Critically Examining the Role of Nanocatalysts in Li–O2 Batteries

Feb 8, 2018 - In lithium–oxygen (Li–O2) batteries, nanocatalysts have been widely employed as a means to suppress the large recharge overpotential...
0 downloads 14 Views 867KB Size
Subscriber access provided by Gothenburg University Library

Letter

Critically Examining the Role of Nanocatalysts in Li-O Batteries: Viability towards Suppression of Recharge Overpotential, Rechargeability and Cyclability 2

Raymond Albert Wong, Chunzhen Yang, Arghya Dutta, Minho O, Misun Hong, Morgan L. Thomas, Keisuke Yamanaka, Toshiaki Ohta, Keiko Waki, and Hye Ryung Byon ACS Energy Lett., Just Accepted Manuscript • DOI: 10.1021/acsenergylett.8b00054 • Publication Date (Web): 08 Feb 2018 Downloaded from http://pubs.acs.org on February 8, 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 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 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.

ACS Energy Letters 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 21 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

ACS Energy Letters

Critically Examining the Role of Nanocatalysts in Li–O2 Batteries: Viability towards Suppression of Recharge Overpotential, Rechargeability and Cyclability Raymond A. Wong†,‡,§,∇, Chunzhen Yang‡,∇, Arghya Dutta†,‡, Minho O‡, Misun Hong†,‡, Morgan L. Thomas‡,⊥, Keisuke Yamanaka∥, Toshiaki Ohta∥, Keiko Waki§,, and Hye Ryung Byon†,‡,#,* †

Department of Chemistry, Korea Advanced Institute of Science and Technology (KAIST), 291

Daehak-ro, Yuseong-gu, Daejeon 34141, Republic of Korea ‡

Byon Initiative Research Unit, RIKEN, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan

§

Department of Energy Sciences, Tokyo Institute of Technology, 4259 Nagatsuta-cho, Midori-

ku, Yokohama 226-8502, Japan ∥

Synchrotron Radiation (SR) Center, Ritsumeikan University, Kusatsu, Shiga 525-8577, Japan

#

Advanced Battery Center, KAIST Institute NanoCentury, 291 Daehak-ro, Yuseong-gu,

Daejeon 34141, Republic of Korea *Corresponding Author

[email protected]

1 ACS Paragon Plus Environment

ACS Energy Letters 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 21

ABSTRACT: In lithium-oxygen (Li-O2) batteries, nanocatalysts have been widely employed as a means to suppress the large recharge overpotential and to possibly improve cyclability. However, these studies have consistently been mired with ambiguity relating to the possible exacerbation of side reactions, which in turn has questioned the role of such catalysts in Li-O2 cells. Here, to shed light on the behavior of nanocatalysts in Li–O2 batteries, we examine Ru, Pt, Pd, Co3O4 and Au nanoparticles supported on carbon nanotubes, which have been widely employed as promising catalysts. We show that while there can be noticeable reduction in overpotential with catalysts, the facile decomposition of Li2O2 is not accompanied by a decrease in side reactions and as a consequence, there is no notable improvement in rechargeability nor cyclability. Instead, highly active catalysts can exhibit non-selectivity for all oxidation reactions including Li2O2 and the electrolyte. This work underscores the importance of metrics beyond simple consideration of the recharge overpotential and the necessity of pursuing approaches that can promote reversible Li-O2 electrochemistry.

TOC GRAPHICS

2 ACS Paragon Plus Environment

Page 3 of 21 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

ACS Energy Letters

Widely touted as one of the most promising next generation energy storage candidates, lithiumoxygen (Li-O2) batteries have immensely high theoretical specific energy (~3.5 kWh kg-1), which can potentially provide a cost effective means of storing energy.1,2 The ideal aprotic Li-O2 electrochemistry entails the overall reaction 2Li+ + O2 + 2e- ⇆ Li2O2 (Eo = 2.96 V vs Li/Li+), with the reversible formation and decomposition of solid Li2O2 during discharge (DC) and recharge (RC), respectively.3,4 Development has however been hampered by critical challenges related to the poor cyclability and large RC overpotentials (>1 V).3,5-8 The origin of the large RC overpotential has been ascribed to (1) sluggish kinetics of charge transport in the wide-bandgap Li2O2,9,10 and (2) formation of insulating side products of lithium carboxylates and carbonates, due to electrolyte and carbon electrode reactivity with reduced oxygen species (O2-, LiO2, Li2O2).11-13 These unintended side reactions compromise rechargeability as defined by the O2 evolved/consumed ratio, which ideally would be at unity.3 In addition, the incomplete removal and accumulation of side products induces poor cyclability.6 In response, numerous reports of metal and metal oxide nanocatalysts have indeed shown what appears to be a dramatic reduction in RC overpotential for Li2O2 electro-oxidation, which is also often coupled with observations of enhanced cyclability.6,14-18 On the other hand, evidence of low kinetic potentials for Li2O2 oxidation,4,19 ambiguity regarding catalytic mechanisms for the oxidation reaction9,14,15 and exacerbation of side reactions3,10 has further questioned the role of catalysts in Li-O2 cells.20 Specifically, the often-proposed merits of reduced RC overpotentials with nanocatalysts may be largely superficial due to the possibility for enhanced parasitic side reactions.3 For instance, there is a common assertion that if the RC potential is suppressed below 4 V, rechargeability can be improved with side reactions lessened, or even prevented.16,21 We will show that this is not the

3 ACS Paragon Plus Environment

ACS Energy Letters 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 4 of 21

case, even with Ru, which has been extensively reported to effectively suppress the RC overpotential.16,18,21 Another shortcoming is the difficulty in comparing various catalysts reported in the literature in terms of the relative efficacy towards Li2O2 oxidation and cyclability. This arises due to differences in experimental conditions, namely, catalyst loading, type of carbon, binder instability, and test conditions.22,23 As a result, the aforementioned reasons have caused us to investigate nanocatalysts more closely. Herein, we examine the viability of nanocatalysts regarding (1) their effectiveness in the degree of RC overpotential reduction, and (2) the predominant processes occurring in catalyst-containing Li-O2 cells. To accomplish this Ru, Pt, Pd, Co3O4 and Au nanoparticles (NPs) were synthesized and supported on multi-walled carbon nanotubes (CNTs), which were subsequently fabricated into binder-free electrodes. By utilizing in situ on-line electrochemical mass spectrometry (OEMS) and ex situ X-ray, FTIR and 1H NMR spectroscopies, we demonstrate that while RC potential can be suppressed with nanocatalysts, there is no correlation between RC potential and rechargeability nor cycling stability, due to the additional effects of catalysts for side reactions. We have employed a catalyst loading of ~40 wt.% with respect to the CNT mass of 1.2 mg, which were used as-received without chemical oxidation or functionalization (Figure S1, supporting information (SI) for Experimental Section). Transmission electron microscopy (TEM, Figure 1 and Figure S1) and X-ray diffraction (XRD, Figure S1) indicate that the NPs are ~10 nm or less in size, with Ru/CNT at ~1.5 nm, Pt/CNT at 4–6 nm, Pd/CNT at 5-8 nm, and Co3O4/CNT and Au/CNT at 8~11 nm. We note that subsequent control studies on the influence of NP size with Au, Pd and Ru NPs (see supplementary note in SI) indicate that at ~10 nm and lower, reasonable comparisons can be made.

4 ACS Paragon Plus Environment

Page 5 of 21 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

ACS Energy Letters

Galvanostatic cell tests with 0.5 M LiTFSI/tetraglyme were performed with DC-RC at a fixed capacity of 1.2 mAh (Figure 2a), which provides an equal footing of comparison for the subsequent analyses. The DC potential plateaus do not significantly vary at ~2.73 V, indicating the relative insensitivity of Li2O2 formation in the presence of these nanocatalysts.24,25 Focusing on the RC profiles, we observe diverging potentials even at the initial stages of RC. As summarized in Figure 2b, Ru/CNT consistently outperforms in terms of catalytic activity, with the lowest potential from the initial stage of RC and approaches ~3.5 V at a state of charge (SOC) of 75%, leading to a round-trip efficiency of 80%. In contrast, CNT exhibits the lowest round-trip efficiency at 69% followed by Au/CNT at 70%, whereas Pt/CNT, Pd/CNT and Co3O4/CNT are comparable at ~73%. Comparable results with full DC to the cutoff potential of 2.2 V were also acquired (Figures S2-S3). The corresponding scanning electron microscopy (SEM) images show film-like Li2O2 completely covering the catalyst/CNT electrodes following DC and its subsequent elimination after RC (Figure S4). The resulting RC overpotential for the first galvanostatic cycle allows us to assess the order of catalytic activity as Ru > Pt, Pd, Co3O4 > Au and CNT (i.e., absence of catalyst). As a result, we can infer the trend of cycling performance because if lower RC potential corresponds to greater rechargeability, this should equate to enhanced cycling stability. Remarkably, however, the cycling results in Figure 2c-e and Figure S5 show unexpectedly higher cycling stability of CNT and Au/CNT, which maintain their capacity for 36 cycles (>1400 h), despite consistently exhibiting the highest RC potential (Figure 2e). In contrast, Ru/CNT and Pt/CNT show a pronounced suppression of RC potential throughout cycling but significantly earlier capacity fade is observed at 18 and 21 cycles, respectively. The lack of correlation between RC overpotential and cyclability

5 ACS Paragon Plus Environment

ACS Energy Letters 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 6 of 21

indicates that the reduced RC potential is not directly related with rechargeability, thus further analysis is necessary to determine the role of nanocatalysts. The subsequent step was therefore to evaluate the O2 evolved/consumption ratio. In situ pressure monitoring of the first DC reveals comparable overall values of 2.0-2.1 e–/O2 (electrons per oxygen molecule) for all electrodes (Table S2 and Figure S6), which are indicative of the ideal 2e– transfer for Li2O2 formation. The RC process as characterized with OEMS shows predominantly O2 evolution along with CO2 and H2 evolution (Figure 3). As O2 evolution is exclusive to Li2O2 oxidation8,26,27 and the O2 evolution trends and amounts are comparable for all electrodes according to SOC (Figure S7), the suppressed RC overpotential corresponds to the improved catalytic or decomposition efficiency of Li2O2. Upon further inspection of the O2 evolution profiles, two common characteristics for all electrodes include: (1) O2 evolution essentially terminates by ~100% SOC and (2) the integral O2 evolution equates to an average of 3.3-3.6 e-/O2 (0 to 100% SOC), corresponding to O2 evolved/consumed ratios of 57-64% (Figure 4a and Table S2), which indicates the remarkably low rechargeability for all electrodes. The origin of the significant amount of missing O2 (~40%) is addressed with ex situ chemical analyses. The X-ray absorption near edge structure (XANES, Figure S8) and Fourier transform infrared (FTIR, Figure S9) spectra confirm that Li2O2 is the predominant DC product for all electrodes. Further, UV-Vis titration8,28 was performed to quantitatively estimate the Li2O2 yield following DC. With the catalyst-free CNT electrode, Li2O2 yields of ~75% were typically obtained, which is higher than the O2 yield following RC. This indicates the occurrence of approximately ~25% loss of Li2O2 from DC, and a further ~15% missing from RC. Although the measurement of Li2O2 yield with the catalyst containing electrodes was unsuccessful due to the catalyzed decomposition of H2O2 (from the observation of gas bubbles) in aqueous solutions,29,30 we expect an analogous Li2O2 yield

6 ACS Paragon Plus Environment

Page 7 of 21 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

ACS Energy Letters

because of the similar DC potential and comparable O2 yield from RC. The loss of Li2O2 during DC may be attributed to the dissolution of superoxide and LiO2 intermediates forming Li2O2 away from the electrode.31 In addition, Li2O2 can be involved in side reactions.11,30 The surface-sensitive partial electron yield (PEY) mode of the XANES spectra (Figure S8) displays a lithium carboxylates-associated signal at ~532.8 eV after DC, which overlaps with the peak assigned to lithium hydroxide (LiOH). However, the FTIR spectra (Figure S9) also contains pronounced lithium carboxylate-related bands but the absence of the stretching vibration of O-H (~3675 cm– 1

), indicating negligible LiOH. Subsequent 1H NMR provides insight into the relative distribution

of lithium carboxylates (Figure S10). Lithium formate is present in all electrodes due to chemical reactions with Li2O2 and tetraglyme.32 The increased presence of lithium acetate is notable with the noble metal catalysts. Therefore, the observation of these induced chemical/electrochemical parasitic reactions with Li2O2 likely contribute to the reduction of Li2O2 yield.32 More importantly, the additional loss of ~15% Li2O2 from RC does not depend on the presence of catalyst or RC potential as evidenced by the fluctuating O2 evolution rate (Figure 3) for all electrodes throughout Li2O2 decomposition. The O K-edge XANES spectra following 100% SOC shows the π* (C=O) of Li2CO3 arising from the shoulder at 532.8 eV along with lithium carboxylate-related signals, while Li2O2 is no longer observed (Figure S8). 1H NMR shows that the portion of lithium acetate has reduced, but overall, moderate variation of lithium carboxylates is observed in comparison with DC result (Figure S10). In comparison, Li2CO3 is known to be a prominent side product for RC.30,33 We previously reported the appearance of Li2CO3 features with CNT even at 25% SOC,8 and this prevalent side reaction results from the transformation occurring at the Li2O2 interfaces due to the degradation of carbon electrode and electrolyte solution.30,33 The formation of Li2CO3 may be lessened at very low potentials as evidenced by Ru/CNT, which

7 ACS Paragon Plus Environment

ACS Energy Letters 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 21

maintains