Li-Ion Battery with LiFePO4 Cathode and Li4Ti5O12 ... - Springer Link

4 downloads 0 Views 605KB Size Report
Sep 5, 2012 - cell with Ketjen black modified LiFePO4 cathode and an unmodified ... negligible fade after more than 1200 cycles with a capacity of ∼130 ...
Li-Ion Battery with LiFePO4 Cathode and Li4Ti5O12 Anode for Stationary Energy Storage WEI WANG, DAIWON CHOI, and ZHENGUO YANG Li-ion batteries based on commercially available LiFePO4 cathode and Li4Ti5O12 anode were investigated for potential stationary energy storage applications. The full cell that operated at flat 1.85 V demonstrated stable cycling up to 200 cycles followed by a rapid fade. A Li-ion full cell with Ketjen black modified LiFePO4 cathode and an unmodified Li4Ti5O12 anode exhibited negligible fade after more than 1200 cycles with a capacity of ~130 mAh/g at C/2. The improved stability, along with its cost-effectiveness, environmental benignity, and safety, make the LiFePO4/Li4Ti5O12 combination Li-ion battery a promising option for storing renewable energy. DOI: 10.1007/s11661-012-1284-4  The Minerals, Metals & Materials Society and ASM International (outside the USA) 2012

I.

INTRODUCTION

THE concerns over the environmental consequences from burning fossil fuels, along with their resource constrains, have spurred growing interest in renewable energy generated from sources such as solar and wind. However, the intermittent nature of the alternative energy sources necessitates the use of the effective energy storage system to dispatch the varied renewable power.[1–3] One potential technology is use of Li-ion batteries that are now being extensively developed for vehicle applications.[3] Unlike the portable and vehicle applications that have stringent requirements on energy/ power densities due to their weight and volume constrains, the stationary energy storage of renewable power puts more emphasis on cost-effectiveness, cycle stability, safety, etc. Accordingly, our efforts have focused on Li-ion battery chemistries using cost-effective anodes and cathodes of excellent structural and chemical stability and thus capable of a long cycle life. Particularly, we investigated the Li-ion batteries made from LiFePO4 cathode and Li4Ti5O12 anode using commercial electrolyte. Li4Ti5O12 demonstrates a specific capacity of 175 mAh/g with a flat discharge/charge profile at 1.55 V and is known for its ‘‘zero’’ straining Li-insertion/ deinsertion behavior and thus long cycle life.[4] The cycling voltage of Li4Ti5O12 with respect to lithium metal is relatively higher than that of graphite, which leads to a lower energy density by narrowing down the useful potential of the full cell. However, compared with graphite, Li4Ti5O12 possesses significant advantages for stationary applications in terms of long cycling stability WEI WANG and DAIWON CHOI, Scientists, are with the Pacific Northwest National Laboratory, 902 Battelle Boulevard, P.O. Box 999, Richland, WA 99354. Contact e-mails: [email protected], [email protected] ZHENGUO YANG, CEO, is with UniEnergy Technologies, LLC, 4333 Harbour Pointe Blvd SW, Unit A, Mukilteo, WA 98275. Manuscript submitted March 14, 2012. Article published online September 5, 2012 METALLURGICAL AND MATERIALS TRANSACTIONS A

and safety by avoiding the formation of unsafe solid electrolyte interface (SEI) layer caused by electrolyte decomposition at lower voltage (4000 deep cycles) for stationary energy storage systems. We have previously reported self-made LiFePO4 as the choice of cathode in a design of Li-ion battery for stationary energy storage over 700 cycles.[3] This article reports the improvement in long-term cyclability of a Li-ion battery consisting of LiFePO4 as cathode and Li4Ti5O12 as anode, in which commercial available materials were used with least modification for cost-effectiveness.

II.

EXPERIMENTAL

LiFePO4 and Li4Ti5O12 powders were obtained from commercial vendors. Ten weight percent Ketjen black was added to LiFePO4 and milled in a planetary mill for VOLUME 44A, JANUARY 2013—S21

8 hours (Retsch 100CM) at 400 rpm. In a subsequent study, single-walled carbon nanotubes (SWNTs) and multi-walled carbon nanotubes (MWNTs) were used as additives to further improve the conductivity of the active materials, for which 10 wt pct of each carbon additive was dispersed with LiFePO4 and Li4Ti5O12 powders, respectively, in isopropanol solvent using high power ultrosonication. The solvent was then allowed to evaporate in order to collect the powder for battery test. Microstructures were analyzed by FE-SEM (FEI Nova 600, FEI, Hillsboro, OR). The cathode or anode active material, super P, and poly(vinylidene fluoride) binder were dispersed in N-methylpyrrolidone solution in a weight ratio of 80:10:10. The slurry was then coated on the Al foil current collector and dried overnight in a vacuum oven at 393 K (120 C). The performance of the LiFePO4, Li4Ti5O12 half-cells and full cells was evaluated with Arbin battery tester (College Station, TX) in 2325 coin cell (National Research Council, Ottawa, Canada) configuration using 1M LiPF6 in EC/DMC (2:1) (ethyl carbonate/dimethyl carbonate) electrolyte at room temperature. Charge and discharge were carried out between 4.5 and 2 V for LiFePO4, and 1 and 2.5 V for Li4Ti5O12 half-cells and LiFePO4/Li4Ti5O12 full cell. All cells were first subjected to rate capability study in galvanostatic modes starting from C/10 rate and were then ramped up to 10 C rate. The cells were subjected to extended cycles at C/2 rate afterward. Various C rate current densities were calculated based on theoretical capacity for cathode and anode, respectively. The weight ratio of LiFePO4 and Li4Ti5O12 in full cells was carefully balanced with ~2 mg/cm2 of loading. The specific capacity, energy, and power density of the full cells were calculated based on the weight of the limiting electrode (one with lesser weight).

III.

RESULTS AND DISCUSSION

Figure 1 shows the scanning electron microscopy (SEM) images of as-received LiFePO4 and Li4Ti5O12 along with LiFePO4 modified with Ketjen black. The particle size of the Li4Ti5O12 is around 2 lm from the SEM analysis. In comparison, LiFePO4 has a much smaller particle size in the nanometer range. From the SEM image, the Ketjen black appears homogeneously dispersed with LiFePO4 particles after milling. LiFePO4, Li4Ti5O12, and LiFePO4 milled with Ketjen black (denoted hereafter as LFP, LTO, and LFPKB, respectively) were studied at different C rates, as shown in Figure 2. Typical charge-discharge curves of LFP and LFPKB in Figure 2(c) show a characteristic flat potential ~3.45 V vs Li, corresponding to the two-phase Li extraction/insertion process. The specific capacities of the pure LFP achieved 151, 128, and 63 mAh/g at C/5, 1 C, and 10 C rates, respectively. Compared to the pure LFP, LFPKB with 10 pct Ketjen black exhibited a significant improved rate capability, reaching 166, 159, and 138 mAh/g at C/5, 1 C, and 10 C, respectively, as shown in both Figures 2(a) and (c). The Ketjen black coating is believed to enhance the conductivity of the pure LFP through its high surface area carbon coating S22—VOLUME 44A, JANUARY 2013

Fig. 1—SEM images of LTO, LFP, and LFPKB.

promoting the mass transport of Li ions and electrons.[9] As shown in Figures 2(b) and (d), LTO half-cell demonstrates a flat voltage at ~1.55 V, indicating the two-phase reaction. The specific charge capacities were 158, 109, and 60 mAh/g at C/10, 1 C, and 5 C rates, respectively. The potential curves for charge/discharge are mostly symmetrical even at 5 C, suggesting a good quality of the raw LTO powder in terms of purity and particle size distribution compared to the self-made LTO in Reference 6. METALLURGICAL AND MATERIALS TRANSACTIONS A

200

200

Specific capacity (mAh/g)

C/10

C/5

C/2

C

150

2C

5C

C/10 C/5 10C

150

C/2

C/2 C

100

2C

100

5C

Charge-LFOKB Disharge-LFOKB Charge-LFO Discharge-LFO

50

50 Discharge-LTO Charge-LTO

0

0 0

5

10

15

20

25

30

35

0

5

10

15

20

25

Cycle number

Cycle number

(a)

(b)

4.5

30

35

2.5

Voltage V (vs.Li)

4.0 LFO C/10 1C 10C LFOKB C/10 1C 10C

3.5

3.0

2.5

2.0

2.0 LTO C/10 1C 5C

1.5

1.0 0

50

100

150

200

0

50

100

150

Specific capacity (mAh/g)

Specific capacity (mAh/g)

(c)

(d)

200

Fig. 2—Electrochemical cycling at various C rates: (a) LFP and LFPKB, (b) LTO and charge/discharge voltage profiles at various C rates, (c) LFP and LFPKB, and (d) LTO.

The electrochemical cycling performance of the full cell consisting of LTO anode and LFP or LFPKB cathode, hereafter denoted as LFP-LTO and LFPKBLTO, respectively, were also studied, as shown in Figures 3(a) and (c). The full cell was fabricated in such a way as to closely balance the capacities of the cathode and anode. The voltage profile in Figure 3(c) shows a plateau ~1.85 V at C/10, indicating two-phase lithium insertion/extraction in both electrodes. At the C/5 rate, a specific capacity of 145 and 147 mAh/g (calculation based on the weight of the limiting electrode) was obtained for LFP-LTO and LFPKB-LTO, respectively. The rate performance of the full cell is much worse than the LFP and LFPKB half-cell, but is comparable to that of the LTO half-cell. The LFPKB-LTO cell demonstrated a better rate performance particularly at higher rates of 5C and 10C. At 10C, LFPKB-LTO delivered 24 mAh/g, while LFP-LTO achieved only 9 mAh/g attributed to the carbon black coating on the LFP particles. As shown in Figure 3(b), the Ragone plots of LTO, LFP-LTO, and LFPKB-LTO almost overlap with each other with LFP and LFPKB showing much better rate performance at higher rate. This suggests that the rate capability of the full cell is limited by the LTO anode. METALLURGICAL AND MATERIALS TRANSACTIONS A

While high rate capability may not be necessary for stationary applications, a better rate performance indicates a lower internal resistance, which leads to exothermic irreversible heat generation Qirr = Igt + I2Rt (I: current, g: absolute value of electrode polarization, R: Ohmic resistance, and t: time).[10] An electrode with lesser heat generation renders the battery cell better safety performance in long-term cycling and a longer cycling life due to prevention of thermal runnaway. The cycling performance of the full-cell battery at C/2 rate is shown in Figure 3(d). The LFP-LTO full-cell demonstrates a stable cycling up to ~200 cycles followed by a rapid fading in subsequent cycles. On the other hand, LFPKB-LTO exhibited a much improved cycling stability for over 1200 cycles with a fading rate of 0.009 pct per cycle. Although the detailed fading mechanism is unclear at present, the Ketjen black coating appears to be a key factor contributing to a much improved cycling stability, which may be due to less heat generation resulting from reduced internal resistance. The capacity fading mechanism in olivine structured cathode materials is often attributed to the formation of cracks and subsequent pulverization induced by the volumetric change during intercalation/deintercalation.[11] Zhi et al. suggested another possible reason VOLUME 44A, JANUARY 2013—S23

1/10C 150

1/5C 1/2C

1/2C 1C 2C

100

5C

Charge LFOKB-LTO Discharge LFOKB-LTO Charge LFO-LTO Discharge LFO-LTO

50

10C

Power density (W/Kg)

Specific capacity (mAh/g)

200

0

1000

LFO-LTO LFOKB-LTO LTO LFO LFOKB

100

10

0

5

10

15

20

25

30

35

10

100

Cycle number

1000

Energy density (Wh/Kg)

(a)

(b)

LFO-LTO C/10 1C 10C LFOKB-LTO C/10 1C 10C

Voltage

2.0

1.5

Specific capacity (mAh/g)

2.5 150

1/2C

100

Charge LFOKB-LTO Discharge LFOKB-LTO Charge LFO-LTO Discharge LFO-LTO

50

0

1.0 0

50

100

150

200

Specific capacity (mAh/g)

200

400

600

800

1000

1200

Cycle number

(d)

(c)

Fig. 3—(a) Electrochemical cycling and (c) charge/discharge voltage profiles of full cell at various C rates. (b) Ragone plot of different cells. (d) Cycling performance of full cell.

for the residual HF attack on the iron leading to a loss of contact between the LiFePO4 particles and conductive carbon.[12] The full cell configuration could further complicate the cause of capacity decay adding the diffusion-induced stress and fracture near the separator region.[13] Nevertheless, the capacity fading of LiFePO4 seems to largely rely on the electronic contact within the electrode. In this regard, the addition of Ketjen black coating serves an important role in forming a contact network to not only facilitate the transport of electrons and Li ion reducing the diffusion-induced stress, but also to buffer the volume change during cycling by mitigating cracking. Extended cycling performance of the LiFePO4/Li4Ti5O12 combination cell in cylindrical configuration, which is more suitable for >1000 cycles, will be performed compared to the coin cell setup. In an effort to further improve the conductivity, LFP and LTO powders with various carbon additives were prepared through simultaneous dispersion of 10 wt pct SWNTs and MWNTs with LFP and LTO powders, respectively, in isopropanol solvent using high power ultrasonication. The solvent was subsequently evaporated to collect the powder. Figure 4(a) demonstrated the rate capability of the half-cells using LFP mixed with 10 wt pct carbon nanotubes as the active materials. Comparison between Figures 4(a) and 1(a) suggests that the rate capability of the commercial LFP can be S24—VOLUME 44A, JANUARY 2013

significantly improved thorough the simple carbon additive approach. The LFP with 10 wt pct SWNT composite cathode delivered ~60 mAh/g at 100 C rate and ~12 mAh/g at even higher rate of 300 C, which is better than most of the synthesized LFO reported so far.[14] Obviously, the improvement in the LFP rate depends on the selection of carbon additives, as the LFP with 10 wt pct of MWNTs exhibited a lower capacity at the same rate, which may have originated from the higher surface area and better conductivity of the SWNTs compared with MWNTs. Similar improvement was observed in the LTO with 10 wt pct of SWNTs, as shown in Figure 4(b), in which close to ~20 and ~30 mAh/g of capacities were demonstrated at rates of 300 C and 200 C, while the LTO without SWNTs showed no capacity. The full cells of the LFP and LTO with different carbon additives are currently being tested and are expected to deliver better rate capabilities. The results will be submitted for publication in due course.

IV.

CONCLUSIONS

LiFePO4 was modified with Ketjen black to improve cycling performance by minimizing the internal resistance that leads to irreversible heat generation. Preliminary electrochemical test based on a full cell from METALLURGICAL AND MATERIALS TRANSACTIONS A

150

10C

100

200

Charge 10%-SWNTs Discharge 10%-SWNTs Charge 10%-MWNTs Discharge 10%-MWNTs

50C 100C

LTO-10%-SWNTs Charge LTO-10%-SWNTs Discharge LTO Charge LTO Discharge

1C

100C

50 200C 300C

Specific capacity (mAh/g)

Specific capacity (mAh/g)

200

10C

150

100 50C 100C

100C

50

200C 300C

0

0 0

100

200

300

400

500

600

Cycle Number

(a)

0

20

40

60

80

100

Cycle Number

(b)

Fig. 4—Electrochemical cycling performance of the LFP and LTO with different carbon additives: (a) LFP at various C rates and (b) LTO at different C rates.

modified LiFePO4 cathode and unmodified Li4Ti5O12 anode showed excellent cycling performance. The cost related to the Li-ion battery is minimized by using commercially available materials with least modification. Thus, LiFePO4 and Li4Ti5O12 Li-ion batteries appear promising for stationary energy storage. The Ketjen black coating was found to be critical for longterm cycling stability of the LiFePO4 and Li4Ti5O12 full cell. The details of the morphology and microstructure and the optimization of the Ketjen black coating will be further studied and reported in a future publication.

ACKNOWLEDGMENTS This work was supported by the Assistant Secretary for Energy Efficiency and Renewable Energy, Office of Vehicle Technologies of the United States Department of Energy, under Contract No. DE-AC02-05CH11231, Subcontract No. 24134 under the Batteries for Advanced Transportation Technologies (BATT) Program. Pacific Northwest National Laboratory is a multiprogram national laboratory operated by Battelle Memorial Institute for the United States Department of Energy under Contract No. DE-AC05-76RL01830.

METALLURGICAL AND MATERIALS TRANSACTIONS A

REFERENCES 1. Z. Yang, J. Zhang, M.C.W. Kintner-Meyer, X. Lu, D. Choi, J.P. Lemmon, and J. Liu: Chem. Rev., 2011, vol. 111 (5), pp. 3577– 3613. 2. T. Xu, W. Wang, M. Gordin, D. Wang, and D. Choi: JOM-US, 2010, vol. 62 (9), pp. 24–31. 3. D. Choi, D. Wang, V.V. Viswanathan, I. Bae, W. Wang, Z. Nie, J. Zhang, G.L. Graff, J. Liu, Z. Yang, and T. Duong: Electrochem. Commun., 2010, vol. 12, pp. 378–81. 4. Z. Yang, D. Choi, S. Kerisit, K.M. Rosso, D. Wang, J. Zhang, G. Graff, and J. Liu: J. Power Sources, 2009, vol. 192, pp. 588–98. 5. V.V. Viswanathan, D. Choi, D. Wang, W. Xu, S. Towne, R.E. Williford, J. Zhang, J. Liu, and Z. Yang: J. Power Sources, 2010, vol. 195, pp. 3720–29. 6. A. Jaiswal, C.R. Horne, O. Chang, W. Zhang, W. Kong, E. Wang, T. Chern, and M.M. Doeff: J. Electrochem. Soc., 2009, vol. 156, p. A1041. 7. A.K. Padhi: J. Electrochem. Soc., 1997, vol. 144, p. 1188. 8. M. Dubarry and B.Y. Liaw: J. Power Sources, 2009, vol. 194, pp. 541–49. 9. X. Li, F. Kang, X. Bai, and W. Shen: Electrochem. Commun., 2007, vol. 9, pp. 663–66. 10. Q. Huang, M. Yan, and Z. Jiang: J. Power Sources, 2006, vol. 156, pp. 541–46. 11. D. Wang, X. Wu, Z. Wang, and L. Chen: J. Power Sources, 2005, vol. 140, pp. 125–28. 12. X. Zhi, G. Liang, L. Wang, X. Ou, J. Zhang, and J. Cui: J. Power Sources, 2009, vol. 189, pp. 779–82. 13. J. Christensen: J. Electrochem. Soc., 2010, vol. 157, p. A366. 14. W. Zhang: J. Electrochem. Soc., 2010, vol. 157, p. A1040.

VOLUME 44A, JANUARY 2013—S25