Engineering aggregation-induced SERS-active porous Au@ZnS multi-yolk–shell structures for visualization of guest species loading

Abstract

Herein, we present aggregation-induced surface-enhanced Raman scattering (SERS)-active hierarchical structures that effectively capture guest species loading in hollow nanocaged materials. The developed SERS-active probe by in situ aggregation was constructed of porous Au@ZnS multi-yolk–shell structures derived from Au@ZIF-8 multi-core–shell structure precursors. The porous ZnS shells characterized by nitrogen adsorption–desorption isotherm enable guest species such as hydrophobic 4-MBA and hydrophilic R6G molecules to shuttle based on their size. As a result of effective localization of the electric field by engineering hotspots in metallic aggregates, the intensity of these SERS reporter molecules adsorbed on Au NPs exhibited a dramatic enhancement during aggregation of Au NPs induced by electrolyte. Although the aggregation of Au NPs was able to be characterized by TEM and the change of plasmonic band from UV-vis absorption, some guest species that were incapable of inducing assembly of Au NPs were difficult to identify by both techniques. After taking into consideration other factors, including the sequence of guest species and incubation time, the variation of SERS signals related to individual Au NPs and their aggregates was applied to visualization of different types of molecules with Raman activity and ion loading in porous nanocages.

---

Introduction

Porous nanocaged materials with some extraordinary features including low density, open channels on the shells and high permeability have attracted great attention as nanoreactors and nanocontainers for catalysis, drug loading and delivery, and so on.¹ Hence, the development of a plain system for understanding the shuttling behavior of different types of guest species between interiors and exteriors of the materials in detail is necessary. Surface-enhanced Raman scattering (SERS) is often applied to the detection of Raman-active molecules adjacent to plasmonic metal engineered into specific structures with sharp corners or small gaps to localize the electric field.² The use of metal encapsulated porous nanocages (yolk–shell structures) could be an effective tool for monitoring molecule diffusion into interiors to contact with metal by SERS measurement, but non-Raman-active ions could not be detected. The ability to enhance the electric field by spherical plasmonic metal nanoparticles (NPs) such as Au NPs is limited, but their aggregates are available for effective coupling at inter-NP gaps.³ Owing to the assembly of colloidal metal NPs easily triggered by adjusting the medium environment,⁴ porous multi-yolk–shell nanostructures should meet these needs by in situ aggregation of metal spherical NPs inducing SERS variation.

Yolk–shell structures are generally derived from their core–shell precursors by self-sacrificed template strategies. Despite the developed template techniques to synthesise various yolk–shell structures, including amorphous silica, carbon, and polymers as shells for yolk–shell structures,⁵ the preparation of multi-yolk–shell structures of controlled size and well-defined shape is still a great challenge.⁶ To address the key issues in synthesizing the heterostructures, the exploration of dexterous strategies to access their multi-core–shell precursors are necessary. Metal–organic frameworks (MOFs) with tunable components and topological structures have become intriguing templates for the preparation of porous nanocages of metal compounds via solution phase processing routes.⁷ On the other hand, since MOFs constructed with organic ligands and metal ions by coordination bonds could tolerate many deficiencies during the crystal growth process, MOF-based multi-core heterostructures could be easily synthesized.⁸ According to the nature of MOF structures, a benefit of the MOF-based template method is the ease of forming porous structures due to the relatively lower packed atom density of MOFs compared to that of inorganic compounds. Therefore, MOFs can be employed as potential sacrificial templates for the synthesis of porous multi-yolk–shell materials.

The aim of this study is to track molecules and ion loading in hollow structured porous materials by SERS measurement. Herein, we chose porous ZnS nanocages as hollow nanocaged materials and aggregates of spherical Au NPs for a SERS probe. The simultaneous occurrence of the dissolution of MOF templates and the growth of metal compound grains presumably leads to a collapse of the hollow structures. This would bring a critical challenge in the preparation of yolk–shell structures. To simplify the complicated process to fabricate the heterostructures, we firstly investigated zeolitic imidazolate framework-8 (ZIF-8) as a self-sacrificed template to prepare porous ZnS nanocages, and explore the formation process in detail. On this basis, the use of multi-core Au@ZIF-8 heterostructures as precursors synthesized well-defined porous Au@ZnS multi-yolk–shell structures. The aggregation-induced SERS activity of the yolk–shell structures was applied to visualizing the loading of hydrophobic 4-mercaptobenzoic acid (4-MBA), hydrophilic rhodamine 6G (R6G), and sodium chloride (NaCl) electrolyte as guest species in ZnS nanocages (Fig. 1).

Fig. 1 Schematic illustration of preparation strategy to porous Au@ZnS multi-yolk–shell structures for visualization of guest species shuttling by SERS measurement.

Experimental section

Materials and characterization methods

All chemical reagents were used as received. Polyvinylpyrrolidone (PVP, M_w: 55 000) and 4-mercaptobenzoic acid were purchased from Sigma Aldrich, and all other chemicals and solvents were Sinopharm Chemical Reagent Co., Ltd, China. Copper specimen grids (300 mesh) were purchased from Beijing XXBR Technology Co.

Transmission electron microscopy (TEM) images were collected on a Tecnai F30 operated at 300 kV. X-ray diffraction (XRD) was preformed with a Rigaku D/Max 2400 automatic powder X-ray diffractometer with Cu-Kα radiation (λ = 1.5418 Å). N₂ sorption isotherm at 77 K was carried out with a Micromeritics 3Flex Surface Characterization Analyzer after being degassed in a vacuum at 200 °C for 12 h. Surface-enhanced Raman scattering (SERS) spectra were monitored with a Jobin Yvon LabRAM HR Evolution Raman Spectrometer. UV-vis absorption spectra were obtained from a TU-1900 spectrophotometer.

Synthesis of ZnS nanocages derived from ZIF-8 rhombic dodecahedrons

ZIF-8 rhombic dodecahedral nanocrystals 300 nm in size were prepared by a modified method. In a typical procedure, identical amounts of Zn(NO₃)₂ (400 mL, 7.5 mM) and 2-methylimidazole (Hmim, 400 mL, 7.5 mM) methanolic solutions were mixed in a 1 L conical flask and kept at room temperature for 24 h without disturbance. The final ZIF-8 powder collected and dried in a vacuum at 100 °C for 12 h was used for the preparation of ZnS nanocages.

ZnS nanocages were prepared via a self-sacrificed template route. ZIF-8 nanocrystals (19.6 mg) were dispersed into a mixture of acetonitrile (50 mL) and water (25 μL) in a 100 mL round-bottomed flask, and heated to 100 °C with gentle stirring. After that, thioacetamide (2 mL, w/v: 50.16 mg mL⁻¹) was added, and this step was repeated five times at intervals of 1.5 h. The reaction was terminated after 8 h. The resultant sample was collected, washed and dried in a vacuum for subsequent characterization.

Synthesis of porous Au@ZnS multi-yolk–shell structures derived from multi-core Au@ZIF-8 heterostructures

Au NPs (d = 15 nm) were synthesized firstly via a citrate-reduced method. In a typical procedure, after deionized water (100 mL) in a round-bottomed flask was heated to boiling at 110 °C with magnetic stirring, chloroauric acid aqueous solution (HAuCl₄, 1 mL, 1 wt%) was added and heated for 15 min. Afterwards, trisodium citrate aqueous solution (3 mL, 1 wt%) was introduced to the flask in one shot. The resulting mixture was heated for another 30 min and cooled down to room temperature naturally.

Multi-core Au@ZIF-8 heterostructures were firstly synthesized via a seed-mediated process. Au NP colloidal solution (100 mL) was added to PVP aqueous solution (100 μL, w/v: 25.6 mg mL⁻¹) to form PVP modified Au NPs. PVP–Au NPs (1 mL) were concentrated to about 5 μL using centrifugation at 17 000g, and dispersed into 7 mL of methanol. Hmim solution (1.5 mL, 15 mM, in methanol) and Zn(NO₃)₂ (1.5 mL, 15 mM, in methanol) were added in a shot with gentle shaking to form a uniform solution. The resulting solution was kept at room temperature for 24 h without disturbance. The final product was collected and dried in a vacuum for subsequent characterization and utilization of a sacrificial template.

Porous Au@ZnS multi-yolk–shell structures were synthesized by the same procedure as for the preparation of ZnS nanocages. The final product was collected, washed and dried in a vacuum for characterization.

SERS measurements and visualization of aggregation of Au NPs

All SERS spectra of solid samples were performed using a Jobin Yvon LabRAM HR Evolution Raman Spectrometer with a 600 l mm⁻¹ grating. SERS measurements were acquired with 633 nm of a He–Ne laser with a power of 1.7 mW and 514 nm of an Ar⁺ laser with a power of 2.5 mW to avoid the photothermal effect of Au NPs to burn porous Au@ZnS multi-yolk–shell structures. The ×50 NA = 0.5 long working distance objective was used to collect the Raman spectra with an acquisition time of 30 s.

Before being used, porous Au@ZnS multi-yolk–shell structures (5 mg) were dispersed in chloroform/n-hexane (50 mL, v/v = 1 : 1) for 12 h with gentle stirring and washed with ethanol several times before being dried in a vacuum to enhance the sensitivity of the SERS probes. The yolk–shell structures (1 mg) were dispersed into a SERS reporter solution, such as 4-MBA ethanolic solution (1 mL, 10⁻⁵ M, 10⁻⁴ M) and R6G aqueous solution (1 mL, 10⁻⁶ M), and incubated for 2 h before NaCl aqueous solution (100 μL, 1 M) was added with incubation for an additional 2 h. After that, the SERS probe was collected and dried at room temperature naturally for SERS measurement.

Results and discussion

Preparation and characterization of ZnS nanocages

Rhombic dodecahedral ZIF-8 nanocrystals with a size of about 300 nm were synthesized by mixing equivalent amounts of Zn(NO₃)₂ and Hmim in methanol and were incubated for 24 h at room temperature, characterized by transmission electron microscopy (TEM, Fig. 2a and b). ZnS nanocages were fabricated by using ZIF-8 as precursor via a conventional solution procedure. ZIF-8 powders were dispersed into acetonitrile and heated to boiling after adding a trace of water (25 μL); thioacetamide solution in acetonitrile as the S source was introduced into ZIF-8 colloidal solution in five equal portions at 1.5 h intervals to yield the completed ZnS nanocages. ZIF-8 rhombic dodecahedral nanocrystals were completely transferred into ZnS nanocages with rhombic dodecahedral shells. With careful observation of the TEM images (Fig. 2d and e), we found that the nanocages were assembled by tiny ZnS nanocrystals (2–3 nm in diameter) which was confirmed by their crystal lattices corresponding to {111} facets of ZnS (JCPDS 00-065-1691) from a high-resolution TEM (HRTEM) image (Fig. S1, ESI†). The different contrast in the TEM images reveals the hollow structures with 20–40 nm shell thickness. The ZnS nanocages exhibit rhombic dodecahedral geometry with a slight shrinkage compared to the ZIF-8 nanocrystals. The sodalite topology of ZIF-8 and the zinc blende crystalline structure of ZnS were characterized by X-ray diffraction (XRD) patterns without other impurity diffraction peaks (Fig. 2c and f). The broad diffraction peaks of ZnS reveal that nanocages comprise small ZnS nanocrystals, which is in agreement with the analysis of the HRTEM images. As a result, to achieve well-defined porous yolk–shell structures, it is desirable to precisely adjust the balance between the dissolution rates of the templates and the growing rates of the products.

Fig. 2 TEM images of (a, b) ZIF-8 rhombic dodecahedra and (d, e) porous ZnS rhombic dodecahedral nanocages with different magnification; XRD patterns of (c) ZIF-8 and (f) ZnS.

To clearly understand the formation process of ZnS nanocages with rhombic dodecahedral geometry, time-dependent experiments were performed. The yolk–shell structures of ZIF-8/ZnS composites were captured by TEM at different time periods (Fig. S2, ESI†). On addition of one portion of S source to the preparation system with a reaction time of 1.5 h, the polyhedral contour of ZIF-8 nanocrystals became blurred and the rhombic dodecahedral shells of the ZnS formed were distorted even slightly collapsed. This could be attributed to the thin-shell structures that cannot sustain the rhombic dodecahedral frameworks during the drying process on the TEM grid. When three portions of S source at 1.5 h intervals were added, ZIF-8 was further dissolved into smaller nanocrystals and thick-shell ZnS with a high mechanical stability were found. Hence, formation of the ZnS nanocages is based on a sacrificial template-directed chemical transformation of ZIF-8 polyhedral nanocrystals. That is to say, thioacetamide molecules are hydrolyzed to produce S²⁻ ions that diffuse to the external surfaces of ZIF-8 and react with Zn²⁺ ions in terminated surface planes, and newly-formed (ZnS)_n clusters aggregate into primary nanocrystallites. After exhausting the surface Zn²⁺ ions, S²⁻ ions subsequently diffuse into interiors from the newly formed ZnS shells and seize the surface Zn²⁺ ions on remanent ZIF-8 for ZnS nanocrystal growth, ultimately resulting in the formation of ZnS nanocages.

Albeit the transformation process is thermodynamically driven by the decrease of the total energy from the difference of lattice energy between ZnS and ZIF-8, the assembly shapes of ZnS are decided by the kinetics of growth related to the amount of reactants, and the intrinsic properties of the materials. The concentration of S²⁻ ions produced by hydrolysis of thioacetamide, depending on the amount of water and thioacetamide, strongly affects the dissolution rate of ZIF-8 nanocrystals. Incomplete ZnS nanocages were found in the final product, accompanied with exfoliated lamellar structures from broken nanocages, as well as many scattered ZnS nanocrystal aggregates while the volume of water increased to 50 μL under otherwise identical conditions (Fig. S3, ESI†). Furthermore, as the volume of water in the reaction system increased, ZnS nanocages gradually disappeared and many more ZnS nanocrystal aggregates were present. Additionally, if all the thioacetamide was introduced into the reaction system in one shot, the assembled hierarchical hollow structures of ZnS were incapable of being synthesized. Therefore, the precursor templates could be damaged due to the fast dissolution of ZIF-8 in a high concentration of S²⁻ ions, and a large amount of (ZnS)_n clusters aggregate into primary nanocrystallites in bulky solution as nuclei, accounting for the irregular ZnS aggregate formation. To achieve ZnS nanocages with rhombic dodecahedral geometry, the production rate of S²⁻ ions needs to be controlled precisely.

In contrast to the pioneering work on using ZIF-8 as precursor templates to prepare ZnS in ethanolic solution,⁹ if we use ethanol instead of acetonitrile as reaction medium, a few completed ZnS nanocages can be formed. Occurrence of this phenomenon could be attributed to the different stability of ZIF-8 rhombic dodecahedra in different solvents during the transformation process. The external surface of ZIF-8 is the layer of Zn²⁺ ions that are bound into the bulk crystal structure by three Hmim linkers,¹⁰ thus the unsaturated coordination Zn²⁺ ions can further interact with N atoms of acetonitrile molecules, leading to a higher energy barrier for S²⁻ ions attacking Zn²⁺ ions. Therefore, the degradation rate of ZIF-8 in acetonitrile solvent is slower than that in ethanol solvent, facilitating the formation of completed ZnS polyhedral nanocages.

Preparation and characterization of porous Au@ZnS multi-yolk–shell nanostructures

After understanding the formation of porous ZnS nanocages derived from ZIF-8 NPs, porous Au@ZnS multi-yolk–shell structures can be synthesized using the same procedure. The multi-core Au@ZIF-8 heterostructures were firstly synthesized by a seed-mediated growth route.¹¹ To achieve SERS-active porous Au@ZnS yolk–shell heterostructures by in situ aggregation in solution, the amount of polyvinylpyrrolidone (PVP) needed to be decreased. PVP modified spherical Au (PVP–Au) NPs (Fig. S4, ESI†) were dispersed into a diluted methanolic solution containing Zn(NO₃)₂ and Hmim and incubated at room temperature for 24 h without disturbance. On the basis of MOF special structures that can tolerate many more defects, the morphology of most heterostructures, loading tens of Au NPs in each ZIF-8 NP, still retained rhombic dodecahedra in shape with a mean size of 300–400 nm, accompanied with a few multifaceted shapes (Fig. 3a and b). Few Au@ZIF-8 rhombic dodecahedra can be achieved if the reactant concentration is increased, such as with 12.5 mM (Fig. S5, ESI†). The XRD pattern of Au@ZIF-8 (Fig. 3c) is similar to that of ZIF-8, except that a weak diffraction peak located at 38.1° is assigned to the {111} facet of Au (JCPDS 00-004-0784). The transformation of the core–shell structures into Au@ZnS multi-yolk–shell structures occurred while adding thioacetamide in batches to the suspension of core–shell structures and refluxing for 8 h. TEM images (Fig. 3d and e) reveal that rhombic dodecahedral nanocages constructed by ZnS shells with 20–40 nm in thickness contain tens of Au NPs. The size and shape of the Au NPs are not apparently changed after transformation. A few broken nanocages with scattered Au NPs at their outsides can be found in the TEM images. Multifaceted Au@ZIF-8 nanocrystals were unfavorable for the formation of complete yolk–shell structures (Fig. S5, ESI†). The crystalline structures of the final products were characterized by XRD (Fig. 3f), and no other impurity could be found except for zinc blende ZnS and cubic Au.

Fig. 3 TEM images of (a, b) multi-core Au@ZIF-8 composites and (d, e) porous Au@ZnS multi-yolk–shell structures with different magnification; XRD patterns of (c) Au@ZIF-8 and (f) Au@ZnS structures.

To affirm the porous structures about Au@ZnS composites, nitrogen adsorption–desorption isotherms at 77 K were used to analyze the specific surface area and average pore diameter (Fig. 4a and b). Type III nitrogen sorption isotherm behavior and non-closed hysteresis loops for Au@ZnS revealed the adsorption was reversible and the adsorbate remained in the porous ZnS shells at a pressure near zero. The size distribution of pores was from larger than one nanometer to tens of nanometers, and smaller micropores were not able to be found, indicating that ZIF-8 was completely converted. Hydrophobic 4-MBA and hydrophilic R6G were chosen as SERS reporter molecules in order to study in detail the molecular diffusion in the novel heterostructures according to both molecules in width smaller than the smallest pores in size.

Fig. 4 (a) Nitrogen sorption isotherm at 77 K (adsorption: solid; desorption: hollow) and (b) pore size distribution of porous Au@ZnS multi-yolk–shell structures.

Aggregation-induced SERS-activity for visualization guest species loading

SERS-active probes for enhancing the cross section of SERS reporter molecules were dependent on the aggregation of spherical Au NPs to form many hotspots during the SERS measurement process. Ligand exchange of spherical Au NPs in bulky solution could lead to their aggregation, driven by hydrophobic effects.¹² The porous Au@ZnS multi-yolk–shell structures were dispersed into 4-MBA ethanolic solution (1 mL, 10⁻⁴ M), and incubated for 2 h. After that, the suspension separated by centrifugation was allowed to dry naturally in air. The SERS spectra of solid powders were acquired using a Jobin-Yvon LabRAM spectrometer under a 633 nm He–Ne laser with a low power of 1.7 mW at the sample while avoiding severe photothermal effects. The 4-MBA Raman signals are hardly observed in the SERS spectrum (Fig. 5a-1).

Occurrence of the phenomenon is possibly caused by three possible scenarios: weak enhanced near field at 633 nm excitation, few hotspots, and few reporter molecules locating at hotspots. According to the UV-visible absorption of Au NPs (d = 15 nm), the peak of the plasmon band is located at 525 nm (Fig. S4, ESI†). Thus, we chose the excitation frequency of 514 nm laser for near-field enhancement based on their effective spectral coupling.¹³ Enhancement of the SERS intensity has not yet been found. On this basis, we speculate that aggregation of Au NPs in yolk–shell structures did not occur under the present conditions. To obtain a large amount of hotspots for trace detection, we further added electrolyte, such as sodium chloride (NaCl) that can induce aggregation of metal NPs, to the mixture of Au@ZnS and 4-MBA in ethanol after incubation for 2 h.¹⁴ Owing to ZnS shell scattering and absorption, the plasmon bands of Au@ZnS composites in solution related to assembly of Au NPs cannot be captured by UV-vis absorption spectroscopy. In the SERS spectrum (Fig. 5a-3), the strong bands at 1080 and 1589 cm⁻¹ attributed to ν₁₂ and ν_8a aromatic ring vibrations of 4-MBA were found, but other vibration modes of 4-MBA were too weak to be observed.¹⁵ The enhancement of SERS intensity was attributed to the formation of Au aggregates induced by NaCl (50 μL, 1 M in water) or 4-MBA with long incubation times. The latter was ruled out based on no obvious change in the SERS intensity when increasing the incubation time of Au@ZnS heterostructures towards 4-MBA to 4 h (Fig. 5a-2).

On the other hand, owing to the stability of PVP–Au NPs retarding their aggregation, porous Au@ZnS multi-yolk–shell structures before being used were dispersed in a mixed solvent of n-hexane and chloroform with gentle stirring to reduce PVP molecules on Au NPs. The treated yolk–shell structures (TEM image shown in Fig. 5c) were incubated in 4-MBA ethanolic solution, and then added to NaCl aqueous solution (100 μL, 1 M in water) to induce Au NPs assembly, confirmed by the TEM image of the linear chains formed (Fig. 5b). After aggregation, the hotspots were formed near the attached areas between Au NPs. 4-MBA molecules trapped in the hotspots led to a dramatic enhancement of Raman signals in the SERS spectrum (Fig. 5a-5). Apart from the aromatic ring vibrations, the bands at 1135 and 1178 cm⁻¹ are assigned to C–H deformation modes; two bands at 844 and 1401 cm⁻¹ correspond to the COO⁻ deformation mode and stretching vibration, respectively. No bands corresponding to the –SH group were found in the SERS spectrum. As expected, some new bands at 930 and 996 cm⁻¹ also appeared presumably due to 4-MBA adsorbed on the surface of Au NPs.¹⁵ Moreover, on dilution of the 4-MBA in the system from 10⁻⁴ to 10⁻⁵ M, the Raman scattering signals of 4-MBA with slightly attenuated intensity can be detected. This result indicates that 4-MBA molecules under this concentration are enough to assemble on Au NPs into monolayers (Fig. 5a-4) although some 4-MBA molecules could adsorb on the ZnS shells. The change of 4-MBA SERS intensity is very sensitive to the amount of hotspots induced by NaCl, and thus can be used to visualize 4-MBA and NaCl diffusing into ZnS nanocages.

Fig. 5 (a) Measured SERS spectra of porous Au@ZnS multi-yolk–shell structures incubation in 4-MBA (10⁻⁴ M) for (1) 2 h and (2) 4 h, (3) sample 1 incubating NaCl (50 μL, 1 M) for 2 h, and treated by n-hexane/chloroform incubation in (4) 10⁻⁵ M and (5) 10⁻⁴ M of 4-MBA for 2 h before incubating NaCl (100 μL, 1 M) for additional 2 h; (b) and (c) TEM images of porous Au@ZnS multi-yolk–shell structures treated prior to and after NaCl addition, respectively. All scale bars are 50 nm, and all SERS spectra were carried out at the excitation wavelength of 633 nm.

To further understand other factors affecting the SERS intensity, we performed some parallel experiments involving the introduction sequence of 4-MBA and NaCl, as well as the incubation time. The variation of SERS intensity was characterized from the aromatic ring vibrations of 4-MBA at 1589 cm⁻¹. Albeit the dramatic change of SERS spectra was not observed (Fig. 6a), the SERS intensity was attenuated when aggregation induced by ions is prior to the adsorption of 4-MBA. The phenomena could be attributed to adsorption of 4-MBA molecules on hot spots, as well as the negative effect of slightly increased gap distance of hot spots due to 4-MBA implantion (Fig. 6b). Moreover, the results reveal that the diffusion rate of ions is higher than that of 4-MBA molecules. On the other hand, if the incubation time of porous Au@ZnS multi-yolk–shell composites towards NaCl was diminished to 10 min after incubating in 4-MBA ethanolic solution for 2 h, the intensity of SERS signals in the controlled experiments were slightly stronger than those without addition of NaCl; but while increasing incubation time to 30 min, the characteristic SERS bands of 4-MBA appeared. When the incubation time was further increased to 1 h, the SERS intensity was dramatically enhanced, indicating that the concentration of NaCl between the interior and exterior was balanced. In contrast, Au NPs in solution were aggregated immediately after adding 4-MBA or NaCl. Hence, porous ZnS shells served as a spacer to control guest species loading and release.

Fig. 6 SERS spectra of porous Au@ZnS multi-yolk–shell structures (a) incubated in solutions with different introduction sequences of 4-MBA (1 mL, 10⁻⁴ M) and NaCl (100 μL, 1 M) and (c) incubation in NaCl with different times; schematic illustration of (b) 4-MBA molecules in the hotspots (upper) before and (lower) after Au NP aggregation and (d) Au NP aggregation with different incubation times towards NaCl.

On the basis of porous Au@ZnS multi-yolk–shell composites, TEM and solid UV-vis absorption can also be used to visualize the diffusion behavior of hydrophobic molecules and electrolytes as guest species that enable Au NP aggregation. Nevertheless, in respect to hydrophilic molecules such as R6G that cannot induce Au NPs assembly, both techniques are incapable of visualization of this type of molecule loading. Aggregation-induced SERS activity can be used to detect the diffusion of R6G into ZnS nanocages. When porous Au@ZnS yolk–shell structures were incubated into R6G aqueous solution (1 mL, 10⁻⁶ M), the reporter molecule was allowed to diffuse into the interior of the heterostructures theoretically based on the correlation of its width and pore size of ZnS. However, Au NPs did not exhibit the strong Raman enhancement (Fig. 7a). As such, the major issue should be the lack of hotspots in the SERS system (Fig. 7c). After introducing NaCl aqueous solution (100 μL, 1 M) to the system, aggregation of Au NPs occurred (Fig. 7b), and the intense Raman signals corresponding to the characteristic signals of R6G are observed (Fig. 7a). The strong bands at 1125, 1182, 1309, 1360, 1506, 1571, 1597, 1646 cm⁻¹ arose from C–C stretching vibrations, which is similar to the resonance Raman of R6G.

Fig. 7 SERS spectra of (a) porous Au@ZnS multi-yolk–shell structures kept in R6G for 2 h before and after introducing NaCl with an incubation time of 2 h; TEM images of samples with (b) and without (c) addition of NaCl. All scale bars are 50 nm.

Conclusions

In summary, to visualize various guest species loading into porous nanocaged materials, we have synthesized aggregation-induced SERS-active porous Au@ZnS multi-yolk–shell structures based on sacrificial template-directed chemical transformation of multi-core Au@ZIF-8 heterostructures. Exploration of ZnS nanocages derived from ZIF-8 rhombic dodecahedra was to simplify the complicated preparation process of Au@ZnS composites. Porous ZnS shells characterized by N₂ sorption isotherms allowed guest species smaller than ∼10 Å in width to shuttle. On the basis of the difference of ability to enhance electric fields between individual Au NPs and their aggregates, a dramatic variation of the SERS intensity induced by the assembly of Au NPs related to hotspots was applied to the visualization of guest species diffusion into the interior to contact with Au NPs. In combination with Au aggregation induced by ions and Raman signals provided by reporter molecules, aggregation-induced SERS activity can be considered as an effective tool for the detection of ion and molecule loading in hollow structures. A detailed understanding of the diffusion process of guest species will benefit the directional construction of porous nanocaged materials with desirable performances.

Acknowledgements

We gratefully acknowledgement financial support from the Foundation for Innovative Research Groups of the National Natural Science Foundation of China (21421005), the Specialized Research Fund for the Doctoral Program of Higher Education of China (021005), and the Fundamental Research Funds for the Central Universities (DUT15LK04).

Notes and references

1. (a) S.-H. Wu, C.-Y. Mou and H.-P. Lin, Chem. Soc. Rev., 2013, 42, 3862; (b) J. Dou and H. C. Zeng, J. Am. Chem. Soc., 2012, 134, 16235.
2. (a) S. Schluecker, Angew. Chem., Int. Ed., 2014, 53, 4756; (b) M. D. Sonntag, J. M. Klingsporn, A. B. Zrimsek, B. Sharma, L. K. Ruvuna and R. P. Van Duyne, Chem. Soc. Rev., 2014, 43, 1230.
3. (a) M. I. Stockman, Phys. Today, 2011, 64, 39; (b) S. M. Nie and S. R. Emery, Science, 1997, 275, 1102; (c) A. M. Michaels, J. Jiang and L. Brus, J. Phys. Chem. B, 2000, 104, 11965.
4. (a) G. Chen, Y. Wang, M. Yang, J. Xu, S. J. Goh, M. Pan and H. Chen, J. Am. Chem. Soc., 2010, 132, 3644; (b) H. X. He, H. Zhang, Q. G. Li, T. Zhu, S. F. Y. Li and Z. F. Liu, Langmuir, 2000, 16, 3846; (c) J. F. Li, Y. F. Huang, Y. Ding, Z. L. Yang, S. B. Li, X. S. Zhou, F. R. Fan, W. Zhang, Z. Y. Zhou, D. Y. Wu, B. Ren, Z. L. Wang and Z. Q. Tian, Nature, 2010, 464, 392.
5. (a) J. C. Park and H. Song, Nano Res., 2011, 4, 33; (b) W. Zhou, Y. Yu, H. Chen, F. J. DiSalvo and H. D. Abruna, J. Am. Chem. Soc., 2013, 135, 16736; (c) H. Sun, X. Shen, L. Yao, S. Xing, H. Wang, Y. Feng and H. Chen, J. Am. Chem. Soc., 2012, 134, 11243.
6. J. Liu, S. Z. Qiao, J. S. Chen, X. W. Lou, X. Xing and G. Q. Lu, Chem. Commun., 2011, 47, 12578.
7. X.-Y. Yu, L. Yu, H. B. Wu and X. W. Lou, Angew. Chem., Int. Ed., 2015, 54, 5331.
8. (a) Z. Li and H. C. Zeng, Chem. Mater., 2013, 25, 1761; (b) G. Lu, S. Li, Z. Guo, O. K. Farha, B. G. Hauser, X. Qi, Y. Wang, X. Wang, S. Han, X. Liu, J. S. DuChene, H. Zhang, Q. Zhang, X. Chen, J. Ma, S. C. J. Loo, W. D. Wei, Y. Yang, J. T. Hupp and F. Huo, Nat. Chem., 2012, 4, 310.
9. Z. Jiang, H. Sun, Z. Qin, X. Jiao and D. Chen, Chem. Commun., 2012, 48, 3620.
10. P. Y. Moh, P. Cubillas, M. W. Anderson and M. P. Attfield, J. Am. Chem. Soc., 2011, 133, 13304.
11. L. Chen, Y. Peng, H. Wang, Z. Gu and C. Duan, Chem. Commun., 2014, 50, 8651.
12. M. Grzelczak, J. Vermant, E. M. Furst and L. M. Liz-Marzan, ACS Nano, 2010, 4, 3591.
13. X. Wang, M. Li, L. Meng, K. Lin, J. Feng, T. Huang, Z. Yang and B. Ren, ACS Nano, 2014, 8, 528.
14. T. Chen, H. Wang, G. Chen, Y. Wang, Y. Feng, W. S. Teo, T. Wu and H. Chen, ACS Nano, 2010, 4, 3087.
15. A. Michota and J. Bukowska, J. Raman Spectrosc., 2003, 34, 21.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra04432h

---