Shell-by-shell synthesis of multi-shelled mesoporous silica nanospheres for optical imaging and drug delivery

Shell-by-shell synthesis of multi-shelled mesoporous silica nanospheres for optical imaging and drug delivery

Biomaterials 32 (2011) 556e564 Contents lists available at ScienceDirect Biomaterials journal homepage: Shell-...

2MB Sizes 0 Downloads 60 Views

Biomaterials 32 (2011) 556e564

Contents lists available at ScienceDirect

Biomaterials journal homepage:

Shell-by-shell synthesis of multi-shelled mesoporous silica nanospheres for optical imaging and drug delivery Chih-Chia Huang, Wileen Huang, Chen-Sheng Yeh* Department of Chemistry, National Cheng Kung University, Tainan 701, Taiwan

a r t i c l e i n f o

a b s t r a c t

Article history: Received 11 August 2010 Accepted 31 August 2010 Available online 27 September 2010

Self-templated synthesis involving interior channel wall protection as well as outermost surface passivation was crucial to successful synthesis of multi-shelled mesoporous silica nanospheres. The shell-by-shell fabrication of double- and triple-walled mesoporous silica nanospheres downsized to w100 nm. The multishelled mesoporous silica can be built as rattle-type or hollow structures with w110 nm of double-shelled and w140 nm of triple-shelled sizes. Notably, the shell-to-shell distance can be tuned by controlling the etching period from the self-templation processes without changing the multi-shelled size or interior core diameter. The multi-shelled mesoporous nanostructures provide a platform for the development of a multifunctional vector by the inclusion of functional species into shell-to-shell cavities and porous shells. The encapsulation of the fluorophore and drug in shell-to-shell space and mesoporous shells showed that multi-shelled silica spheres can be used in dual-modality for imaging and drug co-delivery vectors through the appropriate selection of pH-dependent molecules. The in vitro evaluation in triple-shelled silica indicated that an anti-cancer doxorubicin (DOX), loaded in the outer periphery space, was successfully carried and released in the cytoplasm, then entered nuclei while fluorescein FITC (primarily distributed in inner periphery space) was effectively encapsulated inside the spheres. The double- and triple-shelled nanospheres consistently provided imaging probes with visible tracking capability in vitro and in vivo. Crown Copyright Ó 2010 Published by Elsevier Ltd. All rights reserved.

Keywords: Multi-shelled nanoparticles Silica Self-templation Optical imaging Drug delivery

1. Introduction Multi-shelled spheres formed as rattle-type or hollow structures are unique hierarchical structures comprise of multiple concentric shells with different diameters [1e3]. With the void spacing between the shells ideally suited to load distinct functional species, the fantastic architectures stimulate technological importance because of their potential applications in controlled delivery, confined nanoreactors, and catalysis. Template synthesis has been an effective means of manufacturing such shell-in-shell architectures [4e10]. The choice of template composition and size, and the corresponding preparation strategy strongly influence the final designed multishelled materials. Because of its easy fabrication and derivatization, silica composition is often used as a sacrifice template, either treated as core support [5,6] or as an intermediate layer between the shells [7,8], to develop other composites in multi-shelled structures. For example, the w250 nm of silica acting as a sacrifice core was employed to generate double-shelled SnO2 [5] and coaxial [email protected]

* Corresponding author. E-mail address: [email protected] (C.-S. Yeh).

[6] hollow structures. With the same composition, the ellipsoidal double-shelled SnO2 created by the introduction of silica as removal layer was developed by Archer et al. [7]. The development of shell-byshell synthetic technology for multi-shelled silica with rattle-type and hollow nanospheres remains an attractive challenge. Shell-byshell construction of multi-shelled nanostructures itself is representative of the advancement of nanofabrication technology. An additional challenge to the manufacture of multi-shelled structures is achieving control of the space between the shells. For a certain particle size, tuning shell-to-shell distance is critical in varying the physical and chemical features of multi-shelled nanospheres and may provide better understanding of how to control the local chemical micro/nano environments. Self-templated synthesis involving surface-protected etching provides an effective approach for creating rattle-type or hollow structures from solegel derived colloids because of their porosity [11e13]. With appropriate protection of the outermost surface, a selected etchant is able to dissolve the unprotected interior while sustaining the protected outer layer. Although self-templated strategy presents an elegant route to create single-shelled rattle-type or hollow silica nanostructures, there has been difficult to form multishelled silica colloids. The carefully optimizing outermost layer by the

0142-9612/$ e see front matter Crown Copyright Ó 2010 Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2010.08.114

C.-C. Huang et al. / Biomaterials 32 (2011) 556e564

increase of thickness was necessary in order to sustain extensive etching for the formation of multi-shelled nanostructures. As shown by Yin et al., the resulting shell thickness was at least 100 nm for double-shelled nanoparticles [13]. Therefore, the creation of the complex building block like multi-shelled nanostructures with intact sizes near to 100 nm remains a challenge. In this study, the shell thickness could be down to 10e20 nm to yield double- and triplewalled silica nanospheres. Additionally, the presented strategy was developed to directly build up the mesoporous shell with the tunable pore sizes by the selection of different alkytrimethylammonium bromide (CnTAB) in chain lengths, which is different from Yin and coworkers methods by the introduction of base etching forming mesoporous silica layer. We found that interior channel wall protection with APTES modification as well as outermost surface with PVP passivation was crucial to successful synthesis of multi-shelled mesoporous silica. Briefly, the inner mesoporous silica shells were protected by APTES prior to fabricate outer mesoporous silica layers. Subsequently, the as-prepared outer silica layers were passivated with PVP and treated with base etchant to from shell-in-shell nanostructures. The multi-shelled mesoporous silica can be built as rattletype or hollow structures with w110 nm of double-shelled and w140 nm of triple-shelled sizes. Notably, the shell-to-shell distance can be easily tuned by controlling the etching period from the selftemplation processes without changing the multi-shelled size or interior core diameter. For the multi-shelled structures, previous studies primarily focused on the development of synthetic technologies. However, a unique aspect of these multi-shelled materials is their attractive interior architectures. It would be anticipated to encapsulate different functional molecules into respective shell-toshell space, showing that multi-shelled silica spheres can be used in dual-modality for imaging and drug delivery vectors.

2. Experimental section 2.1. Materials All reagents were of analytical purity and used without further purification: tetraethyl orthosilicate (TEOS, 98%, Acros), polyvinylpyrrolidone (PVP, Mw. 4,0000, SigmaeAldrich), styrene (stabilized 99%, Acros), 4-styrenesulfonic acid sodium salt hydrate divinylbenzene (SigmaeAldrich), Potassium peroxodisulfate (K2S2O8, 98%, Showa), sodium bicarbonate (NaHCO3, J. T. Baker), ethanol (EtOH, 99.9%, J.T. Baker), aminopropyleethoxysilane (APTES, 99%, Acros), sodium hydroxide (NaOH, Fullin), ammoniun hydroxide solution (NH3, 33%, SigmaeAldrich), tetradecyltrimethylammonium bromide (C14TAB, 99þ, Acros), hexadecyltrimethlammonium bromide (C16TAB, 99þ, Acros), octadecyltrimethylammonium bromide (C18TAB, 98%, Aldrich), doxorubicin hydrochloride (DOX$HCl, 98%), fluorescein 5(6)-isothiocyanate (FITC, approx 90%, Sigma), and indocyanine green (ICG, MARK).


2.2.3. Preparation of double-shelled silica nanospheres Preparation of double-shelled silica nanospheres: The double-shelled silica nanoparticles were synthesized by coating a mesoporous silica layer on the resultant [email protected] nanoparticles via a CTAB surfactant-assisted reaction followed by a selftemplating etching process. In a typical synthesis, 3 mL of the as-obtained [email protected] nanoparticles solution was mixed with 5.75 mL of pure water, including with CTAB (8.7 mM) and NaOH (2.17 mM). Then 100 mL TEOS was added as a droplet under a vigorous stirring and then the mixture solution was reacted at 55  C for 3 h. Consequently, [email protected] covered with mesoporous silica layer nanoparticles was generated. The supernatant was removed by centrifugation and the white precipitate was collected. After washing the white precipitate with water three times, the final product was re-dispersed in the 10 mL of PVP solution (1 mM) to modify the mesoporous silica with PVP polymer coverage. The mixture solution was heated at 100  C again for 3 h. After cooling to room temperature, 8 mL of the PVP-modified silica nanoparticles solution was taken and reacted with 1 mL of 0.25 M NaOH for 1 h, resulting in a final 27.8 mM NaOH. TEM images demonstrated that the resulting nanoparticle exhibited a double-shelled configuration. The resulting double-shelled silica nanospheres were subsequently dispersed in ethanol (10 mL), after which 10 mL of APTES was added under stirring for a 2 h reaction. After washing, the final product was stored in 8 mL deionized water before loading the drug or next silica layer coating. The shell-in-shell hollow silica nanoparticles could be formed by a calcination process at 580  C for 2 h. 2.2.4. Preparation of triple-shelled silica nanospheres Preparation of triple-shelled silica nanospheres: The synthesis procedure to prepare triple-shelled silica nanospheres was similar to the preparation of the 2nd mesoporous silica shell, except for the addition of 200 mL of TEOS in the synthesis of 3rd mesoporous silica layer. Once again, the triple-shelled hollow silica nanoparticles were formed after calcination at 580  C for 2 h.

2.3. Encapsulation of fluorescence and anti-cancer drugs in multi-shelled silica nanoparticles Molecules encapsulated into inner cavities of multi-shelled silica vehicles were prepared as follows. The FITC drugs were loaded into rattle-type double-shelled silica nanostructures by physically mixing the materials (8 mL) with an aqueous solution of FITC (2.6 mg/mL, 80 mL) to achieve saturation absorption at room temperature for 24 h. Then, the FITC-loaded double-shelled nanostructures were purified by centrifugation/washing processes with water more than three times. The amount of FITC-loaded in double-shelled nanostructures was calculated from the difference between the initial amount of FITC and the residue in the supernatants collected from centrifugation and subsequent washing processes. The fluorescent signal of FITC (excitation at 450 nm and emission at 515 nm) was recorded using a fluorescence spectrophotometer. We used FITC-loaded silica nanostructures (31.7 mg mg1) as a starting material and repeated the self-templeting process (as described above) to synthesize triple-shelled silica nanostructures. In the course of 3rd shell preparation, the loss of FITC was carefully quantified to estimate the residual amount of FITC-loaded within multi-shelled silica nanoparticles. Subsequently, the FITC-loaded triple-shelled silica nanostructures (8 mL) were physically mixed with a DOX solution (4 mg/mL, 80 mL) to achieve saturation adsorption at room temperature for 24 h. To determine the amount of DOX in the triple-shelled silica nanospheres, the supernatants of the residual drug molecules were collected and subjected to signal quantification. The fluorescent signals of DOX (excitation at 480 nm and emission at 585 nm) were recorded using a fluorescence spectrophotometer. In the triple-shelled silica nanospheres, the loading amount of FITC and DOX were 14.6 mg mg1 and 30.8 mg mg1, respectively.

2.2. Synthesis of multi-shelled silica nanoparticles 2.2.1. Synthesis of polystyrene nanoparticles as a template Synthesis of polystyrene nanoparticles as a template: Poly(styrene-co-styrene sulfonate) (denoted as PS) nanoparticles were prepared by following the previous methods with slight modification [14,15]. 0.025 g of NaHCO3 and 0.1 g of 4-styrenesulfonic acid sodium salt hydrate were pre-dissolved in 50 mL deionized water and heated at 75  C under vigorous stirring. 5 mL of styrene solution was subsequently added. After reacting for 1 h, the initiator K2S2O8 (0.025 g) was added to start polymerization under magnetic stirring for 18 h. 2.2.2. Preparation of single-shelled silica nanospheres Preparation of single-shelled silica nanospheres: To prepare single-shelled silica nanoparticles, we dispersed 0.2 mL of as-prepared PS solution (w12.4 mg in dry powder) into 10 mL of ethanol solution, including 10 mL of APTES. After vigorously stirring (2 h), 100 mL of TEOS and 250 mL of NH3 were then introduced into the above solution accompanying by stirring and reacted at room temperature. The solution became milky when it reacted overnight. The white color product of [email protected] nanoparticles was collected using centrifugation (9500 rpm) and rinsed several times with ethanol. The as-obtained product was [email protected] nanoparticles which were dispersed into 8 mL pure water before further use. We obtained single-shelled hollow silica nanoparticles after removing PS core with calcination at 580  C for 2 h.

2.4. In vitro release studies of multi-shelled silica nanostructures We used FITC-loaded rattle-type double-shelled silica nanostructures (31.7 mg mg1) as the starting materials to synthesize triple-shelled nanospheres (as described above). In the course of 3rd shell preparation, the loss of FITC was carefully quantified to estimate the residual amount of FITC within triple-shelled nanospheres. Subsequently, the as-prepared FITC-loaded silica nanospheres were resuspended in 25 mL PBS at pH ¼ 4.3 and 7.2. The released FITC was evaluated by fluorescence detection (excitation at 450 nm and emission at 515 nm). The drug release behavior of FITC-loaded rattle-type double-shelled silica nanostructures was carried out in 25 mL PBS at pH ¼ 4.3 and 7.2 as well. For DOX-loaded in rattle-type triple-shelled silica nanostructures, the doubleshelled silica nanospheres without inclusion of FITC as the starting materials to synthesize triple-shelled silica nanospheres. Once the fabrication of 3rd shell was completed, the triple-shelled nanostructures (8 mL) were mixed with a DOX solution (4 mg/mL, 80 mL) at room temperature for 24 h. Subsequently, the as-prepared DOX-loaded silica nanospheres were resuspended in 25 mL PBS at pH ¼ 4.3 and 7.2). The released DOX was estimated by the fluorescent signal of DOX (excitation at 480 nm and emission at 585 nm).


C.-C. Huang et al. / Biomaterials 32 (2011) 556e564

2.5. Cellular imaging by the dark-field illumination and laser confocal microscopes

2.7. Cytotoxicity analysis

A549 cells (1.2  104 cells per well) were seeded onto an eight-well chamber slices. After incubation overnight (37  C, 5% CO2), the medium was carefully aspirated and replaced with 300 mL of a new medium (DMEM, 10% FBS, 1% penicillinestreptomycin) containing 40 mg/mL of DOX/FITC co-loaded silica vehicles. The cells were incubated for 1 h, 3 h, and 24 h, at which point the medium was carefully removed. The wells were each rinsed twice with 300 mL of PBS. 150 mL of 4% paraformaldehyde was added, and the cells were fixed for 30 min. The nucleus was stained using DAPI (blue). Subsequently, sample slices were observed under the enhanced dark-field illumination system which was established using a Olympus microscope consisting of a CytoViva 150 dark-field condenser accompanied with a halide light source. Fluorescent imaging was analyzed and monitored with a specialized 61002 (DAPI/FITC/Texas red) emitter and exciter by fluorescence microscopy by the same Olympus microscope. To measure A549 cells treated with ICG-loaded silica vehicles, we used confocal laser scanning microscopy (Digital Eclipse C1si; Nippon Instruments Corporation, Tokyo, Japan) analysis to capture DAPI fluorescence with 405 nm CW-laser and ICG images with an 808 nm CW-laser.

For the MTT assay [16], an A549 (human alveolar basal epithelial cancer cells) cell line (4000 cells/well) was cultured in a 96-well microplate with 100 mL of Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum (FBS) at 37  C in 5% CO2/95% air for 1 day. Both rattle-typed and hollow triple-shelled silica nanospheres with various concentrations (0e40 mg/ml) were loaded separately into 96-well microplates one day after they had been cultured. The culture medium was then removed and replaced with 100 mL of the new culture medium containing 10% MTT reagent. The cells were then incubated for 4 h at 37  C to allow formazan dye to form. The culture medium in each well was then removed, and dimethyl sulfoxide (DMSO) (200 mL/well) was added for an additional 10 min of incubation. After the cells had been centrifuged, the resulting formazan in each well was transferred to an ELISA plate. The quantification determining cell viability was done using optical absorbance (540/650 nm) and an ELISA plate reader.

2.6. In vivo imaging measurements Male BALB/c mice (6e8 weeks old) were provided by the National Health Research Institutes (NHRI) (Tainan, Taiwan). All animals received humane care in compliance with the institution’s guidelines for the maintenance and use of laboratory animals in research. All of the experimental protocols involving live animals were reviewed and approved by the Animal Experimentation Committee of National Taiwan University. Whole body images of mice were acquired before and after an intraperitoneal injection of 25 mg/kg and 50 mg/kg. All images were analyzed and collected with a FITC filter with IVIS Imaging System.

2.8. Particles characterization Electron micrographs of the samples were taken using transmission electron microscopes (TEM) at 200 kV (JEOL 2010; Jeol USA, Inc., Peabody, MA, and CM-200; Philips Research Europe, Eindhoven, The Netherlands). A drop of the sample was placed on a Cu mesh coated with an amorphous carbon film and then the solvent was evaporated in a vacuum desiccator. Field emission scanning electron microscope (FESEM) images of the nanoshells on the Cu plate substrates were taken using an FE-SEM at 10 kV (XL-40 FEG; Philips). The crystalline structures were identified using an X-ray diffractometer (XRD-7000S; Shimadzu Corporation, Tokyo, Japan) with copperpotassium alpha (CuKa) radiation (l ¼ 1.54060 Å) at 30 kV and 30 mA. IR spectra were measured using a KBr plate in a Fourier transformation infrared (FTIR) spectrometer (200E; Jasco International Co., Ltd., Tokyo, Japan). Nitrogen (N2) adsorption

Fig. 1. a) Schematic illustration of the preparation of rattle-type and hollow multi-shelled silica nanoparticles. TEM images of b) single-, c) double-, and d) triple-shelled silica nanoparticles before (left) and after (right) removal of PS cores. SEM images of e) double- and f) triple-shelled silica nanoparticles.

C.-C. Huang et al. / Biomaterials 32 (2011) 556e564


Fig. 2. TEM images of a) [email protected]@silica with APTES modification of 2nd silica layer treated with NaOH for 4 h and b) [email protected]@[email protected]@silica with the 2nd silica shell passivated with PVP treated with NaOH for 1 h. measurements were taken at 77 K using an accelerated surface area and porosimetry analyzer (ASAP 2010; Micrometrics Instrument Corporation, Norcross, GA) with Brunauer-Emmett-Teller (BET) calculations for the surface area. A fluorescence spectrophotometer (F-2500; Hitachi Koki Co., Ltd., Tokyo, Japan) and a UVeVis spectrophotometer (8452A; HewlettePackard Company, Palo Alto, CA) was used to record photoluminescence (PL) characteristics. The emission of ICG-loaded silica vehicles was measured using a spectrophotometer (Jasco V-670, Japan), with wavelengths ranging from 740 to 900 nm at an excitation of 708 nm. To visualize the cells and nanoparticles, we used light scattering/fluorescence microscope Olympus IX70 with CytoViva Adapter (Aetos Technologies, Inc., Auburn, AL, USA).

3. Results and discussion 3.1. Preparation and characterization of multi-shelled silica nanospheres Fig. 1a shows the preparation processes of the multi-shelled mesoporous silica nanospheres. One of our research interests is to design multi-shelled structures with sizes near or below 100 nm for biomedical applications. Thus, a surfactant-free emulsion copolymerization of styrene and 4-styrenesulfonic acid was used to generate 60 nm poly(styrene-co-styrene sulfonate) (denoted as PS) nanospheres (Fig. S1a). FTIR analysis (Fig. S1b) confirmed vibration peaks including w3027 cm1, w2925 cm1, w1600 cm1, and w1470 cm1 corresponding to Ar-H stretching, CeH stretching, and C]C stretching in benzene ring, respectively, from the polystyrene [13e15]. The peaks around w3450 cm1, w1190 cm1, and w1030 cm1 could be assigned to the eOH symmetric stretching and asymmetric and symmetric SO 3 vibrations, respectively [15e17]. No reaction side product residues were observed, such as CO 3 appeared at w1539 cm1, 860 cm1, and 729 cm1 [18]. Because sulfonation yielded eSO 3 groups (49.6 mV of surface charge), deposition of silica layers onto PS nanospheres were completed by the addition of APTES, followed by TEOS for solegel condensation (Fig. 1b). Upon completing silica coating, FTIR spectra (Fig. S2a) of the [email protected] composites exhibited the typical bands of silica vibrations in the range of 400e1200 cm1 with [SiO4] tetrahedron at 1100 and 470 cm1, SieOH vibration at 945 cm1 and SieOeSi bending mode at 800 cm1 [19,20]. The failure of silica layers deposition occurred in the absence of APTES. In fact, the inclusion of APTES turned out to improve resistance against corrosion in the generation of shell-in-shell structures as well as to assist silica coating. The thicker silica shells of [email protected] nanospheres were obtained by simply increasing the amount of TEOS (see Fig. S3 Taking w15 nm of 1st shell thickness as an example, a mesoporous silica layer (w20 nm) was subsequently decorated on the top of 1st silica shell by the introduction of C16TAB

surfactant and TEOS through a hydrolysisecondensation reaction yielding [email protected]@silica. Subsequently, PVP polymer was capped on the outermost silica surface, followed by the addition of a NaOH etchant. Because of PVP surface protection against OH etching, the dissolution mostly occurred in the unprotected interior. The selftemplated synthesis successfully fabricated double-shelled silica nanospheres with the 2nd shells as a mesoporous structure with the creation of the O-ring like cavity as [email protected]@[email protected] (Fig. 1c). The void distance between the shells was determined as w10 nm. To be mentioned that the addition of CnTAB surfactants has allowed us to adjust the porosity of the mesoporous layer by the change of CnTAB in chain lengths. For example, the different alkyl chain lengths of CnTAB (n ¼ 14, 16 and 18) were examined to build up 2nd silica layers ([email protected]@silica), followed by PVP passivation and base etching procedure. Fig. S4a and b show the TEM images of [email protected]@[email protected] nanospheres from C14TAB and C18TAB, respectively. Subsequently, the [email protected]@[email protected] particles were subject to extraction of CnTAB surfactants by APTES/ethanol mixture to expose mesoporous pores [21e23]. The N2-sorption isotherm analysis was conducted and BJH analysis of pore size distributions show the pore population mainly located at 1.7e3 nm for C14TAB and C16TAB surfactants, and appeared in the range of 3e8 nm for C18TAB (Fig. S5). The increase of the alkyl chain lengths of the CnTAB surfactants enlarged pore diameters. This can be explained in term of the increase of micelles sizes in the course of formation of mesoporous structures as the alkyl chain lengths prolonged [24]. In our attempt, we found that the addition of APTES could improve resistance against eOH etching. For example, the presence of APTES (originally introduced in 1st shell) was able to sustain w15 nm of the 1st silica shell in the course of the 2nd shell fabrication. Accordingly, the 2nd shell was modified with additional APTES prior to depositing the 3rd mesoporous silica layer. Once the 3rd mesoporous silica layers (w21 nm) were complete ([email protected]@[email protected]@silica), PVP passivation with the subsequent etching process was then performed to form a 3rd nanoshells ([email protected]@[email protected]@[email protected]). The tripleshelled silica nanospheres were manufactured with a distance of 11 nm between the 2nd and 3rd shells, and 3rd shell thickness of 10 nm (Fig. 1d). APTES plays a crucial role to against etchant and facilitates shell-in-shell fabrication. We have employed CTAB surfactants to build up mesoporous shells. Thereby, the inner shells were protected by APTES prior to fabricate outer silica mesoporous layers. Subsequently, the as-prepared outer silica layers were passivated with PVP and treated with base etchant to form shell-in-shell nanostructures. Previous studies of the modification of mesoporous silica have shown that surface silyl modification using APTES achieved the removal of


C.-C. Huang et al. / Biomaterials 32 (2011) 556e564

Fig. 3. TEM images of rattle-type double-shelled nanospheres etched by NaOH under a) 30 min, b) 1 h, and 2 h d) The enlarged images for shell-to-shell distance under different etching period.

surfactants, e.g. CTAB. That is, APTES molecules were able to access the interior surface of the pore channels and reacted with silanol groups of mesoporous silica walls, providing surface coverage for the channel walls [21e23]. Additional experiments were carried out to support the function of APTES in creation of shell-in-shell mesoporous nanostructures. Following the same preparation scenario, the [email protected] (1st shell) was synthesized and formed a mesoporous silica layer on the top of 1st silica shell ([email protected]@silica). Subsequently, we have replaced PVP using APTES to modify the 2nd silica layer. The resulting particles were incubated with NaOH for 4 h. TEM image (Fig. 2a) shows that the 2nd mesoporous silica layers treated with APTES were able to resist eOH etching and remained intact. To provide additional evidence, the [email protected]@[email protected]@silica particles were prepared as well. The 2nd silica shell was first passivated with PVP instead of APTES modification, followed by deposition of the 3rd mesoporous silica layer, and then the process of PVP passivation on the outermost surface. Once again, the [email protected]@[email protected]@silica particles were treated with NaOH. Fig. 2b indicates that the original 2nd shells were completely removed and the

[email protected]@[email protected]@silica was converted to [email protected]@[email protected] yielding a yolk-shell configuration. Without APTES modification, the 2nd silica shells were unable to against base etching. The creating void space was w31 nm which is consistent with the distance between 1st and 3rd silica shells of [email protected]@[email protected]@[email protected] (tripleshelled nanospheres). SEM images display the double- and triple-shelled morphology from the broken nanospheres (Fig. 1e and f). Importantly, the raised resistance resulting against corrosion from APTES provided an advantage that allowed us to tune the shell-to-shell distance. Taking rattle-type double-shelled nanospheres as an example, control over the thickness of the 2nd shell can be achieved simply by changing the etching period while the 1st shell’s thickness is maintained. Etching times of 30 min, 1 h, and 2 h produced respective shell thickness of 11.8, 10, and 7.2 nm, corresponding to 8.4, 10.2, and 13 nm in shell-to-shell distance (Fig. 3). The rattle-type nanospheres can be transformed to hollow multishelled structures through a calcination process at 580  C for 2 h to

Fig. 4. Configurations of rattle-type and hollow multi-shelled silica nanospheres using w15 nm of 1st shell thickness as an example.

C.-C. Huang et al. / Biomaterials 32 (2011) 556e564




Volume adsorbed(cm3/g, STP)

single-shelled silica nanospheres double-shelled silica nanospheres triple-shelled silica nanospheres









1000 single-shelled silica nanoparticles 155.2 m2/g


double-shelled silica nanoparticles 243.9 m2/g


triple-shelled silica nanoparticles 329.2 m2/g

400 200 0 0.0


0.4 0.6 P/P0



Fig. 5. XRD patterns and BET analysis of single-, double-, and triple-shelled hollow silica nanospheres.

remove the PS cores (Fig. 1aed). After calcination, the multi-shelled silica nanospheres exhibited a slight shrinkage. The detailed configurations of rattle-type and hollow silica can be seen in Fig. 4 (with more TEM images provided in Figs. S6 and S7). The resulting multishelled silica nanospheres were amorphous structures as determined by XRD measurements (Fig. 5a). Energy dispersive spectroscopy (EDS) indicated the presence of Si and O elements (Fig. S8). Additionally, FTIR analysis (Fig. S2b) of as-prepared hollow triple-shelled silica structures provided the relevant silica vibration modes in the range of 400e1200 cm1. To be noted that the following N2-sorption isotherm measurements giving surface area and pore size values

were obtained from hollow multi-shelled silica structures after calcination. BET plots (Fig. 5b) showed type-IV mesoporosity of hollow multi-shelled structures and the surface areas were determined as 155.2 m2/g (single-shelled), 243.9 m2/g (double-shelled) and 329.2 m2/g (triple-shelled). The hysteresis loops occurred in the range of 0.6e1.0 P/P0, indicating an H2-type of hysteresis loops. The presence of H2-type hysteresis represents the inhomogeneous distribution of pore sizes, originated from an interconnected network of pores. BJH analysis showed the mesoporosity of the 1st, 2nd, and 3rd shells (Fig. S9). The single-shelled hollow silica nanoparticles displays pore sizes mainly below 8 nm. The pore size distributions

Fig. 6. a) Fluorescent measurement of ICG-loaded double-shelled silica (rattle-type) and time-course confocal images of A549 cells treated with ICG-loaded double-shelled silica at 1 and 3 h. The blue color represents DAPI stained nuclei. The inset shows the single cell images. b) FITC-loaded double-shelled silica (rattle-type) nanospheres emitting green fluorescence (25 mg/kg and 50 mg/kg) were administrated by intraperitoneal injection and fluorescence signals were observed for a period of 1 h.


C.-C. Huang et al. / Biomaterials 32 (2011) 556e564

Fig. 7. Fluorescence and dark-field images of A549 cells incubated with DOX/FITC-loaded rattle-type triple-shelled silica as a function of time. The blue color represents DAPI stained nuclei and N indicates the nucleus.

peak at w11 nm accompanied with smaller pores less than 8 nm for hollow double- and triple-shelled silica structures. The former ones correspond to the shell-in-shell space and the latter pores would be attributed to the porosity of the silica shells. 3.2. In vitro cytotoxicity and in vitro and in vivo imaging Prior to study their potential biomedical applications, MTT assays were conducted to evaluate the cytotoxicity. The A549 cells were treated with rattle-type and hollow triplet-shelled silica structures showing no apparent toxicity with cells survival greater than 94% in a dosage range of 0e1000 mg/mL (Fig. S10) after 1 day incubation. The multi-shelled mesoporous nanostructures provide a platform for the development of a imaging vector by the inclusion of functional species into shell-to-shell cavities and porous shells. To evaluate the carriage of the periphery space, the rattle-type double-shelled nanostructures, without the removal of the core templates, were examined by the encapsulation of ICG molecules, a near-infrared fluorophore in biological window [25,26]. The obtained multi-shelled nanospheres were all modified with APTES to extract CTAB surfactants from the interior channel walls prior to the encapsulation of molecules. It should be mentioned that the entrapped fluorophores were expected to distribute in the space between the shells as well as in the porous shell because of the mesoporosity of the 2nd shell. Fig. 6a displays the ICG-loaded double-shelled silica emitting a 806 nm wavelength, showing no shift relative to free ICG in emission peak. The ICG-loaded silica was treated with A549 cancer cells, and then inspected by a confocal microscope. The double-shelled silica nanospheres were

loaded with 23.6 mg ICG per gram of nanospheres. The confocal microscope revealed the time-dependent images of the uptake of silica nanospheres (Fig. 6a). The fluorescence of ICG (shown in red) can be clearly seen in the cytoplasm. The red displays mostly appeared as discrete dots indicating ICG mainly resided inside the double-shelled silica. The similar results showing dot-shaped fluorescent imaging, which is distinct from the leakage free fluorophores spread over the cytoplasm, have also been observed from other particle-based carriers [27e30]. More ICG-loaded nanospheres internalized into the cells with increased incubation time. To further test the capability of double-shelled silica to carry fluorophores, we conducted preliminary in vivo examinations. Due to facility limitation, a green emitting fluorescein FITC dye was encapsulated in silica (Fig. 6b). Additional evidence of FITC-loaded in silica was provided by the BET analysis, where the surface area of the double-shelled nanospheres decreased from 204 m2/g to 140 m2/g after FITC was added to the silica (Fig. S11). The nude mice received FITC-loaded silica with two dosages of 25 mg/ kg and 50 mg/kg via intraperitoneal injection. Fig. 6b shows the luminescence images collected with an emission filter (485e505 nm). The higher dosage displayed stronger intensity and a broader fluorescence area from the injection sites. Both dosages exhibited steady illumination during a 1 h observation period. 3.3. Drug delivery evaluated by dark-field microscope analysis and in vitro drug release profiles With this evidence of the double-shelled nanospheres capacity for carrying and imaging, we next inspected the triple-shelled silica to

C.-C. Huang et al. / Biomaterials 32 (2011) 556e564

Molecule released / %




80 pH = 7.2 pH = 4.3

60 40 DOX 20

pH = 7.2 pH = 4.3

0 0

FITC released / %




300 t/h






pH = 7.2 pH = 4.3

0 0

50 t/h


Fig. 8. In vitro release profiles of a) FITC and DOX molecules individually encapsulated in rattle-type triple-shelled silica nanostructures in PBS solution (pH ¼ 4.3 and 7.2) at 37  C and b) FITC encapsulated in rattle-type double-shelled silica nanostructures in PBS solution (pH ¼ 4.3 and 7.2) at 37  C.

co-deliver FITC fluorophore and an anti-cancer doxorubicin (DOX) drug as bifunctional imaging and chemotherapeutic agents. Once again, rattle-type nanospheres were used in studies. In this case, the double-shelled nanospheres were fabricated first to entrap FITC dye. Following the same preparation to build a 3rd silica shell, DOX was then loaded into the triple-shelled nanospheres containing FITC. The loading amount of FITC and DOX were 14.6 mg mg1 and 30.8 mg mg1, respectively, in triple-shelled nanospheres. The DOX/FITC-loaded triple-shelled silica exhibited fluorescence covering from 480 nm to 640 nm (see Fig. S12). Both FITC and DOX were then co-delivered to A549 cancer cells. A transmission-mode dark-field microscope collected light scattering signals combined with fluorescent modality to visualize the A549 cells treated with triple-shelled silica (Fig. 7). The internalization of silica in the cytoplasms was observed in the early stage of 1 h. The green fluorescence of FITC and red emission of DOX displayed as discrete spots indicating both molecules remained inside the silica shell-in-shell structures. Although the release of FITC cannot be inhibited, most of the green fluorescence manifested as discrete spots when the incubation time was prolonged to 3 h. In terms of release, significant DOX, on the other hand, escaped from the silica and colored the cytoplasm in red. At 24 h, the red DOX had entered nuclei while FITC mainly remained as green spots in the cytoplasms. Time-course images clearly showed that the anti-cancer DOX drug was carried to and released in the cytoplasms, followed by nuclei penetration. Multi-shelled silica nanospheres seem to simultaneously achieve imaging and drug delivery capabilities through the appropriate choice of pH-dependent molecules. According to previous studies [31], the w110 nm of mesoporous silica


nanoparticles crossed the cell membranes through clathrin-dependent endocytosis and were transported to acidic endosomes/lysosomes (pH 4.5). FITC is a proteolytic acid with three pKa values of 2.2, 4.4, and 6.7 [32]. It is possible that FITC with poorer solubility (caused by increased protonation of COO groups at lower pH environment) limited their release into the cytoplasm. Meanwhile the increased protonated eNH2 groups on DOX (pKa: 8.6) [33] at acidic endosomes/ lysosomes improved aqueous solubility and resulted in significant release of DOX. We have conducted parallel experiments to monitor timedependent release profiles for multi-shelled nanospheres in PBS solution at pH ¼ 4.5 and 7.2 (37  C). Fig. 8a demonstrated that rattletype triple-shelled silica particles carried FITC and DOX individually. Following the aforementioned preparation processes, the doubleshelled nanospheres were fabricated first to entrap FITC dye, then building 3rd silica shell. At pH ¼ 4.3, the release of FITC was restrained and remained constantly with <11% in the course of 377 h. On the other hand, the release efficiency was gradually to increase and reached a plateau (82%) over 288 h at pH 7.2. To examine DOX release behavior, DOX was loaded into the triple-shelled nanospheres after fabrication of a 3rd silica shell without inclusion of FITC. At pH ¼ 4.3, DOX showed a faster release and reached 29% during the first 48 h, followed by gradual increase to 42% after 377 h. Meanwhile, the amount released was only 8.9% in the course of 48 h and then leveled off (13.6%) at pH ¼ 7.2. These behaviors provide the evidence that FITC exhibits poor release while DOX favors faster escape from multishelled silica nanospheres at acidic condition. Finally, we carried additional experiments to examine the release of FITC from doubleshelled silica nanospheres (Fig. 8b). With the same consequence as those observed in triple-shelled silica, FITC release was limited at acidic pH ¼ 4.3. Significant FITC molecules were released and achieved 75% in the course of 48 h at pH ¼ 7.2. As a comparison between double-shelled and triple-shelled nanospheres, it took w216 h to achieve 75% of FITC release in triple-shelled nanospheres. Therefore, using triple-shelled silica with fluorophores loaded inside the inner cavity has potential to serve for bio-labeling tracking.

4. Conclusion With the introduction of surface passivation and interior channel modification, multi-shelled shell-by-shell silica nanospheres were successfully fabricated using a self-templated approach. Interior modification by including APTES allowed controllable tuning of the shell-to-shell distance. Such a synthetic approach could be applied to solegel derived colloids for the formation of multi-shelled nanospheres. The multi-shelled structures were also demonstrated to simultaneously act as imaging probes and perform drug release by the appropriate selection of pH-dependent molecules.

Acknowledgement This work was supported by the National Science Council, Taiwan. We thank Dr. Jang-Yang Chang for kind help with in vivo fluorescence in National Health Research Institutes (NHRI).

Appendix. Supplementary data TEM images, BET analysis, FTIR spectra, EDS, MTT assay, and UVeVis spectra are provided. Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j. biomaterials.2010.08.114.


C.-C. Huang et al. / Biomaterials 32 (2011) 556e564

Appendix Figures with essential color discrimination. Figs. 4, 6e8 in this article are difficult to interpret in black and white. The full color images can be found in the online version, at doi:10.1016/j. biomaterials.2010.08.114.

References [1] Yang M, Ma J, Zhang C, Yang Z, Lu Y. General synthetic route toward functional hollow spheres with double-shelled structures. Angew Chem Int Ed 2005;44:6727e30. [2] Gu D, Bongard H, Deng Y, Feng D, Wu Z, Fang Y, et al. An aqueous emulsion route to synthesize mesoporous carbon vesicles and their nanocomposites. Adv Mater 2009;21:833e7. [3] Lou XW, Archer LA, Yang Z. Hollow micro-/nano-structures: synthesis and applications. Adv Mater 2008;20:3987e4019. [4] Sun Y, Wiley B, Li ZY, Xia Y. Synthesis and optical properties of nanorattles and multiple-walled nanoshells/nanotubes made of metal alloys. J Am Chem Soc 2004;126:9399e406. [5] Lou XW, Yuan C, Archer LA. Shell-by-shell synthesis of tin oxide hollow colloids with nanoarchitectured walls: cavity size tuning and functionalization. Small 2007;3:261e5. [6] Lou XW, Li CM, Archer LA. Designed synthesis of coaxial [email protected] hollow nanospheres for highly reversible lithium storage. Adv Mater 2009;21:2536e9. [7] Lou XW, Yuan C, Archer LA. Double-walled SnO2 nano-cocoons with movable magnetic cores. Adv Mater 2007;19:3328e32. [8] Li G, Shi Q, Yuan SJ, Neoh KG, Kang ET, Yang X. Alternating silica/polymer multilayer hybrid microspheres templates for double-shelled polymer and inorganic hollow microstructures. Chem Mater 2010;22:1309e17. [9] Chen D, Li L, Tang F, Qi S. Facile and scalable synthesis of tailored silica ‘‘nanorattle’’ structures. Adv Mater 2009;21:3804e7. [10] Wu XJ, Xu D. Formation of yolk/SiO2 shell structures using surfactant mixtures as template. J Am Chem Soc 2009;131:2774e5. [11] Zhang Q, Zhang T, Ge J, Yin Y. Permeable silica shell through surface-protected etching. Nano Lett 2008;8:2867e71. [12] Zhang Q, Lee I, Ge J, Zaera F, Yin Y. Surface-protected etching of mesoporous oxide shells for the stabilization of metal nanocatalysts. Adv Funt Mater 2010;20:2201e14. [13] Zhang Q, Ge J, Goebl J, Hu Y, Lu Z, Yin Y. Rattle-type silica colloidal particles prepared by a surface-protected etching process. Nano Res 2009;2:583e91. [14] Zeng F, Sun Z, Wu S, Liu X, Wang Z, Tong Z. Preparation of highly charged, monodisperse nanospheres. Macromol Chem Phys 2002;203:673e7. [15] Brijmohan SB, Swier S, Weiss RA, Shaw MT. Synthesis and characterization of cross-linked sulfonated polystyrene nanoparticles. Ind Eng Chem Res 2005;44:8039e45.

[16] Huang CC, Khu NH, Yeh CS. The characteristics of sub 10 nm manganese oxide T1 contrast agents of different nanostructured morphologies. Biomaterials 2010;31:4073e8. [17] Wang Y, Shi Y, Xu X, Liu F, Yao H, Zhai G, et al. Preparation of PANI-coated poly (styrene-co-styrene sulfonate) nanoparticles in microemulsion media. Colloids Surf A Physicochem Eng Aspects 2009;345:71e4. [18] Li IF, Su CH, Sheu HS, Chiu HC, Lo YW, Lin WT, et al. Gd2O(CO3)2$H2O particles and the corresponding Gd2O3: synthesis and applications of magnetic resonance contrast agents and template particles for hollow spheres and hybrid composites. Adv Funct Mater 2008;18:766e76. [19] Yu M, Lin J, Fang J. Silica spheres coated with YVO4: Eu3þ layers via solegel process: a simple method to obtain spherical core-shell phosphors. Chem Mater 2005;17:1783e91. [20] Mayerhöfer TG, Shen Z, Leonova E, Ed, én M, Kriltz A, Popp J. Consolidated silica glass from nanoparticles. J Solid State Chem 2008;181:2442e7. [21] Lin HP, Yang LY, Mou CY, Liu SB, Lee HK. A direct surface silyl modification of acid-synthesized mesoporous sililca. New J Chem 2000;24:253e5. [22] Kimura T, Saeki S, Sugahara Y, Kuroda K. Organic modification of FSM-type mesoporous silicas derived from kanemite by silylation. Langmuir 1999;15:2794e8. [23] Liu YH, Lin HP, Mou CY. Direct method for surface silyl functionalization of mesoporous silica. Langmuir 2004;20:3231e9. [24] Beck JS, Vartuli JC, Roth WJ, Leonowicz ME, Kresge CT, Schmitt KD, et al. A new family of mesoporous molecular sieves prepared with liquid crystal templates. J Am Chem Soc 1992;114:10834e43. [25] Kim DE, Jaffer FA, Weissleder R, Tung CH, Schellingerhout D. Near-infrared fluorescent imaging of cerebral thrombi and bloodebrain barrier disruption in a mouse model of cerebral venous sinus thrombosis. Blood Flow Metab 2005;25:226e33. [26] Niemann C, Henthorn T, Krejcie T, Shanks C, Enders-Klein C, Avram M. Indocyanine green kinetics characterize blood volume and flow distribution and their alteration by propranolol. Clin Pharmacol Ther 2000;67:342e50. [27] Bae Y, Fukushima S, Harada A, Kataoka K. Design of environment-sensitive supramolecular assemblies for intracellular drug delivery: polymeric micelles that are responsive to intracellular pH change. Angew Chem Int Ed 2003;42:4640e3. [28] Gillies ER, Frechet JMJ. pH-Responsive copolymer assemblies for controlled release of doxorubicin. Bioconjug Chem 2005;16:361e8. [29] Lai PS, Lou PJ, Peng CL, Pai WN, Yen MY, Huang TH, et al. Doxorubicin delivery by polyamidoamine dendrimer conjugation and photochemical internalization for cancer therapy. J Control Release 2007;122:39e46. [30] Hovorka O, St’astny M, Etrych T, Subr V, Strohalm J, Ulbrich K, et al. Differences in the intracellular fate of free and polymer-bound doxorubicin. J Control Release 2002;80:101e17. [31] Huang DM, Hung Y, Ko BS, Hsu SC, Chen WH, Chien CL, et al. Highly efficient cellular labeling of mesoporous nanoparticles in human mesenchymal stem cells: implication for stem cell tracking. FASEB J 2005;19:2014e6. [32] Fuh MRS, Burgess LW, Hirschfeld T, Christian GD, Wang F. Single fiber optic fluorescence pH probe. Analyst 1987;112:1159e63. [33] Adams DJ. The impact of tumor physiology on camptothecin-based drug development. Curr Med Chem Anticancer Agents 2005;5:1e13.