Introduction

Consumption of foods containing phytochemicals with potential antioxidant properties can reduce the risk of human diseases such as cancer, atherosclerosis, arthritis, diabetes, and other aging diseases (Temple 2000). Efforts have been undertaken to isolate, characterize, and extract antioxidant from natural plant sources. Extraction is the initial step in the isolation of bioactive components from plant materials. The aim of an extraction process is to obtain the maximum concentration of target compounds and of the highest antioxidant activity of the extracts (Spigno et al. 2007). Extraction is influenced by the chemical nature of the compounds, the extraction technique employed, and the presence of interfering substances (Chirinos et al. 2007).

The solvent extraction has been widely used to extract bioactive components from plants. Solvent extraction is a process designed to separate soluble antioxidant compounds by diffusion from a solid matrix (plant tissue) using a liquid matrix (solvent). It is noted that a solvent system for extraction is selected according to the purpose of extraction such as preparation or analysis, the nature of interested components, the physicochemical properties of the matrix, the availability of reagents and equipments, cost, and safety concerns (Yu et al. 2002). The commonly used solvents for extracting antioxidant were methanol, ethanol, and acetone either singly or in combination with aqueous (Lim et al. 2007; Thaipong et al. 2006; Tachakittirungrod et al. 2007; Kahkonen et al. 1999; Velioglu et al. 1998; Zielinski and Kozlowska 2000). The polarities of the different organic solvent greatly influence the selection of a specific solvent for the extraction of a specific group of bioactive compounds.

Extraction of antioxidant is influence by the extraction technique employed (Chirinos et al. 2007). Various techniques have been applied to extract antioxidants from plant materials and other foodstuffs. The techniques commonly used were shaking (Jimenez-Escrig et al. 2001; Lapornik et al. 2005; Xu and Chang 2007), homogenization at high speed (Arnao et al. 2001; Naczk et al. 1992), ultrasonic (Palma and Taylor 1999), maceration (Ahn et al. 2002; Contini et al. 2008), stirring (Alothman et al. 2009), and microwave-assisted extraction (Hemwimon et al. 2007). The selections of particular extraction technique depend on the simplicity of the extraction technique and its convenience.

Guava (Psidium guajava L.) belongs to the Myrtaceae family and is widespread throughout the tropical and subtropical areas. Guava is rich in antioxidants compounds and contains a high level of ascorbic acid ranging from 174.2 to 396.7 mg/100 g fresh fruit (Thaipong et al. 2006). Myricetin and apigenin were reported to be 549.5 and 579.0 mg/kg dry weight, respectively (Koo and Mohamed 2001). The objectives of this research were to investigate the effect of different extraction solvents and techniques on the antioxidant activity of pink-flesh guava fruit using the following assays Folin–Ciocalteu index (FCI), 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical-scavenging assay, and ferric-reducing antioxidant power (FRAP). Each of the above assays measures different aspects of the antioxidant activity of the fruit extracts.

Materials and Methods

Chemicals

Folin–Ciocalteu phenol reagent, ferric chloride (FeCl3·6H2O), and HCl were obtained from Merck (Darmstadt, Germany) and 2,2-diphenyl-1-picrylhydrazyl (DPPH), 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ), gallic acid and Trolox, and sodium acetate trihydrate were purchased from Sigma (USA). Sodium carbonate was purchased from RDH (Germany) while glacial acetic acid was from Mallinckrodt Baker (USA). All chemicals and reagents used in the study were of analytical grade.

Samples Collection and Preparation

Pink-fleshed guava fruits were collected from Setiawan in the state of Perak, North Malaysia. Samples were transfer in ice on the same date to the Food Analysis Laboratory, University Kebangsaan Malaysia. The estimated time of transportation was about 4 h. For the purposes of this study approximately 20 fruits pooled sample portion (taken from a 60 sample lot) were stored in freezer below −25°C until analysis.

Extraction of Antioxidants

Guava fruits were crushed (while still frozen) in a food processor to produce uniform slurry. The extraction procedure was conducted with (0.5 g) samples and 10 mL extracting solvent using one of the following techniques:

  1. 1.

    Homogenization for 1 min under high speed (24,000 rpm) using high-performance disperser (T 25 digital ULTRA-TURRAX®, IKA, Germany)

  2. 2.

    Shaking for 1 h (300 rpm) using shaker (Intertech, Taiwan)

  3. 3.

    Ultrasonic extraction for 1 h at ultrasonic bath (Soniclean, Thebarton, Australia)

  4. 4.

    Mixing with magnetic stirrer for 1 h (1,000 rpm) using magnetic stirrer (Heidolph, MR3001, K, Germany)

  5. 5.

    Maceration of the sample in the extraction solvent for 1, 2, or 3 days

All extracted samples were centrifuged using tabletop centrifuge (Kubota, Japan) for 10 min at 2,580×g. The supernatants were collected for further analysis.

In the second part of this work, the effect of different types of solvent was studied using the best technique selected in the first part. Solvents systems used were absolute methanol, ethanol, acetone, and their aqueous solutions at 50%, 70%, and 100% concentrations. All tests were performed at room temperature.

Determination of Folin–Ciocalteu Index

The FCI used were based on Slinkard and Singleton (1977). Pink-flesh guava fruit extract (0.1 mL), gallic acid (standard calibration), or extracting solvent was placed in a separate 10 mL vials, followed by the addition of 0.4 mL water and 0.5 mL diluted Folin–Ciocalteu reagent. The mixture were swirled and allowed to stand for 5 min followed by the addition of 1 mL of 7.5% (w/v) of sodium carbonate and samples were mixed. Solutions were allowed to stand for 2 h at room temperature and the absorbance was read at 765 nm wavelength using spectrophotometer (UNIKAM, UK). Results were expressed as milligrams of gallic acid equivalents per 100 g of fresh sample (mg GAE/100 g of FW).

Determination of Ferric Reducing/Antioxidant Power

The antioxidant capacity of each sample was estimated according to adapted procedure of Benzie and Strain (1996) with some modifications. FRAP reagent was prepared as using 300 mM acetate buffer, pH 3.6 [3.1 g sodium acetate trihydrate, plus 16-mL glacial acetic acid made up to 1 l with distilled water]; 10 mM TPTZ (2,4,6-tri(2-pyridyl)-s-triazine), in 40 mM HCl; and 20 mM FeCl3 6H2O in the ratio of 10:1:1 to give the working reagent. FRAP reagent, 3,900 µL, prepared freshly and warmed at 37°C, was mixed with 100 µL test sample, standards, or extraction solvent as reagent blank. After 30 min the absorbance was measured at 595 nm wavelength. The result was expressed as milligrams of Trolox equivalents per 100 g of fresh sample (mg TE/g of FW).

Determination of Radical-Scavenging Activity

The decrease of the absorption at 516 nm wavelength of the DPPH solution after addition of the blank or sample extract was measured in a cuvette. An aliquot (3,900 μL) of methanolic DPPH solution (24 mg/L) was mixed with 100 μL of a sample solution (50 mg/mL). The absorption was monitored at the start and at 30 min. The percentage of DPPH scavenging activity was calculated using the following equation: \( {\hbox{Radical - scavenging}}\;{\hbox{activity}} = \left[ {{\hbox{Abs}}\;516\;{\hbox{nm}}\left( {t = 0} \right) - {\hbox{Abs}}\;516\;{\hbox{nm}}\left( {t = t\prime } \right) \times 100/{\hbox{Abs}}\;516\;{\hbox{nm}}\left( {t = 0} \right)} \right] \).

Statistical Analysis

Data collected were analyzed statistically using MINITAB® (14.20) software. Correlation analyses was performed using Pearson’s correlation coefficient (r).

Results and Discussion

Effect of Extraction Techniques on the Antioxidant Activity of Pink-Flesh Guava

Results of antioxidant activity using different extraction techniques are shown in Table 1. The results showed that FCI, DPPH, and FRAP varies with the extraction techniques. The average efficiency of the extraction techniques for FCI and DPPH values in pink-flesh guava fruit showed that both ultrasonic and homogenization were significantly higher (P < 0.05) than other techniques. For FRAP values, ultrasonic was significantly higher (P < 0.05) than other techniques including homogenization. Maceration for 24 h showed significantly (P < 0.05) lower values for FCI, DPPH, and FRAP assay.

Table 1 Effect of extraction methods on the antioxidants activities from pink-flesh guava fruits determined by Folin–Ciocalteu index (FCI)a, DPPH radical-scavenging activityb, and ferric-reducing antioxidant power (FRAP)c
Full size table

When comparing the time needed to achieved the extraction, ultrasonic takes longer time (1 h) while homogenization needs about 1 min to get the same activity for the extract. Extraction with magnetic stirrer technique (1,000 rpm for 1 h), in which extraction was enhanced by stirring effect, or shaking (300 rpm for 1 h) resulted in significantly (P < 0.05) high antioxidant activity compared to maceration for 24 or 48 h but lower than maceration for 72 h. Even though, stirrer and shaker were based on mixing samples and solvents under effect of speed, yet the results are lower than homogenization. Probably this is due the very high speed of homogenization (24,000 rpm) compared with the speed for shaking (300 rpm) and stirring (1,000 rpm). Thaipong et al. (2006) extracted antioxidant from guava fruits using homogenization without reporting the speed. Alothman et al (2009) used the magnetic stirrer at a speed of 1,100 rpm for 3 h for the extraction of antioxidant from different tropical fruits including guava.

The antioxidant assays using maceration (although these values increased significantly with extraction time) were significantly (P < 0.05) lower than those obtained for homogenization extraction, although the same solvent concentrations and the same temperature (room temperature) were used in both cases. Contini et al. (2008) used long maceration at room temperature for hazelnut by-products. Indicated that the use of long maceration at room temperature appeared to be as efficient as hot-extraction with regard to phenolic concentration of the extract. Tachakittirungrod et al. (2007) used maceration for 48 h (×3) for the extraction of antioxidant from different plants including guava. However, the results of this study showed that maceration technique resulted in the lowest value (P < 0.05) of antioxidant activities compared to other techniques studied.

Extraction solvents behave differently with different extraction techniques. The mechanism of ultrasonic extraction involves two types of physical phenomena namely diffusion through the cell walls and washing out (rinsing) of the cell contents once the walls are broken (Vinatoru 2001). Ultrasonic enhanced extraction by intensification of mass transfer and easier access of the solvent to the cell (Jacques et al. 2007). Under the influence of homogenization technique, high speed (the key advantages of homogenization) plays an important role in extraction efficiency. High-speed mixing is known to affect the morphological changes in the plant sample matrix that caused the product to release more readily and thus enhances the extraction process. In maceration, the efficiency of extraction depends mainly on the solubility of the compound in the solvent, the mass transfer kinetics of the product, and the strength of the solute/matrix interactions. In maceration one would expect to obtain the same recovery (as homogenization) at much longer time. However, maceration after 3 days resulted in significantly (P < 0.05) lower value of antioxidant activities compared to homogenization. Extracts that macerated for a longer period lost their antioxidant activities, possibly due to exposure to unfavorable conditions such as light and oxygen (Hemwimon et al. 2007) that lead to chemical degradation or to long-term oxidation.

Extraction time under different homogenization speed was reported (Fig. 1). Extraction at 24,000 rpm for 3 min showed the highest FCI value followed with 18,000 rpm for 3 min. However, when 18,000 rpm was used for 3 min the DPPH radical-scavenging value was higher (P < 0.05) compared to 24,000 rpm for 3 min. However, homogenization at 24,000 rpm showed better repeatability in both FCI (2.32%) and DPPH (0.07%) assays compared to homogenization at 18,000 rpm.

Fig. 1
figure 1

Effect of homogenization speed (rounds per minute, rpm) and time (minute) on the antioxidants activities from pink-flesh guava fruits determined by Folin–Ciocalteu index (FCI) reported as milligrams of gallic acid equivalent per 100 g fresh weight and DPPH radical-scavenging activity reported as % compare to DPPH without added samples

Full size image

Correlations and Repeatability for Antioxidant Assays of Extraction Techniques

The repeatability of the each extraction techniques was performed by calculating the relative standards deviations (RSD %). All extraction techniques showed good repeatability since all RSD % were lower than 10% for all assays. The relative standard deviations for FCI were between 1.31% for maceration (24 h) and 5.65% for magnetic stirring. As for DPPH the relative standard deviations were between 0.22% for maceration (72 h) and 2.92% for maceration (48 h), while for FRAP, the relative standard deviations were between 0.89%for maceration (48 h) and 7.89% for magnetic stirring.

The correlations (Table 2) between polyphenols (FCI) and values for antioxidant activity (DPPH and FRAP) were high. Generally, the correlation between FCI and FRAP were higher when compared to the correlation between FRAP and DPPH or FCI and DPPH. In maceration (48 h), the correlation between FCI and FRAP or DPPH were the highest (1.000). Using magnetic stirrer or maceration (24 h) showed the lowest correlation (0.864) between FCI and FRAP while homogenization showed the lowest correlation (0.692) between FCI and DPPH. Homogenization also showed the highest correlation (1.000) between DPPH and FRAP. Shaking extract showed the lowest correlation (0.744) between FRAP and DPPH. Data in the literature about the relation between concentration of phenolic compounds and antioxidant activity are contradictory. Some authors observed high correlations (Kahkonen et al. 1999; Luximon-Ramma et al. 2003; Thaipong et al. 2006; Tachakittirungrod et al. 2007) while others showed no or weak direct correlation (Akowuah et al. 2005).

Table 2 Pearson’s correlation coefficients of antioxidant activities of various extraction techniques
Full size table

Effect of Extraction Solvent on the Antioxidant Activity of Pink-Flesh Guava

Comparative study was carried out to establish extractive efficiency of various solvents on the antioxidant activity of pink-flesh guava fruit (Table 3). The results showed that FCI, DPPH, and FRAP values were sensitive to extraction solvents whereby in pure solvents, acetone gave the highest extraction efficiency followed by methanol, water, and ethanol, respectively. Aqueous organic solvents were found to give the highest values. Both, 50% acetone and 70% acetone were the best solvents for obtaining extracts with higher antioxidant activities in pink-flesh guava. However, with 50% ethanol the FCI and FRAP values were significantly (P < 0.05) lower than both 50% and 70% acetone, unlike DPPH value where the three solvents showed no significant differences (P < 0.05).

Table 3 Effect of different extraction solvents on the antioxidants activities from pink-flesh guava fruits determined by Folin–Ciocalteu index (FCI)a, DPPH radical-scavenging activityb, and ferric-reducing antioxidant power (FRAP)c
Full size table

Comparing antioxidant activities from this study and other published data is difficult due to the fact that the content of antioxidant compounds can be influenced by extracting solvent, variety, source of the materials (geographical location). Lim et al. (2007) used 50% ethanol for the extraction of Malaysian seeded and seedless white guava fruits. Thaipong et al (2006) extracted antioxidant from Thai pink guava fruits with methanol. Luximon-Ramma et al (2003), Jimenez-Escrig et al. (2001), and Vasco et al. (2008) extracted antioxidant from Spanish and Mauritius guava fruits using two solvents system consist of acetone and methanol. Tachakittirungrod et al. (2007) extracted antioxidants with 95% ethanol from Thai guava fruit, stem, and leaf. Alothman et al. (2009) compared different solvents for the extraction of antioxidant from Thai seedless guava fruit. They reported that 90% acetone extracts showed higher antioxidant activity than other solvents and water showed the lowest activity. As can be seen from these results, the efficiency of solvents to extract the antioxidant compounds differ among different fruits and among different varieties of the same fruit. It is very hard to develop a standards extraction solvent suitable for the extraction of all plant antioxidant compounds.

Effect of Solvent Polarity on the Antioxidant Activity of Pink-Flesh Guava

Variations in the value of antioxidant activities of different extracts might attribute to the change in relative polarity of different solvents used. As found in this investigation, in a mixture with no aqueous content, the extraction efficiency was low and unfavorable. It is clear that the addition of some amount of water enhances the extraction efficiency. The values for polarity index of a mixture of two solvents, P m , were calculated from Eq. 1 (Hemwimon et al. 2007).

$$ {P_m} = {\emptyset_1}{P_1} + {\emptyset_2}{P_2} $$
(1)

where Ø 1 and Ø 2 are the volume fractions of solvents 1, and solvent 2, respectively, and P 1 and P 2 are polarity indices of solvent 1 and solvent 2, respectively.

Table 4 showed the polarities of the solvent used in this study. As can be seen from the table, the increased in the ratio of water increases the polarity index of the mixture. Each solvent seemed to have distinct specificities in the extraction of antioxidants. This fact is in accordance with polarity of the solvent used for the extraction and its solubility and content of antioxidants in the fruit (Turkmen et al. 2006). Thus, there appears to be an optimal solvent composition for extraction using homogenization. From the results showed in Table 3 along with Table 4, it may be suggested that a certain degree of increase in the solvent polarity (up to 50% water) could enhance the solubility of antioxidant compounds in the mixture.

Table 4 The polarities indices of organic solvent–water mixtures
Full size table

Correlations and Repeatability for Antioxidant Assays of Extraction Solvents

All solvents for all antioxidant assays used in this study showed good repeatability as relative standards deviation (RSD %) whereby the lower RSD% the better the repeatability. However, the organic aqueous solvents mixture showed the better repeatability compared to pure organic solvents or water. The relative standard deviations for FCI were between 2.29% for 100% ethanol and 8.78% for 100% acetone. As for DPPH the relative standard deviations were between 0.22% for 70% acetone and 2.92% for water. The relative standards deviation for FRAP were between 1.01% for 50% acetone and 6.89% for 100% acetone.

The results of FCI, DPPH, and FRAP assays used in the present investigation were compared and correlated with each other (Table 5). The higher the DPPH and FRAP values the higher the FCI. It is logical that antioxidant activities were related to the active component in the extract. For FCI and FRAP the highest correlation (0.995) was observed in 100% acetone and water, while 100% methanol showed the lowest value. FCI and DPPH in 50% methanol showed the highest correlation (0.981) while 70% acetone showed the lowest value (0.725). As for FRAP and DPPH the correlation in 50% methanol showed the highest correlation (0.950) while 70% ethanol showed the lowest value (0.589). However, the chosen solvent for antioxidant extraction (50% acetone) showed high correlation between the antioxidant assays compared to the second choice (70% acetone).

Table 5 Pearson’s correlation coefficients of antioxidant activities of various extraction solvents
Full size table

Effect of Repeated Extraction on the Antioxidant Activity of Pink

Fresh pink guava samples were extracted two times to determine the effect of repeated extraction of 50% and 70% acetone on the antioxidant activity. The results showed that first extraction results in significantly higher (P < 0.05) recovery than the second extraction (Fig. 2). Repeated extraction allowed for additional recovery but the time, cost, and potential error introduced while performing a second extraction would not be justified. Second extraction required more chemicals that may contribute negatively towards the environment. Furthermore, the potential error introduced while performing a second extraction would not be justified. Therefore, it was unnecessary to use a second extraction to improve recovery.

Fig. 2
figure 2

Effect of repeated extraction on the antioxidants activities from pink-flesh guava fruits determined by Folin–Ciocalteu index (FCI) reported as milligrams of gallic acid equivalent per 100 g fresh weight, DPPH radical-scavenging activity reported as % compare to DPPH without added samples, and ferric-reducing antioxidant power (FRAP) reported as milligrams of Trolox equivalent (TE) per gram fresh weight

Full size image

Conclusion

This study indicated that the extracts obtained from pink-flesh guava fruits have remarkable antioxidant activities, the extent of which depends on the extraction technique and solvent. Homogenization technique is a simple procedure (one-step extraction) which gives high antioxidant activities while requiring the shortest extraction time (less than 3 min) when compared with the other extraction techniques. Antioxidant extraction depends on the solubility of antioxidant compounds of plant material in the extraction solvent. Acetone was the best solvent compared to methanol, ethanol, or water. The addition of water to organic solvent increased the effectiveness of the extraction. The 50% aqueous acetone was used in this study for best recovery of antioxidant compounds. This aqueous organic mixing allows more scope in the choice of solvent to be used in an extraction process possibly leading to an economic process and improved environmental, health, and safety considerations.