Introduction

Diseases and pests in agriculture have played a major role in causing famines historically1. As a consequence, there has been a significant surge in the use of chemical pesticides over the last couple of decades in order to ensure food security for the ever-growing human population2. While chemical pesticides help maintain food security, they pose significant risks to human health, biodiversity3, and the ecosystem4. Additionally, their prolonged use contributes to the development of pesticide resistance over time5. There is a pressing need to switch to more environmentally-friendly techniques to prevent the immediate and long-term effects of pesticide exposure.

In order to develop alternative pest management techniques, extensive research has already been conducted to improve the performance of chemical pesticides by controlling the release rate6, reducing the toxicity of solvents7, and increasing retention on the leaf surface8,9,10. Significant research revolving around using purely mechanical methods11,12,13 and also the increasing potential of nanoparticles use14,15,16,17 for pest control has been explored. In the meantime, integrated pest management has emerged to be an ecologically conscious alternative involving the coordinated use of various strategies such as biological control with natural enemies, cultural practices, and selection of less harmful chemicals.

In this regard, biomimetics remains an untapped realm for the design of innovative non-toxic pesticides that can be incorporated easily into the currently used techniques of integrated pest management13,18,19. Biomimetics refers to the development of new materials based on imitating the functioning of natural systems and processes20. Nature has proven to be far superior in several areas in comparison to human capabilities, owing to the process of evolution. Natural self-defense is one such area where plants have developed various mechanisms to protect themselves from herbivores. These include physical defense mechanisms like the presence of thorns and spines, and chemical defense mechanisms where toxins can either repel or kill attacking pests or even attract their natural predators21.

Trichomes are one such natural plant defense mechanism present in some plants like wild tomatoes, potatoes etc., which are hair-like structures formed as a result of extensions from the leaf epidermis22. Trichomes can be classified into glandular or non-glandular based on their structure and defense mechanism. Glandular trichomes release chemical compounds from their heads that can resist the attack of pests. A notable example is evident in wild potatoes, which feature two distinct types of glandular trichomes denoted as type A and B. Incoming pests usually encounter the longer type B trichomes, which release an adhesive substance, thus hindering the motion of the insect. As a natural reflex action, the insect tries to escape, and during this struggle, it comes into contact with the shorter type A trichomes owing to their high density and eventually gets immobilized by its secretions23,24. This intricate defense mechanism showcases the dynamic interplay between trichome structure and chemical composition. The chemical makeup of these adhesive secretions from different trichomes varies across plant types. In the Solanum genus (wild potatoes) for instance, phenolic compounds and polyphenol oxidase enzymes contribute to the adhesive properties of its trichomes24. In carnivorous Roridula plants, a resin-based water-insoluble mucin is employed for pest capture, highlighting the adaptability of nature to produce different secretions25,26. Another kind of carnivorous plants, the Drosera species as shown in Fig. 1a, adopts a similar strategy to trap pests. These trichomes also secrete a sticky mucilage to capture their prey along with enzymes for them to digest the captured prey. The glistening drop is suspected to act as an attractant to flying pests, thus unknowingly leading them to their death. In contrast to the resins present in Roridula, the Drosera exudates were found to consist of a 4 wt% acidic polysaccharide with a molecular weight greater than 2 MDa composed of sugars like arabinose, xylose, mannose, galactose and glucuronic acid, myo inositol, inorganic cations like Ca2+, K+, Na+, and a notable water fraction of 95%27,28,29,30,31,32. Upon further investigation of this secretion, there is speculative consideration that its constituents may manifest as a natural deep eutectic solvent (NaDES)33,34.

Fig. 1: Overview of inspiration and proposed biomimetic concept.
figure 1

a Natural trichomes of Drosera capensis. b Components of the sugar-based natural deep eutectic solvent used in this study. From left to right: glucose, fructose, sucrose, water, all components mixed (1:1:1:11 mole ratio), and NaDES formed after heating at the same ratio. c Schematic illustration of the proposed trichome biomimetics and trapping mechanism.

Full size image

Deep eutectic solvents (DES) are a class of non-toxic and eco-friendly solvents currently emerging as an alternative to traditional solvents. Often considered to be a sub-class of ionic liquids, DES exhibits unique properties typically formed by the combination of hydrogen bond acceptors (HBA) and hydrogen bond donors (HBD). The combination of these components in a specific molar ratio results in a mixture with low volatility and a melting point significantly lower than that of each individual component, forming a liquid at room temperature as shown in Fig. 1b. NaDES are a sub-class of DES, where the individual components are derived from natural substances such as sugars, amino acids, water, organic acids and bases. Pioneering work on sugar-based NaDES, was performed by Choi and coworkers, based on the presence of high concentrations of sugars in several plant saps observed through NMR-based plant metabolomics35. Water, which can also be added as a constituent in the preparation of NaDES, is strongly bound and does not evaporate easily. Thus, such NaDES compositions could act as excellent components to hold water in the trichome biomimetics.

Drawing inspiration from the adhesive mucilage of Drosera, we developed a non-toxic and biodegradable physical pesticide by combining the eco-friendly characteristics of NaDES with the naturally sourced biopolymer, hyaluronic acid (HA). To the best of our knowledge, this is the first account of a water-based biomimetic adhesive designed for pest control that closely mimics the natural composition of plant mucilage, consisting of a sugar-rich material and a long-chain polysaccharide. Existing synthetic adhesives for pest control, such as sticky paper traps, consist of pre-coated surfaces that capture insects away from the plant. In contrast, our material is directly applied to plant surfaces, placing the adhesive exactly at the main site of action where pests interact with the plant, providing a more effective and adaptable solution, along with the added benefits of biodegradability. These trichome biomimetics are meticulously designed and optimized using rheological measurements to be sprayed on greenhouse plants directly to mimic the high density of natural trichomes as shown in Fig. 1c. These biomimetic materials maintain a stable composition influenced by the relative humidity (RH) of the surroundings, thus exhibiting stable adhesive properties for up to a week. Indentation experiments were performed to quantify the adhesion performance. This proposed alternative was found to act as an effective deadly trap for the target pest - western flower thrips (Frankliniella occidentalis), owing to their ability to be sprayed in a high density facilitating trapping by multiple contacts with the insect’s body in addition to the inherent adhesiveness of the material itself. The water-based system demonstrates a reversible response to RH changes, regaining adhesive properties after being exposed to elevated humidity and reverting to the desired levels. Importantly, the system is designed to be washable. This key attribute addresses concerns about sticky residues on produce, making it an attractive and consumer-friendly solution. This innovative approach not only offers a promising alternative to conventional pesticides, but also aligns with the growing emphasis on sustainable agricultural practices.

Results and Discussion

Material composition

The primary consideration in developing an eco-friendly alternative to traditional pesticides is the careful selection of ingredients to ensure environmental compatibility and biodegradability. For this purpose, a sugar-based NaDES was chosen, drawing inspiration from the natural exudates of Drosera trichomes, which comprise a mixture of various naturally occurring sugars. A combination of glucose, fructose, sucrose, and water in a molar ratio of 1:1:1:11 (GFSH) was selected since this NaDES was found to maintain its water content consistently over time. In contrast, mixtures with different molar ratios, namely glucose, fructose, and water at 1:1:11 (GFH) and fructose and water at 1:11 (FH), were observed to undergo water loss as shown in Supplementary Fig. S4, indicating the presence of large amounts of free unbound water in the material. The inclusion of three sugars in GFSH and its resulting higher viscosity in comparison to GFH and FH was also expected to aid the adhesive properties of the trichome biomimetics. While sugars are recognized for their inherent stickiness, their tack arises mainly from their high viscosity and resulting capillary forces, and thus, all the applied stress gets dissipated quickly. To create a robust adhesive capable of effectively immobilizing insects, the material must be able to sustain an applied stress. Consequently, a naturally sourced polysaccharide, hyaluronic acid, was incorporated to impart elasticity to the material. In Drosera trichome exudates, the polysaccharide present naturally was found to be greater than 2000 kDa. However, to balance the requirement for long molecular chains with processability, 750–1000 kDa was chosen as a suitable compromise. The optimization of these viscous and elastic components, along with water, is essential to ensure that the critical attributes enabling these environmentally-friendly trichome biomimetics to function as innovative alternatives are met. This includes their ability to be sprayed into fine droplets, regulate water content and adhesiveness over time, and trap target pests - topics that will be elaborated upon in subsequent sections.

Viscoelastic properties

Sprayability and easy processability are integral aspects of the material to be feasible for acting as an alternative to conventional pesticides. The material must be finely sprayed as discrete droplets, ensuring uniform coverage while preventing nozzle clogging. It is also crucial to find the right balance between viscosity and elasticity to achieve good adhesiveness and insect trapping after spraying. To impart these characteristics in our trichome biomimetics, it was aimed to grasp and better understand the dynamic behavior of these systems using rheological measurements. Rheology provides insights into the viscoelastic properties that govern flow behavior and adhesion, which influence sprayability and efficiency. As shown in Fig. 2, the linear rheological properties of samples with increasing concentrations of HA, namely 0.5, 1, and 2 wt%, in the prepared water to GFSH ratios, namely 0:100, 25:75, and 35:65, were compared. As depicted in Figs. 2a–c, within the measured frequency window, we observe that all samples with 0:100 solvent ratio act as viscoelastic solids irrespective of the polymer concentration. Additionally, for the same solvent ratio of 0:100, across panels 2a–c, we notice with increase in HA content, there is a rise in the moduli values, broadening of the entanglement plateau, and a consequent shift of the characteristic crossover frequency, ωe (ωe = 1/τe, where τe is entanglement relaxation time that depicts the onset of physical constraints among the polymer chains) towards higher values. These features all indicate that an increase in HA concentration makes the system more solid-like and less dynamic.

Fig. 2: Viscoelastic behavior of HA in GFSH samples.
figure 2

Linear viscoelastic spectra of HA dissolved in GFSH expressed in terms of storage modulus G′ and loss modulus G″ as a function of the angular frequency ω for a 0.5 wt%, b 1 wt%, and c 2 wt% HA. Grey lines indicate the entanglement plateau, and open circles indicate the characteristic crossover frequency, ωe, for 0:100 solvent ratio. The viscoelastic spectra normalized across the angular frequency ω with respective solvent viscosity (ηsolvent) across corresponding polymer concentrations of 0.5 wt%, 1 wt%, and 2 wt% are depicted in d, e, and f, respectively.

Full size image

As seen in Fig. 2a, at 0.5 wt% HA content, the solvent ratio of 25:75 reduces the moduli values and results in a completely different rheological response, i.e., a viscoelastic liquid as opposed to the solid behavior observed at 0:100 ratio. Additionally, at 25:75 solvent ratio, ωe is no longer visible in the frequency window, instead, the system depicts a crossover frequency ωd (ωd = 1/τd, where τd is the longest relaxation time, where the polymer chain has fully overcome its physical constraints). Moreover, on further increase in the fraction of water, at 35:65 solvent ratio, we observe an additional drop in the moduli values and a shift of ωd towards higher frequencies. Subsequently, as seen in Fig. 2b-c, similar dynamic behavior of decrease in moduli values and appearance of ωd is also observed for 1 wt% and 2 wt% with a change in solvent ratio.

Additionally, by normalizing the frequency axis with the respective solvent viscosity (ηsolvent) values (Supplementary Fig. S5a) across Fig. 2a – c, one can study the overall behavior of the polymer chains without the solvent. Interestingly, Figs. 2d–f depict the normalized mechanical spectra of HA dissolved in GFSH. Impressively, the viscoelastic response across different dilutions gives rise to master curves that depict dynamic behavior over six decades of frequency. The water molecules seem to solely lubricate and soften the friction between two HA monomeric units, facilitating faster dynamics without changing the mode of relaxation or causing any other change in the interaction between the HA chains. Simply put, the collapse of the data points onto a single master curve upon normalization with solvent viscosity implies that the addition of water to GFSH solely reduces the viscosity of the medium in which the HA chains are dissolved. Moreover, across Fig. 2d–f, a clear broadening of the entanglement plateau is observed, as expected with increasing polymer content. In addition, the plateau almost depicts power law behavior instead of being frequency-independent, implying the polydisperse nature of the HA polysaccharide, which is also indicative of fewer entanglements per chain. Moreover, on studying the longest relaxation time, τd (=1/ωd) and plateau modulus, Gp as a function of HA concentration (Supplementary Fig. S5b), the observed scaling is evident of an entangled neutral polymer in θ-solvent. In conclusion, HA when dissolved in GFSH behaves as a non-associating polymer, with the addition of water merely diluting the system without imparting any physical changes to the HA chains.

Having studied the mechanical spectrum and phase behavior of HA in GFSH, we can better understand the role of water dilutions and can now move towards non-linear rheology to optimize the material processability and sprayability. Figures 3a–c portray the flow curves of HA dissolved in GFSH across varying polymer concentrations. A clear observation across all panels is the shear-thinning behavior of the systems. A characteristic feature of samples with 0:100 solvent ratio is the relatively high shear viscosity values, at high shear rates, which hinder the final application of spraying. Although the shear rates experienced by a material flowing through a nozzle just before atomization have been reported to be relatively high (>1000 s−1)36, the viscosities reached are still not low enough due to a weak shear-thinning behavior. Hence, these samples do not satisfy the required criteria since they get ejected rather than atomized, even at such a high shear rate with high applied pressures. As seen in Fig. 2, the addition of water to the solvent mixture progressively lowers the viscosity, facilitating better sprayability to the material. The 1H NMR spectra (400 MHz) of varying ratios of water:GFSH solvent were used to identify the maximum ratio that could be investigated while retaining the eutectic properties, as shown in Supplementary Figs. S6 and S7. As already seen in Fig. 2, the addition of water to the system immediately lowers the friction between the HA monomers, resulting in steep shear thinning slopes as can be seen from Fig. 3a–c. It is evident from the panels that with solvent dilution, very low viscosities can be attained at higher shear rates, implying more promising application of the material with a sprayer. Consequently, the materials were successfully sprayed with an air gun with a pressurized inlet (SATA JET 1000 B RP spray gun) as shown in Fig. 3d–f.

Fig. 3: Shear flow behavior and sprayed pattern of HA in GFSH samples.
figure 3

Flow curves for increasing HA concentration in GFSH depicting steady shear viscosity η as a function of the shear rate \(\dot{\gamma }\) for a 0.5 wt%, b 1 wt%, and c 2 wt% HA. Pattern observed after spraying samples with d 0.5 wt%, e 1 wt%, and f 2 wt% HA in 35:65 water:GFSH ratio. Scale bars: 1 mm.

Full size image

Adhesive properties

Tack properties were measured using indentation measurements to compare the adhesiveness of the materials in terms of three parameters derived from the force vs. displacement curves:

  • Peak pull-off force – The maximum force recorded during retraction of the probe.

  • Retraction length – The distance from the start of retraction to the point of sample detachment from the probe.

  • Adhesive energy – The integrated area under the force vs. displacement retraction curve at a baseline of 0 mN.

A representative curve is shown in Supplementary Fig. S8 in which the different parameters are indicated.

Aliquots of 10 µL for each sample were pipetted and allowed to rest at 60% RH. Adhesion data for these samples were measured over different time points in a week. Upon comparison of the peak forces measured for increasing HA concentration over a period of one week in Fig. 4a–c, we observe that there is a gradual increase in peak forces as we move from 0.5 wt% to the 2 wt% HA samples. This is expected since an increase in HA concentration leads to a stronger elastic network, thus increasing the cohesiveness of the material and subsequently, the peak pull-off forces. The addition of water for each concentration series results in a decrease of the peak force for the 25:75 and 35:65 samples on day 0 (solid bars in Figs. 4a–c), since excess water increases the fluidity of the material leading to a system that cannot sustain applied forces as well. The measured forces in these cases for HA in 25:75 and 35:65 solvent ratios are mostly dominated by capillary forces arising between the liquid sample and the probe. The significance of dilutions becomes apparent when the adhesive properties are followed over time. Evaporation of excess water in the following days results in an increase in the total HA concentration in the material and thus, leads to a significant increase in the peak forces, observable for all samples in solvent ratios of 25:75 and 35:65. This evaporation process continues until all samples reach the same water content as the eutectic composition at 60% RH, as can be seen in Fig. 6a for the 0.5 wt% HA sample having an equilibrium water content of ~37%. In contrast, the samples prepared with a 0:100 water:GFSH ratio have consistent peak forces over the test period since the NaDES does not have much free water to lose and maintains its material properties over time. The retraction length in Fig. 4d–f clearly indicates the transition from a liquid-like to a solid-like material with an increase in polymer concentration. At HA concentrations of 0.5 wt%, a long fibril of >2.3 cm can be observed holding on to the probe, leading to no detachment in the measured deformation range (due to instrument limitations) as shown in Supplementary Movie S1 as an example. In contrast, at 2 wt%, the material becomes too cohesive and fails adhesively at very short distances without any fibrillation.

Fig. 4: Adhesion measurements of HA samples across varying solvent ratios.
figure 4

Maximum force measured from the retraction curve as a function of varying ratios of water:GFSH solvent for a 0.5 wt%, b 1 wt%, and c 2 wt% HA. Maximum length of sample attachment to the probe before debonding for d 0.5 wt%, e 1 wt%, and f 2 wt% HA, and the adhesion energy obtained from the area under the force vs. displacement curve for g 0.5 wt%, h 1 wt%, and i 2 wt% HA in GFSH. Error bars represent standard deviation. * indicates samples which did not detach from the probe in the experimental window. Legend in panel a applies to all panels.

Full size image

The energy of adhesion, shown in Fig. 4g–i, takes into account the peak forces measured and the length at which the material debonds from the probe, giving an overall inference about the adhesive strength of these materials. This is calculated as the area under the retraction curves in Supplementary Figs. S9–S11. As can be seen, the trends observed in the adhesive energy graphs are dominated by the retraction lengths indicating that more liquid-like materials give rise to higher energies while high HA concentrations of 2 wt% results in a decrease in the overall energy. We also observe this phenomenon in Fig. 4h and i, where the materials in 25:75 and 35:65 solvent ratios even show a slight reduction in adhesive energies as compared to the original 0:100 solvent since evaporation of water over time leads to an increase in the HA concentration, making the material stiffer than the 0:100 sample. On the contrary, in Fig. 4g, the materials with higher fractions of water eventually overtake the adhesive energy of the 0.5 wt% HA sample in a 0:100 solvent ratio. This means that evaporation of water in this case results in a subsequent increase in cohesiveness, leading to a material that can sustain the applied load. These results indicate that there is an optimum HA concentration between 1 wt% and 2 wt% beyond which the energy of adhesion starts to decrease due to the material becoming more solid-like.

Interfacial effects over time

Although the peak forces remain comparable over days for each sample, the retraction length, and subsequently the area under the curve, changes significantly, especially for the 1 wt% HA sample as can be seen from Fig. 4e, h. This is contrary to the expectation that, since there is essentially no change in the composition of the material, the adhesive properties should remain similar as well. One plausible hypothesis suggests that HA chains may undergo degradation at room temperature over time, resulting in the formation of smaller fragments. It is not unreasonable that this degradation process could potentially change the tack properties of the material, attributed to a synergistic effect arising from a mixture of molecular weights. Longer chains may contribute to increased entanglements and material stiffness, while shorter fragments, characterized by a shorter relaxation time, could enhance the material’s viscous properties. To investigate this hypothesis, linear shear rheology experiments were conducted on the 1 wt% HA sample over three days, exposed to 60% RH. However, this hypothesis was swiftly refuted, as illustrated in Fig. 5a, since the frequency sweeps of the material over time displayed overlapping profiles, indicating no discernible change in its rheological properties. However, it is possible to note that no terminal region has been attained in the probed frequency range (Fig. 5a), suggesting slow relaxation dynamics. To this end, Fig. 5b depicts the stress relaxation dynamics of samples at various polymer concentrations and at 0:100 solvent ratio. As can be observed in the same Figure, the required relaxation time for materials in 0:100 solvent ratio is over 15 minutes, which could mean that the fresh samples measured on day 0, immediately after pipetting, were not fully relaxed compared to those measured in the following days. Probe tack tests for 1 wt% HA in 0:100 solvent ratio were also compared from the rheometer in Supplementary Fig. S12 where a sample that is allowed to relax for 5 minutes even before the contact phase of the probe tack test gave rise to larger stress vs. strain curves in comparison to an instant retraction without any waiting or contact time. This is attributed to the material not having sufficient time to relax and erase its mechanical history upon loading, subsequently leading to poor contact and thus, lesser adhesive properties. Upon comparing the retraction curves of the same sample on the custom-made indentation setup in Fig. 5c, we again observe that no change in water content still gives rise to completely different retraction curves on day 0 and day 1. Apart from the bulk relaxation mechanism discussed above for samples measured on day 0, we believe that a structural reorganization of the HA chains could also be induced over time due to the exchange of water molecules either in the case of hydration or drying at the surface leading to an interfacial effect, thus also explaining the differences observed in the 0:100 solvent ratio in Fig. 5e37,38. While the exploration of this phenomenon could benefit from additional experiments and advanced scattering techniques, such investigations are reserved for future studies, as they fall beyond the scope of the present work.

Fig. 5: Time-dependent rheological and adhesive properties of HA in GFSH samples.
figure 5

a Linear rheology of 1 wt% HA in 0:100 water:GFSH solvent ratio (bulk material) over time. b Stress relaxation of HA in 0:100 water:GFSH solvent ratio (bulk material). c Retraction curves for 1 wt% HA in GFSH fresh (day 0) and after one day of being exposed to 60% RH (day 1). The inset shows the water content of the samples at both time points. Error bars represent standard deviation.

Full size image

Stability and dynamics of water-based trichome biomimetics

The developed water-based system offers some key advantages as well as disadvantages. As shown in Fig. 6a, samples eventually reach the same water content due to evaporation of the excess water that is not bound in the system, helping the material return to its eutectic point, where the water content is in equilibrium with the external RH. This unique property of the trichome biomimetics aids us in achieving a sprayable material that can regain its adhesiveness depending on the external RH. As a result of this equilibrium established with the surrounding environment, these trichome biomimetics can regulate their adhesive properties with respect to the surrounding environment which is important to ensure long-term effectiveness of the material. For very low HA concentrations of 0.5 wt%, cycling of the RH between 60% and 80% results in a pronounced change in both adhesive energy as well as peak pull-off forces from the retraction curve. As seen in Fig. 6b, the material is shown to have a significant decrease in the absolute values of peak pull-off force at 80% RH due to excess water being taken in, corresponding to a decrease in the adhesion energy since the material loses its cohesiveness. However, once the material is brought back to surroundings with 60% RH, the sample loses this excess water and regains its tack properties. We also observe this at higher HA concentrations of 1 wt% in Supplementary Fig. S13, however, it is to be noted that the overall adhesive energy does not vary as significantly as the 0.5 wt% sample. This could be attributed to the fact that at a low RH of 60%, the 1 wt% material debonds from the probe adhesively with a very short fibril as seen from the retraction curve in Supplementary Fig. S10 while at a high RH of 80%, the uptake of water leads to a decrease in peak force, but a corresponding small increase in area under the curve due to the formation of a weak fibril which results in the adhesive energies being comparable. Although reversible, this trend of loss of tack at high humidities indicates that these specific compositions could only be effective in a narrow RH range where these materials are not too dry nor too wet to trap pests. However, an interesting inference could be derived for higher HA concentrations of 2 wt% at 80% RH in Supplementary Fig. S14, resulting from the usual decrease in peak pull-off force as observed earlier. Since this material is on the stiffer side at 60% RH, tuning the composition further could result in improved adhesive properties at higher humidities due to the plasticization effect of excess water, which could possibly enlarge the operational RH range.

Fig. 6: Humidity-driven recovery and washability of trichome biomimetics.
figure 6

a Water content of 0.5 wt% HA dissolved in increasing water:GFSH ratio over a period of one week. Red dotted line indicates the equilibrium water content attained at 60% RH. b Area under the force vs. displacement retraction curve (red stars) and maximum force measured during retraction (green squares) exhibiting recovery of sample adhesiveness for 0.5 wt% HA in GFSH upon cycling of relative humidity between 60% and 80%. Error bars represent standard deviation. Confocal microscopy of 0.5 wt% HA in GFSH when c freshly sprayed samples on day 0, d followed by spraying with water for 3 seconds, and e then sprayed with water for 10. Scale bars: 100 µm.

Full size image

In order to not have consumers deal with sticky produce or plants, it is essential that these materials can be washed off the plants. For the 0.5 wt% HA sample, confocal microscopy images in Fig. 6c show a dispersion of relatively closely spaced droplets in comparison to bigger puddles for 1 wt% HA in Supplementary Fig. S15 and a web-like pattern for the highest concentration of 2 wt% in Supplementary Fig. S16. Upon briefly spraying with water for 3 seconds in Fig. 6d, the samples follow a similar trend to their behavior at high RH by absorbing water and merging, resulting in much bigger and sparsely spaced droplets/puddles. However, upon contact with much larger amounts of water in Fig. 6e, they can be completely washed off leaving minimal residues. It is to be noted that an increase in HA concentration in the sample results in a higher residue. Consequently, a greater amount of washing will be necessary to thoroughly eliminate the samples, as illustrated Supplementary Figs. S15 and S16.

Phytotoxicity and growth rate assessment

A crucial aspect of developing an alternative to chemical pesticides is the necessity for the material not to negatively affect the plant itself. Relative Growth Rate (RGR) is a measure of how fast a plant is growing relative to its current size. Phytotoxic substances can affect a plant’s ability to take up nutrients or carry out photosynthesis, leading to changes in growth rates. Phytotoxicity experiments revealed that no significant developmental changes were observed between the control and the treated group for three different plant species. RGR shown in Table 1 even indicates an improvement in growth rate in the treated group, which could be attributed to the presence of sugars acting as an energy source for the plants and improving their stress tolerance39. Leaf Area Ratio (LAR) can act as a measure to understand the relative useful area available for photosynthetic activity in plants40. High LAR values indicate that a plant invests a large proportion of its resources in leaf production. However, under conditions of stress or phytotoxicity, a plant might reduce its leaf area as a strategy to conserve resources. As shown in Fig. 7a, the LAR values in comparison to the control groups are slightly lower for chrysanthemums, which could indicate a small reduction in photosynthetic area. When comparing the control group with treated strawberries and paprika, the LAR values remain comparable, indicating no phytotoxic effects in these species due to the applied trichome mimics. During the experimental period of 2.5 weeks, no leaf loss or dead plants were observed on any plants. In addition, with the exception of some spots present on both the control and treated plants, no browning of the leaves or fungal growth on the leaves was observed. No additional morphological differences were observed between the two groups, indicating no negative effect of the samples on plant growth and development.

Table 1 Relative Growth Rate (RGR) values for three tested species over an experimental period of 2.5 weeks
Full size table
Fig. 7: Phytotoxicity assessment and thrips immobilization.
figure 7

a Leaf Area Ratio (LAR) value for chrysanthemums, strawberries, and bell peppers indicating no/minimal phytotoxic effects. Statistical analysis by Welch’s t-test, **p < 0.01. Error bars represent standard deviation. b Results of Petri-dish thrips assay after exposing samples to 60% RH for one day depicted in a stacked bar chart. c Thrips immobilized by 0.5 wt% HA sample due to multiple entrapments of the thrips by sprayed droplets. d Fibril formation observed when trapped thrips attempts to escape from 0.5 wt% HA sample. e Thrips larva immobilized on 0.5 wt% HA sample due to full sample contact with the abdomen. Red arrows in c and d indicate multiple points of contact on the body essential for entrapment.

Full size image

Behavioral response and immobilization of thrips

To evaluate effectiveness upon application, Petri dish assays using thrips as target pests were conducted. From the behavioral monitoring of the thrips within ten minutes after release in the Petri dish, the 0.5 wt% HA sample hindered the movement of the thrips and significantly decelerated their motion. Fibrils could be observed after contact with the sample when thrips attempted to escape, as shown in Fig. 7c (Supplementary Movie S2). An important aspect to note is the necessity of multiple contact points (Supplementary Movie S3) as shown in Fig. 7d or contact over a large area such as the abdomen in Fig. 7e, which can lead to complete immobilization in contrast to thrips, which could escape if only a single leg made contact. This is in striking similarity to the densely packed manner of natural trichomes which is crucial for trapping insects, in addition to the inherent stickiness of the exudate itself. For a medium HA concentration of 1 wt% in contrast to the 0.5 wt% sample, fibril formation was not observed, and they were not trapped upon initial contact as quickly. However, once trapped through multiple contacts, none of the thrips could escape. The frequency of thrips being free in the monitored time period was quite high for this sample, which could be attributed to either low tack or inability to spray the material into finely dispersed drops (shown in Fig. 3e), resulting in a lack of multiple contact points forming. For the highest HA concentration of 2 wt%, thrips were observed to walk over the sample owing to very low tack and high material stiffness (Supplementary Movie S4). Although these differences in trapping effectiveness are reflected in Fig. 7b, they become less apparent here in comparison to the adhesion data in Fig. 4. All samples appear to have performed relatively well in terms of trapping efficiency and it becomes difficult to derive a concrete conclusion on which sample is the most effective. This indicates a crucial bottleneck in such application testing which would be to optimize sprayability to get a finer dispersion such that the droplet sizes and density of all materials remain comparable to study the effect on adhesiveness independently.

Conclusions

This work presents a bio-based adhesive as a form of non-toxic pest control by taking inspiration from the composition of the Drosera trichome exudates. We demonstrate a proof-of-concept system using natural deep eutectic solvents and hyaluronic acid. The adhesiveness of these trichome mimics are tunable by varying the concentration of hyaluronic acid, which influences the elasticity of the material. Low polymer concentrations of 0.5 wt% resulted in liquid-like samples with the formation of long fibrils greater than ∼2.3 cm in length upon extension. Higher polymer concentrations of 2 wt% led to materials with higher cohesiveness accompanied by adhesive failures at low strain values. Sprayability was improved by diluting up to 35% with water, breaking some hydrogen bonds in the solvent, but restoring adhesion upon evaporation of the excess water. Besides having no phytotoxic effects on the tested plant species, a slight improvement in the relative growth rate was observed for all species which could be attributed to the presence of sugars in the material acting as an energy source for plant growth. Application testing using western flower thrips as model pests resulted in at least 40% trapping efficiency, with the revelation of multiple contact points crucial for complete immobilization, a trait commonly observed with natural trichomes as well. In conclusion, this innovative approach acts as a significant step forward in addressing the challenges posed by chemical pesticides. Taking inspiration from nature’s existing defense mechanisms offers a promising avenue for the development of physical pesticides, enabling us to move towards more sustainable pest management practices.

Materials and Methods

Preparation of NaDES

The NaDES used in this study was composed of D-glucose (Sigma-Aldrich; Darmstadt, Germany), D-fructose (Boom B.V; Meppel, Netherlands), sucrose (Sigma-Aldrich; Darmstadt, Germany), and water in the molar ratio of 1:1:1:11. The individual sugars and water were weighed out and stirred together for 1 hour at 65 °C until a clear liquid was obtained. The mixture was then flushed with nitrogen to prevent water absorption during storage and then allowed to cool down to room temperature. The solvent was used for preparation of samples if no crystallization was observed after one day. All experiments using the NaDES were performed within a maximum storage time of one week. The prepared NaDES is referred to as GFSH (corresponding to G-glucose, F-fructose, S-sucrose, and H-water) in the following sections.

Preparation of trichome mimics

Hyaluronic acid sodium salt (HA) with an average molecular weight 750-1000 kDa (Glentham Life Sciences Ltd; Corsham, UK) was dispersed in 5 g of solvent to prepare final concentrations of 0.5 wt%, 1 wt%, and 2 wt% HA by vortexing for 10 s followed by centrifuging for 4 min at 3000 g. Three different solvent ratios were studied with a composition of 0% water+100% GFSH (denoted as 0:100), 25% water+75% GFSH (denoted as 25:75), and 35% water+65% GFSH (denoted as 35:65). The mixture was then heated to 65 °C to facilitate dissolution of the polymer by the solvent. Samples were removed from the heating plate when they appeared to be homogeneous (1 hour stirring time). The vials were centrifuged at 3000 g for 5 min immediately when hot to remove trapped air bubbles in the sample. The samples were then allowed to cool down to room temperature for a day before being used for further experiments.

Rheological measurements

The viscoelastic properties of the HA/GFSH samples were investigated using an Anton Paar MCR302e rotational rheometer. Measurements were performed using a cone and plate geometry (25 mm diameter, 1° cone angle). Experiments were started after the normal force of the loaded sample was less than 0.1 N to ensure full relaxation. All measurements were performed at 20 °C. Silicone oil was used as a trap to prevent the interface from drying. The sample was loaded into the rheometer and a dynamic strain sweep (DSS) was carried out at 10 rad s-1 ranging from 0.1–500% and 500–0.1% to determine an appropriate strain value in the nonlinear and linear viscoelastic regime for the pre-shear protocol. Consequently, the sample was rejuvenated through a dynamic time sweep (DtS) at 10 rad s−1 with a strain value of 550% for 30 s (until steady-state values were attained) to remove all the mechanical history of the sample. The sample was then subjected to aging with a DtS at 0.1% for 100 s (also sufficient to attain steady-state) to reconstruct the microstructure of the sample as shown in Supplementary Figs. S1 - S3.

Linear rheology

DSS at 100 rad s−1 was conducted to determine the linear viscoelastic (LVE) regime. Frequency sweeps were then performed over a range of angular frequencies from 100 – 0.1 rad s−1 to study the viscoelastic spectra of the materials.

Non-linear rheology

The loaded sample was subjected to rejuvenation and aging following the protocol described previously. Start-up of shear rate tests were conducted between 0.01 s−1 and 0.1 s−1 to determine the steady state values of the shear viscosity. Subsequently, flow curves were constructed over shear rates of 0.01 −100 s−1.

Proton Nuclear Magnetic Resonance (1H NMR)

1H NMR experiments in D2O were performed on a Bruker Avance III HD spectrometer operating at 400 MHz using a standard 5 mm broadband Smart probe regulated at 25 °C. Chemical shifts are shown in parts per million (ppm) from tetramethylsilane referenced to the residual isotopomer solvent signal (HOD).

Indentation measurements

A custom-made indentation setup as described by Boots et al. was used to quantify the tack and energy of adhesion of the HA/GFSH samples as shown in Fig. 841. The custom setup had a humidity-controlled chamber where 10 µL droplets were prepared and allowed to equilibrate at 60% RH. Measurements were performed using a 2 mm sphere as a probe and a 10 g Futek load cell. The probe was brought into contact with the sample at a fixed speed of 100 µm/s until a pre-load force of 0.2 mN was applied on the samples. The probe was then retracted at the same speed until complete detachment of the sample or until the instrument’s limits were reached. A Toolcraft USB microscope was used to record the approach and retraction of the probe for different samples. Additionally, reversibility experiments to study changes in adhesive properties were conducted using the same indentation protocol, with samples first exposed to 60% RH for 24 hours, followed by 24 hours at 80% RH, alternating between these conditions.

Fig. 8: Custom-built setup for adhesion measurements under controlled humidity.
figure 8

Schematic illustration of the custom-made indentation setup. The relative humidity inside the measurement chamber is continuously monitored using a digital hygrometer which then instructs the designed humidity generator to turn on or off. The dry air entering the humidity controller mixes with the produced water vapor and is then sent to the chamber.

Full size image

Water content quantification

Gravimetric analysis was used to study the water content of the samples. 10 µL droplets were weighed (W0) and placed in a controlled relative humidity (if required) for a specified number of days after which their weight was recorded (WRH). The samples were then placed in a vacuum oven at 70 °C overnight after which the weight was recorded for the final time (Wdry).

Water content of the samples on day 0 (unexposed to 60% RH) was determined using the formula:

$${{\rm{Water}}}\; {{\rm{content}}}\; {{\rm{on}}}\; {{\rm{day}}}\,0\left( \% \right)=\left[\frac{{W}_{0}-{W}_{{{\rm{dry}}}}}{{W}_{0}}\right]\times 100$$
(1)

Water content of the samples exposed to 60% RH was determined using the formula:

$${{\rm{Water}}}\; {{\rm{content}}}\,\left( \% \right)=\left[\frac{{W}_{{{\rm{RH}}}}-{W}_{{{\rm{dry}}}}}{{W}_{{{\rm{RH}}}}}\right]\times 100$$
(2)

Fluorescence imaging

Owing to the transparency of the HA/GFSH samples, rhodamine B (Sigma-Aldrich; Darmstadt, Germany) was used to visualize sprayed samples with better contrast. Due to the high viscosity of the samples, direct dissolution of the dye in a bulk volume resulted in aggregation. Stock solutions were first prepared with rhodamine B at 0.6 mg mL−1 in the HA/GFSH samples which were then diluted 50 times using the remaining material by stirring until homogeneous. Samples were then sprayed on 9 cm diameter polystyrene Petri dishes using a SATAjet 1000 B RP spray gun with an inlet pressure of 3 bar and a top pressure of 1.2 bar and a 1.6 mm spray nozzle. Samples were observed using a ZEISS 710 Laser Scanning Microscope equipped with a LD Plan-Neofluar 20x/0.4 air-type objective, illuminated using HeNe (543 nm) laser. Fresh samples were imaged within 30 minutes of being sprayed. The sprayed Petri dishes were kept at 60% RH and 20 °C for 24 h followed by spraying distilled water for 3 seconds. They were then allowed to equilibrate at 60% RH for 6 hours to lose excess water and regain the initial water content following which they were imaged to study the changes in droplet appearance after RH changes. The protocol was repeated, but the samples were now sprayed with distilled water for 10 s instead and then imaged after 6 h to study their washable nature.

Phytotoxicity study

Phytotoxicity experiments were carried out using three types of plants and the 0.5 wt% HA sample in a water:GFSH ratio of 35:65. 40 cuttings were used for both Chrysanthemum ‘Baltica’ (chrysanthemums) and Fragaria × ananassa ‘Favori’ (strawberries). The individual cuttings were placed in a mixture of potting soil and vermiculite in 1 L pots and grown for 4 weeks before the experiments. Seeds were first germinated in potting soil for the Capsicum annuum ‘Yolo Wonder’ (bell peppers), and after two weeks, individual seedlings were placed in 1 L with potting soil. The tests were carried out with 50 plants (5 weeks old). All plants were reared at 21 °C constant with 8 hours dark, 16 hours of light, at 60% RH. For the chrysanthemums, each plant was sprayed from top to bottom with a 120° rotation in between each spray, such that the whole plant was covered with 3 sprays, resulting in a total of ~14 g sample being applied on each plant. For the strawberry plants, ~3 g of sample was sprayed for each leaf petiole containing a set of 3 leaflets. This was repeated until the whole plant was covered with the sample. Since the bell pepper plants were much smaller in size, only ~4 g of the sample was applied to each plant. All samples were sprayed using a SATAjet 1000 B RP spray gun with an inlet pressure of 3 bar and a top pressure of 1.2 bar, and a 1.6 mm spray nozzle. The experiment was carried out for a period of two weeks during which the plants were checked every day for irregularities, such as leaf loss, damage to the leaves, and brown spots.

Leaf Area Ratio (LAR) and Relative Growth Rate (RGR) are physiological parameters used in plant ecology and physiology to assess plant growth and performance. These parameters were calculated as follows:

$${{\rm{LAR}}}=\frac{A}{W}$$
(3)

where A is the total leaf area of a plant (cm2) and W is the total dry matter weight (g);

$${{\rm{RGR}}}=\frac{{{\mathrm{ln}}}{W}_{2}\,-{{\mathrm{ln}}}{W}_{1}\,}{{t}_{2}\,-\,{t}_{1}\,}$$
(4)

where W1 and W2 are plant dry weights (g) at times t1 and t2 (days).

Effect on target pests

For efficacy testing, western flower thrips, Frankliniella occidentalis (Pergande), were cultivated in large numbers using chrysanthemum cut flowers of the Baltica yellow variety in a climate-controlled room set at 25 °C, 60% RH, and a 16:8 light-to-dark photoperiod. Samples were sprayed on 9 cm polystyrene Petri dishes (denoted as day 0) and were left open at 20 °C and 60% RH for 24 h after which the behavior of 3 thrips larvae, of varying ages, per sample was monitored for 10 minutes after they were introduced in to the Petri dish (denoted as day 1). To quantify thrips immobilization, young thrips larvae were added to each of these Petri dishes (on day 1) and were then sealed and left back in the climate chamber. The number of thrips were counted the next day (denoted as day 2) and were classified into 3 categories namely, Stuck (if found to be immobilized in the droplet), Free (if found to be alive and outside the sample), or Dead (if found to be dead outside the droplet). 3 Petri dishes were prepared and quantified in a similar way for each sample, adding up to 15 total thrips per sample.