Sol–gel synthesis and electrochemical properties of c-axis oriented LiCoO₂ for lithium-ion batteries

Abstract

To improve the performance of LiCoO₂ as a cathode for lithium-ion batteries, the electrochemical properties of c-axis oriented LiCoO₂ were studied. LiCoO₂ samples with controllable intensity ratios of peak (003) to peak (104) were synthesized via a resorcinol–formaldehyde sol–gel method, followed by an air-controlled high temperature treatment. Electrochemical measurements showed that LiCoO₂ with a low degree of c-axis orientation exhibited better performance than LiCoO₂ with a high degree of c-axis orientation. X-ray absorption spectroscopy was used to reveal the texture–property relationship between the different products. The variances could be attributed to the different electrochemical active sites created and the diffusion lengths for the lithium ions related to c-axis orientation.

---

Introduction

As a solution to the worldwide energy crisis, lithium ion battery (LIB) technology has been well developed and widely used for portable electronic devices in the past twenty years. However, the performance requirements are higher for the application of LIBs in high-power transportation systems such as hybrid electric vehicles (HEVs) or electric vehicles (EVs).^1,2 In current LIB technology, the capacity as well as the transportation of Li⁺ is mainly determined by the cathode material, which makes the developments of cathode materials extremely crucial.^3,4 LiCoO₂ was the first commercially successful cathode for LIB, and it still represents more than 90% of the cathode material market today.² To improve the performance of LiCoO₂, in addition to transition metal doping,⁵ new structure design^6–8 has proved to be an effective method in recent years. Crystalline LiCoO₂ adopts the layered α-NaFeO₂-type structure (space group R3̄m) with lithium and cobalt ions ordered in the octahedral sites of alternating (111) planes, which can be typically described as a hexagonal unit cell with a_hex = 2.816 Å and c_hex = 14.08 Å.⁹ The electrochemical lithium insertion/removal behaviour of layered LiCoO₂ strongly depends on the orientations of the crystal due to its anisotropic crystallographic structure.^10,11 LiCoO₂ is apt to crystallize with c-axis orientation because the (003) plane has the lowest surface energy with the highest atomic density of all the planes.^12–14 Thus, it is critical to investigate the effects of c-axis orientation on the electrochemical performance of LiCoO₂.

The vast majority of reports on c-axis oriented LiCoO₂ mainly focus on LiCoO₂ thin film electrodes for all-solid-state lithium microbatteries.^10,13–15 In the case of LiCoO₂ composite powder electrodes, some early studies^16,17 reported c-axis oriented LiCoO₂ crystals preferentially dominated by (003) planes, and explained that the stronger (003) intensity led to a lower degree of cation disorder. Since cation mixing is more likely to occur in LiNiO₂,¹⁸ LiNi_0.5Mn_0.5O₂,^19,20 and Li–Co–Ni–Mn–O layered compounds,^21,22 more evidence is needed to confirm the hypothesis above. Recently, there have also been reports of LiCoO₂ nanoplates⁴ and micro-sized flake-like LiCoO₂ particles²³ with preferentially exposed (001) planes. However, the effects of the degree of c-axis orientation on the electrochemical behaviour of LiCoO₂ have not yet been investigated.

In the present work, we synthesized LiCoO₂ with different degrees of c-axis orientation via a facile resorcinol–formaldehyde (RF) sol–gel method,^24–26 followed by an air-controlled calcination process. The X-ray absorption fine structure (XAFS) technique, including X-ray absorption near-edge structure (XANES) and extended X-ray absorption fine structure (EXAFS), was then used to reveal the texture–property relationship of c-axis oriented LiCoO₂ cathode materials, and the effects of the degree of texture on the electrochemical performance of the as-prepared products were discussed.

Experimental

Preparation of materials

In a typical synthesis, 0.005 mol Co(CH₃COO)₂·4H₂O and LiCH₃COO·2H₂O containing a 5 mol% excess of Li were dissolved in 25 ml of anhydrous alcohol which included 0.025 mol resorcinol and 0.05 mol formaldehyde (38.5% in water, methanol stabilized), yielding a light pink suspension. After 3 h of vigorous agitation at room temperature, the homogeneous solution was heated to 60 °C in a water bath under magnetic stirring until viscous, and then dried completely in an electric oven at 90 °C. Finally, the gels were placed in alumina crucibles (30 ml, Al₂O₃ ≥ 95%) and calcined at 900 °C for 2 h in a muffle furnace (Zhonghuan, China) both covered and uncovered in air to obtain fine powders. All chemical reagents used in this experiment were of analytical grade and were used without further purification.

Characterization of materials

The phases and structures of the as-prepared products were characterized by X-ray powder diffraction (XRD) using a Rigaku Dmax2200 X-ray diffractometer with Cu Kα radiation (λ = 1.5406 Å) at 40 kV and 40 mA. Data were recorded in the range of 10° to 70° with a scan rate of 6° min⁻¹. Morphological studies were conducted using a field-emission gun environmental scanning electron microscope (Quanta 250 FEG). Elemental maps (Fig. S3 and S4, ESI†) were obtained by energy dispersive X-ray spectroscopy (FESEM/EDS, JSM-7500F). Transmission electron microscopy (TEM) studies were performed using a JEOL JEM-2100F microscope. The Co K-edge X-ray absorption fine structure (XAFS) spectroscopy was measured at the 1W1B-XAFS beam line at the Beijing Synchrotron Radiation Facility (BSRF), China. The storage ring of BSRF was operated at 2.5 GeV with a maximum current of 200 mA. A Si (111) double-crystal monochromator was used to monochromatize the radiation. The XAFS signals were collected in transmission mode at room temperature. The XAFS spectra were analyzed using Athena and Artemis from the IFEFFIT 1.2.11 software package,²⁷ and were then fitted in R-space with theoretical models based on the crystal structure of LiCoO₂ using FEFF6.²⁸ The k values used to fit the Co K-edge EXAFS ranged from 2 to 14.35 Å⁻¹ to minimize noise.

Electrode fabrication and electrochemical tests

The working electrodes were prepared by mixing 80 wt% active material, 10 wt% acetylene black, and 10 wt% polyvinylidene fluoride (PVDF). The mixture was then dissolved in N-methyl-2-pyrrolidinone (NMP), coated on pure aluminum disks (14 mm in diameter) and dried at 120 °C for 10 h. The working electrodes were compressed at 3 MPa before assembly. Electrochemical measurements were carried out via CR2032 type coin cells with metallic lithium as the counter/reference electrode, Celgard 2325 as the separator, and a solution of 1 M LiPF₆ dissolved in a mixture of ethylene carbonate (EC), dimethyl carbonate (DMC), and ethyl methyl carbonate (EMC) (1 : 1 : 1, v/v/v) as the electrolyte. The cells were then assembled in an argon-filled glove box (Dellix, China). The galvanostatic discharge–charge tests were performed on a Land CT2001A battery testing system (Jinnuo, China) at specific current densities from 3.0 to 4.2 V (vs. Li/Li⁺) at room temperature. The electrochemical impedance measurements were conducted on a CHI660D electrochemical workstation (Chenhua, China) at an AC voltage of 5 mV amplitude in the range of 100 kHz to 0.01 Hz.

Results and discussion

Fig. 1 shows the powder XRD patterns of the products prepared in covered and uncovered crucibles, respectively. The patterns can be indexed to the standard pattern of LiCoO₂ (JCPDS card no. 75-0532). Both LiCoO₂ samples show a remarkably strong peak (003) and a relatively weak peak (104), indicating a preferred c-axis orientation (texture). The intensity ratios of the (003) peak to the (104) peak (I₀₀₃/I₁₀₄), which can be used as an indication of the degree of c-axis orientation,¹⁷ are 4.0 and 16.5. Hereafter, the LiCoO₂ samples with I₀₀₃/I₁₀₄ values 4.0 and 16.5 are denoted as l-LCO (LiCoO₂ with low degree of c-axis orientation) and h-LCO (LiCoO₂ with high degree of c-axis orientation), respectively. A controlled experiment was conducted to show how the degree of c-axis orientation could be adjusted by this simple method (Fig. S1, ESI†). It appeared that the LiCoO₂ samples prepared in covered crucibles showed a much higher degree of c-axis orientation.

Fig. 1 XRD patterns of LiCoO₂ prepared in (a) covered and (b) uncovered crucibles.

Fig. 2 shows the SEM images of both l-LCO and h-LCO. The l-LCO sample is composed of several individual irregular particles (blue dashed line) that are fused together (Fig. 2a). Apparently, the whole polycrystalline particles are somewhat fragmental, since the texture of each part varies. Contrastingly, the grains are intact in h-LCO, with clear laminar morphology (yellow dashed line, Fig. 2b). A layered structure can be clearly observed in inset no. 1 of Fig. 2c. The frontal plane could be the (003) plane. In Fig. 2d, the frontal plane is confirmed as the (003) plane, which is perpendicular to both the set of (110) planes with a crossing lattice interdistance of 0.14 nm and the set of (100) planes with a lattice interdistance of 0.25 nm.¹² The morphological analysis was in good agreement with the XRD results, where h-LCO showed a higher I₀₀₃/I₁₀₄ value than l-LCO.

Fig. 2 SEM images of (a) l-LCO and (b) h-LCO; (c) TEM images of an individual h-LCO particle; (d) HRTEM images of inset no. 2 in (c).

The difference in the degree of c-axis orientation could possibly be attributed to controlled air flow rates during heat treatment. It is generally accepted that the equilibrium form of a crystal tends to possess a minimal total surface energy.²⁹ During the calcination, the driving force of the atomic movement was derived from the tendency to minimize the surface energy. As shown in Fig. 3, when the crucible was covered (with only a tiny gap to allow air through), the flow rate of air was dramatically reduced, which caused a large amount of carbonaceous residues to remain due to incomplete combustion of the RF resin. After that, these residues served as nucleation centers for LiCoO₂ during the calcination, and finally completely transformed to volatile products. LiCoO₂ crystallized on the surface of the residues and facilitated the formation of grains with exposed (003) planes, which are quite stable, having the lowest surface energy.^12–14 In this way, h-LCO with exposed (003) planes was obtained. Conversely, when the crucible was uncovered, l-LCO was prepared with little assistance from residue nucleation centers during the calcination process.

Fig. 3 Schematic illustration for the formation process of LiCoO₂ with different degrees of c-axis orientation.

Next, the electrochemical performances of both products were measured (Fig. 4). Fig. 4a shows typical charge–discharge curves at the 1st, 10th and 50th cycles, measured at a current rate of 0.1 C in the range of 3.0–4.2 V. The initial discharge capacities of l-LCO and h-LCO at 0.1 C were 148.1 and 120.5 mA h g⁻¹. After 50 cycles, the discharge capacities remained at 145.7 and 90.8 mA h g⁻¹ (corresponding to capacity retentions of 98.4% and 75.4%), respectively. It seemed that the reversible capacity of h-LCO faded faster than that of l-LCO. Fig. 4b shows the cycling performances and Coulombic efficiencies for l-LCO and h-LCO. The Coulombic efficiencies at the first cycle are relatively low for l-LCO (88.4%) and h-LCO (86.6%), which could be attributed to the high charge capacity during the formation of the solid electrolyte interphase (SEI) layer.³⁰ The average Coulombic efficiency of l-LCO from the 2nd to the 50th cycle is 97.2%, a little lower than that of h-LCO (98.4%). Fig. 4c illustrates the discharge curves at different rates. l-LCO and h-LCO delivered rate capacities of 97.4 and 40.5 mA h g⁻¹ at 2 C, which means that 62.1% and 34.1% of the capacities at 0.1 C were retained. The inset in Fig. 4c shows the electrochemical performances of l-LCO, m-LCO (LiCoO₂ with a medium degree of c-axis orientation, Fig. S2, ESI†) and h-LCO as a function of the I₀₀₃/I₁₀₄ value. Clearly, both the cycling and rate performances degraded with increasing I₀₀₃/I₁₀₄ value, i.e. the degree of c-axis orientation. The impedance spectra of the materials are shown in Fig. 4d, consisting of a semicircle (charge-transfer resistance) and an inclined line (Warburg impedance).³¹ The result of the fitting fits well with the experimental data. The charge-transfer resistance R_ct of h-LCO is 54 Ω, higher than that of l-LCO (38 Ω). As shown above, it was found that LiCoO₂ with a low degree of c-axis orientation exhibited better electrochemical performance than that with a high degree of c-axis orientation.

Fig. 4 (a) The charge–discharge curves at the 1st, 10th and 50th cycles for l-LCO and h-LCO; (b) cycling performances and Coulombic efficiency for l-LCO and h-LCO at 0.1 C; (c) rate capacities of l-LCO and h-LCO at different rates stepwise from 0.1 to 2 C (inset: specific capacity vs. intensity ratio I₀₀₃/I₁₀₄); (d) the impedance spectra (both experimental data and fitted curves) of l-LCO and h-LCO (inset: equivalent circuit used in fitting the impedance data, R_s, series resistance; R_ct, charge-transfer resistance; Z_w, Warburg impedance; CPE, constant phase element).

X-ray absorption fine structure (XAFS) spectra were acquired to reveal the structure–performance relationships, as such spectra can provide information on the local environment around the center atoms.^32–35 Fig. 5a shows the normalized XANES spectra at the Co K-edge of l-LCO and h-LCO with three characteristic peaks, A, B and C. The pre-edge absorption peak A corresponds to the electric dipole-forbidden transition of the 1s electron to the unoccupied 3d orbital in Co³⁺. The B and C peaks represent the dipole-allowed 1s to 4p transitions with and without a shakedown process that originate from a ligand to metal charge transfer.^36,37 The difference in the position of peak C between these products is negligible, which indicates that there is no difference in the cobalt oxidation state, and cobalt is located in the octahedral sites.³⁸

Fig. 5 (a) The normalized XANES spectra at Co K-edge of l-LCO and h-LCO; (b) Fourier transforms of Co K-edge k³-weighted EXAFS spectra of l-LCO and h-LCO, shown with both experimental and fitting curves.

Fourier transforms (FTs) of Co K-edge k³-weighted EXAFS spectra of l-LCO and h-LCO are demonstrated in Fig. 5b. The FT peaks at around 1.9 Å and 2.8 Å correspond to single scattering contributions from the nearest O atoms and Co atoms.^35,36 Compared with cobalt, the contribution of the lithium coordination shell can hardly be seen due to the small scattering factor of the Li atom. As listed in Table 1, the structural parameters of the interatomic distances R and the Debye–Waller factors σ² for the first two coordination shells were obtained from nonlinear least squares fitting. The Debye–Waller factor σ² corresponds to the mean square relative atomic displacement of the interatomic distance R for each bonding pair due to local disorder. The bond lengths R of both the Co–O and Co–Co shells in h-LCO are slightly shorter than those in l-LCO, which is indicative that there is a slight structural contraction in h-LCO without significant changes in the R3̄m structure.³⁸ The Debye–Waller factors for l-LCO (σ² = 0.00337 Ų and 0.00301 Ų for Co–O and Co–Co) are larger than those for h-LCO (σ² = 0.00312 Ų and 0.00271 Ų for Co–O and Co–Co), consistent with the greater amount of disorder³⁵ in LiCoO₂, which has a low degree of c-axis orientation.

Table 1 Structural parameters of the curve fitting results for Co K-edge EXAFS^a

Sample	Shell	N	R (Å)	σ^2 (Å^2)
a N is the coordination number; R is the interatomic distance; σ^2 is the Debye–Waller factor.
l-LCO	Co–O	6	1.922	0.00337
	Co–Co	6	2.820	0.00301
h-LCO	Co–O	6	1.920	0.00312
	Co–Co	6	2.817	0.00271

---

Combined with the scanning electron microscope morphological analysis (Fig. 2), these results indicate that the CoO₂-sheets in h-LCO can be considered to be intact and ordered; moreover, the sheets are distorted and fragmental in l-LCO. These differences in structure could possibly have a significant impact on the electrochemical properties of LiCoO₂ as a cathode material for Li-ion batteries. As illustrated in Fig. 6, with a large amount of O²⁻ and Co³⁺ ions blocking the pathway, Li ions can only move in two-dimensional directions between the CoO₂ slabs.^4,10,13,39 In addition, the diffusion time of Li⁺ in LiCoO₂ is proportional to the square of the diffusion length.^40,41 With fragmental CoO₂-sheets, l-LCO (Fig. 6a) provides more electrochemical active sites and shorter diffusion channels for Li⁺ to migrate during the insertion/removal process than h-LCO (Fig. 6b). The lower charge-transfer resistance R_ct could be attributed to the anisotropy in electron conductivity, which is much better along the a–b-axis than along the c-axis.¹¹ This could be a possible explanation for the fact that l-LCO presents relatively better electrochemical performance, including better rate performance and cyclability, than h-LCO. On the other hand, the higher average Coulombic efficiency of h-LCO may indicate superior reversibility between well-ordered CoO₂ sheets.

Fig. 6 Schematic illustration of the diffusion of lithium ions in LiCoO₂ with (a) low and (b) high degrees of c-axis orientation.

Conclusions

We investigated LiCoO₂ with different degrees of c-axis orientation to analyse the relationship between its texture and electrochemical properties in composite electrodes. LiCoO₂ samples with different I₀₀₃/I₁₀₄ ratios were prepared via a sol–gel method followed by air-controlled high temperature treatment. Electrochemical measurements show that LiCoO₂ with a low degree of texture exhibited better performance: l-LCO and h-LCO exhibited rate capacities of 97.4 and 40.5 mA h g⁻¹ at 2 C, corresponding to capacity retentions of 62.1% and 34.1% compared to those at 0.1 C, respectively. As confirmed by crystallographic structure analysis and XAFS results, LiCoO₂ with a low degree of c-axis orientation creates more electrochemical active sites and shorter diffusion distances for lithium ions, which is favourable for the electrochemical performance of LiCoO₂. Moreover, it should be noted that a high degree of c-axis orientation could possibly indicate particles with intact layered structures. To achieve better performance, oriented growth control would be an important strategy in the structure design of layered electrode materials.

Acknowledgements

This work was financially supported by the National Basic Research Program of China (2013CB934004) and the National Natural Science Foundation of China (11079002).

Notes and references

1. Y. X. Tang, Y. Y. Zhang, J. Y. Deng, D. P. Qi, W. R. Leow, J. Q. Wei, S. Y. Yin, Z. L. Dong, R. Yazami, Z. Chen and X. D. Chen, Angew. Chem., Int. Ed., 2014, 53, 13488–13492.
2. X. L. Xiao, X. F. Liu, L. Wang, H. Zhao, Z. B. Hu, X. M. He and Y. D. Li, Nano Res., 2012, 5, 395–401.
3. J. H. Yan, X. B. Liu and B. Y. Li, RSC Adv., 2014, 4, 63268.
4. B. Xu, D. Qian, Z. Y. Wang and Y. S. Meng, Mater. Sci. Eng., R, 2012, 73, 51–65.
5. J. Liang, D. H. Wu, M. Hu, Y. Tian, J. P. Wei and Z. Zhou, Electrochim. Acta, 2014, 146, 784–791.
6. M. Okubo, E. Hosono, J. Kim, M. Enomoto, N. Kojima, T. Kudo, H. S. Zhou and I. Honma, J. Am. Chem. Soc., 2007, 129, 7444–7452.
7. H. L. Chen and C. P. Grey, Adv. Mater., 2008, 20, 2206–2210.
8. D. S. Wang, X. L. Ma, Y. G. Wang, L. Wang, Z. Y. Wang, W. Zheng, X. M. He, J. Li, Q. Peng and Y. D. Li, Nano Res., 2010, 3, 1–7.
9. K. Mizushima, P. C. Jones, P. J. Wiseman and J. B. Goodenough, Mater. Res. Bull., 1980, 15, 783–789.
10. Y. Iriyama, M. Inaba, T. Abe and Z. Ogumi, J. Power Sources, 2001, 94, 175–182.
11. H. T. Zhang, P. J. Baker and P. S. Grant, J. Am. Ceram. Soc., 2010, 93, 1856–1859.
12. G. Z. Wei, X. Lu, F. S. Ke, L. Huang, J. T. Li, Z. X. Wang, Z. Y. Zhou and S. G. Sun, Adv. Mater., 2010, 22, 4364–4367.
13. J. B. Bates, N. J. Dudney, B. J. Neudecker, F. X. Hart, H. P. Jun and S. A. Hackney, J. Electrochem. Soc., 2000, 147, 59–70.
14. Y. Yoon, C. Park, J. Kim and D. Shin, J. Power Sources, 2013, 226, 186–190.
15. J. Xie, N. Imanishi, T. Matsumura, A. Hirano, Y. Takeda and O. Yamamoto, Solid State Ionics, 2008, 179, 362–370.
16. M. Yoshio, H. Tanaka, K. Tominaga and H. Noguchi, J. Power Sources, 1992, 40, 347–353.
17. W. P. Tang, H. Kanoh and K. Ooi, Electrochem. Solid-State Lett., 1998, 1, 145–146.
18. T. Ohzuku, A. Ueda, M. Nagayama, Y. Iwakoshi and H. Komori, Electrochim. Acta, 1993, 38, 1159–1167.
19. J. Bréger, N. Dupré, P. J. Chupas, P. L. Lee, T. Proffen, J. B. Parise and C. P. Grey, J. Am. Chem. Soc., 2005, 127, 7529–7537.
20. J. Bréger, Y. S. Meng, Y. Hinuma, S. Kumar, K. Kang, Y. Shao-Horn, G. Ceder and C. P. Grey, Chem. Mater., 2006, 18, 4768–4781.
21. C. H. Lei, J. Bareño, J. G. Wen, I. Petrov, S. H. Kang and D. P. Abraham, J. Power Sources, 2008, 178, 422–433.
22. B. Xu, C. R. Fell, M. F. Chi and Y. S. Meng, Energy Environ. Sci., 2011, 4, 2223–2233.
23. T. Wei, R. Zeng, Y. M. Sun, Y. H. Huang and K. Huang, Chem. Commun., 2014, 50, 1962–1964.
24. K. M. Shaju and P. G. Bruce, Adv. Mater., 2006, 18, 2330–2334.
25. H. M. Yang, X. J. Cui, Y. Q. Deng and F. Shi, RSC Adv., 2014, 4, 59754.
26. Y. Kong, X. Shen, S. Cui and M. H. Fan, RSC Adv., 2014, 4, 64193.
27. B. Ravel and M. Newville, J. Synchrotron Radiat., 2005, 12, 537–541.
28. S. I. Zabinsky, J. J. Rehr, A. Ankudinov, R. C. Albers and M. J. Eller, Phys. Rev. B: Condens. Matter Mater. Phys., 1995, 52, 2995–3009.
29. A. F. Wells, Structural Inorganic Chemistry, Clarendon Press, Oxford, 5th edn, 1984.
30. B. R. Wu, Y. H. Ren, D. B. Mu, X. J. Liu, G. C. Yang and F. Wu, RSC Adv., 2014, 4, 10196.
31. Y. Mizuno, E. Hosono, T. Saito, M. Okubo, D. Nishio-Hamane, K. Oh-ishi, T. Kudo and H. S. Zhou, J. Phys. Chem. C, 2012, 116, 10774–10780.
32. J. Bareño, M. Balasubramanian, S. H. Kang, J. G. Wen, C. H. Lei, S. V. Pol, I. Petrov and D. P. Abraham, Chem. Mater., 2011, 23, 2039–2050.
33. S. Permien, H. Hain, M. Scheuermann, S. Mangold, V. Mereacre, A. K. Powell, S. Indris, U. Schürmann, L. Kienle, V. Duppel, S. Harm and W. Bensch, RSC Adv., 2013, 3, 23001.
34. L. Simonelli, N. L. Saini, M. M. Sala, M. Okubo, I. Honma, T. Mizokawa and G. Monaco, Appl. Phys. Lett., 2013, 103, 083111.
35. L. Maugeri, L. Simonelli, A. Iadecola, B. Joseph, M. Okubo, I. Honma, H. Wadati, T. Mizokawa and N. L. Saini, J. Power Sources, 2013, 229, 272–276.
36. T. Nedoseykina, S. S. Kim and Y. Nitta, Electrochim. Acta, 2006, 52, 1467–1471.
37. T. Okumura, Y. Yamaguchi, M. Shikano and H. Kobayashi, J. Mater. Chem., 2012, 22, 17340–17348.
38. A. V. Chadwick, S. L. P. Savin, R. Alcántara, D. F. Lisbona, P. Lavela, G. F. Ortiz and J. L. Tirado, ChemPhysChem, 2006, 7, 1086–1091.
39. S. I. Zabinsky, J. J. Rehr, A. Ankudinov, R. C. Albers and M. J. Eller, Phys. Rev. B: Condens. Matter Mater. Phys., 1995, 52, 2995–3009.
40. P. G. Bruce, B. Scrosati and J. M. Tarascon, Angew. Chem., Int. Ed., 2008, 47, 2930–2946.
41. H. J. Zhang, C. C. Wong and Y. Wang, Cryst. Growth Des., 2012, 12, 5629–5634.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra06571b

---