Quantitative Visualization of Salt Concentration Distributions in Lithium

ure 3c shows the slopes Δdp (rad/m) obtained by a linear fitting at electrolyte bulk region (x = 300-700 µm). As shown in Fig. 3c, the slope at C/2 ...
1 downloads 13 Views 639KB Size
Subscriber access provided by READING UNIV

Communication

Quantitative Visualization of Salt Concentration Distributions in LithiumIon Battery Electrolytes during Battery Operation using X-ray Phase Imaging Daiko Takamatsu, Akio Yoneyama, Yusuke Asari, and Tatsumi Hirano J. Am. Chem. Soc., Just Accepted Manuscript • Publication Date (Web): 15 Jan 2018 Downloaded from http://pubs.acs.org on January 15, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of the American Chemical Society is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Quantitative Visualization of Salt Concentration Distributions in Lithium-Ion Battery Electrolytes during Battery Operation using Xray Phase Imaging Daiko Takamatsu, Akio Yoneyama, Yusuke Asari and Tatsumi Hirano Research&Development Group, Hitachi, Ltd., Hitachi-shi, Ibaraki 319-1292, Japan.

Supporting Information Placeholder ABSTRACT: A fundamental understanding of concentrations of salts in the lithium-ion battery electrolytes during battery operation is important for optimal operation and design of lithium-ion batteries. However, there are few techniques that can be used to quantitatively characterize salt concentration distributions in the electrolytes during the battery operation. In this paper, we demonstrate that in operando X-ray phase imaging can quantitatively visualize the salt concentration distributions that arise in electrolytes during battery operation. From the quantitative evaluation of the concentration distributions at steady-states, we obtained the salt diffusivities in electrolytes with different initial salt-concentrations. Because of no restriction on samples, and high temporal and spatial resolutions, X-ray phase imaging will be a versatile technique for evaluating electrolytes, both aqueous and non-aqueous, of many electrochemical systems.

X-ray phase imaging is a technique that can be used to visualize density differences by detecting the X-ray phase-shift (dp) caused by an X-ray passing through a sample.7 The amount of dp is proportional to the density of the medium on the optical path, as described later. As the cross-section of dp is about 1000 times larger than that of the absorption in a hard X-ray region for light elements, the X-ray phase-contrast imaging has an extremely high-densityresolution compared with ordinary absorption-contrast imaging.8 Due to this advantage, it can be used to visualize small density differences in samples composed of light elements, such as biological soft tissues and organic materials.9 Since the electrolytes of LIBs are also composed of light elements, the X-ray phase imaging is expected to be effective in visualizing small density differences in

(a) S

X-ray

Reference beam M2

Optimal operation and design of lithium-ion batteries (LIBs), particularly in high-power applications such as vehicles, requires knowledge of transport properties in electrolytes during battery operation. As the electrolytes used in commercial LIBs exhibit large concentration polarization due to comparatively low Li+ transference number and salt diffusivity, the polarization is largely related to the transport properties.1,2 For example, it is known that internal resistance of the LIBs temporarily rises by repeating high-rate charge/discharge cycles and that it decreases by stopping the battery operation and rest. Since such a reversible resistance rise occurs only during the battery operation, this phenomenon is speculated to be related to the salt concentration distribution in the electrolyte. However, there are few techniques that can be used to quantitatively characterize salt concentration distributions in the electrolytes during the battery operation. In situ nuclear magnetic resonance (NMR)3,4 and magnetic resonance imaging (MRI)5,6 have been used for quantitatively characterizing electrolyte profiles. However, the cells used in NMR/MRI are in special shapes (e.g. extremely vertically long cells), and this may influence the ionic transport. To quantitatively characterize the salt concentration distributions in the electrolytes during battery operation, it is necessary to establish a novel in operando technique that can accommodate cells similar to actual LIBs with high spatial and temporal resolutions.

1st crystal

M1

A

Cell 2nd crystal

Interference beams

Galvanostat (GS)

(b)

z

(CE)

GS

X-ray camera

Cu Li metal

1 mm

y x

Electrolyte LiFePO4

(WE) L

Al

X-ray

4 mm

Figure 1. (a) Schematic view of two-crystal X-ray interferometer setup. (b) Schematic illustration of spectro-electrochemical cell used for in operando X-ray phase imaging consisting of LiFePO4 composite electrode as a working electrode (WE), Li metal as a counter electrode (CE), and electrolyte. The cell is connected to a galvanostat (GS). Continuous phase-shift images of electrolyte in a square region of red dashed lines were analyzed.

ACS Paragon Plus Environment

Journal of the American Chemical Society

ray phase-shift images of the electrolyte region in the 1.0 charging process acquired at 0 20 80 100 the points A to F of Fig. 2a (t A: 300s (SOC 4%) F 4.0 A B C D 0.5 E B: 600s (SOC 7%) = A: 300, B: 600, C: 1000, D: C: 1000s (SOC 12%) 1700, E: 3000, and F: 7500 s). 0.0 D: 1700s (SOC 20%) 3.5 Since the phase shifts are esE: 3000s (SOC 36%) -0.5 timated as changes from the F: 7500s (SOC 90%) 3.0 first image of t = 0 s (without -1.0 0 2000 4000 6000 8000 an applied current), the 0 200 400 600 800 1000 (CE) (WE) phase-shift change (Δdp) retime (s) x (m) (b) flects the change in electrolyte density caused by charg(CE) Li metal 1 B: 600s C: 1000s D: 1700s E: 3000s F: 7500s ing from the initial state. Figy A: 300s ure 2c displays Δdp profiles x along x-axis of the images 0 from the CE side to the WE 400 μm side shown in Fig. 2b. The (WE) LiFePO4 Charge time -1 Δdp profiles shown in Fig. 2c were averaged on y-axis because they showed almost Figure 2. (a) Constant-current (0.06 mA, C/2 rate) charge profile and (b) corresponding extracted phaseuniform distribution (see Fig. shift change images of electrolyte. Since the phase shifts are estimated as changes from the first image of S2b in SI). Before applying t = 0 s (without an applied current), the phase-shift change (Δdp) reflects the change in electrolyte density current, variation of dp was caused by charging from the initial state. (c) Δdp profiles along x-axis averaged on y-axis of (b). significantly small (see Fig. an electrolyte during battery operation. Eastwood et al. recently S2c in SI), indicating that the electrolyte density is uniform in the demonstrated the use of the X-ray phase imaging to visualize a lithcell. When constant-current charge began, the Δdp in the vicinity ium-metal microstructure formation in a symmetric lithium/lithium of the WE increased while that near the CE decreased. From Fig. cell upon application of a current by distinguishing qualitative lev2c, the Δdp profiles gradually increases until the point D (t = 1700 els of density.10 However, methods to quantitatively evaluate small s), while the profiles were linear and unchanged after the point E (t and continuous changes in density using the X-ray phase imaging = 3000 s). This suggests that the density gradient in the electrolyte techniques have not yet been established, although the small and became the steady state within 3000 s at the C/2 rate. continuous changes are of importance to characterize the salt conFigure 3a displays the Δdp profiles during constant-current charge centration distributions in the electrolyte. In this paper, we first deat 0.06 mA (0.4 mA/cm2, C/2 rate) and 0.36 mA (2.4 mA/cm2, 3C velop an in operando X-ray phase imaging technique for visualizrate), and Fig. 3b illustrates the corresponding voltage profiles. The ing and quantifying the salt concentration distributions in the elecprofiles at 3C rate shows steeper slopes than those at C/2 rate. Figtrolytes. We also demonstrate estimation of salt diffusivities from ure 3c shows the slopes Δdp (rad/m) obtained by a linear fitting at the observed concentration profiles. electrolyte bulk region (x = 300-700 µm). As shown in Fig. 3c, the Figure 1a is a schematic view of the experimental setup of the Xslope at C/2 rate soon became constant (reaching steady state conray phase imaging system. The system is consisting of a two-crystal dition), while that at 3C rate increased monotonically during the (1st/2nd crystals) X-ray interferometer to conduct fine imaging uscharging and reached upper limit voltage before reaching steady ing synchrotron radiation, which is similar to those described elsestate condition. The capacity at 3C rate charging (0.04 mAh) was where.11-14 The incident X-ray is divided at the first wafer (S), then as small as 30% of that at C/2 rate (0.14 mAh), suggesting that the reflected by the second (M1) and third (M2) wafers, and recomsalt concentration gradient at high power application can be a limbined to generate two interference beams at the fourth wafer (A). iting factor of capacity. The high temporal and spatial resolutions A spectro-electrochemical cell was installed as a sample on the opof our method enable us to visualize dynamics of electrolyte dentical path and the cell was connected to a galvanostat (GS) to be sity during high-rate charge/discharge cycles (see Fig. S3 and charged and discharged. A schematic illustration of the cell used in movie S1 in the SI). this study is shown in Fig. 1b. A two-electrode cell is composed of To evaluate the salt distribution during the charging from the X-ray a LiFePO4 composite electrode as a working electrode (WE), lithphase imaging, we conducted a quantitative evaluation of the -3 ium metal as a counter electrode (CE), and 1.0 mol dm (M) phase-shift changes at the steady state condition of C/2 rate chargLiClO4 in a 1:2 volumetric mixture of ethylene carbonate (EC) and ing. In the X-ray phase imaging, the refractive index decrement  ethyl methyl carbonate (EMC) as an electrolyte. The distance beis used as imaging contrast and the dp caused by passing through tween WE and CE was set as about 1 mm in this study (see Fig. S1 the sample is given by in the Supporting Information (SI) for details of cell conditions and

(c)

∆dp (rad)

SOC (%) 40 60

dp (rad)

(a) Voltage (V)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 5

the experimental setup). To investigate the salt concentration distribution in the electrolyte during battery operation, we focused on the continuous phase-shift images of a red square region hereafter. Figures 2a shows a galvanostatic charge profile of the cell during the X-ray phase imaging by applying a current of 0.06 mA (0.4 mA/cm2, corresponding to C/2 rate). The upper horizontal axis denotes state of charge (SOC). The plateau appearing at 3.5 V (vs Li/Li+) indicates that the electrochemical delithiation of LiFePO4 well proceeded in the cell used in this study. Figure 2b shows X-

dp 

2L





(1)

where λ is the wavelength of X-ray and L is the sample thickness along the X-ray path.14 The Δdp caused by charging from the initial state can be obtained as follows

dp  Lre   ni Z i

ACS Paragon Plus Environment

i

(2)

Page 3 of 5

(a)

100

0.5 0.0

0

-0.5

3.0 M 1.0 M 0.3 M

-6

dp (rad)

1.0

C x 10 (M)

-6

-100x10

-1.0 200 CE

Voltage (V)

(b)

WE

(CE)

T RS PQ 4.0

3.0

A B

0

C

D

2000

4000 time (s)

C/2 3C

6000

6000 5000 4000 3000 2000 1000 0

8000

C/2 3C

0

2000

4000 time (s)

6000

8000

Figure 3. (a) Phase-shift change (Δdp) profiles along x-axis at the points in (b). (b) Voltage profiles in constant-current charging at C/2 rate (blue) and 3C rate (red). (c) Slopes of Δdp at the electrolyte bulk region (x = 300-700 µm). where re is the classical electron radius, ni is the number density in a unit volume, and Zi is the atomic number of each atoms con tained in the medium. We use change in the nominal salt concentration (ΔC) from the initial state, and we consider that change in the density (Δni) is proportional to ΔC. From this assumption, we can obtain the correlation between Δdp and ΔC as follows,

 n  dp  A  C , A  Lre    i Z i i  C 

400 600 x (m)

800 (WE)

Figure 4. The phase-shift changes (Δdp) at the steady-state conditions during constant-current (0.06 mA , C/2 rate) charging for initial salt-concentrations of 0.3, 1.0, and 3.0 M. The right vertical axis indicates estimated changes in the nominal salt concentrations (ΔC).

E

3.5

(c) Slope of dp (rad/m)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

(3)

Herein, we consider A with only the salt-concentration change at constant volume. Substituting concrete numerical values into the constant terms in Equation 3, we obtain A = 15628 (rad/M). The calculation procedure for obtaining A is described in SI-3 in SI. The experimentally observed phase-shift gradient (∂∆dp/∂x) under the steady-state condition was 1385 rad/m, as shown in Fig. 2c. By substituting the observed ∂∆dp/∂x into Equation 3, one can determine the salt concentration gradient (∂C/∂x) as 0.089 M/m. Because of the steady-state condition and the linear distribution of the salt concentration, diffusivity of the salt D is estimated to be 3.1 × 10-10 m2/s (see SI-4 in SI for the details). Although the salt diffusivity of the specific electrolyte used in this study (1.0 M LiClO4 in EC/EMC=1/2) has not been reported, the estimated value is comparable to diffusivities in similar electrolytes reported in the literature: 2.6 × 10-10 m2/s and 1.0 × 10-10 m2/s at 1.0 M LiClO4 in propylene carbonate (PC)15,16 This supports the validity of the quantification on the salt distribution changes in this study. Finally, we demonstrated the determination of the salt diffusivities of different initial salt-concentrations. Figure 4 shows the phaseshift profiles under steady-state conditions during constant-current

(0.06 mA, C/2 rate) charge of the electrolytes with initial salt-concentrations of 0.3 (blue) and 3.0 M (red) with that of 1.0 (black) for comparision. The estimated ΔC using Equation 3 is also shown on the right vertical axis of Fig. 4. The slope of the concentration gradient was steeper for higher initial salt-concentration electrolyte. The slopes give the ∂C/∂x of 0.044 and 0.120 M/m for C = 0.3 and 3.0 M, respectively. The estimated diffusivities were 6.6 × 10-10 and 2.4 × 10-10 m2/s for C = 0.3 and 3.0 M, respectively. It is clear that the diffusivity is smaller for the higher salt concentration. This dependency on the salt concentration is consistent with the trend obtained by other conventional methods.4,16 One of the advantages of the present technique is no limitation on the electrode materials and electrolyte salts. The reason to use LiClO4 as the electrolyte salt in this study was its chemical stability against the contact with the atmosphere even if the cell sealing was incomplete. We confirmed that other salts, such as LiPF6, LiBF4, and LiBETI, can also be used (see Fig. S4 in the SI). Furthermore, even if a separator or a gel electrolyte is present between the electrodes, this method can be applied as long as X-ray can pass through them and their geometries keep unchanged. Temporal (a few seconds) and spatial (a few µm) resolutions of our system are enough high compared with other conventional methods (see Table S2 in the SI). In conclusion, we reported for the first time that, using the in operando X-ray phase imaging, quantitative evaluation of the salt concentration distributions in the electrolytes during the battery operation with the high temporal and high spatial resolutions. This technique would be useful for optimizing operational conditions, including transient conditions and design of LIBs, particularly in high-power applications such as vehicles. We envisage that the Xray phase imaging will become a versatile tool for evaluating electrolytes, both aqueous and non-aqueous, of many electrochemical systems, which will further our understanding of the dynamic behavior of electrolytes in actual applications.

ASSOCIATED CONTENT Supporting Information Cell assembly and experimental conditions, dynamic behavior during high-rate chaege/discharge cycles, quantification of X-ray phase-shift imaging, diffusivity estimation, salt dependences, and comparison table between several analytical methods.

AUTHOR INFORMATION

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Corresponding Author [email protected]

Notes The authors declare no competing financial interests.

ACKNOWLEDGMENT This study was carried out under Proposal No. 2012B7601 and 2013A7600, which were approval of the High Energy Accelerator Research Organization. The authors specially thank Dr. Yukinori Koyama (National Institute for Materials Science) for a careful reading of the manuscript and excellent editorial suggestions. We also acknowledge Dr. Haruo Akahoshi (Hitachi, Ltd.) for fruitful discussions, and we thank H. Haruna, M. Kanno, M. Hirooka and K. Kitagawa (Hitachi, Ltd.) for their support of sample preparations.

REFERENCES (1) Nyman, A.; Zavalis, T. G.; Elger, R.; Behm, M.; Lindbergh, G. J. Electrochem. Soc. 2010, 157, A1236-A1246. (2) Nyman, A.; Behm, M.; Lindbergh, G. Electrochimica Acta 2008, 53, 6356-6365. (3) Klett, M.; Giesecke, M.; Nyman, A.; Hallberg, F.; Lindström, R. W.; Lindberg, G.; Furó, I. J. Am. Chem. Soc. 2012, 134, 14654-14657. (4) Krachkovskiy, S. A.; Pauric, A. D.; Halalay, I. C.; Goward, G. R. J. Phys. Chem. Lett. 2013, 4, 3940-3944. (5) Chang, H. J.; Ilott, A. J.; Trease, N. M.; Mohammadi, M.; Jerschow, A.; Grey, C. P. J. Am. Chem. Soc. 2015, 137, 15209-15216. (6) Krachkovskiy, S. A.; Bazak, J. D.; Werhum, P.; Balcom, B. J.; Halalay, I. C.; Goward, G. R. J. Am. Chem. Soc. 2016, 138, 7992-7999. (7) Momose, A.; Takeda, T.; Itai, Y.; Hirano, K. Nat. Med. 1996, 2, 473475. (8) Momose, A.; Fukuda, J. Med.Phys. 1995, 22, 375-379. (9) Momose, A. Jpn. J. Appl. Phys. 2005, 44, 6355-6367. (10) Eastwood, D. S.; Bayley, P. M.; Chang, H. J.; Taiwo, O. O.; Vila-Comamala. J.; Brett, D. J. L.; Rau, C.; Withers, P. J.; Shearing, P. R.; Grey, C. P.; Lee, P. D. Chem. Commun. 2015, 51, 266-268. (11) Yoneyama, A.; Momose, A.; Seya, E.; Hirano, K.; Takeda, T.; Itai, Y. Rev. Sci. Instrum. 1999, 70, 4582-4586. (12) Yoneyama, A.; Momose, A.; Koyama, I.; Seya, E.; Takeda, T.; Itai, Y.; Hirano, K.; Hyodo, K. J. Synchrotron Rad. 2002, 9, 277-281. (13) Yoneyama, A.; Takeda, T.; Tsuchiya, Y.; Wu, J.; Lwin, T. T.; Koizumi, A.; Hyodo, K.; Itai, Y. Nucl. Instrum.Methods Phys. Res. A 2004, 523, 217222. (14) Yoneyama, A.; Hyodo, K.; Takeda, T. Appl. Phys. Lett. 2013, 103, 204108. (15) Sullivan, J. M.; Hanson, D. C.; Keller, R. J. Electrochem. Soc., 1970, 117, 779-780. (16) Georen, P.; Lindbergh, G. Electrochim. Acta 2004, 49, 3497-3505.

ACS Paragon Plus Environment

Page 4 of 5

Page 5 of 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Table of Contents

ACS Paragon Plus Environment

5