Controllable morphology transition from vesicular to worm-like to vesicular multilamellar mesoporous silica induced by β-cyclodextrin

Abstract

Changes in the morphology of mesoporous silica from vesicular to worm-like to vesicular multilamellar were controlled by adding appropriate amounts of β-cyclodextrin to mixed cetyltrimethylammonium bromide/didodecyldimethylammonium bromide surfactant aggregates.

---

Mesoporous silica materials, especially multi-shelled forms, have attracted much interest because of their tunable pore size, morphological properties, large surface area, and large pore volume.¹ Thus, mesoporous silica materials are suitable candidates for several potential applications in catalysis, adsorption and separation, sensors, and controlled drug release as well as templates for other materials.² The controlled assembly of multi-shelled mesoporous silica with different morphologies has gained increased research attention worldwide.³ Multi-shelled mesoporous silica materials with three or more shells are expected to exhibit improved performances than their single- or double-shelled counterparts for applications in drug delivery with prolonged release time or heterogeneous catalysis. Multi-shelled mesoporous silica with different morphologies can be fabricated using single surfactants, mixtures of cationic and anionic surfactants, or nonionic and ionic surfactants as templates.⁴ Zhou et al.^3e prepared vesicle-like silica with a hierarchical structure, using 1,3,5-triisopropylbenzene (TIPB) as a hydrophobic agent and the triblock copolymer Pluronic P123 as a template. Gu et al.⁵ synthesized multilayer silica vesicles with cetyltrimethylammonium bromide (CTAB) and C₃F₇O(CFCF₃CF₂O)₂CFCF₃CONH(CH₂)₃N⁺(C₂H₅)₂CH₃I⁻ as co-templates. TIPB in the co-surfactant systems of CTAB and anionic sodium dodecyl sulfate (SDS) acts as an expander that interacts and enlarges the total volume of hydrophobic chains. Thus, stable vesicular silica morphologies are constructed.⁶ We previously reported one approach to obtain silica materials from mesoporous spheroids featuring worm-like porous silica to onion-like hollow silica with ordered multilamellar shells. We used didodecyldimethylammonium bromide (DDAB) and CTAB as co-surfactants.⁷ The structural properties could be more easily controlled in these methods under different conditions, such as swelling agents, solvents, and manipulation of synthesis conditions.⁸ However, the morphology of multi-shelled mesoporous silica prepared with a double surfactant is mostly confined to the vesicles. These preparation methods can only control the number of layers, particle size, and wall thicknesses but cannot realize the transition among different morphologies.

Host–guest chemistry does provide such possibility.⁹ Cyclodextrins (CDs) are oligosaccharides most commonly composed of 6, 7, or 8 glucosidic units bearing the names α-, β-, and γ-CD, respectively. The outer surface of CDs has abundant hydrophilic hydroxyl groups, facilitating the water solubility of CDs. The inner cavity is composed of glucoside methylene groups, resulting in the nonpolar hydrophobic characteristic of the cavity. The hydrophobic cavity is an ideal harbor in which molecules with poor water solubility can shelter their hydrophobic parts. The surfactant possesses a hydrophobic tail, which can insert the nonpolar hydrophobic cavity of CDs to form the surfactant/CD-bound composite. CDs can form host–guest complexes with most surfactants with high binding constants by integrating surfactants into the CD cavities.¹⁰ Jiang et al.¹¹ reported that β-CD and SDS could self-assemble into lamellar structures. The lamellae transformed to microtubes then to vesicles upon dilution. Xu et al.¹² reported that the characteristic length of the aggregates significantly increased in 93 mM aqueous solutions of (1,1,3,3-tetramethylbutyl)phenyl-polyethylene glycol upon β-CD addition. The observed aggregate growth (micelle to vesicle transition) was induced by β-CD. Nan et al.¹³ synthesized high-quality ordered MCM-41 silica via the self-assembled supermolecules of β-CD with CTAB as a structure-directing agent. Morphologies of the samples changed from elliptical to spherical to irregular with the addition of β-CD.

Synthesis of multilamellar mesoporous silica with different morphologies from CDs and a mixed cationic/cationic surfactant system has not been reported yet. In this paper, structural transition of the β-CD recognition-controlled phase in the mixed cationic/cationic surfactant is achieved. The transformation of vesicular silica with six to seven layers → worm-like silica with three to four layers → vesicular silica with three to four layers is observed by increasing the quantity of β-CD. This study reports for the first time the controlled synthesis of multilamellar vesicular silica (MLV) and multilamellar worm-like silica (MLW) with β-CD/CTAB/DDAB under mild conditions and convenient operation. Five samples were prepared by varying the quantity of β-CD from 0.05 g to 0.25 g. The as-prepared silica samples were labeled as MLS-n, where n = 0.05, 0.1, 0.15, 0.2, and 0.25, and the quantity of β-CD were 0.05, 0.1, 0.15, 0.2, and 0.25 g, respectively. The whole process produced two kinds of morphologies, namely, MLV and MLW. The diameters of MLV and MLW were estimated to be 60–140 and 60–90 nm, respectively, with interlamellar voids of ca. 2–4 nm and shell thicknesses of ca. 2–4 nm. The number of layers of MLV and MLW varied from six to seven to three to four, and β-CD was added to control the surfactant phase structural transition in this system.¹⁴ The application of prepared MLV and MLW in drug loading and release was also investigated, using metformin hydrochloride (MH) as model drug.

Fig. 1 and S1 (ESI†) show the high-resolution transmission electron microscopy (HRTEM) images of the as-synthesized MLV and MLW products induced by different quantities of β-CD. Vesicular silica with six to seven layers and a size of 80–90 nm was prepared with 0.142 g of CTAB and 0.112 g of DDAB as a structure directing agent, as previously prepared by our group, without β-CD.⁷ The addition of small amounts of β-CD (0.05 g) had negligible effects on the morphology of the sample. The morphology of the particles remained vesicular silica with six to seven layers, although the diameters slightly increased from 80–90 nm to 90–140 nm, as shown by the HRTEM images in Fig. 1a and d. However, a drastic change occurred in the morphology from MLV to MLW when the quantity of β-CD was increased to 0.1 g (Fig. 1b and e). Almost all MLVs were converted into MLW. MLW had a worm-like shape with width of 60–90 nm and length of 300–500 nm, a multilamellar structure with three to four layers, and a large cavity. This unique structure is clearly shown by the insets in Fig. 1e. Further increase in the quantity of β-CD from 0.1 g to 0.15 g increased the number of MLV. Significant amounts of MLV with three to four layers coexisted with minor amounts of MLW with three to four layers (Fig. 1c and f). Compared with MLS-0.05, the diameters of MLS-0.15 decreased to 60–100 nm, whereas the number of layers varied from six to seven to three to four. Compared with MLS-0.1, the length of MLS-0.15 decreased from 300–500 nm to 250–350 nm, but its width was almost constant. The number of MLV increased until MLV with 60–100 nm diameter dominated all the images as the quantity of β-CD was further increased from 0.15 g to 0.2–0.25 g (Fig. 1g–j). The number of layers of MLV was three to four, although their size distribution did not evidently change, similar to MLV-0.15. In summary, the interlamellar voids of MLV and MLW were almost unchanged at ca. 2–4 nm. The phase structural transitions from MLV to MLW to MLV can be controlled by the addition of β-CD because of β-CD molecular recognition. Therefore, β-CD is a key parameter in manipulating morphology.¹⁵

Fig. 1 HRTEM images of MLS-0.05 (a and d), MLS-0.1 (b and e), MLS-0.15 (c and f), MLS-0.2 (g and h) and MLS-0.25 (i and j) at different magnifications. The inset is the corresponding magnified images of (e).

FESEM micrographs of the samples are illustrated in Fig. 2 and S2 (ESI†). The samples showed a spheroidal morphology with diameters of 60–140 nm or a worm-like morphology with widths of 60–90 nm and lengths of 250–500 nm. Several spherical vesicular silicas were broken as shown in the FESEM micrographs in Fig. 2d and e, indicating that the particles are hollow spheroids. Furthermore, some ruptured worm-like silicas also had hollow structures as shown in Fig. 2b (inset).

Fig. 2 FESEM images of MLS-0.05 (a), MLS-0.1 (b), MLS-0.15 (c), MLS-0.2 (d), and MLS-0.25 (e), the inset is the corresponding magnified image of (b). The circles in (d) and (e) point to the ruptured vesicular silicas.

The sample structure was further investigated by performing small-angle X-ray diffraction (SAXRD) on the calcined samples. The results are shown in Fig. S3 (ESI†). All samples had one broad peak at approximately 2θ = 2.2°.¹⁶ Their corresponding d-spacing values were approximately 4.0 nm, which were in good agreement with the TEM results of the lamellar structure or multi-layered vesicular silica structure.⁷ However, only one peak for all samples suggested that mesoporous silica with sponge-like wall structure was not well-ordered enough.¹⁷ Compared with other samples, the SAXRD pattern of MLS-0.05 showed a narrower diffraction peak, and diffractions with relatively high intensity were also discernible. This result indicates that the structure of MLS-0.05 was more orderly.

The nitrogen adsorption–desorption isotherms and corresponding pore size distributions determined from the adsorption branch were also recorded and are shown in Fig. S4 (ESI†). All samples exhibited typical type IV isotherms,¹⁸ indicating the presence of mesopores. H4-type hysteresis loops at the relative pressure (P/P₀) in the range of 0.45–0.99 were observed for each sample,¹⁹ suggesting the adsorption of N₂ molecules in the hollow voids.²⁰ Notably, the pore size distribution was extremely narrow, with only one peak centered at approximately 2.7 nm in the pore size distribution curves. The pore sizes, surface areas, and pore volumes of the five samples are summarized in Table S1 (ESI†) for comparison. Compared with other samples, MLS-0.05 showed the largest BET surface area (602 m² g⁻¹) and smallest total adsorption pore volume (0.70 cm³ g⁻¹). Slight differences were observed among MLS-0.1, MLS-0.15, MLS-0.2, and MLS-0.25.

The results described above for MLV and MLW materials obtained with β-CD of different quantities were analyzed. Then, we propose a mechanism for the formation of the MLV and MLW framework in terms of the β-CD-induced phase transformation of the CTAB/DDAB aggregates from spherical vesicles to worm-like vesicles to spherical vesicles again as shown in Scheme 1. The β-CD molecules play an important role in accurately modulating the structures of mesoporous silica in the nonstoichiometrical cationic/cationic surfactant systems.²¹ Hydrophobic and electrostatic interactions between the CTAB and DDAB (molar ratio of CTAB : DDAB = 1 : 0.625) surfactants in the water solution system drove the formation of the stable spherical vesicles with six to seven layers.⁷ Adding a small amount of β-CD (0.05 g) to the system resulted in negligible effects on the morphology of vesicles. Aggregate morphologies were still spherical vesicles with six to seven layers. Increasing the quantity of β-CD from 0.05 g to 0.1 g resulted in the selective transfer of the major component CTAB of mixed surfactant systems from the spherical vesicle aggregates to form host–guest complexes by including CTAB hydrophobic tails into β-CD cavities.^14,15 The selective binding of β-CD shifted the surfactant compositions in the aggregates. Thus, aggregate surface charge density decreased, and spherical vesicles became elongated, inducing the transition from spherical vesicles to worm-like vesicles with a decline in the number of worm-like vesicle layers. Further increase in the quantity of β-CD from 0.1 g to 0.15 g led to the combination of more CTAB. Decreasing the amount of CTAB led to the effective reduction in the worm-like vesicle inter-bilayer repulsive forces to favor small vesicle aggregates. In this process, spherical vesicles could coexist with worm-like vesicles. Continuous increase in the amount of β-CD from 0.15 g to 0.2–0.25 g resulted in the transformation of all aggregates to spherical vesicles again. The spherical vesicles were smaller than that at 0.05 g of β-CD. Increasing the amount of β-CD led to the transformation of aggregated CTAB/DDAB surfactants from spherical vesicles with six to seven layers to worm-like vesicles with three to four layers to smaller spherical vesicles with three to four layers. Finally, the hydrolysis and condensation of TEOS and subsequent calcination resulted in the removal of the organic components, producing MLV and MLW.

Scheme 1 Schematic representation of the proposed assembly mechanism for the formation of MLV and MLW.

Previous studies showed that the adsorption and release of drugs from different mesoporous silica materials are mainly diffusion-controlled in relation to particle size, BET surface areas, pore size, and the surface properties of mesoporous silica.²² The N₂ adsorption–desorption isotherms and the corresponding BJH pore size distribution of the samples determined from the adsorption branch after MH adsorption are shown in Fig. S5 (ESI†). The S_BET, V, and D of the samples after MH adsorption are listed in Table S1 (ESI†). In contrast to the parent MLS-n, the surface areas and pore volume of MH-loaded MLS-n decreased. This characteristic indicates that MH was successfully loaded into the lamellar channels and the cavity of MLS-n from the micropores in the SiO₂ shells.⁷ The MH loaded on MLS-0.05, MLS-0.1, MLS-0.15, MLS-0.2, and MLS-0.25, which were calculated using eqn (S1) (ESI†) were 29.1 wt%, 45.7 wt%, 42.1 wt%, 30.8 wt%, and 40.6 wt%, respectively. These results indicated that the amount of drug loading was dependent on cavity size and number of layers of the samples. MLS-0.1 presented the highest drug loading capacity among the samples, which was attributed to its largest cavities. However, MLS-0.05 showed the lowest drug loading capacity because it had the most number of layers.

The MH release profiles of MLS-n in PBS (pH 6.8) are shown in Fig. S6 (ESI†). The diffusion of drug molecules from the pores was largely dependent on the nature of the interaction of the drug molecule with the pore and the intrinsic mobility of the drug molecules inside the pores.²³ The release profile generally showed a slow and steady release pattern. This trend may be attributed to the multilayer structure of vesicular silica and worm-like silica. The first-order model was applied to evaluate the mechanism that controls the release kinetic process. The results are shown in Fig. S7 (ESI†). The first-order kinetics model showed good linear plots for MH released from MLS-n which the first-order kinetics model fitted well to the MH release. The parameters of release kinetic model were determined and are tabulated in Table S2 (ESI†). The first-order rate constant (k₁) can be calculated from the rate equation log C_t = log C_o − k₁ × t/2.303.²⁴ MLS-0.05-MH showed the lowest k₁ values, indicating slowest release rate.

For the first time, we demonstrated the morphological changes of multilamellar silica microstructures using mixed CTAB/DDAB surfactant as a structure-directing agent and β-CD as an inducing agent. β-CD is a key parameter in manipulating the morphology of silica. Increasing the amount of β-CD resulted in the morphological transformation of silica from MLV to MLW then to MLV. Small amounts of β-CD (0.05 g) did not change the MLV morphology of the particles. Morphology drastically changed from MLV to MLW when the quantity of β-CD was increased from 0.05 g to 0.1 g. The majority of MLV particles were converted into MLW. The number of MLV conversely began to increase, and significant amounts of MLV coexisted with minor amounts of MLW when the quantity of β-CD was further increased from 0.1 g to 0.15 g. Continuous increase in the amount of β-CD from 0.15 g to 0.2–0.25 g increased the number of MLV until MLV dominated the morphology. The present work can provide a novel perspective to fine-tune microstructures in self-assembling systems. These mesoporous materials can be applied in controlled release and selective adsorption–desorption, among others.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Grant no. 51372124, 51572134, 51503108), the Natural Science Foundation of Shandong Province (Grant no. BS2015CL018), and the Program for Scientific Research Innovation Team in Colleges and Universities of Shandong Province.

Notes and references

1. (a) E. M. Johansson, J. M. Córdoba and M. Odén, Microporous Mesoporous Mater., 2010, 133, 66; (b) Y. X. Zhao, C. G. Gao, Y. X. Li and T. M. Zhang, Microporous Mesoporous Mater., 2009, 124, 42; (c) H. M. Chen and J. H. He, Chem. Commun., 2008, 37, 4422.
2. (a) H. M. Lin, G. S. Zhu, J. J. Xing, B. Gao and S. L Qiu, Langmuir, 2009, 25, 10159; (b) C. C. Lu and K. H. Liao, Sens. Actuators, B, 2010, 143, 500.
3. (a) X. Gu, C. L. Li, X. H. Liu, J. W. Ren, Y. Q. Wang, Y. L. Guo, Y. Guo and G. Z. Lu, J. Phys. Chem. C, 2009, 113, 6472; (b) M. Bellino, A. Regazzoni and G. A. A. Soler-Illia, ACS Appl. Mater. Interfaces, 2010, 2, 360; (c) G. W. Zhou, K. K. Fung, L. W. Wong, Y. J. Chen, R. Renneberg and S. H. Yang, Talanta, 2011, 84, 659; (d) Y. S. Li and J. L. Shi, Adv. Mater., 2014, 26, 3176; (e) G. W. Zhou, Y. J. Chen, J. H. Yang and S. H. Yang, J. Mater. Chem., 2007, 17, 2839; (f) Y. Hu, J. O. Jensen, W. Zhang, L. N. Cleemann, W. Xing, N. J. Bjerrum and Q. F. Li, Angew. Chem., Int. Ed., 2014, 53, 3675; (g) J. Liu, S. B. Hartono, Y. G. Jin, Z. Li, G. Q. Lu and S. Z. Qiao, J. Mater. Chem., 2010, 20, 4595.
4. (a) B. Sun, G. W. Zhou and H. J. Zhang, Prog. Solid State Chem., 2016, 44, 1; (b) H. L. Wang, G. Agrawal, L. Tsarkova, X. M. Zhu and M. Möller, Adv. Mater., 2013, 25, 1017; (c) Y. Yamauchi, J. Ceram. Soc. Jpn., 2013, 121, 831; (d) K. Zhang, L. L. Xu, J. G. Jiang, N. Calin, K. F. Lam, S. J. Zhang, H. H. Wu, G. D. Wu, B. Albela, L. Bonneviot and P. Wu, J. Am. Chem. Soc., 2013, 135, 2427.
5. X. Gu, C. L. Li, X. H. Liu, J. W. Ren, Y. Q. Wang, Y. L. Guo, Y. Guo and G. Z. Lu, J. Phys. Chem. C, 2009, 113, 6472.
6. C. C. Wu, G. W. Zhou, X. J. Jiang, J. Y. Ma, H. Y. Zhang and H. B. Song, Process Biochem., 2012, 47, 953.
7. Y. Zhang, G. W. Zhou, B. Sun, M. N. Zhao, J. Y. Zhang and F. J. Chen, Chem. Commun., 2014, 50, 2907.
8. (a) B. Yang, K. Edler, C. Guo and H. Z. Liu, Microporous Mesoporous Mater., 2010, 131, 21; (b) H. M. Chen, T. Hu, X. M. Zhang, K. F. Huo, P. K. Chu and J. H. He, Langmuir, 2010, 26, 13556.
9. Y. Zhou, D. S. Wang, S. L. Huang, G. Auernhammer, Y. J. He, H. J. Butt and S. Wu, Chem. Commun., 2015, 51, 2725.
10. (a) G. Crini, Chem. Rev., 2014, 114, 10940; (b) A. D. Wang, J. B. Huang and Y. Yan, Soft Matter, 2014, 10, 3362; (c) L. X. Jiang, Y. Yan and J. B. Huang, Soft Matter, 2011, 7, 10417; (d) N. Funasaki and S. Neya, Langmuir, 2000, 16, 5343; (e) G. G. lez-Gaitano, A. Crespo and G. Tardajos, J. Phys. Chem. B, 2000, 104, 1869.
11. L. X. Jiang, Y. Peng, Y. Yan and J. B. Huang, Soft Matter, 2011, 7, 1726.
12. H. N. Xu, S. F. Ma and W. Chen, Soft Matter, 2012, 8, 3856.
13. Z. D. Nan, X. Xue, W. G. Hou, X. Yan and S. H. Han, J. Solid State Chem., 2007, 180, 780.
14. L. X. Jiang, Y. Yan and J. B. Huang, Adv. Colloid Interface Sci., 2011, 169, 13.
15. L. X. Jiang, M. L. Deng, Y. L. Wang, D. H. Liang, Y. Yan and J. B. Huang, J. Phys. Chem. B, 2009, 113, 7498.
16. P. T. Tanev, Y. Liang and T. J. Pinnavaia, J. Am. Chem. Soc., 1997, 119, 8616.
17. H. M. Lin, F. Y. Qu, X. Wu, M. Xue, G. H. Zhu and S. L. Qiu, J. Solid State Chem., 2011, 184, 1415.
18. L. X. Zhang, P. C. Li, X. C. Liu, L. W. Du and E. K. Wang, Adv. Mater., 2007, 19, 4279.
19. P. Yuan, S. Yang, H. N. Wang, M. H. Yu, X. F. Zhou, G. Q. Lu, J. Zou and C. Z. Yu, Langmuir, 2008, 24, 5038.
20. C. X. Lin, P. Yuan, C. Z. Yu, S. Z. Qiao and G. Q. Lu, Microporous Mesoporous Mater., 2009, 126, 253.
21. D. Wang, P. F. Long, R. H. Dong and J. C. Hao, Langmuir, 2012, 28, 14155.
22. L. Chen, Z. Zhang, X. M. Yao, X. F. Chen and X. S. Chen, Microporous Mesoporous Mater., 2015, 201, 169.
23. A. Nieto, M. Colilla, F. Balas and M. V. Regi, Langmuir, 2010, 26, 5038.
24. M. M. Ayad, N. A. Salahuddin, N. L. Torad and A. A. El-Nasr, RSC Adv., 2016, 6, 57929.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details and characterization data. See DOI: 10.1039/c6ra13259f

---