Degradation of anion exchange membranes used for hydrogen production by ultrapure water electrolysis

Abstract

Solid-state alkaline water electrolysis using a pure water feed offers several distinct advantages over liquid alkaline electrolyte water electrolysis and proton exchange membrane water electrolysis. These advantages include a larger array of electrocatalyst available for oxygen evolution, no electrolyte management, and the ability to apply differential pressure. To date, there have been only a handful of reports on solid-state alkaline water electrolyzers using anion exchange membranes (AEMs), and there have been no reports that investigate loss in system performance over time. In this work, a solid-state alkaline water electrolyzer was successfully demonstrated with several types of polysulfone-based AEMs using a relatively expensive but highly active lead ruthenate pyrochlore electrocatalyst for the oxygen evolution reaction. The electrolysis of ultrapure water at 50 °C resulted in a current density of 400 mA cm⁻² at 1.80 V. We demonstrated that the short-term degradation of water electrolyzer performance over time was largely a consequence of carbon dioxide intrusion into the system and could be easily remedied, while long-term deterioration was a consequence of irreversible AEM polymer degradation.

---

Water electrolysis is an efficient, modular, and reliable method to produce hydrogen from renewable electricity sources for energy storage and/or grid-scale levelling during spikes in electricity production or periods of low demand.¹ Hydrogen is an important commodity for the production of several chemicals (e.g. ammonia via Haber process and hydrogenation reactions), possesses excellent properties as energy carrier for powering fuel cell devices, can be used to enrich natural gas, enhances biogas production via methanation of CO₂, and has a specific heat capacity approximately 15 times higher than air, making it attractive as a coolant (used to cool turbine windings). There are three types of water electrolysis processes commercially available at this time: alkaline (liquid electrolyte) water electrolysis, PEM water electrolysis, and solid oxide water electrolysis.² Alkaline liquid electrolyte water electrolyzers^3–6 and proton exchange membrane (PEM) water electrolyzers^7–9 are usually operated at temperatures in the range of 50 to 100 °C, and are more appropriate for portable and on-site applications.

Alkaline (liquid electrolyte) electrolyzers use concentrated alkali (typically 10 M KOH) to increase the conductivity and to facilitate the following redox reactions:

(Cathode) 2H₂O + 2e⁻ → H_2(g) + 2OH⁻ (1)

E^o_cathode = −0.828 V vs. SHE

(Anode) 4OH⁻ → O₂ + 2H₂O + 4e⁻ (2)

E^o_anode = 0.40 V vs. SHE

(Overall) 2H₂O + 4e⁻ → O₂ + 2H₂ (3)

E^o_overall = −1.228 V

The 10 M KOH electrolyte is very corrosive and requires routine maintenance because it can form insoluble carbonates when exposed to CO₂ from the air. These carbonates can easily precipitate with the mobile potassium ion, and the carbonate precipitates can deposit on the electrodes, thereby compromising electrolyzer performance.¹⁰ Another drawback of using liquid electrolytes is the inability to operate the electrolyzer with differential pressures. In a liquid electrolyte configuration, the pressure on both sides should be the same to prevent gas crossover and mixing, and to avoid the movement of the electrolyte impregnated within the porous cell separator. The production of hydrogen at intermediate pressures (15–30 bar), while releasing oxygen at atmospheric pressure, is highly desirable as it facilitates hydrogen storage and delivery to applications requiring pressurized hydrogen.¹¹ However, it is difficult to do this with a liquid electrolyte system.

Water electrolyzers using solid polymer exchange membrane (PEM) electrolytes, typical proton exchangers, have higher hydrogen production rates and energy efficiencies when compared to alkaline (liquid electrolyte) electrolyzers. Moreover, their design is more compact and suitable for portable or on-site applications.^8,12 Some of the current major drawbacks of PEM electrolyzers include: (i) expensive perfluorinated membranes (i.e. perfluorosulfonic acid membranes – Nafion®, Flemion®, Aciplex-S®) result in higher capital costs compared with alkaline liquid electrolyte water electrolysis stacks, and (ii) expensive noble metal electrocatalysts (i.e., platinum group metals (PGMs)) are needed to carry-out the necessary redox reactions in acidic media.^7,13

Cheaper membranes and electrocatalysts for membrane-based electrolyzers are a current priority for reducing electrolyzer costs; however, it is imperative that reduction in costs do not come at the expense of system performance that is achieved by state-of-the-art PEM architectures. Anion exchange membranes (AEMs) have shown promising results for fuel cells and are an interesting alternative for use in membrane-based water electrolyzers.^14–18 Herein, we evaluate the use of low cost polysulfone (PSF)-based AEMs as the electrolyte separator and electrode binders in a single-cell water electrolyzer operated with ultrapure water (18 MΩ). Solid-state AEM water electrolysis takes advantages of the best features of conventional alkaline water electrolyzers and PEM water electrolyzers. In a solid-state AEM water electrolyzer configuration, no corrosive liquid is used, the membrane architecture provides mechanical integrity for differential pressure operation (similar to PEM electrolyzers), and most important, the alkaline environment facilitates better OER kinetics and allows the use of non-Pt, and possibly, (in future), non-PGM electrocatalysts for the OER.^3,4,19–33

PSF (Udel®) was selected as the polymer to fabricate AEMs because it has the following properties: it produces mechanically strong films, it is inexpensive, it has excellent oxidative stability, and it is straightforward to process into AEMs. PSF was chloromethylated via a Friedel–Crafts reaction following the procedure first described by Avram and coworkers.^34,35

The chloromethylated PSF (CMPSF) was then reacted with a corresponding tertiary amine or 1-methylimidazole to obtain quaternary ammonium or imidazolium anion exchange materials. In this study, we examined several different AEM chemistries that corresponded to different cationic groups functionalized to the benzyl position of PSF. The different cationic groups assessed were: quaternary benzyl trimethylammonium (TMA⁺), quaternary benzyl quinuclidium (ABCO⁺, a.k.a 1-azaoniumbicyclo[2.2.2]octane), and quaternary benzyl 1-methylimidazolium (1M⁺). Fig. 1 illustrates the chemical structure of PSF and the types of cation chemistries assessed. The ESI section† details the procedure for chloromethylation and subsequent AEM formation and the procedure for NMR characterization of PSF AEM materials. Fig. S1 to S5 in the ESI section† provide the ¹H NMR of PSF, CMPSF, and PSF AEMs in the chloride counter ion form (PSF–TMA⁺ Cl⁻, PSF–1M⁺ Cl⁻, and PSF–ABCO⁺ Cl⁻).

Fig. 1 Polysulfone anion exchange membranes (PSF AEMs) with quaternary benzyl ammonium and imidazolium groups (trimethylammonium (TMA⁺), 1-azaoniumbicyclo[2.2.2.]octane (ABCO⁺), and 1-methylimidazolium (1M⁺)).

Membrane electrode assemblies (MEAs) with an active area of 25 cm² were prepared by sandwiching an AEM between two porous media electrodes (PME). Note: no heating or mechanical pressing was applied while sandwiching the AEM between the electrodes. Platinum black was used as the hydrogen evolution electrocatalyst, while a non-Pt-containing lead ruthenate pyrochlore was used as oxygen evolution electrocatalyst. While the OER catalyst does contain a PGM in ruthenium, the cost of Ru is about 8% that of iridium (a standard OER catalyst used in PEM electrolyzers), and the OER catalyst cost is further lowered by the presence of a significant amount of lead. Hence, this work also demonstrates the feasibility of using a less expensive OER electrocatalyst in an electrolyzer. The lead ruthenate pyrochlore reported herein was prepared using the procedure by Horowitz and co-workers.^36–39 The details of the synthesis and characterization of the lead ruthenate pyrochlore are presented in the ESI section,† along with the details on the fabrication and conditioning of membrane electrode assemblies. Furthermore, the ESI section† details the procedures used for water electrolysis testing.

Using the setup and materials described, we were able to effectively electrolyze ultrapure water at reasonable cell voltages (at 1.9 V we achieved current densities ranging from 300 to 500 mA cm⁻² depending on the AEM material used) and without the addition of any salt additives to increase the conductivity of the cell (see Fig. 2 & 3). The ultrapure water was fed to the anode side and recirculated to the storage tank at a flow rate of 400 mL min⁻¹. The water was stored in a reservoir tank at 50 °C with continuous bubbling of nitrogen to remove CO₂.

Fig. 2 Polarization curves for alkaline water electrolysis at 50 °C using PSF–TMA⁺OH⁻ AEM. Impedance spectra acquired after each polarization run was included in the Fig. S9 in the ESI section.†

Fig. 3 ¹H-NMR of pristine (before use in water electrolyzer) and post-mortem AEM water electrolyzer samples after acquisition of 3 polarization runs (approximately 1.5 hours of continuous operation) and after continuous operation at 200 mA cm⁻² for 6 hours.

The performance of the AEM water electrolyzer using PSF–TMA⁺OH⁻ membrane was better than that using PSF–ABCO⁺ OH⁻ or PSF–1M⁺ OH⁻ membranes (see Fig. S6 and S7 and accompanying discussion in ESI†), and it was similar to performances reported by Hickner and co-workers⁴⁰ and Xiao and coworkers,⁴¹ where current densities of 400 mA cm⁻² at 1.8 V were achieved. Additionally, the water electrolyzer performance in this work exceeded the performance by Faraj and coworkers^40,42 where the authors used a radiation grafted low-density polyethylene AEM containing a quaternary ammonium group with a potassium carbonate doped water feed to boost system ionic conductivity. However, complete carbon dioxide exclusion proved to be a difficult proposition due to its ubiquity in the surrounding environment and its solubility in water (0.75 g L⁻¹ at 50 °C).

Although the initial performance of the solid-state alkaline water electrolyzer using PSF–TMA⁺ was promising, it was observed that the overall overpotential at a given current density increased with each subsequent polarization curve run. Fig. 2 shows three polarization curves recorded in series. Note: it took about 30 min to acquire one polarization curve. After collecting each polarization curve, electrochemical impedance spectra (EIS) were immediately collected and analysis of the impedance spectra revealed a progressive increase in the high frequency resistance (HFR) and charge transfer resistance (see Fig. S9 in the ESI section†). The HFR increased 3-fold, from 0.60 Ohm cm² to 1.67 Ohm cm². Initially, it was hypothesized that the decline in electrolyzer performance was due to AEM degradation in alkaline media. It is well known that PSF AEMs degrade in alkaline solutions and in alkaline electrochemical systems.^35,43–46

However, post-mortem NMR analysis of the PSF–TMA⁺ membrane after 3 polarization curves (collected in succession – total run time of about 1.5 hours) did not show any loss of the quaternary ammonium groups or any evidence of backbone degradation (see Fig. 3 – an upfield focus of the ¹H NMR).

The absence of the characteristic backbone degradation peak at 1.2 ppm confirmed the lack of backbone degradation over the course of three successive polarization curves. The integrated area ratio in the ¹H NMR spectrum for protons in methyl groups attached to the backbone (peak “3”) and the protons in the methyl groups bonded to the quaternary ammonium (peak “2”) normalized to the DF value of the CMPSF precursor was the same before and after acquisition of the polarization curves. This confirmed that the quaternary ammonium groups did not degrade over the course of three successive polarization curve runs.

With this knowledge, we decided to assess if the increase of carbonate and bicarbonate anions due to CO₂ ingress was the reason behind increased polarization of the cell.^47,48 Both species have lower ionic mobility than hydroxide and their inclusion into the system lowers the ionic conductivity of the membrane and electrode binders. The experiment to evaluate whether or not carbon dioxide intrusion was the source of system performance loss was performed by systematically eliminating routes through which carbon dioxide could enter the system. In this experiment, water was degassed for 1 hour and the electrolyzer was sealed from the ambient environment in a makeshift glove box. Additionally, the hydrogen exit tube (cathode) was immersed in 1 M potassium hydroxide (KOH) to avoid back diffusion of carbon dioxide into the system. Under these stringent conditions, the system performance and HFR after each polarization curve run (up to 3) remained largely unchanged. See Fig. 4. The HFR acquired subsequent to each polarization curve run with the modified system was 0.60, 0.72 and 0.73 Ohm cm² respectively. This result did suggest that CO₂ ingress leading to carbonate ion formation was indeed responsible for short-term performance decay, and could be readily remedied by minimizing CO₂ dissolution by isolation of the MEA from air and treatment of the water feed (e.g., bubbling with nitrogen).

Fig. 4 Polarization curves for alkaline water electrolysis at 50 °C using PSF–TMA⁺ AEM after minimizing the carbon dioxide exposure to the system.

In separate test, a PSF water electrolyzer was evaluated for a relatively longer term (about 6 hours) at 50 °C at a constant current density of 200 mA cm⁻². During this constant-current test, the cell voltage climbed from 1.6 to 2.4 V. See Fig. S10 in the ESI section.† Post-mortem analysis of the MEA used in the long-term evaluation test revealed no cation site degradation, but it did demonstrate the onset of backbone degradation. See Fig. 3 (¹H NMR after 6 h of water electrolyzer performance). The degradation product observed at a chemical shift of 1.2 to 1.3 ppm, corresponded to quaternary carbon hydrolysis – a degradation mode identified in our previous work.⁴⁶ It was apparent that the stability of the electrolyzer was more sensitive to CO₂ intrusion in the short-term, while long-term performance loss occurred due to (irreversible) degradation of the AEM (especially the backbone). We did not observe any changes in the ¹H NMR spectrum in the interval 5 to 12 ppm, where the aromatic protons appear (see Fig. S11 in the ESI†).

Future work should be aimed at designing the electrolyzer to minimize CO₂ intrusion and to develop AEMs with higher ionic conductivity and better alkaline stability (both backbone and cation). Currently, we are working on the synthesis of AEMs with long alkyl chains (with quaternary ammonium cation attached at the end of the pendant chain) for improved stability in alkaline media. Since the presence of cationic groups in close proximity to the polymer backbone (in AEMs) triggers backbone degradation through ether and quaternary carbon hydrolysis,⁴⁶ the use of spacers could help reduce AEM degradation. The presence of beta hydrogens can compromise cation stability (Hofmann elimination), however the work by Tomoi,⁴⁹ Hibbs,⁵⁰ Pivovar,⁵¹ and Xu⁵² have shown that cation groups attached to long alkyl pedant groups (greater than n ≥ 3 carbons) exhibited substantially improved cation stability in alkaline media. We believe this strategy can be effectively employed for AEM water electrolyzers.

Conclusions

Solid-state alkaline water electrolysis was successfully demonstrated with several types of PSF AEMs containing different quaternary ammonium or imidazolium groups (TMA⁺, 1M⁺, and ABCO⁺). The water electrolyzer with a PSF–TMA⁺ AEM and binder demonstrated the best performance among the AEMs tested. A key goal of the study was to assess water electrolyzer cell stability and performance over time and to determine where the performance losses occurred. We examined the MEA via NMR spectroscopy post-mortem to determine if AEM/binder chemical degradation was the source of performance loss. Our findings suggest that the chemical (in)stability of the PSF backbone in the AEM was more of a concern at longer time intervals. Conversely, short-term system performance losses were ascribed to CO₂ intrusion, highlighting the extreme sensitivity of these systems to CO₂ ingress. Cation degradation was not observed in this study. We have shown that proper cell sealing of the electrolyzer minimized carbon dioxide intrusion, which greatly improved short-term water electrolyzer stability. Future priorities for developing solid-state alkaline water electrolyzers should focus on designing alkaline stable AEMs with higher ionic conductivities, on sealing the system to prevent ingress of carbon dioxide, and operating the cell in a manner designed to decrease the impact of carbon dioxide on performance (e.g. at elevated temperature).

Materials and methods

Preparation of chloromethylated PSF (CMPSF) precursor for PSF AEMs and the synthesis of PSF AEMs with TMA⁺, ABCO⁺ and imidazolium bases were reported in our previous work.³⁵ Additional details can be found in the ESI section.† Lead ruthenate pyrochlores were synthesized following the alkaline solution technique described initially by Horowitz and coworkers.^36,38

All NMR measurements were carried out on a Bruker Avance 360 MHz NMR spectrometer. Samples were prepared by dissolving 30 to 40 mg of polymer in 1 mL of deuterated solvent (dimethylsulfoxide (d-DMSO) or deuterated chloroform (CDCl₃)). Tetramethylsilane (TMS) was added as an internal standard for all samples.

The electrolysis of ultrapure DI water (no salt addition) was carried out in a 25 cm² single fuel cell hardware (Fuel Cell Technologies, Inc). Membrane electrode assemblies (MEAs) with were assembled in the hardware by placing an AEM between two gas diffusion electrodes (without heating or pressing). Pt black was used as the HER electrocatalyst at the cathode, and lead ruthenate pyrochlore as the OER electrocatalyst at the anode. The electrocatalyst loading was approximately 2.5 mg cm⁻². The concentration of binder in the electrodes (PSF–TMA⁺ Cl⁻, PSF–ABCO⁺ Cl⁻ or PSF–1M⁺ Cl⁻) was 30 wt%. Polarization curves were obtained by stepping the current density from 10 mA cm⁻² to 700 mA cm⁻² (10, 25, 50, 100, 200, 300, 400, 500, 600 and 700 mA cm⁻²). The system was held at each current density for 2 min, and the acquisition stopped when the voltage increased above 2.5 V.

Acknowledgements

We wish to thank the Department of Energy Small Business Innovation Research (SBIR) Phase I (“Economical production of hydrogen through development of novel, high efficiency electrocatalysts for alkaline membrane electrolysis” Lead PI: Kathy Ayers – Proton On-Site) for funding this work. We acknowledge the Research Resource Center (RRC) at the University of Illinois at Chicago for access to the 360 MHz Bruker NMR instrument used during this research. We also thank Mr Min-suk J. Jung for assisting in hydroxide ion conductivity measurements of PSF AEMs.

Notes and references

1. E. Guerrini and S. Trasatii, ed. P. Barbaro and C. Bianchini, Catalysis For Sustainable Energy Production, Wiley-VCH, Weinheim, 2009.
2. A. Brisse, J. Schefold and M. Zahid, Int. J. Hydrogen Energy, 2008, 33, 5375.
3. K. Zeng and D. Zhang, Prog. Energy Combust. Sci., 2010, 36, 307.
4. G. Schiller, R. Henne, P. Mohr and V. Peinecke, Int. J. Hydrogen Energy, 1998, 23, 761.
5. H. Wendt, H. Hofmann and V. Plzak, Mater. Chem. Phys., 1989, 22, 27.
6. C. T. Bowen, H. J. Davis, B. F. Henshaw, R. Lachance, R. L. LeRoy and R. Renaud, Int. J. Hydrogen Energy, 1984, 9, 59.
7. K. E. Ayers, E. B. Anderson, C. B. Capuano, B. D. Carter, L. T. Dalton, G. Hanlon, J. Manco and M. Niedzwiecki, ECS Trans., 2010, 33, 3.
8. K. E. Ayers, C. B. Capuano and E. B. Anderson, ECS Trans., 2012, 41, 15.
9. Y. Zhang, C. Wang, N. Wan, Z. Liu and Z. Mao, Electrochem. Commun., 2007, 9, 667.
10. M. S. Naughton, F. R. Brushett and P. J. A. Kenis, J. Power Sources, 2011, 196, 1762.
11. H. Vandenborre, R. Leysen, H. Nackaerts, D. Van der Eecken, P. Van Asbroeck, W. Smets and J. Piepers, Int. J. Hydrogen Energy, 1985, 10, 719.
12. A. T. Marshall, S. Sunde, M. Tsypkin and R. Tunold, Int. J. Hydrogen Energy, 2007, 32, 2320.
13. A. Ursua, L. M. Gandia and P. Sanchis, Proc. IEEE, 2012, 100, 410.
14. S. Gu, J. Skovgard and Y. S. Yan, ChemSusChem, 2012, 5, 843.
15. R. Zeng, J. Handsel, S. D. Poynton, A. J. Roberts, R. C. T. Slade, H. Herman, D. C. Apperley and J. R. Varcoe, Energy Environ. Sci., 2011, 4, 4925.
16. K. Matsumoto, T. Fujigaya, H. Yanagi and N. Nakashima, Adv. Funct. Mater., 2011, 21, 1089.
17. S. Gu, W. Sheng, R. Cai, S. M. Alia, S. Song, K. O. Jensen and Y. Yan, Chem. Commun., 2013, 49, 131.
18. S. Lu, J. Pan, A. Huang, L. Zhuang and J. Lu, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 20611.
19. Y. W. D. Chen and R. N. Noufi, J. Electrochem. Soc., 1984, 131, 1447.
20. A. J. Esswein, M. J. McMurdo, P. N. Ross, A. T. Bell and T. D. Tilley, J. Phys. Chem. C, 2009, 113, 15068.
21. D. E. Hall, J. Electrochem. Soc., 1981, 128, 740.
22. D. E. Hall, J. Electrochem. Soc., 1985, 132, 41C.
23. M. Hamdani, M. I. S. Pereira, J. Douch, A. Ait Addi, Y. Berghoute and M. H. Mendonça, Electrochim. Acta, 2004, 49, 1555.
24. J. Y. Huot, M. L. Trudeau and R. Schulz, J. Electrochem. Soc., 1991, 138, 1316.
25. F. Rosalbino, D. Macciò, E. Angelini, A. Saccone and S. Delfino, J. Alloys Compd., 2005, 403, 275.
26. R. N. Singh, L. Bahadur, J. P. Pandey, S. P. Singh, P. Chartier and G. Poillerat, J. Appl. Electrochem., 1994, 24, 149.
27. R. N. Singh, A. N. Jain, S. K. Tiwari, G. Poillerat and P. Chartier, J. Appl. Electrochem., 1995, 25, 1133.
28. R. N. Singh, S. K. Tiwari, S. P. Singh, N. K. Singh, G. Poillerat and P. Chartier, J. Chem. Soc., Faraday Trans., 1996, 92, 2593.
29. H. Michishita, Y. Misumi, D. Haruta, T. Masaki, N. Yamamoto, H. Matsumoto and T. Ishihara, J. Electrochem. Soc., 2008, 155, B969.
30. C. Fan, D. L. Piron, A. Sleb and P. Paradis, J. Electrochem. Soc., 1994, 141, 382.
31. X. Wu and K. Scott, J. Mater. Chem., 2011, 21, 12344.
32. X. Wu and K. Scott, J. Power Sources, 2012, 206, 14.
33. X. Wu and K. Scott, J. Power Sources, 2012, 214, 124.
34. E. Avram, E. Butuc, C. Luca and I. Druta, J. Macromol. Sci., Part A: Pure Appl. Chem., 1997, 34, 1701.
35. C. G. Arges, J. Parrondo, G. Johnson, A. Nadhan and V. Ramani, J. Mater. Chem., 2012, 22, 3733.
36. H. S. Horowitz, J. M. Longo, H. H. Horowitz and J. T. Lewandowski, ACS Symp. Ser., 1985, 279, 143.
37. J. B. Goodenough and R. Manoharan, Proc. – Electrochem. Soc., 1992, 92–11, 523.
38. H. S. Horowitz, J. M. Longo and J. T. Lewandowski, Mater. Res. Bull., 1981, 16, 489.
39. J. Prakash, D. Tryk and E. Yeager, J. Power Sources, 1990, 29, 413.
40. Y. Leng, G. Chen, A. J. Mendoza, T. B. Tighe, M. A. Hickner and C. Wang, J. Am. Chem. Soc., 2012, 134, 9054.
41. L. Xiao, S. Zhang, J. Pan, C. Yang, M. He, L. Zhuang and J. Lu, Energy Environ. Sci., 2012, 5, 7869.
42. M. Faraj, M. Boccia, H. Miller, F. Martini, S. Borsacchi, M. Geppi and A. Pucci, Int. J. Hydrogen Energy, 2012, 37, 14992.
43. G. Merle, M. Wessling and K. Nijmeijer, J. Membr. Sci., 2011, 377, 1.
44. G. Couture, A. Alaaeddine, F. Boschet and B. Ameduri, Prog. Polym. Sci., 2011, 36, 1521.
45. J. R. Varcoe and R. C. T. Slade, Fuel Cells, 2005, 5, 187.
46. C. G. Arges and V. Ramani, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 2490.
47. K. N. Grew, X. Ren and D. Chu, ECS Trans., 2011, 41.
48. S. Suzuki, H. Muroyama, T. Matsui and K. Eguchi, Electrochim. Acta, 2013, 88.
49. M. Tomoi, K. Yamaguchi, R. Ando, Y. Kantake, Y. Aosaki and H. Kubota, J. Appl. Polym. Sci., 1997, 64, 1161.
50. M. R. Hibbs, J. Polym. Sci., Part B: Polym. Phys., 2012, 51, 1736.
51. H. Long, K. Kim and B. S. Pivovar, J. Phys. Chem. C, 2012, 116, 9419.
52. Z. Zhang, L. Wu, J. Varcoe, C. Li, A. L. Ong, S. Poynton and T. Xu, J. Mater. Chem. A, 2013, 1, 2595.

---

Footnote

† Electronic supplementary information (ESI) available: Detailed experimental methods. See DOI: 10.1039/c3ra46630b

---