constant weight was observed. The recorded moisture content before grinding was 0.012 ± 0.15 g water/g dry. The blended sample was sieved to an average particle size of 105 mm, sealed in a dark container and stored in 4 °C refrigerator before extraction. For the extraction process, 10 g of the sample was extracted using a desired concentration of ethanol, irradiation time, microwave power level, and feed-to-solvent ratio based on the experimental design matrix (Table 1). An enclosed ethos reflux microwave extractor (1000 W, Frequency 2450 MHz, Milestone, Italy) using 3-levels of heating: 2 min of preheating to desire temperature, irradiation based on experimental design, and 2 min of cooling to 30 °C were used. Thereafter, the extract was filtered through Whatman qualitative filter paper No. 1, concentrated to dryness using a rotary evaporator (Buchi Rotavapor R-200 coupled with Buchi Vac V-500 pump, Switzerland) and stored in a 4 °C refrigerator until further analysis. The experimental processes were carried out thrice and the average values were computed.

Table 1 Coded and actual values for the independent variables
Full size table

Chemicals and reagents

Folin–Ciocalteu phenol reagent, aluminium chloride salt, quercetin, 2,2-diphenyl-picrylhydrazyl (DPPH), and 2,2ʹ-Azino-bis-3-ethylbenzothiazoline-6-sulfonic acid (ABTS+·), methanol, ethanol, and sodium carbonate anhydrous were purchased from Sigma-Aldrich (M) Sdn Bhd, Selangor. The distilled water was collected from Faculty of Chemical Engineering and Natural Resources laboratory. All chemicals and reagents used in this study were of analytical grade.

Total phenolic content determination

The total phenolic content in V. amygdalina leaf extract was evaluated using Folin–Ciocalteu reagent following the procedure described by Alara et al. [16]. Concisely, V. amygdalina leaf extract (100 µl) was thoroughly mixed with 0.2 ml of FC reagent. Thereafter, the mixture was allowed to stand in dark for 5 min at room temperature, 0.6 ml of 0.2 mM Na2CO3 solution was then added and the mixture was left to incubate for the next 2 h. The absorbance of the mixture was recorded at 765 nm using a UV–Vis Spectrophotometer (Hitachi U-1800, Japan). Afterward, the sample concentration was calculated using the gallic acid (50–500 mg/ml) standard curve equation (y = 0.0006x + 0.0169, R2 = 0.9903) and the result was expressed as mg gallic acid equivalents per gram of dried weight sample (mg GAE/g d.w.). The total phenolic contents of the extracts were calculated using Eq. (1). Ethanol was used as the blank and the analyses were performed thrice.

$$TPC\,=\,\frac{{c\, \times \,V}}{m}$$
(1)

where c is the sample concentration from the calibration curve (mg/ml), V is the volume (ml) of the solvent used in the extraction, and m represents the weight (g) of the dried sample used.

Total flavonoid content determination

Total flavonoid content in the extract was evaluated using the procedure outlined by Alara et al. [16]. Concisely, 1 ml of 2% AlCl3 dissolved in ethanol was thoroughly mixed with 1 ml of V. amygdalina leaf extract (or standard). The absorbance of the mixture was measured at 420 nm using a UV–Vis Spectrophotometer (Hitachi U-1800, Japan) after it has been left in the dark for 60 min at room temperature. The sample concentration was determined using the quercetin (50–500 mg/ml) standard curve equation (y = 0.112x + 0.178, R2 = 0.9945) and the result was expressed as mg quercetin equivalents per gram of dried weight sample (mg QE/g d.w.). The total flavonoid content in the extract was calculated using Eq. (2). Ethanol was used as the blank and the analyses were repeated thrice.

$$TFC\,=\,\frac{{c \times \,V}}{m}$$
(2)

where c is the sample concentration from the calibration curve (mg/ml), V is the volume (ml) of the solvent used in the extraction, and m represents the weight (g) of the dried sample used.

Antioxidant activities determination

The DPPH and ABTS free radical assays were carried out based on the procedure outlined by Alara et al. [3]. Briefly, 0.2 ml of the extract was mixed with 2 ml of 0.1 mM DPPH solution. After 30 min of incubation, the absorbance of the mixture was determined at 517 nm using a UV–Vis Spectrophotometer (Hitachi U-1800, Japan). Methanol was used as blank and the percentage inhibition was determined using Eq. (3).

For ABTS free radical assays, 2 ml of ABTS stock solution (a mixture of 2.45 mM potassium persulfate and 7 mM ABTS+· solutions at a ratio of 1:1) was diluted with 120 ml methanol. Thereafter, 2.85 ml of the freshly prepared solution was added to 0.15 ml of V. amygdalina leaf extract. The mixture was incubated at room temperature for 120 min and absorbance was measured at 734 nm using a UV–Vis Spectrophotometer (Hitachi U-1800, Japan). Methanol was used as blank and the percentage inhibition was determined using Eq. (3).

$$\% \,DPPH/ABT{S^{+ \cdot }}\,inhibition\,=\,\frac{{{A_{control}} - {A_{sample}}}}{{{A_{control}}}} \times 100\%$$
(3)

where Acontrol is absorbance of the mixture of methanol and DPPH/ABTS+· solution; and Asample is absorbance of the mixture of sample extract and DPPH/ABTS+· solution.

Experimental design

Response surface methodology (RSM) with face central centered composite design (FCCCD) consisting of 30 experimental runs with six centre points was used to optimize the extraction factors, including, irradiation time (X1), microwave power level (X2), feed-to-solvent ratio (X3), and ethanol concentration (X4) which were selected based on the results obtained from two-level factorial screening [16]. The coded levels and range of values of all the four process variables are shown in Table 1. The obtained experimental results were fitted to a second-order polynomial equation as shown in Eq. (4). Four dependent responses presented in this study are YTPC = total phenolic content (mg GAE/g d.w.), YTFC = total flavonoid content (mg QE/g d.w.), YDPPH = percentage inhibition of DPPH (%), and YABTS = percentage inhibition of DPPH (%).

$$Y={\beta _0}+\sum\limits_{{i=1}}^{k} {{\beta _1}{x_1}} +\sum\limits_{\begin{subarray}{l} i=1 \\ i<j \end{subarray} }^{{k=1}} {\sum\limits_{{j=2}}^{k} {{\beta _{ij}}{x_i}{x_j}+} } \sum\limits_{{i=2}}^{k} {{\beta _{ii}}x{{_{i}^{2}}_{}}}$$
(4)

where various xi values are independent variables affecting the dependent responses Y; β0, βi, βii, and βij are the regression coefficients for intercept, linear, quadratic, and interaction terms, respectively; and k is the number of variables.

Model validation

The MAE conditions were numerically optimized for the optimum recovery of TPC, TFC and antioxidant activities based on analysis of variance and 3D response surface plots of the independent variables. The obtained models were verified by comparing the predicted data with experimental data.

LCMS-QTOF analysis of the extract

The phenolic and flavonoid compounds present in the extract at optimized condition were identified by LC-ESI-MS/MS analysis using Q-TOF mass spectrometer equipped with PDA detector (Waters Vion IMS, USA) and a symmetry C18 column of 100 mm × 2.1 mm, 1.8 µm particle size (Waters Acquity UPLC HSS T3, USA). The composition of mobile phase were: A, water with 0.1% formic acid and 100% acetonitrile as solvent B. The gradient elution was 30% A and 70% B (0.69 min), 50% A and 50% B (1.16 min), 50% A and 50% B (1.85 min), 30% A and 70% B (3.47 min), 99% A and 1% B (4.63 min) with the flow rate of 0.6 ml/min and injection volume of 20 µl. Identification of the polyphenols were carried out using SYNAPT mass spectrometer (Waters) coupled with an electrospray ionization operating in negative and positive ion mode, mass range of 100–1000 m/z, spray voltage 4 keV, column temperature of 40 °C, sample temperature of 10 °C, gas flow of 50 l/h, desolvation temperature of 555 °C, and desolvation gas flow rate of 800 l/h. Polyphenols in the V. amygdalina leaf ethanolic extract was characterized with MS/MS fragmentation pattern reference standards.

Fourier transform infrared transmission (FTIR) analysis

Fourier transform infrared transmission is a powerful analyser for identifying different functional groups in plant extracts. V. amygdalina leaf extract at optimum condition was analysed using FTIR (Nicolet iS5 iD7 ATR; Thermo Scientific, Germany) equipped with OMNIC software. The extract was analysed using KBr standard procedure to obtain IR spectra in the scanning wave number ranging from 600 to 4000 cm− 1 with a resolution of 4 cm− 1 [17]. The spectra obtained for the extract was interpreted with a chart for characteristics infrared absorption frequencies of functional groups.

Statistical analysis

The experimental design, analysis of the results and response predictions were done using Design Expert 7.0 software® (Version 7.1.6, Stat-Ease Inc., Minneapolis, USA). The experimental runs and the analyses were performed thrice and the results were expressed as mean ± SD. Paired t test was used to compare the replicate results.

Results and discussion

Fitting of models

Four variables (irradiation time, microwave power, feed-to-solvent ratio, and ethanol concentration) were optimized using RSM for optimal recoveries of phenolic, flavonoid and antioxidant activities from V. amygdalina leaf. The analysis of variance (Table 2) shows that the quadratic models can best represent the experimental data in obtaining the optimal recoveries. All the models were most significant (p < 0.0001). Generally, closer relationship between the predicted and experiment data reflects satisfactory of the generated models. The coefficients in terms of coded variables for the predicted models are shown in Eqs. (5, 6, 7, 8) for TPC, TFC, %DPPH and %ABTS, respectively.

Table 2 Analysis of variance for the response surface quadratic models
Full size table
$$\begin{aligned} Y_{{TPC}} = & 110.43 - 1.23x_{1} - 2.48x_{2} - 0.52x_{3} + 5.03x_{4} - 0.77x_{1} x_{2} - 0.86x_{1} x_{3} - 0.92x_{1} x_{4} \\ - & 0.61x_{2} x_{3} - 2.85x_{2} x_{4} - 0.64x_{3} x_{4} - 2.28x_{1}^{2} - 0.68x_{2}^{2} - 4.23x_{3}^{2} - 6.51x_{4}^{2} \\ \end{aligned}$$
(5)
$$\begin{aligned} Y_{{TFC}} = & 91.78 - 1.34x_{1} - 2.56x_{2} - 0.62x_{3} + 4.95x_{4} - 0.66x_{1} x_{2} - 0.77x_{1} x_{3} - 0.80x_{1} x_{4} \\ - & 0.52x_{2} x_{3} - 2.75x_{2} x_{4} - 0.54x_{3} x_{4} - 2.28x_{1}^{2} - 0.55x_{2}^{2} - 4.13x_{3}^{2} - 6.36x_{4}^{2} \\ \end{aligned}$$
(6)
$$\begin{aligned} Y_{{DPPH}} = & 95.80 - 1.30x_{1} - 2.56x_{2} - 0.62x_{3} + 5.00x_{4} - 0.70x_{1} x_{2} - 0.75x_{1} x_{3} - 0.80x_{1} x_{4} \\ - & 0.51x_{2} x_{3} - 2.76x_{2} x_{4} - 0.56x_{3} x_{4} - 2.34x_{1}^{2} - 0.59x_{2}^{2} - 4.19x_{3}^{2} - 6.44x_{4}^{2} \\ \end{aligned}$$
(7)
$$\begin{aligned} Y_{{ABTS}} = & 97.58 - 1.09x_{1} - 2.20x_{2} - 0.61x_{3} + 4.37x_{4} - 0.64x_{1} x_{2} - 0.62x_{1} x_{3} - 0.63x_{1} x_{4} \\ - & 0.43x_{2} x_{3} - 2.40x_{2} x_{4} - 0.59x_{3} x_{4} - 2.06x_{1}^{2} - 0.59x_{2}^{2} - 3.73x_{3}^{2} - 5.68x_{4}^{2} \\ \end{aligned}$$
(8)

where x1, x2, x3, and x4 are the coded variables for irradiation time, microwave power, extraction temperature, and feed-to-solvent ratio, respectively.

The lack-of-fit test (p > 0.05) was used to assess the ‘fitness’ of the models. It can be clearly seen from Table 2 that the lack-of-fit for TPC (0.1557), TFC (0.2646), %DPPH (0.1196) and %ABTS (0.1142) were insignificant, showing the good fit of the models. More so, the obtained p-value illustrates the suitability of the models to accurately predict the variations. The qualities of the fitted models were evaluated based on the coefficient of determination (R2) which were 0.9897, 0.9873, 0.9885 and 0.9891 for TPC, TFC, %DPPH and %ABTS, respectively. Although, a larger value of R2 value does not always show a sound regression model [14]. Thus, the R2 value should be comparable to adj R2 for a good statistical model. The obtained R2 values were not differed greatly from adj R2 (Table 2). In addition, the significant adequacy of the models was verified at 0.0001% level of probability with the R2, adjusted R2 and predicted R2 of greater than 90%. The coefficient of variation (CV) explains the extent to which data were dispersed. Thus, the CV for TPC, TFC and antioxidant activities from V. amygdalina leaf were lower which show a better reproducibility [18].

Effect of extraction variables on the total phenolic and flavonoid contents

Extraction variables including irradiation time, microwave power, feed-to-solvent ratio, and ethanol concentration significantly influence the recovery yields of total phenolic content and total flavonoid content. Amongst the variables, ethanol concentration and microwave power greatly affected the polyphenols significantly (p < 0.0001) followed by irradiation time (p < 0.05). Similar resulted were obtained for optimal recoveries of polyphenols from Prunus domestica, Punica granatum leaf and Haematococcus pluvialis using microwave-assisted extraction techniques [18,19,20]. In contrary, the feed-to-solvent ratio was insignificant (p > 0.05) in obtaining optimal yields of polyphenols from V. amygdalina leaf. Parallel observations were made when extracting total phenolic compounds from Eucalyptus robusta leaf whereby sample-to-solvent ratio pose the greatest impact on the recovery [21].

Solvent polarity plays an important role because of its ability to extract polyphenols through solubilization [22]. Generally, phenolic compounds have a wide spectrum of solubility in ethanol–water solution as compared to mono-component solvent. The addition of a little amount of water to organic solvent increases its polar medium which can facilitate the extraction of polyphenols [23]. Thus, the permeability of this solvent into plant matrix may require using a higher concentration. In this study, variation in the ethanol concentration from 60 to 80% showed an increased in the TPC and TFC yields where optimal values of 112.32 mg GAE/g d.w. and 93.68 mg QE/g d.w., respectively were obtained at 76% v/v of ethanol concentration. In the same vein, microwave power had a pivotal effect on the TPC and TFC recoveries from V. amygdalina leaf. The linear term was statistically significant on the yields. At microwave power level between 500 and 600 W, optimal yields of polyphenols were obtained. The irradiation time is another variable that influences the optimal recoveries of TPC and TFC from V. amygdalina leaf. The linear term was significant in obtaining optimal yields. The quadratic terms of irradiation time, feed-to-solvent ratio and ethanol concentration were significant (p < 0.05) but microwave power was insignificant (Table 2). All the independent extraction variables interactions contributed significantly to the optimal recoveries of both TPC and TFC except the interaction between microwave power level and feed-to-solvent ratio that was insignificant (p > 0.05, Table 2).

Effect of extraction variables on antioxidant activities

The effect of microwave-assisted extraction on the antioxidant activity of V. amygdalina leaf was evaluated using DPPH and ABTS antioxidant assays, as the antioxidant ability of extracts from plant sample tends to fluctuate depending on the assays used [21]. The models for both assays were significant (p < 0.0001) and all linear variables significantly contributed in the order: ethanol concentration > microwave power > irradiation time > feed-to-solvent ratio (depending on the p-value, Table 2) to the antioxidant activities. All the quadratic terms were significant (p < 0.05) except microwave power in attaining optimal antioxidant inhibition from V. amygdalina leaf. Nevertheless, the interaction between microwave power × feed-to-solvent ratio and feed-to-solvent ratio × ethanol concentration were insignificant (p > 0.05) in obtaining optimum DPPH antioxidant inhibition but only the interaction between microwave power × feed-to-solvent ratio was insignificant for ABTS antioxidant inhibition. As the ethanol concentration increases from 60 to 80%, the antioxidant inhibition tends to increase until ethanol concentration of 76%, where optimal DPPH of 97.71% and ABTS of 99.23% were obtained.

The antioxidant capacity of a plant extract may spring up from the combined or synergetic action of each polyphenol that can as well be evaluated by their chemical structure [22]. Thus, the antioxidant activity can be affected by any change in polyphenol content in the extracts of the plant. This can simply be explained through a positive correlation between total polyphenol content and antioxidant activity of the extract. The obtained results from this study results are in good agreement with the earlier reports whereby the ferric reducing antioxidant power (FRAP), DPPH and total antioxidant capacity of photochemiluminescence (TAC-PCL) scavenging activities were correlated with total polyphenols [23, 24]. In general, the phenolic contents and activities of bioactive compounds from the plant are primarily influenced by the genotypes (species and cultivars of the plant).

Optimization responses and validation of the predictive models

The influences of extraction conditions on the recovery yields of TPC, TFC and antioxidant activities from V. amygdalina leaf were evaluated through 3D response surface plots (Fig. 1). The x-, y- and z- axes of the 3D plot represent two factors and response, for instance, microwave power, irradiation time, and TPC (Fig. 1Aa). The response surface slope is associated with evaluation index implying that the larger the slope, the more the evaluation index increases. More so, the interaction between two variables is reflected in the contour of the plot whereby a rounded contour line shows a significant interaction between the two variables and a distorted contour reflects a weak (or insignificant) interaction (Table 2; Fig. 1). The predicted optimal conditions were: irradiation time of 3.49 min, microwave power level of 558.09 W, feed-to-solvent ratio of 1:10, and ethanol concentration of 75.56%.

Fig. 1
figure 1figure 1figure 1figure 1

Effects of irradiation time, microwave power level, feed-to-solvent ratio, and ethanol concentration on the total phenolic content (a); total flavonoid content (b); DPPH inhibition (c); and ABTS inhibition (d) from Vernonia amygdalina leaf

Full size image

The models were verified by extracting the V. amygdalina leaf under the predicted optimal conditions. For the experimental verification runs (n = 3), the MAE conditions were set at irradiation time of 4 min, microwave power level of 558 W, feed-to-solvent ratio of 1:10, and 76% of ethanol concentration. Thereafter, the TPC, TFC and antioxidant activities of the optimum extract were determined to validate the reliability of the predicted optimal conditions. The validation showed that the optimal extraction conditions gave the desirable TPC (113.76 ± 2.34 mg GAE/g d.w.), TFC (94.08 ± 1.89 mg QE/g d.w.), %DPPH (97.98 ± 0.73%), and %ABTS (99.34 ± 0.56%) which were found to be not significantly different from predicted values at p > 0.05 using a paired t-test [25]. The close agreement between predicted and experimental data confirmed the validity of the generated models.

Identification of phenolic compounds in the extract from V. amygdalina leaf using LCMS-QTOF

Phenolic compounds are inevitable active compounds in medicinal plants since they disrupt the bacterium cell wall, altering its membrane potential and prevent the ATP pool which results in the death of bacterium. The polyphenols profiling of the extract was carried out using LCMS QTOF analysis. The parent compounds being confirmed by their respective daughter ions are shown in Fig. 2. Results from this analysis revealed that the ethanolic extract of V. amygdalina leaf contains appreciable amounts of polyphenol and flavonoid compounds, including, 3-4-O-dicaffeoylquinic acid at m/z 515.1201, tran-ferulaldehyde at m/z 223.0617, 3,7-dihydroxy-2-4-dimethoxyphenanthrene-3-O-glucoside at m/z 431.1351, 1,3,5-O-tricaffeoyl-quinic acid at m/z 677.1511, (±)-isoduartin at m/z 377.1242, coniferol at m/z 179.0719, hematine at m/z 290.0568, caesalpins P at m/z 290.0568, 6-gingerol at m/z 293.1766, ginkgol at m/z 347.2593, 1,3-O-dicaffeoylquinic acid at m/z 555.0914, trigonelline at m/z 138.0551, kaempferol-3-O-β-d-glucopyranoside at m/z 449.1091, kaempferol at m/z 287.0552, kaempferol-3-O-β-d-glucuronide at m/z 463.0875, neoline at m/z 438.2855, ephedradine A at m/z 515.2640, ephedradine C at m/z 555.20901, polycancanthine at m/z 262.1419, hirsutine at m/z 391.1992, corypalline at m/z 194.1182, dauricinoline at m/z 611.3116, and songorine at m/z 358.2371. The combined action of these compounds may be responsible for the antioxidant and other biological activities of the extract. In addition, studies have shown that phenolic compounds in plants exert several biological activities like modulation of cellular enzymes, inhibition of cancerous cell proliferation, cell metabolisms, caspase-dependent pathways and cell survival, that may be independent of their conventional antioxidant mechanisms. For instance, ginkgol has been reported to significantly inhibit the migration, proliferation, and invasion of human tumor cells in a dose-dependent manner [26]. 1,3,5-O-tricaffeoyl-quinic acid has been found to be a potent RNase H inhibitor and antiviral agent when human peripheral blood mononuclear cells infected with HIV-1 was tested through cellular assay [27]. Folwarczna et al. had reported that trigonelline possess anti-diabetic properties whereas kaempferol compounds could scavenge free radical cells [28].

Fig. 2
figure 2figure 2figure 2

LCMS QTOF analysis of an optimized ethanolic extract of Vernonia amygdalina leaf for phenolic and flavonoid compounds identification

Full size image

FTIR spectra interpretation

The FTIR spectra of V. amygdalina leaf extract is shown in Fig. 3. Since the sample was extracted using aqueous ethanol, there could be an ethanol band in the spectra. Thus, characteristic ethanol band would be 3345.51 cm− 1 which shows the stretching vibration of O–H group or O-H wagging of phenolic compounds. This strongly suggested the presence of phenolic compounds in the extract. The observed bands at 2978.11 and 2901.62 cm− 1 would be due to stretching vibrations of CH2 and CH3 groups. A band at 1641.85 cm− 1 could be due to stretching vibration of C=C groups, aromatic ring deformations, and the presence of flavonoids and amino acids. More so, the band at 1452.57 cm− 1 could be attributed to CH2, CH3, flavonoids, and aromatic rings. Bands at 1383.50 and 1326.78 cm− 1 may be due to (CH2) and (CH3) bending of methyl which can be associated with the presence of proteins [29]. The band at 1274.06 cm− 1 could be attributed to the presence of C–O groups of polyols like hydroxyflavonoids [30]. The identified bands at 1085.05 and 1043.58 cm− 1 could be due to the presence of C–O- stretching ester group and/or secondary alcohols. In addition, a band at 877.38 cm− 1 could be due to aromatic ring vibration. Therefore, the observed functional groups in the ethanolic extract of V. amygdalina leaf show the presence of amino acids, lipids, phenolic compounds, flavonoids, carbohydrates, saponins, and tannins.

Fig. 3
figure 3

FTIR spectra of ethanolic extract of Vernonia amygdalina leaf

Full size image

Conclusion

The effect of microwave-assisted extraction conditions on the recovery yields of total phenolic content, total flavonoid content and antioxidant activities of ethanolic extract of V. amygdalina leaf has been investigated using response surface methodology. The results showed that TPC and TFC were significantly affected by ethanol concentration, microwave power and irradiation time. In contrary, antioxidant activities (DPPH and ABTS) were affected by the four considered extraction variables. The optimal yields of TPC (113.76 ± 2.34 mg GAE/g d.w.), TFC (94.08 ± 1.89 mg QE/g d.w.), DPPH (97.98 ± 0.73%), and ABTS (99.34 ± 0.56%) were obtained at the optimal condition, viz, irradiation time of 4 min, microwave power level of 558 W, feed-to-solvent ratio of 1:10, and 76% of ethanol concentration. In addition, the identified compounds from LCMS QTOF and FTIR analysis confirmed the presence of phenolic and flavonoid compounds in the extract. The combined actions of the present compounds may be responsible for its antioxidant activity. Thus, the obtained models can be applied in the scale-up production of polyphenols for usage in food and pharmaceutical industries.