Project Topics
Get Complete Project Material File(s) Now! »
Influence of the solvation shell on the electrolyte physicochemical properties and cation intercalation
It is widely admitted by the battery community that rate capability of alkali metal-ion batteries are influenced by the transport properties of these alkali metal ions in the bulk of the electrolyte, e.g. ionic conductivity and transference number.123 K+ ions are weaker acids than Na+, which are even weaker than Li+, this difference of Lewis acidity results in smaller Stokes’ radii in various solvents. The Figure I-7a depicts their Shannon’s ionic124 radii and Stokes’ radii in propylene carbonate. According to Matsuda et al., their Stokes’ radii in PC increase in the opposite direction of their Shannon’s ionic radii.125 Lithium ion charge is more localized leading to a stronger interaction with solvent molecules and therefore a bigger solvation sheath. Then, the conventional vehicle-type mechanism in dilute electrolytes explains the better ion transport properties of Na+ and K+ compared to Li+.
Quartz Crystal Microbalance theory and associated techniques
The core principle behind Quartz Crystal Microbalance (QCM) is nested in the piezoelectricity effect. Its discovery dates back to the research of René Just Haüy; however, its study and interpretation are ascribed to the work of Pierre and Jacques Curie brothers (1880) on tourmaline, quartz and Rochelle salt.137 They demonstrated the appearance of an electrical field, which is proportional to the applied mechanical stress on these crystalline materials with no inversion symmetry. The converse effect was mathematically deduced by Gabriel Lippmann in 1881138 and the Curie brothers confirmed its existence.139 The practical applications of this phenomenon are several, starting by the development of sonar (sound navigation ranging) by Paul Langevin during the World War I (1917) to detect submarines to nowadays utilization in horology.
Frequency response in a liquid medium – Kanazawa and Gordon equation
The emergence of QCM measurements in the liquid media has stimulated the studies investigating the influence of liquid on the resonant frequency of a resonator, which, later on, have also found implications in QCMs coupled to electrochemical methods.
When the quartz surface is immersed in a liquid, the shear wave propagates from the quartz into the liquid volume in its vicinity. However, the wave is dampened and its amplitude decays moving away from the quartz. Importantly, the presence of the liquid introduces a new “layer” to the system resulting in a shift in resonant frequency, leading to a new baseline, by an amount corresponding to the mass of viscously coupled fluid layer.144,145 The effect of immersing the QCM in liquid results in a decrease in frequency, which can be estimated thanks to the Kanazawa and Gordon equation (1985):146 𝛥𝑓=−𝑓032√𝜌𝑙𝜂𝑙𝜋𝜌𝑞𝜇𝑞Equation II-4.where 𝜌𝑙 and 𝜂𝑙 are the density (g.cm-3) and viscosity of the liquid (Pa.s), respectively.
Oscillator circuit and QCM with motional resistance monitoring
An electrical equivalent circuit composed of equivalent circuit elements that represent its physical parameters under oscillation can model the quartz resonance. The so-called Butterworth-Van Dyke (BVD) equivalent circuit is used to describe an unperturbed (without additional film or electrolyte) quartz resonator, as represented in the Figure II-4a. In the series branch, the inductor, L1 represents the inertial component related to the mass displaced during oscillation, the capacitor, C1 represents the energy stored during oscillation and the resistance, R1 represents the energy dissipated during oscillation, which is also called the motional resistance. This series branch is referred as the motional branch as it defines the electromechanical characteristics of the quartz resonator. The parallel branch, referred as electrical branch, is composed of a capacitor, C0, which arises from the electrical capacitance of the dielectric medium, i.e. the quartz, sandwiched between the two electrodes on both side of the quartz.
QCM experiments, however, involve quartz crystal surface that are either immersed in liquid media, coated with films, or both, giving rise to a more complex electrical equivalent circuit (Figure II-4b). First, the liquid film has an influence on the resonant frequency as noted by the Kanazawa and Gordon equation (Equation II-4), which depends on its density, 𝜌𝑙 and viscosity, 𝜂𝑙. It can be described as the addition of a mass loading density component, LL and a resistive viscosity component, RL of the liquid. Concerning the coatings that are deposited on the conducting electrode of the resonators, two cases can be distinguished. If the coating is “acoustically thin” and rigid, then it will vibrate in phase with the oscillator following the Sauerbrey equation and causing the addition of a mass loading, LF and a dissipation of energy, RF. Apart from this ideal case, the coating can be viscoelastic, defined with its own physical properties: viscosity (loss modulus), density and elasticity (storage modulus). This will result in the addition of an extra mass loading, LV, a dissipation of energy, RV and a stored energy arising from the elasticity, CV to the circuit. Nonetheless, the electrical equivalent circuit of the practical assembly can be generalized to a classical BVD circuit (Figure II-4c).
QCM with dissipation monitoring at different overtones
In addition to previous read-out systems, a relatively recent method to access the key parameters of the resonance is to excite the resonator with a narrow radio frequency pulse having a frequency matching the expected resonance frequency (Figure II-6a). The excitation is intermittently turned off and resonance decays freely, also called the ring-down method or “pinging”. The Figure II-6b demonstrates that the resonance frequency and the dissipation factor (half-bandwidth) can be obtained from the fitting of the current-versus-time trace with a decaying cosine. It can be noted that the obtained signal is linked to the Figure II-5a presented in the previous subsection by a Fourier transform, passing from the time to the frequency domain. This approach was developed by Rodahl et al.149,150 This measurement methodology is the basis of the QCM-D where D is for “Dissipation monitoring” and the term QCM-D is a trademark owned by Q-Sense (Biolin Scientific). In addition, the different overtones can be excited successively.
EQCM-based methods for the study of Solid Electrolyte Interphase
The use of EQCM-based methods for the characterization of batteries is mainly rooted in the better understanding of the Solid Electrolyte Interphase (SEI). As previously introduced (see I.2.b), the SEI is a new phase mainly formed at the surface of the negative electrode due to electrochemical instability of the electrolyte at such negative potentials. The understanding of the physical properties of this newly formed phase and its evolution are of prior importance for the good performance and enhanced safety of the battery.
Aurbach et al. were the first researchers to use the real-time assessment of mass variation provided by EQCM and the Sauerbrey equation (Equation II-3) in 1995, to study the surface film formation over a bare gold electrode in LiPF6 and LiAsF6 in propylene carbonate (PC).152 The 𝑀/𝑧 value (Equation II-7) was calculated during a negative sweep of potentials in PC-based electrolytes containing different salts and contaminants as water or carbon dioxide. Thanks to this procedure, they could demonstrate the potential-dependent formation mechanism of the SEI, with the solvent reduction at low potentials and the LiF film formation at higher potentials creating an inner inorganic layer covered by a polymeric layer. Then, SEI studies have moved to systems that are more realistic, starting by PVD-prepared carbon. Unlike previously employed model electrodes, the mass change appearing during the potential sweep turned out to be also dependent on the lithium ion insertion/extraction, which needs to be untangled from the mass change caused by solely SEI formation. In this context, Kwon and Evans managed to determine the SEI average density proving its mixed organic and inorganic composition with the initial formation of Li2O in fluorine-free electrolytes.153,154 Such a methodology has been expanded to other active materials as electrodeposited Sn film155 and recently drop-cast graphite156 on the surface of EQCM resonators, leading to similar results but dependent on the electrolyte composition: salt and solvent. In the framework of this Ph.D. thesis, a similar strategy has been applied to prove the formation of a passivating layer composed of fluorinated products induced by the water reduction in the trendy “Water-in-Salt” electrolyte157 system but this study will not be described in this manuscript. Although limited number of studies exist so far, the development of EQCM with dissipation monitoring has contributed to assessing viscoelastic properties of the SEI and changes in electrolyte viscosity, which were unreachable (or entangled) with classical EQCM measurements. Yang et al.158,159 attested the beneficial addition of fluoroethylene carbonate to the ethylene carbonate/ethyl methyl carbonate-based electrolyte for the cycling of tin anode. It gave rise to a stiffer and more homogeneous SEI, corroborating the observed better capacity retention. This type of methodology has been applied by the group of Levi and Aurbach to prove the better quality of SEI in LiTFSI than that formed using LiPF6 + vinylene carbonate on high potential Li4Ti5O12 (LTO) anodes.160 Going further in this process, this study opened up the possibility of using EQCM-D to screen different electrolyte compositions leading to a more efficient SEI, and to access to the predictive information for the practical applications in a short time. Kitz and co-workers adopted an analogous strategy, to model bare electrodes, except that, Electrochemical Impedance Spectroscopy (EIS) and Online Electrochemical Mass Spectrometry (OEMS) were added to the tool panel in order to reveal the SEI formation mechanism. In agreement with the literature, the formation of a rigid LiF-rich interphase was shown to occur prior the emergence of a less rigid film of reduced solvent products during a negative sweep of potentials, demonstrating the highly dynamic nature of this interphase, in terms of both mechanical and electrochemical properties.161 The same experiments in presence of water demonstrated that the SEI becomes thicker and more rigid and this was explained by the formation of Li2CO3-rich early interphase preventing the formation of a soft reduced solvent-based film.162 The contribution of VC and FEC additives on SEI were also investigated. Both additives change the SEI composition causing an increase of the SEI shear modulus. However, the Li-ion conductivity in the SEI is increased in the case of VC and hindered for FEC.163 Searching for complementary compositional information, EQCM-D coupled with surface-enhanced Raman spectroscopy evidenced the charging of an electric double layer composed mainly of accumulated Li+ solvated by EC molecules before the SEI formation explaining the SEI mainly composed of products coming from the EC reduction.164
Mechanical properties of the electrodes during cycling
Mechanical and viscoelastic studies are not restricted to the evaluation of the SEI formation. It has also been extended to study the behavior of composite electrodes under battery operation conditions. EQCM-D has the benefit of simultaneous monitoring of dissipation at multiharmonics, over the manual operation of network analyzers at different overtones. Performing experiments of dissipation monitoring by EQCM-D or NAs proved its utility in distinguishing the gravimetric cases (where Sauerbrey equation applies) from those deviate from the “dense and rigid thin layer assumption”. The group of Levi and Aurbach developed the experimental and theoretical background permitting the extraction of the desired features; however, the literature is scarce and almost monopolized. Classical composite electrodes for battery application are multicomponent: the active material, a conductive agent e.g. carbon black and a binder to mechanically hold the different particles. Therefore, deviations from the idyllic “dense and rigid thin layer” condition may occur for the electrodes with different porous geometry and binders, contributing to the f of the resonators in contact with electrolyte by hydrodynamic and viscoelastic effects, respectively. Consequently, in addition to the gravimetric EQCM, two non-gravimetric cases, which involve the hydrodynamic effects (the behavior of porous and rough surface electrode in the electrolyte) and the viscoelastic effects (softening/stiffening of the electrode in the electrolyte) during charge/discharge, enabled the use of EQCM-D as a structural/morphological and mechanical/viscoelastic probe, respectively.
Table of contents :
Acknowledgments
Table of Contents
Broader context and thesis outline
○ From global warming to battery advent
○ Battery market and applications
○ Thesis outline
Chapter I – The battery technology: State of the art
I.1 – Today’s battery to post Li-ion technologies
I.1.a – Concept discoveries and fundamentals
I.1.b – From Li batteries to Li-ion batteries
I.1.c – Post Li-ion battery technologies and their promises
I.2 – In the quest of better batteries
I.2.a – Higher energy materials
I.2.b – Improve the lifetime, the reliability and the power density
I.3 – Chapter conclusions
Chapter II – An overview of the electrogravimetric, hydrodynamic and viscoelastic techniques
II.1 – Introduction
II.2 – Quartz Crystal Microbalance theory and associated techniques
II.2.a – Quartz Crystal Microbalance fundamentals
II.2.b – Gravimetric and viscoelastic read-out methods
II.2.c – Comparison of the different QCM-based read-out methods
II.3 – Electrogravimetric, hydrodynamic and viscoelastic inputs to the battery field so far 55
II.3.a – EQCM-based methods for the study of Solid Electrolyte Interphase
II.3.b – Mechanical properties of the electrodes during cycling
II.3.c – Characterization of the cation insertion in battery compounds
II.4 – EQCM-based strategies developed and employed in this Ph.D. thesis
II.4.a – Classical EQCM-R measurements adapted to battery electrode characterization
II.4.b – Determination of the film hydrodynamic and viscoelastic properties
II.4.c – Dynamic characterization of the electrode-electrolyte interface
II.5 – Chapter conclusions
Chapter III – Making advanced electrogravimetry as an affordable analytical tool for the battery interface characterization
III.1 – Introduction
III.2 – Optimization of the electrode preparation on EQCM resonators
III.2.a – Description of the preparation protocol
III.2.b – Evaluation of the suitable conditions for electrode preparation by EQCM-R
III.3 – Design and validation of airtight EQCM cell testing workbench
III.3.a – Cell presentation
III.3.b – Cell evaluation
III.3.c – Calibration factor estimation
III.4 – Verification of hydrodynamic and viscoelastic properties in electrolyte
III.4.a – Rigidity assessment
III.4.b – In situ hydrodynamic spectroscopy of the composite electrode
III.5 – Electrochemical-gravimetric measurements of an intercalation material
III.5.a – Viscoelastic properties upon cycling
III.5.b – Direct outputs of EQCM-R on an intercalation compounds
III.5.c – New insights on the insertion mechanism at electrode-electrolyte interface
III.6 – Simulation of the interface between the electrolyte and a model electrode
III.6.a – Interfacial density profiles of the different species involved
III.6.b – Energy landscape and solvation from the electrolyte bulk to the interface
III.7 – Chapter conclusions
Chapter IV – Probing the electrode-electrolyte interface of potassium-ion battery – Aqueous vs. Non-aqueous electrolytes
IV.1 – Introduction
IV.2 – Phase preparation and characterization
IV.2.a – Morphological and structural description
IV.2.b – Water incorporation in the structure lattice
IV.2.c – Thickness control of the prepared films
IV.3 – Electrochemical and electrogravimetric analyses
IV.3.a – Rate capability assessment
IV.3.b –Determination of the rate-limiting step
IV.3.c – First evidence of the potassium solvation shell at the interface
IV.4 – Final evidence of the involved species at the EEI and their associated interfacial kinetics
IV.4.a – Identification of the nature of species and their dynamics at EEI
IV.4.b – Study of the interfacial kinetics associated to the involved species
IV.5 – Chapter conclusions
Chapter V – Elucidating the origin of the electrochemical capacity in proton-based battery: a framework to assess the water contribution at the interface
V.1 – Introduction
V.2 – Phase characterization
V.2.a – Morphological and structural description
V.2.b – Stoichiometry determination
V.3 – Electrochemical analyses
V.3.a – Cycling behavior
V.3.b – Assessment of the pseudocapacitive mechanism
V.4 – Electrochemical-gravimetric investigation
V.4.a – Validation of the gravimetric regime
V.4.b – Kinetic comparison between the two regions
V.4.c – Evidences of the water participation in the insertion mechanism
V.5 – Species determination and associated kinetics at the EEI
V.5.a – Identification of the nature of species and their dynamics at EEI
V.5.b – Outcomes of the kinetic study
V.6 – Chapter conclusions
General conclusions
Materials & Methods
M.1 – Material preparation and characterization
M.1.a – Used and synthetized battery materials
M.1.b – Material characterization
M.2 – Classical electrochemical characterization
M.2.a – Electrode preparation
M.2.b – Classical electrochemical setup
M.2.c – Electrochemical cycling
M.2.d – Diffusion coefficient determination
M.3 – Electrochemical-gravimetric analysis
M.3.a – EQCM electrode preparation
M.3.b – Airtight EQCM cell testing workbench
M.3.c – Verification of hydrodynamic and viscoelastic properties in electrolyte
M.3.d – Electrochemical-gravimetric measurements
M.3.e – Ac-electrogravimetry
M.4 – Classical Molecular Dynamics simulations
Glossary
References
