A sulfur doped carbon nanotube as a potential catalyst for the oxygen reduction reaction

Abstract

In this work, the catalytic effect of a sulfur doped carbon nanotube (SCNT) in a four-electron oxygen reduction reaction (ORR) was studied using theoretical methods. The structures of all reaction species, intermediates and their related transition states via possible mechanistic pathways were optimized and their energies were extracted to obtain the energy profile of the whole process. The results showed the high efficiency of SCNT (and to a lesser extent, CNT) as a catalyst of this process and SCNT could catalyze this process from the beginning to the end by releasing 15.22 eV energy. In addition, each individual step was energetically favorable and during the reaction progress, the energies of intermediates were decreased constantly. Because of the special properties of the sulfur atom, the doping of sulfur atoms into the CNT structure, make it a more useful candidate than CNT to catalyze chemical processes.

---

Introduction

Fuel cells are important resources for clean energy. They provide high power densities with large efficiencies and without producing large amounts of environmental pollutants.^1–4 Two important processes in fuel cells are oxidation of fuel (such as hydrogen, methanol, ethanol, …) at the anode and reduction of oxygen (oxygen reduction reaction or ORR) at the cathode.^5,6 The details of these processes will determine the general efficiency of a fuel cell.⁷ Therefore, many efforts have been made to enhance each of these processes by changing their variable parameters. Undoubtedly, one of the most important parts of each fuel cell, that has direct effects on its efficiency, is the employed catalyst.⁸ Because of the more generality of the cathode process (ORR), like many other scientists, we have focused on this reaction, especially on its catalyst. When the scientists started their studies related to this topic, metal catalysts (especially Pt, Pd and their alloys) have been mostly used as cathode (and also anode) catalysts in the fuel cells because they exhibited the highest electrocatalytic activities in primary experiments.^9–12 However, they showed several disadvantages that could limit their application in the large-scale processes such as CO poisoning, limited resources, poor durability and high-cost.¹ Therefore, much efforts have been made to find new resources for the cathode's catalyst of fuel cells. In this line, non-metal catalysts have been attracted more interests.^13–15 Recently, using both computational and experimental studies, it was reported that doped carbon nanostructures (nanotube, graphene and fullerene) could play a role of metal-free cathode catalyst in fuel cells with remarkable properties such as the high electrical conductivity and oxidation stability as well as higher activity, less price and higher methanol tolerance versus common metal catalysts.^16–19 However, despite the reported studies related to the use of P,¹⁹ N,^20–23 and B²⁴ doped carbon materials as cathode catalysts in fuel cells, only some reports about the use sulfur-doped graphene (alone or along with N) in this process have been observed.^25–30 In addition, the potencies of phosphorus and nitrogen co-doped CNT,³¹ nitrogen doped CNT,^32,33 nitrogen doped graphene,^34,35 boron doped CNT,^36–38 boron and nitrogen doped graphene,³⁹ nitrogen doped fullerene⁴⁰ sulfur and nitrogen dual-doped graphene,⁴¹ nitrogen and fluorine doped graphene,⁴² iron encapsulated nitrogen and sulfur co-doped graphene,⁴³ iron and nitrogen doped CNT,⁴⁴ nitrogen and sulfur doped graphene–CNT composites⁴⁵ and sulfur and nitrogen co-doped graphene oxide⁴⁶ have proven by experimental or theoretical (or both) studies. Therefore, the possibility of using doped CNTs or doped graphenes as a catalyst in ORR has been accepted by the scientific communities. However, any computational reports related to the use of sulfur-doped carbon nanotube (SCNT) as ORR catalyst has not been observed while because of the high polarizability and especial properties of sulfur atom, promising properties for sulfur doped nanostructures and especially SCNTs could be expected.^47–50 It seems useful to examine the ability of SCNT as a potential catalysts of ORR in fuel cells by theoretical studies. Therefore, in continuation of our previous theoretical studies related to the applications of sulfur doped carbon nanostructures,^51,52 we have decided to study of the possibility of using SCNT as ORR catalyst. In this line, the full mechanistic details for possible pathways of this process and its energy profile should be worked out clearly.

In the present work, DFT calculations have been employed to study of the energetic and mechanistic details of ORR using SCNT as a potential catalyst. SCNT has electrocatalytic active sites on its sulfur atom that could enhance the reduction process. The energy profiles of all pathways were calculated and the minimum energy path (MEP) was determined. The possible pathways for a four-electron mechanism (O₂ + 4H⁺ + 4e⁻ → 2H₂O) on the surface of SCNT were analyzed.²⁶ Moreover, in addition to the energies of all involved species, their molecular orbital properties (HOMO and LUMO energies), energy gaps (E_g), chemical potentials (μ), and electrophilicity indexes (ω) were calculated.⁵³ The results of this study and the details of computations will be presented in the next sections.

Method

The mechanistic details of ORR have been evaluated by computational studies using Gaussian 09 program.⁵⁴ Density Functional Theory (DFT) method has employed because its reliable results and lesser complexity against increasing system size, which is an advantage for DFT methods.^55,56 Moreover, among various DFT method, CAM-B3LYP method was used because of considering the long-range interaction and producing more reliable data.^39,57,58 In combination with CAM-B3LYP method, 6-311G(d) basis set⁵⁹ was used to optimize the structure and calculate molecular properties. The structures of transition states for each step were obtained and optimized by applying Schlegel's synchronous-transit-guided quasi-Newton (QST3) method, which was started from the fully optimized structures of reactants and ended on the fully optimized structures of products. The transition states were verified by frequency calculations to ensure that they were first order saddle points with only one negative eigenvalue. Additionally, intrinsic reaction coordinate (IRC) calculations proved that each reaction linked correct products to reactants. Rate constants were calculated by canonical transition state theory using Eyring equation.⁶⁰ All the structures were optimized without any symmetric restriction or pre-defined conformational structures. The absence of imaginary frequency showed that the structure was true minima in its respective levels of theory.

Results and discussion

To start the study, four-electron reduction of oxygen to water molecules (O₂ + 4H⁺ + 4e⁻ → 2H₂O) has been considered in our calculations. A zigzag (5,0) CNT (with chirality angle = 0 and tube length = 10), consisted of 60 carbon atoms and saturated it's both ends with 10 hydrogen atoms has been considered as the employed CNT model. For SCNT model, only one carbon atom in the middle of the CNT model was replaced with sulfur atom and the structures of both models were optimized first. Then, these models were considered to examine their abilities as ORR catalyst according to the described mechanism. For simplicity, only we will talk about the SCNT model during the discussion and the results related to CNT model will be compared with those of SCNT in some parts of this section. The whole process, started from SCNT and molecular oxygen and ended to SCNT and two molecules of water were designed in three different paths (these paths are similar in some steps). All steps consisted of three reactants (SCNT, O₂ and H⁺), 10 intermediates (I1–I10), 5 transition states and two products (SCNT and H₂O) to show the catalytic effect of SCNT on this process. To define the intermediates, van der Waals (VDW) interactions between SCNT and each reaction part were described by dashed lines and covalent interactions between them were defined with solid lines. Oxygen reduction reaction (ORR) on SCNT could proceed by a direct four-electron pathway. The general procedure for these reactions, consisted of all possible intermediates and pathways was shown in Fig. 1. To start the study, the structure of SCNT consisted of one sulfur atom for each reaction step was optimized and then, calculations related to all intermediates steps were performed. The optimized structures of all intermediates, transition states, and products using SCNT as catalyst were depicted in Fig. 2 and the related structures using CNT as catalyst were shown in Table S1 (at ESI†). For each step, the structures of reactants, transition state and products were optimized and their energies were extracted to calculate the relative energies and the energy profiles for reaction pathways. The relative energies for all involved species in presence of both catalysts (SCNT and CNT) were listed in Table 1. The relative energies were listed for each catalytic model separately, the sum of energy of reactants was considered as zero and all energies were compared with this energy. Moreover, the adsorption energies (E_ad) were calculated for some species according to the eqn (1). The negative adsorption energy shows the adsorbate would energetically adsorbed on the surface of SCNT.

E_ad = E_SCNT–O2 − (E_SCNT + E_O2) (1)

Fig. 1 The general procedure for the reaction process of ORR using SCNT as catalyst.

Fig. 2 Optimized structures of the catalyst and all intermediates of studied ORR reaction.

Table 1 Relative energies (eV) for all species involved in SCNT and CNT catalyzed ORR via four-electron mechanism

Name	Molecules	Relative E	Molecules	Relative E
R, cat	SCNT + O2	0	CNT + O2	0
I1	SCNT⋯O2	−0.22	CNT⋯O2	−1.51
I2	SCNT–O2	−1.85	CNT–O2	−2.48
I3	SCNT⋯OOH	−4.82	CNT⋯OOH	−4.64
I4	SCNT⋯OOH^+	−4.82	CNT⋯OOH^+	−4.17
I5	SCNT–OOH^+	−3.95	CNT–OOH^+	−4.48
I6	SCNT–OOH	−6.40	CNT–OOH	−5.29
I7	SCNT⋯OH–OH	−7.76	CNT⋯OH–OH	−7.73
I8	HO–SCNT–OH	−11.50	HO–CNT–OH	−10.57
I9	H2O⋯SCNT–OH	−14.69	H2O⋯CNT–OH	−13.69
I10	H2O⋯SCNT⋯H2O	−16.15	H2O⋯CNT⋯H2O	−15.74
P, cat	SCNT + 2H2O	−15.22	CNT + 2H2O	−15.22

---

In the first step of the reaction using SCNT as catalyst, the O₂ adsorbed (via VDW interaction) on the surface of SCNT as schematically shown in Fig. 1 and 2 that this complex was shown as SCNT⋯O₂ (I1). This should be mentioned that for all structures, the most stable conformation was found and employed in the study. This complex could convert to SCNT–OOH (SCNT in covalent bond with hydroperoxide, I6) by the adsorption of one proton and one electron in three ways. Then, producing SCNT–O₂ (I2) and the simultaneous adsorption of proton and electron to produce SCNT–OOH (I6) is the first way. The adsorption of proton and electron to produce SCNT⋯OOH (I3) and then converting it to I6 is the second way of this mechanism. Moreover, the adsorption of one proton to produce SCNT⋯OOH⁺ (I4), converting it to SCNT–OOH⁺ (I5) and the adsorption of one electron to produce I6 is the third way. In continuation of the reaction, all paths obey the unique route (from I6), consisted of adsorption of proton and electron to produce SCNT⋯OH–OH (I7, two OH connected to each other and have VDW interaction with SCNT via one O₂), converting it to HO–SCNT–OH (I8, two OH separately bonded to SCNT), adsorption of proton and electron to produce H₂O⋯SCNT–OH (I9, producing the first water molecule as product) and finally adsorption of proton and electron to produce H₂O⋯SCNT⋯OH₂ (I10, producing the second water molecule) that this molecule.

Finally, I10 releases two water molecules and starts its catalytic role in another cycle. The same conditions were considered for the reaction using CNT as catalyst for comparison. The simulations shows that all OOH⁺, O₂ and OOH species could be adsorbed on the surface of SCNT, preferably at carbon atom adjacent to the doped sulfur atom, as shown Fig. 2. However, the adsorption energy of O₂ on the surface of SCNT is −0.22 eV, which is smaller than that for OOH (−0.99 eV) and both of them have much smaller adsorption energies than OOH⁺ (−12.97 eV). For CNT, the adsorption energies of OOH, O₂ and OOH⁺ are respectively −0.81, −1.51 and −12.32 eV. In addition, both O₂ and OOH release more energies when they bonded to SCNT covalently (respectively −1.85 and −2.57 eV) while the covalent bonding of OOH⁺ to SCNT is less favorable than its VDW bonding (−12.11 eV for covalent and −12.97 eV for VDW bonding).

By considering the data listed in Table 1, all data showed the exothermic processes for the most of steps and totally for the general reaction using SCNT as catalyst. Therefore, the simulations demonstrate ORR could spontaneously occur using SCNT as catalyst in a four-electron pathway. To present more detailed explanations, it should be noticed that first step (SCNT + O₂ → SCNT⋯O₂) releases 0.22 eV and the second step (SCNT⋯O₂ + H⁺ + e⁻ → SCNT–OOH) releases 6.18 eV. Among three different pathways for step 2, routes 1 (SCNT⋯O₂ → SCNT–O₂ and SCNT–O₂ + H⁺ + e⁻ → SCNT–OOH) and 2 (SCNT⋯O₂ + H⁺ + e⁻ → SCNT⋯OOH → SCNT–OOH) seems more suitable than route 3 (SCNT⋯O₂ + H⁺ → SCNT⋯OOH⁺ → SCNT–OOH⁺ and SCNT–OOH⁺ + e⁻ → SCNT–OOH) because it consisted of increasing in the energy of the reaction. Of course it is better to ignore the SCNT–OOH⁺ and consider the direct conversion of SCNT⋯OOH⁺ to SCNT–OOH without increasing in the energy of intermediate. The third step (SCNT–OOH + H⁺ + e⁻ → SCNT⋯OH–OH) releases 1.36 eV, the forth step (SCNT⋯OH–OH → HO–SCNT–OH) releases 3.74 eV and the fifth step (HO–SCNT–OH + H⁺ + e⁻ → H₂O⋯SCNT–OH) releases 3.19 eV. Finally, the sixth step (H₂O⋯SCNT–OH + H⁺ + e⁻ → H₂O⋯SCNT⋯OH₂) releases 1.46 eV but releasing two water molecules from SCNT need to 0.93 eV energy because of the exclusion of VDW interactions between SCNT and water molecules. The general reaction (SCNT + O₂ + 4H⁺ + 4e⁻ → 2H₂O + SCNT) is energetically favorable by −15.22 eV that confirms the efficient catalytic effect of SCNT in this reaction. This energy is equal to the simple ORR reaction or ORR reaction in presence of CNT or each catalyst because the catalyst (and its energy) could be deleted from the both sides of the reaction equation. By observing the results using CNT as catalyst, all steps are energetically favorable but the most of steps are energetically less favorable than the same step using SCNT and the whole process (from the reactants to I10) is also less favorable (−15.74 eV for CNT and −16.15 eV for SCNT). To better understanding the energy profiles of these reaction pathways, the graphical presentations for both cases (SCNT and CNT) were depicted in Fig. 3. Therefore, it seems that SCNT is more suitable catalyst for ORR reaction than CNT. Moreover, there are some reported experimental studies that showed the efficiency of SCNT in ORR. In the first report, Chen and coworkers studied ORR on graphene–carbon nanotube composites doped sequentially with nitrogen and sulfur that showed its excellent electrochemical stability, along with significant ORR when sulfur atom used as a co-dopant.⁶¹ In another work, SCNT were used alone as catalysts for ORR in alkaline medium.⁶² In the last report, by controlling the active sites of SCNT-graphene, its efficiency in ORR was highly improved.⁶³ In all of these reports, sulfur atom plays an important role for catalyzing ORR.

Fig. 3 The diagram for different reaction pathways of ORR using SCNT (top) and CNT (below) as catalyst.

To confirm the catalytic ability of SCNT in ORR process, in addition to the energy of intermediates, it seems useful to obtain the structures of transition states and barrier energies. Therefore, the structures of related transition states were found (optimized, shown in Fig. 2), the barrier energies were extracted for the best path (via I2, as described previously) and the results were listed in Table 2. According to these BE values, all barriers except conversion of I7 to I8 are less than 1 eV and this pathway is theoretically possible and all ΔE values are prominently negative. In addition, for I7 to I8 step, there is enough released energies from the previous steps to provide enough energies and make this step possible. Meanwhile, the graphical presentation of full reaction pathway, consisting intermediates and transition states, was shown in Fig. 4. This figure shows a continuous decreasing energy during the reaction path.

Table 2 The calculated energy differences between reactants and products (ΔE, in eV) and barrier energies (BE, in eV) for each step of ORR reaction pathway using SCNT as catalyst

Step	Transition state	ΔE	BE
I2 to I6	TS-2–6	−3.92	0.33
I6 to I7	TS-6–7	−1.36	0.32
I7 to I8	TS-7–8	−3.74	1.65
I8 to I9	TS-8–9	−3.18	0.88
I9 to I10	TS-9–10	−1.47	0.60

---

Fig. 4 The full energy diagram of ORR (consisted of all involved intermediates and transition states) using SCNT as catalyst.

Since the frontier molecular orbitals (HOMO and LUMO, FMOs), their energies and the related quantities help us about the reactions, at the final part of this study, the properties of frontier molecular orbitals have been studied. The shapes of HOMO and LUMO for all ten intermediates of both reactions (using SCNT and CNT) were shown in Fig. S2 and S3.†

Moreover, the energies of HOMO and LUMO (in eV), the energy gap between them (E_g) and the chemical potentials have been calculated for all structures and the results for SCNT's reaction were listed in Table 3. Considering the shape of FMOs indicated that in the most of intermediates (except in 2 final cases, I9 and I10), the electron densities in both HOMO and LUMO are placed and the region near doped sulfur atom and our reactants or intermediates that make them active in the reaction process.

Table 3 The energies of HOMO and LUMO, their energy gaps (E_g) and chemical potentials for all intermediates. All energy values are reported in eV

Molecules	E HOMO	E LUMO	Egap	μ
SCNT⋯O2	−0.208	−0.107	0.101	−0.261
SCNT–O2	−0.212	−0.115	0.097	−0.269
SCNT⋯OOH	−0.232	−0.099	0.133	−0.282
SCNT⋯OOH^+	−0.234	−0.105	0.129	−0.286
SCNT–OOH^+	−0.225	−0.104	0.121	−0.277
SCNT–OOH	−0.191	−0.068	0.122	−0.225
SCNT⋯OH–OH	−0.203	−0.094	0.109	−0.250
HO–SCNT–OH	−0.200	−0.084	0.116	−0.242
H2O⋯SCNT–OH	−0.215	−0.087	0.128	−0.258
H2O⋯SCNT⋯H2O	−0.200	−0.100	0.100	−0.250

---

The HOMO energies are between −0.191 and −0.234 eV, the LUMO energies are between −0.068 and −0.115 eV and the energy gaps are between 0.097 and 0.133 eV. The smaller energy gaps of SCNT and these intermediates make them more reactive and more effective than CNT in this process. In general, there is not any meaningful relation or order regard to these values and all of these species have nearly similar electronic properties.

Moreover, the chemical potential values are in the range of −0.162–0.235 and by neglecting some irregularities, the chemical potentials decreased during the reaction process and it reaches from −0.221 in I1 to −0.208 in I10.

Conclusion

In summary, DFT method was employed to study the catalytic effect of SCNT and CNT in four-electron oxygen reduction reaction as an important process in fuel cells. The whole processes were considered and different pathways were examined. All reaction species, intermediates and transition states were investigated and their energies were extracted to obtain the energy profile for the reaction. The results on this process show the highly efficient effect of SCNT as a catalyst of this process. The doping of S in carbon nanotube introduces unpaired electrons and causes local high spin density, resulting in high electrocatalytic performance for the ORR that make SCNT as useful candidate to consider it as catalyst in many chemical process even it seems to be more effective than CNT. Finally, the population analyses were shown that S doping could result in a smaller energy gap than CNT and thus lead to improve its catalytic activity.

Acknowledgements

We are grateful National High-Performance Computing Center (NHPCC) at Isfahan University of Technology (http://nhpcc.iut.ac.ir, Rakhsh supercomputer) for providing computational facilities for performing this work. This work has been supported by research affair of Isfahan University of technology.

Notes and references

1. B. C. H. Steele and A. Heinzel, Nature, 2001, 414, 345–352.
2. R. A. Costa and J. R. Camacho, J. Power Sources, 2006, 161, 1176–1182.
3. J. M. Andujar and F. Segura, Renewable Sustainable Energy Rev., 2009, 13, 2309–2322.
4. H. A. Gasteiger, S. S. Kocha, B. Somapalli and F. T. Wagner, Appl. Catal., B, 2005, 56, 9–35.
5. G. J. K. Acres, J. Power Sources, 2001, 100, 60–66.
6. C. Lamy, A. Lima, V. LeRhun, F. Delime, C. Coutanceau and J. M. Leger, J. Power Sources, 2002, 105, 283–296.
7. H. Liua, C. Songa, L. Zhanga and J. A. Zhang, J. Power Sources, 2006, 155, 95–110.
8. E. Antolini, J. Power Sources, 2007, 170, 1–12.
9. M. Lefevre, E. Proietti, F. Jaouen and J. P. Dodelet, Science, 2009, 324, 71–74.
10. J. Yang, D. J. Liu, N. N. Kariuki and L. X. Chen, Chem. Commun., 2008, 3, 329–331.
11. J. L. Fernández, V. Raghuveer, A. Manthiram and A. J. Bard, J. Am. Chem. Soc., 2005, 127, 13100–13101.
12. B. Lim, M. Jiang, P. H. Camargo, E. C. Cho, J. Tao, X. Lu, Y. Zhu and Y. Xia, Science, 2009, 324, 1302–1305.
13. G. Che, B. B. Lakshmi, E. R. Fisher and C. R. Martin, Nature, 1998, 393, 346–349.
14. K. Gong, P. Yu, L. Su, S. Xiong and L. Mao, J. Phys. Chem. C, 2007, 111, 1882–1887.
15. A. Kongkanand, S. Kuwabata, G. Girishkumar and P. Kamat, Langmuir, 2006, 22, 2392–2396.
16. Z. Yang, H. Nie, X. A. Chen, X. Chen and S. Huang, J. Power Sources, 2013, 236, 238–249.
17. K. P. Gong, F. Du, Z. H. Xia, M. Durstock and L. M. Dai, Science, 2009, 323, 760–764.
18. F. Gao, G. L. Zhao, S. Yang and J. J. Spivey, J. Am. Chem. Soc., 2013, 135, 3315–3318.
19. Z. W. Liu, F. Peng, H. J. Wang, H. Yu and W. X. Zheng, Angew. Chem., Int. Ed., 2011, 50, 3257–3261.
20. L. Zhang, J. Niu, L. Dai and Z. Xia, Langmuir, 2012, 28, 7542–7550.
21. L. Zhang and Z. Xia, J. Phys. Chem. C, 2011, 115, 11170–11176.
22. Y. Sun, C. Li and G. Shi, J. Mater. Chem., 2012, 22, 12810–12816.
23. R. L. Liu, D. Q. Wu, X. L. Feng and K. Mullen, Angew. Chem., Int. Ed., 2010, 49, 2565–2569.
24. L. Yang, S. Jiang, Y. Zhao, L. Zhu, S. Chen, X. Wang, Q. Wu, J. Ma, Y. Ma and Z. Hu, Angew. Chem., 2011, 50, 7132–7135.
25. L. Zhang, J. Niu, M. Li and Z. Xia, J. Phys. Chem. C, 2014, 118, 3545–3553.
26. Z. Yang, Z. Yao, G. Li, G. Fang, H. Nie, Z. Liu, X. Zhou, X. Chen and S. Huang, ACS Nano, 2012, 6, 205–211.
27. J. Liang, Y. Jiao, M. Jaroniec and S. Z. Qiao, Angew. Chem., 2012, 124, 11664–11668.
28. M. A. Hoque, F. M. Hassan, M. H. Seo, J. Y. Choi, M. Pritzker, S. Knights, S. Ye and Z. Chen, Nano Energy, 2016, 19, 27–38.
29. H. R. Jiang, T. S. Zhao, L. Shi, P. Tan and L. An, J. Phys. Chem. C, 2016, 120, 6612–6618.
30. S. Yang, L. Zhi, K. Tang, X. Feng, J. Maier and K. Müllen, Adv. Funct. Mater., 2012, 22, 3634–3640.
31. D. Yu, Y. Xue and L. Dai, J. Phys. Chem. Lett., 2012, 3, 2863–2870.
32. Z. Mo, S. Liao, Y. Zheng and Z. Fu, Carbon, 2012, 50, 2620–2627.
33. Z. Chen, D. Higgins and Z. Chen, Carbon, 2010, 48, 3057–3065.
34. L. Feng, Y. Chen and L. Chen, ACS Nano, 2011, 5, 9611–9618.
35. R. Imran-Jafri, N. Rajalakshmi and S. Ramaprabhu, J. Mater. Chem., 2010, 20, 7114–7117.
36. L. Yang, S. Jiang, Y. Zhao, L. Zhu, S. Chen, X. Wang, Q. Wu, J. Ma, Y. Ma and Z. Hu, Angew. Chem., 2011, 123, 7270–7273.
37. L. Yang, S. Jiang, Y. Zhao, L. Zhu, S. Chen, X. Wang, Q. Wu, J. Ma, Y. Ma and Z. Hu, Angew. Chem., 2011, 50, 1–6.
38. R. Du, N. Zhang, J. Zhu, Y. Wang, C. Xu, Y. Hu, N. Mao, H. Xu, W. Duan, L. Zhuang, L. Qu, Y. Hou and J. Zhang, Small, 2015, 11, 3903–3908.
39. S. Wang, L. Zhang, Z. Xia, A. Roy, D. W. Chang, J. Baek and L. Dai, Angew. Chem., 2012, 51, 4209–4212.
40. F. Gao, G. Zhao, S. Yang and J. J. Spivey, J. Am. Chem. Soc., 2013, 135, 3315–3318.
41. J. Liang, Y. Jiao, M. Jaroniec and S. Z. Qiao, Angew. Chem., 2012, 51, 11496–11500.
42. S. Jiang, Y. Sun, H. Dai, J. Hu, P. Ni, Y. Wang, Z. Li and Z. Li, Nanoscale, 2015, 7, 10584–10589.
43. B. P. Vinayan, T. Diemant, R. J. Behm and S. Ramaprabhu, RSC Adv., 2015, 5, 66494–66501.
44. D. H. Lee, W. J. Lee, W. J. Lee, S. O. Kim and Y. H. Kim, Phys. Rev. Lett., 2011, 106, 175502.
45. D. C. Higgins, M. A. Hoque, F. Hassan, J. Y. Choi, B. Kim and Z. Chen, ACS Catal., 2014, 4, 2734–2740.
46. J. Xu, G. Dong, C. Jin, M. Huang and L. Guan, ChemSusChem, 2013, 6, 493–499.
47. H. Tavakol and D. Shahabi, J. Phys. Chem. C, 2015, 119, 6502–6510.
48. H. Tavakol and A. Mollaei-Renani, Struct. Chem., 2014, 25, 1659–1667.
49. S. Yang, L. Zhi, K. Tang, X. Feng, J. Maier and K. Müllen, Adv. Funct. Mater., 2012, 22, 3634–3640.
50. H. L. Poh, P. Šimek, Z. Sofer and M. Pumera, ACS Nano, 2013, 7, 5262–5272.
51. H. Tavakol and F. Hassani, Struct. Chem., 2015, 26, 151–158.
52. H. Tavakol and F. Hassani, Sens. Actuators, B, 2014, 196, 624–630.
53. R. G. Parr, L. V. Szentpaty and S. Liu, J. Am. Chem. Soc., 1999, 121, 1922–1924.
54. M. J. Frisch, G. W. Trucks and H. B. Schlegel, et al., Gaussian 09 (Revision A.02), Gaussian Inc, Wallingford, CT, 2009.
55. H. Tavakol, Int. J. Quantum Chem., 2011, 111, 3717–3724.
56. H. Tavakol, J. Mol. Struct.: THEOCHEM, 2010, 956, 97–102.
57. T. Yanai, D. P. Tew and N. C. Handy, Chem. Phys. Lett., 2004, 393, 51–57.
58. R. Kobayashi and R. D. Amos, Chem. Phys. Lett., 2006, 420, 106–109.
59. M. J. Frisch and J. A. Pople, J. Chem. Phys., 1984, 80, 3265–3269.
60. R. I. Masel, Chemical Kinetics and Catalysis, Wiley-Interscience, New York, 2001.
61. D. C. Higgins, M. A. Hoque, F. Hassan, J. Y. Choi, B. Kim and Z. Chen, ACS Catal., 2014, 4, 2734–2740.
62. W. Li, D. Yang, H. Chen, Y. Gao and H. Li, Electrochim. Acta, 2015, 165, 191–197.
63. A. M. El-Sawy, I. M. Mosa, D. Su, C. J. Guild, S. Khalid, R. Joesten, J. F. Rusling and S. L. Suib, Adv. Energy Mater., 2016, 6, 1501966.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra11447d

---