Vertically aligned iron oxide nanotube arrays and porous magnetite nanostructures as three-dimensional electrodes for lithium ion microbatteries

Abstract

The simple preparations of vertically aligned iron oxide nanotube arrays and porous magnetite nanostructures and their application as three-dimensional electrodes for lithium ion microbatteries are introduced. Nanotube arrays are well formed across a very large area with a ∼40 nm nanotube diameter. The nanotubes change to be porous nanostructures like nano-pillar arrays after heat treatment under 10% H₂. These two structures allow the omission of conductive agents and binder in the battery assembly. In addition, the structural advantages for lithium ion and electron transport are exhibited through excellent cyclic stability and rate capability even at very high current density, which can complete one charge/discharge in a few minutes.

---

Introduction

In recent years, the demands for high energy and power density materials have been growing fast. Among them, lithium ion batteries have been extensively investigated by many researchers to meet the increasing energy and power demands of portable electronics such as lap-top computers and cellular phones. They are also considered as power sources in hybrid vehicles and electric vehicles, where high rate capability becomes one of the key issues.¹

Graphite is currently used for the anode material due to its cyclic stability, high electronic conductivity, and low cost. Nevertheless, the theoretical capacity (372 mA h g⁻¹) of graphite is moderate and the practical capacity of graphite already approaches very close to the theoretical capacity. Moreover, the limited rate capability of graphite is one of the reasons to search for alternative anode materials.

After a report on the increased reactivity of nano-sized transition metal oxides, the materials have received much attention for advanced anode materials.² This conversion reaction has been applied not only to transition metal oxides but also to transition metal sulphides, nitrides, phosphides and fluorides.³ Among them, iron based materials have shown many advantages as an alternative anode, which include high capacity, low cost, and low toxicity. The main drawback of these materials is poor cyclic stability resulting from the large volume change during cycling, which is common for transition metal oxides, sulphides, nitrides, phosphides and fluorides. Many solutions have been proposed and developed to solve the problem. A solution is three-dimensional nanosturctured electrodes.⁴ They can endure the strain without pulverization and aggregation of materials during the cycling. Also, it is very advantageous in terms of energy density due to there being no necessity for binder and conductive agents, and also provides higher power density because the electrons are supplied directly from the current collector without passing additional interfaces. Furthermore, three-dimensional electrodes lead to higher surface areas and shorten the lithium ion path.

Metal oxide nanotube arrays directly connected on a metal foil such as aluminium, titanium and tantalum have been synthesized by the electrochemical anodization method.⁵ However, the successful fabrication of iron oxide nanotube arrays through anodization of iron foil was developed much later than those of aluminium and titanium oxide. They have been applied in water photooxidation and electrochemical capacitors.⁶

Herein, we report iron oxide nanotube arrays (IONAs) and porous magnetite nanostructures (PMNs) prepared by the anodization method as the anodes for lithium ion batteries. They are very attractive and interesting for the following two reasons. First, IONAs and PMNs both have convenient structures for alleviating volume expansion and inducing fast ion and electron transport. As a result, the electrochemical tests indicate that they have a superior cycle stability and rate capability. Second, the slurry-making step which is necessary in classical electrode manufacturing can be omitted.

Experimental section

Preparation of samples

Vertically aligned iron oxide nanotube arrays were fabricated by a simple sonoelectrochemical anodization method.^6a Anodization was performed with iron foil (Goodfellow, 99.99+%, 0.025 mm thick) in 250 mL of solution, which was composed of 0.5 wt% NH₄ and 3 vol% DI water in ethylene glycol, with the assistance of ultrasonification. In order to use the iron substrate as a current collector for battery tests, the back side of the very thin iron foil was mounted on a Teflon frame during anodization (Fig. 1). Porous magnetite nanostructures were prepared by annealing iron oxide nanotube arrays at 500 °C in 10% H₂ for 1 h.

Fig. 1 Photograph of the Teflon holder used for anodization. The backside of the iron foil is mounted on the Teflon frame to prevent reaction when using it directly as a current collector.

Characterization

X-Ray diffraction (XRD) patterns were obtained with a Rigaku D/Max-3C diffractometer (Cu-Kα radiation source, λ = 0.15418 nm). X-Ray photoelectron spectroscopy (XPS, SIGMA PROBE, ThermoFisher Scientific, UK) was performed using a multipurpose surface analysis system operating at base pressures <10⁻¹⁰ mbar. Al Kα (1486.6 eV) was used as the X-ray source. Field emission scanning electron microscopy (FE-SEM) investigations were carried out using a JSM-6700F, SU70 and SUPRA 55VP. Transmission electron microscopy (TEM) images were obtained on a JEOL EM-2010 microscope. Samples for TEM investigations were prepared by scratching the surface of the foil with a razor and then dispersing them in ethanol with the assistance of ultrasonification. After that, they were dropped on a copper TEM grid.

Electrochemical characterization

In order to use the prepared samples as a working electrode directly, they were dried under vacuum at 120 °C for 8 h before assembling the electrochemical cells. Coin cells (2016 type half cell) were assembled in an argon-filled glove box with lithium metal foil as the counter and reference electrode and 1.0 M LiPF₆ dissolved in ethylene carbonate and diethyl carbonate (1 : 1 in volume ratio) as the organic electrolyte. The voltage window for electrochemical testing was between 0.01 and 2.5 V (vs. Li/Li⁺).

Results and discussion

X-Ray diffraction (XRD) patterns of IONAs (as-anodized) and PMNs (10% H₂) are shown in Fig. 2. IONAs and PMNs have less crystallinity compared to annealed samples in air, oxygen, and argon. (Fig. S1, ESI†). The relatively strong peaks of IONAs indicate the presence of iron oxyfluoride and hematite. In contrast, only small peaks related to magnetite are observed in PMNs. The intensities of the iron oxyfluoride phase decrease when IONAs are calcined at 500 °C in air, oxygen, and argon. The phases of annealed samples in air, oxygen, and argon are clearly found to be mixtures of magnetite and hematite with different ratios. The inset of Fig. 2 shows the colours of iron foil, IONAs, an annealed sample in air, and PMNs, which are silver, reddish gold, dark red, and black, respectively. This observation supports that the primary phases of IONAs, the annealed sample in air, and PMNs are iron oxyfluoride, hematite, and magnetite, respectively.

Fig. 2 X-Ray diffraction (XRD) patterns of IONAs (as-anodized) and PMNs (10% H₂) (JCPDS card no. 33-0664, 19-0629 and 70-1522 for hematite, magnetite and iron oxyfluoride, respectively). The inset shows photographs of iron foil, IONAs, an annealed sample in air, and PMNs.

X-Ray photoelectron spectroscopy (XPS) is performed to obtain further information about the chemical composition and oxidation state of iron on the surface (Fig. 3). In the survey spectrum of PMNs (Fig. 3a), there are the peaks related to Fe, O, and C. However, the F 1s peak is clearly visible in IONAs, which is from iron oxyfluoride. These survey spectra indicate that both IONAs and PMNs have no impurities expect carbon contaminated in air. Fig. 3b shows a high-resolution scan across Fe 2p of IONAs and PMNs. The small satellite peak is observed in IONAs. This implies that the oxidation state of iron is close to +3 such as that in Fe₂O₃. Also, the binding energies of Fe 2p_3/2, Fe 2p_1/2, and the satellite peak are 710.9, 724.4 and 719.0 eV, respectively, which are very close to the reported peak positions of iron oxyfluoride or Fe₂O₃.⁷ On the other hand, the Fe 2p spectrum of PMNs contains a point of inflection in Fe 2p_3/2 peak due to the overlap of Fe²⁺ and Fe³⁺ peaks, which is seen in the Fe 2p spectra of Fe₃O₄. In addition, the peak positions of Fe 2p_3/2 and Fe 2p_1/2 are 710.6 and 724.2, respectively, and are characteristic of Fe₃O₄.⁸ These XPS results are consistent with the previous results in XRD.

Fig. 3 X-Ray photoelectron spectroscopy (XPS): (a) survey spectra and (b) high resolution spectra of the Fe 2p region in IONAs and PMNs.

The morphologies of IONAs and PMNs are analyzed by scanning electron microscopy (SEM) and transmission electron microscopy (TEM) (Fig. 4). Low-magnification top-view SEM images of IONAs and PMNs prove that the structures are well-formed across the large area (Fig. 4a and 4b). From the high-magnification top-view and cross-sectional SEM images of IONAs, the average diameter and length of the nanotubes are found to be around 40 nm and 2 μm, respectively (Fig. 4c and 4d). The morphology of PMNs is bundled nanopillar arrays (Fig. 4e and 4f). Interestingly, although tubular structures change to be solid structures, the one dimensional structures still remain. A transmission electron microscopy image of IONAs clearly shows that the thickness of the tube wall is around 10 nm (Fig. 4g). Energy dispersive X-ray (EDX) spectra were obtained at low-magnification top-view of IONAs and PMNs for 100 s (Fig. 4h). The IONAs and PMNs both have C, O and Fe, but only the IONAs have a high atomic percent of fluorine (∼16.6%). This also agrees with previous results.

Fig. 4 SEM images (a–f): low-magnification top-view of (a) IONAs and (b) PMNs. (c) High-magnification top view and (d) cross-sectional view of IONAs. (e) High-magnification top view of PMNs. (f) Low-magnification side view of PMNs. (g) TEM image of IONAs. (h) EDX spectra of IONAs (top) and PMNs (bottom).

Scheme 1 describes the lithiation and de-lithiation processes of IONAs as anodes for lithium ion batteries. During the discharge process, fluorine doped iron oxide converts to an Fe metal cluster embedded in the Li₂O and LiF matrix, which can act as a buffer.^2,7,9 The electron directly moves through the current collector, and the ion diffusion length is reduced due to the large surface area, leading to high rate performance. The nanotubes have void space both inside and outside, which allows them to endure the volume contraction and expansion efficiently without aggregation and pulverization. These principles and advantages can also apply to PMNs in a similar fashion.

Scheme 1 Schematic illustration of the lithiation and de-lithiation processes of vertically aligned iron oxide nanotube arrays as anodes for lithium ion batteries.

For electrochemical tests, the voltage window was set between 0.01 and 2.5 V (vs. Li/Li⁺). In this voltage range, pure iron metal can be used as a current collector with this electrolyte without any dissolution.¹⁰ Additionally, in many previous reports for three-dimensional electrodes, the areal capacity is used instead of specific capacity to present the capacity because of limitations of the loading level and the difficulty in separating the mass of the current collector and the active material.¹¹ It is difficult to determine the exact value of full capacity. In such case, estimate methods such as ‘signature curve’ can be used to determine the full capacity.¹² In order to evaluate the current density of 1 C rate for this study, full capacity is simply estimated by measuring the second discharge capacity at a very low current density of 5 μA cm⁻². As a result, current densities of 234.48 μA cm⁻² and 65.84 μA cm⁻² are equivalent to 1 C for IONAs and PMNs, respectively. The substrate area is used to calculate the areal capacity in this study.

Charge and discharge voltage profiles of IONAs and PMNs at a current density of 400 μA cm⁻² are shown in Fig. 5. In the first discharge curve, two main potential plateaus around 1.6 and 0.8 V (vs. Li/Li⁺) and one main potential plateau around 0.8 V (vs. Li/Li⁺) are observed in IONAs and PMNs, respectively, which are in good agreement with the previous reports.^7,13 The areal capacities of PMNs are much lower than those of IONAs. It can be explained by particle loss during the heat treatment. In addition, the theoretical capacities of both iron oxyfluoride and hematite are higher than that of magnetite when the same amounts of Fe are given.

Fig. 5 Charge and discharge profiles of (a) IONAs and (b) PMNs for the initial 10 cycles at a current density of 400 μA cm⁻².

Cycle performances of IONAs and PMNs at a current density of 400 μA cm⁻² are shown in Fig. 6. The initial Coulombic efficiencies of IONAs and PMNs at a current density of 400 μA cm⁻² are 60.9 and 70.8%, respectively. The irreversible capacity loss at the first cycle is mainly due to electrolyte decomposition and the formation of a solid electrolyte interface (SEI) layer. IONAs have a lower initial Coulombic efficiency than PMNs, which can be attributed to the irreversible lithium reaction with the fluoride component to LiF. The cyclic stabilities of both IONAs and PMNs are excellent at a current density of 400 μA cm⁻², which is a high rate for the conversion of reaction materials. The first charge capacities of IONAs and PMNs are 197.54 and 56.96 μA h cm⁻², respectively. They have higher areal capacities and cyclic stabilities at high current density compared to those of other three-dimensional electrodes in previous reports.¹¹ For IONAs, 21.0% of the charge capacity decreases until 30 cycles and 13.7% decreases during the subsequent 40 cycles. In the case of PMNs, 16.8% of the charge capacity decreases until 30 cycles and only 3.5% decreases during the subsequent 40 cycles, which indicates a higher stability than that of IONAs. After 100 cycles at a current density of 400 μA cm⁻², the SEI covers both samples (Fig. S2, ESI†). However, the porous structures are still clearly seen in PMNs, which contribute to the high cyclic stability.

Fig. 6 (a) Cycle performance of IONAs at a current density of 400 μA cm⁻² (top) and rate performance of IONAs (bottom). (b) Cycle performance of PMNs at current density of 400 μA cm⁻² (top) and rate performance of PMNs (bottom).

Rate performances from the very low current density of 5 μA cm⁻² to the high current density of 3200 μA cm⁻² of IONAs and PMNs are shown in Fig. 6. The rate capabilities of IONAs and PMNs are excellent even at the high current density of 3200 μA cm⁻². The areal charge capacities of IONAs and PMNs at the very high current density of 3200 μA cm⁻² are 49.33 and 27. 56 μA h cm⁻², which are 23.2 and 46.2% of the charge capacities at very low current density of 5 μA cm⁻², respectively. In particular, PMNs at the current density of 3200 μA cm⁻² take less than 90 s to complete one charge or discharge. The areal capacities are recovered at a current density of 200 μA cm⁻² after cycling at the very high current density of 3200 μA cm⁻².

Conclusions

In summary, we prepared the IONAs by a simple anodization method, which afterwards were heat treated to produced the PMNs. The IONAs and PMNs both have suitable morphologies and structures as electrodes for lithium ion batteries. Conventional lithium ion battery components, such as the current collector, binder, and conductive agent, and the slurry-making step are not required for these materials. The large free space can alleviate the severe volume change during cycling, which leads to a good cyclic stability. Finally, the direct contact of active materials to the current collector and the large surface area leads to high electron and lithium ion conductivities, respectively. In particular, PMNs have an excellent rate capability, where they are stable at the extremely high current density of 3200 μA cm⁻², which can complete one charge or discharge in a couple of minutes.

Acknowledgements

This work is supported by the WCU (World Class University) program through the National Research Foundation of Korea and the Institute of Basic Science (IBS) program funded by the Korean Ministry of Education, Science, and Technology and the Fundamental R&D Program for Technology of World Premier Materials funded by the Ministry of Knowledge Economy, Republic of Korea.

References

1. (a) K. Kang, Y. S. Meng, J. Breger, C. P. Grey and G. Ceder, Science, 2006, 311, 977; (b) S. Baek, S.-H. Yu, S.-K. Park, A. Pucci, C. Marichy, D.-C. Lee, Y.-E. Sung, Y. Piao and N. Pinna, RSC Adv., 2011, 1, 1687.
2. P. Poizot, S. Laruelle, S. Grugeon, L. Dupont and J.-M. Tarascon, Nature, 2000, 407, 496.
3. (a) J. Cabana, L. Monconduit, D. Larcher and M. R. Palacin, Adv. Mater., 2010, 22, E170; (b) Y. Deng, Z. Li, Z. Shi, H. Xu, F. Peng and G. Chen, RSC Adv., 2012, 2, 4645.
4. (a) C. K. Chan, H. Peng, G. Liu, K. Mcilwrath, X. F. Zhang, R. A. Huggins and Y. Cui, Nat. Nanotechnol., 2008, 3, 31; (b) T. Djenizian, I. Hanzu and P. Knauth, J. Mater. Chem., 2011, 21, 9925; (c) W. Wang, M. Tian, A. Abdulagatov, S. M. George, Y.-C. Lee and R. Yang, Nano Lett., 2012, 12, 655; (d) Y. Li, B. Tan and Y. Wu, Nano Lett., 2008, 8, 265.
5. (a) L. Pu, X. Bao, J. Zou and D. Feng, Angew. Chem., Int. Ed., 2001, 40, 1490; (b) J. Gong, C. Lin, M. Ye and Y. Lai, Chem. Commun., 2011, 47, 2598; (c) N. K. Allam, X. J. Feng and C. A. Grimes, Chem. Mater., 2008, 20, 6477; (d) K. Xie, J. Li, Y. Lai, Z. Zhang, Y. Liu, G. Zhang and H. Huang, Nanoscale, 2011, 3, 2202.
6. (a) S. K. Mohapatra, S. E. John, S. Banerjee and M. Misra, Chem. Mater., 2009, 21, 3048; (b) T. J. Latempa, X. Feng, M. Paulose and C. A. Grimes, J. Phys. Chem. C, 2009, 113, 16293; (c) K. Xie, J. Li, Y. Lai, W. Lu, Z. Zhang, Y. Liu, L. Zhou and H. Huang, Electrochem. Commun., 2011, 13, 657.
7. L. Yu, H.-X. Wang, Z.-Y. Liu and Z.-W. Fu, Electrochim. Acta, 2010, 56, 767.
8. T. Yamashita and P. Hayes, Appl. Surf. Sci., 2008, 254, 2441.
9. N. Pereira, F. Badway, M. Wartelsky, S. Gunn and G. G. Amatucci, J. Electrochem. Soc., 2009, 156, A407.
10. (a) S.-T. Myung, Y. Sasaki, S. Sakurada, Y.-K. Sung and H. Yashiro, Electrochim. Acta, 2009, 55, 288; (b) C. Iwakura, Y. Fukumoto, H. Inoue, S. Ohashi, S. Kobayashi, H. Tada and M. Abe, J. Power Sources, 1997, 68, 301.
11. (a) H. S. Kim, S. H. Kang, Y. H. Chung and Y.-E. Sung, Electrochem. Solid-State Lett., 2010, 13, A15; (b) G. F. Ortiz, I. Hanzu, P. Lavela, J. L. Tirado, P. Knauth and T. Djenizian, J. Mater. Chem., 2010, 20, 4041; (c) S. K. Cheah, E. Perre, M. Rooth, M. Fondell, A. Harsta, L. Nyholm, M. Boman, T. Gustafsson, J. Lu, P. Simon and K. Edstrom, Nano Lett., 2009, 9, 3230; (d) G. F. Ortiz, I. Hanzu, T. Djenizian, P. Lavela, J. L. Tirado and P. Knauth, Chem. Mater., 2009, 21, 63; (e) N. A. Kyeremateng, V. Hornebecq, P. Knauth and T. Djenizian, Electrochim. Acta, 2012, 62, 192; (f) W.-H. Ryu, D.-H. Nam, Y.-S. Ko, R.-H. Kim and H.-S. Kwon, Electrochim. Acta, 2012, 61, 19.
12. (a) M. Doyle, J. Newman and J. Reimers, J. Power Sources, 1994, 52, 211; (b) P. L. Taberna, S. Mitra, P. Poizot, P. Simon and J.-M. Tarascon, Nat. Mater., 2006, 5, 567.
13. (a) Y. Piao, H. S. Kim, Y.-E. Sung and T. Hyeon, Chem. Commun., 2010, 46, 118; (b) H. S. Kim, Y. Piao, S. H. Kang, T. Hyeon and Y.-E. Sung, Electrochem. Commun., 2010, 12, 382.

---

Footnote

† Electronic supplementary information (ESI) available: X-Ray diffraction patterns of annealing samples in air, oxygen and Ar and SEM images after cycling. See DOI: 10.1039/c2ra22162d

---