pubs.acs.org/NanoLett
Si/TiSi2 Heteronanostructures as HighCapacity Anode Material for Li Ion Batteries Sa Zhou, Xiaohua Liu, and Dunwei Wang* Department of Chemistry, Merkert Chemistry Center, Boston College, 2609 Beacon Street, Chestnut Hill, Massachusetts 02467 ABSTRACT We synthesized a unique heteronanostructure consisting of two-dimensional TiSi2 nanonets and particulate Si coating. The high conductivity and the structural integrity of the TiSi2 nanonet core were proven as great merits to permit reproducible Li+ insertion and extraction into and from the Si coating. This heteronanostructure was tested as the anode material for Li+ storage. At a charge/discharge rate of 8400 mA/g, we measured specific capacities >1000 mAh/g. Only an average of 0.1% capacity fade per cycle was observed between the 20th and the 100th cycles. The combined high capacity, long capacity life, and fast charge/discharge rate represent one of the best anode materials that have been reported. The remarkable performance was enabled by the capability to preserve the crystalline TiSi2 core during the charge/discharge process. This achievement demonstrates the potency of this novel heteronanostructure design as an electrode material for energy storage. KEYWORDS TiSi2, nanonet, Li ion battery, nanostructure
W
ith the theoretical specific capacity limit of 4200 mAh/g, crystalline Si (c-Si) represents a particularly appealing candidate as the anode material for Li ion batteries. The unmanageable volumetric expansion of Si upon Li+ insertion, however, results in drastic and fast capacity fading due to structural and electronic degradation, dampening the prospect of exploiting the high capacity Si possesses.1 To solve this issue, Si-based nanostructures such as nanoparticles, thin films, nanowires, and most recently nanotubes have been studied.2–8 Similar to cases where bulk Si is involved, pulverization and electronic contact degradation keep the capacity life of anodes made of nanoparticles that contain Si short.9 Thin film Si offers high specific capacity, good capacity retention, and fast charge/discharge rate,10–12 but it suffers a major drawback of low active material content. While the anisotropic nature of Si nanowires acts positively to accommodate the volumetric changes upon Li+ insertion and extraction, the complete lithiation of Si nanowires nevertheless impedes charge transport in the longitudinal direction, limiting the charge/discharge rate and capacity life.5,13 Evidently, the realization of high capacity, long capacity life and fast charge/discharge rate requires the ability to accommodate the volumetric change while maintaining superior charge transport, a goal best met by composite nanomaterials.6,14,15 Carbon nanotubes16 and nanofibers,17 for instance, have been studied as the inactive component to facilitate charge transport in and from active Si. Here we report a new material of Si/TiSi2 that answers to the above challenges. The material includes highly conduc-
tive TiSi2 nanonets (NNs) as the structural support as well as the component to facilitate effective charge transport. Nanoscale Si particles are attached to the NNs as the medium to react with Li+. Compared to the competing structures, the Si/TiSi2 heteronanostructure offers two distinct advantages, ease of interfacing Si with TiSi2 and superior charge transport through TiSi2. The former is enabled by the similarities between TiSi2 and Si crystal structures, and the latter is ensured by the capability to selectively insert Li+ into Si only. We show that fast charge/discharge without significant capacity fading can be achieved. At the charging rate of 8400 mA/g, we observed >99% capacity retention per cycle at the level of >1000 mAh/g over 100 cycles. In a typical synthesis, the two-dimensional TiSi2 NNs were first grown on Ti foil in a chemical vapor deposition (CVD) reactor with TiCl4 and SiH4 as the precursors.18,19 Although every individual TiSi2 NN is two-dimensional by nature, they stand perpendicular to the supporting substrate and pack with a high density, which is ideal as a scaffold for Si deposition. Without exposing the as-grown TiSi2 NNs to ambient air, we next flew through 80 sccm (standard cubic centimeter per minute) SiH4 (10% in He) and 100 sccm H2 at 650 °C for 12 min (Ptotal ) 15 Torr) to directly grow Si nanoparticles (approximately ∼20 nm in diameters) on the TiSi2 NN surfaces. A rapid thermal processing (RTP) treatment at 900 °C for 30 s in forming gas (5% H2 in N2) was then performed to conclude the synthesis process. The weight of the as-produced Si/TiSi2 nanostructures was calculated by measuring the supporting substrate on a microbalance (Sartorius R200D) before and after the synthesis. The information was later used to calculate the specific capacity. The as-produced nanostructure was characterized by a scanning electron microscope (SEM, JEOL JSM 6340F) and a transmission electron microscope (TEM, JEOL 2010F).
* To whom correspondence should be addressed. E-mail:
[email protected]. Tel.: 617-552-3121. Fax: 617-552-2705. Received for review: 10/7/2009 Published on Web: 02/11/2010 © 2010 American Chemical Society
860
DOI: 10.1021/nl903345f | Nano Lett. 2010, 10, 860–863
An STM-TEM sample holder (Nanofactory Instruments AB) was used in conjunction with the TEM to measure the conductivity of the nanostructure. Once synthesized, the Si/ TiSi2 heterostructures was made into an electrode in a glovebox and tested in a half-cell configuration in the glovebox (see Supporting Information for more details). Throughout the fabrication process, the electrode was only exposed to air during sample transferring between apparatuses to a total of less than 5 min. All electrochemical measurements were conducted in a sealed box that was placed in the glovebox at room temperature. A CHI 600C Potentiostat/ Galvanostat was used for these measurements. The gravimetric capacity was calculated by dividing the amount of charges inserted or extracted from the sample by the measured weight. The resulting value represents the capacity of the overall Si/TiSi2 nanostructure. The morphology of typical as-made Si/TiSi2 nanostructures is shown in Figure 1. We emphasize two characteristics of the products, (i) the uniformity of the Si coating, and (ii) the particulate nature of the Si coating; both are important to the realization of high capacity and good cyclability as will be discussed later in this paper. The deposition of Si is sensitive to the reaction conditions. Reaction conditions different from those reported produced either uniform thin film Si coating that was polycrystalline or amorphous Si coating. On the basis of the TEM characterizations, we estimate that Si accounts for ∼75% of the total weight. Therefore, the capacity as reported below would be increased by up to ∼33% if only Si was considered. With the nanostructures successfully produced, our next goal was to identify experimental conditions that will allow for selective Li+ insertion into Si but not TiSi2. The selectivity will prevent potential structural degradation of the conductive TiSi2 core. We achieved this objective by limiting the state of charge (SOC) and discharge (SOD) through voltage control. Electrochemical potential spectroscopy (EPS) suffices for this purpose.20,21 As shown in Figure 2a, the peak in the EPS at 60-70 mV corresponds to Li+ reacting with TiSi2 while that at 120 mV is caused by the lithiation of c-Si. Understanding the nature of the lithiation of TiSi2 nanonets will require further studies. In line with the reports of bulk metal silicides lithiation processes,22,23 we suggest that the lithiation may proceed on the site of Si. C-Si is typically transformed into amorphous Si (a-Si) after the first discharging, leading to a broad peak beginning at ∼240 mV (Figure 2a blue line).24 The difference between the lithiation potentials of Si and TiSi2 as shown in the EPS data (Figure 2) permits the selection of operation potential ranges. Figure 2b illustrates how the range of operation potentials influences the capacity life of Si/TiSi2 heteronanostructures. When the operation potentials were set between 0.150-3.00 V, no lithiation reactions occurred on TiSi2. As a result, the capacity was maintained at a level of ∼1100 mAh/g during the first 50 cycles of charge/discharge. In contrast, when the operation potential range was increased to 0.090-3.00 V, © 2010 American Chemical Society
FIGURE 1. (a) Schematic of the Si/TiSi2 heteronanostructure. Si nanoparticles are deposited on highly conductive TiSi2 nanonets to act as the active component for Li+ storage. (b) A low-mag TEM picture manifests the particulate nature of the Si coating on TiSi2 NNs. (c) A high-mag TEM picture and the selected area electron diffraction pattern reveal the crystallinity nature of the TiSi2 core and the particulate Si coating. (d) The crystallinity of TiSi2 and Si (highlighted by the dotted red line) is shown in this lattice-resolved TEM picture. 861
DOI: 10.1021/nl903345f | Nano Lett. 2010, 10, 860-–863
FIGURE 2. (a) Electrochemical potential spectroscopy of TiSi2 and Si/TiSi2 heterostructures. For clarity, only the portion corresponding to charging is shown here, with arbitrary offsets in the dQ/dV axis. The peaks in the shaded region correspond to Li+ insert to TiSi2. The peak denoted by the red square is due to Li+ insertion into c-Si, and that by the blue circle is due to Li+ insertion into a-Si. (b) The capacity retention is improved by choosing a higher cutoff potential. Charge rate: 8400 mA/g.
FIGURE 3. (a) A TEM graph of the as-prepared Si/TiSi2 heteronanostructure reveals the crystalline nature of both the TiSi2 core and the Si shell. (b) After 20 cycles of continuous charge/discharge, the Si shell is transformed into amorphous while the crystalline nature of the TiSi2 core is preserved. Scale bars: 20 nm. (c) The superior conductivity of the TiSi2 core also survives the charge/discharge processes. The resistivity is calculated as ∼10 µΩ·cm.
the effect of TiSi2 lithiation showed up. Although this reaction is less significant than that between Si and Li+, it nonetheless caused the degradation of TiSi2, presumably due to stressrelated pulverization, which manifested itself as a quick fading in the measured capacity after 40 cycles of charge/ discharge. The effect of TiSi2 degradation-induced capacity fading became more obvious when the operation potential range was further expanded to 0.030-3.00 V. Note that the higher stability at higher cutoff potentials was achieved at the expense of the specific capacity. For example, at the same charge/discharge rate (8400 mA/g), the initial capacity measured with a 30 mV cutoff potential was ∼50% higher than measured with a 150 mV cutoff potential due to the extended lithiation between 30 and 150 mV. We also note that higher specific capacity was measured when slower charge/discharge rates were used. In light of the critical importance of TiSi2 to our design, we paid close attention to the morphology and conductivity of the TiSi2 core at different stages of the process, for example, as-prepared, after the initial charge/discharge and after repeated charge/discharge processes. Both the crystallinity and the conductivity were well preserved when the cutoff potential was set as 150 mV (Figure 3). The intact TiSi2 core serves dual functionalities, structural support and charge transporter. Upon Li+ insertion, the TiSi2 core provides electrons to counteract the cation-insertion-induced charge imbalance, allowing for rapid Li+ incorporation. Similarly, TiSi2 also facilitates the electron collection and transport © 2010 American Chemical Society
during Li+ extraction. The particulate nature of the Si coating is key to the successful implementation of our design. The space between adjacent Si particles permits the volumetric expansion when the Li-Si alloy (e.g., Li14Si5) is formed. Control experiments showed that the capacity faded more rapidly when a uniform Si coating was used (Supporting Information, Figure S4). Attempts to directly coat a-Si on TiSi2 failed to produce the same result, that is, high capacity and long cycling life, although similar results have been obtained on Si nanowires and carbon nanofibers.13,17 We suggest that the key to the high performance lies in the quality of the Si coating. Consistent with the literature,13,17 our results indicate that uniform a-Si is more desirable than uniform c-Si. With the roles of TiSi2 and Si in the nanostructure understood and experimental parameters fixed, we were able to characterize the capabilities of the Si/TiSi2 nanostructures in retaining capacities after repeated charge/discharge cycles. As shown in Figure 4, the first charging capacity of 1990 mAh/g was obtained with a charging rate of 1300 mA/ g. During this step, c-Si was converted to a-Si, and the phase transformation resulted in a large drop in the capacity upon discharge to 1182 mAh/g. It is important to perform this step at a slow pace (1300 mA/g). Otherwise rapid capacity fading would be resulted as a result of pulverization. This process 862
DOI: 10.1021/nl903345f | Nano Lett. 2010, 10, 860-–863
Supporting Information Available. Experimental details of material syntheses, structural characterizations, electrochemical measurements, and electrical measurements; discussions of the influence of charge/discharge rates and morphologies of Si coating on the capacity and capacity retention. This material is available free of charge via the Internet at http://pubs.acs.org. REFERENCES AND NOTES (1) (2) (3)
FIGURE 4. Charge capacity and Coulombic efficiency of the Si/ TiSi2 heteronanostructure with 8400 mA/g charge/discharge rate tested between 0.150 and 3.00 V.
(4) (5)
continued in the first 10 cycles as evident by the continuous capacity fade and the Coulombic efficiency increase. The capacity change after the first 10 cycles was minimum. For example, the charging capacity at the 23rd cycle was 1026 mAh/g and that at the 100th cycle was 937 mAh/g, corresponding to a fade of 8.7%, or ∼0.1% per cycle. In principle, there is still room for capacity improvement by, for example, optimizing the Si-coating thickness and the ratio of Si to TiSi2. A thicker Si coating would lead to higher specific capacity, but probably would also lead to poorer capacity life. Nevertheless, the reported anode capacity falls in a desirable range.25 Further increase of the anode capacity would contribute little to the overall cell capacity improvement without breakthroughs in increasing the capacities of the cathode and the electrolyte.25 In summary, we have synthesized a novel Si/TiSi2 heteronanostructure as the anode material for Li+ batteries. In this structure, Si acts as an active component to store and release Li+ while TiSi2 serves as the inactive component to support Si and to facilitate charge transport. The differences between their electrochemical potentials in reacting with Li+ permit the selection of the operation potentials to keep TiSi2 intact. At a fast charge/discharge rate of 8400 mA/g, superior capacity retention of >99% per cycle at the level of 1000 mAh/g over 100 cycles were obtained. Although bulk TiSi2 and other silicides have been studied as anode materials for Li ion batteries, this is the first report of using nanostructured TiSi2 for similar applications.22,23 The essence of our design lies in the combination of a complex conductive core that does not participate in the lithiation process and a reactive coating that acts as the Li+ insertion and extraction medium. It shall be possible to extend the design other extensively studied materials as well, such as Ge,26 Sn,27 SnO228–30 and transition metal oxides.31
(6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29)
Acknowledgment. The research is funded by Boston College. We thank Y. Lin for experimental assistance and insightful discussions. We also thank S. Shepard and D. Z. Wang for their technical support.
© 2010 American Chemical Society
(30) (31)
863
Kasavajjula, U.; Wang, C.; Appleby, A. J. J. Power Sources 2007, 163 (2), 1003–1039. Tarascon, J. M.; Armand, M. Nature 2001, 414 (6861), 359–367. Li, H.; Huang, X.; Chen, L.; Wu, Z.; Liang, Y. Electrochem. SolidState Lett. 1999, 2 (11), 547–549. Ohara, S.; Suzuki, J.; Sekine, K.; Takamura, T. J. Power Sources 2003, 119-121, 591–596. Chan, C. K.; Peng, H.; Liu, G.; McIlwrath, K.; Zhang, X. F.; Huggins, R. A.; Cui, Y. Nat. Nanotechnol. 2008, 3 (1), 31–35. Graetz, J.; Ahn, C. C.; Yazami, R.; Fultz, B. Electrochem. Solid State Lett. 2003, 6 (9), A194–A197. Arico, A. S.; Bruce, P.; Scrosati, B.; Tarascon, J. M.; Van Schalkwijk, W. Nat. Mater. 2005, 4 (5), 366–377. Park, M.-H.; Kim, M. G.; Joo, J.; Kim, K.; Kim, J.; Ahn, S.; Cui, Y.; Cho, J. Nano Lett. 2009. Ryu, J. H.; Kim, J. W.; Sung, Y. E.; Oh, S. M. Electrochem. Solid State Lett. 2004, 7 (10), A306–A309. Wang, G. X.; Ahn, J. H.; Yao, J.; Bewlay, S.; Liu, H. K. Electrochem. Commun. 2004, 6 (7), 689–692. Morita, T.; Takami, N. J. Electrochem. Soc. 2006, 153 (2), A425– A430. Takamura, T.; Ohara, S.; Uehara, M.; Suzuki, J.; Sekine, K. J. Power Sources 2004, 129 (1), 96–100. Cui, L. F.; Ruffo, R.; Chan, C. K.; Peng, H. L.; Cui, Y. Nano Lett. 2009, 9 (1), 491–495. Mao, O.; Turner, R. L.; Courtney, I. A.; Fredericksen, B. D.; Buckett, M. I.; Krause, L. J.; Dahn, J. R. Electrochem. Solid State Lett. 1999, 2 (1), 3–5. Kim, H.; Cho, J. Nano Lett. 2008, 8 (11), 3688–3691. Shu, J.; Li, H.; Yang, R.; Shi, Y.; Huang, X. Electrochem. Commun. 2006, 8 (1), 51–54. Cui, L.-F.; Yang, Y.; Hsu, C.-M.; Cui, Y. Nano Lett. 2009, 9, 3370– 3374. Zhou, S.; Liu, X. H.; Lin, Y. J.; Wang, D. W. Angew. Chem., Int. Ed. 2008, 47, 7681–7684. Zhou, S.; Liu, X.; Lin, Y.; Wang, D. Chem. Mater. 2009, 21 (6), 1023–1027. Thompson, A. H. J. Electrochem. Soc. 1979, 126 (4), 608–616. Cui, L.-F.; Ruffo, R.; Chan, C. K.; Peng, H.; Cui, Y. Nano Lett. 2008, 9 (1), 491–495. Netz, A.; Huggins, R.; Weppner, W. Ionics 2001, 7 (4), 433–439. Netz, A.; Huggins, R. A.; Weppner, W. J. Power Sources 2003, 119-121, 95–100. Hatchard, T. D.; Dahn, J. R. J. Electrochem. Soc. 2004, 151 (6), A838–A842. Yoshio, M.; Tsumura, T.; Dimov, N. J. Power Sources 2005, 146 (1-2), 10–14. Chan, C. K.; Zhang, X. F.; Cui, Y. Nano Lett. 2007, 8 (1), 307– 309. Winter, M.; Besenhard, J. O. Electrochim. Acta 1999, 45 (1-2), 31–50. Brousse, T.; Retoux, R.; Herterich, U.; Schleich, D. M. J. Electrochem. Soc. 1998, 145 (1), 1–4. Idota, Y.; Kubota, T.; Matsufuji, A.; Maekawa, Y.; Miyasaka, T. Science 1997, 276 (5317), 1395–1397. Meduri, P.; Pendyala, C.; Kumar, V.; Sumanasekera, G. U.; Sunkara, M. K. Nano Lett. 2009, 9 (2), 612–616. Poizot, P.; Laruelle, S.; Grugeon, S.; Dupont, L.; Tarascon, J. M. Nature 2000, 407 (6803), 496–499.
DOI: 10.1021/nl903345f | Nano Lett. 2010, 10, 860-–863
