Oxygen redox reaction in Ionic Liquid and Ionic Liquid-like based

Subscriber access provided by BUFFALO STATE

Energy Conversion and Storage; Plasmonics and Optoelectronics

Oxygen redox reaction in Ionic Liquid and Ionic Liquid-like based electrolytes: a scanning electrochemical microscopy study Irene Ruggeri, Catia Arbizzani, Stefania Rapino, and Francesca Soavi J. Phys. Chem. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.jpclett.9b00774 • Publication Date (Web): 29 May 2019 Downloaded from http://pubs.acs.org on May 29, 2019

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.

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 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 Letters

Oxygen Redox Reaction in Ionic Liquid and Ionic Liquid-like

Based

Electrolytes:

A

Scanning

Electrochemical Microscopy Study Irene Ruggeri, Catia Arbizzani, Stefania Rapino*, Francesca Soavi* Department of Chemistry Giacomo Ciamician – Alma Mater Studiorum Bologna University

*[email protected]; *[email protected]

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry 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 27

Abstract Improving the stability of the cathode interface is one of the critical issues for the development of high-performance Li/O2 batteries. The most critical feature to address is the development of electrolytes that mitigate side reactions that bring about cathode passivation. It is well known that superoxide anion (O2- ) formed during O2 reduction plays a critical role. Here, we propose the scanning electrochemical microscopy (SECM) as analytical tool to screen the electrolyte of Li/O2 batteries. We demonstrate that by SECM it is possible to evaluate the stability of O2- and of the cathode to the passivation process occurring during the oxygen redox reaction. Specifically, we report about a study carried out at glassy carbon electrode in 1-butyl-1-methylpyrrolidinium bis(trifluoromethanesulfonyl)imide (PYR14TFSI) and lithium bis(trifluoromethanesulfonyl)imide (LiTFSI), and tetraethylene glycol dimethyl ether with LiTFSI, the latter ranging from salt insolvent to solvent-in-salt region.


2

Page 3 of 27 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


TOC graphics

Keywords ORR in Li/O2 battery, Scanning Electrochemical Microscopy (SECM), Solventin-salt electrolyte, in-operando investigation, Superoxide ion stability


3


Page 4 of 27

Li/O2 batteries, leveraging the use of the lightest available electrode materials, specifically Li metal and O2, achieve the high practical specific energy of 1000 Wh kg-1. The energy content is 4 times higher than that of the most optimized Li-ion cell commercially available today (250 Wh kg-1) (1). For this reason, much research efforts are being spent to bring the Li/O2 system into reality (2). The oxygen redox reaction (ORR) in aprotic Li/O2 batteries consists at first in the formation of lithium superoxide (LiO2). This product gives disproportion and generates both lithium peroxide (Li2O2) and O2, according to Equation 1: 2LiO2Li2O2+O2

(Eq. 1)

and/or proceeds by the second electrochemical reduction step, according to Equation 2: LiO2+Li++e-Li2O2

(Eq. 2)

The solid and insulating Li2O2 species clogs the electrode surface, limiting the discharge capacity and causing high recharge overpotential. Additionally, the superoxide radical ion (O2·-) might nucleophilically attack the organic solvent, the carbon electrode and anion. Furthermore, it has been reported that O2·- lifetime in the electrolyte is strongly related to the Lewis acidity of the electrolyte environment (3), (4), (5). Indeed, the superoxide is a soft Lewis base that is stabilized by soft Lewis acid cations like the bulky pyrrolidimium of ionic liquids. At the contrary, the peroxide is a hard Lewis base that readily couples with strong Lewis acids like the Li+ cation. Therefore, lithium salts promote O2·dismutation to Li2O2, and, hence, decrease the lifetime of the superoxide (Eq. 1). The Lewis acidity of Li+ can be softened by the use of high donor number (DN) solvents, and this, in turn, improves superoxide lifetime. The different stability of the superoxide anion affects the Li2O2 formation mechanisms and hence the cycling performance of Li/O2 battery, too. Indeed, when ORR is carried out in electrolytes where superoxide is unstable (e.g. in presence of unsolvated Li+), O2·- quickly dismutates ·to Li2O2 at the electrode surface. This is the case of the so-called surface growth mechanism that fosters the passivating coatings on the electrode surface accelerating the cell death. On the other hand, during ORR in electrolyte solutions that are able to stabilize the superoxide anion (e.g. when the cation is considered as a soft Lewis acid), the O2·- produced during ORR can diffuse far from the electrode. In these conditions, at the early stage of the process, Li2O2 forms in solution and then subsequently precipitates as big clusters on the cathode surface. This mechanism, which is defined as Li2O2


4

Page 5 of 27 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


solution formation mechanism, results in a delayed passivation of the electrode with a positive impact on the battery performance (6). It has also to be mentioned that the above-described processes are often paralleled by unwanted side reactions associated to the chemical reactivity of O2·- and peroxide with the electrolyte and carbon electrode. The interpretation of the cathode insulating products formed during Li/O2 battery discharge is challenging. While the mechanism of the oxidation of solid Li2O2 on conductive carbon electrode is still under investigation, it has been found that Li2O2 nature and morphology affect the overpotential of the recharge step in a Li/O2 batteries (7), (8). Glymes, dimethyl sulfoxide (DMSO) and ionic liquid (IL)-based electrolytes have been investigated for a use in Li/O2 batteries (6), (9). As aforementioned, organic solvents with high DN should be preferred because they decrease the Li+ Lewis acidity and, therefore, promote Li2O2 formation in solution. Solvated Ionic Liquid (SIL) solutions, where the salt to solvent molar ratio is higher than 1, are also gaining much attention for the peculiar IL-like structures and chemical physical properties. Indeed, in SIL solutions of lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) and glymes, the glyme molecules coordinate the Li+ ions to give large complex cations that are weakly coupled with the bulky TFSI anion, with no (or negligible) free solvent molecules (10). By an electrochemical mass spectroscopy study, Kwon et al. demonstrated that the LiTFSI-triglyme based SIL has a larger anodic stability compared to conventional low concentrated solutions (typically 1 mol L-1) and that its use as electrolyte in Li/O2 preserves side reactions (11). The decreased amount of free molecules of solvent in SIL suppresses also the incessant parasitic reaction between the electrolyte and the anode, with a positive effect on the lithium solid electrolyte interphase (SEI) stability (12), (13). The ORR in tetraethylene glycol dimethyl ether (TEGDME)-LiTFSI SILs has been investigated by Messaggi et al. who reported that solvent-in- salt (SIS) (or SIL) electrolytes favor the Li2O2 solution formation mechanism and brings about weakly adsorbed Li2O2 on the electrode surface (14). The reason is that in SILs based on TEGDME and LiTFSI, the Li+ Lewis acidity is decreased by the TEGDME coordination. Consequently, SIL tends to stabilize O2•- in solution as for the case of Li+-free IL, without quickly


5


Page 6 of 27

proceeding thought the superoxide disproportion (Eq. 1) that generates the passivating peroxide. Indeed, in aprotic ILs like 1-butyl-1-methylpyrrolidinium bis(trifluoromethanesulfonyl)imide (PYR14TFSI), ORR mainly proceeds through the mono-electronic and quasi-reversible O2/O2·redox process thanks to the soft Lewis acidity of the PYR14+ cation, as referred before with glassy carbon (GC) or gold working electrode (14), (15), (16), (17). In 2017, the ORR mechanism was also studied in conventional (0.03 mol kg-1) and high concentrated (0.5 mol kg-1) NaTFSIPYR14TFSI electrolytes for Na/O2 batteries by Pozo-Gonzolo et al. The authors investigated by a computational study the arrangement of the different species based on the complex O2·--PYR14+xNa+y-TFSIz that differ in the NaTFSI salt concentration. Furthermore, by a rotating ring-disk electrode measurement and using a pressure cell (e-/O2 ratio during a Na/O2 battery discharge), they found that in both electrolytes the main product is the Na2O2 peroxide and in the high concentrated one, the Na2O2 is formed via NaO2 superoxide, which is however insoluble. This hence excludes an ORR with Na2O2 formation from solution via O2·- dissolution in the concentrated NaTFSI-PYR14TFSI electrolyte (18). It is thus clear that in order to deeply understand the mechanism and reaction that occur during the battery operation, combining analytical and spectroscopic techniques with the electrochemical ones is recommended (19), (20). S. Freunberger, B.D. McCloskey and coworkers greatly contributed to get insights into the reversibility of the Li/O2 processes mostly by differential electrochemical mass spectroscopy (DEMS), which is an analytical technique that combines electrochemical experiments with the mass spectrometry (3), (4), (6). Among all the different techniques, the scanning electrochemical microscopy (SECM) is a powerful tool to understand processes that occur at a solid substrate interface. The use of an additional electrode, usually being a Pt ultramicroelectrode (UME) placed close distance from the substrate, enables to detect species formed at the substrate and that might diffuse towards the electrolyte. The Pt UME, hence, acts as a probe of the products. SECM applications are definitely wide, ranging from biological analysis to electrocatalysis (for fuel cell and ORR investigation) (21), (22), (23). Indeed, the ability to probe also non-conductive surfaces makes SECM a feasible and non-invasive method for study the charge transfer and redox activity in a wide range of systems (24). Additionally, SECM can generate images of surface reactivity and chemical kinetics and can be coupled to atomic force microscopy too, for the investigation of various interfacial phenomena, providing local quantitative and electrochemical information of materials (25), (26). SECM has been also


6

Page 7 of 27 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


employed to study several electrochemical processes in Li-ion, semi-solid flow and Li/O2 battery, thus demonstrating to be a unique and powerful in-operando and in-situ analytical tool for electrode investigation (27), (28), (29). G. Wittstock et al. in 2015 have studied the O2 permeation through gas diffusion layer (GDL) of different thickness and its flux from the substrate to the lithium perchlorate (LiClO4)-DMSO electrolyte. An oxidation pulsed procedure was also interestingly proposed to clean up the Pt probe from Li2O2. The O2·- intermediate species was then detected by fluorescence microscopy and by local detection at defined distances from the GDL substrate electrode (30), (31). Here, we report for the first time about the use of SECM to investigate the effect of Li+ salt concentration on ORR and Li2O2 formation in conventional organic electrolyte, SIL and IL. Specifically, ORR is investigated in conventional TEGDME - 0.5 m LiTFSI and, in TEGDME – 5 m LiTFSI SIL electrolyte and in PYR14TFSI IL, with and without LiTFSI salt. We demonstrate that SECM is a powerful tool to get indication on the change of conductivity and passivation of Li/O2 battery cathode upon discharge operation (i.e. during the O2 reduction). Most importantly, our work paves the way to the use of SECM to the fundamental understanding of the different Li2O2 formation mechanisms in Li/O2 batteries. When SECM operates in Substrate Generation/Tip Collection (SG/TC) mode, the species generated at the substrate can be collected at the UME. The current detected at the tip relates to the presence and the amount of these species. In turn, tip response depends on the concentration, life time and diffusion rate of the species generated at the substrate. Here, SG/TC has been used to detect at the Pt UME, the O2•- produced upon O2 reduction at the GC substrate. The latter simulates a Li/O2 battery cathode. The use of SECM in


7


Page 8 of 27

SG/TC mode allows the study of the stability of the species generated (O2•-) at the functioning electrode in the specific electrolyte. The collection is operated at several close distances from the substrate and reports on the presence of the analysed species. If the species is detected, it means that it is able to diffuse in the electrolyte and it is stable enough to foster a “in solution” formation of Li2O2, instead of promoting a direct “grown on surface” passivation layer. In fact, according to Eq. 1 and Eq. 2, O2•- is the ORR intermediate and, in case of a long life-time (high stability in the electrolyte), it can to diffuse far from the GC in the electrolyte bulk and be detected at the Pt tip. Figure 1 reports the results of the SG/TC mode study, where CVs are applied to the GC substrate while the Pt UME probe is kept at a fixed potential (3.50 and 3.00 V vs. Li+/Li) for the detection of the O2•- generated, depending on the solution, at the substrate electrode. Figure 1a and Figure 1c show the voltammograms at 20 mV s-1 of the GC substrate and Figure 1b and Figure 1d report the simultaneous response of the Pt tip placed 5 μm above the GC (dPt/GC). Experiments were run in O2-saturated PYR14TFSI IL, PYR14TFSI - 0.1 M LiTFSI electrolyte and TEGDME - 0.5 m LiTFSI and TEGDME - 5 m LiTFSI electrolytes. The cathodic cut-off in the CVs was 1.70 V vs. Li+/Li and 1.30 V vs. Li+/Li in pure PYR14TFSI IL and PYR14TFSI - 0.1 M LiTFSI respectively, while it was 2.00 V vs. Li+/Li in the LiTFSI-TEGDME based electrolytes. The anodic cut-off was 4.00 V vs. Li+/Li in all the investigated electrolytes. The CVs of GC in the different electrolytes started from 4.00


8

Page 9 of 27 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


V vs. Li+/Li to the cathodic cut-off and then proceeded to the reverse scans (please follow the numbered dashed arrows). Figure 1a and Figure 1b compare the CVs of the GC and Pt probe responses in PYR14TFSI IL, with and without LiTFSI. In Li+ free O2-saturated PYR14TFSI, the GC CV (Figure 1a) shows two peaks attributed to the mono-electron quasi-reversible ORR that involves the O2/O2•- redox couple (red curve) (14). The cathodic peak at ca. 2.20 V vs. Li+/Li is due to O2 reduction to O2•superoxide and the anodic peak at ca. 2.50 V vs. Li+/Li is due to O2•- oxidation to O2. Hence, the superoxide starts to be generated at the GC below 2.50 V vs. Li+/Li (Figure 1a). More interestingly, according to Figure 1b, at the same time, an anodic current starts to increase at the Pt UME, reaching a first peak when the CG reaches ca. 2.00 vs. Li+/Li. This anodic current can be attributed to the oxidation at the Pt tip (kept at 3.50 V vs. Li+/Li) of the O2•– produced at the GC substrate. The absence of a clear, defined Pt current peak can be related to the formation of not-stoichiometric Li-O species at the Pt tip and/or impurities and water traces in the electrolyte (5). Indeed, the measurements were run in ambient atmosphere. Combining Figure 1a and Figure 1b confirms literature findings, i.e. that in Li+-free IL the O2•- formed at the GC is sufficiently stable to be re-oxidized during the anodic scan (Figure 1a). The data also indicate that O2•- lives enough in the IL to be collected at the Pt tip placed 5 μm far from the GC (Figure 1b). At the contrary, in PYR14TFSI – 0.1 M LiTFSI electrolyte, the O2•- is not stable and quickly gives chemical disproportion to O2 and Li2O2. This is clearly appreciable in the GC CV reported Figure 1a (blue curve). Indeed, when


9


Page 10 of 27

LiTFSI is added, the voltammogram has a drastic change. The quasi-reversible signature is lost and the cathodic peak potential is shifted to more positive values because the electrochemical step, i.e. the reduction of O2 to O2•-, is followed by a chemical reaction, i.e. the chemical disproportion of O2•- to Li2O2. In addition, the CV peak currents decrease. The anodic peak at 2.50 V vs. Li+/Li, related to O2•- oxidation, is no more detectable and is substituted by a small peak at ca 3.00 V vs. Li+/Li, likely due to peroxide oxidation to O2 (5), (14). At the same time, Figure 1b shows that in PYR14TFSI – 0.1 M LiTFSI the current detected at the Pt tip (at 3.50 V vs. Li+/Li) is very flat and does not reveal an appreciable signal due to the oxidation of O2•-. This means that no O2•- can be detected at least at a distance of 5 μm from the GC. According to our hypothesis, a fast dismutation of O2•- is occurring at the GC surface preventing the diffusion of such species in the electrolyte. The absence of the signals at the Pt tip can also be due to the passivation of GC and/or Pt by Li2O2 or other products formed after unwanted, side reactions involving the electrolyte.


10

Page 11 of 27 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. SG/TC test in O2-saturated solutions in Teflon-SECM cell: (a, c) CVs of GC at 20 mV/s and (b, d) Pt currents with EPt = 3.50/3.00 V vs. Li+/Li recorded during the GC CV scans with dPt/GC of 5 μm in (a, b) PYR14TFSI with (blue) and without LiTFSI (red) and (c, d) in TEGDME - 0.5 m LiTFSI (blue) and TEGDME - 5 m LiTFSI (red).

Figure 1c and Figure 1d show the SG/TC mode study in the O2-saturated TEGDME - 0.5 m LiTFSI and TEGDME - 5 m LiTFSI electrolytes. The Pt UME was kept at 3.00 V vs. Li+/Li during the CVs of the substrate for the O2•- collection. In Figure 1c, the GC CVs are similar to those obtained in PYR14TFSI – 0.1 M LiTFSI, with a cathodic wave due to the mono-electron reduction of O2 to O2•-, followed by the chemical disproportion of O2•- to


11


Page 12 of 27

Li2O2. The anodic peak is related to peroxide oxidation to O2. Notably, the cathodic peak current is higher in the TEGDME-5 m LiTFSI solution than in the TEGDME - 0.5 m LiTFSI one. This can be explained considering that when LiTFSI concentration is increased, i.e. in TEGDME - 5 m LiTFSI, the Li+ ion complexation by glyme molecules starts to be effective and Li+ Lewis acidity is softened by glyme coordination. The [Li(glyme)1]+ complexes make the superoxide more stable and the chemical dismutation to Li2O2 less pronounced, and promote Li2O2 formation in solution rather than on the GC electrode surface. Consequently, GC electrode passivation is mitigated and cathodic currents in TEGDME - 5 m LiTFSI are higher than in TEGDME - 0.5 m LiTFSI electrolyte (14). Figure 1d reports the Pt UME currents recorded at 3.00 V vs. Li+/Li while performing CVs at the substrate. In TEGDME - 5 m LiTFSI, an anodic current is recorded when the substrate is scanned below 2.70 V vs. Li+/Li, which we attribute to the oxidation of O2•formed at GC (red curve). On the contrary, in TEGDME - 0.5 m LiTFSI no signal is detected, therefore supporting the idea of a very fast disproportion of O2•- to Li2O2 and fast GC passivation (blue curve). Figure 1c and Figure 1d, suggest that the increase of the LiTFSI concentration up to 5 m stabilizes the O2•- anion and mitigates GC passivation. Notably, the comparison of the data reported in Figure 1b and Figure 1d shows that, while in PYR14TFSI – 0.1 M LiTFSI the Pt probe does not detect any species at 5 μm from the GC substrate, it instead reveals the O2•- presence in the TEGDME - 5 m LiTFSI. This supports the idea that O2•- is more stable in TEGDME - 5 m LiTFSI than in the PYR14TFSI – 0.1 M LiTFSI. This can be translated in the conclusion that the GC substrate in TEGDME - 5m LiTFSI is freer from passivation of insulating species than in the IL-based solution.


12

Page 13 of 27 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


In order to better study the O2•- stability and its diffusion through the TEGDME - 5 m LiTFSI electrolyte, the distances from the GC substrate of the Pt tip were changed up to 100 μm and the results are reported in Figure 2. The CVs were collected from 3.00 V vs. Li+/Li, having 2.00 and 4.00 V vs. Li+/Li as cathodic and anodic cut-off, respectively. The GC and the Pt were electrochemically cleaned between the different steps by a quick CV between 2.00 V and 4.00 V vs. Li+/Li. The figure shows that O2•- species is detectable up 25 μm. It means that the superoxide anion life-time is long enough to enable diffusion from the GC substrate to the bulk electrolyte up to such a distance.

Figure 2. SG/TC mode of Pt UME kept at 3.00 V vs. Li+/Li at different UME/substrate distances (dPt/GC = 5 μm, 25 μm and 100 μm) while the GC substrate is scanned at 20 mV s-1 in O2-saturated TEGDME - 5 m LiTFSI electrolyte.

To get further insight into the GC passivation in the glyme-based conventional salt-insolvent TEGDME - 0.5 m LiTFSI solution and in the SIL TEGDME - 5 m LiTFSI solution,


13


Page 14 of 27

SECM in feedback mode has been carried out, using the dissolved O2 as redox mediator (Figure 3). In feedback mode, the current at the tip is detected while it moves closer to the surface (approach curve). If the substrate has a conductive surface the tip current enhances, vice versa decreases in case of passivation. It is thus possible to determine if a GC surface is electrically conductive or becomes insulating during some electrochemical tests. A potential step (PS) at 2.00 V vs. Li+/Li for 1h was applied at the GC substrate allowing the O2 to reduce. Figure 3a and Figure 3b show the feedback curves of the Pt tip before (solid lines) and after (dashed lines) the PSs in TEGDME - 0.5 m LiTFSI (blue) and TEGDME - 5 m LiTFSI, (red) respectively. In the y axis the IT/IT,∞ values are reported, where IT are the currents recorded at the tip during the approach curve to the substrate and IT,∞ is the current recorded at the tip in the bulk. Using this normalization, the currents are independent from the tip radius and redox mediator concentration. The d value is instead the spacing between tip and substrate and a is the disk radius of the Pt probe UME.


14

Page 15 of 27 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. Approach curve in feedback mode of the Pt before (solid line) and after (dashed line) 1h-PS at 2.00 V vs. Li+/Li constant potential applied at the GC in the O2-saturated TEGDME - 0.5 m LiTFSI (a) and TEGDME - 5 m LiTFSI (b). GC current profiles vs. time during the PSs in TEGDME - 0.5 m LiTFSI in (blue curve) and TEGDME - 5 m LiTFSI (red curve): only the first 600 s are shown (c).


15


Page 16 of 27

The two solid curves collected before the electrochemical test indicate that the current at the tip increases for d close to 0. This is an indication of the conductive nature of the GC substrate. After the PS, in TEGDME - 0.5 m LiTFSI, the UME current decreases, suggesting that GC becomes less conductive and that a passivation layer forms on GC (Figure 3a). Interestingly, in TEGDME - 5 m LiTFSI, the UME currents almost overlap before and after PS. This indicates that GC keeps its conductive behaviour even after the PS, and that its passivation is delayed (Figure 3b). The delayed passivation of GC in TEGDME - 5 m LiTFSI vs. TEGDME – 0.5 m LiTFSI is further supported by the substrate PS currents that keep higher values in the former solution than in the latter (Figure 3c). In conclusion, SECM is a very powerful tool to get insight into ORR at Li/O2 battery cathodes, that, however, requires the use of many, complementary techniques: spectroscopies, microscopies and electrochemical analyses. Here we demonstrate for the first time to our knowledge the use of SECM as a tool to study the effect of electrolyte composition on ORR, with a focus on IL and IL-like electrolytes based on PYR14TFSI, TEGDME and LiTFSI. The results clearly demonstrate the different stability of O2•- and of the GC during ORR in the investigated media. SECM clearly shows that Li+ affects O2•- lifetime. In Li+-free IL, O2•- is stable and diffuses from the GC to the Pt tip. On the contrary, O2•- is not detectable in solution both in PYR14TFSI - 0.1 M LiTFSI in and TEGDME - 0.5 m LiTFSI, indicating that it can be involved in side reactions even bringing to GC passivation. An interesting result has been obtained in TEGDME - 5 m LiTFSI. While the CV at GC well compares


16

Page 17 of 27 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


to that in PYR14TFSI - 0.1 M LiTFSI and is representative of irreversible ORR with formation of Li2O2 at the electrode surface, SECM suggests that in this electrolyte O2•- is stable enough to be detected at the Pt tip. This finding would suggest for an “in solution” mechanism formation of Li2O2 in such electrolyte. The TEGDME - 5 m LiTFSI electrolyte also enables a delayed passivation of the GC substrate. This study demonstrates that SECM is a powerful operando technique for screening the best electrolytes to be used in Li/O2 batteries in which the Li2O2 solution formation mechanism and a delayed electrode passivation would be desirable to extend the cell life time. Experimental Methods The 0.1 M LiTFSI in 1-butyl-1-methylpyrrolidinium bis(trifluoromethanesulfonyl)imide (PYR14TFSI, Solvionic) solution and the 0.5m and 5m solutions of lithium bis (trifluoromethane)sulfonimide (LiTFSI, Sigma-Aldrich, ≥99.0%) and tetraethylene glycol dimethyl ether (TEGDME, Sigma-Aldrich, ≥99.0%) were prepared and stored in glove box (MBraun, Ar atmosphere, H2O and O2

Oxygen redox reaction in Ionic Liquid and Ionic Liquid-like based

Recommend Documents