The (Electro)Chemistry of Ethylene Carbonate, Water and HF at the Negative Electrode in Li-ion Batteries

doi:10.21203/rs.3.rs-1950688/v1

Download PDF

Article

The (Electro)Chemistry of Ethylene Carbonate, Water and HF at the Negative Electrode in Li-ion Batteries

https://doi.org/10.21203/rs.3.rs-1950688/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Compared to aqueous electrolytes, the fundamental understanding of the chemical and electrochemical processes occurring in non-aqueous electrolytes in general is far less developed. This is no different for Li-ion battery (LiB) electrolytes, where many questions regarding the solid electrolyte interphase (SEI) on the anode side remain unanswered, including its chemical composition, the mechanism of formation and the impact on LiB performance. Here, we present a detailed experimental and theoretical study of the electrochemistry of ethylene carbonate (EC) and its chemical relationship with trace amounts of water and HF across a vast range of electrode materials, from well-ordered single crystals to realistic graphite electrodes. We reveal the electrocatalytic nature of EC, HF and water electroreduction at all interfaces. Moreover, we show that these reactions are connected in a closed cycle by chemical reactions, that take place either at the interface or in the bulk of the electrolyte. For the first time, we unveil the catalytic role of water in EC electroreduction and demonstrate that the composition of the SEI depends predominantly on the balance between the (electro)chemistry of EC, water and HF.

Increased demand for the utilization of renewable energy sources is driving the rapid development of energy storage technologies. With a huge market in portable electronic devices and the transportation sector, Li-ion batteries (LiBs) are on the forefront of energy storage for mobile applications^1,2. Despite their wide use and decades of research, however, LiBs still face enormous challenges related mostly to the loss of performance over time^3–5. It is commonly agreed that one of the most important drivers of efficiency loss at the anode side of the LiB is the solid electrolyte interface (SEI)^6–8. It is also accepted that the SEI is formed by the reduction of the electrolyte components in the first charging cycle, resulting in a barrier to further electrolyte decomposition that still allows for good Li⁺ ion conductivity^9–12. Typical LiB electrolytes consist of lithium hexafluorophosphate LiPF₆ in linear (ethyl-methyl carbonate EMC or dimethyl carbonate DMC) and cyclic (ethylene carbonate EC) carbonate solvents^13–15. As a result of these electrolyte constituents, the most frequently reported components of the SEI are LiF, Li₂CO₃, LiOH, Li₂C₂O₄ and lithium ethylene di-carbonate (LEDC)^9,16−18, which gives us the answer to “What?” the SEI is comprised of. Yet, depending on the report, different combinations or different ratios of these SEI compounds have been found^11,19−22. In some cases, even the chemical nature of some of the compounds has been questioned. Most recently, it has been suggested in a detailed study by Wang et al. that lithium ethylene mono-carbonate (LEMC), rather than LEDC is most likely the main organic component of the SEI²³. In addition, one can also find several contradictory reports on “Where?” at the interface these compounds can be found^7,11,24−26. Much less common, however, are the answers to “How?” and “Why?” these SEI-forming reactions happen, at least at the atomic/molecular level. Compared to aqueous electrolytes, the fundamental understanding of the chemical and electrochemical processes occurring in non-aqueous electrolytes in general is far less developed. In the case of LiB electrolytes, such understanding is impeded by the enormous complexity introduced by the number of possible competing and interrelated (electro)chemical reactions, determined by the components of the electrolyte and further complicated by the nature and morphology of the electrode material¹³. In fact, the understanding of the role of the electrode surface, potential and chemical composition of the electrolyte in the SEI formation is completely absent and as a result, the physical and chemical properties of SEI are a subject of an ongoing and vigorous debate, oftentimes pointing to substantially different nature, structure and origin of the main SEI components ⁹.

The employment of model systems, which has advanced the fundamental understanding of aqueous systems has rarely been rarely attempted for Li-ion battery systems. By reducing the number of parameters and hence the complexity it is possible to bring theory closer to experiment. We have recently employed our surface science-based approach to study the electrochemistry of two common components of LIB electrolytes HF and H₂O as well as their chemical relationship^27,28. We have demonstrated that potential dependent reorganization of the double layer enables the electrocatalytic transformation of these two species into their corresponding reaction products.

The study reported herein focuses on the electrochemical reduction of EC and its chemical relationship with trace amounts of water. To avoid any interference with HF, which is the most reactive species in LiPF₆-based LiB electrolytes, the experiments were performed in LiClO₄/EC electrolyte. The products of EC reduction are first investigated on model metal systems, i.e. single crystals of Au(111), Ir(111), Pt(111), Cu(111) and graphene on Pt(111). The knowledge gained from these model systems was then utilized on more complex carbon-based materials ranging from basal plane and edge-exposed highly ordered pyrolytic graphite (HOPG) to real graphite anodes. We show that the chemical nature of the most abundant organic component of the SEI is LEDC. More importantly, the experimental and computational results demonstrate the electrocatalytic nature and the role of water in LEDC formation. We show, that the electrocatalytic trend in the presence of water tightly follows the trend in electroreduction of water itself, which depends on the nature of the substrate and the density of surface defects. Finally, we investigate the behavior of the SEI created in LiClO₄/EC/H₂O electrolyte in the presence of HF, creating a closed cycle between interfacial and bulk (electro)chemical reactions of the three most important components of LIB electrolytes, i.e. HF, H₂O and EC. It is the balance between these reactions that ultimately determines the composition and the morphology of the SEI and, depending on the experimental conditions, leads to vastly different outcomes even in the same electrolyte.

Electrochemistry of H₂O and EC on metal single crystals

We start by investigating the electrochemical reduction of 1.3 M LiClO₄/EC electrolyte on metal single crystal electrodes. As a representative of metal surfaces, we focus here predominantly on Au(111). This surface served as our model system for the present study as well as the reference system for establishing our complete methodology that was then used on other samples. Similar results/trends, which were observed on Pt(111), Ir(111) and Cu(111), are summarized in the Supplementary Information. The choice of the single solvent electrolytes over LiPF₆/EC/EMC (LP57, commonly used in LiB electrolytes) was done to minimize the complexity of the system and allow the isolated study of the electrochemical response of each individual component. Furthermore, LiClO₄/EC was chosen specifically to avoid interference from HF, which is always present in LiPF₆ based electrolytes in millimolar concentrations as an impurity. As reported in our recent study²⁷, HF is the most reactive component in those electrolytes and gets electrochemically reduced before any other electrolyte component, effectively overriding all other processes that happen at more negative potentials.

Figure 1a shows a typical voltammogram for the Au(111) surface in 1.3 M LiClO₄/EC electrolyte, recorded at 1mV/s. Two overlapping peaks are observed in the cathodic scan at 1.7 and 1.5 V (all potentials are reported vs. Li/Li⁺). The anodic scan is rather featureless, indicating irreversibility of the process. The shape of the curve, as well as the second scan (see Figure S1), indicate a partial passivation of the surface, which is complete only after several scans. The AFM image after the first scan to 1 V clearly shows the presence of a granular film on the gold surface (Fig. 1f). It seems that the porosity of the film gives sufficient access of the electrolyte to the surface leading to diminished, but still significant currents in subsequent scans. As shown in Fig. 1b we observed a gradual increase of the reduction currents when increasing the water content in the electrolyte demonstrating that water is reduced in the 1.25–1.75 V potential window on Au(111) via the hydrogen evolution reaction (HER) (1):

$${\text{H}}_{\text{2}}\text{O+}{\text{e}}^{\text{-}}\text{+}{\text{Li}}^{\text{+}}\text{→}\text{LiOH}\text{+}\frac{\text{1}}{\text{2}}{\text{H}}_{\text{2}}$$

However, even in a dried electrolyte (water content 5 ppm), we still observed significant current response in this potential region. Moreover, the reaction order analysis shown in Figure S2 as log(i) vs log ([H₂O]) gives a reaction order of ~ 0.6. This is indicative of either multiple parallel reactions or of a multistep reaction taking place at the surface.

In order to further illuminate the nature of these reactions we have probed the chemical composition of the porous film at the electrode surface by means of Fourier transform infrared (FTIR) spectroscopy and X-ray photoelectron spectroscopy (XPS). Snapshots of the surface film composition were taken at 5 different potentials: 2 V, 1.75 V, 1.5 V, 1.25 V and 1 V. The evolution of FTIR and C 1s XPS spectra with increasingly negative potentials are shown in Figs. 1c and 1e, respectively. No detectable amount of any species was observed at or positive of 1.75 V consistent with no film observed on AFM images at 1.75 V. At 1.5 V several bands start to appear in the FTIR spectra and become increasingly stronger at 1.25 V and 1.0 V. We first note the absence of strong vibrations around 1800 cm^− 1; the region with the most prominent vibrational modes of the EC:LiClO₄ solvate (Figure S3). This indicates that even with gentle washing of our samples (see experimental section for details), we were able to remove most of the residual electrolyte from the sample surface and thus prevent any potential interference in the interpretation of the FTIR and XPS data. Next, we identified two sets of bands, which we attribute to two compounds forming the solid film observed on the Au (111) surface. The two broad absorption peaks at 1524 cm^− 1 and 1436 cm^− 1 and a sharp peak at 876 cm^− 1 (red dashed lines in Fig. 1c) give a close match with literature data^29,30 as well as the reference spectra of a Li₂CO₃ thin film (Figure S3). The second set of bands (black dashed lines in Fig. 1c) is assigned to a lithium alkyl carbonate R-CO₃Li, with uniquely characteristic vibrational modes for this group of compounds at 1665 cm^− 1, and 1318 cm^− 1 that belong to O-C = O asymmetric stretching and CH₂ wagging. It is commonly accepted by the LiB community that the Li-alkyl carbonate obtained by electrochemical reduction of ethylene carbonate is LEDC^9,31,32. More recently, however, this “common knowledge” has been disputed by Wang et al. in a rigorous study of complex interconversion equilibria between various R-CO₃Li compounds in DMSO. This study suggests that LEMC is the most likely component of the SEI²³. We will return to the discussion of the exact chemical nature of the alkyl carbonate later in the manuscript. At this point we continue with the analysis of the surface film, with FTIR data pointing to Li₂CO₃ and R-CO₃Li as the main two constituents of our SEI on the Au(111) surface. This interpretation is further corroborated by XPS data, which shows increased intensities of lithium carbonate and C-O functionalities at 290.0 eV and 286.5 eV, respectively, as the potential limit of the scan is decreased from 2 to 1 V. (Fig. 1e). These functional groups are further consistent with an alkyl carbonate or a mix of Li₂CO₃ with alkyl carbonate. We also note these are distinct from carbonate signals arising from the solvent, which are expected at higher binding energies of 291.0 eV as discussed further below. The appearance of the carbonates on the Au(111) surface exactly follows the current profile in the voltammogram, with significant amounts of carbonate(s) appearing only below 1.75 V.

Having previously established, that the observed current in the 1.75 V – 1 V potential range at least partially corresponds to water reduction, we were curious to see how the addition of water to the electrolyte affects the SEI formation. The AFM in Fig. 1g shows that a much thicker film of high porosity is formed. Furthermore, a change in morphology to from granular to more fiber-like is observed. Surprisingly, the chemical composition of the film did not significantly change, still predominantly consisting of Li₂CO₃ and R-CO₃Li. However, having deposited much more material, all the signals previously observed in the FTIR and XPS spectra for the SEI formed in the “low water content” electrolyte, were significantly accentuated. No new species were detected by either technique and no significant presence of LiOH was detected, as would have been expected in the case of amplified water reduction via reaction (1). On other metal surfaces, we made strikingly similar observations.

On all metals, a SEI comprised predominantly of R-CO₃Li and Li₂CO₃ is observed exactly in the potential range of water electroreduction (Figure S4). Because the HER from water commences at different potentials on individual metals, a trend is observed that precisely matches the HER activity trend on these metals, i.e. Ir ~ Pt > > Au > Cu. This trend, which closely follows the work function of pure metals, arises due to the necessity to establish a high enough coverage of Li⁺ at the electrode surface. The Li⁺ stabilizes the activated complex [Li⁺--OH₂--Li⁺], promoting the rate of reaction (1). These results suggest that water is not only involved in a parallel reaction of H₂ and LiOH formation but is actively participating in the reactions producing the two main components of the SEI on gold surface, i.e. R-CO₃Li and Li₂CO₃.

With substantially more material at the surface, it becomes easier to interpret the FTIR and XPS spectra, as the peak intensities of the main SEI components increase while potential impurities stay in the “background”. Again, the bands associated with Li₂CO₃ are clearly visible at 1505 cm^− 1 (note that a small shift from 1524 cm^− 1 is observed with thicker films) and 1436 cm^− 1 and 876 cm^− 1, while the complete set of vibrational frequencies arising from R-CO₃Li can now be seen well discerned at 1665 cm^− 1, 1450 cm^− 1, 1408 cm^− 1, 1345 cm^− 1, 1312 cm^− 1, 1110 cm^− 1, 1084 cm^− 1 and 829 cm^− 1 (Fig. 1d; for detailed assignments of all the bands see Figure S5 and Table S1). The latter set of absorption peaks give an exact match with the LEDC synthesized by Wang et al. in their recent study²³. Moreover, the absence of characteristic strong vibration modes at 1063 cm^− 1 and 3383 cm^− 1, that should be observed in compounds with C-OH functionality, suggests that the alkyl carbonate in our SEI is LEDC and not LEMC. The XPS data in Fig. 1e further support this claim, with the CO₃ vs C-O peak intensity ratio in favor of the carbonate group, consistent with a mixture of LEDC and Li₂CO₃.

Based on the above experimental evidence, we now turn to a possible reaction mechanism leading to the observed SEI composition. There are several possible routes for the electrochemical reduction of EC, involving multiple chemical and electrochemical steps. Most of them have been well documented in the literature^13,33−35. It seems that the first electron transfer creates a radical intermediate, which undergoes the ring opening via two possible paths, marked as path A and path B in Fig. 2. This short-lived intermediate can then undergo further chemical transformation or electron transfer, leading to a limited number of possible gaseous, soluble and insoluble products, including CO, CO₂, ethylene, lithium glycolate, lithium oxalate, Li₂CO₃ and LEDC (see reactions (2)-(7) in Fig. 2). By detecting evolved gases on the graphite anode during initial charge, Onuki et al. have shown that EC gets reduced by both paths A and B³³, most likely due to the sterically open nature of the cyclic EC molecule. Our FTIR and XPS analysis of the SEI on metals clearly confirms the formation of Li₂CO₃ and LEDC, presumably through path B in Fig. 2. As explained later in the text, we were able to confirm Onuki’s findings on graphite powder samples i.e., the EC reduction proceeds through both path A and B. Due to the low surface area of the metal samples, however, the evolved gases are below detection limit of on-line electrochemical mass spectrometry (OEMS), at least in the low water content electrolyte. This changes in the presence of water, where H₂ from reaction (1) and ethylene from reaction (6) can be detected on gold and platinum (Figure S11). Most importantly, however, none of the routes 1–6 provide any explanation for how water could be involved in the reaction mechanism and enhance the formation of Li₂CO₃ and LEDC as observed experimentally. We therefore suggest a third possible path C, which begins via reaction (8) with a nucleophilic attack of EC by OH⁻ generated from the electroreduction of water in reaction (1)

$$\text{E}\text{C}+{\text{L}\text{i}}^{+}+\text{O}{\text{H}}^{-}\to \text{L}\text{E}\text{M}\text{C}$$

Note that the “activated complex” in path C is actually LEMC, suggested by Wang et al. In fact, this compound is not a reduced form of EC at all, as it is obtained by mere recombination of EC and LiOH. Assuming a fast formation of OH⁻ in reaction (1) and a rate determining reaction with EC in reaction (8) would give a reaction order of 2/3 with respect to water, a plausible explanation for the observed value of 0.6. EC ring opening through OH⁻ driven hydrolysis is well documented³⁶ and can lead to complete decomposition to CO₂ and glycolate through reaction (9) or enter an electrochemical reduction via reaction (10), producing the intermediate that leads to Li₂CO₃ and LEDC.

In order to evaluate their relative favorability, we performed DFT calculations of the energetics of each path outlined in Fig. 2. We found that path B is more favorable than path A on all metals investigated (Figure S12). However, the EC ring-opening barrier observed in simulations following path B for metals cannot explain the catalytic EC reduction trends observed experimentally. Therefore, EC reduction in the presence of water was considered on the different metals as well. Following the H₂O reduction trends investigated previously (see section S3 in the SI), we probed the effect of three possible surface species on EC electroreduction: *Li, *LiOH and *Li₂OH (where * represents the metal active site). Figures 3a-c show the free energy diagrams for EC reduction on Au(111) involving *Li, *LiOH and *Li₂OH respectively. In Fig. 3a, EC reduction (EC ring-opening following path B in Fig. 2) shows a very high barrier (0.92 eV) for breaking the carbon-oxygen bond when only *Li is involved. Figure 3b and 3c illustrate that the presence of OH from water reduction stabilizes the EC ring-opening structure, forming the LEMC (HOCH₂CH₂CO₃Li) intermediate. From the LEMC intermediate the energy decreases towards LEDC as long as *LiOH or *Li₂OH species are present. Similar effects were observed for other metals, like Pt(111), Cu(111) and Ir(111) shown in Figure S15. We note that *LiOH is also involved in the HER from H₂O on metal surfaces, (see section S3 in the SI), indicating the connection between H₂O reduction and EC reduction. All the findings support that the presence of H₂O on all metal surfaces promotes LEDC formation from EC reduction. In addition, Figs. 3d-f show that the potential for the electrochemical response measured in experiments is correlated one-to-one with the adsorption energy of *LiOH and thereby to the work function of the metal surfaces. This suggests that the experimentally observed activity trends for EC reduction (Ir ~ Pt > > Au > Cu) can be described by the work function of the metals, since it determines the adsorption energy of *LiOH, which catalyzes EC towards LEDC. Note that this does not preclude the reaction to still partially proceed through paths A and B as well. However, as seen from the reaction mechanisms and their energetics, it is unlikely that LEMC is stable on the electrode surface as it can either chemically decompose through reaction (9) or electrochemically transform via reaction (10).

Electrochemistry of H₂O and EC on model carbon systems

To bridge the gap between model metal systems and real graphitic samples, we proceed to explore the H₂O and EC electrochemistry on model carbon systems, including epitaxially grown graphene with low surface defect density (LDG), and highly oriented pyrolytic graphite with either basal or edge plane exposed to the electrolyte (b-HOPG and e-HOPG). These surfaces provided different degrees of complexity in terms of the amount of defects (mostly grain boundaries and steps) found on each, following the order LDG < b-HOPG < < e-HOPG.

We first look at SEI formation on LDG samples. As shown by AFM imaging (Figure S7) LDG offers the closest possible approximation to a defect free well-ordered graphite surface, with large, ~ 10µm grains separated by rare grain boundaries. Note that no significant Li intercalation is expected due to the predominant “basal plane nature” of this sample, providing a good opportunity to study SEI formation deconvoluted from lithium intercalation. The electrochemical response of LDG, shown in Fig. 4a, is dominated by a sharp reduction peak observed at 0.5 V in the negative sweep from 3.2 V – 0.2 V. Unlike on the Au(111) surface, the graphene surface is almost completely passivated after the first sweep. AFM imaging of the surface after the first sweep confirms that a denser, more compact SEI is found on the surface (Fig. 4f). The same analysis protocol as in the case of metal samples is employed for the graphene sample, with snapshots of the surface film composition taken at 5 different potentials: 1 V, 0.8 V, 0.6 V, 0.4 V and 0.2 V. In contrast to the Au (111) surface, we first notice, that in addition to the two familiar sets of bands described below, the FTIR spectra reveal a third set of vibrations at 1864 cm^− 1, 1798 cm^− 1, 1481 cm^− 1, 1392 cm^− 1, 1160 cm^− 1, 1070 cm^− 1 and 971 cm^− 1, that start appearing at much more positive potentials, around 1 V. These belong to the EC:Li⁺ solvate (see comparison with reference spectra in Figure S6). Interestingly, these bands do not disappear even with extensive washing and are certainly not coming from the SEI on the graphene surface. Instead, our AFM image in Fig. 4d clearly shows the formation of bright pockets mostly around grain boundaries/defects. These bubble-like structures have been reported before on basal HOPG surface in several studies^37–39 and are thought to belong to the Li⁺:EC solvate, trapped between the graphene layers. This is confirmed by our FTIR results. The solvate bands disappear from the FTIR spectra after exposure to UHV, consistent with evacuation of the solvent from graphene multi-layer host and with the XPS spectra that show no significant solvent presence at any potential. As seen from the evolution of the FTIR spectra, solvate intercalation into the graphene structure depends on potential, however, only a tiny charge is passed between 1.0–0.8 V, while a significant intercalation of the solvated Li⁺ is observed. This seems to suggest that intercalation precedes the charge transfer, but a more detailed discussion of this phenomenon is beyond the scope of this paper. The main SEI formation process starts just negative of 0.6 V with the appearance of absorption peaks at 1524 cm^− 1, 1436 cm^− 1 and 876 cm^− 1, belonging to Li₂CO₃ and the signature vibrational modes at 1667 cm^− 1and 1318 cm^− 1, that belong to LEDC. Note that other LEDC peaks are somewhat obscured due to overlap with the solvent absorption bands. Nonetheless, the formation of LEDC/LiCO₃ is further confirmed by the C1s XPS spectra (Fig. 4g) which show the evolution of peaks at 290.0 eV and 286.5 eV just negative of 0.6 V, where we also start observing significant currents in the voltammogram. As stated previously, these belong to CO₃ and C-O functionalities, respectively. We conclude, therefore, that the electrochemistry of EC reduction is the same on graphene as it is on metals, albeit displaced by ~ 1 V towards more negative potentials, resulting in roughly similar SEI compositions, but with somewhat different morphologies.

In analogy to the experiments above on metals, we observe increases both in the reduction peak currents and the SEI thickness upon addition of 1000 ppm of water to the electrolyte (AFM in Fig. 4f). Again, significant differences in morphology are observed, with the film displaying more fabric-like properties. Both Li₂CO₃ and LEDC signals in FTIR and XPS C1s spectra also increase, confirming that the proposed reaction scheme on metals (Fig. 2) translates to these carbon systems as well. The main difference, however, is in the extent of the promoting effect that the same concentration of water has on graphene substrates. On Au (111), the increase in the reduction current is multiple times higher than on LDG. Similarly, a higher amplification of the LEDC and Li₂CO₃ signals in FTIR spectra is observed on gold compared to graphene, which can be linked to the amount of the two compounds produced during the potential sweep. It follows, that the graphene surface falls exactly on the same trend line as all the metals, i.e., the higher the work function of the electrode material, the more positive the adsorption potential of Li⁺ onto the surface and the more positive the reduction potential of EC and H₂O. In addition, the better the catalytic ability of the surface for H₂O reduction, the more pronounced is the enhancement of LEDC and Li₂CO₃ content in the SEI through path C in our reaction scheme.

To conclude the section on model carbon systems, we briefly look at the basal and edge plane HOPG. In addition to grain boundaries, these surfaces also contain a significant density of steps. A lot has been published on these two systems^{19,22,24,25,37−42}, so we just summarize our main observations here. As can be seen from Figure S8, the electrochemical response on basal HOPG depends on the quality of HOPG and the quality of individual cleavage, always displaying a peak at 0.5 V, but often showing additional peaks at 0.7 and 0.8 V. As shown in the accompanying AFM images, these two additional features appear on more defected HOPG surfaces. As in the case of the LDG, we can see an extensive solvate intercalation has already taken place before any significant currents. In the case of HOPG, the EC:Li⁺ co-intercalation is so extensive that it is now visible even with XPS (in spite of removal of a significant portion of the solvent in UHV). Typical EC R-CO₃ functionality is observed at ~ 1eV higher binding energy compared to LEDC and Li₂CO₃, i.e., at 291.0 eV. At 0.8 V, the HOPG terraces still look pristine. As we move towards more negative potentials, we eventually see the reaction commence at steps/defects and finally terraces at 0.5 V. AFM images at 0.2 V clearly show complete coverage of the terraces, while XPS data confirms the LEDC/Li₂CO₃ composition of the SEI. Finally, we look at the edge HOPG surface. Here, the massive current observed between 1.0 and 0.2 V belongs to lithium intercalation. Note that the processes of solvate intercalation as well as SEI formation, which is represented only by a tiny current, are superimposed on the Li-intercalation, making the deconvolution of individual processes practically impossible. This is part of the reason why SEI studies on real systems are plagued by so many interferences and ambiguous outcomes.

Electrochemistry of H₂O and EC in real systems

To complete the investigation of the H₂O-EC electrochemistry, we monitored the SEI formation on real, high-surface-area graphitic samples. As mentioned above, we believe that FTIR, XPS and electrochemical data on these complex samples can often give misleading results due to the interferences from the EC:Li⁺ solvate within the graphite structure as well as the trapped electrolyte in the SEI and chemicals used in the preparation of the graphite composite electrodes. OEMS, however, allows the quantitative monitoring of evolved gases during the SEI formation during the potential scan, while avoiding most of the interferences encountered in other techniques. In Fig. 5, we summarize the gas evolution data during the first potential scan in SMG-A5 graphite/Li cell. Because of the 100–500 times higher surface area of the graphite electrode compared to metal mesh or crystal electrodes, the gas evolution during the SEI formation is much more clearly visible compared to the metal mesh electrodes. Several observations are noteworthy: In LiClO₄/EC electrolyte with low water content, we detected three gases, namely H₂, CO and ethylene C₂H₄ that start evolving roughly at 0.75 V (Fig. 5a). Considering the high number of defects in the graphite powder, this potential correlates well with the potential where we first observe the formation of SEI on basal and edge HOPG. While hydrogen is produced through water reduction via reaction (1), CO and C₂H₄ are formed in the “EC cycle” in reactions (2), (6) and (7) through path A and B (see reaction scheme in Fig. 2), together with (CH₂OLi)₂, Li₂CO₃ and LEDC. Note that roughly equal amounts of CO and C₂H₄ (460 and 420 ppm, respectively) are produced in the first charging cycle in the low-water-content electrolyte. This suggests that the energetics of path A and B for this system are roughly the same. Interestingly, we also observe changes in the concentration of CO₂, which is inherently present in this electrolyte and is found in the headspace (200–300 ppm) at open circuit voltage (OCV). Upon polarization CO₂ is consumed at around 0.75 V, where the other gases are formed. We suggest that the most likely reaction responsible for the consumption of this native CO₂ is reaction (11), formation of Li₂CO₃ through neutralization of CO₂ with OH⁻, formed through the reduction of water in reaction (1). Most importantly, in the high-water-content electrolyte a 9-fold increase in ethylene production is detected, while CO formation is cut in half (Fig. 5b). This result complements the results obtained on model metal and carbon systems, which display a 5-10-fold increase of LEDC formation in the presence of 1000 ppm of water via reaction (6) as a coproduct of C₂H₄ and confirms the relationship between the water and EC reaction paths. As proposed above, OH⁻ created in the water reduction acts as a catalyst for reaction (6) by providing an energetically more favorable path C. While the initial amount of OH⁻ is “buffered” by inherent CO₂, the excess is available to form intermediate LEMC, which either decomposes into CO₂ and glycolate or is electrochemically transformed into LEDC and C₂H₄. This results in the observation of two opposite trends regarding CO₂ concentration, with an initial decrease followed by gradual and steady in concentration increase. Finally, in Fig. 5c we show the gas evolution LiPF₆/EC electrolyte. One of the characteristics of this electrolyte is the very low content of water, due to its quantitative reaction with PF₆⁻ to HF^27,43,44 via reaction (14) in the reaction scheme in Fig. 2. Therefore, as in the case of low-water-content LiClO₄/EC electrolyte, the contribution of reaction path C to the overall product formation is significantly diminished, pushing the gas evolution ratio CO:C₂H₄ in favor of CO.

In general, on all the systems investigated in the study, we find that the (electro)chemistry closely follows the same reaction mechanism. In disagreement with some previous reports⁴², we clearly show the electrocatalytic nature of the SEI formation, as well as a clear link between model and real systems.

The interaction of HF, H₂O and EC electrochemical cycles in Li-ion battery electrolytes

Considering that most commercial Li-ion battery electrolytes are based on LiPF₆ mixtures with cyclic and linear organic carbonates, it makes sense to establish a link between our findings in LiClO₄ electrolyte in this study and the commonly used LIB electrolytes. In our previous study, we have followed the formation of HF from water impurities and its electrocatalytic transformation to H₂ and LiF in LiPF₆ electrolyte²⁷. By exposing the SEI made in LiClO₄ to HF containing LiPF₆ we now link the HF, EC and H₂O reaction paths together. We notice that in 1.3 M LiPF₆/EC electrolyte, both carbonates are no longer detectable with FTIR after 60 s. In order to follow a slower evolution of the FTIR and XPS signals of the SEI, we diluted the LIPF₆/EC electrolyte with LiClO₄/EC electrolyte in a 1:10 ratio. Figure 6a shows the gradual decrease in LEDC and Li₂CO₃ vibrational modes upon exposure of the SEI on Au (111) surface to LiPF₆ electrolyte. Even after 5 min the AFM images (Fig. 6c) still show the presence of a film on the surface, albeit with a different morphology. As determined by XPS (Fig. 6d), this film is pure LiF, in agreement with our report on SEI composition in LiPF₆ electrolyte in a flooded cell. Most significantly, no carbonates are detected by XPS. Note that for clarity, the black curves in Fig. 6g are charge referenced differently from one another, even though they are from the same dataset, due to significant differential charging when a thick, highly insulating LiF layer is formed on the electrode surface. This phenomenon has been observed previously, and a detailed discussion about charge referencing of theses complex surfaces is provided in the SI for the interested reader. These results suggest that upon exposure to a sufficient amount of HF, both LEDC and Li₂CO₃ get quantitatively transformed into LiF via reactions (12) and (13).

$$L{i}_{2}C{O}_{3}+2HF\to C{O}_{2}+{H}_{2}O+2LiF$$

$$(C{H}_{2}C{O}_{3}Li{)}_{2}+2HF\to 2C{O}_{2}+HOC{H}_{2}C{H}_{2}OH+2LiF$$

The interaction of the three chemistries is best represented schematically and is summarized in Fig. 7. It shows the delicate balance between the chemistries of the three major players/electrolyte components that, for the most part, determine the composition, as well as morphology of the SEI.

At the electrode/electrolyte interface, which is represented by the triangle in Fig. 7, each of the three components, i.e., HF, H₂O and EC (shown in the corners of the triangle), enters its own electrochemical transformation. In this and our previous two studies, we demonstrated that all three electrochemical reactions are electrocatalytic in nature, an important fact that has often been overlooked. As a consequence, these reactions take place at vastly different potentials depending on the electrode material or different surface sites on the same material (e.g. terraces, steps, point defects).

The overall electrochemistry for these reactions, however, does not change from system to system, which allows the employment of much less complex model systems. These systems are by default closer to the ideal systems used in computational efforts such as DFT, which brings experiments closer to theory. Our studies of these three reactions show, that the electrocatalytic nature of these reactions does not stem only from the typical interaction of the reactants with the electrode surface in the bond breaking and bond making process, but also, at least in the case of HF and H₂O, from the potential dependent structure of the double layer. Specifically, the adsorption of Li⁺ onto the electrode surface, which allows the formation of energetically favorable activated complexes, is related to the work function of the electrode material, establishing a 2 V potential window between the most active Ir and the least active graphene surface. From a phenomenological point of view, this is a unique manifestation of the potential in electrocatalysis.

The complexity of the SEI formation and composition, however, does not end with the three electrocatalytic reactions described above. These electrochemical reactions are connected in a closed cycle by chemical reactions, that take place at the interface or in the bulk of the electrolyte as shown on the sides of the triangle in Fig. 7. Water in the bulk of the electrolyte enters the chemical reaction with LiPF₆ to form HF (reaction (14)). HF from the bulk of the electrolyte attacks the LEDC and Li₂CO₃ in the SEI and transforms them into LiF (reactions (12) and (13)). Finally, LiOH, the product of electroreduction of H₂O, serves as a catalyst for the electrochemical formation of LEDC and Li₂CO₃ (reactions (8), (10)), which has been shown for the first time in this study. We note that other reactions are possible and have been reported in the literature, but their products are either soluble and hence not present in the SEI or just not as abundant as the main components described above. The composition of the SEI therefore depends predominantly on the balance between the (electro)chemistry of EC, water and HF. This balance is heavily influenced by the experimental conditions used in a particular study, leading to vastly different outcomes, i.e. SEI compositions, even in the same electrolyte. While our study does not directly address the relationship between the SEI composition and cell performance, it gives us a blueprint of how to create SEIs of any composition and morphology, and thus link it to the cell performance in the future. We believe that our findings, while addressing important problematics relevant for LiB, point to much broader electrochemical phenomena, which are of general importance for our fundamental understanding of electrochemical interfaces.

Extended surface electrode preparation and electrochemical measurement

Pt(111), Ir(111), Au(111), and Cu(111) electrode surfaces having ~ 0.283 cm² geometric area (discs of 6 mm diameter and 4 mm length, Princeton Scientific Corp.) were prepared by inductive heating (EASYHeat, Ambrell) for 7 min at 1323 K for Pt, 5 min at 1073 K for Au and Cu, and 10 min at 1473 K for Ir in an controlled atmosphere (97% Ar/3%H2 gas mixture, Airgas). The crystal discs were slowly cooled down (ca. 7 min) to the laboratory temperature (~ 294 K) under the same atmosphere condition. The electrode surface was protected by a water droplet before being exposed to the laboratory’s atmosphere, then carefully assembled into the rotating disk electrode (RDE) configuration (a polypropylene thin sheet was used to support the electrode disc during the RDE assembling, with the electrode surface facing towards the sheet protected by a thin water film). After electrode was assembled, the RDE (the electrode surface always protected by a water droplet) was transferred into the glove box (< 0.80 ppm of oxygen gas, < 0.5 ppm of moisture) antechamber and let the water droplet dry completely (under low-vacuum condition) before being transferred into the glovebox’s main compartment. Electrochemical measurements were performed inside a glovebox using a three-electrode glass electrochemical cell. Au wire was used as a counter electrode, separated from the working electrode compartment by a frit. Li/Li + electrode was used as a reference electrode. The electrode surface was being immersed into the electrolyte at controlled potential (3.2 V). All potentials are given vs. Li/Li + electrode. All measurements were controlled using a potentiostat (PGSTAT 302N, Metrohm Autolab). Electrochemical measurements were done with scan rate of 1 mV/s in the negative direction.

Preparation of carbon samples

Epitaxially grown graphene with low surface defect concentration (LDG) was grown on Pt(111) heated to 1473 K in methane flow by chemical vapor deposition (CVD).

The HOPG ZYA type crystal with basal exposed plane (Princeton Scientific Corp.) was assembled into RDE and freshly cleaved with adhesive tape inside the glove box prior to each experiment. The HOPG ZYA type crystal with edge exposed plane (Princeton Scientific Corp.) was cut parallel to the edge surface with a stainless-steel blade to expose fresh surface and subsequently washed with THF, dried and inserted into RDE.

Chemicals

1.3 M LiClO₄/EC and 1 M LiPF₆/EC were prepared from EC (Gotion), LiClO₄ (Battery grade, Sigma-Aldrich) and LiPF₆ (BASF). Tetrahydrofuran (THF, Sigma-Aldrich) was used for rinsing the electrodes after electrochemical measurements.

Water determination: Water content in electrolytes was measured using the Karl Fischer titration system (Metler-Toledo) placed inside an Ar-filled glovebox with H₂O level below 0.5 ppm.

FTIR Measurements

Fourier-transformed infrared spectroscopy (FTIR) was used to probe the species present at the electrode surface after SEI formation. We utilized ex-situ grazing angle external reflectance FTIR (Thermo-Fisher Nicolet iS50 with Pike Technologies VeeMax III accessory) at 65º incidence angle (most sensitive to the surface species) with p-type polarized light (ZnSe polarizer positioned at the incident beam). After electrochemical SEI formation, the electrode was carefully rinsed with THF to remove any electrolyte residue (without compromising the SEI) and sealed in a glass jar prior to transferring to the FTIR instrument inside a glovebox. After transfer, the glass jar was opened, and the electrode surface was positioned in a way that guarantees the infrared light is directly reflected into the detector. The MCT detector (with spectral resolution of 2 cm^− 1) of the FTIR instrument was cooled with liquid argon to provide enough sensitivity for detection of surface species. The electrode surface prior to electrochemical cycling served as baseline for absorbance determination. Band assignment was performed with assistance of established literature^9,23,29−32.

AFM Measurements

The atomic force microscopic (AFM) measurements were performed on a Bruker Dimension Icon AFM at tapping mode with a silicon cantilever (RFESPA-75, Bruker). The experiments were conducted in a vibrational isolation enclosure for imaging. All images were taken with scan rate of 0.6 Hz, and scan sizes of 1 µm and 5 µm. After the images were acquired, the Nanoscope Analysis software was used to process the images. Prior to the AFM measurements, all samples (electrode surfaces) were rinsed with THF to remove any residual electrolyte.

XPS Measurements

XPS measurements were performed using a Specs PHOIBOS 150 hemispherical energy analyzer using a monochromated Al Kα X-ray source. The load-lock of the analytical UHV system is connected directly to an Ar-filled glove box, enabling the loading of samples without any exposure to ambient atmosphere. Survey spectra were measured using a pass energy of 40 eV at a resolution of 0.2 eV/step and a total integration time of 0.2 sec/point. Core level spectra were measured using a pass energy of 20 eV at a resolution of 0.05 eV/step and a total integration time of 0.5 sec/point. A charge neutralizing electron flood gun was used for some samples to attempt to address differential charging between highly insulating LiF and conductive substrates. Deconvolution was performed using CasaXPS software with a Shirley-type background and 70 − 30 Gaussian-Lorentzian peak shapes. Differential charging of SEI layers relative to the conductive substrates required different strategies depending on substrate and SEI chemistry. Au(111) samples cycled in 1.3 M LiClO₄/EC electrolytes and reacted with LiPF₆ electrolytes were primarily charge referenced using sp³ carbon in the C 1s at 284.8 eV. Au(111) samples with higher amounts of LiF exhibited differential charging that was not easily corrected using charge neutralization, requiring an alternative charge referencing strategy using the position of sp³ carbon for C 1s data and LiF (685.5 eV in the F1s) for the F 1s data. Pristine HOPG samples used the position of graphitic carbon at 284.0 eV, while cycled graphene/HOPG samples used the position of Li carbonates in the Li 1s at 55.4 eV (consistent with the position measured on Au(111) samples) to control for differential charging as a function of SEI thickness. See the Supplementary Methods for a more detailed discussion of charge referencing, differential charging and charge compensation on these samples.

OEMS measurements:

OEMS experiments were performed in an all PEEK OEMS cell similar to the cell developed by TU Munich⁴⁵.Lithium foil (99.9% purity, Albemarle) was used as the counter electrode (CE) and 2 glass fiber disks (VWR, 205 µm thickness, 15 mm diameter) were used as separator. The working electrode (WE) was placed on the separator and 1 ml electrolyte was put on the electrode; the electric contact for the WE was made with a Pt wire (99.99% purity, Goodfellow, 0.1 mm diameter). The PEEK cell design allows for experiments without contact of the electrolyte with stainless steel.

All pieces of cell hardware were dried for at least 12 h at 60°C in dynamic vacuum before usage. The glass fiber separators were dried for 5 h at 300°C under dynamic vacuum in a glass oven (Buchi, Switzerland).

The graphite electrodes for the OEMS measurements consisted of 97 wt% SMG-A5 graphite powder (Hitachi), 1.5 wt% carboxymethyl cellulose binder (CMC Sunrose MAC200, NPI, Japan) and 1.5 wt% styrene-butadiene rubber binder (SBR, Zeon, Japan), the materials were mixed in a planetary orbital mixer (Thinky, USA) at 2000 rpm and 50 mbar for 10 min. The resulting ink was blade-coated onto a 20 µm thick, porous Celgard separator (H2013, Celgard, USA) at a wet-film thickness of 50 µm using an automatic coater (RK Print, UK).

This configuration allows for a reasonably short diffusion time for gases produced by the graphite electrode to the flow-restricting capillary (on the order of several minutes), which connects the OEMS cell with the mass spectrometer, which is a requirement for measurements with the OEMS system⁴⁵

After assembling the cell and attaching the cell to the OEMS system, the cell was purged with argon and subsequently kept at open circuit voltage (OCV) for 4 h in order to obtain stable OEMS background signals. After this, a potential scan from OCV to 0.3 V vs. Li/Li⁺ followed by a scan to 3.0 V vs. Li/Li⁺ was performed with a scan rate of 0.2 mV/s. All potentials shown are referenced vs. Li/Li⁺ CE. To quantify the amount of evolved gases, the ion currents at different m/z values (I_z) were normalized by the signal for the ³⁶Ar isotope and calibrated with calibration gases containing 2000 ppm of CO, C₂H₄, and H₂ in pure argon (further details can be found in Metzger et al.⁴⁶).

DFT Calculations

The calculations are performed using the GPAW code^47,48. The ASE package⁴⁹ has been used to handle the crystal structures. The calculations are performed using revised Perdew-Burke-Ernzerhof (RPBE)⁵⁰ as exchange-correlation functional. The calculations for these intermediates’ adsorption energies are performed using a 5x4x3 unit cells with the last bottom layer kept frozen to reproduce the bulk of the slab and the two top layers let free to relax as well as the adsorbate. We use a 3 × 3 × 1 Monkhorst-Pack k-point (Gamma) grid and the wave functions were represented on a uniform real-spaced grid with 0.18Å grid-spacing. A vacuum of minimum 10Å on z direction was employed. (For more details see Supplementary)

Data availability: All data is available from the authors upon reasonable request

Acknowledgements

This research was sponsored by BMW Technology Corporation. Support from Peter Faguy at the Vehicle Technologies Office (VTO), Office of Energy Efficiency and Renewable Energy, U.S. Department of Energy, is also gratefully acknowledged. The research was conducted at Argonne National Laboratory - a US Department of Energy Office of Science laboratory operated by UChicago Argonne under contract number DE-AC02-06CH11357.

Author contributions

M.Z. and D.S. conceived and designed the experiments. M.Z., D.S., J.G.C., D.H., P.F.B.D.M., B.G., P.P.L. and B.G. performed the experiments. I.E.C., H.W., K.L.S. and J.R. performed the calculations. M.Z., D.S., J.G.C., P.F.B.D.M., I.E.C., F.M., J.R., I.E.C., R.J. and N.M.M. discussed the results and wrote the paper.

Conflicts of interest

There are no conflicts of interest to declare.

Goodenough, J. B. & Park, K. S. The Li-ion rechargeable battery: A perspective. J. Am. Chem. Soc. 135, 1167–1176 (2013).
Zubi, G., Dufo-López, R., Carvalho, M. & Pasaoglu, G. The lithium-ion battery: State of the art and future perspectives. Renew. Sustain. Energy Rev. 89, 292–308 (2018).
Palacín, M. R. Understanding ageing in Li-ion batteries: A chemical issue. Chem. Soc. Rev. 47, 4924–4933 (2018).
Kabir, M. M. & Demirocak, D. E. Degradation mechanisms in Li-ion batteries: a state-of-the-art-review. Int. J. Energy Res. 41, 1963–1986 (2017).
Etacheri, V., Marom, R., Elazari, R., Salitra, G. & Aurbach, D. Challenges in the development of advanced Li-ion batteries: A review. Energy Environ. Sci. 4, 3243–3262 (2011).
Gauthier, M. et al. Electrode-Electrolyte Interface in Li-Ion Batteries: Current Understanding and New Insights. J. Phys. Chem. Lett. 6, 4653–4672 (2015).
Edström, K., Herstedt, M. & Abraham, D. P. A new look at the solid electrolyte interphase on graphite anodes in Li-ion batteries. J. Power Sources 153, 380–384 (2006).
Aurbach, D., Zinigrad, E., Cohen, Y. & Teller, H. A short review of failure mechanisms of lithium metal and lithiated graphite anodes in liquid electrolyte solutions. Solid State Ionics 148, 405–416 (2002).
Verma, P., Maire, P. & Novák, P. A review of the features and analyses of the solid electrolyte interphase in Li-ion batteries. Electrochim. Acta 55, 6332–6341 (2010).
Soto, F. A., Ma, Y., Martinez De La Hoz, J. M., Seminario, J. M. & Balbuena, P. B. Formation and Growth Mechanisms of Solid-Electrolyte Interphase Layers in Rechargeable Batteries. Chem. Mater. 27, 7990–8000 (2015).
Peled, E. & Menkin, S. Review—SEI: Past, Present and Future. J. Electrochem. Soc. 164, A1703–A1719 (2017).
An, S. J. et al. The state of understanding of the lithium-ion-battery graphite solid electrolyte interphase (SEI) and its relationship to formation cycling. Carbon N. Y. 105, 52–76 (2016).
Xu, K. Electrolytes and interphases in Li-ion batteries and beyond. Chem. Rev. 114, 11503–11618 (2014).
Xu, K. Nonaqueous liquid electrolytes for lithium-based rechargeable batteries. Chem. Rev. 104, 4303–4417 (2004).
Zhang, S. S., Jow, T. R., Amine, K. & Henriksen, G. L. LiPF 6 ± EC ± EMC electrolyte for Li-ion battery. J. Power Sources 107, 18–23 (2002).
Aurbach, D. Review of selected electrode-solution interactions which determine the performance of Li and Li ion batteries. J. Power Sources 89, 206–218 (2000).
Wang, A., Kadam, S., Li, H., Shi, S. & Qi, Y. Review on modeling of the anode solid electrolyte interphase (SEI) for lithium-ion batteries. npj Comput. Mater. 4, (2018).
Liu, T. et al. In situ quantification of interphasial chemistry in Li-ion battery. Nat. Nanotechnol. 14, 50–56 (2019).
Leroy, S., Martinez, H., Dedryvère, R., Lemordant, D. & Gonbeau, D. Influence of the lithium salt nature over the surface film formation on a graphite electrode in Li-ion batteries: An XPS study. Appl. Surf. Sci. 253, 4895–4905 (2007).
Chattopadhyay, S. et al. In situ X-ray study of the solid electrolyte interphase (SEI) formation on graphene as a model Li-ion battery anode. Chem. Mater. 24, 3038–3043 (2012).
Owejan, J. E., Owejan, J. P., Decaluwe, S. C. & Dura, J. A. Solid electrolyte interphase in Li-ion batteries: Evolving structures measured in situ by neutron reflectometry. Chem. Mater. 24, 2133–2140 (2012).
Leroy, S. et al. Surface film formation on a graphite electrode in Li-ion batteries: AFM and XPS study. Surf. Interface Anal. 37, 773–781 (2005).
Wang, L. et al. Identifying the components of the solid–electrolyte interphase in Li-ion batteries. Nat. Chem. 11, 789–796 (2019).
Peled, E. et al. Composition, depth profiles and lateral distribution of materials in the SEI built on HOPG-TOF SIMS and XPS studies. J. Power Sources 97–98, 52–57 (2001).
Tsubouchi, S. et al. Spectroscopic Characterization of Surface Films Formed on Edge Plane Graphite in Ethylene Carbonate-Based Electrolytes Containing Film-Forming Additives. J. Electrochem. Soc. 159, A1786–A1790 (2012).
Heiskanen, S. K., Kim, J. & Lucht, B. L. Generation and Evolution of the Solid Electrolyte Interphase of Lithium-Ion Batteries. Joule 3, 2322–2333 (2019).
Strmcnik, D. et al. Electrocatalytic transformation of HF impurity to H2 and LiF in lithium-ion batteries. Nat. Catal. 1, 255–262 (2018).
Castelli, I. E. et al. The role of an interface in stabilizing reaction intermediates for hydrogen evolution in aprotic electrolytes. Chem. Sci. 11, 3914–3922 (2020).
Pasierb, P., Komornicki, S., Rokita, M. & Rȩkas, M. Structural properties of Li2CO3-BaCO3 system derived from IR and Raman spectroscopy. J. Mol. Struct. 596, 151–156 (2001).
Wijaya, O. et al. A gamma fluorinated ether as an additive for enhanced oxygen activity in Li-O2 batteries. J. Mater. Chem. A 3, 19061–19067 (2015).
Zhuang, G. V., Xu, K., Yang, H., Jow, T. R. & Ross, P. N. Lithium ethylene dicarbonate identified as the primary product of chemical and electrochemical reduction of EC in 1.2 M LiPF 6/EC:EMC electrolyte. J. Phys. Chem. B 109, 17567–17573 (2005).
Lanz, P. & Novák, P. Combined In Situ Raman and IR Microscopy at the Interface of a Single Graphite Particle with Ethylene Carbonate/Dimethyl Carbonate. J. Electrochem. Soc. 161, A1555–A1563 (2014).
Onuki, M. et al. Identification of the Source of Evolved Gas in Li-Ion Batteries Using [sup 13]C-labeled Solvents. J. Electrochem. Soc. 155, A794 (2008).
Zhuang, G. V., Yang, H., Blizanac, B. & Ross, P. N. A study of electrochemical reduction of ethylene and propylene carbonate electrolytes on graphite using ATR-FTIR spectroscopy. Electrochem. Solid-State Lett. 8, 441–445 (2005).
Strehle, B., Solchenbach, S., Metzger, M., Schwenke, K. U. & Gasteiger, H. A. The Effect of CO 2 on Alkyl Carbonate Trans-Esterification during Formation of Graphite Electrodes in Li-Ion Batteries. J. Electrochem. Soc. 164, A2513–A2526 (2017).
Metzger, M., Strehle, B., Solchenbach, S. & Gasteiger, H. A. Hydrolysis of Ethylene Carbonate with Water and Hydroxide under Battery Operating Conditions. J. Electrochem. Soc. 163, A1219–A1225 (2016).
Matsuoka, O. et al. Ultra-thin passivating film induced by vinylene carbonate on highly oriented pyrolytic graphite negative electrode in lithium-ion cell. J. Power Sources 108, 128–138 (2002).
Jeong, S.-K., Inaba, M., Abe, T. & Ogumi, Z. Surface Film Formation on Graphite Negative Electrode in Lithium-Ion Batteries: AFM Study in an Ethylene Carbonate-Based Solution. J. Electrochem. Soc. 148, A989 (2001).
Inaba, M. et al. Electrochemical scanning tunneling microscopy observation of highly oriented pyrolytic graphite surface reactions in an ethylene carbonate-based electrolyte solution. Langmuir 12, 1535–1540 (1996).
Domi, Y. et al. Electrochemical AFM Observation of the HOPG Edge Plane in Ethylene Carbonate-Based Electrolytes Containing Film-Forming Additives. J. Electrochem. Soc. 159, A1292–A1297 (2012).
Domi, Y. et al. In situ AFM study of surface film formation on the edge plane of HOPG for lithium-ion batteries. J. Phys. Chem. C 115, 25484–25489 (2011).
Alliata, D., Kötz, R., Novák, P. & Siegenthaler, H. Electrochemical SPM investigation of the solid electrolyte interphase film formed on HOPG electrodes. Electrochem. commun. 2, 436–440 (2000).
Lux, S. F. et al. The mechanism of HF formation in LiPF6 based organic carbonate electrolytes. Electrochem. commun. 14, 47–50 (2012).
Plakhotnyk, A. V., Ernst, L. & Schmutzler, R. Hydrolysis in the system LiPF 6 - Propylene carbonate - Dimethyl carbonate - H 2O. J. Fluor. Chem. 126, 27–31 (2005).
Tsiouvaras, N., Meini, S., Buchberger, I. & Gasteiger, H. A. A Novel On-Line Mass Spectrometer Design for the Study of Multiple Charging Cycles of a Li-O 2 Battery. J. Electrochem. Soc. 160, A471–A477 (2013).
Metzger, M., Strehle, B., Solchenbach, S. & Gasteiger, H. A. Origin of H 2 Evolution in LIBs: H 2 O Reduction vs. Electrolyte Oxidation. J. Electrochem. Soc. 163, A798–A809 (2016).
Enkovaara, J. et al. Electronic structure calculations with GPAW: A real-space implementation of the projector augmented-wave method. J. Phys. Condens. Matter 22, (2010).
Mortensen, J. J., Hansen, L. B. & Jacobsen, K. W. Real-space grid implementation of the projector augmented wave method. Phys. Rev. B - Condens. Matter Mater. Phys. 71, 1–11 (2005).
Larsen, A. H. The Atomic Simulation Environment - A Python library for working with atoms. J. Phys. Condens. Matter 29, 1–21 (2012).
Hammer, B., Hansen, L. B. & Nørskov, J. K. Improved adsorption energetics within density-functional theory using revised Perdew-Burke-Ernzerhof functionals. Phys. Rev. B - Condens. Matter Mater. Phys. 59, 7413–7421 (1999).

There is NO Competing Interest.

ECH2OSIfinalvNC.docx
Supplementary Information File

Download PDF

Version 1

posted

You are reading this latest preprint version

The (Electro)Chemistry of Ethylene Carbonate, Water and HF at the Negative Electrode in Li-ion Batteries

Status:

Version 1

Abstract

Figures

Introduction

Results And Discussion

Electrochemistry of H₂O and EC on metal single crystals

Electrochemistry of H₂O and EC on model carbon systems

Electrochemistry of H₂O and EC in real systems

The interaction of HF, H₂O and EC electrochemical cycles in Li-ion battery electrolytes

Methods

OEMS measurements:

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1

The (Electro)Chemistry of Ethylene Carbonate, Water and HF at the Negative Electrode in Li-ion Batteries

Status:

Version 1

Abstract

Figures

Introduction

Results And Discussion

Electrochemistry of H2O and EC on metal single crystals

Electrochemistry of H2O and EC on model carbon systems

Electrochemistry of H2O and EC in real systems

The interaction of HF, H2O and EC electrochemical cycles in Li-ion battery electrolytes

Methods

OEMS measurements:

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1

Electrochemistry of H₂O and EC on metal single crystals

Electrochemistry of H₂O and EC on model carbon systems

Electrochemistry of H₂O and EC in real systems

The interaction of HF, H₂O and EC electrochemical cycles in Li-ion battery electrolytes