Charge Curves of Lithium-Ion Battery

Sep 19, 2016 - We show that basic electrochemical relationships, that is, the Nernst equation and the Butler–Volmer equation, are able to reproduce ...
0 downloads 0 Views 669KB Size
Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

On the Asymmetry of Discharge/Charge Curves of Lithium-Ion Battery Intercalation Electrodes Florian Hall, Sabine Wussler, Hilmi Buqa, and Wolfgang G Bessler J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.6b07949 • Publication Date (Web): 19 Sep 2016 Downloaded from http://pubs.acs.org on September 22, 2016

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.

The Journal of Physical Chemistry C 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 27

The Journal of Physical Chemistry

Hall, Wußler, Buqa, Bessler, J. Phys. Chem. C., revised (09/2016) 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

On the Asymmetry of Discharge/Charge Curves

of Lithium-Ion Battery Intercalation Electrodes Florian Hall †, Sabine Wußler ‡, Hilmi Buqa ‡, and Wolfgang G. Bessler *,†



Institute of Energy Systems Technology (INES) Offenburg University of Applied Sciences Badstrasse 24 77652 Offenburg Germany ‡

Leclanché GmbH Industriestraße 1 77731 Willstätt Germany

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry

Page 2 of 27

Hall, Wußler, Buqa, Bessler, J. Phys. Chem. C., revised (09/2016) 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

Abstract.

Nickel cobalt aluminum oxide (NCA) based lithium-ion battery electrodes exhibit a distinct asymmetry in discharge/charge behavior towards high bulk stoichiometry (low state of charge). We show that basic electrochemical relationships, that is, the Nernst equation and the ButlerVolmer equation, are able to reproduce this behavior when a two-step reaction mechanism is assumed. The two-step mechanism consists of (1) lithium-ion adsorption from the electrolyte onto the active material particle surface under electron transfer, and (2) intercalation of surfaceadsorbed lithium atoms into the bulk material. The asymmetry of experimental half-cell data of an NCA electrode cycled at 0.1 C-rate can be quantitatively reproduced with this simple model. The model parameters show two alternative solutions, predicting either a saturated (highlycovered) or a depleted surface for high bulk lithiation.

1

Introduction Lithium nickel cobalt aluminum oxide (LiXNi0.8Co0.15Al0.05O2, NCA) is widely used as

lithium-ion battery cathode material, as it provides a high capacity, good rate capability and low cost 1,2. Figure 1a shows experimental discharge/charge curves of an NCA cathode versus Li/Li+. The data exhibit a distinct asymmetry of the discharge and charge potential in the range of 3.03.6 V (i.e., at high lithium stoichiometry or low state of charge). This behavior was also observed, although not discussed, by other groups 3–6. To the best of our knowledge, the origin of the asymmetry of discharge/charge curves of NCA electrodes has not yet been explained. In this paper we show that an asymmetry in discharge/charge potential follows from fundamental electrochemical relationships (Nernst equation, Butler-Volmer equation) applied to

ACS Paragon Plus Environment

2

Page 3 of 27

The Journal of Physical Chemistry

Hall, Wußler, Buqa, Bessler, J. Phys. Chem. C., revised (09/2016) 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

an intercalation material consisting of lithium atoms and vacancies. The asymmetry is pronounced when a two-step reaction mechanism consisting of adsorption and intercalation is assumed. We follow here a classical approach of physical chemistry: A macroscopic observation (here: asymmetry of discharge/charge potential) gives rise to a postulated microscopic mechanism (here: two-step adsorption/intercalation reaction). The postulated mechanism is verified by comparing its prediction to the macroscopic observation, recognizing that other postulated mechanisms may also explain the observation. Experimental evidence is shown in Section 2. The behavior of single-step and two-step mechanisms is derived from electrochemistry theory in Section 3. Section 4 demonstrates the application of the theory to the experimental data. Finally, Section 5 discusses the findings and alternative mechanisms.

ACS Paragon Plus Environment

3

The Journal of Physical Chemistry

Page 4 of 27

Hall, Wußler, Buqa, Bessler, J. Phys. Chem. C., revised (09/2016)

a) charge

discharge

b)

discharge

charge

c) 0.1 C 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

0.5 C 1C

Figure 1: Experimental half-cell potentials of a) LiNi0.8Co0.15Al0.05O2 (NCA) and b) Li4Ti5O12 (LTO) versus Li/Li+, as well as c) full-cell voltage of an NCA/LTO lithium-ion battery.

ACS Paragon Plus Environment

4

Page 5 of 27

The Journal of Physical Chemistry

Hall, Wußler, Buqa, Bessler, J. Phys. Chem. C., revised (09/2016) 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

2

Experiment Large format full-cell measurements were carried out using commercial (Leclanché)

lithium-ion battery pouch cells consisting of lithium titanate (Li4Ti5O12, LTO) as negative electrode, lithium nickel cobalt aluminum oxide (LiNi0.8Co0.15Al0.05O2, NCA) as positive electrode and 30 µm Al-oxide ceramic laminated type separator (produced at Leclanché). The Al-oxide particles are embedded in a polyvinylidene difluoride (PVDF, Arkema, France) matrix. The cells have a nominal capacity of 16 Ah and nominal voltage of 2.3 V. The cells were cycled (BasyTec XCTS) at 20 °C (CTS Type T −40/200 climate chamber) using a constant current/constant voltage (CCCV) protocol for both charge (2.7 V/0.1 C cutoff) and discharge (1.7 V/0.1 C cutoff) at different C-rates. In this study, lithium titanate (Li4Ti5O12, LTO) with 0.7 µm (d50) average particle size and lithium nickel cobalt aluminum oxide (LiNi0.8Co0.15Al0.05O2, NCA) with 12.7 µm (d50) average particle size were used as active materials as received from supplier (Toda Kogyo, Japan). Slurries for LTO and NCA electrodes were processed with innovative water based binder solutions7. In order to enhance the electrode conductivity, Super P (Imerys Graphite & Carbon, Bodio, Switzerland) as conductive additive was dispersed in addition to active materials and homogenized for 1 h. During the aqueous slurry preparation with CMC (carboxy-methyl cellulose) and SBR (styrene butadiene rubber) type binder, the viscosity of the slurry was carefully controlled in order to get a uniform dispersion (no agglomeration) between the binder and the electrode materials. The resulting slurry was mixed thoroughly, and coated in pilot-line production on a 12 µm copper foil (LTO electrode) and 20 µm aluminum foil (NCA electrode). By the optimization of slurry and processing conditions, mechanically stable and flexible standard electrodes were manufactured. The final thickness of electrodes after calendering was 97 µm (LTO electrode) and 75 µm (NCA electrode) with an electrode density of 1.5 g/cm3 for

ACS Paragon Plus Environment

5

The Journal of Physical Chemistry

Page 6 of 27

Hall, Wußler, Buqa, Bessler, J. Phys. Chem. C., revised (09/2016) 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

LTO and 2.1 g/cm3 for. The electrode loading allows the continuous symmetric charging and discharging at 4-5 C as a rate capability. For half-cell tests, electrodes having 1.3 cm2 geometrical areas were punched from large production sheets. The half cells were cycled at room temperature using a computer controlled cell capture cycling device (CCCC; Astrol Electronic AG). The electrochemical measurements were performed in titanium two-electrode Swagelok cells that were vacuum dried before assembly at 120 °C overnight. The internal arrangement of the cell consisted of a) a working electrode (LTO or NCA), b) 1 mm thick soft glass-fiber separator type EUJ116, Hollingsworth & Vose, Ltd., England and c) a 0.75 mm thick lithium foil (Alfa Aesar, Johnson Matthey GmbH, purity 99.9 %) as counter electrode. Before use, the glass fiber separator was dried at 120 °C under a reduced argon pressure and stored in an argon atmosphere. The LiPF6 salt based electrolytes contained less than 10 ppm of water, as determined by Karl-Fischer titration. The cells were assembled in an argon filled glove box, with less than 1 ppm of oxygen and water. The cell components were put under a light pressure with a spring pressure of approximately 2 kg/cm2. The galvanostatic measurements of both (LTO and NCA) electrodes were performed at low current density (0.1 C-rate for the first cycle). The cells containing NCA electrodes were cycled between 3.0 V and 4.3 V vs. Li/Li+. In the first cycle, the cells were charged with a C/10 rate to 4.3 V, kept at 4.3 V until C/20 rate, and then discharged to 3.0 V at C/10. The experiments with LTO electrodes were performed under comparable conditions but within the voltage limits of 2.0 V and 1.0 V vs. Li/Li+. Experimental results are shown in Figure 1 for both half cells (panels a, b) and the full cell (panel c). The NCA electrode (Figure 1a, positive electrode in the full cell) shows a pronounced asymmetry towards low voltages (high lithium stoichiometry), as discussed above.

ACS Paragon Plus Environment

6

Page 7 of 27

The Journal of Physical Chemistry

Hall, Wußler, Buqa, Bessler, J. Phys. Chem. C., revised (09/2016) 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

The LTO electrode (Figure 1b, negative electrode in the full cell) shows a plateau at 1.56 V and an asymmetrical behavior towards both high and low stoichiometries, as described and interpreted before8. The asymmetry of the NCA half cell translates into the full-cell data (Figure 1c), while the LTO electrode is balanced to within its plateau.

3 3.1

Theory Single-step charge-transfer reaction

An intercalation material can be chemically described as two-species system, consisting of lithium atoms and vacancies

9,10

, which we denote as Li[NCA] and V[NCA], respectively. Here

and in the following, we use square brackets to denote the bulk host material (solvent). Sites of the intercalation host can either be occupied, Li[NCA], or free, V[NCA]. This is shown schematically in Figure 2. An intercalation material is therefore an example of a lattice gas

11

,

although the present model does not make further use of lattice gas theory.

ACS Paragon Plus Environment

7

The Journal of Physical Chemistry

Page 8 of 27

Hall, Wußler, Buqa, Bessler, J. Phys. Chem. C., revised (09/2016) 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

a) e–[graphite] Li+[electrolyte]

V[NCA] Li[NCA] b) e–[graphite] Li+[electrolyte]

Li(s)

(s) V[NCA] Li[NCA]

Figure 2: Schematic illustration of an intercalation material with a) single-step charge-transfer reaction and b) two-step mechanism consisting of adsorption with charge transfer and intercalation. The electron and lithium ion originate from conductive additive and electrolyte, respectively, the microstructures of which are not explicitly resolved in this figure. Square brackets denote bulk species, round brackets denote surface species.

The half-cell reaction can be described in a single-step mechanism (Figure 2a) according to Li+[electrolyte] + e–[graphite] + V[NCA] ⇄ Li[NCA] .

(1)

The thermodynamics of this reaction, described in the form of its equilibrium potential   , is given by 9,10

ACS Paragon Plus Environment

8

Page 9 of 27

The Journal of Physical Chemistry

Hall, Wußler, Buqa, Bessler, J. Phys. Chem. C., revised (09/2016) 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

  = − where, 









ln 





 ,

(2)

is the bulk stoichiometry of lithium in the intercalation material (i.e., the

concentration of intercalated lithium over the total concentration of intercalated lithium and vacancies), R is the ideal gas constant, T the temperature, F Faraday’s constant, and Δ" # the free enthalpy of the reaction at standard conditions (



= 0.5) given as

# # # Δ" # = '  − '( )*+,-).+ − '/



− '#01,234+ ,

(3)

where ' # are the chemical potentials of the involved species at standard conditions. Equation (2) is the well-known Nernst equation

12

applied to a two-species system under assumption of a

lithium reference electrode 9. The kinetics of reaction (1) is given by the Butler-Volmer equation 12, ;

5 = 5 # 6exp : 0.5, resulting in an S-shaped dependency of coverage on bulk stoichiometry. For

case (c), the S-shaped behavior is shifted towards lower bulk stoichiometries. The discharge/charge curves in all cases show an asymmetrical behavior (upper panels of Figure 4). Case (a) shows the identical behavior as the single-step mechanism due to the twospecies behavior of the surface. Cases (b) and (c) show a strongly pronounced asymmetry, the charge and discharge branches exhibiting a difference of up to ca. 1 V and particularly wide plateaus of both discharge and charge branches. In conclusion, the assumption of a two-step mechanism results in a complex interdependence between bulk and surface thermodynamics, surface coverage, and kinetics. As result, discharge/charge curves show a highly nonlinear behavior with pronounced asymmetry.

Li(s)

/ 1

Potential E vs. Li/Li

+

/ V

c)

Surface coverage

Li(s)

/ 1

Potential E vs. Li/Li

+

/ V

b)

Surface coverage

Li(s)

/ 1

Potential E vs. Li/Li

+

/ V

a)

Surface coverage

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

Figure 4: Results from the two-step reaction mechanism for three different choices of thermodynamic parameters. The upper panels show equilibrium and half-cell potentials, the lower panels show corresponding surface coverages.

ACS Paragon Plus Environment

14

Page 15 of 27

The Journal of Physical Chemistry

Hall, Wußler, Buqa, Bessler, J. Phys. Chem. C., revised (09/2016) 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

4

Results As the single-step mechanism shows an asymmetric discharge/charge behavior, it is

worthwhile investigating if this mechanism can explain the experimental NCA half-cell curve. Here and in the following, we assume that the voltage range of the experimental cycles (4.3…3.0 V vs. Li/Li+) correspond to a bulk stoichiometry range of 



= 0.25 … 1.0 3,9,10,13,14. Figure

5 shows the experimental data and calculations with the single-step model. For the calculations, # '



and 5/5 ## were adjusted to the experimental data at a bulk stoichiometry of 



=

0.5. Clearly, the model behaves very different to the experiment. Particularly, neither the slope nor the asymmetry of the experimental data can be reproduced by the model. The conclusion from this result is that NCA does not behave as ideal intercalation material. This was concluded

before by Colclasure et al.9 for both NCA and graphite. The reason is that the host lattice is not an ideal “solvent”, but shows considerable interaction with the intercalated lithium, leading to, for example, steps or slope variations in the half-cell potential. We therefore next tested the two-step mechanism against the experimental data. To this # # ## goal, we performed a simultaneous fit of '  (  ), '(T) (  ), and 5/5 . Note

that, here, the thermodynamic data were assumed to depend on   . The chemical potentials

of the other species (VNCA, (s), Lil electrolyte, e graphite) were kept at their reference

value of zero. The fitted value for the exchange current density is 5/5 ## = 0.14 . Results of the

calculations are shown in Figure 6. Experimental and calculated discharge/charge curves are shown in Figure 6a). The model is able to fully reproduce the experimental data with a deviation of < 0.2 % (average) and < 18 % (maximum). In particular, both the slope and the asymmetry are # reproduced. The slope is reproduced because chemical potential '



is not constant as in the

ACS Paragon Plus Environment

15

The Journal of Physical Chemistry

Page 16 of 27

Hall, Wußler, Buqa, Bessler, J. Phys. Chem. C., revised (09/2016) 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

single-step model, but is assumed to depend on   , thus allowing the fit to follow the experimental data. More importantly, the asymmetry is also reproduced. This is due to the twostep mechanism which results in a strong nonlinearity of