Synthesis of mesoporous silica post-loaded by methyl eugenol as an environment-friendly slow-release bio pesticide

Salicylaldimine, furfuralimine and benzaldehyde imine were adopted to modify mesoporous silica (MCM) respectively denoted as Sal-MCM, Fur-MCM and Ben-MCM before loading methyl eugenol (Me) for pesticide delivery. Me was adsorbed by Schiff base mesoporous silica without destructing regular hexagonal pore structure verified by the characterization results. DSC result implied that Me in amorphous state which was distributed in the pores of the mesoporous silica. The loading content of Me-Sal-MCM, Me-Fur-MCM and Me-Ben-MCM 67.89%, 73.34% and 73.84% which was higher than Me-MCM without modification (67.35%).Because the electrostatic interaction and π-π interaction between Schiff base and Me strengthened the adsorption capacity of the carrier. And the electrostatic interaction played a more important role in interaction between Me and Schiff base modified mesoporous silica. As a result, Schiff base modified sustained release system also has significantly longer sustained release time with a sequence of Me-Sal-MCM > Me-Ben-MCM > Me-Fur-MCM in release speed in negative correlation with the electric potential sequence. The behaviors of their sustained release performance can be fitted by First order kinetic model before Schiff base modification. After modification, their sustained release behaviors were consistent with Korsmeyer-Peppas equation with non-Fickian diffusion mechanism indicating that main impact on the release process after modification was no longer mainly controlled by the difference of the concentration. Finally, the highest lure rate of the modified MCM (Me-Fur-MCM) equals to the 73% of the pure Me due to its highest BET surface area and strongest interaction with Me among the three Schiff base modified samples. Therefore, the environment-friendly slow-release bio pesticide with long service life was prepared to reduce the damage on the environment caused by pesticide.


Preparation of salicylaldimine, furfuralimine and benzaldehyde imine grafted on APTES.
According to the literature 19,20 , 4.42 g of APTES, 2.44 g of salicylaldehyde and 100 mL of ethanol were added into a flask and refluxed at 95 °C for 3 h. Ethanol was removed through rotary evaporation. 20 mL of dichloromethane was added, then the products washed with deionized water 3 times. The organic layer was extracted and standing for 12 h. Then the product was filtered to remove dichloromethane to attain salicylaldimine. The furfuralimine and benzaldehyde imine were prepared though the same method illustrated above while salicylaldehyde was replaced by furldehyde and benzaldehyde respectively.

Preparation of MCM, Sal-MCM, Fur-MCM and Ben-MCM.
According to previous research 15 , co-condensation method was adopted to prepare salicylaldehyde modified mesoporous silica (Sal-MCM. 1.0 g of CTAB, 100 mL deionized water and 70 mL of ammonia were added to the flask to be dissolved at 60 °C with stirring. And 5 g of TEOS was added to the solution dropwise. 1 hour later, 1 g of as synthesized salicylaldimine was added and kept on reacting for 6 h before being crystalized at room temperature, filtered, washed and dried. Finally, the template was removed by ethanol to attain Sal-MCM Using this approach, the final products Fur-MCM and Ben-MCM (i.e salicylaldimine replaced by furfuralimine and benzaldehyde imine, respectively) were prepared. In comparision, MCM was prepared as mentioned above without adding salicylaldimine.
Loading of methyl eugenol. According to previous study 19 , the supported Me was prepared via impregnation. The mesoporous silica were activated under vacuum at 80 °C for 12 h. And 0.25 g of samples was immersed in 250 mg Me at room temperature for 1 h, then filtered and dried. The samples obtained were denoted as Me-Sal-MCM, Me-Fur-MCM and Me-Ben-MCM respectively according to the different carrier. And the loading content was calculated through TG method.
Measurements. The small angel X-ray diffraction (SAXD) were performed using a Bruker AXS D8 X-ray diffractometer (Bruker, Germany) with Cu radiation (λ = 1.5418 Å) and a graphite monochromator at 25 °C, 40 kV, and 30 mA. The measurements were scanned at 2°/min (angular range 2θ = 0.5~10°) in 0.02° step size. FTIR spectra were recorded in the region 4000-400 cm −1 by a Spectrum100 Fourier infrared spectrometer (PerkinElmer, USA) using the KBr squash technique. The gold particles were sprayed on the surface of samples under protection of N 2 and the samples were characterized by an S4800 scanning electron microscope (Hitachi, Japan) to observe the surface topography. TEM observation was conducted on a FEI Tecnai G2 F20 transmission election microscope (Thermo Fisher Scientific, USA). BET surface area of samples was determined by N 2 adsorption isotherms at 77 K, operated on Quadrasorb SI adsorption equipment (Quantachrome, USA). The samples were degassed at 200 °C for 12 h in vacuum before N 2 adsorption experiment. A Q200 differential scanning calorimeter (TA Instruments, USA) was used to conduct differential scanning calorimetry and detect the crystalline degree of the mesoporous silica in the particles over a heating range of 0~200 °C and a heating rate of 10 °C/min under the protection of N 2 (flow rate 50 mL/min). Thermal gravity (TG) measurements of these samples were carried out on an SDT-Q600 thermogravimetric analyzer (TA Instruments, USA) analyze the heat stability of particles over the heating range of 40~600 °C under the condition of N 2 flow rate 50 mL/min and heating rate 10 °C. X-ray photoelectron spectra (XPS) were recorded on a ESCALAB250XI spectrometer (Thermo Fisher Scientific, USA) under a vacuum of ~2 × 10 −7 Pa. Charging effects were corrected by adjusting the main C 1 s peak to a position of 284.8 eV. The zeta potential of the samples was investigated with a Zetasizer Nano ZS (Malvern, UK) in water at pH 7 through ultrasonic dispersion.
where y is the experimental accumulated rate of Me (%) at time t, respectively. K is the rate constant for the kinectic model. The correlation coefficient (R 2 ) and the release exponent (n) were used to determine the best-fit kinetic model and the mechanism of the drug release.
Attraction of bactrocera dorsalis test. According to the literature 8,19 , A certain among of male Bactrocera Dorsalis with sexual maturation were transferred to a 45 cm × 45 cm × 45 cm cage. Steiner trap contained 100 mg sample was placed in the center of the cage. And the number of Bactrocera Dorsalis trapped in the Steiner trap was recorded to calculate the lure rate.

Results and discussion
Structure characterization. SEM and TEM images of MCM samples before and after modification were depicted in Figs. 1 and 2. The regular hexagonal pore structure and the structure integrity from MCM 24 was well-maintained 25 after modification and the orderings of the (100) and (110) was not affected by Schiff base co-condensation. The particle size of the four samples were 833 nm, 789 nm, 701 nm and 763 nm respectively for MCM, Sal-MCM, Fur-MCM and Ben-MCM measured from the SEM images. And the surface of MCM become rough with small particles formed on the surface as the SEM images shown in Fig. 1 after modification. Also a layered of shell structure appeared on the surface as the TEM images from Fig. 2 showed due to the precipitation of silane coupling agent 26 . Figure 3 depicted the SAXD patterns of the samples. Four characteristic peaks at 0.860°, 2.439°, 4.000° and 4.540° were ascribed to (100), (110), (200) and (210) crystal faces respectively. And the crystal faces of (100) and (110) were also shown in the TEM images as illustrated in Fig. 2 representing for the regular hexagonal pore structure in accordance with SEM and TEM images. However, the (100) diffraction of the modification samples was shown to further shift toward smaller 2θ value due to the block of the pore by Schiff bass. The 2θ value shift of (200) and (210) crystal faces also happened for Fur-MCM. What's more, the peaks ascribed to (200) and (210) crystal faces disappeared for Sal-MCM and Ben-MCM which proved that benzene ring was introduced to the system and decreased its degree of orderliness 27 .
As shown in Fig. 4a, the N 2 adsorption/desorption isotherms of MCM, Sal-MCM, Fur-MCM and Ben-MCM belong to Langmuir IV (the slope of it was decreasing) which indicated that their pore size was relatively small and also confirmed by the pore size distribution results calculated by DFT method 28 as shown in Fig. 4b. The N 2 adsorption/desorption isotherms in Fig. 4 do not overlap at relative pressures <0.2 were due to N 2 chemisorbed by silanol on the surface of the pores which confirmed by our previous researches 18,26 . The rapid shift of N 2 adsorption isotherms caused by capillary condensation inside the pore for MCM disappeared after modification. And the BET surface, pore size and pore volume decreased after modification due to the pore block by Schiff base. And the BET surface of Sal-MCM, Fur-MCM and Ben-MCM after modification were 225.5 m 2 /g, 413.4 m 2 /g and 274.8 m 2 /g respectively which illustrated the difference in the degree of pore block by three different Schiff base as shown in Table 1 which was much higher than our previous researches (<200 m 2 /g) due to the templates removed from the samples. Fur-MCM had the highest BET surface among and largest pore size and pore volume among the Schiff base modified MCM indicating that the surface of the pores from MCM were homogeneously grafted by furalimine probably due to its strongest interaction towards to the substrate.
The particle diameters and Zeta potential studies of MCM, Sal-MCM, Fur-MCM and Ben-MCM were investigated at pH 7 as listed in Table 2. The particle size calculated by DLS method was bigger than the particle size shown in SEM images due to the agglomeration and solvation effect since DLS measurements are conducted in solution according with previous research 29 . The particle size of Sal-MCM, Fur-MCM and Ben-MCM decreased after modification. The decrease in particle size was caused by the effect of Schiff base on TEOS hydrolysis during the condensation. And the sufficient Si-OH groups on the surface MCM before modification made it Zeta potential negative. After modification, Zeta potential of the samples shifted to be positive which confirmed the introduction of Schiff base into the mesoporous samples. Among them, Fur-MCM has the highest Zeta potential leading to its strongest electrostatic interaction with Me in consistent with BET results.
The   www.nature.com/scientificreports www.nature.com/scientificreports/ symmetric and nonsymmetrical C-H stretching vibration bands from amino group for Schiff base which indicated the Schiff base was successfully grafted to MCM. After loading Me, the characteristic peaks of Me located at 3060, 2990, 2930, 1635, 1590 and 1510 cm −1 which proved that the Me was successfully adsorbed by mesoporous silica 31 . The IR spectra of Me showed a peak at 1724 cm −1 for carbonyl group which was gradually disappearing after loaded on mesoporous silica due to the interaction with MCM substrate. Intensity of peaks between 2800    www.nature.com/scientificreports www.nature.com/scientificreports/ and 3100 cm −1 ascribed to the C-H stretching band was increased significantly showing that Me was successfully impregnated with mesoporous silica.
The surface elements chemical states were observed with XPS analysis as shown in Fig. 6 and Table 3. The binding energy (BE) negative shift of Si 2p as shown in Fig. 6a,b were observed after Schiff base modification due to the reaction between Si-OC 2 H 5 from Schiff base and Si-OH from MCM 15  Thermogravimetric analysis (TG) was used to investigate the thermal stability. As shown in Fig. 7, the loss in mass for MCM below 100 °C was caused the evaporation of water adsorbed by the samples 32 in accordance with DSC results as shown in Fig. 8. The significant loss in mass occurred within the temperature range 160-350 °C which was caused by the decomposition of Schiff base on the mesoporous silica for Sal-MCM and Fur-MCM. While for Ben, 20% mass still existed even above 350 °C indicating that the mass loss was caused by the carbon-  www.nature.com/scientificreports www.nature.com/scientificreports/ base modification which strengthened the interaction with mesoporous silica by π-π interaction between the benzene ring from Me and aromatics from Schiff base and electrostatic interaction between Schiff base with positive charge and Me with negative charge. And Fig. 9 also depicts the sustained release curves with a sequence of Me-Sal-MCM > Me-Ben-MCM > Me-Fur-MCM in release speed in negative correlation with the electric potential sequence shown in Table 2  Kinetics study. To further understand the sustained release mechanism, the data of sustained release of Me and Me loaded in various mesoporous silica were fitted to different kinetic models ( Table 4). The sustained release curves of Me and Me-MCM were in accordance with first-order kinetic equation proving the barrier-free diffusion of the drug. While Me-Sal-MCM, Me-Fur-MCM and Me-Ben-MCM were in consistent with Korsmeryer-Pappas kinetic equation. The diffusion coefficients K 1 of the fitting equations were between 0.45~1 controlled by a non-Fickian diffusion mechanism 33,34 . The sustained release behavior of MCM changed after modification because the difference of the concentration was no longer the main factor on controlling the release performance due to the strengthened interaction between Me and the substrate. Therefore, the service life of Me was prolonged.   www.nature.com/scientificreports www.nature.com/scientificreports/ Me. In summary, the sustained release system get rid of the organic solution and increases the Me service life without affecting its performance significantly.

Sustained release test.
As depicted in Fig. 11, the drug release process of mesoporous silica was illustrated by characterization results and kinetic study in this paper. Schiff base was the link between modified mesoporous silica and Me by π-π interaction between the benzene ring from Me and aromatics from Schiff base and electrostatic interaction between Schiff base with positive charge and Me with negative charge. And the electrostatic interaction played a more important role in interaction between Me and Schiff base modified mesoporous silica. As a result, the release performance of Me was proved and the service life of Me would be significantly prolonged in practical usage. www.nature.com/scientificreports www.nature.com/scientificreports/ conclusions In conclusion, Schiff base (salicylaldimine, furfuralimine and benzaldehyde imine) modified mesoporous silica was prepared by co-condensation method. Then Me was post-loaded in the mesoporous silica after removing the templates in the pores. The regular hexagonal pore structure was well-maintained without agglomeration after modification. The existence of interaction between Schiff base and Me was confirmed by the characterization

Release model Materials
Parameters  www.nature.com/scientificreports www.nature.com/scientificreports/ results. And the negative correlation is found between sustained speed of Me and zeta potential of the samples indicating that the electrostatic interaction played a more important role in the interaction between Me and Schiff base modified mesoporous silica. Me is distributed homogeneously in amorphous state in the pores of the mesoporous silica confirmed by DSC results. The loading content of Me-Sal-MCM, Me-Fur-MCM and Me-Ben-MCM 67.89%, 73.34% and 73.84% which was higher than Me-MCM without modification (67.35%). Their sustained release curves could be described by Korsmeyer-Peppas equation in consistence with non-Fickian diffusion mechanism after Schiff base modification. Finally, the attraction of Bactrocera Dorsalis test showed the highest lure rate of the modified MCM (Me-Fur-MCM) equals to the 73% of the pure Me. In short, this sustained release system can avoid the usage of organic solution for dissolution and prolonged its service life without affecting its performance significantly through the enhancement of interaction between drug and carrier through Schiff base modification.

Data availability
The data generated or analyzed during the current study are available from corresponding author upon reasonable request.