液滴在超疏水植物叶面的沉积：实验和分子动力学模拟

曹冲，张裴，曹立冬，刘铭鑫，宋玉莹，陈鹏，黄啟良,＊，韩布兴

1中国农业科学院植物保护研究所，北京 100193

2中国科学院化学研究所，胶体、界面与化学热力学实验室，北京 100190

1 Introduction

Chemical pesticide plays an important role in the control of pests for the security of food production1.Studies showed that nearly 50% of pesticides were lost in the process of leaf spray due to bounce and splash in the target crop leaves2,3, which brought about various issues by the loss of chemical pesticides4-10.Enhancing the pesticide droplet deposition on the target leaf is an effective strategy to decrease pesticides loss, and the issue has been extensively studied in recent years5-9,11-13.

The deposition of pesticide droplets on the leaf surface of target plants involves the physical and chemical properties of the liquid and solid surface of target plants leaf14-16.Generally speaking, there are three results of dynamic deposition of pesticide droplets on target: splashing, rebound and deposition14,17.Only effective deposition of pesticide droplets works in pest control when pesticide droplets adhere to the leaf surface of target crops.

The efficient deposition of pesticide droplets on the target leaf surface could be achieved by regulating pesticide liquids properties such as surface tension, viscosity and droplet size14,17,18.Adding surfactants to the liquid is a common and effective approach to regulating liquid properties19,20.However, the addition of surfactant always increase the wetting spread on the plant pageviareducing the surface tension of pesticide liquid,resulting in the decrease of droplet size, which is favorable to drift from the leaf surface21,22.In addition, the use of traditional surfactants which are difficult to be biodegraded resulted in added pollution to the environment.Developing efficiently green surfactant systems is desirable and remains challenges.

Bio-based surfactants have several advantages involving nontoxicity, biodegradability, biocompatibility, and so on.In addition, the bio-based surfactants derived from biomass was reported to promote plant growth and improve the quality of soil in agriculture.However, there are few reports about the application of bio-based surfactants in the pesticide formulation and spray industry.In this work we propose a new and green system to enhance droplet deposition on superhydrophobic/hydrophobic leaf surfaces by using a bio-based surfactant.Here,a bio-based surfactant SAAS-C12 which was derived from sorbitol (structure in Fig.S1a, Supporting Information)23behaved highly efficient deposition behaviors on the hydrophobic surface with low concentration (0.25% (w, mass fraction)) together with a small quantity of glycerol (0.001%(w)).Control experiments and MD calculation displayed.

2 Experimental and computational section

2.1 Materials

The cabbage (Brassica oleracea L.), quinoa (Chenopodium album L.) and citrus (citrus L.) plants were provided by Institute of Plant Protection, Chinese Academy of Agricultural Sciences.The green onion (Allium fistulosum L.) was purchased from supermarket.Glycerol with a purity of 99% was obtained from Shanghai Macklin Biochemical Co.Ltd.The sorbitolalkylamide surfactants SSAS-C12and glycerol based surfactants BAPO-C12were prepared and purified as described previously23,24.The synthesis method of DSSAS-C12 was detailed in the supplementary information.

2.2 Characterization of the plant leaf surfaces

The microstructure of the plant leaf surfaces was obtained using an SU8010 scanning electron microscope (SEM) at 10.0 kV.The contact angles of the aqueous solutions on the plant leaf surfaces were calculated with a high-speed optical measuring device OCA 20 (DataPhysics Instruments GmbH, Germany).2 μL droplets of water was injected onto the plant leaf surface and the contact angle values were measured by sessile drop method and calculated using circle-fitting methods by analyzing the drop contour.

2.3 Characterization of the aqueous solutions

The viscosity values of all aqueous solutions were investigated with a viscometer DV-III ULTRA (Brookfield,USA) programmable rheometer at 25 °C.The dynamic surface tension was determined with an automatic maximum bubble pressure tensiometer (Krüss BP100, Germany) under the condition temperature of 25 ± 1 °C.The measurement range of the time window measurements was from 10 ms to 100 s with a capillary diameter of 0.210 mm.

2.4 Diffusion coefficient measurement

Diffusion coefficients were measured by the pulsed field gradient NMR method (PFG-NMR method) on a Quantum-I spectrometer operating at proton resonance frequency of 400 MHz.A standard 90°-180° pulse sequence and stimulated echo pulse sequence were used for the measurement in the surfactants and surfactant/glycerol binary system.

2.5 Impact experiments

The dynamic processes of the droplet impacting on the plant leaf surfaces were recorded with a high-speed camera(FASTCAM Mini UX100 Photron) at a rate of 8000 frames per second with a shutter speed of 1/20000 s.The impact velocity of the droplet was adjusted by adjusting the height of the droplet free fall.

All experiments were repeated 3 times at 25 ± 1 °C, and the error was within 3%.

2.6 Molecular dynamics simulations

In order to investigate the dynamic adsorption behavior of surfactant/glycerol to the solid surface, the CG molecular dynamics (MD) simulations were performed.The surfactant concentration was 0.25% (w) and the concentration of glycerol was 0.02% (w).All CG MD simulations were performed using GROMACS with Martini force field25,26.A solid surface of 24 nm × 24 nm representing cabbage (Brassica oleracea L.) leaf surface was constructed following the same chemical compositions as provided in previous study27,28.Nine cone-like obstacles (6 nm in diameter of bottom surface and 3 nm in height) were constructed on solid surface to describe the surface roughness.The constructed solid surface is parallel toXY-plane and was kept frozen in all CG MD simulation systems.The periodic distance in theZ-axis, which is perpendicular to the solid surface, is set to 40.0 nm, which is sufficiently large so that interactions between simulated molecules and the periodic image of the constructed surface in the top plane can be ignored.

For three surfactants and glycerol molecules, the CG force field was developed based on a three-to-one mapping scheme,meaning that each three connected heavy atoms and the associated hydrogens were represented by a single CG interaction site (bead).In all CG MD simulations, the initial configurations of CG surfactants, glycerol molecules and water molecules on solid surface was built based on the following three steps (Supporting Information).The equations of motion were integrated using a classical velocity Verlet leap-frog integration algorithm with a time step of 10 fs accelerating CG MD simulations of modelling systems at extended spatiotemporal scales with a modest computational cost.A cutoff distance of 1.6 nm was set for short range van der Waals interactions and realspace electrostatic interactions.The Particle-Mesh Ewald (PME)summation method was employed to handle long range electrostatic interactions in reciprocal space with an interpolation order of 5 and a Fourier grid spacing of 0.2 nm.All simulation systems were first energetically minimized using a steepest descent algorithm, and thereafter annealed gradually from 600 K to room temperature within 20 ns.All annealed simulation systems were equilibrated in an isothermal-isobaric(NPT) ensemble for 20 ns of physical time maintained using a Nosé-Hoover thermostat and a Parrinello-Rahman barostat with time coupling constants of 0.4 and 0.2 ps, respectively, to control temperature at 300 K and pressure at 1 atm (1 atm = 100 kPa).CG MD simulations were further performed in a canonical(NVT) ensemble for 40 ns, and simulation trajectories were recorded at an interval of 100 fs for further structural and dynamical analysis.

3 Results and discussion

3.1 Dynamic deposition behaviors of SSASC12/glycerol on various superhydrophobic plant leaf surfaces

The binary additives consisted of 0.25% (w) SSAS-C12and 0.001% (w) glycerol were used to investigate the deposition on various leaf surfaces.The deposition of droplets on the leaf surfaces was recorded by a high-speed camera at a velocity of 2.62 m·s-1.The diameter (D0) of droplets is about 2.25 ± 0.02 mm.Fig.1 showed the dynamic behaviors of high-speed drops impacting on superhydrophobic cabbage (Brassica oleracea L.)leaf surfaces.The cabbage leaf surface was characterized by SEM (Fig.1b), which showed a water contact angle of 151.8° ±2.0° (Fig.1c), indicating that the cabbage leaf surface was superhydrophobic.Fig.1d-g showed the impact behaviors of various droplets, including water, 0.001% (w) glycerol, 0.25%(w) of SSAS-C12and the binary system SSAS-C12/glycerol on the cabbage leaf surfaces.The results in Fig.1d, e showed that water and glycerol droplets impacted on the cabbage leaf surface and retracted quickly, then shatter and rebound behaviors occurred (Supporting Information, Movies S1).For 0.25% (w)SSAS-C12droplet (Fig.1f), the receding splash and rebound behaviors were inhibited by certain degrees.However, the droplet with 0.25% (w) SSAS-C12retracted violently, leaving several tiny droplets on the cabbage leaf surface.After adding 0.001% (w) glycerol into 0.25% (w) SSAS-C12solution (Fig.1m), the droplets including SSAS-C12/glycerol binary system completely spread on the superhydrophobic cabbage leaf surface leaving a large wetting area (Supporting Information, Movies S1).Then we changed the droplet impact velocity up to 3.55 m·s-1and adjusted the tilted angles of leaf surface, similar deposition efficiency was obtained on the leaf surfaces(Supporting Information, Movies S1).Furthermore, the aqueous droplets containing SSAS-C12/glycerol would spread and deposit properly on the cabbage leaf surface with spray impact(Supporting Information, Movies S1).

Fig.1 Dynamic behaviors of high-speed drops impacting on superhydrophobic cabbage (Brassica oleracea L.) leaf surfaces.

We then used other plant leaves to confirm the general effectiveness of SSAS-C12/glycerol binary additives in enhancing droplets deposition on superhydrophobic or hydrophobic surfaces as shown in Fig.2.The plant leaves contain green onion (Allium fistulosum L.) (Fig.2a), quinoa(Chenopodium album L.) (Fig.2b) and citrus (citrus L.) (Fig.2c),which are typical target crops that use pesticide.The results showed similar retention properties on different plant leaves surfaces.All impacting water droplets on the superhydrophobic or hydrophobic leaves exhibited fully bounce.As expected, the droplets with SSAS-C12/glycerol not only showed fully inhibited rebound and deposited on the leaves but also finally obtained a large wetting area (Supporting Information, Movies S1).In order to verify the feasibility of the SSAS-C12/glycerol system in agricultural application, we used micro-emulsions of 5% lambda-cyhalothrin to carry out spray experiment on the cabbage leaf surface (Supporting Information, Movies S2).The results showed that the pesticide liquid solution with microemulsions of 5% (w) lambda-cyhalothrin diluted 1500 times with SSAS-C12/glycerol additive can be well deposited on the cabbage leaf surface.

Fig.2 Dynamic behaviors of high-speed drops impacting on various leaf surfaces.

To further investigated the droplet dynamics of SSASC12/glycerol on the plant leaf surfaces, we studied the widely applied normalized maximum contact diameterDt/D0, whereD0represented the initial radius of droplets,Dtwas the contact spreading diameter6,18.In the falling stage, we regarded the movement of droplets as free falling.The moment when the droplets contacted the leaves was recognized as the initial time(0 ms).It was observed from Fig.3 that all theDt/D0values of droplets reached the maximum at about 1.5 ms, and theDt/D0value of the SSAS-C12/glycerol drop held the maximum value finally, showing good spreading and efficiency deposition (Fig.3a).Generally speaking, the addition of surfactants mainly reduces the surface tension or changes the dynamic surface tension of liquid droplets to increase the effective deposition of droplets on the superhydrophobic surfaces19,29,30.Here, the addition of glycerol did not change the dynamic surface tension of the surfactant solution obviously (Fig.3b).According to the Kelvin-Helmholtz instability29,Kmax= 2ρaUr2/3γ(ρais the air density,Uris the relative velocity between gas and liquid,γis the liquid surface tension), with a constant surface tension, the key to improving stability is the speed of retraction, and it is necessary for the droplets to reach a larger degree of wetting and spreading on the hydrophobic surface in a short time20.In our study, local pinning is obtained for SSAS-C12(Fig.1f), and the totally pinning is observed for SSAS-C12/glycerol with the maximum spreading (Fig.1g and Supporting Information,Movies S1).As shown in Fig.3b, the dynamic surface tension of SSAS-C12solution exhibited no significant change after the addition of glycerol with low to high concentration (0.01% (w)).Fig.3c showed that SSAS-C12/glycerol aqueous solution can effectively inhibit the bouncing behavior of droplets on the plant leaf surface with varied tilted angles (Supporting Information,Movies S1).In addition, variations in the viscosity of the solution (Supporting Information, Fig.S2) and even the contact angle on the superhydrophobic surface were not observed(Supporting Information, Fig.S3).

Fig.3 The dynamic surface tension and impact behavior of SSAS-C12, glycerol, and SSAS-C12/glycerol binary additive.

3.2 Investigation of the structure differences of surfactants on the dynamic deposition behaviors

To explore the impacting and deposition mechanism of the SSAS-C12/glycerol system, two additional surfactants 1,4-bis(dodecylamino)butane-2,3-diol (denoted as DSSAS-C12)and 1,3-bis(dodecylamino)propan-2-ol (denoted as BAPO-C12)with hydroxy groups variations were studied for comparison,and the structures were presented in Fig.S1b and 1c (Supporting Information).The dynamic impact and deposition process of the two surfactants and the mixture of surfactants with 0.001% (w)glycerol on the cabbage leaf surface were studied respectively(Supporting Information, Fig.S4 and Fig.S5).The results showed that the 0.25% (w) DSSAS-C12and 0.25% (w) DSSASC12with 0.001% (w) glycerol binary droplets eventually deposited on the cabbage leaf surfaces as small broken droplets(Supporting Information, Movies S3), even when the impact speed is lower than 1.4 m·s-1(Supporting Information, Movies S3).We chose 0.05% (w) BAPO-C12for comparison due to the solubility of BAPO-C12.0.05% (w) BAPO-C12with 0.001% (w)glycerol binary system exhibited the phenomenon of bouncing after impacting, even if the glycerol concentration was as high as 0.05% (w) (Supporting Information, Movies S4) and the impact velocity was lower than 1.4 m·s-1(Supporting Information, Movies S4).Furthermore, the aqueous droplets containing DSSAS-C12/glycerol (Supporting Information,Movies S3) and BAPO-C12/glycerol (Supporting Information,Movies S4) could not spread and deposit on the cabbage leaf surface with spray impact.We also have investigated the physicochemical property of DSSAS-C12/glycerol and BAPOC12/glycerol systems, including the viscosity (Supporting Information, Fig.S6), dynamic surface tension (Supporting Information, Fig.S7) and contact angle on the leaf surface(Supporting Information, Fig.S8).It is found that the addition of glycerol has not influenced the viscosity of surfactant solutions.Also, the dynamic surface tension and the contact angle on the leaf surface have not been changed markedly.The results showed that the SSAS-C12/glycerol binary solution had lowest dynamic surface tension and contact angle on the leaf surface among three surfactant/glycerol systems.

In many reported surfactant-enhanced deposition, the surfactant concentration tend to about 1% (w) or more.As revealed in Fig.1 and Fig.2, the SSAS-C12/glycerol binary additive can enhance the droplet deposition with SSAS-C12concentration as low as 0.25% (w), which concentration is much lower than that in the reported system3,19,31.Therefore, the dynamic property of SSAS-C12/glycerol binary system must play an important role in controlling the droplets deposition during the high-speed impacting process.

3.3 Physicochemical properties of droplets and dynamic deposition mechanism

The physicochemical properties of the SSAS-C12/glycerol system were considered.It was reported that the influence of viscosity was more important than dynamic surface tension in obtaining uniform spread during drop impact process31.For system of 0.25% (w) SSAS-C12with 0.001% (w) glycerol, the shear viscosities of aqueous solutions are shown in Fig.S2(Supporting Information).It was found that the addition of glycerol has not changed SSAS-C12solution viscosity, and all aqueous solutions shear viscosities were nearly equal to 1.0 mPa·s.Songet al.19considered that the aggregation morphology of the surfactant was the key factor to determine the impact and retention on the superhydrophobic leaf surface.In our study, the addition of glycerol has not affected the aggregation morphology of SSAS-C12(Supporting Information, Fig.S9).The contact angle on the cabbage leaf surface has not changed significantly with 0.001% (w) glycerol addition (Supporting Information, Fig.S3).However, when the concentration of glycerol was higher than 0.001% (w), the SSAS-C12/glycerol binary system droplets would improve the droplets deposition on the plant leaf surface(Supporting Information, Movies S4).It was proved that there was some special interactions between SSAS-C12and glycerol,which changed the diffusion and adsorption processes of SSASC12in the bulk and the solid/liquid interface.

Subsequently, we investigated the diffusion ordered NMR spectroscopy (DOSY) (Supporting Information, Fig.S10),which was a pulsed field gradient NMR spectroscopy that enabled measurement of translational diffusion of dissolved molecules32.DOSY is a powerful tool for diffusion coefficient to provide direct information on molecular dynamics, including intermolecular interactions32,33.According to the Stokes-Einstein equationD=KBT/(6πηrH), whereDis the diffusion coefficient,Tis the absolute temperature,KB is the Boltzmann constant,ηis the viscosity andrH is the hydrodynamic radius32.Fig.4 showed the diffusion coefficientsDsand the hydrodynamic radiusrHof SSAS-C12in the SSAS-C12/glycerol binary systems.The addition of glycerol led toDsincreasing andrHdecreasing.It demonstrated that adding glycerol to the SSASC12 solution decreased its hydration radius and increased its diffusion rate.Both SSAS-C12molecules and glycerol molecules possess hydroxyl groups, and there are strong hydrogen bonds between them.In fact, the surfactant SSAS-C12itself form selfassociationviahydrogen bonds at this concentration (0.25%(w))23.Due to hydrogen bonds between SSAS-C12and glycerol molecules, the self-association of SSAS-C12can be damaged and SSAS-C12/glycerol complex can be formed.In this way, the hydrodynamic radius of SSAS-C12/glycerol complex was smaller than that of SSAS-C12self-association, and the diffusion coefficient showed a faster diffusion rate.It was possible that hydrogen bond interactions between SSAS-C12and glycerol led to a twisting of headgroup in SSAS-C12molecules.Meanwhile,we have investigated the effect of hydroxy groups on the diffusion coefficientsDs.The diffusion coefficients of DSSASC12/glycerol and BAPO-C12/glycerol systems were exhibited in Table S1 and Table S2 (Supporting Information).The addition of glycerol led to theDsdecreasing andrHincreasing for both DSSAS-C12/glycerol and BAPO-C12/glycerol systems.DSSASC12 and BAPO-C12 molecules differed from SSAS-C12molecules in the number of hydrophilic hydroxyl groups.The reduction in the number of hydroxyl groups bought about a decrease in hydrogen bond interaction sites.

Fig.4 Diffusion coefficients and hydrodynamic radius of SSAS-C12 as a function of glycerol concentration in the SSAS-C12/glycerol binary systems at 25 °C.

Based on these findings, molecular dynamics (MD)simulations were further utilized to understand the adsorption and deposition mechanism.MD simulations provided insights into the molecular-level mechanism34.Time sequence of typical snapshots of how SSAS-C12adsorbed to the solid interface with/without glycerol (Supporting Information, Fig.S11)revealed that the addition of glycerol could facilitate SSAS-C12diffuse and adsorb to the solid surface, leading to accumulation of SSAS-C12molecules at liquid/solid interface.SSAS-C12molecule distribution relative to distance from the solid surface further confirmed that SSAS-C12molecules tended to easily accumulate at the solid surface with glycerol addition(Supporting Information, Fig.S11d).It is worthwhile mentioning that the glycerol addition promotes SSAS-C12diffusion to minimize the system energy (Supporting Information, Fig.S11c).To unveil the underlying mechanism,molecular dynamics (MD) simulations were also performed for DSSAS-C12/glycerol system and BAPO-C12/glycerol system showed in Fig.5.The results revealed that SSAS-C12molecules adsorbed and well covered the solid interface, and DSSAS-C12molecules partly covered the solid interface but so few BAPOC12adsorbed to the solid interface, which agreed with the trend of contact angle (Supporting Information, Fig.S8) and the spreading state (Fig.1).The SSAS-C12/glycerol binary system also had lowest energy (Fig.5c).Furthermore, based on the distribution relative to distance from the solid surface, it appeared that SSAS-C12molecules tended to readily adsorb to the solid surface with glycerol addition (Fig.5d).

Fig.6 illustrated the interactions between SSAS-C12and glycerol which enhanced the droplet adsorption and deposition on the superhydrophobic leaf surface.The droplet impacted on the superhydrophobic leaf surface and reached its maximum spreading area due to the droplet inertia force.In this case, new water/air and water/solid interfaces were brought out.The final impacting and spreading states were determined by the surfactant diffusion and adsorption from bulk to the newly formed water/air and water/solid interfaces.Surfactant with faster diffusion rate can quickly adsorb at the water/air and water/solid interfaces and make the superhydrophobic surfaces hydrophilic.

It was possible that hydrogen bond interactions between SSAS-C12and glycerol led to a twisting of headgroup in SSASC12molecules.Fig.6a showed a molecule arrangement in which the headgroups of SSAS-C12formed two five-membered rings through intermolecular hydrogen bonds without glycerol addition.In this case, the two alkyl chains of SSAS-C12chose to arrange in two orientations similar to the arrangement of sorbitol, showing open “arms”.The additive of glycerol destroyed the balance and disrupted the arrangement of the fivemembered rings due to the hydrogen bonds between SSAS-C12and glycerol, which gave rise to a twisting in the headgroup and the alkyl chains of SSAS-C12oriented towards the same side.In this way, the SSAS-C12/glycerol complex possessed smaller size and faster diffusion rate (Fig.4).The water and glycerol droplets easily rebounded and splashed since they cannot infiltrate the micro/nano structures of the superhydrophobic leaf surface (Fig.5b1-b3).Due to the large shape and slow diffusion rate of SSASC12self-associations, the micro/nano structure of the superhydrophobic surface cannot be completely infiltrated during the short contact period from spreading to retracting,resulting in droplets fragmentation (Fig.5c1-c3).For SSASC12/glycerol binary additive (Fig.5d1-d3), the SSASC12/glycerol complex with the alkyl chains orienting toward the same side can diffuse and adsorb to the newly formed water/air and water/solid interfaces rapidly.Because of the fast diffusion rate, SSAS-C12/glycerol effectively diffuse and adsorb to the air/liquid/solid interface.In this case, it can change the wettability of solid surface when the droplet contacts the solid surface, thus leading to hardly receding behavior and a large wetting area19.In this way, the high-speed impacting droplets can firmly and quickly deposit on the superhydrophobic leaf surface.

Fig.5 MD simulations of surfactants with different hydroxy diffusion behaviors on the solid surface with glycerol.

Fig.6 Schematic illustration for enhancing the droplet impact deposition on a superhydrophobic leaf surface by SSAS-C12/glycerol binary additive.

4 Conclusions

In summary, we proposed a green and efficient system to enhance droplets deposition on superhydrophobic/hydrophobic leaves with low concentration.The binary system of SSASC12/glycerol is capable to inhibit the droplet bouncing and splashing through fast surfactant diffusion, resulting in the improvement of retention and deposition.The mechanism of surfactant/glycerol droplets spreading and deposition was also studied, which gave insights on the hydrogen bonding for droplet impacting and deposition on superhydrophobic/hydrophobic leaf surfaces.And the MD simulations indirectly proved the effect of droplet impacting and deposition on superhydrophobic/hydrophobic solid surface through the energy of surfactant/glycerol system and the distribution of surfactant molecules.It is known that low retention of pesticides on superhydrophobic/hydrophobic plants leaf surface is a major problem in pesticides spray.This work not only provides a constructive way to overcome the bouncing behavior of droplets advances but also prompts us to verify the importance of H-bond interactions in the screening of agricultural surfactants.

Supporting Information:available free of chargeviathe internet at http://www.whxb.pku.edu.cn.