Defect engineering of highly stable lanthanide metal–organic frameworks by particle modulation for coating catalysis†

Yifa Chen‡ ^a, Shenghan Zhang‡ ^a, Fan Chen ^a, Sijia Cao ^a, Ya Cai ^a, Siqing Li ^a, Hongwei Ma ^a, Xiaojie Ma ^a, Pengfei Li *^ab, Xianqiang Huang *^c and Bo Wang *^a
aBeijing Key Laboratory of Photoelectronic/Electrophotonic Conversion Materials, Key Laboratory of Cluster Science, Ministry of Education, School of Chemistry, Beijing Institute of Technology, Beijing 100081, P. R. China. E-mail: bowang@bit.edu.cn
^bAdvanced Research Institute of Multidisciplinary Science, Beijing Institute of Technology, Beijing 100081, P. R. China
^cShandong Provincial Key Laboratory of Chemical Energy Storage and Novel Cell Technology, School of Chemistry & Chemical Engineering, Liaocheng University, Liaocheng, 252059, P. R. China

First published on 13th November 2017

---

---

Introduction

The development of heterogeneous catalysts with well-defined macroscopic shapes that avoid complicated separation and recycling steps during operation is an urgent task to achieve sustainable catalytic synthesis with high economic and energy efficiency.¹ There is great interest in finding new, highly efficient multifunctional catalysts for the exploration of heterogeneous catalysts applicable in batch or continuous-flow catalysis.² By virtue of their crystallinity, tailorability and high surface area, metal–organic frameworks (MOFs) have attracted great interest and found many important applications like gas storage/separation and catalysis.³ MOFs show the advantages of periodically distributed metal centers, ordered porosities and tunable functional groups, which make MOFs promising candidates for heterogeneous catalysis.⁴

However, the brittle nature of MOFs, where MOFs easily break down into small particles or powders, hinders their long-term usage as catalysts. The catalyst loss of MOF catalysts is inevitable when MOF catalysts are applied in fluid reactions. Pipe clogging will also occur during long reaction cycles.⁵ To alleviate this problem, general processing methods for the fabrication of MOF crystals into pellets or beads under high pressure with high energy consumption are developed.⁶ However, these pellets or beads only slow down the pulverization process and the most accessible catalyst is confined to the surface with low MOF utilization. Hence, novel methods for the fabrication of MOFs into more robust forms are highly desirable. In particular, facile and versatile MOF coatings that can easily modify the inner surfaces of pipes or containers like tubes, autoclaves and beakers into highly active batch or continuous-flow reactors are long sought after.

In order to apply MOFs as robust coatings for efficient and diverse catalytic systems, the desired properties for MOF catalysts are proposed as follows: (1) small particle sizes (preferably in the nanometer range) to provide uniform dispersion;⁷ (2) unsaturated metal sites for high performance catalysis;^3b (3) high stability that can endure harsh conditions.⁸ To date, it is still a challenge to combine most of the requirements together to make a nano-sized MOF with finely-tuned defects that are stable for industrial applications. MOFs with unsaturated/open metal sites, which generally lack chemical and/or thermal stability,⁹ still pose a challenge for size control and further processing. Moreover, the design of a suitable nano-sized MOF catalyst for batch or continuous-flow catalysis needs overall consideration: structural stability, catalytic site engineering, and more importantly, catalyst fabrication.

Herein, we report a coordination modulation method for lanthanide MOFs, and a highly efficient coating catalyst for Knoevenagel condensation based on a newly designed highly stable nano-sized Ce-MOF through defect engineering and electrospinning fabrication (Fig. 1). We design a Ce-BTB MOF (BIT-58, BIT stands for Beijing Institute of Technology) with Ce³⁺ as the metal center and 1,3,5-tris(4-carboxyphenyl)benzene (BTB) as the ligand to satisfy the above mentioned criteria. Ce³⁺ is widely used as a Lewis acid catalyst and has a versatile redox behavior with a high coordination number and large charge density, which provides the potential to keep high stability with open metal sites. The crystal size and defects of BIT-58 are controlled by a versatile coordination modulation method¹⁰ utilizing various nitrogenous heterocyclic modulators (e.g., 1-methylimidazole (1-mIM)). This method is also suitable for rare earth MOFs with distinct structures. Furthermore, a BIT-58 nanoparticle-based coating presenting high catalytic activity and excellent reusability is fabricated on the surface of a container through electrospinning for batch catalysis (Fig. 1).

Results and discussion

Structure and stability of BIT-58

BIT-58 is synthesized by a solvothermal method. Ce(NO₃)₃·6H₂O and BTB are heated in a DMF/MeOH/H₂O mixed solution at 85 °C for 24 h, which affords light brown crystals. The powder X-ray diffraction pattern (PXRD) shows that BIT-58 is isoreticular to La-BTB (Fig. S1, ESI†).¹¹ A structural model for BIT-58 is constructed by Rietveld refinement of the BIT-58 PXRD data (Fig. 2a and b, additional information is provided in the ESI S2†). The obtained BIT-58 possesses a one-dimensional channel with a pore diameter of about 1 nm. Ce in the structure is trivalent as shown by X-ray photoelectron spectroscopy and is coordinated by nine oxygen atoms as the CeO₉ cluster (Fig. S2, ESI†).¹² In the framework, two adjacent CeO₉ clusters give an edge-sharing connecting model. Each CeO₉ cluster connects with the adjacent one by two shared bridging oxygen atoms from two distinct BTB molecules and three CeO₉ clusters act as a basic unit (Fig. 2a). Besides the two shared bridging oxygen atoms, six other oxygen atoms in the cluster are from the carboxyl groups of three BTB. The last oxygen atom is from the coordinating solvent molecule. As the solvent molecule is removable, BIT-58 is an unsaturated MOF and has the potential for efficient catalysis.

Remarkably, BIT-58 exhibits exceptionally high chemical and thermal stability. A wide range of solvents are tested. For example, the integrity of BIT-58 remains unchanged in common organic solvents (e.g., chloroform) for more than three days (Fig. 2c). In addition, the BIT-58 remains intact in water at 25 °C or even at 100 °C for at least three days (Fig. 2c). Especially, BIT-58 is stable under highly acidic or basic conditions with a wide pH range from 1 to 12. Moreover, BIT-58 is stable at temperatures as high as 490 °C under N₂ and 400 °C under air as shown by variable temperature PXRD (Fig. 2d and S3, ESI†) and thermogravimetric analyses (TGA) (Fig. S4, ESI†).

Particle modulation effect with various modulators

The high stability of BIT-58 lays the foundation for further processing. We attempt to adjust the size of BIT-58 with 1-mIM. 1-mIM might serve as a monodentate ligand that competes with the coordination sites and controls the growth of MOF crystals.¹³ 1.2 mL 1-mIM is added to a 16 mL reaction mixture (volume ratio, 0.07), and the average particle size of BIT-58 is largely reduced from ∼25 μm to ∼30 nm with crystallinity unchanged (Fig. 3a, b and S5, ESI†). For convenience, the 1-mIM tuned BIT-58 nanoparticles are denoted as nano-BIT-58. Furthermore, a series of nitrogenous heterocyclic and carboxyl-containing ligands are applied to adjust the particle sizes of BIT-58. In particular, we found that most nitrogenous heterocyclic ligands can drastically decrease the crystal sizes of BIT-58, while carboxyl-containing ligands show little or no particle-size-decreasing effect under the same conditions (Fig. S6–S9 and Table S1, ESI†). Besides, the size-tuning effect of most nitrogenous heterocyclic modulators is positively correlated with their polarizability (Table S1, ESI†).¹⁴ Remarkably, nitrogenous heterocyclic modulators containing both one (e.g., pyridine or pyrrole) and two nitrogen-coordination sites (e.g., imidazole or 2-methylimidazole) present higher size-decreasing effects with larger polarizability values (Fig. S6 and Table S1, ESI†). For example, for modulators with one nitrogen-coordination site, the particle sizes of nano-BIT-58 tuned with 2,4-dimethylpyrrole (polarizability, 12.02 ų), 1,2-dimethylimidazole (11.76 ų), 1-methylimidazole (10.01 ų), pyridine (9.65 ų), and pyrrole (8.2 ų) show an increasing size range from ∼26 nm to ∼760 nm (Table S1, ESI†). Furthermore, the steric hindrance effect of the ligands also plays an important role in coordination modulation. 2,6-Lutidine and benzimidazole show weak particle size modulation ability, even though their polarizabilities (13.47 and 14.51 ų) are the highest among selected ligands with one and two nitrogen-coordination sites, respectively. This might be attributed to the steric hindrance of the two neighboring methyl groups in 2,6-lutidine and benzene in benzimidazole (Table S1, ESI†).

The crystal of nano-BIT-58 is imaged by transmission electron microscopy (TEM) and its crystal lattice is further measured by selected-area electron diffraction (SAED) analysis. The diffraction pattern is clearly visible with a spacing of 3.62 Å attributed to the (4 2 3) crystal plane (Fig. 3c). Elemental mapping analysis of nano-BIT-58 reveals that the Ce³⁺ ions in the particle are uniformly distributed (Fig. 3d). We further adjust the amount of 1-mIM added to the reaction. The average particle size shows a U-shaped curve by changing the amount of 1-mIM from 0 to 9 mL (volume ratios, 0 to 0.36) (Fig. 3g, h and S6, ESI†). The smallest particle size (about 30 nm) is achieved with 1.2 mL 1-mIM (volume ratio, 0.07). To prove the generality of this method, we further extend this method to different rare earth MOFs. In particular, the crystal sizes of Ce-NDC,^15a Ce-BTC,^15b,c Eu-BTC,^15d Dy-BTC^15d and La-BTB¹¹ with different structures are also tuned using the same method with 1-mIM as the modulator (Fig. S11 and S12, ESI†).

Characterization of the defects and catalysis investigation

The porosities of the nano-BIT-58 and BIT-58 are investigated by N₂ sorption at 77 K. In particular, nano-BIT-58 (∼30 nm) with hierarchically distributed porosity exhibits a slightly higher BET surface area of 1169 m² g⁻¹ than BIT-58 (1075 m² g⁻¹) (Fig. 3f and S9, ESI†). The micropore volume of nano-BIT-58 (0.39 cm³ g⁻¹, <2 nm) is almost the same as that of BIT-58 (0.40 cm³ g⁻¹, <2 nm) (Fig. 3f). Remarkably, the mesopore volume of nano-BIT-58 (0.35 cm³ g⁻¹, 2–40 nm) has an about seven times enhancement in contrast to BIT-58 (0.05 cm³ g⁻¹, 2–40 nm). The large enhancement of the mesopore volume might be beneficial to mass transport of the substrates in catalytic reactions, which also indicates that nano-BIT-58 is a more efficient catalyst compared with BIT-58. What is more, the chemical (stability in the pH range from 1 to 12) and thermal stabilities (stability up to 450 °C in N₂ and 390 °C in air) of nano-BIT-58 are almost unchanged compared with those of BIT-58 (Fig. S14 and S15, ESI†).

With a stable nano-sized BIT-58 in hand, the exposed defects in the MOF nanocrystals are carefully characterized by NH₃-temperature programmed desorption (NH₃-TPD) (Fig. S16, ESI†). The TPD peak at about 100 °C reflects the physical sorption of NH₃ on MOFs, and this peak is wider in nano-BIT-58 than that of BIT-58 crystals with a 1.53 times increase in desorbed NH₃ amount (nano-BIT-58, 3.92 and BIT-58, 2.56) (Fig. 3e). TPD peaks higher than 200 °C attributed to different acid sites are either broader or more intense in nano-BIT-58 compared with BIT-58. NH₃-TPD curves show that nano-BIT-58 presents a total acid site amount of 7.06 which is 10.5 times that of BIT-58 (0.67) due to the enhanced defects like more Lewis acid sites (e.g., unsaturated metal sites) and Brønsted acid sites (e.g. uncoordinated carboxyl groups) exposed.¹⁶

The largely exposed metal sites and drastically enhanced mesopore volume imply that the nano-BIT-58 is a superior catalyst.¹⁷ As a proof of concept, nano-BIT-58 with a particle size of about 30 nm exhibits higher catalytic efficiency than BIT-58 in a Knoevenagel condensation catalysis of benzaldehyde and malononitrile (Fig. 4a).¹⁸ After workup, the structural integrity of nano-BIT-58 is maintained (Fig. 4b). The conversion efficiency of nano-BIT-58 reaches 100% at 6 h, which is superior to that of BIT-58 (78%, 6 h) (entry 1, Table 1 and Fig. S17, ESI†). Moreover, we have extended the substrates to various benzaldehyde derivatives (i.e. 4-chlorobenzaldehyde, 4-bromobenzaldehyde, 4-methoxybenzaldehyde, 3,5-dimethoxybenzaldehyde, 1-naphthaldehyde and 9-anthraldehyde) with diverse molecule sizes. The catalytic performance improvement is shown by the efficiency ratio of the nano-BIT-58 catalyzed reaction to the BIT-58 catalyzed reaction, which is more significant in the catalysis of larger benzaldehyde derivatives with a similar substitution pattern. For example, 4-chlorobenzaldehyde (4.521 Å × 9.428 Å) and 4-bromobenzaldehyde (4.521 Å × 9.718 Å) have 2.9 and 1.9 times improvement in catalytic performances (entries 2 and 3, Table 1), respectively. 3,5-Dimethoxybenzaldehyde (7.248 Å × 8.074 Å) shows a 7.3 times increase compared with 3.2 times for 4-methoxybenzaldehyde (4.521 Å × 9.31 Å) (entries 4 and 5, Table 1). Furthermore, the difference in catalytic performances is more significant when the derivatives are 1-naphthaldehyde (7.240 Å × 7.417 Å) and 9-anthraldehyde (7.244 Å × 9.756 Å) with larger molecule sizes tested, in which nano-BIT-58 has 2.9 and 14.0 times improvement in conversion efficiency, respectively (entries 6 and 7, Table 1). Although the pore size distribution is broadened, nano-BIT-58 still shows substrate selectivity toward substrates with a suitable molecular size. The catalytic conversion efficiency with a molecular size smaller than 8 Å is almost quantitative compared to that with the larger reactants.

---

As the crystal size of nano-BIT-58 is controlled in the nanometer range, it affords perfect manufacturing properties. We successfully fabricated nano-BIT-58 into a film (70 wt% nano-BIT-58 in polyacrylonitrile (PAN)) with intact internal topology through electrospinning (Fig. 4e and S18, ESI†).¹⁹ The obtained film displays a nanofiber-interweaved network and nano-BIT-58 particles are uniformly dispersed on the nanofibers (about 200 nm) (Fig. 4c). The porosity of nano-BIT-58 in the film is accessible as demonstrated by the N₂ sorption test with a S_BET of 409 m² g⁻¹ (Fig. S19, ESI†). Besides, the film can hold 0.05 MPa pressure in a stretching resistance test (Fig. S20, ESI†). With high mechanical strength and readily accessible pores, this film is directly used in the condensation reaction. It presents 100% catalytic conversion efficiency in just 6 h and is easily recycled at least three times without losing its activity and stretching robustness (Fig. 4d, e, S20 and S21, ESI†). It is noteworthy that we successfully manufactured a large-scale film (14 cm × 15 cm) through position-controlled electrospinning (Fig. S22, ESI†). Furthermore, a portable electrospinning device is used to directly deposit this film on the surface of various substrates (i.e. beaker, aluminum foil, plastic mesh, nickel foam, glass cloth, non-woven fabric and cardboard) (Fig. 4f and S23, ESI†). The wet nanofibers after electrospinning with trace amounts of solvents are attached to the surface of the substrate with interfacial interactions.²⁰ Taking a beaker for example, nano-BIT-58/PAN as a powerful catalyst coating directly turns a beaker with no activity into a highly active one and exhibits a 95% catalytic conversion efficiency in the reaction (Fig. 4f). After the catalysis, the product is easily separated by pouring out and the coated beaker can be directly used in the next cycle without loss of activity.

Conclusions

In summary, we report a general coordination modulation method and explore a newly designed coating catalyst through defect engineering and fabrication of a highly stable Ce-MOF (BIT-58). The particle sizes of BIT-58 are well tuned in a large size range and the strategy is applicable to various rare earth MOFs. By decreasing the particle sizes from micrometers to nanometers, the defects of nano-BIT-58 are drastically exposed (10 times increased total acid site amount; 7 times higher mesopore volume) and the catalytic performance is largely increased (7 or even 14 times increase in conversion efficiency) compared with BIT-58. Nano-BIT-58 can be further fabricated into films or powerful coatings (suitable for various substrates) with excellent catalytic activity and reusability for batch catalysis. This general strategy, accompanied by the exploration of new MOFs with high stability, would herald an era in promoting the industrial application of MOFs.

Conflicts of interest

There are no conflicts to declare.

Author contributions

B. Wang, P. Li and X. Huang designed the experiments. Y. Chen, S. Zhang and F. Chen conducted the experiments. S. Cao, Y. Cai and S. Li contributed to the syntheses of MOFs. H. Ma and X. Ma helped the work of data refinement. B. Wang, Y. Chen, S. Zhang and P. Li wrote the manuscript.

Acknowledgements

This work was financially supported by the 973 Program 2013CB834704; the Provincial Key Project of China (Grant No. 7131253); the National Natural Science Foundation of China (Grant No. 21471018, 21404010, 21201018, and 21490570); the 1000 Plan (Youth).

Notes and references

1. A. Corma, H. Garcia and F. X. L. I. Xamena, Chem. Rev., 2010, 110, 4606–4655 CrossRef CAS PubMed .
2. (a) Y. F. Chen, X. Q. Huang, S. H. Zhang, S. Q. Li, S. J. Cao, X. K. Pei, J. W. Zhou, X. Feng and B. Wang, J. Am. Chem. Soc., 2016, 138, 10810–10813 CrossRef CAS PubMed ; (b) A. Sachse, R. Ameloot, B. Coq, F. Fajula, B. Coasne, D. De Vos and A. Galarneau, Chem. Commun., 2012, 48, 4749–4751 RSC .
3. (a) J. R. Li, R. J. Kuppler and H. C. Zhou, Chem. Soc. Rev., 2009, 38, 1477–1504 RSC ; (b) J. Lee, O. K. Farha, J. Roberts, K. A. Scheidt, S. T. Nguyen and J. T. Hupp, Chem. Soc. Rev., 2009, 38, 1450–1459 RSC ; (c) S. M. J. Rogge, A. Bavykina, J. Hajek, H. Garcia, A. I. Olivos-Suarez, A. Sepúlveda-Escribano, A. Vimont, G. Clet, P. Bazin, F. Kapteijn, M. Daturi, E. V. Ramos-Fernandez, F. X. L. I. Xamena, V. Van Speybroeck and J. Gascon, Chem. Soc. Rev., 2017, 46, 3134–3184 RSC ; (d) Q. H. Yang, Q. Xu and H. L. Jiang, Chem. Soc. Rev., 2017, 46, 4774–4808 RSC ; (e) J. W. Zhou, R. Li, X. X. Fan, Y. F. Chen, R. D. Han, W. Li, J. Zheng, B. Wang and X. G. Li, Energy Environ. Sci., 2014, 7, 2715–2724 RSC ; (f) Z. J. Zhang, Z. Z. Yao, S. C. Xiang and B. L. Chen, Energy Environ. Sci., 2014, 7, 2868–2899 RSC ; (g) Y. Q. Lan, H. L. Jiang, S. L. Li and Q. Xu, Adv. Mater., 2011, 23, 5015–5020 CrossRef CAS PubMed ; (h) A. Cadiau, Y. Belmabkhout, K. Adil, P. M. Bhatt, R. S. Pillai, A. Shkurenko, C. Martineau-Corcos, G. Maurin and M. Eddaoudi, Science, 2017, 356, 731–735 CrossRef CAS PubMed .
4. (a) L. Q. Ma, C. Abney and W. B. Lin, Chem. Soc. Rev., 2009, 38, 1248–1256 RSC ; (b) L. Zhu, X. Q. Liu, H. L. Jiang and L. B. Sun, Chem. Rev., 2017, 117, 8129–8176 CrossRef CAS PubMed .
5. A. U. Czaja, N. Trukhan and U. Müller, Chem. Soc. Rev., 2009, 38, 1284–1293 RSC .
6. L. D. O'Neill, H. F. Zhang and D. Bradshaw, J. Mater. Chem., 2010, 20, 5720–5726 RSC .
7. (a) W. J. Rieter, K. M. L. Taylor, H. Y. An, W. L. Lin and W. B. Lin, J. Am. Chem. Soc., 2006, 128, 9024–9025 CrossRef CAS PubMed ; (b) H. Bunzen, M. Grzywa, M. Hambach, S. Spirkl and D. Volkmer, Cryst. Growth Des., 2016, 16, 3190–3197 CrossRef CAS .
8. (a) C. H. Wang, X. L. Liu, N. K. Demir, J. P. Chen and K. Li, Chem. Soc. Rev., 2016, 45, 5107–5134 RSC ; (b) T. Devic, C. Serre, N. Audebrand, J. Marrot and G. Férey, J. Am. Chem. Soc., 2005, 127, 12788–12789 CAS .
9. (a) N. Al-Janabi, P. Hill, L. Torrente-Murciano, A. Garforth, P. Gorgojo, F. Siperstein and X. L. Fan, Chem. Eng. J., 2015, 281, 669–677 CrossRef CAS ; (b) Y. L. Hu, C. Y. Song, J. Liao, Z. L. Huang and G. K. Li, J. Chromatogr. A, 2013, 1294, 17–24 CrossRef CAS PubMed .
10. T. Tsuruoka, S. Furukawa, Y. Takashima, K. Yoshida, S. Isoda and S. Kitagawa, Angew. Chem., Int. Ed., 2009, 48, 4739–4743 CrossRef CAS PubMed .
11. J. G. Duan, M. Higuchi, S. Horike, M. L. Foo, K. P. Rao, Y. Inubushi, T. Fukushima and S. Kitagawa, Adv. Funct. Mater., 2013, 23, 3525–3530 CrossRef CAS .
12. F. Larachi, J. Pierre, A. Adnot and A. Bernis, Appl. Surf. Sci., 2002, 195, 236–250 CrossRef CAS .
13. J. Cravillon, R. Nayuk, S. Springer, A. Feldhoff, K. Huber and M. Wiebcke, Chem. Mater., 2011, 23, 2130–2141 CrossRef CAS .
14. (a) K. J. Miller, J. Am. Chem. Soc., 1990, 112, 8533–8542 CrossRef CAS ; (b) T. N. Olney, N. M. Cann, G. Cooper and C. E. Brion, Chem. Phys., 1997, 223, 59–98 CrossRef CAS .
15. (a) Z. Wang, C. M. Jin, T. Shao, Y. Z. Li, K. L. Zhang, H. T. Zhang and X. Z. You, Inorg. Chem. Commun., 2002, 5, 642–648 CrossRef CAS ; (b) M. M. Peng, M. Ganesh, R. Vinodh, M. Palanichamy and H. T. Jang, Arabian J. Chem.,  DOI:10.1016/j.arabjc.2014.11.024; (c) J. H. Luo, H. W. Xu, Y. Liu, Y. S. Zhao, L. L. Daemen, C. Brown, T. V. Timofeeva, S. Q. Ma and H. C. Zhou, J. Am. Chem. Soc., 2008, 130, 9626–9627 CrossRef CAS PubMed ; (d) B. L. Chen, Y. Yang, F. Zapata, G. N. Lin, G. D. Qian and E. B. Lobkovsky, Adv. Mater., 2007, 19, 1693–1696 CrossRef CAS .
16. (a) H. X. Jiang, Q. Y. Wang, H. Q. Wang, Y. F. Chen and M. H. Zhang, ACS Appl. Mater. Interfaces, 2016, 8, 26817–26826 CrossRef CAS PubMed ; (b) Y. Y. Pan, B. Z. Yuan, Y. W. Li and D. H. He, Chem. Commun., 2010, 46, 2280–2282 RSC ; (c) Q. Y. Li, S. Jiang, S. F. Ji, D. Shi and H. K. Li, J. Porous Mater., 2015, 22, 1205–1214 CrossRef CAS ; (d) J. C. Jiang and O. M. Yaghi, Chem. Rev., 2015, 115, 6966–6997 CrossRef CAS PubMed .
17. (a) Z. L. Fang, B. Bueken, D. E. De Vos and R. A. Fischer, Angew. Chem., Int. Ed., 2015, 54, 7234–7254 CrossRef CAS PubMed ; (b) F. Vermoortele, B. Bueken, G. Le Bars, B. Van de Voorde, M. Vandichel, K. Houthoofd, A. Vimont, M. Daturi, M. Waroquier, V. Van Speybroeck, C. Kirschhock and D. E. De Vos, J. Am. Chem. Soc., 2013, 135, 11465–11468 CrossRef CAS PubMed .
18. W. Jiang, J. Yang, Y. Y. Liu, S. Y. Song and J. F. Ma, Inorg. Chem., 2017, 56, 3036–3043 CrossRef CAS PubMed .
19. (a) R. Ostermann, J. Cravillon, C. Weidmann, M. Wiebcke and B. M. Smarsly, Chem. Commun., 2011, 47, 442–444 RSC ; (b) W. E. Teo and S. Ramakrishna, Nanotechnology, 2006, 17, R89–R106 CrossRef CAS PubMed ; (c) Y. Y. Zhao, Z. M. Zhang, L. M. Yu and Q. W. Tang, Synth. Met., 2016, 212, 84–90 CrossRef CAS .
20. J. W. Xu, C. Liu, P. C. Hsu, K. Liu, R. F. Zhang, Y. Y. Liu and Y. Cui, Nano Lett., 2016, 16, 1270–1275 CrossRef CAS PubMed .

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c7ta09036f

‡ These authors contributed equally to this work.

This journal is © The Royal Society of Chemistry 2018