The effect of incorporating herbal (Lippia javanica) infusion on the phenolic, physicochemical, and sensorial properties of fruit wine

Abstract The use of medicinal herbs in food processing to improve food quality and human health is growing in sub‐Saharan Africa. Herbal infusions have perceived medicinal benefits. This study investigated the effect of incorporating L. javanica extract on the phenolic, physicochemical, and sensory properties of a Uapaca kirkiana fruit‐based wine. The fruit and L. javanica were analyzed for proximate, pH, total soluble sugars (TSS), total sugar, titratable acidity, phenolics, and antioxidant activity (AOA). The prepared substrate was fermented at optimized fermentation temperature of 23°C, pH of 3.4, inoculum concentration of 9.5% (v/v) (Saccharomyces cerevisiae), and L. javanica extract concentration of 10% (v/v). The produced wine had a pH, total sugar, TSS, total acidity, and vitamin C content of 3.1 ± 0.2, 3.0 ± 0.1 g/L, 4.8 ± 0.1%, 5.9 ± 0.01 g/L, and 4.4 ± 0.1 mg/100 g, respectively. The wine had 12.2 ± 2.1 mg GAE/g, 0.06 ± 0.01 mg/g, and 1.8 ± 1.1 mg CE/g total phenols, tannins, and flavonoids, respectively. The alcohol, free sulfur dioxide (SO2), AOA, and color values of the produced wine were 10.2 ± 0.1 alcohol by volume (ABV)%, 58.1 ± 1.2 mg/L, 73.1 ± 0.1 EC50 mg/L, and 40.4 ± 2.1% yellow, respectively. The flavor, color, and overall acceptance of the produced wine were rated as “good” and were significantly different (p < .05) from control wine. The addition of L. javanica extracts enhanced total phenol, color, and sensory properties of the wine. The utilization of U. kirkiana fruit by incorporating L. javanica infusion can reduce postharvest losses and improve nutrition and health.


| INTRODUC TI ON
In recent years, there has been an increase of interest in herbal infusions due to the effective promotion of their medicinal properties (Bhebhe et al., 2015). An underutilized indigenous herbal plant, Lippia javanica, has been made popular in many communal homes of Zimbabwe as a tisane (Maroyi, 2017). Lippia javanica (Burm.f.) Spreng belongs to the family Verbenaceae ( Figure 1a) and is the tidiabetic (Arika et al., 2015), antioxidant , antimicrobial (Viljoen et al., 2005), and anticancer (Fouché et al., 2008) activity of L. javanica. L. javanica is currently being used as a herbal plant with many ethnomedicinal applications such as in asthma, bronchitis, chest pains, colds, cough, diarrhea, and fever (Maroyi, 2017). Lippia javanica plant has no health effect associated with its consumption in humans. Lippia javanica is used as a food additive in Kenya (Kipkore et al., 2014) and leafy vegetable in India (Narzary et al., 2015). Conversely, the use of mature L. javanica plant as a food additive and in food blends is still limited in Zimbabwe.
Fruits provide energy, minerals, vitamins, and phytochemicals, and their consumption improves the physiological functions and overall health in humans (Kaur et al., 2019). Uapaca kirkiana belongs to the genus Uapaca (Figure 1b) of the family Euphorbiaceae (World Agroforesty Centre, 2021) and subfamily Phyllanthaceae (Ecocrop Database, 2021) and is an underutilized indigenous fruit tree (IFT) that is well adapted to the miombo ecological zone in sub-Saharan Africa (Chawafambira et al., 2020). The fruit is known from different countries by different names as wild loquat in English, nkusu/ msuku in Tanzania, Zambia, and Malawi, and mahobohobo (Ndebele language) or mazhanje (Shona language) in Zimbabwe (Chawafambira et al., 2020). The fruit tree is found distributed in semidry and wet areas of Zimbabwe and it produces fruits which ripen from October to February (Chawafambira et al., 2020). Besides, Stadlmayr et al. (2013) reported the proximate composition of the fruit pulp as ash (1.1 g/100 g), carbohydrates (28.7 g/100 g), calories (523 kcal/ kJ), fat (0.4 g/100 g), fiber (2.3 g/100 g), proteins (0.5 g/100 g), vitamin C (16.8 mg/100 g), and water (72.6 g/100 g).
Fruits wines have recently increased public interest as healthenhancing functional foods (Tsegay & Gebremedhin, 2019). Studies have revealed the possibilities of developing health-enhancing functional fruit wines through the incorporation of medicinal plants with fruit juice (Fracassetti et al., 2019;Lee & Chen, 2016). Globally, fermentation of grape and apple fruits is commonly used to commercially produce fruit wines and the fermentation of U. kirkiana fruit juice with added L. javanica leaf extracts has the potential to produce functional fruit wines. The production of fruit wines from other fruits other than grapes is gaining customer interest because they are considered as functional foods (Velić et al., 2018).
In Zimbabwe, U. kirkiana fruits have not yet been incorporated into industrial wine production. Also, L. javanica has not been incorporated in fruit wines. Maroyi (2017) reported the potential of using L. javanica in new food product development. Therefore, this study aimed at investigating the effect incorporating L. javanica infusion on the phenolic, physicochemical, and sensorial properties of fruit wine.

| Sample collection
Ripe U. kirkiana fruits that had fallen on the ground were collected from Bikita communal area. Bikita is a dry area located 20. 5 S° 31.37° E in Agro farming region 4 in Zimbabwe. U. kirkiana fruit trees are found distributed in the forests and some are domesticated at homes. L. javanica was collected from plants that were growing in the forest and were sun-dried.

| Preparation of Lippia javanica extracts (infusion)
The extraction process was conducted following the water extraction method. Distilled water was used in the extraction process because it was found a better extraction solvent for phenolic compounds (Bhebhe et al., 2015). Sun-dried L. javanica leaves were sorted and washed (twice) to remove any dirt particles. The leaves were then boiled in a stainless pot with 1L distilled water at 65-70°C for 8 min. The leaves were then strained, and the liquid extract was cooled and analyzed.

| Preparation of Starter culture
Commercial yeast Anchor Wine Yeast (N96 Saccharomyces cerevisiae bayanus) was purchased from a local supermarket. The yeast was prepared according to the manufacturer's specifications which were 20-30 g/hl. A weight of 0.3 g dry anchor wine yeast cell was hydrated with 50 ml of warm distilled water at 35°C and added into the yeast extract peptone dextrose (YEPD) media. The mixture was then diluted to 150 ml using sterilized distilled water. This mixture was then incubated for 24 hr at 28°C and shaked at 44 × g in an Orbital digital rotary shaker (KS501, Thermo Fisher Scientific). The incubation temperature was used because yeast grows well at room temperature range of 25-28°C but not at high temperature above 30°C. The mixture was then transferred into a 1-L Erlenmeyer flask which was cleaned with cotton wool that was soaked in 1% sulfur dioxide and had 500 ml of the sterilized U. kirkiana juice pulp. The mixture was adjusted by the addition of 200 g/L sugar, 70 mg/L sodium thiosulfate, and lemon juice to a pH of 3.4. The mixture was then incubated for 36 hr at 28°C and ready for use in the fermentation process.

| Fermentation of must to wine, clarification, and filtration
Fermentation, clarification, and filtration of the fruit wine samples were carried out according to the method described by Tsegay and Gebremedhin (2019) with slight modification. Primary fermentation was initiated in wine by the addition of the starter culture.
One (a) liter capacity fermentation vessels (Erlenmeyer flasks) were prepared by soaking in a 1% sulfur dioxide (19 g/L sodium metabisulfite) solution for 1 hr and then rinsed with distilled water. A 1 L of prepared U. kirkiana juice was then poured into the fermentation vessels under aseptic conditions. The sugar content of the substrate was then adjusted by addition of table sugar (200 g), pH of 3.4 (using 0.5 g/L fresh lemon juice), and sulfur thiosulfate (70 mg/L).
The adjusted pH of 3.4 was used in this study because it inhibits spoilage bacteria and wine yeast can metabolize at this pH level.
L. javanica extract and inoculum (propagated yeast) were added at a concentration of 10% (v/v) and 9.5% v/v of 8.2 Log CFU/mL viable cells, respectively. This was done to ensure rapid onset of fermentation of the must which had high sugar content. The control experiments had no L. javanica extract added. The fermentation vessels were closed using cotton wool that was soaked in 1% sulfur dioxide.
Secondary fermentation was done in an air-tight flask at 23°C and monitored for 10 days. This temperature was used to promote a steady growth of yeast cells and controls the levels of volatile compounds. After fermentation, the wine was chilled to 5°C, racked with minimum exposure to the air, and clarified. The fermented wine was then filtered using a sterilized cotton cheesecloth. The wine was then packaged into 330-mL sterilized glass bottles and then allowed to mature at 15-24°C. Mineral content and carbohydrate content were analyzed using inductively coupled plasma optical emission spectrometer (ICP-OES) (Agilent 5100, Agilent Technologies) and difference method, respectively (AOAC, 2005). The vitamin C (ascorbic acid) content was determined by the dichlorophenolindophenol (DCPIP) titration method (AOAC, 2005). The total sugar content was determined using the sulfuric acid method described by Debebe et al. (2018).

| pH and TSS
The pH and TSS of the fruit juice and the prepared wine were determined according to AOAC (2005)  During the pH and TSS measurement, 10 ml of fruit juice and wine samples were analyzed.

| Total acidity
The Total acidity (TA) was determined according to a standard method by AOAC (2005). The samples were dissolved in 100 ml distilled water and titrated against 0.1 M NaOH solution. The production of pink color was recorded as the endpoint using phenolphthalein as an indicator.

| Extraction of phenolic compounds
Dried U. kirkiana fruit pulp and plant material of the L. javanica leaves were ground into a powder and sieved through a < 500 microns sieve into powder. Extraction of the soluble phenolic compounds was done using methanol: acetone: water (7:7:1 v/v/v), and the supernatant (extract) was stored at 4°C for analysis.

| Total phenol
Total phenolic content was determined using a method described by Harbone (1998) and Makkar (1999). A 50 µl sample of extracts, fruit pulp extract, and wine was diluted with 1L distilled water. Sodium carbonate (2.5 ml) and 1 N Folin-C reagent (500 µl) and were then added, and the mixture was incubated at 25°C for 40 min. The absorbance of the mixture was recorded at 725 nm using a Spectronic Genesys spectrophotometer (Genesys 5, Thermo Fisher Scientific) against a methanol blank. A calibration equation (y = 0.0013x + 0.0177; R 2 = .9982) was produced using standard gallic acid (Merck) and was used to calculate the total phenolic content in micrograms of gallic acid equivalence (GAE) as (µg GAE/ml Dry weight (DW) of the sample.

| Tannin binding assay
Tannins were determined following the method adapted from Makkar (1999). One gram of polyvinylpolypyrrolidone (PVPL) was dissolved in 1 ml distilled water. A 1 ml sample was then added, and this mixture was vortexed and incubated for 15 min at 4°C. The mixture was then centrifuged at 1,107 g using a Microyn Digital Bench-top Centrifuge. The total content of phenolic compounds in the supernatant was determined by measuring the absorbance at 725 nm using a Spectronic Genesys spectrophotometer (Genesys 5, Thermo Fisher Scientific). The tannin content was calculated as follows: where A 1 = total content of phenolic compounds before binding with PVPL and A 2 = total content of phenolic compounds after binding with PVPL.

| Flavonoid content
Flavonoid content was determined using the vanillin assay (AOAC, 2005), where 0.5 ml of the sample was measured and distilled water was added to make 1 ml in a test tube. The mixture was mixed with 1:1 v/v methanol-HCl (2.5 ml) and then 0.5 g/25 ml vanillin reagent (2.5 ml) was added. The mixture was then vortexed and left to stand for 20 min. The absorbance was done at 500 nm using a Spectronic Genesys spectrophotometer (Genesys 5, Thermo Fisher Scientific) and 50% methanol as a blank. The total flavonoid (proanthocyanidins) content was calculated and expressed as catechin equivalent (CE) per g dry weight.
2.7.7 | Antioxidant activity (DPPH free radical scavenging) The radical scavenging activity was determined according to the AOAC method (AOAC, 2005). In the original experiment, absorbance was done at 515 nm. A 3.9 ml solution of 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical solution (0.06 mm) was mixed to 100 μl sample and incubated in the dark for 1 hr at 25°C. Ascorbic acid was used as a reference compound. The absorbance was determined at 517 nm using a Spectronic Genesys spectrophotometer (Genesys 5, Thermo Fisher Scientific) after calibration with methanol. The DPPH radical scavenging activity was expressed in terms of EC50 (effective concentration at which the DPPH radical was scavenged by 50%).

| Color of the wine
The color of the prepared fruit wine was measured using a UV-Vis spectrophotometer (Genesys 10S, Thermo Scientific) with 1 cm path length cuvettes at absorbances 420 nm, 520 nm, and 620 nm following a method described by Yildirim (2006). Color density (A 420nm + A 520nm ), color intensity (IC) (A 420nm + A 520nm + A 620nm ), and shade (A 420nm /A 520nm ) were used to describe the color of the wine using distilled water as a blank. The levels of blue (% B) red (% R), and yellow (% Y) were determined as A 620nm × 100/IC, A 520nm × 100/IC, and A 420nm × 100/IC, respectively.

| Free sulfur dioxide
The free sulfur dioxide was determined following a method de- where 1.05 g is the amount of carbon dioxide produced for every gram of ethanol produced and 0.79 g/ml is the density of ethanol alcohol.

| Sensory evaluation
In the sensory evaluation process, 30 semitrained panelists were and were also instructed not to discuss their results during the sensory evaluation process.

| Data analysis
All results were expressed as mean ± SD, and analysis was carried out using Statistical Package for the Social Sciences (SPSS) package  such as calcium, potassium, phosphorus, copper, iron, magnesium, and zinc (Mahlangeni et al., 2018). The presence of these essential minerals in the leaves has made the plant important in human nutrition as herbal tea and food additive (Maroyi, 2017). More so, the significant concentrations of these minerals could enhance the nutritional and curative properties of L. javanica. L. javanica had a higher radical scavenging activity of 64.2 ± 0.1 mg/L than U. kirkiana (35.1 ± 0.1 mg/L) (Table 1). This could be explained by the presence of higher total phenolic content in comparison with U. kirkiana fruit which is a popular indigenous fruit tree that is consumed by many people in Zimbabwe (Chawafambira et al., 2020). Additionally, Shikanga et al. (2010) reported a higher phenolic content of 14.8 mg GAE/g DW in leaf extracts of L. javanica.

| Physiochemical properties of fruit juice, Lippia javanica leaf extract, and wines
The pH, °Brix, and total acidity of fruit juice, herbal infusion, and the prepared wines are shown in Table 2. The fruit juice and L. javanica leave extract had pH of 4.4 ± 0.05 and 5.8 ± 0.1, respectively. These pH values indicated that further pH adjustment was required before they were used as substrates for wine fermentation. This resulted in fresh lemon being added in this experiment to adjust to a pH of 3.4. The low pH observed in the wine could also be attributed to the physiological action of yeasts, lactic acid bacteria, and organic acids that are produced during alcoholic fermentation (Tsegay & Gebremedhin, 2019). The pH differences between the wine with herbal infusions and control wine sample were significantly different at p < .05.
The total soluble solids (TSS) of L. javanica extract and U. kirkiana fruit juice were 4.1 ± 1.01% °Brix and 22.3 ± 0.12% °Brix, respectively. Katsvanga et al. (2007) noted that fruits that grow in humid conditions with warm nights often have higher TSS levels and lower fruit acidity, which could explain the observed results on total soluble solids and TA on U. kirkiana fruit juice used in this study. The fruits were collected from a semiarid and hot area with maximum temperatures >29°C. Also, the TSS could be due to the effect of solar radiation during fruit ripening stage, which promotes the increase in the conversion of starch to sucrose, and reduction in sugars (Alston, 1992 This was due to the solubilization of dissolved solids such as sugars as they were broken down via a series of biochemical reactions to produce alcohol, organic acids, and carbon dioxide (Tsegay & Gebremedhin, 2019).
The total acidity was lower in L. javanica extracts (0.12 ± 0.10) but increased to 0.31 ± 0.01 in the prepared wine. The high acid content in the U. kirkiana fruit could be attributed to the presence of organic acids (Chen et al., 2009). Vertuani et al. (2002) supported the explanation and reported the occurrence of tartaric, citric, malic, succinic, and ascorbic acid in ripe fruit pulps. The organic acid composition is determined by the fruit species, climate and geomorphological property of soil (Velić et al., 2018). As a result, the acidity of fruit used in winemaking affects the acidity of the fruit wine. The wine with L. javanica extract had a lower total titratable acidity of 5.9 g/L when compared to wines produced from orange fruit (6.3 g/L) and kiwi fruit wine (13.6 g/L) (Kelebek et al., 2009;Soufleros et al., 2001). This could be attributed to the differences in concentrations of organic acids especially citric acid which is present in most citrus fruits. Velić et al. (2018) reported the presence of malic acid as a major organic acid present in apple (6.2 g/L), blackberry (3.5 g/L), and cherry (6.8 g/L) wines.
The L. javanica leaves had higher antioxidant activity than the fruit as shown in Table 1. The free radical scavenging activity of L. javanica can be explained by the presence of phenolic compounds which contain reactive properties in their chemical structures for free radical scavenging activity (Chanda andDave, 2009). Muchuweti et al. (2006) reported that ethanolic leaf extract of L. javanica had a 74.4% inhibition of the DPPH radical. Also, Bhebhe et al. (2016) found similar activity in L. javanica leaves.

| Alcohol, color, sulfur dioxide, and phenolic content of wine
The total phenolic composition of the fruit, L. javanica, and the produced wines is shown in Table 3 (Critchfield et al., 1996) and luteolin 12 has been found to fight against herpes simplex virus in the presence of another flavonoid kaempferol (Kumar & Pandey, 2013).
Also, the increase in total phenol in the wine could be attributed to the presence of nonphenolic compounds in the fermenting substrate, that is, the bound soluble conjugated or insoluble phenolic acids in the cell walls of raw materials which could be converted into phenolic compounds by the action of esterase enzyme produced by the yeast (Tchabo et al., 2017). This improvement in phenolic content in the produced wine with L. javanica is in agreement with other researched done on the fermentation of apple-herb wine (Lee et al., 2013).
The color properties of the produced wine are shown in Table 3.
Results on wine color indicate that shade/brightness of the wine is higher than color intensity and color density. The observed color in the wine might have been produced from the reactions of anthocyanins and phenolic compounds present in the L. javanica extract during fermentation as well as during the wine aging (Tsegay and Gebremedhin (2019)). The yellow color of the U. kirkiana fruit juice and the added light brown color of L. javanica extract could have produced the yellowish color of the wine during the fermentation process (Table 3).
Fruit juice substrates with low sugar content produce wine with lower alcohol content after fermentation. In the study, the sugar content was adjusted by adding sucrose to the must. The produced wine had alcohol (ABV %) of 10.2 (Table 3). This could be explained by the production of ethanol during the fermentation process. In wine production, ethanol is the main alcohol produced and it normally determines the alcohol content. The alcohol content was higher than wines fermented from apple tea substrate (7.9 to 8.19% v/v) (Kumar et al., 2016). The differences in the composition of the substrate and method of the fermentation process used could have affected the alcohol content of the wines.
Sulfur dioxide is a common preservative still being used in wine production as an antioxidant and antimicrobial (Velić et al., 2018).
Free sulfur dioxide content was 58. TA B L E 3 Alcohol, color, and sulfur dioxide (SO 2 ) content of wine free SO 2 in the prepared wine was higher than the wine produced from orange fruit juice (8.2 mg/L) (Kelebek et al., 2009)  inhibit the growth of some wild yeast species and bacteria related to wine spoilage. Sulfur dioxide is also important in inhibiting the activity of polyphenol oxidase during wine production (Guerrero & Cantos-Villar, 2015). As a result, oxidative and unwanted spontaneous fermentation processes are both controlled by the action of sulfur dioxide (Ribereau-Gayon et al., 2006). Additionally, sulfur dioxide reduces the rate of phenolic polymerization and loss of color during wine aging (Santos et al., 2012). However, consumption of SO 2 in wines can have adverse health effects such as abdominal pain and diarrhea (Ferrer-Gallego et al., 2017). It was therefore important in this study to regulate the addition of potassium thiosulfate (70 mg/L), a source of SO 2 in the wine.
The antioxidant activity was 72.1 ± 0.1 and 46.3 ± 0.2 in wine with added L. javanica extract and control, respectively. This could be attributed to the presence of dietary phytonutrients such as phenolics and vitamins. Shikanga et al. (2010) reported an antioxidant activity with an EC 50 value of 358 μg/mL for L. javanica leaf infusions and a total phenolic content of 14.8 mg GEA/ml of dry weight. More so, the observed high antioxidant activity in the L. javanica leaf extract and subsequently in the produced wine could be attributed to the high levels of verbascoside 2 (1.5 mg/g dry weight) found in the leaves of L. javanica (Olivier et al., 2010).

| Sensory evaluation
Demographic characteristics have shown that men were the majority of participants that participated in the sensory evaluation (Tables 4). The majority of the sensory evaluation participants were aged 30-39 (n = 13). ists rated flavor and color of the produced wine with added L. javanica extracts as "good" compared control wine. The taste of the wine with added L. javanica extracts was rated as "average." The control wine was rated "average" for flavor and color.
The mean scores for the sensory attributes of the produced wine with added L. javanica extracts and control wine samples are presented in of the wine might have been influenced by the low residual sugar content, titratable acidity, and pH of the produced wine. This was supported by Tsegay and Gebremedhin (2019)
Furthermore, the chemical and phenolic composition in the wine was significantly different (p < .05). More so, the alcohol (10.2 ± 0.1 ABV %,), free sulfur dioxide (58.1 ± 1.2 mg/L), antioxidant activity (73.1 ± 0.1 EC 50 mg/L), and color intensity (40.4 ± 2.1 yellowish) were significantly different (p < .05) from control wine. This study noted the importance of the addition of L. javanica extracts as it complements vitamin C and flavonoids in the produced wine. The addition of L. javanica extract was able to improve the sensorial attributes of the wine. Also, it was noted that the wine had acceptable levels of sulfur dioxide that do not cause adverse health effects when consumed. Future research must evaluate the volatile acids, organic acids, and cytotoxicity analysis of the L. javanica extracts and the prepared wine.

ACK N OWLED G M ENTS
The author would like to thank the Chinhoyi University of Technology, Department of Food Science and Technology and Standards Association of Zimbabwe, for assistance with their laboratories. The participation by panelists is greatly appreciated.

CO N FLI C T S O F I NTE R E S T
The authors declare there is no conflict of interest.

E TH I C A L S TATEM ENT
This study does not involve any human or animal testing.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data of this research are available and will be provided on request by the authors. TA B L E 6 Mean scores for the sensory evaluation of wine with Lippia javanica and control wine samples (n = 30)