Optimization of Edible Coating Formulation Using Response Surface Methodology for Delaying the Ripening and Preserving Tomato ( Solanum lycopersicum ) Fruits

,


Introduction
Edible coatings are thin layers of edible component, which are applied to the surface of foods either by dipping or brushing, in addition to or as a replacement for the natural protective coating [1][2][3][4]. Tey are low-cost, derived from renewable sources, biodegradable, and have specifc gas transmission properties. Edible coatings are mainly composed of polysaccharides such as starch and Gum Arabic which can be applied to provide a selective barrier to oxygen, carbon dioxide, and moisture, thus preserving fruits and vegetables from rapid spoiling [5].
Several studies have shown the efectiveness of these components in prolonging postharvest fruit shelf life and maintaining fruit quality [6][7][8]. Gum Arabic and starch are hydrocolloids and hydrophilic polymers with hydroxyl groups in their molecular structure. Tese hydroxyl groups could help scavenge free radicals during the browning process. To date, these hydrocolloids (Gum Arabic and starch) have been widely used in coating formulations for fruits preservation [9]. Te efectiveness of these compounds for preservation could be further booster by combining them with bioactive substances from agricultural byproducts [10].
Samuchaya et al. [11] reported that dried cofee leaves powder (Cofea arabica), as a by-product, has high polyphenol and procyanidins contents and high free radical scavenging capacity. Cofee leaves have been confrmed to be a food source for humans, owing to its richness in phenolic compound and its high antioxidant activity. Tus, cofee leaves could be used as a component of edible coating formulations, in addition to starch and Gum Arabic, to prolong the shelf life of fruits.
Tomato (Solanum lycopersicum L.) is a plant of the Solanaceae family. It is cultivated for its fruits commonly known as tomatoes which are consumed worldwide [12]. Tomatoes have a high nutritional value owing to their richness in vitamins, minerals, natural antioxidant compounds, and amino acids. Several other healthpromoting substances have been found in tomatoes, such as carotenoids, folic acid, ascorbic acid, lycopene, and β-carotene, which have been correlated with reduced risk of cancer and some heart diseases in humans [14]. Tomato is a fruit with a short shelf life (ranging from one to two weeks) [15]. It is classifed as a climacteric and highly perishable fruit with a high respiratory peak associated with a high rate of ethylene production at postharvest [16]. Low-temperature storage is inappropriate for tomatoes because this technique of conservation leads to the loss of some quality parameters such as color, frmness, favor, and appearance, all of which afect the commercial interest of the fruit. Previous studies reported that at room temperature, the shelf life of tomato fruits can be extended from 19 to 30 days by applying diferent coating formulations notably, the edible coating based on Gum Arabic [8,16], cassava starch [17], pectin and chitosan [18], and ethanol [19]. Te experimental results suggested that coating formulations should be sought to further extend the shelf life of tomatoes. To this end, the combination of cofee leaf extract, starch, and Gum Arabic in a formulation could better preserve tomato fruits. Te response surface methodology (RSM) has been proven to be a powerful tool to determine the efects of each factor and the interactions between them, which allow the optimization of the processes [20]. Terefore, the objective of this research study was to use the response surface methodology to determine the optimal formulation of edible coating based on cofee leaf extract, Gum Arabic, and starch, in order to increase the shelf life and preserve the quality of tomato fruits.

Plant Material.
Tomato fruits of Rio Grande variety were harvested at the "turning stage" in an experimental farm located in Dschang (5°10′-5°38′N, 9°50′-10°20′E, altitude 1400 m) and immediately transported to the Laboratory of Applied Botany of the University of Dschang, Cameroon, for the experiments. Te "turning stage" is defned as the developmental stage at which a yellowish ring is visible at the distal end of the fruit [21]. Only frm, diseasefree, and uninjured fruits of the same size were selected for experimentation.
Te cofee leaves used in this study were harvested from cofee plants in the locality of Foumbot, a district located in the western Cameroon (5°30′0″-5°30′53″N, 10°37′60″-10°60′50″E). Tese leaves were washed with water and air-dried in the shade to a constant weight. Te Gum Arabic used was purchased from a local market in the city of Garoua (Cameroon). Te starch used for the study was extracted from freshly harvested cassava tubers, following the method described by Belibi [22].

Reagents.
All the chemicals which were used for the study were from Sigma-Aldrich (Germany), with purity > 98%.

Preparation of the Cofee Leaf Extract.
Using a stainless steel Wiley mill, the dried cofee leaves were ground into powder to pass through a 60-mesh screen and used to prepare the extracts for coatings, following the methodology described by Aghofack-Nguemezi et al. [23]. Diferent quantities of cofee leaves powder corresponding to the diferent concentrations tested were macerated in 1500 ml of water/ethanol mix at 1 : 2 (v/v) ratio. Each mixture was left to stand for 3 hours at room temperature and then fltered through a muslin cloth. Te residue from the fltration was then macerated in 750 ml of the same solvent (water/ethanol at 1 : 2 v/v ratio) for one hour after which the mixture was fltered. Te two fltrates were mixed and used as cofee leaf extract (CLE). Te volume of each cofee leave extract was thus 2250 ml.

Coating Preparation.
Te Gum Arabic coating was prepared following the method described by Khaliq et al. [24]. Diferent weights of gum Arabic powder, corresponding to the diferent concentrations tested, were dissolved in 750 ml of distilled water for 60 minutes and then the mixtures were fltered. After fltration, diferent quantities of starch which were needed to have the diferent concentrations tested were added to the fltrate, and then the mixtures were heated to a temperature of 100°C while homogenizing, in order to have the Gum Arabic/starch solutions [20]. After these operations, diferent cofee leave extracts were mixed with the corresponding Gum Arabic/ starch solution to obtain the desired combinations of cofee leave/Gum Arabic/starch concentrations. Te total volume of each coating solution was thus 3000 ml consisting of a mixture water/ethanol (1 : 1, v/v). Bleach was added to each solution at the rate of 230 µl/l in order to disinfect the mixture.

Treatment of the Fruits.
Tomato fruits were thoroughly washed with distilled water and soaked for 25 minutes in the diferent coating solutions T1 to T20. Te control fruits were soaked in distilled water to which bleach was added at the rate of 230 µl/l. Tree replicates each of 8 fruits were used per treatment.

Experimental Plan.
Te response surface methodology (RSM) was performed using Minitab version 2018 software package. Te central composite design (CCD) was used to determine the efects of three independent variables (cofee leaf extract concentration, starch concentration, and Gum Arabic concentration) on the dependent variables (responses) which were percentage of ripening (PR), frmness, pH, physiological loss of mass (PLM), and pigments (Chl a, Chl b, β-carotene and lycopene) contents. Te choice of factors and their levels of variation were made considering data from previous studies [8,17,25]. Te diferent factors and levels (minimum, center, and maximum) are shown in Table 1. Te experimental design generated 20 edible coating formulations from the central composite design in which there were 06 (six) replicates of the points at the center (Table 2) [26]. Te six center points show the repeatability of the method [27].

Empirical Model for the Prediction of Ripening
Parameters. Experimentally obtained data were subjected to multiple regression analysis. Te empirical model was developed by ftting the experimental data obtained from the central composite design device into a second-order polynomial mathematical equation (1). In order to determine the relationship between the independent variables (factors) and the response variables from the mathematical model equation, a second-order polynomial function was developed for each response. Te 3D curves were constructed with two independent variables while keeping the 3rd variable constant.

Evaluation of the Efect of Coating Solutions.
Te efect of the diferent coating solutions was evaluated on the 14th day after treatment (date at which the control fruits started their senescence). Te parameters investigated included the percentage of ripening (PR), frmness, physiological loss of mass (PLM), pH, and pigments (chlorophyll a, chlorophyll b, β-carotene, and lycopene) contents.

Determination of the Percentage of Ripening.
A fruit was considered ripe when it was fully red. Te percentage of ripening at day 14 was calculated according to the following formula: PR � number of ripe fruits total number of fruits × 100, with PR � percentage of ripening.

Measurement of the Firmness.
A penetrometer (GY-2, SAUTER GmbH, Germany) was used to measure the frmness of the tomato pulp. Te tomato epicarp was removed from three locations on the fruit using a razor blade. Te penetrometer was zeroed and the tip head placed on the pulp of the peeled area of the fruit. A cylindrical probe (diameter: 4 mm) of convex type was used to perforate the three previously peeled areas of the tomato. Continuous downward pressure was applied so that the tip sank into the fruit pulp to the depth marked (halfway up) on the tip. Te tip was then removed and the value indicated on the dial of the penetrometer noted. Te frmness values were expressed in Newton (N).

Measurement of the Physiological Loss of Mass.
Tomatoes were weighed using balance (METTLER, PB603-S, Germany) on day 1 to obtain the initial mass and then on day 14 to obtain the fnal mass. Te physiological loss of mass was calculated using the following formula by Athmaselvi et al. [29]: with PLM � Physiological loss of mass.

pH Measurement.
A pH meter (ATC, Lutron PH-221, Taiwan) was used to measure the pH of tomato juice. Te tomato fruit juice was prepared as follows: 15 g of tomatoes were crushed and the crushed material was fltered through muslin cloth into a beaker. Te pH meter probe was then inserted into the juice, and the pH value was directly read on the screen. Tree measurements were made for each treatment.

Determination of Pigments' Contents.
Te determination of pigments' contents was carried out according to the method described by Nagata and Yamashita [30]. Pigments were extracted by grinding 5 g of fresh material in 10 ml of solvent (acetone/hexane in a 4/6 (v/v) ratio). Te mixture was stored at 4°C for 24 hours. Te mixture was fltered through a Whatman paper N°1. Te absorbance of hexanolic extracts was measured at 663, 645, 505, and 453 nm wavelengths using a UV-visible spectrophotometer (BIOBASE, BK-UV1800, China). Te concentrations of the diferent pigments were determined by the following equations: A 663 , A 645 , A 505 , and A 453 are the absorbances at 663 nm, 645 nm, 505 nm, and 453 nm, respectively.
2.6. Analysis of the Data. Statistical analysis of data was carried out using Minitab version 18 software package. Te response surface methodology (RSM) was used for the generation of response surface and for the optimization of process variables. Determination of the coefcients (α) of the diferent models was done by a matrix approach using multiple linear regression [31]. A total of twenty runs were provided by central composite design by RSM. Probability value (model signifcance) was used to assess the quality of the model on the one hand, and the calculation of coefcients of determination (R 2 ) that measure the ftness of the regression model on the other hand [32]. Te regression analysis was used to determine the efects of variables in fstorder, two-factor interactions, and second-order polynomial models [33]. Indeed, when the coefcient of determination (R 2 ) is close to 1 (R 2 ≥ 0.8), the degree of correlation is high between the observed and predicted values, highlighting a reasonable agreement of the model with the experimental results [34]. Variability of terms in the regression equation for each response was determined using analysis of variance (ANOVA). Te signifcance of the linear, quadratic, and interaction efects of the diferent factors; as well as that of each of these coefcients was determined by comparing the observed probability (p value) to a critical probability (p � 0.05).

Optimization and Validation.
To determine the optimal values of the independent variables (cofee leaf extract, starch, and Gum Arabic concentrations), the graphical and numerical optimizations were performed [35]. Te three-    Journal of Food Quality dimension (3D) response surface plots were generated from the ftted model for each dependent variable in order to better observe the interaction efects of cofee leaf extract, starch, and Gum Arabic concentrations on the responses. Te numerical optimization of the response(s) was carried out using the desirability function approach described by Derringer and Suich [36], that is, to have the maximum values of chlorophylls contents and fruit frmness and minimum values of percentage ripening, physiological loss of mass, pH, lycopene, and β-carotene contents. Tukey's comparison test was performed between the predicted and experimental response values to check the adequacy of the fnal model response. Table 3 shows that except the cofee leaf extract concentration, each of the factors had signifcant infuence on responses. Te interaction term between X 1 and X 2 , X 1 and X 3 infuenced the ripening parameters studied. Furthermore, the square (quadratic) terms of all these factors signifcantly infuenced the diferent responses. In addition, the coefcients of determination R 2 showed values all above 80%. Tis would mean that the generated polynomial models are adequate to explain the efects of cofee leaf extract, starch, and Gum Arabic concentrations on the diferent measured parameters [37]. In other words, there is a good agreement between the data experimentally obtained and the data predicted by the software.

Efects of Edible Coating on Chlorophyll a and Chlorophyll b Contents.
Chlorophyll a and chlorophyll b concentrations were higher in treated fruit than in control fruit. Te highest chl a content (0.058 mg/g) was recorded with essays number 5 and 9, while essay number 3 resulted in the highest chl b content (0.034 mg/g) ( Table 2). Figure 1 shows that tomatoes treated with 66.66 g/l leaf extract showed the highest chl a content. On the other hand, chl a content increased with increasing starch concentration. Figure 2 indicates that chl b levels were highest in tomatoes treated with cofee leaf extract and Gum Arabic at the concentrations of 66.66 g/l and 10%, respectively. Each of the three factors signifcantly infuenced chlorophyll a content, while chlorophyll b was only signifcantly infuenced by starch and Gum Arabic concentrations (Table 3). However, the quadratic efect of each factor (X 1 2 , X 2 2 , and X 3 2 ) signifcantly infuenced these two pigments contents. Meanwhile, X 1 X 2 and X 1 X 3 interactions signifcantly infuenced the chl a and b contents. Te coefcients of determination (R 2 ) were greater than 94% for the two pigments and the lack of ft were 0.148 and 0.02 for chl a and chl b, respectively, indicating a well ftted response models. Color change from green to red is an indicator of tomato ripening [38]. Tis change is associated with the degradation of chlorophylls, followed by biosynthesis of carotenoids such as xanthophylls and carotenes [39]. Tis degradation normally occurs during ripening of tomato fruit when stored under ordinary conditions (without treatment and at room temperature), but this degradation can be delayed when the fruit is stored at low temperature or when it is coated and stored. Te coating may have slowed the activity of chlorophyllase, an enzyme that catalyzes the degradation of chlorophylls. Indeed, in fruits treated with cytokine (benzyl-aminopurin), Wang et al. [40] reported a slowdown of chlorophyll degradation as a response to a decrease in chlorophyllase activity. Takamiya et al. [41] showed that the degradation pathway of chlorophyll includes dephytylation, i.e., the removal of the magnesium atom from tetrapyrrole macrocycle. Te dephytylation involves chlarophyllase activity [42]. Te chlorophyll a and b content of the control fruit was very low compared to that of fruit from all treated lots. Tis result is similar to that obtained by Aghofack et al. [23] who showed that coating with cocoa leaf extracts delays the degradation chlorophyll a in tomato fruit. Similar drop of chlorophyll a degradation were reported by Donjio et al. [10] with tomatoes coated with pineapple peel extract and Gum Arabic. During chlorophyll pigment degradation, chl b degrades faster than chl a [43]. Te results of this study confrm this fact as it was observed that chl a was always higher than chl b. Tis shows that the degradation of chl a and chl b are carried out diferently. Indeed, chl b is destroyed either directly by reductase [44] or by conversion to chl a before its degradation [43].

Efects of Edible Coating on Lycopene and β-Carotene
Contents. Lycopene levels ranged from 0.044 to 0.144 mg/g for all coated fruits after 14 days of storage. In the control fruits, the lycopene content was 0.189 mg/g. For β-carotene content, values ranged from 0.043 to 0.109 mg/g in coated fruit, while the value in uncoated fruits was 0.091 mg/g. Lycopene and β-carotene levels were higher in control fruit compared to treated fruit, except the fruits treated with formulations 5, 9, and 12, in which β-carotene levels were slightly higher than in control fruits ( Table 2). Only the starch and Gum Arabic concentrations had signifcant linear efect on these two responses during ripening (Table 3). Te quadratic terms of the three factors signifcantly infuenced the β-carotene content. Figure 3 shows that lycopene content decreases with increasing starch concentration. Te lowest lycopene content was obtained at 10% Gum Arabic. Figure 4 shows that β-carotene content was lowest when cofee leaf extract and starch concentrations were around 66.66 g/l and 40 g/l, respectively. Te lack of ft was signifcant (p < 0.01) for lycopene and not signifcant (p � 0.098) for β-carotene, but the coefcients of determination (R 2 ) were well enough for well-ftted models (Table 3). Lycopene is a carotenoid associated to color change during ripening [45,46]. Previous studies have shown that in ripening fruits chlorophyll degradation and lycopene accumulation are positively correlated [47]. Te present study shows that lycopene and β-carotene contents were lower in treated fruits compared to control fruits, indicating that the coating solutions formulated with cofee leaf extract, starch, and Gum Arabic slowed down the degradation of chlorophyll, and the synthesis of carotenoids as well. Tis is in agreement with Daisy et al. [45] Journal of Food Quality who reported that Gum Arabic-based coating has good ability to delay carotenoid synthesis during fruit ripening.

Efect of Edible Coating on pH.
Tere was very little variation in pH among treatments. Indeed, pH ranged from 4.2 to 4.36 for treated fruit and was 4.26 for untreated fruit. Te results of this work are consistent with those of Adjouman et al. [17] who recorded pH values ranging between 4.29 and 4.31 at 21 days after coating tomato fruits Table 3: Regression coefcients, coefcients of determination (R 2 ), and p value of lack of ft of predicted equations.

Sources
Constant X 1 , X 2 , and X 3 are the linear efect terms; X 1 X 2, X 1 X 3 , and X 2 X 3 are the interaction efect terms; and X 1 2 , X 2 2 , and X 3 2 are the quadratic efect terms; signifcance of model terms at 5%, 1%, and 0.1% are, respectively, indicated by " * ", " * * ", and " * * * ".    with cassava starch. However, Table 3 shows that the efect of linear terms of cofee leaf extract and Gum Arabic on pH was signifcant (p < 0.01). Similarly, the interaction between cofee leaf extract and starch concentrations as well the interaction between cofee leaf extract and Gum Arabic concentrations had signifcant efects (p < 0.05) on fruits' pH. Te quadratic term of starch concentration also signifcantly infuenced fruits' pH. As shown in Figure 5, the pH of tomatoes was lowest when the concentration of cofee leaf extract was close to 66.66 g/l. It decreased with increasing concentration of Gum Arabic. Te lack of ft was signifcant (p < 0.01), but the coefcient of determination (R 2 ) was above 89%, therefore assuring that model is well ftted. Te combined coating with cofee leaf extract, starch, and Gum Arabic slightly increased the pH of tomatoes. A similar increasing trend in pH values of the fenugreek galactomannan and guar galactomannan coated guavas throughout the storage period has been reported in a previous study [48]. Tis variation may be due to changes in titratable acidity which in turn may be attributed to increased citric acid glyoxylase activity during ripening. A reduction in acid content may be due to their conversion into sugar [49]. Indeed, the increase in pH during ripening is attributed to the loss of citric acid [49,50].

Efects of Edible Coating on Physiological Loss of Mass.
Te control fruits had a higher mass loss (7.47%) compared to all treated fruit lots, whose physiological loss of mass ranged between 5.08 and 7.27% (Table 2). Te lowest mass loss (5.08%) was recorded with the coating solution containing 66.66 g/l cofee leaf extract, 65.23 g/l starch and 10% GA. Te physiological loss of mass was signifcantly infuenced by the linear term of starch concentration (p < 0.05). On the other hand, the interaction between starch and Gum Arabic concentrations, as well as the quadratic terms of cofee leaf extract and starch concentrations signifcantly infuenced this response (p < 0.05 and p < 0.01, respectively) ( Table 3). Te lack of ft was not signifcant (p � 0.093) and the coefcient of determination (R 2 ) was more than 84%, indicating a well ftted response model. Figure 6 shows that physiological loss of mass decreased with increasing starch concentration while it increased with increasing Gum Arabic. Compared to the results obtained with control fruits, coating with a combination of cofee leaf extract, starch, and Gum Arabic resulted in a decrease of the physiological loss of mass of tomato fruits. Similar increase of physiological loss of mass in response to increasing the concentration of Gum Arabic in coating solution has previously been reported by Sanchita et Hari [25]. Ali et al. [8] showed that edible coating formulation based on high concentration of Gum Arabic (15%) results in excessive mass loss of tomatoes. It follows from the results obtained that the efectiveness of Gum Arabic in controlling physiological weight loss depends on its concentration. Because at high concentrations, the physiological loss of mass increases. Evapotranspiration and respiration are the physiological phenomena that best explain the mass loss of fresh fruits during their ripening process [51]. Previous studies have shown that starch used as a coating material acted as a semipermeable barrier against gas (CO 2 and O 2 ) and moisture exchange in tomatoes, thus reducing respiration rate, mass loss, and oxidation [52,53]. Water exchange between the indoor and outdoor environments is considered the main cause of postharvest mass loss of fruits. Gum Arabic behaves as a hydrophilic polysaccharide with the polar -OH groups and C-O covalent bond in their structure. However, the covalent bonds between starch and Gum Arabic could play an important role in forming a stabilizing coating macromolecule to improve their barrier property. Terefore, the coating solution containing medium concentrations would have decreased gas exchange thus slowing down the physiological loss of mass of tomato fruits.

Efects of Edible Coating on Firmness.
Te frmness of the control fruits was signifcantly lower (25.95 N) compared to that of the treated fruits, whose values ranged from 33.70 to 39.22 N. Table 3 shows that two linear terms (starch concentration and Gum Arabic concentration) signifcantly (p < 0.01) infuenced frmness, as did the interaction between cofee leaf extract concentration and Gum Arabic concentration. Te quadratic term of starch concentration also infuenced frmness. Figure 7 indicates that fruit frmness increased as the concentration of cofee leaf extract increased and as the concentration of Gum Arabic decreased. Tus, the highest frmness was recorded when leaf extract concentration above 66.66 g/l was associated with low Gum Arabic concentrations. Te lack of ft was not signifcant (p � 0.22) and the coefcient of determination (R 2 ) was more than 85%, indicating a well ftted response model. Compared to the control treatment, the coating formulations based on a combination of cofee leaf extract, starch, and Gum Arabic, slowed down the decrease in frmness of the tomato fruits. Similar results of reduction in frmness loss were reported by Hernandez-Guerrero et al. [54] with mangoes coated with 2% (w/v) starch. Firmness is an important indicator to determine the degree of ripeness of the fruits. Te ripening of the fruits leads to their softening. Degradation of cell structure, cell wall composition, and intracellular compounds lead to the softening of the fruit [55]. Since the frmness of treated fruits was high compared to control fruits, this indicates that the coating slowed down the degradation of cell wall compounds through hydrolysis of pectin. Hydrolysis also leads to softening by converting starch into sugar [56]. Tis biochemical process is due to the action of hydrolases, which are the enzymes that hydrolyze pectin and starch [8].  Table 3 indicates that the linear terms of cofee leaf extract and Gum Arabic concentration signifcantly (p < 0.05 and p < 0.01, respectively) infuenced the ripening rate. Similarly, cofee leaf extract and starch, as well as cofee leaf extract and Gum Arabic interactions also infuenced this response. Only the quadratic term of starch concentration did not have a signifcant (p > 0.05) infuence on the response. Te percentage of ripening was low when the cofee leaf extract concentration was between 33.33 and 66.66 g/l. It decreased with increasing starch concentration ( Figure 8). Te lack of ft was not signifcant (p � 0.690) and the coefcient of determination (R 2 ) was more than 92%, indicating a well ftted response model. Te coating solutions improved the shelf life of tomato fruits compared to uncoated fruits. Generally, the ripening process of tomatoes harvested at the turning stage lasts 5 to 6 days. Tis short green life severely limits its long distance marketing [57]. Te shelf life of the fruit is related to its ripening rate. Ripening encompasses a set of irreversible and unavoidable physiological processes that change the composition of the fruit [57]. Te edible coating has barrier characteristics that reduce the permeability of the fruit surface to oxygen and carbon dioxide, leading to a change in the internal gas composition that in turn reduces oxidative metabolism and increases the shelf life of the fruit [57]. Several works have shown that coating fruits with Gum Arabic would result in signifcant slower ripening rate and increased shelf life [6]. In addition, previous investigations by Donjio et al. [10] showed a signifcant efect of a combination of coatings to delaying the percentage of ripening of tomato fruits.

Optimization.
Chlorophyll a content signifcantly varied with cofee leaf extract, starch, and Gum Arabic concentrations. Te other parameters were mostly signifcantly infuenced by starch and gum concentrations (Table 3).   Optimization was carried out by maximizing chlorophylls (a and b) contents and frmness and minimizing lycopene and β-carotene contents, pH, physiological loss of mass, and percentage of ripening. Te resulting optimal concentrations obtained for cofee leaf extract, starch, and Gum Arabic were 78.5 g/l, 56 g/l, and 9.5%, respectively, with a desirability of 0.71. A new coating formulation based on 78.5 g/l, 56 g/l, and 9.5% cofee leaf extract, starch, and Gum Arabic, respectively, was applied to tomatoes under the same conditions. All parameters were measured on the 14th day after storage. Te experimental data obtained as a result of coating fruits with the optimal formulation as well as the predicted data are shown in Table 4. Te results of the comparative analysis between the experimental and predicted data show that there was no signifcant diference at 5% probability level (p < 0.05).

Conclusions
Te response surface methodology (RSM) has been successfully used to optimize an edible coating formulation based on cofee leaf extract, starch, and Gum Arabic to extend the shelf life or delay the ripening process of tomato fruits at ambient conditions. Te optimum concentration of cofee leaf extract, starch, and Gum Arabic was predicted to be 78.5 g/l, 56 g/l, and 9.5%, respectively, which had satisfactory thickness and was efective in slowing down ripening and associated parameter progressions. Te present study reports an edible coating formulation which is efective in extending shelf life, slowing the ripening process beyond 14 days compared to control fruits which ripened at 100% after 6 days of storage, with the onset of senescence at the 14 th day.

Data Availability
All data used to support the fndings of this study are from previously reported studies and datasets, which have been cited in this manuscript. Furthermore, the processed data will be provided upon request.