Optimization of lemongrass essential oil nanoemulsion formulation and antifungal activity against Colletotrichum musae QB6 causing anthracnose in bananas

2.1 Materials

Lemongrass (Cymbopogon citratus) was collected, and LEO extraction was performed following the procedure detailed in Fig. S8. Bananas (Musa AAA) were harvested, and the fungus C. musae QB6 was cultured, as described in Fig. S9.

Tween® 80, a polysorbate non-ionic surfactant with purity >58.0% (GC), and benzimidazole, with purity >97.5% (HPLC), were employed in the experiments and were purchased from Sigma-Aldrich company. All other chemicals used were of analytical grade and obtained from reputable suppliers.

2.2 Preparation and characterization of LEO Nanoemulsion

Nanoemulsions were prepared by adapting the method of Hunde et al. (Hunde et al., 2023) and Gupta et al. (Gupta et al., 2016) with modifications. The emulsions were formed by combining the oil phase and the water phase, as described in Fig. S10.

The DD and PDI of the LEO-NE were evaluated by dynamic light scattering with non-invasive backscattering using a particle size analyzer (Zetasizer Nano ZS90, Malvern Instruments Ltd., UK).

2.3 Optimization of LEO-NE for DD and PDI

RSM was employed to optimize the preparation conditions of the LEO nanoemulsion, as shown in Table S4. BBD was employed for the optimization study (Table 1) with a total of 17 trials. The collected experimental data were analyzed using Minitab 21.2 software to construct a generalized quadratic regression equation, representing the relationship between the independent variables and the response variables, as follows:

$Y_n=a_o+{\sum }_{i=1}^3{a}_i{X}_i+{\sum }_{i=1}^3{a}_{ii}{X}_i^2+{\sum }_{i=1}^2{\sum }_{j=1+1}^3{a}_{ij}{X}_i{X}_j$

Where:

• Yₙ: The response variables,

• a₀: constant,

• aᵢ, aᵢᵢ, aᵢⱼ: linear, quadratic, and interaction coefficients, respectively,

• Xᵢ and Xⱼ: levels of the independent variables.

2.4 Determination of nanoemulsion stability

The LEO nanoemulsions were stored in transparent glass bottles, wrapped in aluminum foil to prevent direct light exposure to the samples. Systems exhibiting phase separation or creaming were considered unstable. DD and PDI values were measured after storage periods at 0, 30, 60, 90, and 120 days of storage. Electrophoretic mobility was measured using a Zetasizer Nano ZS (Malvern, UK) via laser Doppler velocimetry at 25°C.

2.5 Antifungal activity to anthracnose on banana fruit

Antifungal activity of the optimized LEO-NE on C. musae was evaluated following the method, as designated in Fig. S11. The inhibitory effect was calculated as the percentage (%) of inhibition of the growth rate of the colony diameter, Percentage inhibition of radial growth (PIRG%).

$PIRG\%=\frac{(R_1-R_2)}{R_1}\times 100\%$

Where R₁: radial growth of fungi on control plates, R₂: the radial growth of fungi on treated plates.

Effect of the optimized LEO-NE on C. musae biomass formation in vitro, as described in Fig. S12. The percentage inhibition of fungal biomass growth was calculated based on the dry weight of the biomass (Long et al., 2020).

In antifungal assays of optimized LEO-NE against C. musae in vivo, as detailed in Fig. S13. The diameter of the lesions was measured daily using a digital calliper over a 6-day period, with three replicates for each treatment (five bananas per replicate).

2.6 Statistical analysis

All data were processed and optimized using Minitab software with the experimental design based on the Box-Behnken Design (BBD) method. Statistical significance was determined at a confidence level of p < 0.05.

3.1 Optimization of LEO nanoemulsions by RSM

The experimental data for LEO-NE optimization have been presented in Table 1. The DD (Y₁) ranged from 146.88 to 281.62 nm, which significantly influences the stability, optical properties, and viscosity of nanoemulsions (Sharma et al., 2010; Uluata et al., 2016). The PDI (Y₂), reflecting DD uniformity, varied from 0.19 to 0.55, indicating a stable nanoemulsion system. Regression analysis and ANOVA were employed to evaluate the significance of the model terms, with the R² and adjusted R² values demonstrating strong predictive capability (Table 2). Specifically, the R² values of 0.9633 for DD (Y₁) and 0.9526 for PDI (Y₂), along with their adjusted counterparts of 0.9266 and 0.9157, confirm that the quadratic polynomial model can explain more than 90% of the variability in the response variables. ANOVA results indicated that linear, quadratic, and interaction terms of independent variables significantly influenced both DD and PDI (p < 0.05), justifying the non-linear relationships and the use of a quadratic polynomial model, as visualized in Fig. 1 and Fig. 2. The final equations according to the coded factors for the responses Y₁ and Y₂ produced the following polynomial formulas:

$\begin{array}{c}{Y}_1=2595-270.0X_1-0.0459X_2+14.40X_3\\ +9.223{\text{X}}_1^2+0.03\times {10}^4{\text{X}}_2^2-21.52\times {10}^4{X}_1{X}_2\\ -1.381X_1{X}_3+8.6\times {10}^4{X}_2{X}_3\end{array}$

$\begin{array}{c}{Y}_2=6.167-0.6634X_1-1.36.104X_2+0.0428X_3\\ +217.3\times {10}^4{\text{X}}_1^2-0.004X_1{X}_3+0.02\times {10}^4{X}_2{X}_3\end{array}$

Close

Fig. 1.

Response surface plots of interaction between (a) T-80 concentration and homogenization speed; (b) T-80 concentration and ultrasonication time; (c) ultrasonication time and homogenization speed on DD of the LEO nanoemulsions, and (d) The influence of factors on DD at (X1:X2:X3) of (17.5:11000:15).

Export to PPT

Close

Fig. 2.

Response surface plots of interaction between (a) T-80 concentration and homogenization speed; (b) T-80 concentration and ultrasonication time; (c) ultrasonication time and homogenization speed on PDI of the LEO nanoemulsions; (d) The influence of factors on PDI at (X1:X2:X3) of (17.5:11000:15)

Export to PPT

The DD (Y₁) was primarily influenced by T-80 concentration (X₁) and the homogenization speed (X₂) (Fig. 1a). The smallest DD (<100 nm) was achieved at high T-80 concentrations (∼20 g/L) and homogenization speeds (12,000-14,000 rpm), attributed to enhanced interfacial tension reduction and efficient droplet dispersion. However, exceeding the optimal T-80 level led to system destabilization due to excess micelle formation, while excessive homogenization speeds generated localized heating, promoting droplet re-coalescence. Conversely, lower levels of both factors resulted in larger DD (>300 nm), indicating reduced nanoemulsion stability. These findings align with previous studies on emulsifier concentration and mechanical energy for optimal DD reduction and stability (Sharma et al., 2010; McClements and Jafari 2018). In addition, the interaction between X₁ and X₃ (Fig. 1b) and that between X₂ and X₃ (Fig. 1c) significantly impacted DD, with optimal conditions leading to smaller sizes and excess leading to aggregation or re-agglomeration. This confirms the synergistic effect of shear forces and ultrasonic cavitation facilitated efficient droplet breakdown and uniform dispersion.

The PDI (Y₂) was significantly influenced by the T-80 concentration (X₁) and the homogenization speed (X₂), with the ultrasonication time (X₃ = 0) held constant (Fig. 2a). Increasing the concentration of T-80 from 15 g/L to 20 g/L led to a decrease in PDI due to enhanced surface stabilization and improved droplet dispersion. However, excessively high homogenization speeds (∼14,000 rpm) resulted in an increase in PDI due to localized heating effects, which destabilized the emulsion and caused droplet re-coalescence. The optimal conditions were identified at a T-80 concentration of 20 g/L and a homogenization speed of 12,000 rpm, yielding the lowest PDI (∼0.2), indicating a highly uniform nanoemulsion system. The PDI was significantly affected by X₁ and X₃ (Fig. 2b), which underscores the necessity of optimizing both factors simultaneously to maintain nanoemulsion stability. X₂ and X₃ exhibited a synergistic effect on PDI (Fig. 2c), emphasizing the critical importance of simultaneous optimization of homogenization speed and ultrasonication time to achieve stable nanoemulsions with uniform DD distribution.

The optimization results using the BBD revealed that the ideal conditions for producing LEO nanoemulsions were a T-80 concentration of 17.22 g/L, a homogenization speed of 10,180 rpm, and an ultrasonication time of 20 min (Table 3). Under these parameters, the nanoemulsion achieved a DD of 153.02 nm and a PDI of 0.22, signifying a highly stable and uniform system. The measured zeta potential of -12.0 mV for the optimized nanoemulsion indicates the presence of negative charges on the droplet surface, contributing to electrostatic repulsion among droplets (Fig. S5). These optimized conditions underscore the importance of harmonizing emulsifier concentration, homogenization speed, and ultrasonication time to attain the desired nanoemulsion properties.

3.2 Stability of LEO nanoemulsions

The stability of LEO nanoemulsions is essential for their efficacy in various applications (Gago et al., 2019; Barradas and de Holanda e Silva 2021). Key stability indicators, such as DD and PDI, were evaluated in this study. Nanoemulsions were formulated under optimized conditions of T-80 concentration, homogenization speed, and ultrasonication time. The results, as depicted in Fig. 3, confirm the exceptional stability of LEO nanoemulsions over 120 days, with DD consistently remaining below 200 nm, and PDI stayed under 0.3. The absence of creaming, phase separation, or notable size variation reflects an optimal balance between surface tension reduction and coalescence prevention, achieved through precise control of the emulsifier concentration and processing parameters. These findings are consistent with contemporary research highlighting the critical role of optimized emulsifier levels and energy input in maintaining nanoemulsion stability, making prepared LEO-NE highly suitable for applications in pharmaceuticals, food preservation, and cosmetics (Gupta et al., 2016; McClements and Jafari 2018; Gago et al., 2019; Barradas and de Holanda e Silva 2021).

Close

Fig. 3.

DD and PDI of optimized LEO nanoemulsion samples over storage time.

Export to PPT

A batch of LEO-NE was made especially with these ideal conditions after RSM determined the optimized parameters as a validation experiment. This optimized nanoemulsion batch’s experimentally measured properties at time zero included a DD of 153.2 nm and a PDI of 0.22 (Tab. S5 and Fig. S4), both of which are very close to the values that the RSM model predicted.

3.3 Antifungal activity against anthracnose on banana fruit

Effect of nanoemulsion on the colony diameter growth of C. musae after 10 days of monitoring. The results presented in Table 3 and Fig. 4 demonstrate exceptional antifungal activity against C. musae QB6. The optimized LEO-NE showed complete inhibition of fungal growth with a PIRG of 100% comparable to the synthetic fungicide Benzimidazole. In contrast, non-nanoemulsified LEO partially inhibited fungal growth, with a PIRG of 85.94%. This finding reinforces the growing evidence of nanoemulsions as sustainable and efficient antifungal agents.

Close

Fig. 4.

Growth of C. musae QB6 colonies after 10 days of incubation under different treatments: (a) Distilled water (negative control), (b) Benzimidazole (positive control), (c) Pure LEO, and (d) Optimized LEO nanoemulsion.

Export to PPT

The effect of LEO-NE on biomass formation of C. musae QB6 after 10 days of incubation at 28°C was measured. Figs. 5 and 6 indicate that both the optimized LEO-NE and Benzimidazole exhibited superior antifungal activity compared to pure LEO and the negative control (distilled water). In the absence of antifungal agents, C. musae QB6 biomass reached 259.3 mg, indicating unrestricted growth. The optimized LEO-NE completely inhibited biomass formation, achieving a 100% inhibition rate, comparable to Benzimidazole. In contrast, pure LEO exhibited significant but lower antifungal efficacy, reducing biomass to 30.1 mg with an inhibition rate of 88.39%. These findings align with various studies, which highlight the role of nanoemulsions in enhancing the bioactivity of EOs by improving cellular uptake and reducing required dosages for efficacy (Maurya et al., 2021; Guidotti-Takeuchi et al., 2022).

Close

Fig. 5.

Inhibitory effect of optimized LEO nanoemulsion on C. musae QB6 biomass after 10 days of incubation at 28°C.

Export to PPT

Close

Fig. 6.

Biomass formation of C. musae QB6 after 10 days of incubation in PDB medium under different treatments: (a) Distilled water (negative control), (b) Benzimidazole (positive control), (c) Pure LEO, and (d) Optimized LEO nanoemulsion

Export to PPT

Banana fruit antifungal activity against anthracnose was evaluated in vivo (Fig. 7). Benzimidazole and LEO nanoemulsion-treated bananas showed no outward signs of damage after 6 days, demonstrating total inhibition of fungal growth. In the meantime, bananas treated with pure LEO developed minor lesions by the end of the observation period, indicating delayed but partial inhibition, and significant lesion development was seen in the untreated control (sterile distilled water). By providing a sustainable method for managing fungal diseases in fruits after harvest, this study highlights the potential benefits of using nanoemulsified EOs as efficient and environmentally friendly substitutes for synthetic fungicides.

Close

Fig. 7.

Effect of different treatments on anthracnose lesion development in bananas inoculated with C. musae QB6 in vivo. (a) Distilled water (negative control), (b) Benzimidazole (positive control), (c) LEO, (d) Optimized LEO nanoemulsion

Export to PPT

The enhanced antifungal efficacy of LEO-NE over bulk LEO is likely attributable to its nanoscale properties, including increased surface contact, improved membrane permeability, and stabilization of volatile actives. These physicochemical advantages support sustained inhibition in both in vitro and in vivo settings, as reported in similar studies, reinforcing the potential of nanoemulsions in postharvest disease control.

Nonetheless, several limitations should be acknowledged. The in vivo assays were conducted under controlled laboratory conditions, which may not fully reflect commercial postharvest environments. Additionally, the scope of biological validation and storage stability assessments remains restricted. Future research should address these aspects to better inform real-world applicability.