Subscriber access provided by University of Colorado Boulder
Article
Electrochemical Cross-Talk Leading to Gas Evolution and Capacity Fade in LiNi Mn O/Graphite Full-Cells 0.5
1.5
4
Barbara Michalak, Balázs B. Berkes, Heino Sommer, Torsten Brezesinski, and Jürgen Janek J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.6b11184 • Publication Date (Web): 15 Dec 2016 Downloaded from http://pubs.acs.org on December 17, 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 17
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 Journal of Physical Chemistry
Electrochemical Cross-Talk Leading to Gas Evolution and Capacity Fade in LiNi0.5Mn1.5O4/Graphite Full-Cells Barbara Michalak,† Balazs B. Berkes,*,† Heino Sommer,†,‡ Torsten Brezesinski,*,† and Jürgen Janek†,+ †
Battery and Electrochemistry Laboratory, Institute of Nanotechnology, Karlsruhe
Institute of Technology, Hermann-von-Helmholtz-Platz 1, 76344 EggensteinLeopoldshafen, Germany. ‡
BASF SE, 67056 Ludwigshafen, Germany.
+
Institute of Physical Chemistry, Justus-Liebig-University Giessen, Heinrich-Buff-Ring
17, 35392 Giessen, Germany.
Abstract Continuous destruction of the solid electrolyte interphase (SEI) on the graphite-based negative electrode during cycling operation is a significant degradation mechanism that raises safety concerns and limits the cycle life of LiNi0.5Mn1.5O4 (LNMO)/graphite full-cells. Herein, we report on gassing phenomena which are typically concomitant with SEI destruction processes. Abrupt H2 evolution is observed by differential electrochemical mass spectrometry and pressure measurements at the end of discharge. Using a lithium reference electrode reveals that the gassing, which intensifies with cycling, is caused by an increase in the anode potential. Lithium is irreversibly consumed upon SEI formation, but this loss is not compensated for by the intrinsic degradation of LNMO in the first cycle. When the potential of the anode on discharge increases above approximately 0.9 V, the SEI is instantly damaged, causing gas generation, and eventually capacity fade. We show that this (“mediatorfree”) cross-talk phenomenon can be suppressed to various degrees by either using a precycled graphite electrode or an LNMO material having a higher initial irreversible capacity loss.
1 ACS Paragon Plus Environment
The Journal of Physical Chemistry
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
Introduction A key issue in the development of high-energy-density lithium-ion batteries (LIBs) is the improvement of cell life time. The performance of LIBs not only depends on the properties of the electrode materials used, but also on their interplay during operation. LiNi0.5Mn1.5O4 (LNMO) has a high theoretical specific capacity of 147 mAh/g and shows good long-term stability when cycled against lithium.1,2 However, LNMO cells containing a graphite-based negative electrode suffer from severe capacity fading.3,4 To understand the phenomena leading to degradation in these cells, it is necessary to investigate how the processes occurring on graphite and LNMO affect each other. Such where one electrode has a notable effect on the other (opposite) electrode are summarized under the term “electrochemical cross-talk”.5,6 A well-known example of unfavorable cross-talk is the dissolution of transition metal ions from LNMO and their migration to the graphite anode, where they can poison or damage the solid electrolyte interphase (SEI).7 Another example – which until now has not been discussed in this context and forms the main subject of the present paper – is related to the SEI formation on graphite, which is a basic requirement for stable cycling performance of LIBs.8 The SEI protects the anode from further decomposition reactions, and thus from undesired gas evolution.9,10 However, its formation is associated with an irreversible lithium loss, which, if not compensated for by the cathode, may adversely affect the cell behavior. We note that this kind of cross-talk does not involve any mediators except the electrolyte (in case of metal dissolution, the mediators are the dissolved metal ions themselves). Loss of active lithium as a result of SEI reformation (or repair) on graphite due to deposition of manganese species, for example, is recognized as one of the causes for capacity degradation in LNMO/graphite cells.3-4,6,7 To overcome this issue, a sacrificial lithium source additive (e.g., in the form of stabilized lithium metal powder or common lithium metal)
11,12
or a prelithiated graphite anode6,13 may be used. Although this helps to
increase the cell life time to some degree, it is not practical and capacity fade is still present.6,12,13 Gaseous products are known to be generated during electrolyte decomposition.14-19 The gas evolution is typically significant in the first couple of cycles, but continues with cycling, especially at high potentials. Furthermore, as we will demonstrate in this work, electrode cross-talk leading to the destruction of parts of the graphite SEI is 2 ACS Paragon Plus Environment
Page 2 of 17
Page 3 of 17
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 Journal of Physical Chemistry
also responsible for gassing. In recent years, it has been shown by differential electrochemical mass spectrometry (DEMS) and gas chromatography (GC) that gases like hydrogen (H2), ethylene (C2H4) and carbon monoxide (CO) are predominantly generated on graphite.14,15 Both C2H4 and CO are side products of the SEI formation process, while H2 is formed by reduction of H+ (in the beginning mostly from unavoidable water impurities in the electrolyte, but also from decomposition products). As shown by Bernhard et al. in a recent paper, a well formed SEI can significantly suppress the H2 evolution even in the presence of added water.20 In contrast, carbon dioxide (CO2) evolves at the cathode side through electrolyte oxidation.16,17 Therefore, release of gases like H2 and C2H4 upon prolonged cycling is an indirect sign of SEI destruction. Measurements conducted on three-electrode cells have shown a significant increase in discharge potential of graphite and LNMO during cycling.6,13 Levi et al. explained this by self-discharge of the anode (irreversible lithium loss) mediated by “poisonous” decomposition products from LNMO.6 We note that increased discharge potentials have a strong and adverse effect on the graphite performance. This has been demonstrated for other cell systems, including LiMn2O4/graphite21 at elevated temperatures and LiNi0.8Co0.15Al0.05O2/graphite22, and by theoretical calculations.23 Structural changes, including exfoliation and SEI destruction,10,24 seem to occur at potentials greater than or equal to 0.9 V vs. Li/Li+. However, to the best of our knowledge, there are no studies on LNMO/graphite full-cells that correlate electrochemical data with the gassing behavior during prolonged cycling. Recently, we have shown the pattern of evolving gases for practical LNMO/graphite cells.19 An unexpected release of H2 was observed at the end of discharge. Here, this so far unexplained feature is discussed in detail, which allows more far-reaching conclusions on the capacity fade. The focus is on the correlation between the gassing and the (partial) destruction of the graphite SEI. The gassing behavior was investigated via DEMS and pressure measurements. While the latter only provide information on the total amount of gases in the cell, they can be clearly distinguished by means of DEMS. Using a lithium reference electrode allowed us to monitor the anode and cathode potential separately, and thus to identify one of the key issues with LNMO/graphite cells, namely the potential-driven destruction of the SEI on graphite during discharge. 3 ACS Paragon Plus Environment
The Journal of Physical Chemistry
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
Experimental Section Materials and Electrode Processing. The LNMO-based (BASF SE) positive electrodes contained 88 wt.% cathode material, 3 wt.% conductive carbon black (Super C65, Timcal), 3 wt.% graphite (SFG6L, Timcal) and 6 wt.% polyvinylidene fluoride (Kynar HSV 900). They were prepared by casting an N-ethyl-2-pyrrolidone (Sigma-Aldrich) slurry onto Al foil using a smart coater (KTF-S, Mathis AG) with a drying temperature set to 140 °C, followed by calendering (Sumet Messtechnik) at 32 N/mm and 25 °C. The loading was (2.0 ± 0.1) mAh/cm2. The graphite-based negative electrodes (2.3 mAh/cm2) were received from BASF SE. 40 mm diameter electrodes were cut and dried overnight under vacuum at 100 °C. SelectiLyteTM LP57 (1 mol/dm3 LiPF6 in a 3:7 weight ratio of ethylene carbonate [EC] and ethyl methyl carbonate [EMC]) – with water content less than 2 ppm – was used as the electrolyte. The cells were assembled inside an argon-filled glovebox (MBraun) using porous glass microfiber filter paper (GF/A, Whatman) as the separator and 600 µL of LP57. For DEMS, pressure and three-electrode measurements, home-built hard-case cells were used.25 A small piece of lithium metal served as the reference electrode in three-electrode cells. Precycled graphite-based negative electrodes were obtained from disassembled LNMO/graphite cells after the initial cycle. They were rinsed with EMC and then cycled against pristine LNMO-based positive electrodes. Instrumentation. All cells were cycled at 25 °C using a BioLogic VSP300 potentiostat. The setup used for DEMS has already been described elsewhere.25 The DEMS cell was evacuated and filled with helium for four times to remove the glovebox atmosphere. Helium served as an inert carrier gas and its flow rate was set to 2 cm3/min. The mass spectrometer used was an OmniStar GSD 320 O2 instrument (Pfeiffer Vacuum GmbH) with a detection limit in the ppb range. For pressure measurements, the same cell as that used for DEMS was equipped with a pressure sensor (PAA33X-V-3, Omega). Powder XRD patterns were recorded on a Bruker D8 Advance with a Cu-Kα1 radiation source and a LYNXEYE 1D strip detector. SEM was performed on a LEO 1530 microscope operated at 10 keV. ICPOES was conducted using both a PerkinElmer Optima 4300 DV and Thermo Scientific iCAP 7600. N2-physisorption at 77 K was performed on a Micromeritics ASAP 2020. 4 ACS Paragon Plus Environment
Page 4 of 17
Page 5 of 17
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 Journal of Physical Chemistry
Results and Discussion The voltage profiles of cells containing a standard LNMO-based positive electrode and state-of-the-art graphite-based negative electrode cycled under various conditions (with 5 h open circuit voltage (OCV) period after each charge/discharge cycle) and the corresponding DEMS signals for H2 with m/z = 2 are presented in Fig. 1. We note that only H2 is shown for clarity. The whole spectrum of gases generated during cycling operation can be found elsewhere.19 Fig.
1a shows the
charge/discharge curves and measured H2 signal for a cell cycled between 3.3 V and 4.8 V at different rates ranging from C/10 to 1C. H2 evolution is clearly visible during the charge cycles. Its amount (area under the curve) is notably high in the initial cycles due to reduction of H+ (mainly from water impurities in the electrolyte) at the graphite anode – which is not yet sufficiently protected by an SEI layer. In addition, the m/z = 2 signal is observed at the end of discharge. In the first cycle at C/10, the peak is small, but clearly distinguishable from the background. In the subsequent cycles at higher C-rates, it is still apparent; however, the peak can hardly be separated from the overall MS signal at rates greater than or equal to 1C. Interestingly, the H2 signal increases significantly in the fifth cycle at C/10 and dominates the pattern. The appearance of the peak at the end of discharge suggests that there is a correlation between gas release and potential. Thus, the cut-off potentials were varied in a systematic fashion. Fig. 1b,c shows the voltage profiles of LNMO/graphite cells charged or discharged to different states of charge (SOC) or discharge (SOD) and the corresponding H2 signals from DEMS. As evident, the SOC has no effect on the appearance of H2 at the end of discharge. However, the SOD or, in other words, the discharge cut-off potential strongly affects the H2 evolution. For 80% SOD and below (second to fifth cycle), the peak is gone. When the cell is fully discharged in the sixth cycle, H2 generation is again visible. This indicates that the gassing at the end of discharge can be eliminated by adjusting the cut-off potential on discharge which, however, decreases the reversible cell capacity. The discharge capacity in the second cycle (80% SOD) is only 116 mAh/g, while it is 128 mAh/g for a fully delithiated cell (100% SOD). Overall, we conclude from the data in Fig. 1 that the gas generation is closely connected to the degree of delithiation of graphite and lithiation of LNMO. 5 ACS Paragon Plus Environment
The Journal of Physical Chemistry
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 1. Voltage profiles and the corresponding DEMS signals for H2 for standard LNMO/graphite cells cycled under various conditions: (a) 3.3-4.8 V range at different C-rates, (b) variation of the state of charge and (c) discharge at C/10. Note that 100% corresponds to cut-off potentials of 4.8 V and 3.3 V, respectively. The arrows indicate the appearance of H2 at the end of discharge. In the galvanostatic charge/discharge tests, H2 evolves apparently in the OCV period after the discharge cycles. However, its evolution is not related to the OCV period, but rather the onset is potential-dependent. This is confirmed by the combination of cyclic voltammetry with DEMS; the H2 peak starts to appear at around 3.6 V and reaches its maximum at the cut-off potential of 3.3 V (see Fig. S1). During 6 ACS Paragon Plus Environment
Page 6 of 17
Page 7 of 17
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 Journal of Physical Chemistry
galvanostatic cycling, the potential changes suddenly at the end of discharge, thus not allowing a precise determination of the onset potential. We also note that galvanostatic cycling with no OCV period after discharge would lead to overlapping signals from H2 evolution at the end of discharge and beginning of charge and would mask this effect, which may explain why it has not been noticed earlier.
Figure 2. LNMO and graphite electrode potentials vs. lithium reference and the corresponding
pressure
curves
for
different
cell
systems:
(a)
standard
LNMO/graphite, (b) standard LNMO/precycled graphite and (c) low-surface-area LNMO/graphite. The first four cycles at rates of C/10, C/4, C/2 and C/10 are shown. Shaded areas indicate the 5 h OCV period after discharge. The zoomed-in region shows more details about the pressure increase at the end of discharge.
To gain more insight into the influence of anode and cathode potentials on the H2 evolution at the end of discharge, three-electrode measurements were conducted on different LNMO-based cells. Fig. 2 presents the LNMO and graphite potential vs. lithium reference electrode and the corresponding pressure curves. The voltage profile of LNMO reveals the typical Ni oxidation (4.68 V and 4.75 V) and reduction plateaus (4.67 V and 4.72 V) on charge and discharge.26 As expected, the intercalation and deintercalation of lithium into/from graphite occurs in stages at 0.21 V, 0.12 V and 0.08 V and 0.10 V, 0.14 V and 0.23 V, respectively.10 Fig. 2a shows that because of SEI formation, the initial gas generation in standard LNMO/graphite 7 ACS Paragon Plus Environment
The Journal of Physical Chemistry
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 8 of 17
cells is significant (see ref. 27 for details on the shape of similar pressure curves). The pressure increases by approximately 28 mbar in the first cycle. Gas evolution also occurs in the subsequent cycles, causing a pressure increase of 50 mbar after four cycles (93 mbar after sixty cycles). In the following, the pressure data are correlated with the potentials of LNMO and graphite. The discharge cut-off potential was always kept at 3.3 V. However, as can be seen from Fig. 2a, both the anode and the cathode potential changes significantly during operation. In the initial cycle, the discharge potential of graphite reaches 0.8 V, while it increases to approximately 1.2 V in the fourth cycle. At this point, an abrupt pressure increase occurs (see zoomed-in region of the pressure curve in Fig. 2a), which corresponds to the H2 evolution mentioned above (see Fig. 1a). The pressure remains constant throughout the first OCV period, while gas generation is visible from the pressure curve in the subsequent OCV periods, especially after the fourth cycle. This is also reflected in the voltage profiles: the potential of LNMO is relatively stable in the OCV periods; however, that of graphite becomes unstable with prolonged cycling, thereby indicating that the SEI is damaged and its reformation (or repair) contributes to gas evolution. This effect is much more pronounced when the graphite discharge potential is greater than or equal to 0.9 V. On the basis of these data, it seems clear that the SEI is instantly damaged at such high potentials, thus creating new reactive surfaces and causing reductive electrolyte decomposition. Scheme 1 depicts the effect of potential-driven SEI destruction on graphite.
Scheme 1. Schematic of the potential-driven SEI destruction on graphite during discharge
and
the
associated
H2
evolution
due
to
reductive
electrolyte
decomposition. CC and OCV stand for constant current and open circuit voltage, respectively.
8 ACS Paragon Plus Environment
Page 9 of 17
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 Journal of Physical Chemistry
The shifts in potential are caused by the irreversible lithium loss upon SEI reformation on graphite and other lithium consuming processes. The Coulombic efficiency of a standard LNMO/Li cell is around 94%, while it is only 91% for a graphite/Li cell in the initial cycle. This means more lithium is consumed by graphite than LNMO would “lose” naturally. Consequently, the discharge potential is shifted to more positive values (≥4.2 V) than that of “pristine” LNMO against Li (~3.3 V). In the subsequent cycles, the potential increases even further because of additional lithium loss as a result of SEI reformation (or repair). Overall, this process can be described as a positive feedback loop: the reformation of the SEI consumes lithium that will force the graphite anode to a higher discharge potential in the subsequent cycle, thereby damaging parts of the SEI again. The continuously increasing lack of lithium causes the gradual increase in discharge potential of both graphite and LNMO and accelerates the degradation process. The profound effect that the graphite potential has on the SEI is also apparent from DEMS measurements conducted on graphite/Li cells, in which the discharge cut-off potential (note that, as for full-cells, discharge is also defined as the delithiation of graphite) was varied to simulate the conditions that are thought to create the H2 evolution in the standard LNMO/graphite cells. As can be seen in Fig. 3, the H2 generation is significant in the first charge cycle – mainly due to water impurities in the electrolyte – with a local maximum at the cut-off potential of 10 mV. In the following cycles, however, minor amounts evolve upon charging. During the course of discharge, there is virtually no H2 generation when the cut-off potential is less than or equal to 0.9 V. This is indicative of the presence of a fairly stable SEI. However, discharging to 1.2 V leads to the appearance of a strong m/z = 2 signal, which provides clear evidence of potential-dependent destruction of the SEI (note that “SEIfree” graphite surfaces react instantly with the surrounding electrolyte). The reason why the H2 peak is strong seems to be associated with the high reactivity of H+ and the abundance of reaction sites on the graphite surface. Again, the source of H+ is either water or side products of cathodic reactions like alcohols.28
9 ACS Paragon Plus Environment
The Journal of Physical Chemistry
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 3. Voltage profiles and the corresponding H2 signal from DEMS for a graphite/Li cell discharged to different potentials at C/10. Measurement artifacts are indicated by asterisks. The arrow indicates the appearance of H2 at the end of discharge.
To keep the irreversible lithium loss in the initial cycle at a minimum, a precycled graphite anode was used in combination with a standard LNMO cathode (Fig. 2b). Because the graphite SEI is already present, the pressure increase during the first charge cycle is much lower than in the standard LNMO/graphite cells (12 mbar vs. 19 mbar). However, the increase in the subsequent discharge is similar, thereby indicating that the gas evolution occurs primarily at the cathode side. In contrast to the standard LNMO/graphite system, the discharge potentials of LNMO and graphite are lower during cycling operation. Also, both the electrode potentials and the pressure are relatively stable throughout the OCV periods. The pressure increases by approximately 38 mbar after four cycles (76 mbar after sixty cycles), without any abrupt gas generation after discharge. This is also confirmed by DEMS (see Fig. S2), showing that there is either no or only minor potential-driven SEI destruction when using a precycled graphite-based electrode as the anode. The smaller lithium loss causes smaller shifts in the potential, and thus the balancing is better maintained. The observed pressure increase seems to be predominantly associated with the oxidative electrolyte decomposition at high potentials and changes in the structure and properties of the anode SEI due to, for example, deposition of transition metal species. However, on the long-term scale, these processes will also slowly lead to changes in the LNMO and graphite potentials.
10 ACS Paragon Plus Environment
Page 10 of 17
Page 11 of 17
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 Journal of Physical Chemistry
Figure 4. Low-magnification SEM images of (a) standard LNMO and (b) low-surfacearea LNMO.
Yet another way to prove the role of the irreversible lithium loss in the gas evolution is to substitute the standard LNMO with a material showing a higher degradation in the initial cycle. This material is referred to as low-surface-area LNMO in the following. Fig. 4 presents scanning electron microscopy (SEM) images of both the low-surfacearea and standard LNMO. As seen, the morphology is different and the median – primary and secondary – particle size is larger for the low-surface-area LNMO. And this is most likely also why the Brunauer-Emmett-Teller (BET) surface area is considerably smaller (~1 m2/g vs. 9 m2/g). The Coulombic efficiency of the lowsurface-area LNMO/Li cells is around 89%, which is, in fact, lower than that of the standard LNMO/Li cells (94%). We believe that the irreversible capacity loss is largely caused by the decomposition of impurity phases. These are apparent in the powder X-ray diffraction (XRD) pattern shown in Fig. S3. However, Fig. 2c establishes that the balancing is significantly improved. The irreversibility or, in other words, the intrinsic degradation of LMNO is "sufficient" to compensate for the lithium loss upon SEI formation on graphite. The discharge potentials of LNMO and graphite remain constant at around 4.0 V and 0.7 V, respectively, thereby preventing the potential-driven SEI destruction (during the experiment time of about 225 h, corresponding to 60 cycles). DEMS (see Fig. S4) also indicates no H2 evolution at the end of the discharge cycles. As expected, the pressure increase during the first charge is only a little lower than in the standard LNMO/graphite system (17 mbar vs. 19 mbar). Nevertheless, it is reduced by a factor of more than two in the subsequent discharge cycle (4 mbar vs. 9 mbar), and this seems to be a direct result of the difference in the specific surface area of both LNMO materials. The pressure increases by about 34 mbar after four cycles (66 mbar after sixty cycles), which is 11 ACS Paragon Plus Environment
The Journal of Physical Chemistry
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 12 of 17
much less than in the standard LNMO/graphite cells. Overall, the results can be explained by the lower BET surface area and the fact that severe SEI damage does not occur upon cell discharge. However, the issue of metal dissolution from LNMO is also present in this system. Post mortem inductively coupled plasma optical emission spectroscopy (ICP-OES) analysis revealed that 0.192 wt.% Mn and 0.025 wt.% Ni (based on LNMO) are deposited on the graphite anode after 445 cycles. In contrast, for the standard LNMO/graphite cells, the amounts of Mn and Ni species found are only 0.144 wt.% and 0.022 wt.%, respectively. A possible explanation for this might be that the H+ concentration in the electrolyte is higher because less H2 is generated during cycling (note that the dissolution process is catalyzed by HF in the electrolyte).7,29 Irrespective of metal dissolution, the cycling stability of the lowsurface-area LNMO/graphite cells is better than that of the standard LNMO-based cell systems, which suggests that the cross-talk phenomenon is much more severe than Mn and Ni poisoning of the graphite anode. Fig. 5a shows that 73.1%, 80.4% and 82.7% of the initial discharge capacities are retained in the standard LNMO/graphite,
standard
LNMO/precycled
graphite
and
low-surface-area
LNMO/graphite cells, respectively, after 60 cycles. These data were obtained on three-electrode cells. The capacity retention of conventional coin-type cells (with no lithium reference electrode) using 13 mm diameter electrodes is shown in Fig. 5b. From this, it can be seen that the low-surface-area LNMO/graphite cells exhibit significantly improved stability, with capacity retention of 82% after 445 cycles (compare to 66% for the standard LNMO/graphite cells), corresponding to a capacity fade per cycle of only 0.04%. However, the specific capacity delivered by the lowsurface-area LNMO is lower (134.3 mAh/g vs. 144.6 mAh/g in the first charge cycle), likely because the material is not phase pure. Collectively, these results demonstrate that the potential-driven SEI destruction in LNMO/graphite full-cells has a profound effect on the cycling performance, and this issue can be avoided to some extent via the use of tailored anode and cathode materials.
12 ACS Paragon Plus Environment
Page 13 of 17
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 Journal of Physical Chemistry
Figure 5. (a) Specific charge/discharge capacities from three-electrode cells for different systems: standard LNMO/graphite, standard LNMO/precycled graphite and low-surface-area LNMO/graphite. After the first four cycles, the voltage profiles of which are shown in Fig. 2, were completed, the C-rate was increased to C/2 for the subsequent cycles. (b) Long-term cycling performance of standard LNMO/graphite and low-surface-area LNMO/graphite cells (with no lithium reference electrode). The rate was varied from C/10 to 2C discharge. Except for C/10, where the charge and discharge currents were equal, the rate on charge was C/4. Q represents the specific capacity and n is the cycle number.
Conclusions In summary, we have shown that it is vital to use cathode materials that are capable of compensating for the inevitable loss of active lithium upon SEI formation on graphite to achieve long-term stability for LIB full-cells. Otherwise, especially highvoltage systems like LNMO/graphite will suffer at some point from potential-driven SEI destruction. Its repair or reformation is not only accompanied by gas evolution, but it also leads (again) to lithium loss, and thus to a positive feedback loop. The results from our investigations using a lithium reference electrode demonstrate that the discharge potential of both the LNMO-based positive electrode and the graphitebased negative electrode shifts to more positive values with cycling. Abrupt gas release (mainly H2) is observed by DEMS and pressure measurements as the anode potential on discharge increases above approximately 0.9 V. This effect is more significant for the cell life time than the Mn and Ni dissolution problem. However, the capacity retention of LNMO/graphite cells can be improved by either using (i) a 13 ACS Paragon Plus Environment
The Journal of Physical Chemistry
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
graphite anode with a preformed SEI, (ii) an LNMO material having a higher irreversible capacity loss than graphite in the initial cycle, or (iii) a sacrificial lithium source additive. Overall, this means that proper balancing – taking into consideration the processes occurring on anode and cathode (especially during the formation cycle) and potential electrode cross-talk – is of utmost significance to ensure stable and safe operation of high-energy-density LIBs.
Author Information Corresponding Authors * E-mail:
[email protected] * E-mail:
[email protected] Notes The authors declare no competing financial interest.
Acknowledgements This study is part of the projects being funded within the BASF International Network for Batteries and Electrochemistry. We thank Dr. Thomas Bergfeldt, Dr. Zhirong Zhao-Karger, Dr. Rihab Al-Salman and Dr. Stefan J. Sedlmaier for assistance with ICP-OES, BET, SEM and XRD, respectively.
Associated Content The Supporting Information is available free of charge on the ACS Publications website. Cyclic voltammetric curves, galvanostatic charge/discharge profiles and DEMS data for standard LNMO/graphite, standard LNMO/precycled graphite and low-surfacearea LNMO/graphite cells; and XRD patterns of standard LNMO and low-surfacearea LNMO References (1) Markovsky, B.; Talyossef, Y.; Salitra, G.; Aurbach, D.; Kim, H. J.; Choi, S. Cycling and Storage Performance at Elevated Temperatures of LiNi0.5Mn1.5O4 Positive
14 ACS Paragon Plus Environment
Page 14 of 17
Page 15 of 17
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 Journal of Physical Chemistry
Electrodes for Advanced 5 V Li-Ion Batteries. Electrochem. Commun. 2004, 6, 821826. (2) Zhong, Q. M.; Bonakdarpour, A.; Zhang, M. J.; Gao, Y.; Dahn, J. R. Synthesis and Electrochemistry of LiNixMn2-xO4. J. Electrochem. Soc. 1997, 144, 205-213. (3) Kim, J.-H.; Pieczonka, N. P. W.; Li, Z.; Wu, Y.; Harris, S.; Powell, B. R. Understanding the Capacity Fading Mechanism in LiNi0.5Mn1.5O4/Graphite Li-Ion Batteries. Electrochim. Acta 2013, 90, 556-562. (4) Manthiram, A.; Chemelewski, K.; Lee, E.-S. A Perspective on the High-Voltage LiMn1.5Ni0.5O4 Spinel Cathode for Lithium-Ion Batteries. Energy Environ. Sci. 2014, 7, 1339-1350. (5) Xu, J.; Deshpande, R. D.; Pan, J.; Cheng, Y.-T.; Battaglia, V. S. Electrode Side Reactions, Capacity Loss and Mechanical Degradation in Lithium-Ion Batteries. J. Electrochem. Soc. 2015, 162, A2026-A2035. (6) Levi, M. D.; Dargel, V.; Shilina, Y.; Borgel, V.; Aurbach, D.; Halalay, I. C. Tailoring the Potential Window of Negative Electrodes: a Diagnostic Method for Understanding Parasitic Oxidation Reactions in Cells with 5 V LiNi0.5Mn1.5O4 Positive Electrodes. J. Power Sources 2015, 278, 599-607. (7) Pieczonka, N. P. W.; Liu, Z.; Lu, P.; Olson, K. L.; Moote, J.; Powell, B. R.; Kim, J.H. Understanding Transition-Metal Dissolution Behavior in LiNi0.5Mn1.5O4 HighVoltage Spinel for Lithium Ion Batteries. J. Phys. Chem. C 2013, 117, 15947-15957. (8) Peled, E.; Golodnitsky, D. SEI on Lithium, Graphite, Disordered Carbons and TinBased Alloys. In Lithium-Ion Batteries: Solid-Electrolyte Interphase; Balbuena, P. B.; Wang, Y., Eds. Imperial College Press: London, 2004. (9) Vetter, J.; Novak, P.; Wagner, M. R.; Veit, C.; Moller, K. C.; Besenhard, J. O.; Winter, M.; Wohlfahrt-Mehrens, M.; Vogler, C.; Hammouche, A. Ageing Mechanisms in Lithium-Ion Batteries. J. Power Sources 2005, 147, 269-281. (10) Winter, M.; Besenhard, J. O.; Spahr, M. E.; Novak, P. Insertion Electrode Materials for Rechargeable Lithium Batteries. Adv. Mater. 1998, 10, 725-763. (11) Jarvis, C. R.; Lain, M. J.; Yakovleva, M. V.; Gao, Y. A Prelithiated Carbon Anode for Lithium-Ion Battery Applications. J. Power Sources 2006, 162, 800-802. (12) Hu, L.; Amine, K.; Zhang, Z. Fluorinated Electrolytes for 5-V Li-Ion Chemistry: Dramatic Enhancement of LiNi0.5Mn1.5O4/Graphite Cell Performance by a Lithium Reservoir. Electrochem. Commun. 2014, 44, 34-37. 15 ACS Paragon Plus Environment
The Journal of Physical Chemistry
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
(13) Hoeweling, A.; Glatthaar, S.; Noetzel, D.; Binder, J. R. Evidence of Loss of Active Lithium in Titanium-Doped LiNi0.5Mn1.5O4/Graphite Cells. J. Power Sources 2015, 274, 1267-1275. (14) Imhof, R.; Novak, P. In Situ Investigation of the Electrochemical Reduction of Carbonate Electrolyte Solutions at Graphite Electrodes. J. Electrochem. Soc. 1998, 145, 1081-1087. (15) Onuki, M.; Kinoshita, S.; Sakata, Y.; Yanagidate, M.; Otake, Y.; Ue, M.; Deguchi, M. Identification of the Source of Evolved Gas in Li-Ion Batteries using (13)C-Labeled Solvents. J. Electrochem. Soc. 2008, 155, A794-A797. (16) Wang, H.; Rus, E.; Sakuraba, T.; Kikuchi, J.; Kiya, Y.; Abruna, H. D. CO2 and O2 Evolution at High Voltage Cathode Materials of Li-Ion Batteries: A Differential Electrochemical Mass Spectrometry Study. Anal. Chem. 2014, 86, 6197-6201. (17) Xu, M.; Tsiouvaras, N.; Garsuch, A.; Gasteiger, H. A.; Lucht, B. L. Generation of Cathode Passivation Films via Oxidation of Lithium Bis(oxalato) Borate on High Voltage Spinel (LiNi0.5Mn1.5O4). J. Phys. Chem. C 2014, 118, 7363-7368. (18) Michalak, B.; Sommer, H.; Mannes, D.; Kaestner, A.; Brezesinski, T.; Janek, J. Gas Evolution in Operating Lithium-Ion Batteries Studied In Situ by Neutron Imaging. Sci. Rep. 2015, 5, 15627. (19) Michalak, B.; Berkes, B. B.; Sommer, H.; Bergfeldt, T.; Brezesinski, T.; Janek, J. Gas Evolution in LiNi0.5Mn1.5O4/Graphite Cells Studied In Operando by a Combination of Differential Electrochemical Mass Spectrometry, Neutron Imaging, and Pressure Measurements. Anal. Chem. 2016, 88, 2877-2883. (20) Bernhard, R.; Metzger, M.; Gasteiger, H. A. Gas Evolution at Graphite Anodes Depending on Electrolyte Water Content and SEI Quality Studied by On-Line Electrochemical Mass Spectrometry. J. Electrochem. Soc. 2015, 162, A1984-A1989. (21) Blyr, A.; Sigala, C.; Amatucci, G.; Guyomard, D.; Chabre, Y.; Tarascon, J. M. Self-Discharge of LiMn2O4/C Li-Ion Cells in Their Discharged State - Understanding by means of Three-Electrode Measurements. J. Electrochem. Soc. 1998, 145, 194209. (22) Kleiner, K.; Jakes, P.; Scharner, S.; Liebau, V.; Ehrenberg, H. Changes of the Balancing between Anode and Cathode due to Fatigue in Commercial Lithium-Ion Cells. J. Power Sources 2016, 317, 25-34.
16 ACS Paragon Plus Environment
Page 16 of 17
Page 17 of 17
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 Journal of Physical Chemistry
(23) Xie, Y.; Yuan, C. An Integrated Anode Stress Model for Commercial LixC6LiyMn2O4 Battery during the Cycling Operation. J. Power Sources 2015, 274, 101113. (24) Huang, W.; Frech, R. In Situ Raman Studies of Graphite Surface Structures during Lithium Electrochemical Intercalation. J. Electrochem. Soc. 1998, 145, 765770. (25) Berkes, B. B.; Jozwiuk, A.; Vracar, M.; Sommer, H.; Brezesinski, T.; Janek, J. Online Continuous Flow Differential Electrochemical Mass Spectrometry with a Realistic Battery Setup for High-Precision, Long-Term Cycling Tests. Anal. Chem. 2015, 87, 5878-5883. (26) Terada, Y.; Yasaka, K.; Nishikawa, F.; Konishi, T.; Yoshio, M.; Nakai, I. In Situ XAFS Analysis of Li(Mn, M)2O4 (M = Cr, Co, Ni) 5 V Cathode Materials for LithiumIon Secondary Batteries. J. Solid State Chem. 2001, 156, 286-291. (27) Berkes, B. B.; Schiele; A.; Sommer, H.; Brezesinski, T.; Janek, J. On the Gassing Behavior of Lithium-Ion Batteries with NCM523 Cathodes. J. Solid State Electrochem. 2016, 20, 2961-2967. (28) Metzger, M.; Strehle, B.; Solchenbach, S.; Gasteiger, H. A. Origin of H2 Evolution in LIBs: H2O Reduction vs. Electrolyte Oxidation. J. Electrochem. Soc. 2016, 163, A798-A809. (29) Yamane, H.; Inoue, T.; Fujita, M.; Sano, M. A Causal Study of the Capacity Fading of Li1.01Mn1.99O4 Cathode at 80°C, and the Suppressing Substances of Its Fading. 2001, 99, 60-65.
TOC Graphics
17 ACS Paragon Plus Environment