Quantum-dot-encapsulated core–shell barcode particles from droplet microfluidics

Feika Bian , Huan Wang , Lingyu Sun , Yuxiao Liu and Yuanjin Zhao *
State Key Laboratory of Bioelectronics, School of Biological Science and Medical Engineering, Southeast University, Nanjing 210096, China. E-mail: yjzhao@seu.edu.cn

Received 9th April 2018 , Accepted 19th July 2018

First published on 19th July 2018


The development of robust quantum dot (QD) barcode particles with specific compositions and simple identification is important to meet the demand for high-throughput assays. Here, we present a multiple-inner phase channel capillary microfluidic approach to generate novel QD-encapsulated core–shell barcode particles with distinctive features for multiplexing analysis. By using different QD dispersed polyethylene glycol diacrylate (PEGDA) solutions as the inner phases, the particles were endowed with hydrogel locked QD cores, which could maintain the dispersed status and provide distinctive identification for the particles. The shells of the barcode particles were silica nanoparticle-dispersed ethoxylated trimethylolpropane triacrylate (ETPTA) resin, which could not only improve the stability and biocompatibility of QDs, but also provide functional groups for immobilization of biomolecules due to the assembling of the silica nanoparticles on their surfaces. Due to the advanced emulsification capability of the capillary microfluidic device, double emulsion templates with multiple inner droplet phases and their resultant multicomponent QD-encapsulated core–shell barcode particles could be continually generated. These particles showed remarkable spectral coding capacity in practice, which make them ideal for biomedical applications.


Introduction

Barcode particles have shown a demonstrated value in encoding information about their specific compositions and enabling simple identification for multiplexed high-throughput bioassays.1–7 Many encoding mechanisms have been developed to generate high-throughput and functional barcode particles,8–11 among which fluorescence of organic dyes and QDs is a common mechanism as identification codes.12–14 Compared with the traditional organic dyes, QDs exhibit numerous advantages including high quantum yields, broad absorption spectra coupled to narrow size-dependent emissions, exceptional resistance to photobleaching and size-tunable emission, owing to quantum confinement effects.15–17 QD-barcode particles are in demand due to their water-solubility, nontoxicity and multivariate analysis for biomedical applications. Capsulated by silica, polymers or hydrogels,18,19 QD-barcode particles could realize the above features and prevent the dispersed status from assembling. However, most of the present QD encapsulation processes are complicated and less controllable.20–22 In addition, the resultant QD-barcode particles are with debatable encoding stability due to the leaking of QD elements when the particles are exposed to solutions for a long time. Thus, a rapid and simple way to generate novel and stable QD-barcode particles is still anticipated.

In this paper, we proposed a simple microfluidic approach to generate QD-barcode particles with the desired features. The microfluidic technique offers a versatile platform for the generation of nano- or microsized particles.23–27 The droplet templates are generated and manipulated through immiscible flows inside microchannels in a highly controllable and reproducible manner, and particles can be synthesized by rapid polymerization.28–31 Thus, their production efficiency, which means the generation throughput of particles, and the particle monodispersity are much higher than those of conventional emulsification methods.32–34 In addition, double or multiple emulsion droplets can be generated on demand by implementing multilevel emulsification, through which particles with cores are available.35–37 These microfluidic methodologies have been successfully utilized to generate QD-barcode particles based on single or double emulsion templates.38 However, the recently developed microfluidic QD-barcode particles were composed of a simple hydrogel component or with a hydrogel shell surrounding, both of which were still suffering from QD leakage during their storage.

Thus, we herein employed a multiple-inner phase channel capillary microfluidic device for the generation of QD-encapsulated core–shell barcode particles, as schemed in Fig. 1. The cores of the barcode particles were polyethylene glycol diacrylate (PEGDA) hydrogels locked with different QDs, which could maintain the dispersed status and provide distinctive identification for the particles. The shells of the barcode particles were silica nanoparticle-dispersed ethoxylated trimethylolpropane triacrylate (ETPTA) resin, which could not only improve the stability and biocompatibility of QDs, but also provide functional groups for immobilization of biomolecules due to the assembling of the silica nanoparticles on their surfaces. Benefitting from the extended emulsification capability of the capillary microfluidic device, double emulsions with multiple inner droplet phases could also be continually generated, which provided advanced templates for the creation of multicomponent QD-encapsulated core–shell barcode particles with remarkable spectral coding capacity in practice. These features make the QD-barcode particles described here ideal for biomedical applications.


image file: c8tb00946e-f1.tif
Fig. 1 Schematic illustrations: (a) the capillary microfluidic device for the generation of multi-component double emulsions with inner droplets of different QDs; (b) the procedure of multivariate analysis based on the QD-encapsulated core–shell barcodes.

Results and discussion

In a typical experiment, a water-in-oil-in-water (W/O/W) double emulsion capillary microfluidic device was utilized as a template for QD-encapsulated core–shell barcodes. The inner aqueous phases containing 30% w/w PEGDA were mixed with different QDs (Fig. S1, ESI), and 1% w/w poly vinyl alcohol (PVA), 1% v/v ethyleneoxide propyleneoxide tri-block copolymer (Pluronic F108) and 1% v/v photoinitiator 2-hydroxy-2-methylpropiophenone were pumped through the seven-bore capillary array, respectively. The middle oil phase made up of silica nanoparticle-containing ETPTA resin and 1% v/v photoinitiator 2-hydroxy-2-methylpropiophenone was infused into the middle capillary in the same direction. The outer aqueous solution phase was 2% w/w PVA and flowed via the region between the square capillary and the middle capillary array. When the three phases were pumped into the capillary device, facilitated by hydrodynamic focusing, the inner water droplets were generated at the end of the inner capillary by the middle oil phase and were subsequently encapsulated by the middle oil phase at the orifice of the middle capillary to form monodispersed W/O/W double emulsions. As the seven-bore capillary and the middle capillary were coaxially assembled, the middle oil stream would precisely center the inner aqueous stream. By tuning the velocities of the three phases, the size of the whole particles and the encapsulated core droplets, and the shell thickness of the particles could be adjusted, as shown in Fig. 2a, b and Fig. S2 (ESI). When the flow rate of the outer phase was fixed, the increasing ratio of the flow rates of the inner and middle phases Finner/Fmiddle brought about the increasing radius of the inner droplets Rinner (Fig. 2c). Similarly, under a fixed inner-phase flow rate, the radius of the outer droplets Router increased with the ratio of the flow rates of the middle and outer phases, Fmiddle/Fouter (Fig. 2d). The PEGDA and ETPTA utilized in this study are both photopolymers, which can be immediately polymerized under ultraviolet (UV) irradiation. Thus, by polymerizing the droplet templates, particles with ETPTA resin shells and QD-PEGDA cores were achieved. The generation throughput of QD-encapsulated core–shell barcode particles could achieve 120 per minute. In addition, by using capillary with a smaller orifice and adjusting the flow rate, the particles could be imparted much smaller diameter to be compatible with cytometry.
image file: c8tb00946e-f2.tif
Fig. 2 The relationships between flow rates of the different phases and the diameters of both the cores and the whole droplets. (a and b) The real-time microscopic images of the microfluidic generation process of the droplets with tunable sizes. (c) The radius of the cores Rinner increased from 131 μm to 158 μm with the increase of Finner/Fmiddle. (d) The radius of the whole droplets increased from 267 μm to 289 μm with the increase of Fmiddle/Fouter.

During the generation of the double emulsion droplets, the silica nanoparticles in the middle oil phase would migrate spontaneously towards and assemble at the inner interface between the core and shell, and the outer interface between the shell and the continuous phase simultaneously decreasing the total interfacial energy of the system. The energy reduction is sufficient so the nanoparticles would not be dislodged from the interface by thermal motions. It is worth mentioning that the silica nanoparticles have surface silanol groups (Si–OH), after simple modification with a silane coupling agent, the nanoparticles on the outer interface could bind with desired functional groups for immobilization of biomolecules (Fig. 3a). To investigate the microstructure of the achieved QD-encapsulated core–shell barcodes, they were characterized with a scanning electron microscope (SEM). It was found that the QD-encapsulated core–shell barcodes showed monodispersed sphericity owing to the precise manipulation of the microfluidic approach. As can be seen from the cross-sectional view of a QD-encapsulated core–shell barcode after being cut, highly ordered surface architectures assembled by silica nanoparticles could be observed at the inner interface between the core and the shell (Fig. 3b), and the outer interface between the shell and the continuous phase (Fig. 3c), respectively, which proved the existence of silica nanoparticles and the feasibility of the follow-up experiment.


image file: c8tb00946e-f3.tif
Fig. 3 (a) Scheme of the core–shell barcode particles with silica nanoparticle arrays at the inner and outer surfaces. (b and c) SEM images of (b) the inner surface of a broken particle and (c) the outer surface (denoted as white boxes in the insets). The scale bars are 700 nm in (b and c), 300 μm in the inset.

The fluorescence stability of QD-barcode particles is essential for the following biomedical detection. In recent years, many methods have been put forward to improve the stability of QDs, such as encapsulation by silica, polymers or hydrogels, for the purpose of preventing the dispersed status from assembling. Here, in order to further stabilize the fluorescence intensity of QD-barcode particles, the cores were encapsulated by the resin. The resin shells could provide effective protection in aqueous solution for a long time. In this study, the QDs were locked by PEGDA, and encapsulated subsequently by ETPTA resin. To investigate the stability between the QD-PEGDA barcodes (Fig. 4a) and the QD-encapsulated core–shell barcodes (Fig. 4b), their fluorescence intensity was measured and compared for several days (Fig. 4c) and different pH values (Fig. 4d) using a fluorescence microscope. The results demonstrated that QD-encapsulated core–shell barcodes showed nearly constant fluorescence intensity along time and pH treatment, while QD-PEGDA barcodes showed a tendency of fluorescence decrease owing to the leakage of QDs. These results indicated that the resin shells improved the fluorescence stability of QDs efficiently, which benefited for the accuracy of biomedical detection.


image file: c8tb00946e-f4.tif
Fig. 4 The fluorescence microscopy images of (a) PEGDA encapsulated QD barcodes and (b) QD-encapsulated core–shell barcodes. (c) Time stability and (d) pH stability of the QD-PEGDA barcodes and QD-encapsulated core–shell barcodes. The scale bars are 300 μm.

It was notable that the multicomponent double emulsions with inner droplets of different content could be fabricated by the controllability of microfluidics to increase the coding capacity. The multicomponent double emulsions with inner droplets of different content could be fabricated for the controllability of microfluidics. In this study, three inner fluids containing different QDs were injected through the three channels of a seven-bore capillary array. These channels were nonadjacent, separated by blank capillaries, to prevent the inner phase droplets from fusion. In addition, the injection tube was hydrophobic, so the middle oil phase tended to wet the whole orifice of the injection capillary array during emulsification. Thus, when the aqueous solution phases mixed with different QDs were pumped into the capillary array, the inner droplets were in contact only with the immiscible middle oil phase, forming the desired multicomponent double emulsions. By tuning the flow rates of different phases, the number and components of the inner droplets inside the double emulsions could be adjusted. As the QD-encapsulated core–shell barcodes were identifiable optically using a fluorescence microscope, the QDs with different colours could be used directly as encoding elements without the need for extra decoding methods like fluorescence spectrometry, absorption spectrometry, Raman spectrometry and so on (Fig. 5a). Thus, the achievable codes are precise and controllable because of the individual difference among the particles. In theory, various combinations of QDs with different colours could provide a variety of QD-encapsulated core–shell barcodes when the number of cores is kept constant. Moreover, with the increase of the number of cores and the types of QDs, the encoding number would increase dramatically. Here, the QD-encapsulated core–shell barcode particles are encoded by the multiple characteristic fluorescence emission peaks of different QDs in various cores, which are readily decoded using a fibre optic spectrometer with a mercury lamp source combined with the fluorescence microscope (Fig. 5b). By adjusting the concentrations of inner phases, the photoluminescence intensity (PL Intensity) of each core was about 50 arbitrary units (a.u.) under the same excitation conditions. Thus, the core numbers could be directly utilized for accurate encoding. Because the cores in the particles located randomly and the cores with the same peak position are undistinguishable, the number of possible QD-barcode particles (N) can be expressed with a simple combination calculation as follows:

image file: c8tb00946e-t1.tif
where n is the number of all QD types, m is the number of fluorescence intensity levels, nm is the number of all unique characteristic fluorescence emission peaks and k is the number of the QD-PEGDA cores of the barcode particles. However, because of spectral overlapping, signal-to-noise requirements and fluorescence intensity variations, the actual coding capabilities might be substantially lower. It is better to employ more QD types, rather than more intensity levels, in order to increase the number of valuable codes. Theoretically, it was possible to utilize 4 QD-PEGDA cores with 6 types of QDs and 4 different fluorescence intensity levels, yielding 10626 recognizable codes. Therefore, these easily identifiable QD-encapsulated core–shell barcodes could serve as prominent fluorescence encoding particles for multiplex assays and in other biomedical fields.


image file: c8tb00946e-f5.tif
Fig. 5 (a) The fluorescence microscopy images of QD-encapsulated core–shell barcodes with (i) two cores and (ii) three cores; (b) The fluorescence spectrum of QD-encapsulated core–shell barcodes, the capillary microfluidic with inner droplets of three different QDs. Barcode m:n:k represented that the particles have m blue cores, n green cores and k red cores. The scale bars are 300 μm.

For biomedical applications, the barcodes should possess a large number of distinctive identification features to increase throughput and a biocompatible surface for biomolecule immobilization and reaction. The prepared QD-encapsulated core–shell barcodes from microfluidics have many advantages that make them an excellent choice for biomedical applications. To demonstrate the practical value of the QD-encapsulated core–shell barcodes, two kinds of them with two red cores (barcode 0:0:2) and three blue cores (barcode 3:0:0) were applied for multiplex immunoassays. The exposed silanol groups of silica nanoparticles were silanized with aminopropyl-trimethoxysilane (ATPMS) to introduce amine groups and then coupled with succinic anhydride to obtain carboxyl groups. The carboxyl groups could combine with the amine groups on the antibody to immobilize them. Barcode 3:0:0 particles were bound with anti-rabbit immunoglobulin G (IgG) and barcode 0:0:2 particles were bound with anti-human IgG, respectively. Then bovine serum albumin (BSA) was utilized to block the empty sites of the particles, to prevent undesired adsorption. The two kinds of QD-barcode particles were mixed together and incubated at room temperature in PBS buffer, each of which containing no target antigen, human IgG (at 10 mM), rabbit IgG (at 10 mM), and both targets (both at 5 mM). Because of the specific binding between the antigen and its corresponding fluorescently labelled antibody, it was expected that the fluorescence signals of the antibody would be observed only on QD-barcode particles whose corresponding antigens were present. Fig. 6 shows the result of multiplex antigen detection. In each instance, it was found that only when the corresponding target antigen existed, anti-rabbit IgG or anti-human IgG tagged with fluorescein isothiocyanate (FITC) could combine with the homologous QD-barcode particles. Thus, the unmarked detection of rabbit IgG and human IgG could be visualized by observing the fluorescence microscopy images or quantitatively detected using fibre spectra of the QD-encapsulated core–shell barcodes and the labelling antibody (Fig. S6, ESI), which provides a direct and simple way for their clinical applications.


image file: c8tb00946e-f6.tif
Fig. 6 The fluorescence microscopy image of rabbit IgG and human IgG detection, (a) no target antigen existed, (b) only human IgG existed, (c) only human IgG existed and (d) both the antigens existed. The scale bar is 300 μm.

Experimental

Materials

The material used to prepare the outer phase was 2% polyvinyl alcohol (PVA, MW: 13[thin space (1/6-em)]000–23[thin space (1/6-em)]000, Sigma-Aldrich Co.). The middle oil phase was made up of silica nanoparticle-containing ethoxylated trimethylolpropane triacrylate (ETPTA, Sigma) and the photoinitiator (2-hydroxy-2-methylpropiophenone, 1% v/v). The inner aqueous phase was made up of QD-containing 30% w/w poly(ethylene glycol)diacrylate (PEG-DA, MW: 700), 1% w/w PVA, 1% w/w ethyleneoxide propyleneoxide tri-block copolymer (Pluronic F108, BASF), and 1% w/w photoinitiator (2-hydroxy-2-methylpropiophenone, Sigma). 3-Mercaptopropionic acid (MPA) modified QDs in water with different photoluminescence spectra were purchased from Suzhou Xingshuo Nanotech Co. N-(3-Dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride crystals (EDC), N-hydroxysuccinimide (NHS) and the hydrophobic reagent octadecyltrichlorosilane (OTS) were all derived from Sigma-Aldrich Co. Deionized water was used for all experiments.

Microfluidics

The microfluidic device was assembled using coaxial capillaries on a glass slide, including inner, middle and outer round capillaries and a square glass capillary. The round capillary and the seven-bore annular capillary array were obtained from World Precision Instruments, Inc. The round capillaries were used as middle and outer phase capillaries, while the seven-bore capillary was used as the inner phase. The inner and outer diameters of the round capillaries were 800 μm and 1 mm, respectively. And the square capillary with an inner diameter of 1.05 mm was purchased from VitroCom, Inc. The capillaries were tapered using a capillary puller (Sutter Instrument, P-97) and a microforge (Narishige, MF-830) to facilitate the assembly process. For the inner phases, the capillary array was stretched to reach a diameter of about 200 μm at the orifice. And the diameters of the capillaries for the middle phase and the outer phase were 300 μm and 800 μm, respectively. For the treatment of the capillary array, it was immersed in the hydrophobic reagent octadecyltrichlorosilane (OTS) and incubated for 10 min. Subsequently, the solution was blown out. These capillaries were coaxially assembled in a square capillary. The connectors of the assembled capillaries were sealed with dispensing needles and transparent epoxy resin (Devcon 5 Minute Epoxy) where necessary. The inner and middle phases flowed in the same direction via the injection capillaries in an appropriate arrangement as required. The outer phase flowed via the interstices between the square capillary and the injection capillary array, or between the square capillary and the collection capillary. Each fluid was pumped using a syringe pump (LongerPump, LSP02-2A), and was connected through a polyethylene tube (Scientific Commodities Inc., with inner and outer diameters of 0.86 mm and 1.32 mm, respectively) with a glass syringe (SGE Analytical Science). The polymerization of the double emulsion shells was done using UV-light (UVEC-4II, 365 nm, 40 W) for 5 s after collection.

Detection of antigen

For the studies on the detection of antigen, antibody (10 mM) was immobilized onto the resin shells of the barcode particles with QD-PEGDA cores in the presence of coupling reagents (EDC 5 mM, NHS 0.05 M) at room temperature for 3 hours and 4 °C overnight. After washing with PBS buffer, the QD-barcode particles were used as carriers for antigen antibody reaction. To demonstrate the multiplex capability, two kinds of particles were immobilized with anti-rabbit IgG and anti-human IgG, respectively, and incubated in the same well of antigen solution for hybridization. During this process, the tubes were kept at 37 °C. The unbound antigen was washed away with PBS buffer. Then anti-rabbit IgG and anti-human IgG tagged with fluorescein isothiocyanate (FITC) could combine with the homologous QD-barcode particles.

Characterization

A double emulsion generation process in the collection capillary of the device was observed and recorded in real time using a charged coupled device (CCD, Pike F032B ASG). The outer radius (Router), inner radius (Rinner), and shell thickness of the double emulsions were all measured by using the AOS Imaging Studio V3.4.2 software. Shell thickness was calculated from Router minus Rinner. Each set of experiments was repeated ten times and the average values were obtained. The fluorescence images of the QD-encapsulated core–shell barcodes were obtained using a fluorescence microscope (Olympus CKX41) and were captured using a CCD (Oplenic PSC603). The core–shell structures of the QD-barcode particles after polymerization of the shell were explored by SEM images using a scanning electron microscope (SEM, Hitachi S3000N).

Conclusions

In summary, we have developed a new class of QD barcode particles with core–shell structures by using a multiple-inner phase channel capillary microfluidic approach. The QD dispersed PEGDA was used as the inner phase, and the QDs were locked into the core of the barcode particles after polymerization, and thus they could maintain the dispersed status and high stability. The silica nanoparticle-dispersed ETPTA resin was employed as the shells of the barcode particles, which could not only improve the stability and biocompatibility of QD-distributed cores, but also provide functional groups for immobilization of biomolecules. Benefitting from the advanced emulsification capability of the capillary microfluidic device, a series of double emulsion templates with multiple inner droplet phases and their resultant multicomponent QD-encapsulated core–shell barcode particles could be continually generated. The generated barcode particles showed remarkable spectral coding capacity based on their distinguishable multicomponent QD cores. These features indicated that the QD-encapsulated core–shell barcode particles are ideal for multiplexing biomedical applications.

Conflicts of interest

The authors declare no competing financial interests.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Grant No. 21473029 and 51522302), the NSAF of China (Grant No. U1530260), and the Scientific Research Foundation of Southeast University.

Notes and references

  1. F. Zhang, L. Christiansen, J. Thomas, D. Pokholok, R. Jackson, N. Morrell, Y. N. Zhao, M. Wiley, E. Welch, E. Jaeger, A. Granat, S. J. Norberg, A. Halpern, M. C. Rogert, M. Ronaghi, J. Shendure, N. Gormley, K. L. Gunderson and F. J. Steemers, Nat. Biotechnol., 2017, 35, 852 CrossRef PubMed.
  2. F. Lan, J. R. Haliburton, A. Yuan and A. R. Abate, Nat. Commun., 2016, 7, 11784 CrossRef PubMed.
  3. C. Myhrvold, M. Baym, N. Hanikel, L. L. Ong, J. S. Gootenberg and P. Yin, Nat. Commun., 2017, 8, 14698 CrossRef PubMed.
  4. F. H. Hu, C. Zeng, R. Long, Y. P. Miao, L. Wei, Q. Z. Xu and W. Min, Nat. Methods, 2018, 15, 194–200 CrossRef PubMed.
  5. J. Hou, M. Z. Li and Y. L. Song, Angew. Chem., Int. Ed., 2018, 57, 2544–2553 CrossRef PubMed.
  6. L. R. Shang, F. F. Fu, Y. Cheng, H. Wang, Y. X. Liu, Y. J. Zhao and Z. Z. Gu, J. Am. Chem. Soc., 2015, 137, 15533–15539 CrossRef PubMed.
  7. G. N. Li, G. Yang, P. C. Zhang, Y. Y. Li, J. X. Meng, H. L. Liu and S. T. Wang, Small, 2015, 11, 5642–5646 CrossRef PubMed.
  8. S. Shikha, T. Salafi, J. T. Cheng and Y. Zhang, Chem. Soc. Rev., 2017, 46, 7054–7093 RSC.
  9. Y. J. Zhao, L. R. Shang, Y. Cheng and Z. Z. Gu, Acc. Chem. Res., 2014, 47, 3632–3642 CrossRef PubMed.
  10. Y. K. Leng, K. Sun, X. Y. Chen and W. W. Li, Chem. Soc. Rev., 2015, 44, 5552–5595 RSC.
  11. F. F. Fu, L. R. Shang, F. Y. Zheng, Z. Y. Chen, H. Wang, J. Wang, Z. Z. Gu and Y. J. Zhao, ACS Appl. Mater. Interfaces, 2016, 8, 13840–13848 CrossRef PubMed.
  12. Y. J. Zhao, Y. Cheng, L. R. Shang, J. Wang, Z. Y. Xie and Z. Z. Gu, Small, 2015, 11, 151–174 CrossRef PubMed.
  13. J. Kim, M. J. Biondi, J. J. Feld and W. C. W. Chan, ACS Nano, 2016, 10, 4742–4753 CrossRef PubMed.
  14. Y. K. Leng, W. J. Wu, L. Li, K. Lin, K. Sun, X. Y. Chen and W. W. Li, Adv. Funct. Mater., 2016, 26, 7581–7589 CrossRef.
  15. M. A. Boles, D. S. Ling, T. Hyeon and D. V. Talapin, Nat. Mater., 2016, 15, 141 CrossRef PubMed.
  16. Y. J. Zhao, H. C. Shum, H. S. Chen, L. L. A. Adams, Z. Z. Gu and D. A. Weitz, J. Am. Chem. Soc., 2011, 133, 8790–8793 CrossRef PubMed.
  17. S. Bai, T. Nguyen, P. Mulvaney and D. Y. Wang, Adv. Mater., 2010, 22, 3247–3250 CrossRef PubMed.
  18. S. H. Hu and X. Gao, Adv. Funct. Mater., 2010, 20, 3721–3726 CrossRef PubMed.
  19. T. Song, J. Liu and W. Li, ACS Appl. Mater. Interfaces, 2014, 6, 2745–2752 CrossRef PubMed.
  20. A. Zebibula, N. Alifu, L. Q. Xia, C. W. Sun, X. M. Yu, D. W. Xue, L. W. Liu, G. H. Li and J. Qian, Adv. Funct. Mater., 2018, 28, 1703451 CrossRef.
  21. Y. L. Zhou, P. F. Damasceno, B. S. Somashekar, M. Engel, F. L. Tian, J. Zhu, R. Huang, K. Johnson, C. McIntyre, K. Sun, M. Yang, P. F. Green, A. Ramamoorthy, S. C. Glotzer and N. A. Kotov, Nat. Commun., 2018, 9, 181 CrossRef PubMed.
  22. D. S. Zhang, Y. Jiang, H. O. Yang, Y. J. Zhu, S. J. Zhang, Y. Zhu, D. Wei, Y. Lin, P. P. Wang, Q. H. Fu, H. Xu and H. C. Gu, Adv. Funct. Mater., 2016, 26, 6146–6157 CrossRef.
  23. F. Lan, B. Demaree, N. Ahmed and A. R. Abate, Nat. Biotechnol., 2017, 35, 640 CrossRef PubMed.
  24. Y. R. Yu, F. F. Fu, L. R. Shang, Y. Cheng, Z. Z. Gu and Y. J. Zhao, Adv. Mater., 2017, 29, 1605765 CrossRef PubMed.
  25. D. Guo, Y. Li, X. Zheng, F. Y. Li, S. R. Chen, M. Z. Li, Q. Yang, H. Z. Li and Y. L. Song, J. Am. Chem. Soc., 2018, 140, 18–21 CrossRef PubMed.
  26. D. Koziej, C. Floryan, R. A. Sperling, A. J. Ehrlicher, D. Issadore, R. Westervelt and D. A. Weitz, Nanoscale, 2013, 5, 5468–5475 RSC.
  27. Y. R. Yu, L. R. Shang, W. Gao, Z. Zhao, H. Wang and Y. J. Zhao, Angew. Chem., Int. Ed., 2017, 56, 12127–12131 CrossRef PubMed.
  28. S. S. Datta, A. Abbaspourrad, E. Amstad, J. Fan, S. H. Kim, M. Romanowsky, H. C. Shum, B. J. Sun, A. S. Utada, M. Windbergs, S. B. Zhou and D. A. Weitz, Adv. Mater., 2014, 26, 2205–2218 CrossRef PubMed.
  29. J. Wang, L. Y. Sun, M. H. Zou, W. Gao, C. H. Liu, L. R. Shang, Z. Z. Gu and Y. J. Zhao, Sci. Adv., 2017, 3, e1700004 CrossRef PubMed.
  30. X. Xie, W. X. Zhang, A. Abbaspourrad, J. Ahn, A. Bader, S. Bose, A. Vegas, J. Q. Lin, J. Tao, T. Hang, H. Lee, N. Iverson, G. Bisker, L. X. Li, M. S. Strano, D. A. Weitz and D. G. Anderson, Nano Lett., 2017, 17, 2015–2020 CrossRef PubMed.
  31. L. R. Shang, Y. Cheng and Y. J. Zhao, Chem. Rev., 2017, 117, 7964–8040 CrossRef PubMed.
  32. Y. N. Li, D. Yan, F. F. Fu, Y. X. Liu, B. Zhang, J. Wang, L. R. Shang, Z. Z. Gu and Y. J. Zhao, Sci. China Mater., 2017, 60, 543–553 CrossRef.
  33. Y. Song, U. Shimanovich, T. C. T. Michaels, Q. M. Ma, J. M. Li, T. P. J. Knowles and H. C. Shum, Nat. Commun., 2016, 7, 12934 CrossRef PubMed.
  34. J. Wang, M. H. Zou, L. Y. Sun, Y. Cheng, L. R. Shang, F. F. Fu and Y. J. Zhao, Sci. China Mater., 2017, 60, 857–865 CrossRef.
  35. S. S. Lee, J. Park, Y. Seo and S. H. Kim, ACS Appl. Mater. Interfaces, 2017, 9, 17178–17185 CrossRef PubMed.
  36. T. Y. Lee, M. Ku, B. Kim, S. Lee, J. Yang and S. H. Kim, Small, 2017, 13, 1700646 CrossRef PubMed.
  37. C. B. Chang, J. N. Wilking, S. H. Kim, H. C. Shum and D. A. Weitz, Small, 2015, 11, 3954–3961 CrossRef PubMed.
  38. I. Lignos, R. Maceiczyk and A. J. deMello, Acc. Chem. Res., 2017, 50, 1248–1257 CrossRef PubMed.

Footnote

Electronic supplementary information (ESI) available. See DOI: 10.1039/c8tb00946e

This journal is © The Royal Society of Chemistry 2018