Synthesis and extensive characterisation of phosphorus doped graphite

Abstract

The pyrolysis of 1,2-diphosphinobenzene at 800 °C gives a phosphorus-doped graphite (P-DG) with an unprecedented high phosphorus content, ca. 20 at%. In contrast with previously studied boron and nitrogen doped graphite materials, thorough characterisation and analysis of this material demonstrates that it is extensively disordered and contains substitutional P-atoms along with P=O units in the host graphitic lattice, as well as P₄ molecules trapped between the graphitic sheets. This represents a stabilised form of P₄, which has been shown to covalently bind to lithium as Li₃P in this material.

---

1. Introduction

In recent years the drive to create novel materials from simple, Earth-abundant and inexpensive resources has led to a renewed focus on the chemistry of lighter main group elements of the periodic table. The discoveries in the 1990s and 2000s of new allotropes of carbon have reignited research into carbon-based materials. Much of this interest has focused on low-dimensionality materials in the form of graphene, carbon nanotubes and buckminsterfullerenes. On the other side of the spectrum is graphite, which has long been the basis for the development of a host of new materials. One of the major driving forces in current research on graphite is energy storage, particularly its uses in battery electrodes, hydrogen storage media or electrodes for oxygen reduction reactions (water splitting) and thus the production of hydrogen.^1–5

While the majority of the research on graphite has been concerned with intercalation between the graphitic sheets as a way of tuning properties,^6,7 recent studies have looked primarily at the inclusion of boron or nitrogen substitutionally into graphitic sheets.^8–12 Only limited attempts to dope phosphorus into the graphite lattice have been reported; even though this offers a particularly attractive option for novel lithium anodes.

Phosphorus has the potential to be a high capacity anode in a lithium battery: lithium is stored as Li₃P, corresponding to a theoretical capacity of 2594 mA h g⁻¹, which is nearly 7 times greater than commercially-used graphite.^13,14 However, poor conductivity limits the real capacity of elemental phosphorus, making it commercially undesirable. Previous studies have attempted to decorate the surface of mesoporous carbon in order to improve conductivity and make the phosphorus more stable, or dealt with low dimensionality materials (nanotubes/graphene) however few have attempted phosphorus inclusion in the graphitic lattice, which is what we present here.^15–19

A second application of phosphorus-doped carbon based materials is as metal-free anodes for the oxygen reduction reaction (ORR).^3,18,20–24 The ORR is the key step for the generation of energy in hydrogen fuel cells, which has been touted as one way of reducing/replacing fossil fuels in automotive engines. It is the reaction used in fuel cells to reduce oxygen electrochemically and oxidise hydrogen to generate electricity (producing water as the by-product). Traditionally, the ORR has been catalysed by platinum electrodes, however, these suffer from problems such as CO poisoning (CO binding and blocking the Pt active sites), and most importantly, the prohibitively high cost of Pt metal (∼£22 000 kg at time of writing).^25–28 While there have been some promising advances made using non-precious metal catalysts, a lot of attention has been directed towards carbon-based materials, including doped graphites.²³ Liu et al. have prepared P-DG layers via the pyrolysis of triphenylphosphine and toluene as well as P-doped carbon nanospheres and multi-walled carbon nanotubes (MWCNTs), that have shown good electrocatalytic performance for the ORR.^21,29,30

Synthetic routes to phosphorus-doped graphites (P-DGs) have traditionally involved chemical vapour deposition (CVD) processes utilising triphenylphosphine (TPP) as the source of phosphorus, in conjunction with a carbon source such as toluene. These methods produce phosphorus doped carbons (P-DG) but are limited to phosphorus contents of <6 at%.^20,21,29,31 The CVD approach has also been reported for the production of P-doped nanotubes^22,32,33 and graphene,³⁴ though these have both a low P content and have differing properties to bulk graphite and so are not discussed in detail here.

We have reported previously that bulk samples of the stoichiometric N- and B-doped graphites C₃N and C₃B can be obtained efficiently via the thermolysis of the single-source precursors 1,3-(NH₂)₂C₆H₄ and 1,3-(BBr₂)₂C₆H₄, respectively.^8,9 In this paper we show that this method can also be used to obtain highly doped P-DG with a phosphorus content of ca. 20 at% from the pyrolysis of 1,2-diphosphinobenzene (Fig. 1). Detailed analysis of this P-DG suggests that it is heavily disordered, containing intercalated oxidised phosphorus species as well as some P=O and P substitution within the graphitic lattice, and, importantly, P₄ intercalation between the graphitic sheets.

Fig. 1 Synthesis of the P-DG via pyrolysis of 1,2-diphosphinobenzene.

2. Experimental

2.1 Synthesis of P-DG

In a typical synthesis, 0.15 ml (0.99 mmol) of the precursor 1,2-diphosphinobenzene³⁵ was heated to 800 °C in a sealed quartz tube of 11 mm diameter and 120 mm length under static vacuum and held at this temperature for 1 hour prior to cooling to room temperature. The reaction tube was opened with a diamond saw in air. Initial air exposure resulted in the spontaneous combustion of part of the material, burning with a bright white flame, which suggests the formation and exclusion of residual white phosphorus (P₄) and/or PH₃. The black solid produced was then washed with acetone and matured in hydrofluoric acid (10 wt%, 5 ml overnight, to remove residual Si from the surface) before finally being oven-dried. This material is air stable, even after grinding the samples in air appear not to degrade over time, as no change in the physical characteristics is observed.

2.2 Characterisation of P-DG

The P-DG was investigated extensively using scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDX), X-ray photoelectron spectroscopy (XPS), powder X-ray diffraction (pXRD), Raman spectroscopy, matrix-assisted laser desorption/ionization-time of flight mass spectroscopy (MALDI-TOF), static and MAS (Magic Angle Spinning) ³¹P solid-state nuclear magnetic resonance (NMR) spectroscopy. Resistivity measurements were also undertaken. For full details see ESI.†

3. Results and discussions

3.1 Characterisation of P-DG

SEM images show that the P-DG consists of large flakes that are ∼5 μm thick and several hundred μm² in area (Fig. 2). The faces of each flake are relatively smooth. Occasional extrusions from the surface are observed and are probably produced during graphitization by escaping gases migrating to the surface. We have shown previously that this process can also lead to the formation of multi-walled graphitic microspheres (CMS).⁸

Fig. 2 SEM images of P-DG showing large flakes with a consistent width of ∼5 μm.

Although EDX is of limited use for the quantitative determination of low atomic mass elements, it was of value in showing the presence of carbon and phosphorus in the P-DG samples unambiguously. We turned to XPS in order to determine the C : P ratio (Fig. 3). Argon ion beam milling shows that this ratio decreases as a function of depth. At the surface of the P-DG the ratio is up to 2 : 1, however, this moves towards a value of 6 : 1 (C : P) in the bulk (see Fig. 3b). The overall ratio of C : P is C_4.17P, i.e. the graphite is doped with P atoms to 19.3 at%.

Fig. 3 (a) X-ray photoelectron spectroscopy (XPS) survey scan. (b) Atomic composition determined by X-ray photoelectron spectroscopy (XPS) as a function of depth. The elements present are: carbon (black), oxygen (blue), phosphorus (red). Etch levels indicate number of times Ar ion beam applied to the same spot. (c) High resolution XPS P 2p scan after Ar milling with 3 types of phosphorus environments: P=O (blue) 78%, elemental P (green) 4% and PPh₃ (red) 17% with orange the overall curve fit and black the experimental data. (d) High resolution XPS C 1s scan after Ar milling with 4 different carbon environments: sp² C (red), sp³ C (purple), C–O (blue) and C–P (green), with orange the overall fit and black the experimental data.

A substantial oxygen component is also observed by XPS, which decreases with depth, consistent with the presence of surface oxidation (Fig. 3b). As expected, high-resolution scans of the XPS spectrum in the P 2p region of the initial product of the pyrolysis reaction after air-exposure, and prior to washing, show high levels of P₄O₁₀ (ca. 95%, an expected product of the observed combustion of the residual elemental phosphorus and/or PH₃ in air). Adventitious contamination with oxygen also causes a significant background oxygen level in samples, as shown by XPS analysis of graphite alone.³⁶ This makes it difficult to accurately quantify the at% of O in the sample.

XPS analysis shows that surface P₄O₁₀ is efficiently removed on washing samples further with acetone (see ESI Fig. S3†). Only two phosphorus environments are seen at the surface after treatment with acetone/HF. Judging by their XPS binding energies, these appear to be P=O-like (ca. 80%, i.e., oxidised P-atoms) and elemental phosphorus-like (ca. 5%). These two environments are present throughout the depth in the sample, as evidenced by XPS depth analysis using argon ion beam ablation (ESI Fig. S5†).

A further environment is also found in the bulk material (Fig. 3c), a phosphine-like environment which presumably results from P-atom substitution in to the graphitic lattice (ca. 15%), however this may also be assigned to a mixture of O_xP–C type environments.^18,37–39 A depth profile of the P environments in the P-DG (ESI Fig. S5†) demonstrates no real trend, though it is interesting to note that whilst the amount of elemental P remains relatively constant, the P–C type environment and the P=O one are interlinked, potentially indicating that this is the major species that is being oxidised.

Deconvolution of the C 1s spectra (Fig. 3d) is very revealing. There are four major environments: sp² C, sp³ C, C–O and evidence for C–P bonds, which is in agreement with the P 2p scan. However, this latter environment is not present throughout the sample – different scans at different points on the sample and at different etch levels show some or no evidence for C–P bonds (see ESI Fig. S4†). This suggests that the C–P component is small in the P-DG, but present.¹⁹

The overall graphiticity of the P-DG was demonstrated by both powder XRD (pXRD) and Raman spectroscopy. The increased D (1330 cm⁻¹) to G (1560 cm⁻¹) band ratio in the Raman spectrum indicates a high level of disorder in the material (see Fig. 4). The broad, asymmetric peaks of the pXRD pattern can be attributed to the turbostratic nature of the material: it exhibits good 2D order, but limited 3D order. The peaks at 26° and 44° correspond to the (002) and (101) planes, respectively, with the former indicating a slightly increased interlayer spacing of 3.69 Å compared to graphite (3.36 Å).⁴⁰

Fig. 4 (a) Raman spectrum of P-DG (red) and graphite (black). The D and G bands are seen at 1330 and 1560 cm⁻¹. (b) A powder XRD scan of P-DG shows the turbostratic nature of the material. The peaks at approximately 26° and 44° are the (002) and (101) planes, with the (002) spacing seen to be 3.69 Å.

Further evidence for the presence of phosphorus covalently bonded in the P-DG was provided by MALDI-TOF. In particular, it is clear from this analysis that the P-DG has a distinctly different ionisation ‘fingerprint’ to graphite. Fig. 5 compares the laser power required for ionization of the same carbocation fragments (C_n⁺) for graphite and P-DG. A significantly greater amount of energy is required for a given fragment of the P-DG compared to graphite. If phosphorus or phosphorus-containing species were simply decorating the surface or intercalated exclusively between the sheets of graphite then the same energy signature should be obtained for a specific ion, which is not the case here.

Fig. 5 MALDI-TOF analysis of required laser energy to make the fragments fly. m/z of 72 (dots), 120 (dashes) and 132 (line) are shown for graphite (black) and P-DG (red).

The ionisation of P-DG generates peaks corresponding to C_n, ³¹P_n and C_nP families. Strong signals attributable to phosphorus clusters appear intermittently; a species with m/z = 155, corresponding to ³¹P₅, being particularly characteristic. This strongly suggests the formation of small grains of phosphorus embedded within the carbon framework, in addition to the previously noted P–C bonded fragments.

Using positive MS-MS mode, a small number of peaks corresponding to C_nP have been identified and broken down with the aid of collision induced dissociation (CID) gas to form other C_nP and C_n peaks (see ESI Fig. S6†). Combining this with the ionisation fingerprint, it can be concluded that there is some degree of C–P bond retention and/or formation, the P-DG is not just P decorated on a carbon matrix.

Resistivity measurements were also carried out on flakes of the P-DG (Fig. 6) using a standard four-point lock-in amplification technique from 300 down to 2 K. The resistivity at room temperature was determined to be 0.0029 Ω m, and rises with falling temperature. P-DG is therefore likely to be a semiconductor material; for T ≥ 240 K the resistivity obeys an Arrhenius activation law indicating a band gap of 16 meV. At low temperatures a three-dimensional variable-range hopping model⁴¹ is more appropriate, suggesting a strongly disordered system with localised charge and some cross-linking of phosphorus atoms between the graphitic layers.

Fig. 6 (a) Ratio of the resistivity ρ(T) to its room-temperature value. Inset is Arrhenius plot for T ≥ 240 K. Red markers are data; black line is linear fit, indicating a band gap of 16 meV. (b) The same data at low temperature. Red markers are data; black line is best fit to a variable-range hopping model in three dimensions. Inset is natural logarithm of the resistivity against T^−1/4, showing good agreement with the variable-range hopping model.⁴¹

Solid state NMR (ssNMR) spectroscopy, although often neglected in the study of doped graphites, is a very powerful tool for understanding the environments within the P-DG. 1D ³¹P and ¹H NMR experiments and 2D ³¹P exchange spectroscopy have been used to study the pristine material. Fig. 7 shows Magic Angle Spinning (MAS) ³¹P NMR spectra of pristine P-DG at two different magnetic fields. The spectra contain two major signals at −521 ppm, corresponding to P₄, and a broad signal between 0 and 500 ppm. There are also minor signals at −445 ppm and between −30 and 30 ppm. Comparing ³¹P and ¹H NMR spectra of different pristine batches, we find that these signals correlate with the presence of residual water between 8 and 11 ppm in ¹H NMR (ESI Fig. S7†). Therefore, the −445 ppm signal is tentatively assigned to P₄ in contact with water, and the narrow signals between −30 and 30 ppm to phosphate species dissolved in residual water. The presence of P₄ molecules trapped between the graphitic layers can be seen as a solid-state analogue of the host–guest incorporation of P₄ in supramolecular chemistry.⁴²

Fig. 7 ³¹P MAS NMR spectra of pristine P-DG at (a) 16.4 T and (b) 4.7 T. The insert in (a) is an expansion of the −70 to 70 ppm region for the pristine material. Spinning sidebands are marked with a star.

The line shape of the broad signal between 0 and 500 ppm is essentially independent of the magnetic field, which is typical of inhomogeneous line broadening. The low-field spectrum (Fig. 7b) reflects a shoulder at 210 ppm. Indeed, ³¹P NMR spectra of other P-DG batches suggest that the broad resonance between 0 and 500 ppm consists of three overlapping line shape components at 20, 210 and 370 ppm (see Fig. S7, ESI†). The diagonally elongated shape of these three overlapping signal components in 2D ¹³P NMR spectrum illustrate the inhomogeneous lineshape (ESI Fig. S8a†). The shift and line width of the peak at 20 ppm is similar to that of red phosphorus.⁴³ The two other components are consistent with phosphorus embedded in the paramagnetic carbon matrix.

From all the characterisation data for P-DG discussed above, it is apparent that the material does not behave in a manner analogous to the boron- and nitrogen-doped graphite analogues, where a carbon atom is substituted for a dopant, with retention of sp² hybridisation. Using the complete characterisation data collected by an extensive array of techniques, we have now elucidated the structure of P-DG. Raman spectroscopy, pXRD, XPS and ssNMR all indicate a high degree of disorder and inhomogeneity of the P-DG, so it would unsuitable to propose a periodic global structure for this material. However, it is possible to look at the material on the macro-, nano- and atomic-scales. At the macro-scale, the P-DG consists of large graphitic flakes of several hundred μm² in area and ∼5 μm thick. These flakes are formed of turbostratically arranged nanocrystalline domains, that show substantial 2D ordering, but limited 3D order.

On the atomic-scale there are a number of different environments for both C and P. XPS reveals that carbon is present as sp² C (most likely in C₆ rings), sp³ C (hydrogenated or termination of flakes), oxidised to C–O and forming C–P bonds (Fig. 8, showing the atomic scale environments within P-DG). Phosphorus is found predominately as either elemental P or in an oxidised state, but XPS and MALDI-TOF mass spectrometry also demonstrate it bound to C. This data is in agreement with that observed in the ³¹P ssNMR spectra. P forms very strong bonds to oxygen (P=O 560 kJ mol⁻¹)⁴⁴ and so the heavy degree of oxidation can be explained by initial air exposure, the aqueous work up and from scavenging any residual oxygen from the reaction apparatus.

Fig. 8 Idealised drawings of the different atomic-scale environments that are found in P-DG, including: sp² C, sp³ C as well as C–P and C–O bonds, intercalated P₄ units, intercalated red P, PPh₃-type environment and R₃P=O.

3.2 Li binding to P-DG

The method of Li binding to P-DG was assessed by electrochemical insertion of Li. In graphite Li intercalates between the layers; here the presence of elemental P offers the opportunity for covalent binding as Li₃P.

The ³¹P ssNMR of a lithiated sample reveals a new signal at −250 ppm (Fig. 9a and c). This signal, which appears to grow at the expense of the P₄ peak at −500 ppm, has previously been assigned to that of Li₃P. This environment is also seen in the ⁷Li ssNMR at 7.6 ppm (Fig. 9b and d).¹⁵ The field-independent width (in ppm) of this signal indicates Li sites with symmetric tetrahedral or octahedral coordination. This is consistent with the crystal structure of bulk Li₃P, which contains two types of Li in a tetrahedral coordination.⁴⁵ The remainder of the ⁷Li ssNMR spectra reflect the presence of a predominant fraction of LiPF₆ salt combined with oxidized species such as Li₂PO₂F₂ Li₂O, LiOH and Li_xH_3−xPO₄ species (−0.6 ppm). This signal is broader at lower magnetic fields, consistent with quadrupolar line broadening caused by local electric field gradients at Li positions.

Fig. 9 (a and c) ³¹P and (b and d) ⁷Li MAS NMR spectra of (black) pristine and (red) lithiated P-DG at (a and b) 16.4 T and (c and d) 4.7 T. The new signal arising from lithiation is marked with a vertical arrow, and spinning sidebands with a star. The two ⁷Li NMR spectra are height-normalized w.r.t. the 7.6 ppm signal, whose line shape is practically field independent.

¹⁹F ssNMR spectroscopy confirms the presence of LiPF₆ electrolyte in the lithiated P-graphite, along with oxidised electrolyte species and polytetrafluoroethylene (PTFE) binder (ESI Fig. S8†).

There is a 0.7 V difference in voltage between the insertion and removal plateaus (ESI Fig. S10†), higher than would be expected for Li intercalation in carbons. This overpotential is characteristic of the formation of a lithium compound with covalent character. These results are consistent with the formation of the Li₃P seen in the ssNMR spectra (Fig. 9). A similar profile has been reported for phosphorus decorated on the surface of carbon.¹⁵ No further electrochemical tests were carried out, as the P-DG rapidly degraded upon repeated cycling. These data indicate that the lithium is being covalently bound to the phosphorus of the P-DG as Li₃P.

4. Conclusions

In summary, we have demonstrated a thorough characterisation of a novel phosphorus–carbon graphitic material. We have determined that phosphorus does not behave in a manner analogous to boron- and nitrogen-doped graphites, in that it only partially replaces carbon substitutionally within graphitic layers. This is almost certainly caused by the substantial atomic radius increase relative to carbon. This exclusion from the graphitic lattice has been used to generate stabilised elemental phosphorus that is encapsulated within the graphitic matrix. Furthermore we have found that lithium binds covalently to the embedded phosphorus within this material, which has potential energy storage applications.

Acknowledgements

The authors thank the EPRSC (P. D. M., T. C. K.), the EPSRC Cambridge NanoDTC, EP/G037221/1 (H. G., J. A. C) and the ERC (Advanced Investigator Award for DSW) for financial support. We also thank Keith Parmenter for glass blowing. We thank Evan N. Keyser for many fruitful discussions.

Notes and references

1. Y. Su, Y. Zhang, X. Zhuang, S. Li, D. Wu, F. Zhang and X. Feng, Carbon, 2013, 62, 296–301.
2. Y. Zhang, X. Zhuang, Y. Su, F. Zhang and X. Feng, J. Mater. Chem. A, 2014, 2, 7742–7746.
3. D. Yu, Y. Xue and L. Dai, J. Phys. Chem. Lett., 2012, 3, 2863–2870.
4. W. Wei, H. Liang, K. Parvez, X. Zhuang, X. Feng and K. Müllen, Angew. Chem., Int. Ed., 2014, 53, 1570–1574.
5. I. Shimoyama, T. Hakoda, A. Shimada and Y. Baba, Carbon, 2015, 81, 260–271.
6. M. S. Dresselhaus and G. Dresselhaus, Adv. Phys., 1981, 30, 139–326.
7. T. E. Weller, M. Ellerby, S. S. Saxena, R. P. Smith and N. T. Skipper, Nat. Phys., 2005, 1, 39–41.
8. T. C. King, P. D. Matthews, J. P. Holgado, D. A. Jefferson, R. M. Lambert, A. Alavi and D. S. Wright, Carbon, 2013, 64, 6–10.
9. T. C. King, P. D. Matthews, H. Glass, J. A. Cormack, J. P. Holgado, M. Leskes, J. M. Griffin, O. A. Scherman, P. D. Barker, C. P. Grey, S. E. Dutton, R. M. Lambert, G. Tustin, A. Alavi and D. S. Wright, Angew. Chem., Int. Ed., 2015, 54, 5919–5923.
10. C. Zhang and A. Alavi, J. Chem. Phys., 2007, 127, 214704.
11. A. A. Kuzubov, A. S. Fedorov, N. S. Eliseeva, F. N. Tomilin, P. V. Avramov and D. G. Fedorov, Phys. Rev. B: Condens. Matter Mater. Phys., 2012, 85, 195415.
12. H. Wang, T. Maiyalagan and X. Wang, ACS Catal., 2012, 2, 781–794.
13. W. Li, Y. Yang, G. Zhang and Y.-W. Zhang, Nano Lett., 2015, 15, 1691–1697.
14. A. Ueda, M. Nagao, A. Inoue, A. Hayashi, Y. Seino, T. Ota and M. Tatsumisago, J. Power Sources, 2013, 244, 597–600.
15. C. Marino, L. Boulet, P. Gaveau, B. Fraisse and L. Monconduit, J. Mater. Chem., 2012, 22, 22713–22720.
16. C. H. Choi, S. H. Park and S. I. Woo, J. Mater. Chem., 2012, 22, 12107–12115.
17. R. Li, Z. Wei, X. Gou and W. Xu, RSC Adv., 2013, 3, 9978–9984.
18. C. Zhang, N. Mahmood, H. Yin, F. Liu and Y. Hou, Adv. Mater., 2013, 25, 4932–4937.
19. J. Sun, G. Zheng, H. W. Lee, N. Liu, H. Wang, H. Yao, W. Yang and Y. Cui, Nano Lett., 2014, 14, 4573–4580.
20. D. S. Yang, D. Bhattacharjya, S. Inamdar, J. Park and J. S. Yu, J. Am. Chem. Soc., 2012, 134, 16127–16130.
21. Z. Liu, F. Peng, H. Wang, H. Yu, W. Zheng and X. Wei, J. Nat. Gas Chem., 2012, 21, 257–264.
22. J. Wu, X. Zheng, C. Jin, J. Tian and R. Yang, Carbon, 2015, 92, 327–338.
23. L. Dai, Y. Xue, L. Qu, H.-J. Choi and J.-B. Baek, Chem. Rev., 2015, 115, 4823–4892.
24. C. H. Choi, S. H. Park and S. I. Woo, ACS Nano, 2012, 6, 7084–7091.
25. Johnson Matthey Platinum Price Charts, http://www.platinum.matthey.com/prices/price-charts (accessed November 2015).
26. C. Sealy, Mater. Today, 2008, 11, 65–68.
27. M. Winter and R. J. Brodd, Chem. Rev., 2004, 104, 4245–4269.
28. K. Gong, F. Du, Z. Xia, M. Durstock and L. Dai, Science, 2009, 323, 760–764.
29. Z.-W. Liu, F. Peng, H.-J. Wang, H. Yu, W.-X. Zheng and J. Yang, Angew. Chem., Int. Ed., 2011, 50, 3257–3261.
30. Z. Liu, F. Peng, H. Wang, H. Yu, J. Tan and L. Zhu, Catal. Commun., 2011, 16, 35–38.
31. Z. Jin, Z. Sun, L. J. Simpson, K. J. O'Neill, P. A. Parilla, Y. Li, N. P. Stadie, C. C. Ahn, C. Kittrell and J. M. Tour, J. Am. Chem. Soc., 2010, 132, 15246–15251.
32. I. O. Maciel, J. Campos-Delgado, E. Cruz-Silva, M. A. Pimenta, B. G. Sumpter, V. Meunier, F. López-Urías, E. Muñoz-Sandoval, H. Terrones, M. Terrones and A. Jorio, Nano Lett., 2009, 9, 2267–2272.
33. J. Campos-Delgado, I. O. Maciel, D. A. Cullen, D. J. Smith, A. Jorio, M. A. Pimenta, H. Terrones and M. Terrones, ACS Nano, 2010, 4, 1696–1702.
34. X.-K. Kong, C.-L. Chen and Q.-W. Chen, Chem. Soc. Rev., 2014, 43, 2841–2857.
35. E. P. Kyba, S. T. Liu and R. L. Harris, Organometallics, 1983, 2, 1877–1879.
36. M. A. Mannan, M. Nagano, K. Shigezumi, T. Kida, N. Hirao and Y. Baba, Am. J. Appl. Sci., 2008, 5, 736–741.
37. J. C. Han, a. P. Liu, J. Q. Zhu, M. L. Tan and H. P. Wu, Appl. Phys. A, 2007, 88, 341–345.
38. F. Claeyssens, G. M. Fuge, N. L. Allan, P. W. May and M. N. R. Ashfold, Dalton Trans., 2004, 3085–3092.
39. S. Some, J. Kim, K. Lee, A. Kulkarni, Y. Yoon, S. Lee, T. Kim and H. Lee, Adv. Mater., 2012, 24, 5481–5486.
40. G. E. Bacon, Acta Crystallogr., 1951, 4, 558–561.
41. N. F. Mott, Philos. Mag., 1969, 19, 835–852.
42. P. Mal, B. Breiner, K. Rissanen and J. R. Nitschke, Science, 2009, 324, 1697–1699.
43. A. Bytchkov, F. Fayon, D. Massiot, L. Hennet and D. L. Price, Phys. Chem. Chem. Phys., 2010, 12, 1535–1542.
44. A. H. Sehon and M. Szwarc, Annu. Rev. Phys. Chem., 1957, 8, 439–462.
45. Y. Dong and F. J. Disalvo, Acta Crystallogr., Sect. E: Struct. Rep. Online, 2007, 63, 97–98.

---

Footnote

† Electronic supplementary information (ESI) available: Detailed synthesis of precursor and material, along with further detail for SEM, XPS, MALDI-TOF, ssNMR, resistivity and Li binding measurements. See DOI: 10.1039/c6ra08639j

---