Introduction

Studies show that among the multitudinous biotic factors, particularly phytophagous insect pests are widely recognized as important constraints to agricultural production1. The developing herbivore larva can reduce the yields of crops by feeding and mining the leaf mesophyll, damaging the leaves, and reducing plant photosynthetic capacity2. The consequences of pest infestation in the absence of efficient management could potentially account for approximately 25–30% of agricultural production losses3. To combat continuous threats from various phytophagous insects, plants have evolved a whole arsenal of sophisticated defense mechanisms, including morphological and structural characteristics as well as the synthesis of chemical compounds4. Chemical defense products may range from low molecular weight substances called secondary metabolites (SMs) to peptides and proteins that are effective against insects. SMs are produced through host-pest interactions and actively contribute to plant chemical defense against phytophagous insects with potent antiherbivore properties5. Under normal conditions, SMs are either absent or synthesized in very minimal amounts, while during biotic stress, these metabolites serve as antioxidants, osmoprotectants, antimicrobials, and repellents to protect the plant6. Furthermore, SMs can directly function as toxins by deterring attacks of herbivores or undergo oxidation by enzymes such as peroxidases or polyphenol oxidases, resulting in poisonous byproducts that disrupt insect growth and development processes7.

Phenolics are one of the prominent subgroups of SMs widely distributed in plants. Among the multifarious phenolic SMs, tannic acid (TA) stands out as a weakly acidic polyphenol composed of gallic acid (GA) units connected by ester bonds to glucose8. In terms of function, TA is a mouth-puckering bitter polyphenol that not only shows insect growth inhibitory activity against numerous pests but also functions as a defense agent and exhibits antifeedant activity as a deterrent molecule for herbivorous pests9. It has been reported that TA can be directly synthesized and accumulated from precursors such as aromatic amino acids (e.g., phenylalanine) through the shikimate pathway within plant cells10. Plants can efficiently uptake and metabolize externally supplied phenylalanine (PA), with PA-enriched plants showing high resistance to pests through a combination of priming and induced defense responses, thus positioning PA as an effective insect repellent11,12. Simultaneously, the metabolized residue of GA from TA can influence the development and exhibit potential toxicity to some insects at high concentrations13. In terms of chemical structure, the hydroxyl groups (-OH) in the phenolic act as excellent hydrogen (H) donors and form strong H bonds with H-bond-accepting compounds containing carboxyl, amine, or hydroxyl groups14. In addition, non-covalent π-π interactions (π-π stacking) and hydrophobic interactions can occur between the aromatic rings of phenolic and other aromatic systems or hydrophobic molecules, respectively15. Such prospects serve as the foundation for the application of phenolic in the supramolecular self-assembled nanosystems. Considering both the outstanding functional and structural characteristics provided by natural phenols, phenols functionalized nanosystems with improved physicochemical properties and biological activities are increasingly employed to accomplish the efficient delivery of agrochemicals16.

Pesticides are extensively used in agrochemicals in controlling the outbreak of pests for improving agricultural productivity and maintaining a sufficient food demand17,18,19. However, up to 90% of agrochemicals applied via foliar spray are lost through washoff or runoff, primarily influenced by external environmental conditions (wind, rain, and hot weather) and poor uptake caused by the hydrophobic waxy cuticle barrier on plant leaves20. Meanwhile, the excessive and indiscriminate use of traditional chemical pesticides raises widespread concerns regarding their impact on non-target species, and potential deleterious effects on human health, ecosystems, and biodiversity21,22. There is a growing interest in reducing the reliance on commercial pesticides and exploring alternatives to these agrochemicals. As a result, biopesticides have emerged as a potential alternative to synthetic pesticides and inspiration for developing new active ingredients (AIs)23. Currently, the most promising botanical insecticide for integrated pest management is azadirachtin (AZA), which is a limonoid tetranortriterpenoid derived from neem trees (Azadirachta indica) with excellent bioactivities against more than 600 insect species24. As the most prominent biopesticide, AZA can interfere with various physiological processes and chemical pathways thereby offering toxic effects as a feeding deterrent or insect growth disruptor for pest control in organic agriculture25. On a worldwide scale, the main commercially available formulation of AZA is emulsifiable concentrate (EC) owing to its hydrophobicity. Traditional EC formulations pose a significant risk to the environment and human health due to the use of hazardous solvents like toluene and xylene in some procedures26. Although nanoformulations have been reported to deliver AZA efficiently, there are still problems and limitations such as the usage of nanocarrier materials and harmful adjuvants, low drug loading rates (ranging from 9.23–14.30%), complex preparation process, and potential environmental risks27,28,29. Given the above inevitable defects with existing hydrophobic pesticide applications, the water-based nanosystems without any harmful solvents constructed via self-assembly technology could represent challenging and subversive progress to the conventional formulations and carrier-based nanopesticides30,31,32. Moreover, plant-based products functionalized nanosystems for hydrophobic pesticides could remarkably enhance the effectiveness of agrochemicals while acting as potent regulators to promote plant growth and development33,34,35.

In this study, three plant-based products TA, PA, and GA with different regulative models and physicochemical characteristics (UV shielding capability and hydrophilicity) are envisaged and preselected as versatile “green” building blocks to modify the lipid-soluble active substance AZA to co-assembled into nanosystems for efficient drug delivery, synergistic pest management, and concurrently develop the defensive capacity and increase the production yield of plants (Fig. 1). The possibility, mechanism, preparation conditions, and assembly abilities of co-assembled nanosystems for AZA to three molecules are systematically analyzed and discussed. To gain insights into the stability of AZA nanosystems, hydrodynamic particle size distribution, polydispersity index (PDI) value, ζ-potential, and morphology characterization are performed using dynamic light scattering (DLS) and transmission electron microscopy (TEM). The structure of prepared AZA nanosystems is characterized by high-performance liquid chromatography (HPLC), UV-visible absorption spectrum (UV/vis), and Fourier transform spectroscopy (FTIR). The physicochemical properties, including wettability, deposition, rain retention, and photostability are fully investigated and verified. The controlled release behavior of AZA from co-assembled nanosystems, in vitro and in vivo insecticidal activities, and biosafety of prepared nanosystems are thoroughly estimated. This work is anticipated to provide a convenient method for the application of biopesticides, which can improve the physicochemical properties and maximize the utilization efficiency of pesticides.

Fig. 1: Schematic illustration of design strategy.
figure 1

Construction of water-based nanopesticide systems based on the non-covalent assembly of azadirachtin (AZA) and three plant secondary metabolites (tannic acid, TA; phenylalanine, PA; gallic acid, GA) for long-lasting and synergistic pest management.

Full size image

Results and discussion

Preparation of AZA-based nanosystems

The preparation method of co-assembled nanosystems based on hydrophobic AZA molecules and hydrophilic molecules with different log P (TA, PA, or GA) is shown in Fig. 2. The mixture solution prepared by dissolving AZA and TA, PA, or GA in MeOH or water was slowly introduced into water drop by drop, and then co-assembled into the different suspensions through a solvent exchange reaction. To achieve the successful preparation of water-based nanosystems of AZA, the potential impacts of the molar ratio between AZA and TA, PA, or GA (before blending) on the composition, and dispersing properties of the co-assembled complexes were systematically explored (Fig. 3). As a classical high-molecular-weight polyphenol, the chemical structure of TA involves abundant hydrophilic hydroxyl groups and hydrophobic aromatic rings that could interact with target molecules (Supplementary Table 1)36. Therefore, the amount of TA was held constant, while the amounts of AZA were systematically varied ranging from 1 to 10. Due to the small molecule construction and spatial molecular arrangement, the amount of AZA was held constant, the amounts of PA or GA ranging from 1:5 to 1:0.17.

Fig. 2: Schematic of preparation method.
figure 2

Preparation of co-assembled nanoparticles (NPs) based on AZA, TA, PA, and GA through a solvent exchange method.

Full size image
Fig. 3: Co-assemblies of AZA and TA, PA, GA.
figure 3

a–c Co-assembly rates (Data are presented as mean values ± SE), averaged particle size distributions (hydrodynamic diameter), ζ-potential values (Data are presented as mean values ± SD), and polydispersity index (PDI) values of co-assembled AT NPs with different molar ratios. d–f Co-assembly rates (Data are presented as mean values ± SE), particle size distributions, ζ-potential values (Data are presented as mean values ± SD), and PDI values of formed AP NPs. g TEM and enlarged TEM images of AG suspensions (Scale bar = 500 and 200 nm) at a molar ratio of 1: 0.33. h and i TEM and enlarged TEM images of AT NPs (Scale bar = 500 and 100 nm) at a molar ratio of 1: 8 (TA: AZA) and AP NPs (Scale bar = 500 and 200 nm) at a molar ratio of 1: 4 (AZA: PA). j and k For the further verification and exhibition of a well-defined core-shell structure, enlarged TEM images of AT NPs (Scale bar = 100 nm) at a molar ratio of 1: 2 (TA: AZA) and AP NPs (Scale bar = 300 nm) at a molar ratio of 1: 5 (AZA: PA) were also displayed. Each experiment contained three independent samples. Source data are provided as a Source Data file.

Full size image

As expected, it was observed that TA and PA molecules could utilized as the dynamic cross-linking media for AZA to co-assemble into relatively small aggregates and well dispersed in water, which met the requirement of regulating the dispersion behavior of AZA (Fig. 3). As illustrated in Supplementary Fig. 1a, the combination of water-insoluble AZA and water-soluble TA leads to the successful construction of co-assembly systems. As the molar ratios of TA to AZA increased from 1:1 to 1:10, the nanoparticles became more compact with diminishingly average sizes, the stacking, and dissociation of numerous negatively charged hydroxyl groups, resulting in lower ζ-potential values and excellent stability until the molar ratio of 1:8 (Fig. 3b, c). Simultaneously, the co-assembly rates of the AT NPs gradually increased reaching a maximum of 82.46% with an average size, PDI, and ζ-potential value of about 120 nm, 0.028, and − 32 mV, suggesting a uniformly dispersed stable nanosystem with a balance of hydrophobicity-hydrophilicity at the molar ratio of 1:8 (Fig. 3a–c). However, as the content of PA decreased, the combination of AZA and PA led to significant dehydration, thus resulting in spontaneous precipitated from water at molar ratios ranging from 1:0.5 to 1:0.17 (Supplementary Fig. 1b), suggesting a nonequilibrium state of hydrophobicity-hydrophilicity. The co-assembly rate and ζ-potential values of AP NPs reduced until the construction of water-insoluble aggregates, which correlates with the reduced PA content and the loss of negatively charged carboxy groups (Fig. 3d–f). In contrast to the dehydration, dynamic stable nanosystems with a high co-assembly rate (> 65%), low PDI value (< 0.15), and ζ-potential value (< − 20 mV) were established at the molar ratio of 1: 4 and 1: 3 for AZA to PA (Fig. 3d–f).

Unfortunately, the combination of AZA and GA formed the water-insoluble aggregates due to the strong dehydration at whole molar ratios (1: 5-0.17), and many small irregular aggregates further accumulated and packed closely, leading to the formation of larger aggregates (Fig. 3g and Supplementary Fig. 1c). The formation of water-insoluble aggregates could be attributed to the extremely low compatibility of AZA with GA, which was crucial to the precise regulation of the morphology and structure of co-assembled nanoparticles37. Moreover, an imbalance of hydrophobicity-hydrophilicity appeared in the combination of AZA and GA due to the hydrophobicity properties and low solubilities of GA functionalized assemblies38. According to these construction results, the co-assemblies of AZA molecules to different compounds were associated with the solubilities and interface assembly capacities of building blocks. In consideration of the maximum AZA content and well disperse behavior, the TA and PA were selected as the optimal molecules for the construction of co-assembled AZA nanosystems in an aqueous solution. For the efficient delivery of AZA, the optimal molar ratio of TA to AZA was selected as 1: 8 with co-assembly rate of 82.46% and loading rate of 77.22%, while that of PA to AZA was selected as 4: 1 with co-assembly rate of 70.96% and loading rate of 52.17% (Supplementary Table 1).

Morphological characterization

TEM images of the prepared AT NPs and AP NPs at the optimal molar ratios illustrate regular spherical nanostructure with sizes of ~ 100 and 250 nm, respectively (Fig. 3h, i). As shown by the inset enlarged part of TEM photos, the prepared AT NPs or AP NPs are core-shell structured spherical nanoparticles with a core of AZA and an outer layer of TA or PA. For the further verification and exhibition of a well-defined core-shell structure, enlarged TEM images of AT NPs (TA: AZA = 1: 2) and AP NPs (AZA: PA = 1: 5) were also displayed (Fig. 3j, k). The observation corroborated the classical core-shell structure of AT NPs or AP NPs and the encapsulation of AZA within the core. Apparently, the coating thickness of the sample gradually increased with the amounts of TA or PA, while it gradually decreased as the amount of AZA increased.

Figure 4a shows the morphology and obvious Tyndall phenomenon effect of the co-assembled nanosystems at the optimal molar ratios. Due to the hydrophobic property, the free AZA could not be self-assembled and precipitated to sediments in the deionized water, while the free TA and PA could be dissolved. The average hydrodynamic particle sizes of AT NPs and AP NPs determined by DLS were 117 nm (PDI = 0.029) and 287 nm (PDI = 0.129) (Fig. 4b, c), respectively. The larger sizes than that determined by the TEM images could be attributed to the fact that the Zeta-sizer Analyzer tests the hydrated particle size, whereas TEM typically provides the actual particle size39. The average surface charge values of free AZA, TA, and PA were − 5.97, − 18.3, and − 3.93 mV, respectively, which could be attributed to the negative charge property of hydroxyl and carboxy groups in their molecular structures (Fig. 4d). The more negative surface of AT NPs and AP NPs could be ascribed to the fact that the TA and PA assembled at the outer layer, thereby leading to the stacking and dissociation of multitudinous negatively charged hydroxyl and carboxyl groups40,41. Overall, the small particle size of the nanosystems would endow the large contact area between AZA and the target site, while a negative ζ-potential would also facilitate the translocation through the leaves into the vasculature and further transport to other plant compartments, thus then improving the utilization efficiency and bioavailability of pesticides42.

Fig. 4: Morphological characterization and formation mechanism of AT NPs and AP NPs.
figure 4

a Photographs of the aqueous solutions and Tyndall effect of AT NPs and AP NPs. b and c Hydrodynamic particle size distribution of prepared AT NPs and AP NPs. d ζ-potential values of AZA, TA, PA, AT NPs, and AP NPs (Data are presented as mean values ± SD). e Schematic of detailed possible acting forces driving the formation of the co-assembled AT NPs and AP NPs based on AZA, TA, and PA. Source data are provided as a Source Data file.

Full size image

Formation mechanism

The possible formation mechanism of co-assembled nanosystems is illustrated in Fig. 4e. AZA is a typical hydrophobic compound with low solubility in water at room temperature43. Owing to the high dihydroxyphenyl (catechol) and trihydroxyphenyl (galloyl) content and related multidentate properties of polyphenols, TA was expected to have strong affinity to AZA through multiple non-covalent interactions, such as hydrogen bonding between hydroxyl groups, hydrophobic interactions involving AZA and aromatic rings, and π-π stacking between aromatic rings and carbon-carbon double bond44. When the methanol solution of AZA and TA was added to water, linear macromolecular fragments would be formed through intermolecular hydrogen bonding between hydroxyl groups, then they might spin and intertwine each other to form a spherical three-dimensional structure through the combined actions of hydrophobic and π-π stacking force. In terms structure of PA, it possesses a hydrophobic benzene ring, a carboxyl group, and an amino group, which serve as potential building blocks to trigger molecular co-assembly via hydrophobic interactions, π-π stacking, and hydrogen bonds. Upon adding the mixture of methanol solution containing AZA and PA to the water, the two molecules would form linear macromolecular fragments through hydrogen bonding between the hydroxyl groups of AZA and carboxyl or amino groups of PA. This interaction caused them to rotate and intertwine, ulteriorly forming a spherical three-dimensional structure driven by the combined action of hydrophobic and π-π stacking force. The formation mechanism of the nanosystems was further performed by UV-vis, HPLC, and FT-IR (Supplementary Fig. 2).

Wetting and adhesion properties

Most crop leaves are hydrophobic or superhydrophobic, and the surface of leaves is covered by a hydrophobic waxy layer consisting of macromolecules and lipids (fatty acids, long-chain alkanes, polyols, triterpenoids, etc)45. Consequently, foliar-applied agrochemical sprays need simultaneous good wetting and adhesion properties on such a complex surface of leaves to maximize biological activity. The wetting and adhesion behaviors of pesticide droplets on the target surface were largely affected by the surface tension. As shown in Fig. 5a, the static surface tensions of AZA EC suspensions and deionized water were 64.36 and 72.61 mN/m, which was significantly higher than that of AT NPs (44.7 mN/m) and AP NPs (51.5 mN/m), respectively. This result indicated that the lower surface tension of nanosystems would be beneficial for improving the wet capacity of pesticides46. The spreading performance on hydrophobic Brassica oleracea leaves simultaneously confirmed the good surface activity of the nanosystems (Fig. 5a). The outstanding wetting and spreading capacity could be attributed to the effective coverage and protection of the hydrophobic AZA core by the hydrophilic shell during the co-assembly process. Moreover, the small and uniform size distribution of nanosystems also could contribute to the improvement of dispersion and surface activity, which facilitated their wetting and spreading on the leaf surface.

Fig. 5: Physicochemical properties of AT NPs and AP NPs.
figure 5

a Surface tensions. b Maximum retentions. c Photographs of different leaves (CS: Cucumis sativus; BO: Brassica oleracea and ZM: Zea mays) retentions at different treatments of AT NPs and AP NPs. d Schematic illustration of spray droplets and proposed interaction mechanism of AT NPs and AP NPs on the surface of target crop for the improvement of physicochemical properties in the field. Rain washing (e) and photostabilities (f) of AT NPs and AP NPs. (n = 3 independent experiments; Data are presented as mean values ± SE; Statistical significance was defined by one-way ANOVA followed by Duncan’s multiple comparison tests (p < 0.05); Different letters within each column indicate statistical differences between each pesticide treatment at same conditions). Source data are provided as a Source Data file.

Full size image

In addition to the improved wettability and spreading the increased foliage retention is another essential factor affecting effective utilization47. The retention properties of nanosystems were studied by measuring their max retention amounts on the surface of different crop leaves. Figure 5b, c showed the max retention amounts of nanosystems on the leaves of three model crops: hydrophilic cucumber (Cucumis sativus, CS), intermediate hydrophilic maize (Zea mays, ZM), and hydrophobic wild cabbage (Brassica oleracea, BO), respectively. The retention amounts of AT NPs and AP NPs were significantly improved in comparison to the AZA EC suspension on three crop leaves due to the decreased surface tension and improved spreading performance of the droplets. The co-assembled AT NPs showed better wettability and adhesion properties than AP NPs, which might be due to the more hydrogen bonds formed by phenolic hydroxyl groups on AT NPs with fatty acids and alcohols on the waxy layer of the leaves (Fig. 5d)48. These results suggested that the co-assemblies of TA or PA to the AZA markedly improved the wettability and adhesivity of the nanosystems on the crop leaves, which would be beneficial to achieving the maximum efficacy and prevention of environmental contamination.

Rain erosion resistance

Since rainfall can reduce pesticide retention on crop foliage to a pitiful 0.1–10% after spraying pesticides, rainwater erosion is a significant contributor to the loss of pesticide droplets during deposition49. Therefore, improving the resistance ability to wash off and deposition of pesticide droplets are prospective approaches to enhance the bioavailability of pesticides. Figure 5e shows the retention rates of AZA EC, AT NPs, and AP NPs on maize leaves were 30.29%, 67.86%, and 56.91% after simulated rainfall washout of 50 mL volume, respectively. The results indicated that a substantial amount of AT NPs and AP NPs remained on the leaves, whereas a significantly smaller quantity of AZA EC was retained. The remarkable rain erosion resistance was possibly attributed to the strong hydrogen bonding interactions between hydroxyl, amino, and carboxyl groups of nanosystems with fatty acids, fatty alcohols, and fatty aldehydes on the waxy layer of the leaves (Fig. 5d). Moreover, the smaller particle size of nanosystems could be easily embedded in the microstructure and nano mastoid of the plant leaf surface, making it significantly more resistant to washing than the AZA EC50.

Photodegradation by UV light

Studies showed that the direct exposure of AZA to sunlight could lead to its degradation, and the photodegraded AZA is less potent against pests28. Therefore, the photostabilities of AZA EC and nanosystems were investigated. The results illustrated that the retention rates of AZA under UV light decreased with irradiation time and the degradation curves were fitted well to a first-order kinetic model (Fig. 5f). The photodegradation half-life (DT50) value of AZA EC was 18.50 min, which illustrated that the AZA was unstable under UV light irradiation. The prepared AT NPs and AP NPs prolonged the photodegradation of AZA nearly 1.66 and 3.04 times with DT50 values of 30.62 and 56.19 min, respectively, indicating a significant improvement in the photostability of AZA against UV. The remarkably improved photostability could be ascribed to the UV-shielding capabilities of the absorption maximum of the aromatic ring π-π transitions, which might act as a shield to protect AZA from premature degradation under UV light (Fig. 5f)51.

pH-mediated release of AZA

The midguts of most hemipteran pests are normally slightly acidic with a pH value of 5.0–6.0, while those of phytophagous lepidopteran pests possess a unique alkaline gut (pH level up to 12), which can be exploited as a biological trigger to achieve precise and long-lasting pesticide release52,53. Figure 6 and Supplementary Fig. 3 displayed the release behaviors of AZA from AT NPs and AP NPs under different pH conditions and mathematic models. At pH 7.0, the cumulative release rates of AZA from the AT NPs and AP NPs were only 12.29% and 20.57% at 96 h, indicating good stability at neutral conditions. At pH 5.0 and 9.0, there was a fast release during the first 12 h, and then the release of AZA from the nanosystems showed a gradual process (Supplementary Fig. 3a, b). After 96 h of treatment, the cumulative release rates of AZA from the AT NPs and AP NPs increased gradually with time and reached 65.57%, and 89.31% at a pH value of 5.0, respectively. Due to the attack of acidic conditions, the hydroxyl groups of TA and the amino group of PA were gradually protonated and then led to the instability and breakage of H-bonding, and thus resulted in the pH-induced release of AZA from nanosystems (Fig. 6i)54,55. At pH 9.0, the cumulative release rates of AZA from the AT NPs and AP NPs at 48 h were 76.06% and 93.51%, respectively. This could be explained that above the pKa value (about 8.5) of TA, the AT NPs were gradually disintegrated due to deprotonation of the hydroxyl groups on TA and the breakdown of the physical interactions between AZA and TA at alkaline conditions56.

Fig. 6: Release behaviors of AT NPs and AP NPs.
figure 6

Kinetic models (First-order, Ritger-Peppas, Higuchi, and Peppas-Sahlin) for fitting curves of the release profiles of AZA from AT NPs (a–d) and AP NPs (e–h) under different pH values. i Schematic illustration and proposed mechanism of the release behaviors of AZA from AT NPs and AP NPs in the different targets. (n = 3 independent experiments; Data are presented as mean values ± SE). Source data are provided as a Source Data file.

Full size image

All results showed that more AZA was released from nanosystems under acidic or alkaline conditions in a one-step or stepwise process, indicating a bidirectional pH responsiveness release behavior to microenvironmental stimuli. The release data of AZA from AT NPs (Fig. 6a–d) and AP NPs (Fig. 6e–h) were further quantified by several common mathematical models to address the preliminary kinetics. The results showed that the obtained release curves fitted best for the Higuchi models, in which the regression coefficient values (R2) ranged from 0.879 to 0.995 (Supplementary Table 2). This data suggested that the predominant mechanisms of the release kinetics were dissolution and diffusion, and the accumulative amount of AZA released from nanosystems was proportional to the square root of times57. The diffusion exponent (n) values of AZA from nanosystems fitted by the Ritger-Peppas model were less than 0.45 under acidic or alkaline conditions, suggesting that the release process was controlled by a Fickian diffusion58.

In vitro insecticidal activity

As shown in Supplementary Fig. 4a, a leaf-disc-mediated dipping approach was used to evaluate the in vitro bioactivities of nanosystems against Ostrinia furnacalis. The results suggested that treatments of nanosystems significantly decreased the survival rates of larval on the maize leaves after treatment of 48 h (Fig. 7a) and 96 h (Fig. 7b), with a dose-dependent relationship at a concentration of 25–500 mg/L. The remarkable reduction might be ascribable to the marked antifeedant activity of AZA against O. furnacalis. After 96 h, the mortality of the larval reached 100% at the AZA concentration of 500 mg/L (Supplementary Fig. 4b), which indicated that AZA had no quick-acting effect on the O. furnacalis. The results could be consistent with the previous studies that larvae of O. furnacalis died from severe metabolic and physiological aberrations rather than the direct toxic effect of AZA after the treatment of 2–7 days59. Besides, the decrease in average fresh weight, as well as growth retardation and impaired larval were observed after the treatments of nanosystems at 96 h. There was a significant difference in the inhibition rate of fresh weight of larval among AZA EC, AT NPs, and AP NPs treatments (Fig. 7c), indicating a synergistic effect between the released AZA and TA or PA (Fig. 7d).

Fig. 7: Insecticidal activities of AT NPs and AP NPs.
figure 7

Mortality rates of Ostrinia furnacalis at different concentrations of AZA at 48 h (a) and 96 h (b); c Inhibition rates of fresh weight of pest treated after 96 h. d Schematic illustration and proposed mechanism of the synergistic effect of AT NPs and AP NPs. e and f In vivo control efficacies of AT NPs and AP NPs against Aphis gossypii on the cucumber. (n = 30, numbers of pests for each treatment, independently repeated for four experiments; Data are presented as mean values ± SE; Statistical significance was defined by one-way ANOVA followed by Duncan’s multiple comparison tests (p < 0.05); Different letters within each column indicate statistical differences between each pesticide treatment at same concentrations). Source data are provided as a Source Data file.

Full size image

In vivo control efficacies of AZA

A leaf-spraying approach was used to evaluate the in vivo control efficacies of nanosystems against A. gossypii on cucumber seedlings (Supplementary Fig. 4c). The results demonstrated that the application of the nanosystems could effectively mitigate the infestation of A. gossypii on cucumber seedlings (Fig. 7e, f). The mortalities of A. gossypii on treated cucumber leaves were 59.07 and 69.73% for AT NPs and AP NPs, which were significantly higher than that of AZA EC (43.56%), indicating a synergistic control efficacity. For the treated A. gossypii on cucumber flowers, AT NPs and AP NPs exhibited significantly improved control efficacies than that of AZA EC (70.71%) with mortality of 79.41% and 86.67%, which further corroborated a synergistic effect between the released AZA and TA or PA. The remarkably improved bioactivities could be ascribed to the potentiated wettability, adhesivity, and UV resistance of co-assembled nanosystems. Moreover, the released TA and PA from AT NPs and AP NPs were already identified as potential glutathione transferase activity inhibitors and inducers of plant-insect interaction mediators, which could display synergistic effects with AZA8,60.

Biosafety evaluation

The effects of nanosystems on plant growth and development can be negative or positive depending on their structure, size, and dosage. In this study, the effects of AZA EC, TA, PA, AT NPs, and AP NPs on soybean (Glycine max L.) seedlings at different concentrations under greenhouse conditions were evaluated. As indicated in Fig. 8a, the results demonstrated that there were no morphological abnormalities observed in the above‐ground parts of treated soybean seedlings, which indicated that the application of nanosystems had no negative impact on the plants. The chlorophyll content and height growth of treated crop seedlings were further applied to examine the relevant physiological condition after different treatments. Compared to the blank treatment (CK), there was no significant difference (p > 0.05) in the chlorophyll content and plant height of soybean seedlings under various concentrations of AZA at 7 d (Fig. 8b, c), indicating that nanosystems had no adversely impact on the photosynthesis and growth of plants. The same phenomenon was also demonstrated in the treatments of Chinese cabbage (Brassica pekinensis L.) seedlings (Fig. 8d, e). The combined results revealed that there was no difference between seedlings treated with different concentrations of AZA and the control group, which indicated that nanosystems exhibited high biosecurity and could be used for integrated pest management.

Fig. 8: Effect of AT NPs and AP NPs on crop seedlings.
figure 8

a Picture of soybean seedlings treated with AZA at different concentrations after seven days. The chlorophyll content (b) and plant height (c) of treated soybean seedlings. d Picture of Chinese cabbage seedlings treated with AZA at different concentrations after seven days. e The chlorophyll content of treated Chinese cabbage seedlings. (n = 4, numbers of plants for each treatment, independently repeated for three experiments; Data are presented as mean values ± SE; Statistical significance was defined by one-way ANOVA followed by Duncan’s multiple comparison tests (p < 0.05); Different letters within each column indicate statistical differences between each pesticide treatment at same concentrations). Source data are provided as a Source Data file.

Full size image

In summary, TA, PA, and GA with different regulative models for plants and physicochemical characteristics (UV shielding capability and hydrophilic property) were envisaged and preselected as versatile “green” building blocks to modify the hydrophobic AZA into co-assembled nanosystems. The results showed that AZA was successfully co-assembled with TA or PA into nanoparticles (AT NPs or AP NPs) through a straightforward solvent exchange method without the usage of any potentially hazardous additives. Combined with the co-assembly rate, particle size, ζ-potential, and morphological characteristics, the optimal molar ratio for the successful fabrication of AT NPs was selected as 1: 8 (TA: AZA) with co-assembly rate of 82.46% and loading rate of 77.22%, while that of AT NPs was selected as 4: 1 (PA: AZA) with co-assembly rate of 70.96% and loading rate of 52.17%, respectively. The DLS and TEM images indicated that the obtained AT NPs and AP NPs had a classical core-shell spherical structure with diameters of ~ 120 and 280 nm, low PDI values of 0.029 and 0.129, and high negative charges of − 36.40 and − 21.01 mV, respectively, indicating good stability and dispersibility of nanosystems. The prepared nanosystems exhibited excellent physicochemical properties, including low surface tension, high max retention, excellent rainfastness, and outstanding photostability, which markedly improved the wettability, adhesivity, and retention of AZA on crop leaves. Importantly, the nanosystems showed bidirectional pH responsiveness, enabling their disassembly in both acidic and alkaline solutions, where the protonation of hydroxyl, amino, and carboxyl groups mediates the disassembly kinetics, and subsequently allowed for the controllable release of AZA in a one-step or stepwise process, The in vitro and in vivo insecticidal activities indicated that nanosystems had a synergistic effect against O. furnacalis and A. gossypii. The safety effects on plant growth demonstrated that nanosystems displayed good safety in soybean and Chinese cabbage seedlings. Therefore, the presented findings would provide an environmentally friendly way for the efficient utilization of biopesticides.

Methods

Materials

Asian corn borer (Ostrinia furnacalis) insects used in this study were obtained from the laboratory on a regular artificial diet for multiple generations in the College of Plant Protection, China Agricultural University. O. furnacalis colony larvae were maintained under controlled conditions (24 ± 1 °C, 70 ± 5% relative humidity, and a photoperiod of 16: 8 (L: D)). Aphis gossypii was naturally grown on cucumber seedlings without using any pesticides in a greenhouse at 24 ± 3 °C and 65 ± 5% RH with a photoperiod of 13/11 h (L: D).

Preparation of AZA based co-assembled nanosystems

The methanol or water-dissolved AZA (0.274 g), TA (0.255 g), PA (0.027 g), or GA (0.026 g) solutions (2 mL) were combined with different molar ratios and added drop by drop to an aqueous solution by magnetic stirring at 500 rpm for 5 min. For the removal of the non-assembled AZA, TA, or PA, and organic solvent, the resultant solution was dialyzed for 24 h in dialysis bags. The resulting solution that was left in the dialysis bag was then freeze-dried to obtain AT NPs and AP NPs. The co-assembly rate (CAR) of AT NPs and AP NPs after the dialysis was analyzed by HPLC and calculated according to the following Eq. (1).

$${\mbox{CAR}}\left(\%\right)=\frac{{{\mbox{A}}}_{{{\rm{t}}}}}{{{\mbox{A}}}_{0}}\times 100$$
(1)

where A0 represents the initial amount of AZA, and At represents the amount of AZA retained in AT NPs and AP NPs after the dialysis.

Surface tension and max retention

The surface tension of samples was measured by the platinum plate method on the Surface Tensiometer (Model: DST-30) at ambient temperature61. The deionized water was used to calibrate the instrument. The needle and the dispensing system were regularly rinsed with distilled water to prevent the impact measurements of residues. Simultaneously, the spreading performance of samples on the super hydrophobic Brassica oleracea leaves with the same surface chemistry and roughness was roughly tested. For further investigation of wettability and adhesivity, the maximum retention amounts of samples on different leaves were obtained using micro weighing and dipping methods62. Three crop leaves cucumber (Cucumis sativus, CS), maize (Zea mays, ZM), and wild cabbage (Brassica oleracea, BO) were chosen as the model leaves representing hydrophilic, intermediate hydrophilic, and hydrophobic crop leaves. The initial weight (Wi) of the fresh leaf was computed by an electronic analytical balance, and the leaf was immersed into 20 mL of sample aqueous suspensions with a tweezer for 30 s. Then, the leaf was quickly pulled out from the solutions, hung in the air for 120 s until no liquid drop was dripping, and weighed eventually (We). The treated leaves were photographed, and the coverage area (S) of leaves was quantified using ImageJ software. Three parallel measurements were replicated. The maximum retention (Rm) on the leaf per unit area was calculated by Eq. (2).

$${{\mbox{R}}}_{{{\rm{m}}}}=\frac{{{\mbox{W}}}_{{{\rm{e}}}}-{{\mbox{W}}}_{{{\rm{i}}}}}{{\mbox{S}}\times 2}$$
(2)

Laboratory-scale rain wash-off

The deposition and retention of pesticides on plant surfaces are critical challenges for modern precision agriculture, which directly affect bioavailability, efficacy, and the loss of pesticides in spray applications. Therefore, the rain fastness of the samples on the plant leaf surface was measured quantitatively by simulating the rainwashing method63. Briefly, 100 μL of sample suspension (1 mg/mL) was uniformly smeared onto the surface of fresh maize leaves fixed on the slides and left the leaves to dry at room temperature for 3 h. The treated maize leaves (all at an inclination angle of 30°) were then rinsed vertically at a scouring speed of 10 mL/min by deionized water from a scouring height of 10 cm (to simulate natural rainfall conditions). After washing, the simulated rainwater was collected into a centrifugal tube and detected by HPLC. To ensure accuracy, the experiments were repeated three times and the average value was calculated. The retention rate (RR) of AZA after washout was calculated by the following formula (3):

$${\mbox{RR}}\left(\%\right)=\frac{{{\mbox{C}}}_{0}{\times {\mbox{V}}}_{0}-{{\mbox{C}}}_{{{\rm{t}}}}\times {{\mbox{V}}}_{{{\rm{t}}}}}{{{\mbox{C}}}_{0}\times {{\mbox{V}}}_{0}}\times 100$$
(3)

where C0 and Ct represent the concentration of AZA on the maize leaves before and after rinsing, respectively. V0 and Vt represent the volume of AZA on the maize leaves before and after rinsing.

Photodegradation by UV light

As a biopesticide, AZA is sensitive to UV light and prone to premature photodegradation, which results in short persistence and restricts its effective utilization. Given that TA and PA are natural broad-spectrum sun blockers due to their various UV-absorbing functional groups, the light stabilities of AZA nanosystems were determined by the germicidal lamp method64. Briefly, quartz glass tubes containing 40 mL of AZA water suspension (10 mg/L) were exposed to a UV germicidal lamp (36 W; Wavelength: 254 nm) with a distance of 20 cm. After predetermined intervals, 0.5 mL of the sample was sampled from each tube and the same amount of water was added. The contents of AZA in each sample were detected by HPLC. All of the samples were repeated three times. The first-order model was used to fit the results of the photodegradation experiment and the following Eqs. (4) and (5) were used to calculate the half-life of the degradation value (DT50).

$${{\mbox{C}}}_{{{\rm{t}}}}={{\mbox{C}}}_{0}\,{{{{\rm{e}}}}}^{-{{{\rm{kt}}}}}$$
(4)
$${{\mbox{DT}}}_{50}={{{\mathrm{ln}}}}\,2/{{{\rm{k}}}}$$
(5)

where Ct represents the concentration of the AZA at the moment of irradiation t, C0 represents the incipient concentration of the AZA, and k represents the degradation rate constant.

Controlled release behavior

Depending on the insect species, larval midgut conditions of pests can be very acidic or strongly alkaline with a wide pH range values of 2–1252. The release behaviors of AZA from the AT NPs or AP NPs were investigated with the dialysis method at different pH conditions (pH = 5.0 (slightly acidic), pH = 7.0 (neutral condition), and pH = 9.0 (slightly alkaline)) to mimic the field conditions that AZA nanosystems are exposed to when applied to plant leaves. Briefly, about 5 mL of AT NPs or AP NPs suspension was added into dialysis bags, and then immersed into 100 mL of the release medium (phosphate buffer solutions, pH = 5.0, 7.0, or 9.0) in a 200 mL brown beaker at 25 °C. Then the beaker was sealed with a sealing film to prevent volatilization, and shaken on a shaker with a speed of 100 r/min. At predetermined time intervals, 0.5 mL of the release medium was withdrawn from the beaker, and 0.5 mL of fresh media solution was replenished into the jar. The sampled release medium was filtered through a syringe filter with a 0.22 μm pore size. The cumulative release (CR) concentration of AZA from AT NPs or AP NPs in the different release mediums was determined using HPLC. Three replicates were included in each treatment and averaged. The CR was calculated with Eq. (6) as follows:

$${{{\rm{CR}}}}(\%)=({{{{\rm{C}}}}}_{{{{\rm{t}}}}}{{{{\rm{V}}}}}_{{{{\rm{total}}}}}+\sum \limits_{0}^{{{{\rm{t}}}}-1}{{{{\rm{C}}}}}_{{{{\rm{t}}}}}{{{{\rm{V}}}}}_{{{{\rm{t}}}}})/{{{{\rm{m}}}}}_{0}\times 100\%$$
(6)

where Ct stands for the concentration of AZA released from AT NPs or AP NPs at the time t, Vtotal stands for the total volume of media (100 mL), Vt stands for the volume of the media taken out (0.5 mL) at time t, and m0 stands for the total amount of AZA.

The first-order, Ritger-Peppas, Higuchi, and Peppas-Sahlin models were applied to assess the release data, respectively. The specific Eqs. (7–10) are as follows:

$${{{\rm{First}}}}-{{{\rm{order}}}}\,{{{\rm{model}}}}\!\!:\,\frac{{{{{\rm{M}}}}}_{{{{\rm{t}}}}}}{{{{\rm{Mz}}}}}=1-{e}^{-kt}$$
(7)
$${{{\rm{Ritger}}}}-{{{\rm{Peppas}}}}\,{{{\rm{model}}}}\!\!:\,\frac{{{{{\rm{M}}}}}_{{{{\rm{t}}}}}}{{{{\rm{Mz}}}}}=k{t}^{n}$$
(8)
$${{{\rm{Higuchi}}}}\,{{{\rm{model}}}}\!\!:\,\frac{{{{{\rm{M}}}}}_{{{{\rm{t}}}}}}{{{{\rm{Mz}}}}}={{{{\rm{kt}}}}}^{0.5}+{{{\rm{b}}}}$$
(9)
$${{{\rm{Peppas}}}}-{{{\rm{Sahlin}}}}\,{{{\rm{model}}}}\!\!:\,\frac{{{{{\rm{M}}}}}_{{{{\rm{t}}}}}}{{{{\rm{Mz}}}}}={{{{\rm{k}}}}}_{1}{{{{\rm{t}}}}}^{{{{\rm{n}}}}}+{{{{\rm{k}}}}}_{2}{{{{\rm{t}}}}}^{{{{\rm{2n}}}}}$$
(10)

where Mt/Mz is the percentage of AZA released from the AT NPs or AP NPs at time t; k, k1, and k2 are the kinetic constant; and n is the diffusional exponent that can be related to the mode of the drug transport mechanism. For the spherical samples, n ≤ 0.45 corresponds to a Fickian diffusion, 0.45 < n < 0.89 is related to a non-Fickian or anomalous diffusion, and n ≥ 0.89 is related to a zero-order transport.

In vitro insecticidal activity

A leaf-disc-mediated dipping bioassay approach was used to evaluate the in vitro bioactivities of the developed AZA nanosystems against the O. furnacalis65. Freshly reared and uniform-sized 2nd instar O. furnacalis (n = 30) were collected and starved for 4 h to be used in the experiment. Fresh maize leaves (two months) without using any pesticides were collected, clipped into oblong shapes of a certain size, and impregnated in AZA suspensions with different final concentrations of AZA (25, 50, 100, 250, and 500 mg/L) for 15 s and air dried naturally. Treated leaves were placed over in a culture dish (9 cm) which contained wet filter paper, and the starved test larvae were released. AZA EC at the same concentration gradient and deionized water were used as the positive and negative controls, respectively. Four treatments were repeated independently for every 30 larvae. The mortality data (%) of O. furnacalis were collected and recorded at 48 and 96 h intervals posttreatment, and inhibition levels of pest fresh weight (%) were calculated.

In vivo insecticidal activity

The insecticidal activity of the AZA nanosystems against A. gossypii on cucumber seedlings was investigated using the leaf-spraying method in the greenhouse of China Agricultural University66. The commonly used intermediate concentration of 50 mg/L was adopted to study the insecticidal performance of AZA EC, AT NPs, and AP NPs. The pesticide suspensions were uniformly sprayed onto the leaf surfaces and flowers of cucumbers infested with aphids using a plastic spray tower. The same amount of water with 0.1% Triton-100 was sprayed into the blank control. The experiment was independently repeated quadruple for each treatment, and at least thirty A. gossypii were used per replicate. The mortality of A. gossypii in each treatment was recorded and compared with the control after the treatment of 7 d.

Biosafety

Soybean and Chinese cabbage seedlings were used to evaluate biosafety of the prepared co-assembled nanosystems at concentrations of AZA of 5 and 10 mg/L. The seedlings pots were kept in the greenhouse at 20 ± 5 °C and 70 ± 5% RH. The same amount (20 mL) of water was added to the blank control pots. Four plants were treated each time, and each treatment was independently repeated three times. The effect of AZA on crop growth was investigated 7 d after spraying, the chlorophyll content of leaves was measured using SPAD-502 Plus chlorophyll meter (Konica Minolta C., Japan), and the height from the soil surface to the growing point was measured as the plant height.

Statistical analysis

The SPSS 23.0 statistical analysis software (SPSS, Chicago, IL, USA) was applied for statistical analysis. The data was analyzed by the one-way ANOVA followed by Duncan’s multiple comparison tests (p < 0.05) and expressed as the mean ± standard error (SE) for all the experiments.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.