Chemical reagents

Calcium chloride (CaCl₂), ethanol (EtOH96), glutaraldehyde, sodium alginate (C₆H₇O₆Na; AlgNa), and tween 80 were acquired with Reactivos Química Meyer S.A de C.V. Sodium cacodylate buffer ((CH₃)₂AsO₂Na) was acquired with Sigma-Aldrich Química S.A. de C.V.

Encapsulation efficiency

AlgNa/CaCl₂ treatments corresponding to 0.5/0.8 and 1.0/0.8 were selected to evaluate the encapsulation efficiency. Samples of capsules of 1 g were crushed in the Ultra-Turrax equipment for 5 min at 13,500 rpm and mixed with 40 mL EtOH96. After settling, absorbance was measured at 275 nm in the liquid with an UV-vis spectrophotometer (DR 500, HACH, Germany) and the theoretical concentration of essential oil ( Ct ) was determined with Equation (2). A second set of capsules weighing 1 g was sampled and rapidly washed in 40 mL EtOH96, but without trituration. Absorbance was measured and the superficial oil concentration ( Cs ) was evaluated. The encapsulation efficiency ( η , %) was evaluated with Equation (15).

Microscopic structure

The structure of capsules was observed using a Jeol Scanning Electron Microscope (JMS-035; Jeol Ltd., Akishima, Japan). Capsules were fixed with 3 % glutaraldehyde in 0.1 M cacodylate buffer for 24 h. The liquid was withdrawn and two additional washes with cacodylate buffer were carried out. Then, gradual dehydration was carried out with sequential washes in ethanol between 30 and 100 %. Drying at critical point was performed, a gold coating was applied, and samples were observed in the scanning electron microscope at 15,000x.

Evaluation of fungal inhibition

Orange fruits with fungal development were obtained from a local market. The contaminating fungus was sampled, incubated in Petri dishes with potato-dextrose-agar (PDA) culture medium, and observed with an optical microscope to identify the species that were present. The fungus Penicillium digitatum was selected as a working model. Samples of such fungus were placed in immersion in 1 L of sterilized distilled water and successive dilutions were made up to 10^-5. PDA culture media were prepared in Petri dishes and these were inoculated at the center with the microbial dilution with bacteriological loop. The plates were individually placed in closed 1-L containers. The AlgNa/CaCl₂ treatment corresponding to 1.0/0.8 with immersion of 10 min was used to conduct fungal inhibition tests. Samples of 1.0, 1.5, or 2.0 g were placed in felt bags inside the containers, together with the Petri dishes. A Control treatment was established with Petri dishes not exposed to the essential oil. Fungal growth was observed on the plates for five days.

Data analysis

Behaviors of the essential oil encapsulation and release kinetics were evaluated considering three variation factors: the concentration of sodium alginate, with two levels (0.5 and 1.0 %), the concentration of calcium chloride, with two levels (0.8 and 2.5 %), and the immersion time in the solutions, with four levels (10, 20, 40, and 60 min). Data were subjected to analysis of variance, complemented with mean comparison routines applied with the Tukey's test, performed with a significance level of 0.5. All evaluations were carried out in triplicate.

Physical characteristics of capsules

Capsules of thyme essential oil were formed through ionic gelation with AlgNa and CaCl₂. Chains of D-mannuronic and L-guluronic acids interacted with calcium ions and formed a gel with a reticular structure (^{Ching et al., 2017}). When the AlgNa-Eo emulsion fell into the CaCl₂ solution, there was instantaneous gelation at the interface, which allowed the encapsulation of the Eo. Over time, the calcium ions diffused into each droplet and interacted with the AlgNa (^{Lee & Mooney, 2012}). Gelation occurred by a cooperative union of calcium ions with residues of guluronic acid involving their dimerization. With the addition of calcium, the alignment of two opposite chains of guluronic acid with a hydrophilic cavity was induced and favored the union with calcium ions using oxygen atoms of the carboxylic groups, resulting in egg-box structures (Ching et al., 2017). Scanning electron microscope observation showed several layers of alginate from the outside to the inside, which was not very evident with low gelation times, but it was very noticeable with longer times (Figure 1). Consistent with the report of ^{Bušić et al. (2018}), high porosity was found.

Figure 1 Structure of capsules obtained with ionic gelation from sodium alginate and calcium chloride solutions, with residence times in the calcium solution of 10 (left) and 20 min (right). 

The capsules had a diameter in the range of 2.16 to 2.81 mm, similarly to that obtained by ^{Aceval et al. (2019}). Likewise, they had thickness between 0.20 and 0.46 mm and both dimensions were affected by the composition of the gelation system and by the residence time in the CaCl₂ solution (Figure 2). The increase in AlgNa favored larger encapsulation diameters. Due to the polymeric nature of such compound, the increase in concentration caused higher viscosity, which caused a larger droplet size at the outlet of the dosing elements. Likewise, the capsule diameter tended to increase with higher concentration of CaCl₂, but the contrast was significant only with residence times greater than 10 min.

Figure 2 Dimensions of capsules prepared through ionic gelation. Equal capital letters indicate non-significant difference between means at the same time. Equal lowercase letters indicate non-significant difference between means within a treatment. Error bars indicate standard deviations. 

Although gelation occurs instantaneously, the calcium ions diffuse inwards as time passed, causing a larger capsule size (^{Ching et al., 2017}). However, the time in the calcium solution affected the diameter of capsules in different manner. With 0.5 % AlgNa, the diameter remained in the range of 2.16 to 2.45 mm and, although with immersion of 20 min values were higher, the contrast was considered moderate.

With 1.0 % of AlgNa, the residence time caused increase in diameter with the transition from 10 to 20-40 min of immersion, but a reduction in size was observed with 60 min. With long residence times, the capsules reached a homogeneous size, independently of the composition, with an average value of 2.39 (± 0.04) mm at 60 min and, although the contrast with respect to 10-min immersion was significant (p ≤ 0.05), the difference only varied within 2.55-5.67 %.

In the case of thickness, encapsulation with 2.5 % CaCl₂ caused values of 0.34 (± 0.03) mm, with no significant difference between different times, which suggested that, from the first contact of AlgNa with calcium ions, the gel membrane properties were not affected by additional residence times. On the other hand, with 0.8 % CaCl₂, the thickness varied within 0.20-0.24 mm with 10 min immersion and between 0.35 and 0.44 mm with times of 20-40 min.

The procedure used corresponded to a direct gelation and the thickness of the membrane capsule is proportional to the reaction time in the calcium bath (^{Caporaso & Formisano, 2016}), so the development of the capsule must be stopped by removing the spheres from the solution. Although this coincided with the systems with 0.8 % CaCl₂, it was not observed with 2.5 % concentration. However, similarly to the diameter, with a time of 60 min, the thickness tended to a homogeneous value of 0.34 (± 0.01) mm in all systems, without significant difference between them (Figure 2B).

The time of 60 min was the longest period evaluated at work. As time passes, the porosity of the capsular membrane decreases due to a greater polymer crosslinking with calcium ions, which might have caused contraction and even matrix dehydration due to the expulsion of water.

The behavior with 0.8 % CaCl₂ suggested that, depending on composition, it is feasible to control the diameter and the thickness of capsules. Based on this result, mixtures with that concentration of CaCl₂ were selected to evaluate the behavior of the encapsulation systems. However, as the capsule structure is not rigid and changes with the interaction between ions, a more stable state is reached as time advances, ceasing the dependency to the system composition.

On the other hand, the amount of trapped oil is determined by the dimensions of the capsule core. The lowest Dn values were found in the 0.5/2.5 AlgNa/CaCl₂ system, followed, in ascending order, by that of 0.5/0.8, and finally by those of 1.0/2.5 and 1.0/0.8 systems (Figure 2C). This suggested that the concentration of sodium alginate, acting as a wall component, determined the amount of encapsulated oil. The highest contents of essential oil were obtained in the 1.0/0.8-AlgNa/CaCl₂ system, with immersion time of 10 min derived of a small thickness in a moderate diameter, and with 20 min, which was associated with large thickness in a large diameter capsule.

Systems based on the 0.8 % CaCl₂ solution were evaluated in terms of encapsulation efficiency ( η ). The average values varied in the range of 88.2 to 95.2 %, which indicated that between 4.8 and 11.8 % of Eo remained in the superficial region of capsules and not in the core, despite this, a good encapsulation degree was accepted. The comparison of efficiencies between similar times showed that there was no significant difference (p > 0.05) between systems with 0.5 or 1.0 % of AlgNa. However, in both cases, the efficiency decreased with the transition from 10 to 20-40 min of residence in the calcium solution and then increased again with 60 min of immersion (p ≤ 0.05) (Table 1). This behavior coincided with the variation of capsule dimensions, thus the hypothesis that the structure tended to a more stable state with residence time was strengthened, which favored the retention of essential oil in the nuclear zone. In this regard, ^{Moura et al. (2018)} pointed out that there is an inverse relationship between the efficiency and the dimensions of the capsules.

Kinetics of essential oil release

The release of the essential oil from the capsules had a logarithmic behavior and data were properly fitted to Equation (14), with determination coefficients higher than 0.99 in all cases. Thus, the increase in concentration in the headspace of containers occurred rapidly at the beginning and thereafter the rate of change gradually slowed down, so the presence of Eo tended to a constant value (Figure 3). This behavior was due to the release of Eo was carried out in a closed container and, over time, the concentration gradient between the interior of the capsules and the headspace was gradually reduced, thereby also reducing the release. However, if release develops in systems with higher dimensions such as a greenhouse, it shall be more important the evaluation of the concentration change as a function of the distance from the liberation point.

Figure 3 Thyme essential oil release kinetics affected by the composition of AlgNa-CaCl₂ systems and the immersion time in the calcium solution. 

The constant k4 of Equation (14) represents the maximum concentration that can be reached in the container after a long release time and was not affected by the composition of AlgNa solution, but it was affected by that of CaCl₂, where the highest values were found with 0.8 % and the lowest with 2.5 % concentration (Table 2), although the average difference was less than 1.2 mL m^-3. On the other hand, as the immersion time in the calcium solution increased, the maximum concentration reached in the headspace tended to be higher. However, the contrast was significant only between 10- and 60-min of immersion times, but it was considered without practical importance due to the difference was less than 1.3 mL m^-3.

The constant k5 (s^-1) represents an essential oil release coefficient and expresses the rate of change of concentration in the container per unit of concentration gradient between the interior of capsules and the headspace. The constant k5 (s^-1) was affected by both, the composition of AlgNa and CaCl₂ solutions. The highest values were obtained with the lowest concentrations of such compounds. Likewise, k5 had the lowest value with an immersion time of 10 min, but the highest value was found at 20 min (Table 2).

Although the constant k4 indicates the maximum concentration to be reached in the headspace, the constant k5 is associated with the required time to reach certain concentration. Based on Equation (14), some characteristic release times can be obtained (Equation 16), where t1/2 is the mean essential oil release time (h), while t7/8 and t0.99 are required times (h) to reach 87.5 and 99 % of the maximum concentration, respectively.

Average t1/2 time was 39.57 (±2.86) and was not affected by AlgNa and CaCl₂ nor by the immersion time (Table 2). The average t7/8 and t0.99 were 118.70 (±8.57) and 262.87 (±18.97) h, respectively. Values of t1/2 , t7/8 , and t0.99 tended to be higher with higher concentrations, congruently with the constant k5 , although with significant contrast (p ≤ 0.05) only in the effect of CaCl₂. In addition, the longest required values of t1/2 , t7/8 , and t0.99 were found with immersion of 10 min and the shortest with 20 min, also congruently with k5 , while time requirements were intermediate with 40 and 60 min.

These results indicated that, based on the concentration of AlgNa and CaCl₂ solutions, it is feasible to control the release rate of Eo, but this can only be performed with immersion times between 10 and 20 min within the calcium solution, since different behaviors were not obtained with longer residence times. However, the higher release rate with 20 min was caused by a lower encapsulation efficiency (Table 1), where the superficial oil volatilized faster than that of the nuclear region. In this regard, since the superficial concentration of essential oil cannot be controlled, the 10-min residence time was considered a better alternative to attend the encapsulation of the essential oil.

Potential for fungal inhibition

The formulation AlgNa/CaCl₂/t-1.0 %/0.8 %/10min was used to evaluate the potential of the capsules for fungal inhibition, due to the larger nuclear diameter and high encapsulation efficiency obtained. A searching process was carried out for the required quantity of capsules to inhibit the development of Penicillium digitatum, which had been inoculated on culture plates, through the release of essential oil in their headspace.

The growth of the fungus in the Control treatment, without essential oil exposure, varied from low to moderate on the first day of incubation, from moderate to high on the second, and it was high in all the cases after the third day (Table 3). In contrast, when the inoculated medium was exposed to the presence of 2.0 g of capsules, the released oil inhibited totally the fungus development. In order to find the capsules minimum inhibitory quantity (MIQ), the test was repeated with batches of 1.0 g of capsules and a low to moderate growth was observed in the first two days. A moderate growth was noted afterwards until the end of the five days evaluation period. These results indicated that the MIQ was between 1.0 and 2.0 g. In this sense, a third test with batches of 1.5 g of capsules showed that, with this quantity, a total inhibition of the fungus was also achieved. Although results indicated that the MIQ was between 1.0 and 1.5 g of capsules, the quantity used in the last test, that is, 1.5 g of capsules, was accepted as the minimum required to cause the inhibition of the fungus.

P. digitatum is an important factor that limits the shelf life of citrus fruits and it is known as green mold. Besides, other species of the genus Penicillium limit the consumption of fruits as apple due to the production of mycotoxins (^{Heinmaa et al., 2019}). ^{Talibi et al. (2014}) pointed out that the use of essential oils, particularly those of cinnamon and thyme, have potential to control most rots of citrus in postharvest, including green mold, blue mold, and sour rot.

The proposed mechanism of action of essential oils against microorganisms includes the affectation of the cell membrane, causing its disruption and making it more permeable, altering ion transport processes, thus modifying the concentration of electrolytes like K⁺, Ca²⁺, and Na⁺. Also, it has been suggested the interaction with membrane proteins, which may cause the disruption of the functions of mitochondria, which conduces to an imbalance in ATP concentration and, eventually, the death of the cell (^{Mutlu-Ingok et al., 2020}).

^{Talibi et al. (2014}) pointed out that the use of essential oils in citrus fruits can be made difficult by possible problems of phytotoxicity or alteration of sensory attributes. In this regard, the results of the present work showed that a direct contact with the product is not required, since the essential oil is released in the headspace. This use based on the release of essential oil vapors in the headspace was consistent with the use of biopolymeric films, where ^{Peretto et al. (2014}) incorporated carvacrol and methyl cinnamate at 0.75 %, and these compounds were vaporized from there and achieved fungal control in strawberry fruits.

Also, ^{Martínez-Romero et al. (2007}) reported the high effectiveness of an essential oil through the use of carvacrol vapors at a concentration of 1000 mL m^-3 in the headspace of grapefruits. Similarly, ^{López-Gómez et al. (2021}) released vapors of eugenol, bergamot, and grapefruit essential oil at a concentration of 100 mL m^-3 in a modified atmosphere system that allowed the quality of mushroom slices to be maintained for 12 d.

The use of essential oils to reduce microbial growth is normally based on direct contact through biopolymeric coatings, but the results of this work showed that the use of vapors with controlled release can be done through encapsulation. In this regard, a release analysis based on Equation (13), with the parameters of the AlgNa/CaCl₂/t-1.0/0.8/10min system showed that the concentration of Eo in the headspace that is required to cause inhibition was less than 100 mL m^-3, which contrasted with the concentrations that must be used with biopolymeric coatings, fluctuating between 500 and 1000 mL m^-3 (^{Valle-Ortiz et al., 2019}), indicating that the release of essential oils from encapsulation systems constitutes a better strategy to control the fungus development.

Nowadays, the evidence for using essential oils to reduce losses caused by microbial development is growing. However, it is necessary to attend several aspects to improve the potential for using such substances in the conservation of horticultural products. In this regard, ^{Mutlu-Ingok et al. (2020}) pointed out the following aspects that may part of future researches for using vapors of essential oils: the quantity to be used should consider not only the minimum inhibitory concentration (MIQ) but also the maximum sensorially permitted concentration (MSC) to avoid the modification of the quality attributes of the product, for which designed mixtures of different essential oils may be applied, considering the synergistic effect between them. The determination of MIQ must also consider the control of both, the growth and mycotoxin production of fungi, the part of the plant to be protected, the geographical aspects, and the extraction method of the essential oil. Authors explained that it is also necessary to investigate the antifungal actions not only against monocultures, but also against polycultures. In addition, new strategies for improving the stability of essential oils are required. In relation to this, an encapsulation strategy as that showed in the present work can be adequate, together with the modeling of the kinetics release, which can allow determining the quantities to be used in a better form, once the MIQ and the MSC have been established.