Optimizing the carbon coating on LiFePO₄ for improved battery performance

Abstract

Core–shell structures as LiFePO₄@carbon with a continuous and uniform carbon coating were achieved by means of the in situ polymerization of dopamine. Systematic control of the coating layer identified that a 5 nm carbon coating produces the best battery performance. Our results provide conclusive evidence for an optimal carbon coating for polyanion-type cathode materials.

---

The surface modification of polyanion-type cathode materials has been known as an indispensible step for their application in lithium ion batteries (LIBs).^1–3 A carbon coating layer has proved to be highly favorable to combat the inherently low electron conductivity of LiFePO₄ (LFP). Interestingly, despite the fact that contributions from carbon have been widely-confirmed,^4–6 little attention has been paid to systematic investigation on the surface coating layers. Therefore, the synthesis efforts for better cathode materials usually become a tedious process of trials and efforts. It's highly demand that a critical condition of the coatings layer be identified,⁷ so that an explicit guidance can be available for the surface control of electrode materials. For example, a recent research on silicon anode confirms that a 7 nm of silica coating turns out to be the most favorable thickness for its application in LIBs.⁷

To achieve a fair assessment on the coating effect of carbon, a uniform and continuously-distributed surface layer should be guaranteed,⁸ which has been a challenge for the surface treatment of cathode materials. It necessitates a homogeneous formation of carbon species right on the particle surface without damaging the pristine materials. In this way a model system can be built and the surface layer can be readily tuned to achieve an optimal electrochemical performance. As a matter of fact, such a continuous and uniform carbon coverage should also be favorable for application of cathode materials. The effective protection of carbon coating can prevent the direct contact between the active material and the electrolyte during the highly-unstable cycling process,⁹ avoiding the capacity fading and safety problems.^10,11 Herein we present a facile method to coat LiFePO₄, a typical polyanion-type cathode material we selected for demonstration, with a highly uniform and thickness-controllable carbon layer. A typical core–shell structure as LiFePO₄@carbon (LFP@C) can be prepared nicely and easily with the coating thickness being precisely controlled. Further investigation reveals that the battery performances of these cathode materials are directly relative to the surface coating layer. An optimal coating thickness for an improved electrochemical performance is thus determined.

The formation of LFP@C composites is schematically illustrated in Scheme 1. Dopamine is known as a bio-inspired building block for surface coating and it was proposed that the coexistence of catechol and amine groups make it crucial to achieve a good adhesion to different kinds of materials.¹² The continuous oxidation and self-polymerization of dopamine form thin films of polydopamine^13,14 and can adhered onto the LFP particles. In this way, a uniform coating can form with no phase separation between polydopamine and LFP nanoparticles. The surface polydopamine is then transferred into conductive carbon by high temperature under a reductive atmosphere.

Scheme 1 Schematic illustration of the synthesis of LFP@C composites and the coating mechanism of dopamine.

Fig. 1a shows a transmission electronic microscopy (TEM) image of the pristine LFP particles. The high resolution TEM (HRTEM) characterization on a randomly-selected particle reveals that it is highly crystalline with a free surface borderline (inset in Fig. 1a). After a two-step process, which briefly includes the above-mentioned polydopamine coating and its following graphitization, a uniform coating layer can form. Fig. 1b shows a typical TEM image of the coated sample (denoted as LFP@C1), which is synthesized when a low dopamine concentration (5.27 mM) is used in the polymerization step. Obviously, LFP surfaces have changed and a tiny thin film shows up. A thorough TEM examination on the particles didn't find any separate carbon particles. In benefit of the special growth mechanism of polydopamine, the carbon species have been successfully attached onto the LFP surface to form a core–shell structure. HRTEM observation shows that the coating layer is very uniform on the thickness, which is measured to be around 2.3 nm as shown in the inset in Fig 1a. Despite of its thinness, the coating turns out to be continuous and covers all the LFP nanoparticles.

Fig. 1 (a–e) TEM images of pristine LFP (a), LFP@C1 (2 nm coating) (b), LFP@C2 (5 nm coating) (c), LFP@C3 (8 nm coating) (d) and LFP@C4 (15 nm coating) (e); (f) XRD patterns of LFP, LFP@dopa2, and LFP@C2, respectively.

Further investigation shows that the carbon thickness can be easily changed by simply adjusting the concentrations of the starting materials or the length of the polymerization time. For example, when we double the concentration of dopamine to 10.54 mM while halved the amount of LFP, thicker carbon layer can be achieved as shown in Fig. 1c (sample denoted as LFP@C2). HRTEM image confirms the formation of ca. 5 nm surface layers (inset in Fig. 1c). A close look on the surface carbon shell reveals the existence of irregular lattice fringes of carbon due to the graphitization process. Raman characterization of LFP@C2 also confirms the emergence of carbon in the coated sample. As shown in Fig. S1,† the peak at 1593.07 cm⁻¹ and 1345.60 cm⁻¹ are Raman fingerprints for the G and D band¹⁵ of carbon respectively. It shows that the graphitization process at a medium temperature of 750 °C has formed partially graphitic carbon.

Similarly, 52.73 mM of dopamine and 38.03 mM LFP can produce an 8 nm carbon layer as shown in Fig. 1d (LFP@C3). Following this protocol, we can produce an even thicker coating as long as enough dopamine is supplied for the surface growth. Meanwhile, it is also possible to form even thicker surface layer by using a seeded-growth methodology. Fig. 1e shows a 15 nm carbon coating when we use a pre-synthesized LFP@polydopamine sample (denoted as LFP@Dopa) is used as seeds for a second growth. According to our detailed TEM observations, all four levels of thickness for LFP@C1–4 are highly uniform and continuous on all the particles. No separate carbon spheres have been found in the tested samples, indicating that the carbon species are all for coating with no obvious phase separations.

We also did different characterizations for a better understanding of the polymerization process. For an LFP@Dopa sample collected right after polymerization, Fourier-transfer infrared (FT-IR) spectroscopy clearly confirms the formation of polydopamine¹⁶ (Fig. S2†). TEM characterization on different LFP@Dopa samples locate these newly-formed polymers on the surface of LFP nanoparticles, while having a much thicker shell as compared to their carbonized counterparts. As shown in Fig. S3,† the surface polydopamine layers are around 3.8 nm, 9 nm, 18 nm, and 25 nm in thickness, respectively. Obviously, the graphitization at high temperature has compressed the surface layer due to the mass loss during high temperature treatment. X-ray diffraction (XRD) characterizations on samples of LFP, LFP@Dopa2 and LFP@C2 shows that no discernible change on crystalline structure has occurs (Fig. 1f). All these three samples show almost identical patterns, whose peaks can be well indexed to an orthorhombic space group of LFP powder (JCPDS no. 81-1173). Our further scanning electron microscopy (SEM) characterization shows that there is no obvious change on the shape and size of the LFP nanoparticles after the coating process (Fig. S4a–e†).

We consider these surface-controlled samples a good model system to study the coating effect inspired by its unprecedented uniformity of carbon layer on LFP surface. Detailed electrochemical characterizations are then carried out and Fig. 2a shows typical charge/discharge curves for different samples we tested. The lacking of carbon in the pristine LFP sample leads to a discharge capacity of only 62.2 mA h g⁻¹ even at a low discharge current of 0.1C (1C = 170 mA g⁻¹). Its discharge voltage plateau of the pristine LFP is also lower than the equilibrium voltage of 3.4 V. As expected, the coating strategy can effectively alleviate the conductivity issue and even a 2.3 nm carbon coating in LFP@C1 gives a discharge capacity of 141 mA h g⁻¹ at 0.1C. Meanwhile, the flat discharge plateau at 3.4 V can be totally revealed due to the existence of the surface coating layer, indicating a much reduced polarization during electrochemical process. In our test, a 5 nm carbon shell coating in LFP@C2 shows an improved discharge capacity of 147.9 mA h g⁻¹. However, further increase in coating thickness cannot bring more benefits as far as the battery performance is concerned. For example, the LFP@C4 sample, which has a very thick carbon shell of 15 nm, only shows a discharge capacity of 28 mA h g⁻¹ at 0.1C. Rate capabilities of these coating-controlled electrodes of LFP@C1–4 have also been tested. As shown in Fig. 2b, the LFP@C2 sample shows an obvious advantage over the other ones.

Fig. 2 Electrochemical performance of the samples: (a) comparison of the first charge/discharge curves of LFP, LFP@C1–4 at 0.1C rate, (b) comparison of the rate-capability among LFP, LFP@C1–4 and LFP-C, (c) comparison of the cyclability between LFP@C2 and LFP, (d) EIS spectra for LFP and LFP@C1–4 with the frequency range of 100 kHz to 100 MHz.

To have a better understanding on the coating effect, the experiment by using electrochemical impedance spectroscopy (EIS) is carried out on all the electrodes we tested. Fig. 2d shows the Nyquist plots for different thickness-controlled samples. The radius of the semicircle in the high-to-medium frequency range indicates the charge transfer and surface film resistance (R_ct), which is related to lithium ion interfacial transfer between the electrolyte and the active material.¹⁷ The inclined line in the low frequency region corresponds to the Warburg impedance, which is related to Li-ion diffusion within the LFP bulk materials.¹⁸ For the four LFP@C samples, the lithium-ion's diffusions inside the bulk material are similar, but the diffusions through the interface between LFP and electrolyte are different from each other, which could be reflected by R_ct. Among all the four samples with thickness-controlled coatings, LFP@C2 has the least value of R_ct than other three. It suggests that the charge-transfer interface reaction occurs more easily with the carbon coating thickness of 5 nm than with other thicknesses, which also explains well its lowest polarization or its superior electrochemical performance. Therefore, a 5 nm thick carbon layer exhibits the optimal balance between electronic conductivity and Li⁺ ionic diffusion for as-used LFP. This result is in good agreements with the rate-capability data obtained from battery performance tests. It is also noteworthy that the core–shell structured sample shows much better battery performance as compared to a simple mixture of LFP and carbonized polydopamine (denoted as LFP-C).

The core–shell structured sample turns out to be pretty stable during cycling. Fig. S4f† shows a TEM image of LFP@C2 after being cycled for 90 times. The carbon layer is almost the same as that of a fresh one before cycling. We can still observe a continuous and uniform coating on the LFP particles. There is no detectable shedding of carbon layer and the thickness remains around 5 nm as revealed by the HRTEM image (inset in Fig. S4f†). Our battery tests also disclose that the LFP@C2 sample has good cyclability. As shown in Fig. 2c, there is no obvious fading of discharge capacity after 50 cycles, and the coulombic efficiency was around 100%, which is promising for their practical applications in lithium ion batteries.

In summary, a core–shell structure of LFP@C has been synthesized with help of the in situ polymerization of dopamine. The carbon shell is very uniform on the surface with its thickness being readily tuned at different levels. These surface-controlled samples of LFP@C are then used as a model system to evaluate the coating effect of carbon. Electrochemical characterizations reveal that a ca. 5 nm carbon coating on LFP obtains the best battery performance, probably due to a good balance between electronic conductivity and Li⁺ transport. Our results provides explicit conclusion on an optimized surface carbon layer on LFP sample. The experiment could also offer referential information for improving the electrochemical performance of other kinds of electrode materials.

Acknowledgements

This work was supported by the major State Basic Research Program of China (973 program: 2013CB934000, 863 program: 2013AA030800), the National Natural Science Foundation of China (Grant no. 21373238), and the Chinese Academy of Sciences.

Notes and references

1. Y. Wang, Y. Wang, E. Hosono, K. Wang and H. Zhou, Angew. Chem., Int. Ed., 2008, 47, 7461.
2. D. Wang, H. Buqa, M. Crouzet, G. Deghenghi, T. Drezen, I. Exnar, N.-H. Kwon, J. H. Miners, L. Poletto and M. Grätzel, J. Power Sources, 2009, 189, 624.
3. S.-M. Oh, S.-T. Myung, J. B. Park, B. Scrosati, K. Amine and Y.-K. Sun, Angew. Chem., 2012, 124, 1889.
4. S. L. Bewlay, K. Konstantinov, G. X. Wang, S. X. Dou and H. K. Liu, Mater. Lett., 2004, 58, 1788.
5. A. A. Salah, A. Mauger, K. Zaghib, J. B. Goodenough, N. Ravet, M. Gauthier, F. Gendron and C. M. Julien, J. Electrochem. Soc., 2006, 153, A1692.
6. X. Liao, Z. Ma, L. Wang, X. Zhang, Y. Jiang and Y. He, Electrochem. Solid-State Lett., 2004, 7, A522.
7. S. Sim, P. Oh, S. Park and J. Cho, Adv. Mater., 2013, 25, 4498.
8. C. Lei, F. Han, D. Li, W. C. Li, Q. Sun, X. Q. Zhang and A.-H. Lu, Nanoscale, 2013, 5, 1168.
9. H. Li and H. Zhou, Chem. Commun., 2012, 48, 1201.
10. W. Zhang, X. Wu, J. Hu, Y. Guo and L. Wan, Adv. Funct. Mater., 2008, 18, 3941.
11. X. W. Lou, C. M. Li and L. A. Archer, Adv. Mater., 2009, 21, 2536–2539.
12. H. Lee, S. M. Dellatore, W. M. Miller and P. B. Messersmith, Science, 2007, 318, 426.
13. A. Postma, Y. Yan, Y. Wang, A. N. Zelikin, E. Tjipto and F. Caruso, Chem. Mater., 2009, 21, 3042.
14. S. M. Kang, J. Rho, I. S. Choi, P. B. Messersmith and H. Lee, J. Am. Chem. Soc., 2009, 131, 13224.
15. W. Li, D. Chen, Z. Li, Y. Shi, Y. Wan, G. Wang, Z. Jiang and D. Zhao, Carbon, 2007, 45, 1757.
16. J. Jiang, L. Zhu, X. Li, Y. Xu and B. Zhu, J. Membr. Sci., 2010, 364, 194.
17. Z. Qin, X. Zhou, Y. Xia, C. Tang and Z. Liu, J. Mater. Chem., 2012, 22, 21144.
18. J. Wolfenstine, U. Lee and J. L. Allen, J. Power Sources, 2006, 154, 287–289.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details. See DOI: 10.1039/c3ra47702a

---