Template-based formation of ultrasound microbubble contrast agents

Abstract

Precisely controlling microbubble size is critical for medical ultrasound imaging, where large microbubble contrast agents may lead to pulmonary microvascular embolization. In this study, we reported a facile method to fabricate the laminated ultrasound contrast agents with narrow size distribution. First, we prepared monodisperse silica particles (∼1.0 μm in diameter) as a core template. Then an amphiphilic shell was coated on the SiO₂ particles surface via amine–anhydride modification. After etching the SiO₂ cores by HF, the formed hollow structured particles were further PEGylated with free carboxylate residues. After dialysis and freeze-drying, the as-prepared microbubbles showed a narrow size distribution (diameter = 1.1 μm, PDI = 0.36). The microbubbles showed excellent stability for enhanced ultrasound contrast imaging for at least 20 minutes, resulting from the observed firm-shelling through electron microscopy. The laminated shells of the microbubbles have an average thickness of 100 nm composed of multiple aliphatic composites that efficiently reduce the leakage of inert gas. Further, these microbubbles also showed good biocompatibility when co-culturing with BNL CL2 cells in vitro. Based on the above evidence, this improved method for fabricating uniform ultrasound contrast agents can further translate into many biomedical applications.

---

1. Introduction

X-ray, computed tomography (CT), ultrasound (US), and magnetic resonance imaging (MRI) are used for medical diagnostic imaging and therapeutic guidance. Ultrasound stands out for being portable and of relatively low price for examination, non-invasive, real-time imaging and convenience for monitoring patients, making it an appealing imaging tool.^1,2 However, ultrasound's major shortcoming is its inferior-quality image when compared to other imaging modalities. Ultrasonic imaging has weak echogenicity difference between various tissues and often obstructs diagnostic accuracy. Ultrasound contrast agents are used to improve the image quality^3,4 which helps the diagnostician to distinguish between normal and abnormal conditions.

The development of contrast agents for US imaging came about as the result of an accidental discovery in the late 1960s that the presence of gas bubbles in the circulation could significantly enhance ultrasound signal intensity.^5,6 US contrast agents have different formulations like solid particles in suspension, liquid droplets (emulsion), gas bubbles, encapsulated gases or liquids.^7,8

Currently, gas-filled echogenic microbubbles are mostly used as a US contrast agent to improve the US image quality.^9–12 A variety choice of gas core such as air, perfluorocarbon and sulfur hexafluoride¹³ have been used to provide compressible and echogenic properties. Also, a variety choice of the microbubbles shell such as albumin, lipid bilayer, biocompatible polymers give the elastic and gas-protective functions. A surfactant or polymer^14–17 stabilized gas microbubbles to become the most effective type of contrast agent available for ultrasound sonography. Such microbubbles are prepared by sonication or emulsification methods^8,11,18,19 which typically result in broad size distribution from one to tens of micrometers.^11,18,19 However, the oversize microbubbles could have difficulties for passing through the pulmonary capillary bed.^11,20 Recently, scientists proposed new ideas to improve the size uniformity such as microfluidic processing,²¹ inkjet printing,²² coaxial electrohydrodynamic atomization^23,24 but need higher costs. The core–shell methods were used to prepare hollow structures due to controllable size, shape, and chemical reaction, thus expanded its applications.^25–27 Lin et al.²⁷ firstly reported a template-based synthesis that use the removable polystyrene cores to produce hollow bubbles laminated with thin layers of silica. Based on these remarkable findings, method to produce homogeneous microbubbles with good biocompatibility is crucial for ultrasound contrast agent development.

In this study, we propose a silica-template method to create uniform microbubbles without using any organic solvents. Hydrofluoric acid was selected to etch the silica core, leaving the thin shells composed of amphiphilic octenylsuccinic anhydride derivatives. Additional PEGylation is selected for microbubble stability^28,29 and for further eliminating rapid clearance from the mononuclear phagocyte system which ended up in the liver.^1,17 Serum proteins adsorption and uptake by the mononuclear phagocyte system can be drastically reduced.³⁰ PEGylation can be achieved by any different methods such as physical adsorption, covalent grafting, copolymers^17,31–33 or by the simple EDC chemistry in this study. The resultant microbubbles showed good size homogeneity, biocompatibility and echogenicity. The uniform ultrasound contrast agent offers great potential application to medical application.

2. Experimental

2.1. Materials and instrumentation

Methyltrimethoxysilane (MTMS, Sigma-Aldrich), ammonium hydroxide (NH₄OH, 28%, Riedel-deHaen), (3-aminopropyl)trimethoxysilane (APTMS, Sigma-Aldrich), polyvinylpyrrolidone (PVP, M_w ∼ 360 000, Sigma-Aldrich), polyethylene glycol 1500 (PEG1500, Sigma-Aldrich), N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC, Sigma-Aldrich), N-hydroxysuccinimide (NHS, Sigma-Aldrich), hydrofluoric acid (HF, Riedel-deHaën), 1-octenylsuccinic anhydride (OSA, Polysciences), α-methoxy-ω-amino PEG (CH₃O–PEG–NH₂ 2k, Rapp Polymere), Dulbecco's modified Eagle's medium (DMEM, Sigma-Aldrich), festal bovine serum (FBS, Sigma-Aldrich), 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, Sigma-Aldrich) and dimethyl sulfoxide (DMSO, Sigma-Aldrich) were used as received without further purification. All the compounds used were reagent grade chemicals, unless stated otherwise. FTIR spectra were collected at room temperature (on a Jasco FTIR-4100 spectrometer) from samples prepared as pellets with KBr. Thermal gravimetric analyses (TGA) was carried out with a TGA-Q500 (TA Instrument) at a heating rate of 20 °C min⁻¹ under air. Structural characterization was carried out with transmission electron microscopy (TEM on a JEOL 200FX, acceleration voltage: 120 kV) and scanning electron microscopy (SEM, JSM 7600F, acceleration voltage: 10.0 kV).

2.2. Synthesis of amino-modified silica (AMS) particles³⁴

The AMS particles were synthesized via a conventional base-catalyzed sol–gel reaction.³⁴ Briefly, the aqueous solution containing 0.2 g of PEG1500, 0.02 g of PVP and de-ionized water (10 mL) was firstly prepared on the ice bath. The organo-methoxysilane mixture of MTMS (3.61 g) and APTMS (0.56 g) were slowly added into the aqueous medium. An small amount of ammonium hydroxide solution (0.1 mL, 28%) was then added to the aqueous medium under vigorous stirring and the reaction mixture was kept stirring for 1 h to yield uniform silica spheres, as shown in Scheme 1. Finally, the reaction continued for 6 h at room temperature and the reaction mixture was washed several times with ethanol and separated from the aqueous mixture by centrifugal sedimentation to remove the surfactants. Samples were lyophilized using a freeze dryer (FD 5030/8530, Panchum Scientific Corp., Taipei, Taiwan).

Scheme 1 Synthesis of AMS, HOPs, HPS particles.

2.3. Preparation of hollow PEG spheres (HPS)

The silica–OSA core–shell particles were synthesized as follows. First, 0.1 g of AMS particles was dispersed in DMSO with ultrasonic agitation at room temperature for 1 h. Then 0.5 g of OSA was added to the mixed solution and the mixture was stirred to react continuously at the temperature of 80 °C. After stirring for 24 h, the excess OSA were removed by centrifugation (4000 rpm, 10 min) and washed with de-ionized water for three times. The purified OSA-coated silica particles were then freeze-dried for 48 h.

Hollow OSA particles (HOPs) were fabricated by etching the as-synthesized OSA-coated silica particles with 1% hydrofluoric acid. Approximately 0.05 g of OSA-coated silica particles were mixed with 2 mL of 1% HF in a closed vial and then left to react for 24 h. After etching was completed, the resulting HOPs were washed with ultrapure water by centrifugation. Repetitive centrifugation and re-dispersion were performed at least three times to remove possible side reactants as well as remaining HF. The HOPs were then dialyzed against ultrapure water using a membrane (molecular weight cutoff (MWCO): 1000 Da) for 48 hours, followed by freeze-drying.

After drying, the HOPs can be further functionalized with polyethylene glycol (PEG) to get hollow PEG spheres (HPS). PEG was grafted onto the surface using EDC/NHS cross-linkers in phosphate buffer (PBS) for 24 hours on an agitator at room temperature.³⁵ For 0.05 g of HOPs in 5 mL of phosphate buffer (PBS) solution, 0.01 g of EDC, 0.05 g of NHS ester, and 0.005 g of mPEG–NH₂ (2000 Da) were used. The samples was then transferred to deionized water by ultrafiltration using 3 K MWCO filters (Amicon Ultra-15) and lyophilized to obtain a white solid. Eventually, hollow PEG spheres (HPS) were obtained.

2.4. Fabrication of microbubbles (MBs)

The HPS particles were filled with octafluoropropane (C₃F₈) in order to test whether the particles could effectively carry perflurocarbon (PFC) gas for a significant period of time. C₃F₈ was used because of its insolubility in water. In order to make C₃F₈-filled microbubbles,^36,37 DMSO (1.0% (v/v)) was firstly added to microcapsules as cryoprotectant, then the hollow microbubbles were frozen in a −80 °C freezer and then lyophilized to become fully dried form.³⁸ After lyophilization, another vacuum process was performed for evacuating the hollow microbubbles in approximately 10 min and refilled the C₃F₈ gas under a pressure of 1 atm, which allowed the inert gas diffusing into the hollow microbubbles within 2 h. Finally, the degassed water was quickly added to trap the C₃F₈ gas inside the microbubbles.^14,37

2.5. In vitro ultrasound imaging

In vitro ultrasound imaging was performed in the static state. An optically transparent phantom gel plate was made by the 2% agarose gel using a mold of Eppendorf tube (1.0 mL). US imaging was achieved using a 7.5 MHz probe (LA75 type, OPUS 5000, Chang Gung Medical Supplies & Equipment Corp).

2.6. In vitro cell cytotoxicity assay

The cellular biocompatibilities of MBs were evaluated by MTT assays. BNL CL2 cells were seeded in a 96-well plate at a density of 5000 cells per well in 200 μL of medium (DMEM) containing 10% FBS and incubated at 37 °C under conditions of 5% CO₂. Next, the culture medium was replaced by 200 μL of fresh DMEM supplemented with 10% fetal bovine serum (FBS) containing variable concentrations of MBs allowed to grow for another 6 and 24 h, respectively. After incubation, the medium was removed and the cells were washed with PBS. At least five parallel samples were performed in each group. Cells not treated with MBs were taken as the control. Then, 200 μL of 5.0 mM MTT solution was added to each well of the 96-well plate, followed by additional 4 h incubation at 37 °C. The culture medium with MTT was removed and 200 μL of DMSO was added. The absorbance of each well was measured at 570 nm using an enzyme linked immunosorbent assay (ELISA) reader with pure DMSO as a blank. A non-treated cell was used as a control and the relative cell viability (mean% ± SD, n = 5) was expressed as Abs_sample/Abs_control × 100%.

3. Results and discussion

3.1. Morphology and characterization of AMS particles, HOPs, and HPS

In this studying, microbubbles (MBs) with uniform sizes can be readily obtained by using silica microspheres as the core template. A schematic representation of the formation process of MBs is illustrated in Scheme 1. First, the ∼1.0 μm silica particles were synthesized through base-catalyzed sol–gel reaction, followed by heating the mixture with 1-octenylsuccinic anhydride (OSA) in DMSO at 80 °C to furnish the amphiphilic shelling. The silica cores of OSA-functionalized silica particles were removed to become the hollow spheres (HOPs) and the carboxylate surface were further PEGylated through simple EDC/NHS chemistry. HPS was prepared by reaction between HOPs and mPEG–NH₂. We found that the final products (HPS) showed excellent monodispersed properties observing under dark-field microscopy. To further verify the successful synthesis of the results, FTIR spectra of AMS, HOPs and HPS were obtained in Fig. 1.

Fig. 1 FTIR absorption spectra of (a) AMS, (b) HOPs and (c) HPS.

In the FTIR spectra (Fig. 1, curve a) the presence of the characteristic bands of AMS at 785 cm⁻¹ and 2972 cm⁻¹ is attributed to the stretching vibrations of Si–C and C–H, which was characteristic of symmetric deformations in Si–R groups. The spectrum of HOP (curve b) showed the disappearance of the band at 1535 cm⁻¹, while the band at 1750 cm⁻¹ became more pronounced. The characteristic band at 1750 cm⁻¹ was attributed to C=O stretching vibration from the anhydride group (curve b). The peak of Si–O–Si linkages couldn't be found in Fig. 1 curve b, which indicated that the template molecules were removed by the HF solution. The spectra of HPS (curve c) showed a strong band at 3450 cm⁻¹. This broad band is attributed to hydrogen bonded N–H and OH vibrations. The strong peak at 1650 cm⁻¹ has been assigned to C=O stretching associated with the presence of a tertiary amide. The appearance of the OH band near 3360–3560 cm⁻¹ is attributed to the hydrogen bonding. It is observed that PEG encapsulated microcapsules exhibited CH₂ stretching vibrations near 2800 cm⁻¹, which is attributed to the presence of PEG.³⁹

Fig. 2 illustrated SEM images of the synthesized AMS, silica–OSA particles, HOPs and HPS. The AMS surface was smooth surface (Fig. 2a), and the silica–OSA particles surface was covered with many protrusions that probably consisted of 1-octenylsuccinic anhydride (Fig. 2b). After immersing these particles properly in HF solution, silica core was removed immediately. The thin-shelling structure of HOPs was observed according to some broken spheres in the SEM image, as marked by the red arrows in Fig. 2c. The surface of HOPs was coated with mPEG–NH₂ (PEG) to become HPS, where the even roughly surface can attribute to the successful PEGylation. The additional component of 5% PEG shell was also found from the TGA results (Fig. S1†).

Fig. 2 SEM images for particles of (a) AMS, (b) silica–OSA particles, (c) HOPs and (d) HPS.

From the TEM images (Fig. 3a), HPS showed an uniform structure of hollow spheres, indicating that the silica core was successfully dissolved and extracted from the core via permeation through the semi-pervious walls composed of OSA conjugates. Simultaneously, the HPS surface still remained intact after covering with PEG, revealing a good structural stability. Furthermore, the HPS showed a narrow particle size distribution with an average diameter of about 1100 nm. Each HPS composed of 80 nm shell thickness as shown as (Fig. 3b). Moreover, the clear hollow structure of HPS MBs can be easily observed by the dark-field microscopy (CytoViva) (Fig. 3c) with narrow size distribution (Fig. 3d).

Fig. 3 (a and b) TEM images of HPS; (c) dark-field microscopy image of MBs; (d) the particle size distribution of MBs.

TGA was performed to quantitatively determine the composition of AMS, silica–OSA particles and HPS materials, measured under atmospheric air, as shown in Fig. 4. Compared to curve a, it was deduced that silica–OSA particles (curve b) revealed a slightly high organic mass release of about 10% between 400 °C and 600 °C. This weight loss could be due to the weight of 1-octenylsuccinic anhydride. HPS (curve c) displayed a distinct mass-loss profile above 200 °C compared to that obtained in silica–OSA particles (curve b), the 67 wt% loss of HPS (curve c) could be attributed to the SiO₂ be removed. In addition, we also can observe that there is about 6.5 wt% char yield of silica remaining in curve c, in agreement with the SEM-EDX result (Fig. S2†). Here a trace amount of silica can maintain MBs structure stability. However, the TGA result still indicated that HF could effectively etch silica.

Fig. 4 TGA curves of (a) AMS, (b) silica–OSA particles and (c) HPS.

3.2. In vitro ultrasound image

After the microbubbles are prepared, we carry out the ultrasound images two weeks later (Fig. 5 & 6). The ultrasound images of the ultrapure water and different concentration MBs were shown in Fig. 5. Compared with of the de-ionized water (Fig. 5, left), MBs containing groups showed distinctly echogenic signals which depended on the concentrations. Imaging the MBs with concentration from 1.25 to 10 mg mL⁻¹ resulted in maximum signals in 5.0 mg mL⁻¹. The decreasing echogenic signal upon reducing the MBs concentration attributed to the numbers of echogenic MBs source. When the inclusion concentration exceeded 5.0 mg mL⁻¹, the echogenic signals reached to the saturated state under current condition. The stability of microbubbles by different ultrasound exposure time was also investigated (Fig. 6). The echogenic signals gradually reduced upon the continuous ultrasound image capturing. After 20 minutes exposure, over 90% vanished signals may attribute to the leakage of the C₃F₈ gases upon sonication. Fig. 6 presents increasing the time of ultrasound-induced destruction significantly decreased the contrast enhancement, indicating the occurrence of microbubbles destruction. Contrast intensity was then significantly reduced leakage of the C₃F₈ gases upon ultrasound-induced destruction of the microbubbles through prolonged exposure to ultrasound imaging (20 min). This demonstrates the utility of HPS MBs as a US contrast agent. In addition, we also test about the shelf life of these MBs. Fig. 7 illustrates the ultrasound images after two weeks and three months later, respectively. As shown as Fig. 7, the echogenic signal is not much difference. It indicates that the MBs maintain stable at least three months.

Fig. 5 US signals at various concentrations of MBs. 1 through 5: MBs at concentrations of 1.25, 2.5, 5.0, 7.5 and 10 mg mL⁻¹, respectively.

Fig. 6 Stability of MBs under ultrasound exposures. US imaging of MBs “before” and “after” ultrasound-induced MBs destruction. The decrease in contrast enhancement indicates the destruction of MBs. All contrast-enhanced ultrasound imaging signals were significantly reduced after destruction of MBs induced by prolonged exposure of ultrasound imaging (20 min).

Fig. 7 Comparing the US signals of MBs upon long-term storage: 2 weeks versus 3 months.

3.3. Cytotoxicity

Biocompatibility is a key issue for applying ultrasound contrast agent in future biomedical fields. For cell viability study, the biocompatibility of MBs was investigated by MTT assay. Fig. 8 showed that BNL CL2 cells co-cultured with varies concentrations of MBs for 1, 6 and 24 h. As expected, no acute cell death was found upon addition of the MBs concentration range of 0–5.0 mg mL⁻¹. The cells remained more than 95% of their viability even when the concentrations of MBs were up to 5.0 mg mL⁻¹ (Fig. 8), indicating the good biocompatibility of the resulting MBs and ensuring their use as novel ultrasound contrast agent. These also indicate that the as-prepared MBs can be considered to be safe for in vitro and in vivo applications.

Fig. 8 Cell viability of BNL CL2 cells in the presence of different concentrations of MBs.

4. Conclusions

In summary, this research presents novel microbubbles synthesized by using SiO₂ particles as a template. Several advantages are discussed and evaluated in this study. Firstly, MBs with narrow size distribution can only be controlled by the core template of silica particles. Secondly, the OSA molecules can form an amphiphilic shelling which offer the intra-bubble gas stability and whole bubbles integrity. The property is similar to lipid-based microbubbles.¹⁵ At the same time, the MBs also possess excellent biocompatibility from the cell viability test. Furthermore, properly selection of PEG molecules on MBs can be easily incorporated to versatile antibodies, nucleic acid, or functional particles for multi-modal biomedical image application. Finally, this general method of synthesizing uniform MBs might have many potential applications in medical diagnostics.

Acknowledgements

The authors gratefully acknowledge the financial support from the Ministry of Science and Technology (MOST, Taiwan) under grant number MOST 103-2221-E-033-014-MY3, 103-2622-E-033-011-CC3 and 102-3011-P-033-004. The authors would like to thank the Center for Biomedical Technology at Chung Yuan Christian University for supporting this research.

References

1. J.-M. Correas, L. Bridal, A. Lesavre, A. Méjean, M. Claudon and O. Hélénon, Eur. J. Radiol., 2001, 11, 1316–1328.
2. A. Raisinghani and A. N. DeMaria, Am. J. Cardiol., 2002, 90, 3–7.
3. E. Quaia, Eur. J. Radiol., 2007, 17, 1995–2008.
4. D. Cosgrove, Eur. J. Radiol., 2006, 60, 324–330.
5. R. Gramiak and P. M. Shah, Invest. Radiol., 1968, 3, 356–366.
6. F. Kremkau, R. Gramiak, E. Carstensen, P. Shah and D. Kramer, Am. J. Roentgenol., 1970, 110, 177–183.
7. K. J. P. Jonathan Ophir, Ultrasound Med. Biol., 1989, 15, 15.
8. E. Stride and M. Edirisinghe, Soft Matter, 2008, 4, 2350–2359.
9. K. H. Min, H. S. Min, H. J. Lee, D. J. Park, J. Y. Yhee, K. Kim, I. C. Kwon, S. Y. Jeong, O. F. Silvestre, X. Chen, Y.-S. Hwang, E.-C. Kim and S. C. Lee, ACS Nano, 2015, 9, 134–145.
10. C. F. C. Shengping Qin and K. W. Ferrara, Phys. Med. Biol., 2009, 54, 31.
11. E. G. Schutt, D. H. Klein, R. M. Mattrey and J. G. Riess, Angew. Chem., Int. Ed., 2003, 42, 3218–3235.
12. J. G. Fabian Kiessling and M. Palmowski, Theranostics, 2011, 1, 8.
13. J. G. Riess, Curr. Opin. Colloid Interface Sci., 2003, 8, 8.
14. D. M. El-Sherif and M. A. Wheatley, J. Biomed. Mater. Res., Part A, 2003, 66, 347–355.
15. S.-T. Kang, J.-L. Lin, C.-H. Wang, Y.-C. Chang and C.-K. Yeh, J. Mater. Chem. B, 2015, 3, 5938–5941.
16. E. Kang, H. S. Min, J. Lee, M. H. Han, H. J. Ahn, I.-C. Yoon, K. Choi, K. Kim, K. Park and I. C. Kwon, Angew. Chem., Int. Ed., 2010, 49, 524–528.
17. R. Díaz-López, N. Tsapis, M. Santin, S. L. Bridal, V. Nicolas, D. Jaillard, D. Libong, P. Chaminade, V. Marsaud, C. Vauthier and E. Fattal, Biomaterials, 2010, 31, 1723–1731.
18. K. Pancholi, E. Stride and M. Edirisinghe, J. Drug Targeting, 2008, 16, 494–501.
19. D. E. Goertz, M. E. Frijlink, M. M. Voormolen, N. De Jong and A. F. W. van der Steen, Ultrasonics, 2006, 44, E131–E134.
20. K. Ferrara, R. Pollard and M. Borden, in Annual Review of Biomedical Engineering, Annual Reviews, Palo Alto, 2007, pp. 415–447.
21. J. I. Park, E. Tumarkin and E. Kumacheva, Macromol. Rapid Commun., 2010, 31, 222–227.
22. H. A. Z. De-Cai Bao, X.-D. Liu, Y.-J. Zhao, X.-J. Ma and Q. Yuan, J. Dispersion Sci. Technol., 2007, 28, 6.
23. U. Farook, H. B. Zhang, M. J. Edirisinghe, E. Stride and N. Saffari, Medical Engineering & Physics, 2007, 29, 749–754.
24. U. Farook, E. Stride, M. J. Edirisinghe and R. Moaleji, Med. Biol. Eng. Comput., 2007, 45, 781–789.
25. J. Liu, F. Liu, K. Gao, J. Wu and D. Xue, J. Mater. Chem., 2009, 19, 6073–6084.
26. X. W. Lou, L. A. Archer and Z. Yang, Adv. Mater., 2008, 20, 3987–4019.
27. P.-L. Lin, R. J. Eckersley and E. A. H. Hall, Adv. Mater., 2009, 21, 3949–3952.
28. V. C. F. Mosqueira, P. Legrand, A. Gulik, O. Bourdon, R. Gref, D. Labarre and G. Barratt, Biomaterials, 2001, 22, 2967–2979.
29. R. Gref, Y. Minamitake, M. T. Peracchia, V. Trubetskoy, V. Torchilin and R. Langer, Science, 1994, 263, 1600–1603.
30. D. E. Owens Iii and N. A. Peppas, Int. J. Pharm., 2006, 307, 93–102.
31. G.-b. Hong, J.-x. Zhou and R.-x. Yuan, Int. J. Nanomed., 2012, 7, 2863–2872.
32. Y. Yao, M. Zhang, T. Liu, J. Zhou, Y. Gao, Z. Wen, J. Guan, J. Zhu, Z. Lin and D. He, ACS Appl. Mater. Interfaces, 2015, 7, 18369–18378.
33. F. M. Veronese and G. Pasut, Drug Discovery Today, 2005, 10, 1451–1458.
34. C.-C. Hsu, K.-C. Chang, T.-C. Huang, L.-C. Yeh, W.-T. Yeh, W.-F. Ji, J.-M. Yeh and T.-Y. Tsai, Microporous Mesoporous Mater., 2014, 198, 15–21.
35. J. Yang, J. Lee, J. Kang, K. Lee, J.-S. Suh, H.-G. Yoon, Y.-M. Huh and S. Haam, Langmuir, 2008, 24, 3417–3421.
36. H. P. Martinez, Y. Kono, S. L. Blair, S. Sandoval, J. Wang-Rodriguez, R. F. Mattrey, A. C. Kummel and W. C. Trogler, MedChemComm, 2010, 1, 266–270.
37. H.-Y. Huang, S.-H. Hu, S.-Y. Hung, C.-S. Chiang, H.-L. Liu, T.-L. Chiu, H.-Y. Lai, Y.-Y. Chen and S.-Y. Chen, J. Controlled Release, 2013, 172, 118–127.
38. D. Pegg, in Cryopreservation and Freeze-Drying Protocols, ed. J. Day and G. Stacey, Humana Press, 2007, pp. 39–57.
39. P. B. Subhas Ghosh, Int. J. Org. Chem., 2012, 2, 366–370.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra09316g

---