Antifungal Activity of Gelatin-Tapioca Starch Film and Coating Containing Copper Nanoparticles against Colletotrichum gloeosporioides Causing Anthracnose

-is study aimed to fabricate a nontoxic coating containing copper nanoparticles (CuNPs) to protect fruits from pathogenic Colletotrichum gloeosporioides causing anthracnose on several tropical fruits. We used a green approach, in which CuNPs were synthesized by reducing CuSO4 with ascorbic acid in the presence of gelatin and glycerol as the capping agents. -e formation of CuNPs was confirmed by UV-vis absorption spectra of the reaction mixture, which showed a surface plasmon resonance peak at 578–594 nm. -e x-ray diffraction spectrum of the CuNPs indicated the presence of mostly metallic copper with some minor impurities of Cu2O, CuO, and Cu(OH)2. Transmission electron microscopy (TEM) images and dynamic light scattering studies showed that the sizes of 90% of CuNPs were in 100–300 nm range. A 30–50 nm capping layer of gelatin surrounding CuNPs can be observed in the TEM images. Comparing FTIR spectra of the used reagents and CuNPs confirmed the depletion of ascorbic acid, as well as the gelatin layer protecting CuNPs. -e synthesized CuNPs showed dose-dependent antifungal activity against C. gloeosporioides with 100% growth inhibition at 200 ppm copper. Gelatinized tapioca starch was then added to the CuNPs solution to obtain a film-forming mixture to produce stand-alone composite films on Petri dishes and coatings on mangoes. C. gloeosporioides could not grow on the surface of nutrient agar in contact with the films containing 245 ppm CuNPs, while they grew normally on control films without CuNPs. For the in vivo antifungal tests on mangoes, both the control and the CuNPscontaining coatings equally inhibit fungal growth, possibly due to the low oxygen permeability of the protein and starch components in the films. -is study thus demonstrated the potential applications of composite coatings using biodegradable polymers that contain CuNPs in postharvest protecting fruits from phytopathogenic fungi.


Introduction
Mango (Mangifera indica L.) is one of the most popular tropical fruits due to its strong and attractive aroma, delicious taste, high nutritional value, and bright yellow colours of peel and flesh [1]. However, anthracnose caused by Colletotrichum gloeosporioides is one of the most severe diseases causing postharvest losses in mango during storage and transportation. Anthracnose can cause 30-60% damage, and sometimes up to 100% for mangoes produced under wet or humid conditions. Besides mango, anthracnose also contributes significantly to preharvest and postharvest losses of other crops, such as cashew, papaya, pomegranate, guava, and acid lime [2].
Many methods have been developed to fight anthracnose in fruits, but each of them has its strengths and weaknesses. Chemical control by direct spraying fungicides on fruits is highly effective but may result in resistance of the pathogens and high levels of residue [3]. Many natural essential oils are emerging as effective and safe biofungicides, but they are volatile, and their high contents may affect the intrinsic aroma and taste of the protected fruits [4]. Heat treatment by dipping fruits in hot water is a simple, nontoxic, and moderately effective method but can easily cause overheat damage to the fruits [5]. Gamma irradiation or biological antagonists have high cost and low effectiveness and require specific legislation [6,7]. Until now, a sequential or simultaneous combination of hot water and fungicide is the most effective commercial postharvest treatment for mango anthracnose.
Among the emerging techniques to fight anthracnoses in fruits, surface coatings incorporated with antimicrobial agents are a promising solution with several additional advantages: a glossy appearance of the fruit surface, reduced weight loss of fruit, delayed ripening, controlled release of active components, and extended storage life [8]. e fruit coatings can be natural polymers, such as chitosan, starch, gelatin, agar, alginate, or synthetic polymers, such as polyethylene, carboxymethyl cellulose, and copolymers [9]. e polymeric matrix of fruit coatings should desirably be nontoxic, biodegradable, transparent, and tunable in barrier properties against water vapour, CO 2 , oxygen, and ethylene.
It should be noted that although films and coatings are usually used as synonyms, they refer to different concepts according to their purpose and utilization. A film is a standalone thin layer of material that can be used as cover, wrap, or separation layer. On the other hand, coatings are films formed directly on the surface of the product that they are intended to be utilized with. Speaking another way, the coating is a part of the final product.
For antimicrobial coatings, antimicrobial agents are added to the packaging materials to avoid direct contact and control the release of the preservatives. Many antimicrobial agents were successfully incorporated into food coatings and exhibited strong inhibition of bacterial and fungal growth: commercial preservatives [10,11], fungicides [12], plant extracts [13], essential oils [14], and nanoparticles [15,16].
Metal nanoparticles possess unique properties like physicochemical, electrical, optical, and biological, which are much different from their bulk metals. Extensive studies have proved the antimicrobial activity of metal nanoparticles. is property is due to their small size and high surface-to-volume ratio, which enhance their interactions with a broad spectrum of microorganisms. Silver nanoparticles are the most investigated nanoparticles with high biocidal effects on bacteria, fungi, and viruses [17]. However, the high price of silver significantly hinders its application range. e low price and the high antimicrobial activity of copper nanoparticles (CuNPs) make them an attractive alternative to silver nanoparticles in antimicrobial packaging for the food industry [18]. Metallic copper and copper compounds have a long history of being used as antibacterial and antifungal agents in agriculture and medicine [19,20]. CuNPs synthesized by Ocimum sanctum leaf extract showed growth inhibition activity against 11 fungal phytopathogens, including C. gloeosporioides [21]. During its antifungal actions, CuNPs change the colour, morphology, form, texture, and density of the fungal mycelia [22]. Among the approaches for CuNPs synthesis, chemical reduction is the most common and convenient. In this approach, copper (I or II) salts are reduced by medium-to-strong reducing agents, such as NaBH 4 , hydrazine, and hypophosphite, which are usually toxic [23]. Moreover, to produce CuNPs, the reduction of copper ions must be carried out in the presence of capping molecules, which adsorb onto surface of the formed CuNPs and prevent them from growing further in size.
To be applied for fruit coatings, nontoxic substances must be used in the fabrication of CuNPs and active coatings. erefore, in this study, we used a green method of CuNPs synthesis: reducing CuSO 4 with ascorbic acid (AA) in the presence of gelatin and glycerol as dual capping agents and stabilizers for CuNPs. e one-pot technique was used to fabricate CuNPs-containing films and coatings: freshly synthesized CuNPs solution was combined with gelatinized tapioca starch to obtain a film-forming mixture. To evaluate the antifungal activity against C. gloeosporioides, CuNPs and the stand-alone films containing CuNPs were tested in vitro on Petri dishes, while the coatings were tested in vivo on mangoes.

Chemicals and Materials.
Gelatin, glycerol, copper sulfate, and ascorbic acid were purchased from Xilong Scientific LTD (China) with analytical grades. Tapioca starch and mangoes were purchased from a local supermarket in Ho Chi Minh City, Vietnam. e nutrient broth and agar were purchased from Himedia (India).

Preparation of CuNPs Solution.
CuNPs were synthesized based on a published method with some modifications [24]. An aqueous solution of 0.01 M CuSO 4 (50 mL) containing 4 g of gelatin and 3 g of glycerol was prepared and magnetically stirred for the complete dissolution of gelatin. After that, the solution was heated to 80°C and 50 mL of 0.05 M AA was added. e reaction mixture was kept at 80°C with continuous stirring for 90 min and then cooled to room temperature.

Formation of Composite Films Containing CuNPs.
Tapioca starch was dispersed in distilled water (5%, w/v) and gelatinized at 95°C for 30 min. e gel was cooled to 50°C and the CuNPs solution was added to obtain a mixture with 245 ppm of copper. e mixture was stirred for 1 h while cooling to room temperature. Every 7.0 mL of the filmforming mixture was poured onto a Petri dish (diameter 9.0 cm) and dried at 45°C for 48 h in a convection oven. e dried films were peeled off and conditioned in a chamber with a constant relative humidity of 80% (using a saturated solution of NaCl) at least 48 h before further experiments.
A control film was prepared with the same procedure, but the added CuNPs solution was replaced by the same volume of distilled water.

CuNPs Characterization.
Absorption UV-vis spectra of the CuNPs solution were recorded from 400 to 800 nm using a UH5300 spectrophotometer (Hitachi, Japan).
To obtain solid CuNPs for FTIR characterization, the CuNPs solution was vacuum-dried at 60°C for 6 h and the obtained solid was finely ground and mixed with KBr to a final KBr/solid ratio of 200. ATR-FTIR spectrum of the mixture was recorded from 4000 to 400 cm −1 with a resolution of 2 cm −1 on a Jasco FT/IR 4700 spectrometer (Germany).
For XRD characterization, CuNPs were centrifuged at 10,000 rpm for 30 min and washed twice with deionized water. XRD spectrum of CuNPs was recorded on a D8-Advance x-ray diffractometer (Bruker, Germany) with 2θ from 20°to 90°. e operation parameters were 40 kV potential, 40 mA current, and Cu-Kα wavelength of 1.5406Å. e size distribution of CuNPs in the solution was recorded using the dynamic light scattering (DLS) method on a HORIBA LA-920 (Japan).
Transmission electron microscopy (TEM) was used on a JEM-1400 (JEOL, Japan) to record the size and shape of CuNPs in solution at a voltage of 100 kV.
Scanning electron microscope JEOL 5410 LV (Japan) was used to observe the cross-sections and surfaces of the films on Petri dishes and the coatings on mango.

In Vitro Antifungal Activity of CuNPs Solution.
In vitro antifungal activity of CuNPs against C. gloeosporioides was evaluated based on a published method [22]. e CuNPs solution was mixed with 25 mL of melted nutrient broth (NB) agar medium to reach predetermined copper concentrations of 0, 50, 100, 150, and 200 ppm. e mixtures were poured onto Petri dishes and left cooling to room temperature. Mycelia of C. gloeosporioides were then put onto the centre of each Petri dish. e dishes infected with C. gloeosporioides were incubated at 30°C for 10 days.
e antifungal tests were also conducted for 3 control dishes containing each of the following instead of CuNPs: (i) 500 ppm CuSO 4 ; (ii) 2200 ppm AA; and (iii) 20% mango flesh (to confirm the growability of the fungus on mango). e size of each fungal colony was evaluated as the average of two perpendicular diameters. e percent inhibition of radial growth (IRG) of C. gloeosporioides was calculated as follows: where d 0 and d 1 are the average diameters of the colonies in the dish containing no copper and that containing CuNPs, respectively. All dishes were triplicated.

In Vitro Antifungal Activity of Composite Films
Containing CuNPs. One-week-old fungal mycelia of C. gloeosporioides were dispersed in 1.0 mL of sterile distilled water and then spread onto the surface of NB agar dishes. Films containing 245 ppm CuNPs were cut 3 × 3 cm and put on the agar surface infected with the fungus. e blank was an infected agar dish without films and the control was a dish with the tapioca starch-gelatin film not containing CuNPs. e radial mycelial growth of C. gloeosporioides colonies on these Petri dishes was measured for 48 h. All Petri dishes were triplicated.

In Vivo Antifungal Activity of Coatings on Mango.
e in vivo antifungal tests were adapted from a published method with some modifications [25]. Mangoes (Moringa oleifera) at moderate ripeness (4 fruits/kg) were washed successively with tap water and then with 70% ethanol to disinfect the fruit surface. e sterile fruit surface was pricked with a sterile needle to form 9 small holes of 3 mm depth. Each hole was subsequently injected with 20 µL of C. gloeosporioides mycelia in sterile water and left 2 h for drying. e first 50 mL portion of the film-forming solution was poured and spread carefully on the fruit surface to ensure a uniform coating. After 1 h of partial drying at room temperature, a second 50 mL portion of the film-forming solution was poured on the first layer. e sizes of developing fungal colonies on the fruit surface were measured for 4 days and IRG was calculated according to formula (1). All coated and uncoated mangoes were triplicated.

Formation and Characterization of CuNPs.
Copper(II) ions are reduced by several agents, such as NaBH4, hydrazine, and AA. Because the work orients on applications in food packaging, we chose AA based on its safety to human health and the environment. In the reaction with AA in solution, copper(II) ions are first converted to Cu(OH) 2 , which is then reduced successively to Cu 2 O and Cu, or CuNPs [26]. In our reaction system, gelatin and glycerol played the role of dual capping agents that protect CuNPs from growing in size. At the same time, gelatin also served as one of the polymeric materials in the composite film, and glycerol as the plasticizer for the film [27]. e dark brown-red colour of the reaction mixture after 90 min (inset of Figure 1) demonstrates the formation of metallic copper [24]. Moreover, UV-vis spectra of the reaction mixture at 30-120 min of reaction ( Figure 1) show a peak at 578-594 nm, which is characteristic for surface plasmon resonance of CuNPs [24,28]. e unchanged height of this peak after 60 min implied that the formation of CuNPs almost finished after that time. However, the spectral part with wavelengths below 570 nm still rose after 60 min, demonstrating that some side reaction(s) still occurred. We suggest that the excess AA was being oxidized by oxygen in the air. e size distribution of CuNPs in the reaction mixture after 90 min ( Figure 2) shows that their size ranged from 90 to 510 nm, in which 90% of CuNPs were smaller than 300 nm and 50% of them smaller than 200 nm. ese sizes were higher than those in other studies due to the low molar ratio of AA : CuSO 4 (5 : 1) [28,29]. However, for an applicationoriented study like this, a high excess of AA is cost-ineffective. Moreover, too small CuNPs are not stable against oxidation by air oxygen [30].
TEM images of CuNPs in the reaction mixture ( Figure 3) show that most of the CuNPs are smaller than 300 nm, which agrees with the DLS size distribution. A 20-50 nm thin layer of gelatin-glycerol capping agents surrounds each particle (red arrows). is polymeric layer was formed due to the intermolecular interactions between functional groups of gelatin and glycerol with the surface of CuNPs [31]. We suggest that these intermolecular interactions are electrostatic attractions between negatively charged electron clouds on the surface of metallic CuNPs with positively charged hydrogen atoms in -OH groups in glycerol and protonated -NH 2 groups in gelatin molecules. e capping layer serves as a cushion that prevents the adhesion of CuNPs upon their collisions. TEM images of CuNPs in other studies using polyvinyl pyrrolidone as the capping agent also showed similar thin protecting layers of about 5-8 nm [32,33].
TEM images also show the aggregation of CuNPs into bigger particles. e light areas in many aggregates or particles (yellow arrows) indicate the places where the particles did not "fit" each other upon aggregation, thus forming "holes" in the aggregates. ese holes increase the surface areas of the aggregates and therefore can facilitate the release of antimicrobial copper ions into the medium.
XRD spectrum of the synthesized CuNPs product (Figure 4) shows sharp and high characteristic peaks at 2θ of 43.31°, 50.40°, and 74.17°, which correspond to (111), (200), and (220) planes in standard JCPDS-04-0836 for metallic copper crystals [26]. ese peaks confirmed the formation of mainly copper metal in the reaction between CuSO 4 and AA. Some small peaks of Cu 2 O, CuO, and Cu(OH) 2 impurities in the XRD spectrum [34] were possibly due to surface oxidation of metallic copper by oxygen in the air or incomplete reduction of copper(II) by AA [30,35]. Figure 5 shows the ATR-FTIR spectra of glycerol (a), AA (b), gelatin (c), and the CuNPs-containing powder obtained after vacuum-drying the reactive solution (d).
In the spectrum of glycerol, the strong and broad peak at 3200-3540 cm −1 corresponds to O-H stretching and the double peak at 2950 corresponds to C-H stretching. ese two characteristics peaks of glycerol also appeared on the spectrum of CuNPs powder, hence indicating that glycerol remained in the CuNPs powder and may be involved in the capping layer of CuNPs [36].
In the spectrum of AA, four characteristic peaks at 3523, 3406, 3307, and 3228 cm −1 correspond to four O-H groups in the molecule. ese four peaks disappeared on the CuNPs powder spectrum indicating complete oxidation of AA by copper ions and air oxygen [37]. e 1670 cm −1 peak corresponding to the C-O-C group and the 1757 cm −1 peak corresponding to the lactone C�O stretching in AA also appeared in the CuNPs spectrum [38]. However, looking at the relative intensities of these two peaks in the CuNPs powder spectrum, one can recognize that the height of the 1652 cm −1 peak in CuNPs spectrum has increased significantly. is is due to the presence of the 1659 cm −1 amide I band of the gelatin present as a capping agent for CuNPs [37]. Figure 6 shows the cross-sections of the composite film cast on a Petri dish (a1-e1) and the composite coating on a mango surface (a 2-f 2 ). e average thickness (from 12 positions) of the coating was about 16 ± 1 μm, which was much lower than that of the film of 48 ± 2 μm (Figure 4, d 2 -e 2 ). is was because the low adhesive force between the film-forming solution and the mango surface kept only a thin layer of the former on the latter. e film and the coating surfaces in Figure 6 a 1 and a 2 had uniform morphology and without cracks or holes, possibly because of the flexibility of gelatin and the plasticizing effect of glycerol [39]. However, holes and channels can be observed in the cross-sections of the coating (Figure 6 a2, b2), possibly due to the continuous transmission of water vapour from the fruit to the air.

SEM Images of Cross-Sections of the Composite Films.
Particles of 2-5 μm can be observed in the cross-section of the films (red arrows in Figures 6 b 1 , e 1 , d 2 ). ese sizes are about 10-30 times bigger than those obtained in TEM and DLS analysis of CuNPs sizes.
is result indicates that, during the starch gelatinization and the film formation, CuNPs can aggregate into larger clusters. As discussed in the following section, the antimicrobial activity of the composite films is based on the release of copper ions to the medium. erefore, the aggregation of CuNPs in the film matrix is undesirable, because it reduces the surface area and the releasing ability of CuNP, which in turn reduces the antimicrobial effect of the film. However, larger sizes of CuNP make the release of copper ions more sustained over a longer time and prevent the risk of a high concentration of released copper, which may diffuse into the peel and flesh of the coated fruit. For applications that require smaller CuNPs and more homogeneous films, we suggest adding more polymer components (gelatin and/or starch) before the filmforming step to make the solution more viscous and prevent CuNP aggregation during the film drying.

In Vitro Antifungal Test of CuNPs against C. gloeosporioides.
e 500 ppm CuNPs solution was mixed with the NB medium to final copper concentrations of 100, Journal of Chemistry 150, and 200 ppm. ese CuNPs-containing nutrient media, together with other control nutrient media containing 20% mango flesh, 2200 ppm AA, or 500 ppm CuSO 4 , were used to incubate C. gloeosporioides on Petri dishes. A strong inhibition effect results in smaller colony size, compared to that in the blank control. Figure 7 shows that mango flesh (7b) and AA (7c) exhibit some antifungal activity because their acidity can inhibit the respiration of the fungi [40]. Copper ions from 500 ppm CuSO 4 100% inhibited the growth of C. gloeosporioides (7d), which is their well-known property [20]. Figures 7(e)-7(g) show concentration-dependent antifungal activity of CuNPs with 100% inhibition at 200 ppm of CuNPs. Figure 8 shows the growth dynamics of the C. gloeosporioides colonies on Petri dishes. In the first 3 days, the fungus grew on the mango flesh control faster than on the blank, possibly due to the nutrient ingredients in the mango. e concentration-dependent antifungal activity of CuNPs can be observed on the whole time range of the experiment. Due to the immobility of solid particles in the solid agar medium, the antifungal activity of CuNPs is attributable to the release of copper ions from CuNPs to the medium [41]. e antifungal activity of the CuSO 4 control can confirm this hypothesis. Moreover, many studies have shown the ease of oxidation of CuNPs to copper ions in the presence of air [22,41]. e cytotoxicity mechanisms of copper ions are well studied: they induce the formation of hydroxyl radicals when getting in contact with the cell membrane; these radicals result in the oxidative degradation of the lipids composing the cell membrane, the subsequent leakage of intracellular substances, such as K + ions [42], the alteration of intracellular biochemical processes, and finally cell death [43].
It is known that CuSO 4 begins exhibiting antifungal activity at 5-100 ppm; and in our experiments, CuSO 4 at 500 ppm inhibited 100% the growth of C. gloeosporioides [44]. However, it should be noted that using CuSO 4 in fruit coatings is undesirable due to the possible diffusion of abundant copper ions from the coating to the fruit peel. In contrast, bigger CuNPs are immobilized in the polymer matrix, and the oxidative release of copper ions from the surface of CuNPs is gradual, thus lowering the risk of copper contamination in the fruit peel. In this study, 100% inhibition of C. gloeosporioides growth at 200 ppm or 0.2 mg/mL of copper is high antifungal activity. For the sake of comparison, CuNPs synthesized using AA showed 100% growth inhibition against Fusarium solani, Neofusicoccum sp., and Fusarium oxysporum only at concentrations higher than 0.75 mg/mL.

In Vitro Antifungal Activity of Tapioca Starch/Gelatin
Films Containing CuNPs. Figure 9 shows that C. gloeosporioides grew on the control dish more strongly, compared to the blank dish. e fungus grew under, and even on, the control film because gelatin and starch are nutritional components favouring the fungal growth. In contrast, C. gloeosporioides could not grow on, and even under, the composite film containing 245 ppm CuNPs. is result indicates that CuNPs imparted the antifungal property not only to the composite film, but also to the agar surface. e agar surface acquired antifungal properties possibly because of the diffusion of antifungal copper ions from the film to the agar.

In Vivo Antifungal Activity of CuNPs-Containing Films on Mango Surface.
e coatings on mango in our study were very thin (30-50 μm from Figure 6 a 2 -d 2 ), hence keeping the colour of the fruit peel not affected by the red-brown CuNPs. Moreover, the tapioca starch-gelatin imparts a glossy appearance to the fruit (Figures 10 a, c, and d compared to b). Figure 10(b) shows that, after 4 days of infection, the C. gloeosporioides significantly developed on the surface of uncoated mangoes. At the same time, the mangoes coated with control film (Figure 10(c)) and CuNPs-containing composite film (Figure 10   84.44% and 84.77%, respectively. is result is contrary to that in the in vitro antifungal test for the films (Figure 9), where CuNPs-containing films exhibited higher antifungal activity. is contradictory result can be explained as follows.
Firstly, in the in vivo tests, the coatings completely covered the fruit surface and prevented the oxygen supply from the air to the fungus thanks to good gas-barrier properties of gelatin and starch [45,46]. In the in vitro tests, the films did not tightly cover the agar surface, thus ensuring the aerobic conditions for the fungal growth. erefore, the fungus grew strongly with the control film without CuNPs. Secondly, the wet surface of agar promoted the growth of C. gloeosporioides, while the mango surface was much drier   Journal of Chemistry and not suitable for the fungal growth [47].
is again implies that the antifungal activity of the CuNPs-containing film in the in vitro test was due to the diffusion of copper ions from the film to the agar surface.

Conclusions
CuNPs were synthesized using a green chemical approach, in which copper ions were reduced by AA in the presence of gelatin and glycerol as protecting agents. e identity and sizes of CuNPs were confirmed by XRD, TEM, and DLS results. CuNPs and tapioca starch-gelatin composite films containing CuNPs exhibited high in vitro antifungal activity against C. gloeosporioides causing anthracnose on mangoes.
is in vitro activity is suggested to be due to the diffusion of copper ions from the surface of CuNPs into surroundings. In vivo assay on mango surface showed that the starch-gelatin coatings with and without CuNPs exhibit equally high antifungal activity against C. gloeosporioides due to the low oxygen permeability of the flour and gelatin coating. is study suggests that a non-toxic and edible film based on tapioca starch and gelatin can be used for protecting mangoes from anthracnose. Further studies with proper designs should be continued to separate the antifungal effects of the composite film and the CuNPs on mangoes, as well as to study the physiological and physicochemical changes of the coated mangoes during storage, including copper contents in the peel and the flesh.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.