Engineering Strategies to Improve All-Solid-State

Battery Performance under Low-Pressure Conditions

Benjamin Hennequart

To cite this version:

Benjamin Hennequart. Engineering Strategies to Improve All-Solid-State Battery Performance un-
der Low-Pressure Conditions. Material chemistry. Sorbonne Université, 2023. English. �NNT :
2023SORUS638�. �tel-04524445�

HAL Id: tel-04524445

https://theses.hal.science/tel-04524445
Submitted on 28 Mar 2024

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
 Sorbonne Université
École doctorale – ED 397 – Physique et Chimie des Matériaux

Collège de France
Laboratoire – Chaire de Chimie du Solide et Énergie

Engineering Strategies to Improve All-Solid-State

Battery Performance under Low-Pressure Conditions

Benjamin Hennequart
Doctoral thesis in Materials Science / Chemistry

Directed by Jean-Marie Tarascon and Christophe Lethien

President of the jury: Dr. Mathieu Morcrette

Presented and defended in public on December 18th 2023 in front of the jury,

Dr. Mathieu Morcrette Research director, LRCS Referee

Dr. Valérie Pralong Research director, CRISMAT Referee
Dr. Florian Strauss Group Leader, Karlsruhe Institute of Examiner
Technology
Dr. Damien Dambournet Associate professor, PHENIX, Sorbonne Examiner
Université
Dr. Fanny Bardé Founder and CTO, SOLiTHOR Examiner

Prof. Jean-Marie Tarascon Professor, CSE, Collège de France Director

Prof. Christophe Lethien Professor, IEMN, Université de Lille Co-director
 Sorbonne Université
École doctorale – ED 397 – Physique et Chimie des Matériaux

Collège de France
Laboratoire – Chaire de Chimie du Solide et Énergie

Ingénierie pour Améliorer les Performances des

Batteries Tout-Solide sous Faible Pression

Benjamin Hennequart
Thèse de doctorat en Chimie des Matériaux

Dirigée par Jean-Marie Tarascon et Christophe Lethien

Président du jury : Dr. Mathieu Morcrette

Présentée et soutenue publiquement le 18 Décembre 2023 devant un jury composé de,

Dr. Mathieu Morcrette Directeur de recherche, LRCS Rapporteur

Dr. Valérie Pralong Directrice de recherche, CRISMAT Rapportrice
Dr. Florian Strauss Group Leader, Karlsruhe Institute of Examinateur
Technology
Dr. Damien Dambournet Chargé de recherche, PHENIX, Sorbonne Examinateur
Université
Dr. Fanny Bardé Founder and CTO, SOLiTHOR Examinatrice

Prof. Jean-Marie Tarascon Professeur, CSE, Collège de France Directeur

Prof. Christophe Lethien Professeur, IEMN, Université de Lille Co-directeur
Acknowledgements
First and foremost, I would like to express my sincere gratitude to Jean-Marie Tarascon
for welcoming me in his lab. You have given me your trust to carry out my work with a degree
of freedom beyond what I would have hoped. Your guidance and consistent availability have
played a pivotal role in shaping this research journey. Through your passion and hard work,
you have cultivated an enthusiastic and vibrant lab environment for scientific exploration that
has undoubtedly contributed to my growth as a scientist in every regard.

I would also like to thank Christophe Lethien. Although changes in the project's
direction limited your direct involvement in this thesis, you consistently provided me with
constructive, and when relevant, positive feedback on my presentations and publications.

In many regards, this thesis has greatly benefited from the contributions of various
people whom I would like to express my gratitude to for their time and energy. In particular,
special thanks to Michaël Deschamps for his help in carrying out the NMR study, Ronan
Chometon and Juan Forero-Saboya for their invaluable time spent in performing SEM imaging,
Jacques Louis and Gwenaëlle Rousse for their assistance with crystallography and XRD analysis
and Carine Davoisne who helped me acquire SEM imaging of my air-sensitive samples. Thanks
to Thomas Marchandier and his team in Saint-Gobain for providing me with electrolytes. I am
also grateful to Elisa Quemin for her invaluable help with impedance spectroscopy, and I
extend my thanks to Maria Platonova for her assistance with various experiments. Many
thanks to Romain Dugas for the time dedicated not only to designing low-pressure cells but
also in answering my many scientific questions.

Beyond science, these three years spent in the lab were truly enjoyable. I want to
specifically express my heartfelt gratitude to my incredible PhD colleagues and friends from
the solid-state team, Elisa, Ronan and Tuncay. The atmosphere in the team and these long
hours doing experiment followed by our bar sessions would not have been the same without
you. Of course, I also thank my great officemates, Damien and Romain, for the countless
terrible jokes and shared laughter. Thanks to Benjamin for joining me in our numerous movie
sessions, to Jessica for her invaluable help in dealing with all the administrative nightmares. A
special thanks to Elisa Q., Ronan, Tuncay, Benjamin, Damien, Jacques, Thomas, Alexia, Elisa G.,

i
Clémence and to all the former, current, and new lab members whose contributions to the
lab’s atmosphere have made being a part of it truly enjoyable.

I would like to thank Dr Mathieu Morcrette and Dr Valérie Pralong for their time spent
reviewing this thesis and to Dr Florian Strauss, Dr Fanny Bardé and Dr Damien Dambournet
for being part of the jury.

Enfin, bien sûr, je remercie infiniment mes parents, mes frères et ma sœur pour le
soutien qu'ils m'ont apporté tout au long des hauts et des bas de cette thèse, mais aussi une
mention spéciale à Robin, mon cousin, pour ses rires, ses petits plats, pour m’avoir supporté
et soutenu ces trois années, et particulièrement pendant les derniers mois de cette thèse.

ii
Table of contents
Acknowledgements ......................................................................................... i

Table of contents ...........................................................................................iii

Broader context and thesis outline................................................................. 1

Broader context ...................................................................................................................... 2
Thesis outline ......................................................................................................................... 4

Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-

State Batteries ............................................................................................... 7
1.1 – Overview of Energy Storage and Batteries Technologies ............................................. 8
o General scientific background: the working principle of batteries ................................ 9
o The journey towards the state-of-the-art Li-ion batteries and their limitations ......... 10
1.2 – The Resurgence of All-Solid-State Batteries: An Historical Approach on Solid
Electrolyte Development...................................................................................................... 14
o The potential benefits of all-solid-state batteries ........................................................ 14
o From the early inorganic solid conductor discoveries to the first solid lithium-ion
conductors ........................................................................................................................ 16
o The race towards high performing inorganic solid electrolytes: The rise of different
families ............................................................................................................................. 19
1.3 – Persisting Challenges in All-Solid-State Batteries ........................................................ 22
o Chemical and electrochemical interfacial reactivity in ASSBs ...................................... 22
o Formulation, processing and integration ..................................................................... 27
o The necessity of decreasing the stack pressure for Li metal implementation ............. 33
1.4 – Exploring the Cutting-Edge of All-Solid-State Batteries .............................................. 35
o The latest solid electrolyte developments.................................................................... 35
o Emerging progress in full cell ASSB systems ................................................................. 42
o The progress toward low pressure cycling ................................................................... 44
1.5 – Chapter conclusion ...................................................................................................... 46

Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries

enabled by Halide-based Solid Electrolyte .....................................................49
2.1 – Chapter Introduction ................................................................................................... 50
2.2 – The low pressure cycling of argyrodite-based NMC composites ................................ 51
o A high pressure reference ............................................................................................. 51
o Low pressure performance and the importance of formulation ................................. 53

iii
 2.3 – Li3InCl6: A good candidate for low pressure cycling .................................................... 58
o Electrochemical and chemical stability of Li3InCl6 ........................................................ 58
o Performances of Li3InCl6-based cathode composites under various pressures ........... 59
o The importance of the composite preparation process ............................................... 62
2.4 – Introducing Li3YBr2Cl4: a mixed halide SE with enhanced reduction stability and low
hardness ............................................................................................................................... 63
o Physical, chemical and electrochemical properties of Li3YBr2Cl4 SE............................. 63
o The low pressure performance of Li3YBr2Cl4-based composites .................................. 67
o Discussion on the reason underlying the enhanced low-pressure cyclability of halides
over Li6PS5Cl solid electrolyte........................................................................................... 70
o A first step towards larger pouch cells ......................................................................... 73
2.5 – Attempting stable cycling of “full-halide” ASSB .......................................................... 75
o First strategy: coated NMC particles ............................................................................ 76
o Second strategy: LIC and LYBC dual-SE architecture .................................................... 76
2.6 – Chapter conclusion ...................................................................................................... 77

Chapter 3 – Solid-Electrolyte-Free Cathode Concept for Low-Pressure Cycling

of All-Solid-State Batteries ............................................................................81
3.1 – Chapter Introduction ................................................................................................... 82
3.2 – The O1-TiS2 case: a literature benchmark ................................................................... 83
3.3 – O3-LixTiS2: A Pre-lithiated lamellar phase enabling low pressure cycling ................... 85
o Characteristics and performances in typical composite cathodes ............................... 85
o Enabling SE-free cycling of O3-LixTiS2 through a milling strategy ................................. 87
o Improved performance through lithium ion diffusion coefficient enhancement ........ 92
o The low pressure performance of the SE-free electrode configuration ...................... 98
3.4 – Applying the SE-free concept to other CAMs for low pressure cycling .................... 103
o Exploring other titanium-based sulphide polymorphs ............................................... 103
o Leveraging the ductility of halide-based materials..................................................... 104
3.5 – Chapter conclusion .................................................................................................... 108

Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its

Implementation .......................................................................................... 111
4.1 – Chapter Introduction ................................................................................................. 112
4.2 – Implementing Li metal anodes in low pressure systems .......................................... 113
o Applicability to the halide-based composite system .................................................. 114
o Lithium metal anode in the SE-free cathode system.................................................. 115
4.3 – How accurate is the evaluation of the SE//Li interface? ........................................... 118
o Evaluating the critical current density with our electrolytes ..................................... 118
o Parameters influencing the critical current density ................................................... 120

iv
 o How to improve the measure of the CCD? ................................................................. 122
4.4 – Promising strategies and perspectives to enable the use of lithium metal anodes . 124
o Design of solid electrolytes ......................................................................................... 124
o Alloy anodes ................................................................................................................ 125
o Coatings and interlayers ............................................................................................. 126
o In-situ formed and self-healing coating ...................................................................... 127
o Lithium reservoir-free cell configuration .................................................................... 127
4.5 – Chapter conclusion .................................................................................................... 129

General conclusions and perspectives ......................................................... 131

Appendix..................................................................................................... 137
A2 – Supplementary information for Chapter 2 ................................................................ 138
A2.1 – Materials and methods ....................................................................................... 138
A2.2 – Supplementary figures ........................................................................................ 142
A3 – Supplementary information for Chapter 3 ................................................................ 148
A3.1 – Materials and methods ....................................................................................... 148
A3.2 – Supplementary figures ........................................................................................ 153
A3.3 – Supplementary tables ......................................................................................... 156
A4 – Supplementary information for Chapter 4 ................................................................ 157
A4.1 – Materials and methods ....................................................................................... 157
A5 – Additional supplementary information ..................................................................... 159

References .................................................................................................. 163

v
 Broader context and thesis outline

Broader context and thesis outline

1
 Broader context and thesis outline

Broader context
In a world where economic growth is the prime driving force for development, the
scale of global energy consumption has witnessed an unprecedented surge driven by rapid
industrialization and population growth over the past century and this trend is set to persist
as emerging economies develop and mature. Traditional energy sources, predominantly
relying on fossil fuel combustion have historically dominated the energy landscape due to
their abundance, accessibility, and cost-effectiveness. However, the extensive use of fossil fuel
since the industrial revolution in the late 18th century has already led to a worrying increase
in anthropogenic greenhouse gas (GHG) emissions that today are unequivocally linked to the
ongoing climate change and global rise in temperature1 (Figure 1b). In light of such climate
deregulation crisis, addressing global energy demand requires comprehensive and impactful
effort and policies from governments, industries, and societies worldwide to limit GHG
emissions (Figure 1a). By 2020, the global surface temperature had already increased by 1.1°C
and projections from the most recent Intergovernmental Panel on Climate Change (IPCC)
report1 report that limiting our impact on the environment involves a strong necessity in
transitioning to sustainable and low-carbon energy alternatives and in electrifying the
transportation sector. Both of which have seen their price plummet, making them more
accessible and attractive for widespread adoption (Figure 2).

Figure 1. (a) Projected global greenhouse gases emissions depending on followed scenario and trend with the
current implemented policies (pre-COP26, red) and (b) change in global surface temperature observed from
1850 to 2020 (black line) and simulated evolution due to natural drivers (green) or human and natural drivers
(brown). Adapted from IPCC’s AR6 Synthesis report1.

2
 Broader context and thesis outline

Figure 2. (a) Market cost as compared to fossil fuel cost in 2020 (orange band) and (b) market adoption of
photovoltaic (PV), onshore/offshore wind energies and Li-ion battery pack for EVs evolution from 2000 to
2020. Adapted from IPCC’s AR6 Synthesis report1.

Towards this goal, energy storage systems (ESS) are the basis for this transformation.
For more than 100 years already, mechanical EES such as dams and pumped-storage
hydroelectricity facilities have been used to store excess electricity and balance the load on
the electrical grid.2,3 However, in a world shifting towards renewable but intermittent energy
sources and electrified vehicles, energy storage is becoming increasingly important. In recent
years, electrochemical energy storage devices, such as batteries, fuels cells or supercapacitors
have seen important advancements both in terms of performance and cost. Notably,
rechargeable lithium-ion batteries (LIBs) have emerged as the dominant technology in this
sector, offering high energy densities4,5 (up to 350 Wh.kg-1) and competitive prices1
(decreasing from 921 to 137 $2020/kWh between 2010 and 2020, Figure 2a). Already widely
adopted in mobile devices, their significance has grown even more in the automotive industry
where they play a centre role in the electrification of transportation. Moreover, LIBs are

3
 Broader context and thesis outline

finding increasing application in large-scale grid storage systems, further highlighting their
versatility and impact on shaping the future of energy.

Certainly, the emergence of new applications for lithium-ion batteries, especially in

electric vehicles (EVs), brings new requirements. Two key areas that demand improvement
are energy density and safety.6 Increasing the energy density of batteries is essential for
enhancing the driving range of EVs while reducing their weight and size. Additionally, safety is
of paramount importance, given the potential risks of fires associated with high-power
operations of LIBs in EVs.

The aim of this thesis is to contribute to the advancement of practical high-energy all-
solid-state battery technology and it will be presented as follow.

Thesis outline
Mainly, this thesis will specifically focus on exploring concepts that enable the
reduction of operating pressure of solid-state batteries without affecting their overall
performance. This work is divided into five chapters as detailed below:

Chapter 1 introduces the current Li-ion technology and traces the historical
development of solid electrolytes from the early 1970s to the present, leading to the
resurgence of all-solid-state batteries (ASSBs). The chapter then presents the state-of-the-art
ASSBs and discusses their main challenges, emphasizing on the significance of reducing
operating pressure for their practical application.

Chapter 2 concentrates on the design of cathode composites that can operate at

significantly low pressure by employing a novel family of solid electrolytes known as halide-
based SE. The specific chosen solid electrolytes for investigation in this chapter are Li3YBr2Cl4
and Li3InCl6, selected for their favourable mechanical and electrochemical properties with the
former furnished by Saint-Gobain.

Chapter 3 focuses on the use of a possible new concept in all-solid-state battery known
as the solid-electrolyte-free cathode composite, which allows important enhancement in the
energy density of cathode active material owing to their mixed ionic-electronic conductivity
properties.

4
 Broader context and thesis outline

Chapter 4 is dedicated to the implementation of lithium metal anodes into the systems
discussed in the previous chapters. It is followed with a brief discussion on the controversy on
the technique used for the study of the lithium/SE interface called Critical Current Density
(CCD) determination, which appears to have limitations and could potentially benefit from
improvements to enhance its reliability.

To conclude, the last part provides a comprehensive summary of the findings from this
thesis and offers valuable insights and guidance for future research to build upon the work
presented in this study.

5
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

Chapter 1 – Today’s Battery

Technologies and the Resurgence
of All-Solid-State Batteries

7
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

1.1 – Overview of Energy Storage and Batteries Technologies

In the history of battery technologies, significant breakthroughs led to the
development of modern batteries. Luigi Galvani's first observations of muscular contractions
induced by electrical currents in a frog's leg in 1780 marked an important milestone. He
noticed that when two different metal came simultaneously into contact with a dead frog's
leg, an electrical current flowed between them, causing the frog's leg to twitch; he called this
phenomenon “animal electricity”. Building on Galvani’s discovery, a few years later, in 1800,
Alessandro Volta's invention of the Voltaic pile revolutionized the field by introducing the first
chemical battery (Figure 1.1). The Voltaic pile is a stack of alternated bilayer of zinc and copper
disks, separated by brine-soaked paper disks that demonstrated the production of a steady
current.

Figure 1.1. Voltaic pile on display at the Tempio Voltiano museum in Como, Italy. Picture by I, GuidoB,
reproduced from reference7.

With the combination of Galvani's observations and Volta's invention, the

understanding of electricity and battery principles began to take shape, and the development
of batteries progressed rapidly over the subsequent century. Notably, we can mention the
work of Georges Leclanché and Carl Gassner on “dry cell” design between 1868 and 1888,
which is the foundation of today’s primary alkaline cells, or the development of the lead-acid
batteries in 1859 by Gaston Planté, which revolutionized battery technology as the first

8
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

rechargeable battery. These innovations paved the way for modern batteries and
electrochemical devices we rely on today.

o General scientific background: the working principle of batteries

Electrochemical storage devices consist in several essential components: a positive and

a negative electrode (typically called cathode and anode respectively for batteries), where
chemical energy is stored. They are insulated by an ion-conducting electrolyte that facilitates
ion movement and connected via an external circuit, enabling the flow of electrons outside of
the cell and the harnessing of electrical energy (Figure 1.2).

Figure 1.2. Schematic of a battery. Adapted from reference8.

Among these devices, we can differentiate two classes: the primary and secondary
battery systems. In brief, primary cells, characterised by non-reversible chemical reactions,
are incapable of being recharged after depletion. In contrast, secondary batteries, featuring
reversible reactions, are rechargeable. In secondary systems, the electrochemical reactions
taking place at the electrodes are responsible for energy storage or release (charge or
discharge) depending on the direction of the flow of electron. Battery performance is
determined by various metrics, such as energy density, power density, calendar life, safety,
cost or environmental impact, to name a few. Among these, improvement in energy density

9
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

and specific energy has been the primary driving force behind technological progress in the
field. Also known as volumetric and gravimetric energy respectively, they refers to the amount
of energy stored in a battery and are defined as the product of the cell voltage (U) by the
capacity (Q) of a particular system (Equation 1.1), expressed per unit of volume (Wh.L-1) or
mass (Wh.kg-1).
𝐸(𝑊ℎ. 𝑘𝑔−1 ) = 𝑄(𝑚𝐴ℎ. 𝑔−1 ) ∙ 𝑈(𝑉) Equation 1.1
The average cell voltage (U) is determined by the averaged difference in potential of
the two electrodes of the cell while the capacity (Q) is the total amount of charges
transported, typically expressed as amps per unit of mass of electrochemically active material
(mAh/g) and calculated with the following Equation 1.2.
𝑛𝐹
𝑄(𝑚𝐴ℎ. 𝑔−1 ) = Equation 1.2
3.6 ∙ 𝑀
where n is the number of electron involved in the reaction, F the Faraday constant
(𝐹 = 𝑒𝑁𝐴 ≈ 96485 C/mol) and M the molecular mass of electrode material.

o The journey towards the state-of-the-art Li-ion batteries and their limitations

Early on, in the pursuit of enhancing battery performance, lithium-based systems were
deemed highly promising owing to the low electrochemical potential (-3.04 V vs. SHE),
lightweight nature (6.94 g/mol) and high capacity (3860 mAh/g) of lithium, thereby enhancing
cell voltage and specific capacity. From the 1970s, Li-based battery started being developed
with the discovery of multiple lithium intercalation compounds. Intercalation chemistry, a
term first introduced by Rouxel9,10 in 1971 and discussed for application in electrochemical
cells by Steele and Armand10,11 in 1972, serves as the fundamental mechanism for most of the
current electrode material. It refers to a host/guest solid-state redox reaction involving the
reversible insertion and extraction of guest ions (Li+, Na+, etc.) into and from the crystal lattice
of a host material during the discharge and charge processes of a battery. These insertion and
extraction of ions occur without causing significant structural damage (eg. bond breaking) to
the host material resulting in a high level of reversibility and an extended calendar life. In 1975,
Whittimgham laid the foundation of Li-based battery systems with the first use in a battery of
a Li+-intercalation compound, the layered dichalcogenides TiS2. This material having a
reasonably high lithium potential (around 1.8 V vs. Li+/Li) and allowing for reversible insertion

10
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

and extraction of Li ions was promising for use as a cathode material in Li-based batteries
coupled to a Li-metal anode, following the reactions below.

(𝑎𝑡 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒 𝑒𝑙𝑒𝑐𝑡𝑟𝑜𝑑𝑒) TiS2 + Li+ + e− ⇄ LiTiS2 Equation 1.3

(𝑎𝑡 𝑛𝑒𝑔𝑎𝑡𝑖𝑣𝑒 𝑒𝑙𝑒𝑐𝑡𝑟𝑜𝑑𝑒) Li ⇄ Li+ + e− Equation 1.4

Exxon explored the potential commercialisation of this TiS2/Li metal system while the
similar MoS2/Li metal system was developed and commercialised by Moli Energy in the 1980s.
However, both systems faced challenges, particularly related to safety risks associated with
the use of a lithium metal anode (LMA). During the charging process, the use of a LMA often
leads to uneven plating of lithium on the anode's surface, forming dendrites (Figure 1.3a).
These growing dendrites ultimately cause internal short circuits leading to potential fire
hazards and safety issues.

Figure 1.3. Scheme of (a) a lithium metal based cell with dendrite formation and (b) a typical lithium-ion cell
with a graphite anode. Adapted from reference8.

Despite the setbacks faced by early lithium-based systems, research and development
efforts persevered. In 1980, John B. Goodenough introduced a new transition metal layered
oxide LiCoO2 (LCO), a higher potential oxide cathode compound that held great promise for
battery technology. The pursuit of safer and more efficient battery systems continued, and in
1991, Sony achieved a significant milestone by commercialising the first lithium-ion battery
(LIB). This revolutionary battery, pioneered by Akira Yoshino, incorporated Goodenough's
LiCoO2 cathode and a newly developed carbonaceous anode (Figure 1.3b). The use of carbon-
based intercalation materials for the anode mitigated the safety risks associated with lithium
metal, allowing for the successful commercialisation and widespread adoption of LIBs in

11
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

diverse applications. In recognition of their outstanding contributions on the development of

lithium-ion batteries, M. Stanley Whittingham, John B. Goodenough, and Akira Yoshino were
awarded the Nobel Prize in Chemistry in 2019.

In the subsequent decades following Sony’s battery commercialisation, remarkable

progress was achieved in the development of cathode materials, with notable example being
the NMC/NCA family (LiNi1-x-yMnxCoyO2 and LiNi1-x-yCoxAlyO2). Within this family, the partial
substitution of cobalt in LiCoO2 to other transition metal such as nickel or manganese has led
to significant improvements in specific capacity, structural and thermal stability. This family,
and particularly the NMC622 (LiNi0.6Mn0.2Co0.2O2) and NMC811 (LiNi0.8Mn0.1Co0.1O2)
compositions are now widely utilised in the current generation of Li-ion batteries.12 The next-
generation cathode materials include the Li-rich layered oxides compounds (Li1+xM1-xO2 with
M a transition metal) that show a 25% increase in energy density but still suffer from non-
resolved hurdles such as voltage fade and low energy efficiency.13 However, it is worth
mentioning that, in parallel of the layered oxide, other Co-free compounds have been
developed such as the spinel LiMn2O4 or the polyanionic LiFePO4 (LFP) family phases, thereby
enhancing the battery's specific energy and/or environmental friendliness while
simultaneously addressing the ethical concerns associated with cobalt extraction. In
particular, the latter LFP chemistry have seen an important penetration in today’s market
(Figure 1.4) and a substantial growth is projected for this technology in the coming decades
owing to its use in light electric vehicles.12

100 Other
7% 6%
3% 17%
27% LFP
Light EV battery capacity

80
share by chemistry (%)

60
78% 89% 87%
40 76%
66% High-Ni

20

11% 6% Low-Ni
0 5% 4% 4%
2018 2019 2020 2021 2022
Year
Figure 1.4. Electric light-duty vehicle (EV) battery capacity by chemistry, market share evolution from 2018 to
2022. Low-Ni includes NMC333. High-Ni includes NMC532, NMC622, NMC721, NMC811, NCA and NMCA.
Cathode sales share is based on battery capacity. Adapted from reference12.

12
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

Due to their remarkable energy density14 and decreasing costs1, current Li-ion batteries
have gained widespread adoption in portable electronics and electric vehicles (EVs).1
However, they are now approaching their theoretical limits, with energy densities reaching
approximately 300 to 350 Wh.kg-1, and achieving further significant improvements will
necessitate substantial paradigm shifts. A key constraint lies in the challenges associated with
introducing lithium metal anodes, which have the potential to substantially enhance the
energy density. Although several alternative paths to enhance energy density are currently
under exploration such as anionic redox compounds, Li-sulphur systems or full silicon anodes,
these paths are still in the early stages of research and development. Presently, the most
advanced system, utilising an NMC-type cathode and a Li metal anode, reaches energy density
of up to 575 Wh.kg-1 (Figure 1.5), nearly double that of previous systems but to the expense
of a poor cycle life.15

Figure 1.5. Evolution of the energy density of rechargeable intercalation-type cathode pouch cells in the past
30 years. Green dots highlight cells including a lithium metal anode (or anodeless designs, designated by Cu for
copper current collector). Adapted from reference14.

The pursuit of enhanced energy density by integration of the LMA has revealed the
need to push towards novel battery systems capable of preventing dendrite growth and
ensuring superior safety. Thus, the investigation into all-solid-state battery systems has
emerged as a promising avenue to address these challenges.

13
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

1.2 – The Resurgence of All-Solid-State Batteries: An Historical Approach

on Solid Electrolyte Development
o The potential benefits of all-solid-state batteries

The success of LIBs in powering a diverse range of applications is undeniable.

Nevertheless, with the ever-increasing requirements for increased energy density, improved
safety, and prolonged battery life, researchers have turned their attention to all-solid-state
batteries (ASSBs) as a promising transformative alternative. Unlike conventional LIB cells that
utilise liquid electrolytes as the ionic carriers, ASSBs adopt a solid electrolyte that serves both
as a Li-ion conductor and as a separator, presenting a fundamental shift in their design and
composition that could offer a range of promising benefits.

Notably, by replacing the flammable organic liquid electrolyte to a ceramic solid

electrolyte (SE) (Figure 1.6), ASSBs holds the promise to substantially improve safety by
effectively mitigating fire hazards arising from electrolyte leakage. Moreover, owing to their
solid nature, SEs are anticipated to allow the use of LMAs by effectively create an
impenetrable solid barrier to lithium dendrites, addressing a long-standing challenge in
traditional LIBs. As a result, ASSBs incorporating LMAs are projected to exhibit a substantial
increase in energy density over LIBs. Theoretical estimates propose a potential enhancement
of 70% in volumetric energy density and 40% in gravimetric energy density.16

Figure 1.6. Scheme of a typical lithium-ion battery (right) and a lithium metal all-solid-state battery (left) with
compared volumetric (wvol) and gravimetric (wgrav) energy densities of both systems. CAM stands for cathode
active material. Adapted from reference16.

14
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

ASSB offer another significant advantage in terms kinetics primarily attributed to the
unity transference number (𝑡+ = 1) of such SEs. This unique feature is expected to allow
ASSBs to achieve higher power density compared to their liquid counterparts, which usually
have lower transference numbers due to the presence of mobile anions in solution.

In addition to the aforementioned benefits, this technology offers further advantages,

notably the ability to leverage the wide range of chemistry possibilities of SEs, enabling the
use of higher voltage cathode active material (CAM) and offering a wider temperature range
operation, further contributing to increase energy density and safety, respectively. Another
notable advantage is the effective suppression of the chemical cross talk issues between
electrodes observed in conventional LIBs that arise from the dissolution of species in the liquid
electrolyte (LE) often resulting in self-discharge and battery degradation.

Furthermore, looking at the assembly of the ASSB, the concept of a bipolar stacking
architecture can be considered (Figure 1.7a). It involves the arrangement of cells connected
directly in series within a battery module by placing cathodes and anodes on opposite sides
of a single current collector, the bipolar plate (green plates in Figure 1.7a). In theory, this
design could offer enhanced power capabilities and energy density, as well as reduced costs.17
In addition, owing to the homogeneity of the electron flow through the bipolar plates
compared to regular current collectors (Figure 1.7b and d), it was calculated that the power
capabilities were improved while heat generation upon cycling at high rates reduced.17
However, in practice, its implementation presents significant challenges and it is not clear if
the bipolar architecture would offer significant advantages over the typical parallel stacking
that consists in double-coated current collector cells arranged in parallel and that is already
readily used in LIBs (Figure 1.7c). An in-depth study on the actual benefits of such architecture
over the parallel one is still needed. The main issue of a bipolar system lies in the lack control
of individual cell potential within the stack. As a result, all cells present in the bipolar stack
must be precisely identical in terms of manufacturing and behaviour over cycling. This
uniformity is essential to ensure the proper operation of the entire stack and avoid imbalances
that could potentially compromise the overall efficiency and stability of the system.
Additionally, adapting the manufacturing process of such system remains crucial to prevent
any internal short circuit due to misalignment of the different layers.

15
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

Negative Positive
CC Bipolar plate CC

a. b.
e-

Bipolar cell stack

e-

Li+ e- Li+ e- Li+ e- Li+ e-

e-

– SE + – SE + – SE + – SE +

e-
e- e-

c. d.
Parallel cell stack

e- e-

Li+ Li+ Li+ Li+

e- e- e-

– SE + + SE – – SE + + SE – e-
e-
e-

Figure 1.7. Differences between bipolar and regular parallel cell stacking architecture. Schematics of (a) a
bipolar and (b) a regular parallel monopolar cell architecture with four cells in a pouch cell of identical energy,
and the electron flow difference in the monopolar and bipolar plates. CC stands for current collector.

Given the many advantages mentioned above, ASSBs have received significant
attention in recent years. To understand the development and progress of SEs, it is essential
to explore their historical evolution, which highlights the evolution of the materials and
technologies used in this field.

o From the early inorganic solid conductor discoveries to the first solid lithium-
ion conductors

The field of solid-state ionics and the study of ionic conductors can be traced back to
the early 19th century when scientific knowledge was far from what it is today. The laws of
thermodynamics had not yet been formulated, and the periodic structure of crystals was still
unknown. Concepts such as point disorder and entropy were not yet developed, and the very
existence of atoms and ions was still disputed. It is in that historical context that in just a few
years, between 1831 and 1834, Michael Faraday discovered the motion of mobile ions in both
liquid and solid thus laying the foundation of electrochemistry and solid-state ionics. In 1834,
Faraday's observation of remarkable conduction properties in heated solid Ag2S and PbF2
marks the earliest recorded instance of a transition from poorly conducting to conducting
state in an ionically conducting material, which are now recognised as solid electrolytes.10

16
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

Figure 1.8. Michael Faraday in 1860 (left) and ionic conductivity and heat evolution with temperature of PbF2
(right). Reproduced from references10,18,19.

Over the following century, progress in the study of ionic conductors was relatively
slow, marked by a few notable discoveries. One of these was made by Nernst in 1900, who
described the high conductivity observed in yttria-stabilized zirconia10,20, which was later
confirmed to be an oxygen-ion conductor10,21 During this period, other highly conductive silver
ion conductors were also studied, with notably the silver halides.22 However, it was not until
the late 1960s that a significant turning point was observed in the field of solid-state ionics
when two remarkable fast-ion conductors, rubidium silver iodide (RbAg4I5) and β-alumina
were discovered. The introduction of RbAg4I523 brought promising prospects with its high
silver ion conductivity at room temperature (RT), while β-alumina M1+δAl11O17, unveiled by
Yao and Kummer in 1967, was the first ambient temperature fast ionic conductor, other than
silver, to be discovered. Notably, β-alumina stands out amongst solid conductors,
demonstrating high conductivity for various ionic species such as monovalent Li+, Na+, K+, H3O+
or divalent Mg2+, Ca2+ cations.24 Upon these discoveries, batteries incorporating such
electrolytes into full solid-state cells were developed; however, they encountered challenges
stemming from the low energy density (about 5 Wh.kg−1) in RbAg4I5 cells and the high
brittleness of β-alumina. These obstacles resulted in the gradual discontinuation of their
utilisation without attaining the threshold of commercial viability.

Simultaneously, other lower conducting SE were studied, in particular Li+ conductors

such as lithium nitride25 (Li3N) with its relatively good ionic conductivity (10-4 S/cm) or lithium
iodide (LiI). First studied in the 1930s26 and although exhibiting a low Li+ ion conductivity of

17
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

only about 10-7 S/cm, the latter is of particular importance since it got successfully employed
in a primary non-rechargeable cell for cardiac pacemaker in 1972 in Italy, when the first
lithium-battery-powered (Li|LiI|I2) device was implanted into a human being.27,28 To this day,
several millions of people have benefited from implantable devices powered by diverse
iterations of the Li/I2 battery design.

Figure 1.9. Image of a modern implantable cardiac pacemaker powered by a Li/LiI/I 2 primary cell (left) and
cross view of the device with the battery (right). Reproduced from references29,30.

In the following decades, the exploration of solid conductors for Na+ and Li+ ions led to
significant breakthroughs. Particularly noteworthy was Goodenough’s discovery31 of the fast
sodium ion conduction in NASICON (Sodium SuperIonic CONductor)-type compounds
(Na1+xZr2SixP3-xO12, 0 ≤ x ≤ 3) in 1976. Drawing inspiration from this, in 1986, Subramanian et
al. explored Li-conduction within several similar NASICON-type structure. This pursuit led
them to the identification of the conduction in LiTi2(PO4)3 (2 · 10-6 S/cm at RT) which could
further be improved by partial substitution of Ti by Al to give rise to the compound
Li1.3Al0.3Ti1.7(PO4)3 (LATP) exhibiting a conductivity of 7 · 10-4 S/cm at RT. Concurrently during
this period, other types of SE were under examination and it is within this timeframe that a
separation into different families characterised by diverse compositions, ionic and electronic
conductivities, and mechanical properties, became apparent. Today, we separate these
ceramic SE into mainly three different families: oxide, sulphide and halide SEs. It is worth
highlighting that polymer-based and hybrid polymer/ceramic SE have also been subjects of
investigation since the 1970s;32 however, they do not fall within the scope of this thesis and
will not be addressed hereafter.

Altogether, the amalgam of these historical discoveries opens the way for high
performance SEs, a journey explored in the next section.

18
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

o The race towards high performing inorganic solid electrolytes: The rise of
different families

After these initial advances the research on ionic conductors has continued but at a
slower pace due to the rapid development of liquid Li-ion batteries that took the forefront
and mobilised most of the research efforts.

In the quest for better SE, several key requirements must be met to realise a functional
battery. Among others, a robust solid electrolyte must have a high ionic conductivity
(> 10-3 S/cm at RT), possess a broad electrochemical stability window (preferably ranging from
0 to 5 V vs. Li+/Li), demonstrate sufficient chemical stability with moisture, cathode material
and lithium metal, while also being easy to process. In the decades following the early
discoveries, the main driving force behind SE developments has been the enhancement of
ionic conductivity. In that purpose, several new types of solid electrolytes were then
developed.
SULPHIDES

HALIDES
OXIDES

Liquid electrolyte (10-2)

2011

2021
2013

2022
2008

2019
2007

2018
1993
2000

1992
1990
1981
1976
1976
1967
1957
1930s

Figure 1.10. Chronological development of the three main families of inorganic solid electrolytes towards
superionic Li+ conductivity. Ionic conductivities, denoted in parentheses, are expressed in S/cm.

19
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

The oxide solid electrolyte family was further explored after the NASICON and LATP
works from the 70’s and 90’s. Noteworthy are the perovskite-type Li3xLa2/3-xTiO3 (LLTO),
discovered in 1993, which exhibits an ionic conductivity of around 10-5-10-4 S/cm33,34 and the
garnet-type Li7La3Zr2O12 (LLZO), described in 2007 by Weppner’s group35, showing a
conductivity of 7 · 10-4 S/cm (at 25°C) which was further improved36 in 2013 to > 10-3 S/cm (at
RT) by partial substitution of Zr by Al, Ga, Ta, etc. These oxide SEs often demonstrate good
stability with moisture, exhibit high thermal stability and have wide electrochemical potential
stability window, making them ideal for coupling with high potential CAM. However, although
presenting some advantages, due to their high mechanical rigidity, these oxides are difficult
to process and integration into ASSB cells remains challenging. Moreover, their conductivities
are still lacking behind the ones of their liquid counterparts.

Sulphide solid electrolytes (SSEs) have been under investigation since 1981 when
research began with the substitution of oxygen with sulphur in the glassy Li2O-P2O5 thus
resulting in the formation of the sulphide glass Li2S-P2S5.37,38 This compound and its derivatives
were found to exhibit conductivities in the range of 10-4 S/cm at RT. Throughout the following
decade, no notable improvements were observed until the year 2000, when Kanno’s group
introduced a new family of SSEs, the thio-LiSICON (Lithium SuperIonic CONductors)
Li4-xGe1-xPxS4 (0 < x < 1) family39, which exhibited conductivities of 10-3 S/cm, and it is in 2011
that the same research group made another significant contribution. Unveiling the discovery
of a new SE Li10GeP2S12 (LGPS)40, the first capable of reaching high ionic conductivities
(1.2 · 10-2 S/cm) which are comparable to those of their liquid counterpart (typically around
10-2 S/cm). This discovery marked the rebirth of the ASSB field, which had been suffering from
limited advancements. From then on, the race towards the best performance SSE’s
compounds starts. Other families of SSE were developed over the years. In 2016, Kanno’s
group unveiled another highly conducting SE the Li9.54Si1.74P1.44S11.7Cl0.3, with a LGPS-type
structure, it displayed an exceptionally high conductivity of 25 mS/cm.41 In addition from 2008
onwards, the argyrodite-type SSEs (Li6PS5X, X = Cl, Br, I) were also developed in parallel.42 To
date, this class of SSEs remains the most widely used and studied, with Nazar’s group achieving
the highest conductivity with the Li5.5PS4.5Cl1.5 argyrodite composition. SSEs show high ionic
conductivity and excellent mechanical ductility, allowing for simple processing through

20
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

cold-pressing. However, they suffer from poor chemical and electrochemical stability, as well
as sensitivity to moisture which results in the release of toxic H2S gas.38,43

The third SE class is the halide-based SE. They have been studied since the 1930s,
primarily with the lithium halides (LiX with X = F, Cl, Br, I)27,44 and later in 1976 by Weppner
and Huggins with their work on LiAlCl4, however they were still poorly conductive (10-7-
10-6 S/cm). It is only recently, in 2018, that they have regain much interest with the discoveries
of the Li3YCl6 and Li3YBr6 by Asano and co-workers, which exhibited excellent RT ionic
conductivities of 5.1 · 10-4 and 1.7 · 10-3 S/cm, respectively.45 Since then, the development of
new metal-halide SE, now generally characterised by the formula LiaMXb (where M represents
a metal cation such as In, Zr, Y or Sc, and X a halide), has taken off with the introduction in
recent years of a series of high performing halide SEs.44,46–54 A case in point is the study on the
mixed halide Li3YBr3Cl3, which exhibited an exceptional conductivity of 7.2 · 10-3 S/cm after
hot-pressing at 170°C46 as well as of Li2In1/3Sc1/3Cl4 which revealed an excellent stability with
high potential CAM as demonstrated by Nazar et al.54 Compared to oxide or sulphide SEs,
halide SEs combine the merits of both of their advantages. Notably, they display a broader
electrochemical stability window (up to 6.71 V vs. Li/Li+)55 and a better chemical stability
toward CAM, traits often associated with oxides, while also offering high ionic conductivity at
RT and remarkable deformability, characteristics reminiscent of sulphide-based SE. However,
they still lack in stability towards both moisture and low potential anodes such as Li metal.

Overall, when designing an ASSB, the selection of the SE plays an essential role. This
choice significantly influences the battery's performance by impacting chemical compositions,
compatibility with the CAM and Li, processability, and overall cost. The main distinctions are
outlined in Figure 1.11 below.

Oxides Sulphides Halides

(Chlorides and Fluorides)
High-V stability High-V stability High-V stability

Li Air stability Li Air stability Li Air stability

Li stability Processability Li stability Processability Li stability Processability

Figure 1.11. Key metrics comparison for the three main solid electrolyte families. Adapted from reference56.

21
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

Expanding upon these three core families, other types have been studied in the past
few years. Particularly notable is the emerging lithium-metal-oxyhalide family, which has
introduced some promising high-conductivity compounds. Examples include the glassy Li3OCl
phase, reaching an ionic conductivity of 25 mS/cm at 25°C, or the LiTaOCl4 and LiNbOCl4 (12.4
and 10.4 mS/cm, respectively, at 25°C).57–61 Additionally, nitride-based SE like Li3N and Li5NCl2,
despite exhibiting low conductivities, are still under investigation as they hold the promise of
a stable interface with Li metal, a property lacking in most high conductivity SEs.62,63

1.3 – Persisting Challenges in All-Solid-State Batteries

Given the evolving landscape of SEs and the diversity within the existing families being
discussed, it is essential to understand the remaining challenges associated with designing
ASSBs. Indeed, although being expected to present some transformative advantages
compared to liquid-based electrolytes, the use of SEs have proven difficult on multiple aspects.

o Chemical and electrochemical interfacial reactivity in ASSBs

Since ASSBs are made entirely of solid materials, most of the issues currently
encountered arise from the various Solid-Solid interfaces within the cell. Figure 1.12 illustrates
the main challenges pertaining to ASSB design, which are mainly divided between the cathode
and anode interfaces.

Stack pressure
Cathode interface

Coatings

CEI

CAM
Tortuosity volume
change
Contacts
Cathode composite

Moisture/air
stability

SE separator

Contacts Porosity

Dendrites
Pore Lithium metal
formation Solid electrolyte
SEI
separator

Anode interface
Figure 1.12. ASSB cell illustration and main challenges associated to their development.

22
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

Similar to liquid systems, the high potential of advanced CAMs or the strong reduction
capability of lithium metal raises stability concerns at electrolyte/electrode interfaces. Three
types of interfaces can be formed; they are listed below and illustrated in Figure 1.13:

1. Thermodynamically Stable Interface: This type of interface is inherently stable and

experiences no chemical reactions between the electrolyte and electrode. Typical
examples are the nitride-based SE/Li0 interfaces.64

2. Non-Passivated Mixed-Conductive Interphase (MCI): In this case, the interface is

unstable and a mixed ionic/electronic interphase is formed, thus resulting in the
continuous growth of the interphase until short-circuit.

3. Passivated Kinetically Stable Interphase (SEI): Formed by chemical reactions at the

electrolyte-electrode interface, this interphase is electronically insulative while ideally
being ionically conductive. Therefore, it is self-limiting and passivates the surface
preventing further decomposition. Typically observed for most sulphide SEs in contact
with CAM or Li0 and recently reported for some F-based halide SEs/Li0 interface.48,50,65

Figure 1.13. Illustration of the three types of interfaces that can be formed at the AM/SE interface. Reproduced
from reference66.

Such interphases are formed at the lithium metal anode (LMA) surface and within
electrode composites at interfaces between SE and e--conductive surfaces such as carbon
additives or CAMs. Both the kinetically stable (SEI) and thermodynamically stable interfaces
are beneficial for long-term battery performance. The thermodynamic stability is determined
by the electrochemical potential stability window of each electrolyte and corresponds to the
voltage range within which the SE remains stable. It can be computationally calculated by
density functional theory (DFT) or determined experimentally via cyclic voltammetry (CV)
measurements (Figure 1.14a). Although taking into account the kinetic dimension of such

23
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

reactivity, the latter often leads to overestimation of the stability window. This is primarily
due to the challenge of precisely determining the current threshold at which the system is
considered unstable. Furthermore, an insufficient electronic surface in contact with the SE can
also contribute to overpotential of the reactivity, thereby resulting in an overestimation of the
stability window. The electrochemical potential stability windows of typical SEs used in ASSB
are given in Figure 1.14b.

Figure 1.14. Electrochemical stability window of solid electrolytes. (a) Cyclic voltammetry measurement on a
LPSCl+C/LPSCl/Li cell in the voltage range 0-4.2 V vs. Li/Li+. Reproduced from reference67. (b) Calculated
electrochemical potential window of typical solid electrolytes. Hashed regions correspond to SE that are unstable
and form a mixed conducting interface (MCI) upon contact with lithium metal. Adapted from reference68.

Hence, it is critical to consider this intrinsic stability window when pairing SE with
electrode materials. Since, if upon cycling the AM potential goes beyond the SE stability
window, an interphase will form. For instance, Li6PS5Cl, with its narrow thermodynamic
stability window, displays instabilities at both high and low potential. This results in the
formation of insulative species (left phase equilibrium in Figure 1.15), as per the proposed
mechanism67 below, ultimately giving rise to a self-limiting SEI at both the CAM and Li metal
interfaces.

1 5
In oxidation: 𝐿𝑖6 𝑃𝑆5 𝐶𝑙 → 𝐿𝑖𝐶𝑙 + 2 𝑃2 𝑆5 + 2 𝑆 + 5𝐿𝑖 + + 5𝑒 −

In reduction: 𝐿𝑖6 𝑃𝑆5 𝐶𝑙 + 8𝐿𝑖 + + 8𝑒 − → 𝐿𝑖𝐶𝑙 + 𝐿𝑖3 𝑃 + 5𝐿𝑖2 𝑆

On the other end, halide-based SEs, such as Li3InCl6 or Li3YCl6, usually exhibiting high
oxidation limits and are well suited for high-potential CAM. However, due to their limited
reduction stability and the presence of the central metal cation that can be reduced from Mn+
to the electronically conductive M0, these SEs decompose at low potential into species that
exhibit both ionic and electronic conductivities (as depicted in Figure 1.15). Consequently, this

24
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

decomposition gives rise to a mixed-conductive interphase (MCI) which prevents their use in
direct contact with low-potential anode materials. Such MCI formation is also observed with
other SEs such as LGPS, LLTO or LATP due to their metal cations Ge4+ and Ti+4 that can easily
be reduced.69–71

Figure 1.15. Calculated phase equilibria for (left to right) Li6PS5Cl, Li3YCl6 and Li3ScCl6 and potential of Li metal,
LiIn alloy and high potential CAMs. Calculations based on first-principle calculations (DFT). Stable potential
window of SEs are coloured and electronically conductive species are in red. Adapted from references 48,56,72.

Apart from the intrinsic thermodynamic reactivity of SEs, they also can experience
chemical and electrochemical reactivity against the CAMs themselves when combine into the
cathode composite. Such reactivity is dependent on many factors such as chemical
compositions and their compatibility, state-of-charge (SOC), surface species and so on. Using
techniques such as time-of-flight secondary-ion mass spectrometry (ToF-SIMS) and X-ray
photoelectron spectroscopy (XPS), Janek et al. demonstrated that sulphide-based SEs readily
undergo reactions with transition metal oxides at high potentials. In particular, their study
unveiled that argyrodite-type Li6PS5Cl or β-Li3PS4 SEs react with NMC622, resulting in the
formation of sulphate (SOx) and phosphate (POx) species at the interface, as illustrated in
Figure 1.16 from their work.43,73 This phenomenon was well correlated with a decrease in
battery performance. Thus showing how critical interfaces and their understanding are in
ASSBs.

25
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

Figure 1.16. Electrochemical reactivity of argyrodite solid electrolyte and NMC materials. (a) Galvanostatic
cycling of 1st, 2nd and 100th cycles and (b) capacity retention of a NMC622:LPSCl/LPSCl/LiIn cell. (c) Local fragment
distribution of NiO2– (blue) and POx– (green) for the reference SE and the uncycled and cycled composite cathode.
(d) Three-dimensional reconstruction of the depth profile of the cycled composite cathode. Reproduced from
reference43.

Lastly, SEs can also exhibit reactivity towards their surroundings, typically in response
to air or moisture exposure. Oxide SEs, in particular, tend to demonstrate the highest stability
in this aspect. Conversely, sulphide or halide SEs are more reactive with moist air, resulting in
a decline in ionic conductivity due to structural deterioration and SE decomposition.74–76
Moreover, in the case of sulphide SEs, this reactivity results in the release of toxic H2S gas,
which is a major concern for large-scale production of SE or ASSB.

Altogether, it is clear that understanding the formation and evolution of interfaces in

ASSBs is of crucial importance as it can lead to severe cell performance degradation. However,
while decomposition reactions pose challenges, they can hopefully be mitigated through
chemical design strategies of the interface or interphase. Coatings, widely used in typical
liquid-based LIBs, have emerged as a promising approach to tackle such decomposition issues
in ASSBs. Recent efforts have focused on applying coatings to the CAM to either prevent or at
least slow down decomposition reactions. Notably, coatings like LiNbO3 or Li2ZrO3 have
demonstrated significant success in recent years.77 In addition, coatings on SE has also been
studied mainly as a mean of preventing the environmental reactivity due to air and moisture

26
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

that lead to SE decomposition and loss of ionic conductivity.74,78,79 The straightforward coating
of argyrodite LPSCl SE with Li2CO3, achieved through a CO2 gas phase approach was shown to
greatly improve the moisture stability of the SE as well as its chemical stability towards
NMCs.74

o Formulation, processing and integration

In addition to the aforementioned interfacial reactivity problems, the formulation,

processing and integration of the different components into an ASSB pose significant hurdles.
Among these components, the cathode composite stands out as particularly challenging.

A cathode composite consists of various materials, including the AM, SE, conductive
additives and potentially a polymer binder. This assortment of materials introduces a high
degree of complexity due to their distinct chemical, physical and mechanical properties. It is
therefore extremely important to know the microstructure of these composite in order to
optimise their formulation and processing. An effective composite is characterised by a range
of metrics, including achieving high ionic and electronic percolation, maintaining low
tortuosity, minimising porosity, achieving high AM loading while minimising the amount of
electrochemically inactive components (SE, conductive or polymer additives). Collectively,
these factors should result in high power and energy density ASSBs, while maintaining loading
levels that are suitable for industrial-scale production, therefore enabling competitive
performance compared to conventional lithium-ion batteries (LIBs).

In pursuit of these objectives, the necessity of optimising the individual components

within the composite and their intricate interactions becomes evident. While achieving the
highest SE ionic conductivity is clearly of paramount importance, optimising the particle size
and morphology of the different components in the composite has been shown to also be of
critical importance in insuring high AM utilisation as well as good rate performances.80,81 In
particular the AM/SE particle size ratio significantly influences percolation, and various
theoretical investigations have explored its implications on the energy density of composites
(Figure 1.17).80,82,83

27
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

Figure 1.17. Visualisation of modelled NMC/SE composites and percolation through the electrode thickness
depending on AM/SE particle size ratio and NMC loading. On the right panels, the solid and transparent grey
particles represent the percolated/active NMC particle and inactive NMC respectively. Adapted from reference 80.

In addition to its size, the morphology of the active material (AM) can significantly
impact cell performance. The material may be monolithic or polylithic, which involves an
assembly of small primary particles with varying orientations. Such morphological variations
can alter the mechanical integrity of the particles, leading to crack formation and isolated AM
regions within the cell as a consequence of volume changes and stress accumulation during
cycling (Figure 1.18).

Figure 1.18. SEM images of pristine and cycled NMC as either polylithic or monolithic particles. (a, c, e, g) SEM
images and (b, d, f, h) cross-sectional FIB-SEM images collected from as-prepared (a, b, e, f) and cycled (c, d, g,
h) NMC811 composite cathodes prepared with (a−d) Polylithic-NMC811 and (e-h) Monolithic-NMC811.
Reproduced from reference84.

28
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

Moreover, ensuring close physical contact between AM and SE is crucial to facilitate

ion transport and maximise the AM utilisation, especially for high loading composite cathodes.
The state of interfacial contact is influenced by the properties of the SEs, with a focus on their
mechanical attributes. Specifically, the hardness and elasticity of the SE play a significant role
in composite formulation and processing, impacting the overall quality of the resulting
interface.85 A SE with excessively high rigidity, like oxide SEs, requires a high-temperature
sintering process to establish effective particle contacts within the composite. Conversely,
materials like sulphide or halide SEs exhibit greater ductility85–87, allowing for efficient contact
through straightforward cold pressing. Thus reducing interfacial resistance and simplifying the
process of preparing and assembling ASSB. Additionally, it is important to note that in general,
ductile SEs tend to outperform brittle ones when it comes to withstanding the strains and
stresses generated at the AM/SE interface during cycling due to AM volume change which can
be as high as 7% for NMCs (Figure 1.19).85,88 As a result, micro-cracks or fractures are likely to
form, and utilising the ductile property of some SEs can lead to a more closely integrated
interfacial contact.89,90 Maintaining an intimate interface between SEs and electrodes is of
paramount importance and is usually addressed by subjecting the cell to a significant external
pressure.91 This point will be discussed in more detail in the next section.

Figure 1.19. Volume change of CAM with state-of-charge. (a) The unit cell volume of various CAMs as a function
of the state of lithiation obtained from crystallographic data versus and (b) False colour scanning electron
microscopic image showing the contact loss between NCM (grey particles) and a thiophosphate SE (yellow
particles). Adapted from references92,93.

Overall, it is clear that addressing transport limitations in thick electrodes, which stems
from the discussed percolation and mechanical problems, need to be addressed in order to
achieve performance similar to conventional LIBs. To achieve this, it may be necessary to

29
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

develop and adapt advanced manufacturing techniques that allow precise control over the
electrode microstructure.

When considering the fabrication of such cathode composite, the primary objective is
to achieve a high degree of homogeneity within the composite structure. To realise this,
various techniques are employed, they are usually separated into two types: the dry and wet
processing. Among the former, the simplest method, often used in laboratory-scale research,
is the manual hand grinding approach, utilising a mortar and pestle. However, this method
lacks repeatability due to its dependence on the operator's actions, applied force, and grinding
duration. Additionally, it is not a scalable solution and as an alternative, ball-milling techniques
are commonly adopted. These techniques offer increased homogeneity and repeatability
across different batches, with enhanced overall control. Numerous studies have attempted to
determine the best conditions for preparing composites using this technique.94 In recent
developments, a mechano-fusion technique was reported as a scalable process for producing
composites with high AM loading by improving the contact between the AM and the SE
(Figure 1.20).94

Figure 1.20. Schematic illustration of (a) a typical NCM composite cathode obtained by hand grinding or ball
milling and (b) an NCM@LPSCl composite cathode obtained through a mechano-fusion technique. Reproduced
from reference94.

Various other composite fabrication techniques are available, including wet processes
that employ slurry-based methods to achieve high component homogeneity. These
techniques offer the advantage of using technologies already established in the fabrication of
conventional liquid LIBs, rendering them more scalable. However, adopting solvent-based
approaches raises concerns about the stability of the different components, especially the SE,

30
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

when in contact with such solvents.95–97 Alternatively, SE dissolved in solution can be

infiltrated into porous electrodes prepared by conventional slurry processing, thus filling the
pore with ionically conductive material. The obtained composite electrodes are then
subjected to subsequent solvent evaporation, heat treatment, and cold pressing.98–100

As far as ASSB assembly is concerned, several challenges also emerge at this phase. At
the laboratory scale, ASSBs are constructed as pellets by sequentially compacting the SE
separator and the electrode materials (Figure 1.21a). However, this procedure, which leads to
the creation of thick ASSB pellets with limited energy density, is complex, time-consuming,
and therefore costly to implement. A more practical approach for upscaling ASSBs would
involve the production of sheet-type electrodes (Figure 1.21b-c) thus also replicating
conventional LIB electrodes preparation. To insure the mechanical cohesion of sheet-type
electrode, such process often involves the use of a polymer as binder. Its incorporation can
take two different routes: the slurry-based, where the polymer is dissolve in solution and
mixed with the other components, or the dry route that can involve either extrusion, chemical
or physical deposition techniques for example. The solvent route in the electrode fabrication,
although introducing more complexity, has already been successfully used in small scale
pouch-type cells.96,101

Figure 1.21. Schematic illustration of cell fabrication as: (a) a pellet, (b) a cathode-supported two-layer, and (c)
a sheet-type individual three-layer cell. Reproduced from reference97.

31
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

Moreover, when scaling up from lab scale (< 1 cm²) to pouch cell size cell (tens of cm²),
additional considerations regarding pressure are necessary.

Firstly, upon assembly, a high pressure (> 300 MPa) is often applied to densify the
various components and establish contact between particles. For small cells, this can easily be
achieved through uniaxial pressure (Figure 1.22a, middle). However, as the cell size increases,
utilising the same protocol becomes impractical due to the immense forces required to
achieve such pressure. In that context, other densifying techniques need to be considered.
Calendaring (Figure 1.22a on the left), readily used in typical electrode preparation for LIBs
can be considered; although it can sometimes have limitations in effectively eliminating pores
from the SE and the electrode layers. Another more recent technique is the cold/warm
isostatic press process (Figure 1.22a right), which consists in the application of a uniform
pressure in all directions, effectively reducing voids within the composite to a significant
extent. However, the practical implementation of this method on a larger scale is constrained
by its cost and processing time.

Figure 1.22. (a) Schematic of the three major strategies to achieve densification of pouch-type cells ASSBs. (b)
Schematic of an ASSB cell module with hardware to apply stack pressure, and cell to module conversion
efficiencies vs. stack pressure based on a system of (c) 1 and (d) 20 kWh. Adapted from reference102.

Secondly, during operation, the stack pressure and its homogeneity throughout the
cell surface are also of paramount importance as applying pressure allows the mitigation of
contact loss due to volume changes. In the literature, the electrochemical performances of
ASSBs are frequently assessed at high external pressure (often exceeding 50 MPa) to ensure
close contact between the electrode particles. However, similar to the assembly procedure
discussed above, applying such pressure for operation is not viable, as it would require
excessive forces. This would necessitate the use of additional hardware (thick rigid module
housing, strong spring, etc. as shown in Figure 1.22b) which would greatly imped the energy

32
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

density at the module/pack level (Figure 1.22c-d). Thus the need to introduce better
processing to decrease the necessity for such high pressures.

o The necessity of decreasing the stack pressure for Li metal implementation

The potential benefit of ASSB in increasing the energy density of the cell is possible at
the condition of having a high capacity anode, with the lithium metal anode (LMA) standing
out as the optimal candidate. In introducing LMA in ASSBs, several thermodynamic, kinetic,
and morphological requirements need to be fulfilled.103

One of the primary challenges is the low redox potential of lithium metal which poses
complexities in handling and selecting compatible materials for assembling battery cells.
Lithium metal readily reacts with various components of ambient air (N2, O2, H2O, CO2) and
therefore has a native passivation layers (Figure 1.23) containing compounds such as Li3N,
LiOH, Li2O, and Li2CO3 at its surface104 that are forming even in trace amounts within a dry,
inert atmosphere and get trapped between the LMA and the SE separator upon assembly. This
makes it highly complicated to achieve a pristine metallic surface for homogeneous contact
with the SE, even under seemingly ideal inert atmosphere such as an argon-filled glovebox.
The bulk contamination of lithium metal is also challenging with some reporting non-negligible
oxygen solubility or high sodium impurities that can strongly influence the reactivity of the
metal.103,105 Furthermore, the choice of compatible SE is also of paramount importance with
only few realising a thermodynamically stable interface while others passivate on contact.

Figure 1.23. Different causes for current constriction at a Li/SE interface due to insufficient contact or pore
formation, partial coverage with contaminations or the presence of other charge transfer hindering interlayers,
grain boundaries and protrusions at the interface or surface defects. Reproduced from reference103.

Moreover, lithium is a soft metal with a yield strength of 0.7 ± 0.2 MPa.103 It undergoes
plastic deformation when subjected to pressures exceeding this threshold and, as previously
demonstrated by Doux et al., pressures < 5 MPa are required for proper operation of LMAs
with an Li6PS5Cl SE separator while pressures > 5 MPa leads to the mechanical extrusion of the

33
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

lithium through the SE.91 Therefore, decreasing the pressure is also a prerequisite to LMA
integration. However, with a bare LMA, reports have highlighted the importance in
maintaining some level of pressure to prevent issues like pore formation and contact losses
(Figure 1.23) during stripping, which could potentially result in current constriction and
contribute to cell failure.106 Thus, at this stage finding the right balance in pressure seems to
be crucial in ensuring the stable and efficient operation of lithium metal ASSBs.

Significantly decreasing the stack pressure, while necessary for lithium implementation
and practical applications, is likely to also have a negative impact on the performance of the
positive electrode. As mentioned earlier, to this day, most studies employ extremely high stack
pressures, often exceeding 50 MPa, when cycling ASSBs. This is done in order to mitigate the
electrochemical fatigue that can lead to contact losses and crack formation (Figure 1.24)
within the cathode composite due to the volume changes experienced by the CAM upon
lithiation and delithiation.107 Hence, it becomes crucial to minimise or eliminate sources of
strain within the electrodes and to engineer materials capable of offering efficient stress relief
mechanisms, thereby mitigating the occurrence of fractures and accommodating the
inevitable volume changes of most of the high performance CAMs.

Figure 1.24. Different types of fatigue damage identified in composite cathodes. Reproduced from reference107.

Overall, decreasing the pressure and integrating LMAs into high-energy density cells
still poses several challenges. Most of the current literature, bypass these issues through some

34
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

impractical and misleading tricks. Some of these include raising the temperature to enhance
plating homogeneity, using low current densities, reducing CAM loading to minimise the
current and the amount of plated Li, or increasing the thickness of the separator to prevent
dendrites from penetrating through the SE. However, some more practical interlayer, coating
and alloying strategies have recently shown promising results in increasing lithium stability
with atmosphere and SE as well as in enabling homogeneous plating and stripping even at
higher current densities.101,108–111

1.4 – Exploring the Cutting-Edge of All-Solid-State Batteries

Having developed a comprehensive understanding of the current challenges
associated with ASSB design, it is now important to highlight some innovative strategies that
have been implemented to overcome these obstacles in recent years. The forthcoming state-
of-the-art analysis will focus on three main areas: SE development, noteworthy ASSB full cell
performance and low-pressure strategies.

o The latest solid electrolyte developments

In recent years, there have been significant advancements in the performance of SEs.
Notably, the conductivity and stability of argyrodite SEs have seen remarkable improvements.

The argyrodite-SE, with the most common Li6PS5Cl phase, is showing conductivities of
about 2 · 10-3 S/cm. This conductivity was greatly improved in 2019 when Nazar’s group
reported that halide substitution (Li7-xPS6-xClx with x = 1.0-2.0) could improve the ionic
conductivities. They demonstrated an optimal performance for Li5.5PS4.5Cl1.5 through a simple
solid-state synthesis and obtained a conductivity of up to 9.4 · 10-3 S/cm.112 This new
composition was then employed recently in 2022, to significantly improve NMC/Li-In cell
performance at high rate, low temperature and moderately high loading.113 Notably, they
were able to reach 80 mAh/g at a 5C rate at RT and maintain a capacity of 115 mAh/g at 2C
for over 250 cycles afterwards (Figure 1.25a). They also demonstrated exceptional rate
capabilities at low temperature of -20°C with stable 50.6 mAh/g at a C/2 rate after 100 cycles
as depicted in Figure 1.25b and c. Additionally, extended cycling tests were conducted at 5C
and 10C, yielding relatively stable capacities (Figure 1.25d and e). At 5C, they reached around
100 mAh/g for 4500 cycles with a 99.4% capacity retention, and at 10C, about 75 mAh/g for
10000 cycles with an 82.4% capacity retention.

35
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

Figure 1.25. Rate capability and cyclability evaluation of the NMC622/Li5.5PS4.5Cl1.5/LiIn ASSBs. The rate
capability of the fabricated SSBs working at (a) RT and (b) −20 ℃. Cycling performance under different
temperatures and C-rates: (c) cycled at 1C under −20 °C, (d) 5C and (e) 10C at RT. Reproduced from reference113.

Significantly increasing the ionic conductivity was not the only improvement that
argyrodite received. The SE's stability against environmental factors such as moisture and air,
as well as its compatibility with high potential cathode active materials and lithium was the
subjects of significant improvements as well.74,78 One remarkable achievement was the
development of a protective coating on argyrodite particles.74 This protective layer, formed
through a facile and scalable spontaneous gas-solid reaction with CO2 gas (Figure 1.26a),
resulted in the deposition of a thin Li2CO3 coating on the surface of the argyrodite particles.
This protective coating was able to considerably improve the SE ionic conductivity retention
after exposure to moist air from a 1.8% to an impressive 80.9% (Figure 1.26b) with just a 1.5
hours long gas-treatment that formed a 70 µm-thick Li2CO3 coating. Furthermore, stability
towards CAMs was improved, demonstrating an excellent capacity retention of 89.4% after
2100 cycles owing to the 50 µm-thick layer of Li2CO3 formed by a 1h CO2 treatment. This
coating, although displaying a low conductivity was thin enough to allow sufficient conduction
through the SE (Figure 1.26e).

36
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

Figure 1.26. (a) Schematic illustration of the solid-gas reaction and its application. (b) Ionic conductivity of
treated samples before and after exposure to humid air for 1h and conductivity retentions, (c) initial cycles at
0.1C rate, (d) rate capabilities, (e) long-term cycling and (f) charge profiles with CO2-1h LPSC at different cycles.
Reproduced from reference74.

Since around 2018, the year when halide-based SE resurfaced, their exploration has
yielded noteworthy advancements as well. The discovery of new halides such as Li2InxSc2/3-xCl4
(LiInScCl) or Li3-xZrx(Ho/Lu)1-xCl6 by Nazar’s group which, although both showing ionic
conductivity in the 1-2 mS/cm range, exhibited remarkable cycling stability towards Ni-rich
NMC materials.49,54 Specifically, the former demonstrated a RT ionic conductivity of up to
2.0 mS/cm for the Li2In1/3Sc1/3Cl4 composition and displayed exceptionally long cycling
performance for up to 3000 cycles with over 80% capacity retention at a high upper cut-off

37
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

potential of 4.3 V vs. Li+/Li (Figure 1.27a). Furthermore, they showed remarkable stability of
this SE at cut-off potential of up to 4.8 V vs. Li+/Li as well as stable capacities for high loading
cells of up to 21.6 mgAM/cm² and 52.5 mgAM/cm² for NMC851005 and LCO respectively
(Figure 1.27b-d). With these loadings, they reached aerial capacities of approximatively 3 to
4 mAh/cm², which are suitable for industrial-scale applications. Despite their impressive
performance, it is worth noting that these halide-based SEs, like many others in the literature,
still incorporate expensive and scarce elements such as indium and scandium.

Figure 1.27. Long-term cycling of NMC851005/LiInScCl/LiIn cells (a) at a 3C rate between 2.8 and 4.3 V vs. Li+/Li,
(b) at C/5 between 2.8 and up to 4.8 V vs. Li+/Li and (c-d) high loading cycling of (c) NMC at RT and (d) LCO at
50°C. Reproduced from reference54.

Nevertheless, the Li stability remains one of the main challenges, as halide SEs are
unstable at low potential. For instance, the calculated thermodynamic reduction stability
limits for Li3YCl6 and Li3YBr6 are 0.62 and 0.59 V vs. Li+/Li, respectively. To avoid this issue, most
studies employ a LiIn alloy anode, which potential is of 0.622 V vs. Li+/Li or use an additional
protective layer of stable SE such as argyrodite. However, these strategies introduce
complexity in the cell manufacturing and lower the overall energy density.

38
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

Nonetheless, some striking advancements have been made in that regard. One
noteworthy example dates back to 2021, when a research group employed a partial
fluorination strategy of Li3YBr6 to form Li3YBr5.7F0.3.65 This modified SE demonstrated improved
stability at low potentials while retaining most of the ionic conductivity exhibited by the non-
fluorinated SE (Figure 1.28a and b). Additionally, the modified SE displayed enhanced contact
stability with lithium metal. As depicted in Figure 1.28c and d, stable symmetric cycling was
achieved for 1000 hours, in contrast to the unmodified SE that failed after only 300 hours due
to the formation and growth of the MCI layer. This improvement was attributed to the in-situ
formation of a fluoride-rich interfacial layer, composed of LiF and YFx compounds, during
lithium plating/stripping. This layer passivated the surface and limited the further growth of
the interphase, resulting in the observed stability enhancement.65,114,115

Figure 1.28. (a) Linear sweep voltammetry curve and (b) Arrhenius plots of Li 3YBr6 and Li3YBr5.7F0.3. (c,d) Li
symmetric cell cycling at 0.1 mA/cm² with a capacity of 0.1 mAh/cm² in (c) a Li/Li3YBr5.7F0.3/Li and d)
Li/Li3YBr6/Li symmetric cell configurations. Reproduced from reference65.

Similarly, in 2023, a significant breakthrough was achieved with the discovery of a new
LaCl3-based halide SE with a UCl3-type lattice that exhibited exceptional compatibility with
lithium metal.50 Through Ta doping, the resulting optimised Li0.388Ta0.238La0.475Cl3 SE showed a
high conductivity of 3.02 mS/cm and, when coupled to Li in a symmetric cell (Figure 1.29c),
demonstrated stable operation for more than 5000 hours. Such interface stabilisation was
rationalised by the formation of an insulating LiCl-rich buffer layer upon Li contact

39
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

(Figure 1.29b). They also demonstrated stable NMC532/Li cell operation for more than 100
cycles with an 81.6% retention (Figure 1.29c-e). Moreover, their investigation extended to
various other compositions such as Li0.495Zr0.259Ca0.086La0.432Cl3. Such versatile electrolyte
formulation promises enormous potential for the use of more abundant and cheaper
elements for the design and fabrication of such SE stable with Li metal.

Figure 1.29. (a) Symmetric cell cycling, (b) schematic of interface stabilisation and (c) galvanostatic, (d) dQ/dV
and (e) retention of a NMC532/Li0 cell with the La-based SE. Adapted from reference50.

Apart from these advancements, a surprising concept comprising a full-halide cell was
published by Cheng Ma’s group in 2023.116 It consists in the use of the Ti-based halide Li3TiCl6

40
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

(LTC) as separator, with its ionic conductivity of 1 mS/cm, as well as cathode and anode
materials (LTC+C) playing on the Ti3+/Ti4+ and Ti2+/Ti3+ redox, respectively (Figure 1.30). Such
cell could not reach high capacities nor stability but the concept allows for higher energy
density, as the electrodes are free of the electrochemically inactive SE mass.

Figure 1.30. (a) 1st cycle and (b) retention of a LTC+C/LTC/LTC+C cell. Adapted from reference116.

New classes of solid electrolytes are also attracting attention due to their promising
characteristics. One notable example is the lithium-metal-oxyhalide family, which has recently
shown very promising results. A significant breakthrough comes in 2023 with a publication by
Panasonic presenting two new oxyhalide SEs: LiNbOCl4 (LNOC) and LiTaOCl4 (LTOC). These
oxyhalides display exceptional ionic conductivities, with LNOC and LTOC achieving 10.7 and
12.4 mS/cm, respectively. Additionally, the study highlighted the high-rate capability of these
materials in ASSB full cells, using LNOC as the cathode SE (Figure 1.31b and c).

Figure 1.31. (a) Ionic conductivity comparison between LNOC, LTOC and other reported SEs measured at RT.
(b) First galvanostatic cycle and (c) rate capabilities of LCO/LNOC/LPSCl/Graphite (in red) and
LCO/LYC/LPSCl/Graphite (in green) full cells. Adapted from reference60.

Collectively, recent years have seen remarkable advances in the development of highly
performing SEs. These new SEs have proved capable of operating over a wide range of
potentials, with impressive conductivities, reaching those of traditional liquid electrolytes,
while simultaneously exhibiting robust cycling stability. This progress holds significant promise
for the evolution of solid-state battery technology.

41
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

o Emerging progress in full cell ASSB systems

The integration of various components has indeed posed persistent challenges in the
development of ASSBs. However, some significant progresses in the performance of full cell
are worth highlighting here.

In 2020, a ground-breaking study by Lee et al. presented a significant advancement in

the design and integration of ASSBs.101 The authors reported a new large pouch ASSB cell
comprising a NMC/Li6PS5Cl/VGCF cathode, a Li6PS5Cl separator and an anodeless design
(Figure 1.32a). To combat the Li dendrite issue, they implemented a novel interlayer between
the current collector and the SE, composed of carbon and silver nanoparticles (Ag-C), that
enabled homogeneous Li plating/stripping across a large surface area of a 0.6 Ah prototype
pouch cell (Figure 1.32b). Experimentally, they obtained stable cycling for over 1000 cycles,
retaining 89% of the initial capacity (Figure 1.32c). Notably, their utilisation of the isostatic
pressing procedure (WIP) was one the first to be reported and although disruptive at the time,
one should mention that the cycling temperature used was still high (60°C) which considerably
aid the use of lithium metal.

Figure 1.32. (a) Schematic of an ASSB implementing the Ag-C interlayer. (b) Pouch cell illustration and X-ray
computed tomography of the cell cross section. (c) Capacity retention of the 0.6 Ah pouch cell prototype cycled
at 60°C. Adapted from reference101

More recently, in 2023, the same research group further extended their innovative
approach when they demonstrated the use of this same interlayer in an oxide-based SE cell,
although a liquid electrolyte was still used in the cathode to facilitate Li transport. 108 The

42
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

mechanism underlying the homogeneous plating and stripping aided by the Ag-Graphite
interlayer was recently studied by Peter Bruce’s group.110 They revealed that upon charge, Li
initially electrochemically intercalates into graphite before chemically reacting with silver to
form Li-Ag alloys. This process continues until the complete formation of LiC6 and Li10Ag3;
lithium plating then occurs between the Ag-C interlayer and the current collector.

In 2021, another breakthrough shook the ASSB world when Meng’s group reported the
use of a high energy density silicon anode paired to a Li6PS5Cl separator in a full cell design
ASSB.117 They assembled a cell (Figure 1.33a) with an anode composed of 99.9% micro silicon
(µSi) anode and a high loading NMC811 cathode (25 mgAM/cm²) and demonstrated an
exceptional cycling performance at 1C for 500 cycles, maintaining an 80% capacity retention
(Figure 1.33b). The superior performance achieved in comparison to liquid systems was
attributed to the significant volume expansion of µSi. This expansion, counterintuitively,
played a beneficial role by leading to a high degree of densification of the anode. However, it
is worth noting that despite achieving long cycling, the capacity of the NMC811 was largely
underutilised, only reaching approximately 80 mAh/g, while the applied pressure needed to
be excessively high as well in the order of 400 and 100 MPa for the formation and subsequent
cycles, respectively.

Figure 1.33. (a) Schematic of the NMC811/LPSCl/µSi cell and proposed lithiation mechanism and (b) long
cycling performances at RT. Adapted from reference117.

Another interesting concept, published in 2021, was the use of a SE-free and C-free
cathode by revisiting the layered titanium disulphide.118 Their approach hinged on the mixed

43
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

ionic-electronic conductor property of TiS2, which allowed them to employ the material
without any ionic or electronic additives (SE and carbon), which are commonly incorporated
into conventional composites (Figure 1.34a). As a result, despite operating at the lower
potential range of TiS2, they managed to reach high energy density of up to 414 Wh/kgelectrode
by eliminating any electrochemically inactive components. Furthermore, they conducted
cycling at high loadings of up to 45.6 mgAM/cm², which translated to a practical areal capacity
of approximately 9.43 mAh/cm² (Figure 1.34b). The same group also work on the same
concept for anode material, of which they chose graphite and silicon. 119,120 Using these
anodes, they achieved relatively stable cycling at high areal capacities of nearly 3 mAh/cm²
while also having high energy density at electrode level owing to the SE-free anode design.

Figure 1.34. (a) Cell schematic, (b) galvanostatic cycling and (c) capacity retention of the TiS2/LPSCl/Li0 cell.
Adapted from reference118.

The notion of SE-free cathodes, relying solely on the intrinsic conduction properties of
the AM is of particular interest. This concept is closely related to other studies mentioned
earlier, such as the utilisation of µSi117 anodes and full halide116 cells based on Li3TiCl6. The
potential of this approach in reducing the required stack pressure for ASSBs will also be
investigated within the scope of this thesis.

o The progress toward low pressure cycling

Mitigating the need for high stack pressure during cycling is a crucial challenge in
scaling up ASSBs. In recent years, noteworthy research works have explored various strategies
aimed at addressing this issue.

A problem when attempting to decrease the stack pressure is to maintain particle

contacts and percolation in the cathode composite. In that context, in 2023, a group used a
processing technique that they newly applied to ASSB composite preparation, known as
solvent-free mechano-fusion (MF) process. This process consists in subjecting the AM/SE

44
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

mixture to high shear and compression forces to form a thin and robust coating layer of SE on
the surface of the AM (Figure 1.35a). Through this process, NMC particles were effectively
coated with a thin and uniform layer of Li6PS5Cl SE (Figure 1.35b). The outcomes were
remarkable, as outstanding cycling performance was achieved with composites containing as
little as 20% and 10% SE. Typically, such low SE content would be insufficient to ensure
percolation thus leading to poor AM utilisation and low capacity. The coated NMC@LPSCl
composites, prepared in 80:20 and 90:10 ratios, exhibited initial discharge capacities of 163
and 148.5 mAh/g, respectively, performances significantly surpassing those of the regular
NMC+LPSCl composites which only registered 148 and 38 mAh/g in the same ratios,
respectively (Figure 1.35c-e). Moreover, the improved percolation translated into enhanced
rate capabilities compared to regular composites. Beyond enhancing cycling performance and
rate capabilities, this technique also contributed to the possibility of substantially lowering the
cycling pressure. The authors demonstrated that pressures of 20 or 3 MPa had no impact on
reversible capacity when using MF-prepared composites. In contrast, with bare NMC
composites, lowering the pressure to 3 MPa resulted in a substantial 65% decrease in the first
cycle discharge capacity.

Figure 1.35. (a) Illustration of the mechano-fusion process and (b) surface and cross-section SEM images of the
obtained Li6PS5Cl-coated NMC (NMC@LPSCl) particles. First galvanostatic cycles of bare NMC composite
electrode and NMC@LPSCl composite electrode at 20 and 3 MPa with NMC to LPSCl ratios of (c) 80:20 and (d)
90:10 and (e) capacity retention of the 80:20 ratios. Adapted from reference94.

Another noteworthy contribution to low-pressure cycling of NMC composites was

presented by Peter Bruce's research group in 2022. They capitalized on the high ionic
conductivity and excellent high potential stability of halide SE integrated into the cathode
composite. This enabled them to achieve cycling at remarkably low pressures, as low as 2 MPa.
However, it is important to note that they still required high-temperature operation at 80°C

45
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

to reach similar performance as the high-pressure ones. The mention of the choice of SE, the
composite composition, the temperature and rates highlight how important these factors
become when trying to decrease the cycling pressure.

Figure 1.36. Comparisons of the first cycle performance of the composite cathode NMC811/Li3InCl6/CNF in
65/30/5 weight ratio under different current densities, temperatures and stack pressures. Reproduced from
reference121.

1.5 – Chapter conclusion

In conclusion, Chapter 1 has provided an in-depth exploration of the landscape of
today's battery technologies and the renewed interest in ASSBs. We explored the historical
development of solid electrolytes, tracing the evolution from early inorganic conductors to
the emergence of high-performance lithium-ion conductors.

The potential benefits that ASSBs can offer in terms of safety, energy density, and
overall battery performance were discussed. However, we recognised that despite the
promises, ASSBs come with their own set of challenges. The chemical and electrochemical
interfacial reactivity within these batteries presents a complex set of issues that need to be
addressed for achieving stable and long-lasting performance. The formulation, processing,
and integration of various components, such as cathode composites and solid electrolytes,
add complexity that must be carefully considered. Moreover, the quest for decreased stack
pressures and the integration of high-capacity lithium metal anodes underscore the pressing
need for innovative solutions. We explored strategies to address these challenges, ranging
from new solid electrolyte materials to advanced processing techniques, highlighting the
cutting-edge advancements in the field.

46
 Chapter 1 – Today’s Battery Technologies and the Resurgence of All-Solid-State Batteries

Looking ahead, Chapters 2 and 3 will delve deeper into strategies aimed at lowering
cathode-side cycling pressure, providing an overview of innovative approaches and their
implications. In addition, Chapter 4 will focus on the incorporation of lithium metal anodes
into these previously explored solutions, with the aim of further increasing the energy density
of ASSB systems and paving the way for scalability.

47
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

Chapter 2 – Atmospheric-Pressure
Operation of All-Solid-State
Batteries enabled by Halide-based
Solid Electrolyte1

1
This chapter includes the following publication: Hennequart, B. et al. Atmospheric-Pressure Operation of All-
Solid-State Batteries Enabled by Halide Solid Electrolyte. ACS Energy Lett. 9, 454–460 (2024). And a patent
application (EP23305141, pending)

49
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

2.1 – Chapter Introduction

Despite the potential for high energy density in all-solid-state batteries (ASSBs),
research efforts are still needed to overcome fundamental issues that prevent their
commercialisation. As introduced in the previous chapter, a critical requirement for achieving
the promised energy densities is the use of a lithium metal anode (LMA) that requires the
considerable reduction of cycling or stack pressure.91 However, to date, achieving high
cyclability under low pressures has proven to be challenging. This difficulty stems from a
multitude of interfacial issues that give rise to inadequate contacts in the cathode. Indeed, in
an ASSB, the cathode usually being a composite of cathode active material (CAM), solid
electrolyte (SE) and carbon additive, new challenges emerge as the pressure is decreased.
Most notably is the volume variation that the CAM undergo during charge and discharge cycles
of 6 to 8% for the typical LiNixMnyCo1-x-yO2 (NMC)92, which can lead to loss of contact between
the CAM and SE particles and result in poor cyclability. Consequently, most studies in current
literature on cathodes for ASSBs tend to overlook this issue by conducting cycling experiments
at high stack pressures, typically exceeding 50 MPa. These pressures are definitely
incompatible with real-world applications or Li metal integration.

Current strategies to enable low pressure cycling include increasing cycling

temperature, lowering upper cut-off potential to limit CAM volume changes92, optimising
CAM and SE particle sizes to improve percolation81,80, or using solid electrolyte-free cathode
systems118 to minimise interfaces. More recently Gao et al.121 utilised the chloride-based
Li3InCl6 SE and NMC to reach pressures of 2 MPa at 30°C when having a limited oxidation cut-
off of 4.2 V vs. Li+/Li.

This chapter aims at proposing a new strategy to overcome this pressure limitation
that is based on consideration of chemical and mechanical properties of the SE present in the
composite. We will first benchmark the performance of a typical sulphide-based (NMC:LPSCl)
cathode composite under different pressure conditions and see the effect of the composite
formulation on the cyclability. Then we will explore the halide-based SE Li3InCl6 as a promising
candidate for low-pressure cycling. Finally, we will introduce Li3YBr2Cl4, a mixed halide SE
displaying improved reduction stability and impressive mechanical properties.

50
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

2.2 – The low pressure cycling of argyrodite-based NMC composites

To establish a baseline for future experiments involving various SEs, we conducted a
performance assessment using a commercial argyrodite Li6PS5Cl SE (NEI Corporation, Fine
LPSCl) stated having a particle size of 1 µm but that turns to contain few larger chunks in the
range of 10 to 20 µm, as deduced by SEM measurements (Figure 2.1a). Additionally, through
electrochemical impedance spectroscopy (EIS), we determined an ionic conductivity of
1.62 mS/cm at a pressure of 100 MPa for this SE.

a. b.
RT
200 100 MPa
-Im(Z) (Ohm)

100

0
0 100 200 300
Re(Z) (Ohm)
Figure 2.1. Commercial argyrodite Li6PS5Cl properties. (a) SEM image of the material and (b) conductivity
measurement by EIS.

o A high pressure reference

Initially, we assessed the performance at a high pressure of 100 MPa within a

homemade ASSB cell device as represented in Figure 2.2. This ASSB cell is composed of a LPSCl
separator, a lithium-indium alloy/LPSCl composite anode (LiIn, in a 1/2 atomic ratio to obtain
a stable 0.62 V vs. Li+/Li) and a cathode composite comprising an 4 µm in size uncoated
monolithic LiNi0.6Mn0.2Co0.2O2 (referred to as NMC622 hereafter) active material, the same
LPSCl SE and vapour-grown carbon fibre (VGCF) as a conducting additive.

51
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

Figure 2.2. Schematic of (a) the homemade cell used for ASSB tests and (b) the ASSB pellet stack composition
used for cycling.

The assembly procedure of such solid-state battery involves the sequential

densification at 400 MPa of the powders used for the different layers of the battery, i.e. the
LPSCl separator, the Li0.5In/LPSCl anode and the NMC622:LPSCl:VGCF cathode composites.
Subsequently, the galvanostatic cycling process for these high-pressure cells is conducted at
100 MPa by tightening the six screws at 2.3 N.m using a torque wrench. A more
comprehensive and detailed procedure for the preparation of the composite and the assembly
of the cells is available in the Appendix Section A2.1.

Galvanostatic cycling in the potential window 2.1-3.6 V vs. Li+/LiIn (or 2.7-4.2 V vs.
Li+/Li) was performed on this LPSCl-based ASSB configuration. Figure 2.3 shows the
galvanostatic profile of the first cycle and the capacity retention of this cell subjected to cycling
at C/20 under 100 MPa with a loading of 12 mgNMC/cm².

This system exhibit charge and discharge capacities of 180 and 138 mAh/g on first cycle
resulting in a 77% initial coulombic efficiency (ICE). In addition, from Figure 2.3b, such cell
shows a retention of 88% after 20 cycles. This high irreversibility and poor retention are typical
of sulphide-based composites, known to undergo oxidation reactions at high potential. 67 It is
usually mitigated by the use of a thin coating on the AM particles, such as zirconium, cobalt or
niobium-based oxide coatings. As an example, a Zr-coated NMC622 was also tested in this
configuration and an improved ICE of 81% was obtained (Appendix Figure A2.2). However, in

52
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

our study, only bare NMC622 will be used. Altogether, these achieved results will serve as a
reference for studying the effect of low-pressure conditions on cycling.

200 105
a b
3.5 C/20, RT 100

Coulombic efficiency (%)

Charge Capacity (mAh/g)
Voltage (V vs. Li+/LiIn)

12 mgNMC/cm² 150
100 MPa 95

3.0
90
100

85
2.5
50 Charge Capacity (mAh/g) 80
Discharge Capacity (mAh/g)
Coulombic efficiency (%) 75
2.0 0
0 50 100 150 200 0 5 10 15 20
(Q-Qo) (mAh/g) Cycle number
Figure 2.3. Galvanostatic cycling performance of a NMC622/LPSCl/VGCF cathode composite at a pressure of
100 MPa in a LPSCl-based ASSB architecture. (a) Voltage profile of the first galvanostatic cycle and (b) capacity
retention and coulombic efficiency evolution.

o Low pressure performance and the importance of formulation

To investigate the low-pressure performance, we used the same cell composition as

previously but modified the cycling device to accommodate a broader pressure range. As
illustrated in Figure 2.4, it comprises a frame in which the prepared cell is positioned for
cycling. Pressure monitoring is achieved through a force sensor located beneath the cell, and
pressure adjustments are made by tightening a screw that secures the cell in place during the
experiment. This design is a fixed-gap system having a constant volume but for which the
pressure can vary throughout the cycling process. This is in opposition to a constant pressure
setup that can accommodate volume changes while keeping the pressure nearly constant by
the use of a spring. Noteworthy is that this design is not air-tight and must be employed inside
a glovebox, which exposes it to larger temperature variations and vibrations that can disturb
low pressure performance.

53
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

Figure 2.4. Schematic of the fixed-gap cell design used for low pressure cycling.

Low-pressure cycling experiments were carried out employing the same cycling
protocol, but with an initial pressure set to 10 MPa and a reduced loading of 5 mgNMC/cm².
During the initial cycle, depicted in Figure 2.5, the achieved capacities are similar to those
observed in the high-pressure tests, attaining 137.6 mAh/g on the first discharge, with a
typical ICE of 76.5%. However, affected by this lower stack pressure, the CE decreases from
an average of 99.15% to 94.22%.

200 105
a b
3.5 C/20, RT 100

Coulombic efficiency (%)

Charge Capacity (mAh/g)

5 mgNMC/cm² 150
Voltage (V vs. Li+/Li)

95

3.0
90
100

85
2.5
50 Charge Capacity (mAh/g) 80
Discharge Capacity (mAh/g)
Coulombic efficiency (%) 75
2.0 0
0 50 100 150 200 0 5 10 15 20
(Q-Qo) (mAh/g) Cycle number
Figure 2.5. Galvanostatic cycling performance of a NMC622/LPSCl/VGCF cathode composite at a pressure of
10 MPa in a LPSCl-based ASSB architecture. (a) Voltage profile of the first galvanostatic cycle and (b) capacity
retention and coulombic efficiency evolution.

Subsequently, we further reduced the testing pressure. To this end, we conducted two
formation cycles at 10 MPa, after which the pressure was systematically changed every 3
cycles, ranging from 10 to 0.1 MPa, by adjusting the top screw. Figure 2.6 shows that at
10 MPa, the cell displays a nominal charge capacity of 174 mAh/g with an ICE of 74%, slightly

54
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

lower than before. However, further lowering the pressure resulted in a rapid decay in
reversible capacity, reaching values of 94 and 88 mAh/g at 1 and 0.1 MPa, respectively.

200
a 1st cycle b
3.5

Charge Capacity (mAh/g)

150 100 MPa
Voltage (V vs. Li+/Li)

3.0
100

2.5 10 MPa 5 2 1 0.5 0.2 0.1

50
C/20, RT Charge Capacity (mAh/g)
Cycling pressure (MPa)
5 mgNMC/cm²
0.1 10 Discharge Capacity (mAh/g)
2.0 0
0 50 100 150 200 0 5 10 15 20
Capacity (mAh/g) Cycle number
Figure 2.6. (a) First galvanostatic cycle and last cycle at each pressure and (b) capacity retention of a solid state
cell comprising a LPSCl-based cathode composite (NMC622:LPSCl:VGCF) cell cycled versus a LiIn anode as a
function of the stack pressure (from 10 to 0.1 MPa) and the reference at 100 MPa, cycled at RT and a C/20 rate

This capacity decline can be attributed to various factors, including percolation issues
arising from a poor optimisation of the particles sizes of the AM and the SE. In an earlier report,
Ceder showed both via simulations and experimental data that cathode utilisation in ASSBs is
controlled by percolation, and that a higher ratio of cathode to SE particle size facilitates
greater cathode loading.80 The same rationalisation could explain the deteriorating
performance observed when pressure is reduced, as an insufficiently optimised particle size
ratio may contribute to increased percolation issues and contact losses. To assess the
significance of this metric in low-pressure cycling performance, we conducted experiments at
10 MPa using argyrodite with two different particle sizes. The commercial LPSCl having
particle of approximately 1 µm, with a few larger chunks of around 10 to 20 µm (Figure 2.1a)
was compared to another LPSCl prepared in-house through a solid-state synthesis, referred to
SS-LPSCl hereafter. A detailed procedure can be found in Appendix Section A2.1.

Compared to the commercial LPSCl, this SS-LPSCl displays significantly larger particle
sizes with the largest chunks ranging from 50 to 200 µm; however, its ionic conductivity was
higher at around 3.2 mS/cm.

55
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

a. b.
RT
100 100 MPa

-Im(Z) (Ohm)
50

0
0 50 100 150
Re(Z) (Ohm)
Figure 2.7. Homemade solid-state argyrodite SS-Li6PS5Cl properties. (a) SEM image of the material and (b)
conductivity measurement by EIS.

Figure 2.8 compares the first cycle capacities at C/20 of composite made with either of
these two argyrodites. Note that at high-pressure (100 MPa on the top in Figure 2.8), the
performance is not affected by the particle size. However, decreasing the pressure to 10 MPa
(bottom part in Figure 2.8) we observed that the fine LPSCl has a much higher reversible
capacity compared to the SS-LPSCl. This highlights the importance of having the proper
formulation and of adjusting the particle size ratio between the SE and the AM in the
composite when performing low pressure cycling experiments.
10 MPa 100 MPa

SS-LPSCl 126 mAh/g

Fine LPSCl 127.3 mAh/g

SS-LPSCl 89.5 mAh/g

Fine LPSCl 126 mAh/g

0 50 100 150 200

Capacity (mAh/g)
Figure 2.8. First cycle capacities at C/20 as a function of pressure and SE/AM particle size ratio, either with the
fine or the larger SS-LPSCl. Solid coloured bars corresponds to the first discharge capacity and the hashed parts
to the irreversible on the first cycle. The cumulative value represents the first charge capacity. All data are
averaged over the capacities of 2 or 3 identical cells.

Besides the particle size ratio, the presence of an electronic conductor also significantly
influences the cycling performance when pressure is decreased. Notably, when cycling a
carbon-free composite under high pressure, the percolated NMC network provides a

56
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

sufficiently high electronic conductivity at low C-rates to achieve the full capacity under these
conditions. As our group previously reported, this behaviour can be exploited in sulphide-
based composite to mitigate the catalytic impact of carbon additives on SE decomposition.122
However, as illustrated in Figure 2.9, the decrease in the initial cycling pressure results in a
substantial capacity loss within this carbon-free system, reaching only 26 mAh/g at 10 MPa.
This is notably lower than the 130 mAh/g achieved with the carbon-containing composite
shown in Figure 2.6. This outcome can be explained by the diminishing contacts between the
NMC particles under low-pressure conditions, resulting in a disrupted electronic percolation
network in C-free composites.

Average discharge capacity (mAh/g)

125 a. 10 100 50 25 10 100 MPa 125 b.
Discharge Capacity (mAh/g)

RT
100 C/20 100

75 75

50 50

25 25

0 0
0 20 25 30 35 0 20 40 60 80 100
Cycle number Pressure (MPa)
Figure 2.9. Capacity retention of a C-free composite cycled at various stack pressure from 10 to 100 MPa. (a)
Capacity retention of a NMC/LPSCl composite cycled against a LiIn/LPSCl anode at different pressure and (b)
average discharge capacity for each pressure tested.

Altogether, our exploration of the low-pressure performance of LPSCl-based

composites reveals that their performance deteriorates notably below 10 MPa. Nonetheless,
given the widespread utilisation of LPSCl in high-pressure studies in the literature, these
findings establish a valuable reference point for our future investigations into lower pressure
conditions. Moreover, we have confirmed the importance of optimising the particle sizes and
the necessity of using a conductive additive to provide adequate ionic and electronic
percolation within the composite, particularly in the context of low-pressure measurement
where the percolation issues are exacerbated. Building on that we have decided to explore
different SE chemistries available to us. In that context, we turned towards halide-based SE.

57
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

2.3 – Li3InCl6: A good candidate for low pressure cycling

From the LPSCl systems, we shifted our focus to explore alternative SEs. In particular,
we turned to the halide SE family due to their favourable attributes, including enhanced
oxidation stability, low hardness, and low elastic moduli. These mechanical properties become
especially significant when lowering stack pressures, as ductile SEs tend to outperform brittle
SEs in handling the strains and stresses that arise at the interface between SEs and electrode
active materials during cycling.107 Being the most widely used among the halide-based SE, we
started with the Li3InCl6 phase.

o Electrochemical and chemical stability of Li3InCl6

As previously reported by our group123,124, the Li3InCl6 SE, hereafter referred to as LIC,
is synthesized through a water-assisted method. Detailed synthesis protocols are available in
the Appendix Section A2.1. We previously reported this LIC SE to display an ionic conductivity
of 1.81 mS/cm and to exhibit crystalline particles smaller than 2 µm, which tend to loosely
agglomerate into larger secondary particles.123,124 These characteristics render LIC a suitable
choice as a SE for the preparation of cathode composites. Its conductivity meets the necessary
requirements, and its particle size aligns well with the 4 µm-large NMC particles.
Consequently, a cathode composite consisting of the same monolithic uncoated NMC622 as
previously used, this LIC SE, and VGCF conductive additive was manually prepared through
hand grinding, with a weight ratio of 66.5%, 28.5%, and 5%, respectively.

As discussed in Section 1.3 of the previous chapter, due to their limited reduction
stability, halide SEs decompose at low potentials, resulting in the formation of a detrimental
mixed ionic-electronic interphase that lead to rapid cell deterioration. A straightforward
approach to prevent this issue would be to employ a hetero-structure design, wherein another
solid electrolyte, such as LPSCl, is utilised as the separator (Figure 2.10b). However, even with
such an architecture, there is a notable degradation in cycling performance (Figure 2.10e).
This decay has been ascribed to an increase in resistance at the interface, triggered by the
formation of insulating species which may be catalysed by the electronic conductive species
present at the triple point between LIC, LPSCl, and NMC or C.125,126 As a result, an alternative
strategy involving a dual solid electrolyte separator design, as depicted in Figure 2.10c, is
commonly employed. In this approach, a bilayer consisting of LIC and LPSCl is utilised as the

58
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

separator. While this strategy generally leads to more stable cycling46,48,49,54, it is important to
note that in this particular SE pairing (LIC and LPSCl), a chemical incompatibility has been
identified.125,126 Through ToF-SIMS, Janek’s group reported126 this chemical incompatibility to
result in an increase in the separator resistance by the formation of In2S3, LiInS2 and LiCl
decomposition products. Our research group has previously demonstrated the feasibility of
implementing a nanometric interlayer composed of Li3PO4 or Al2O3, fabricated via atomic layer
deposition (ALD), to prevent any reactivity between the two SEs.125 However, for the sake of
simplicity, in the subsequent study, a dual-SE architecture will be employed without any
protective coating. Consequently, it is anticipated that there will be a continuous degradation
of capacity over time.

Figure 2.10. Cycling performances of the different cell architecture, i.e. conventional, hetero-structure and
dual solid electrolyte cell architecture. (a-c) Schematics of the tested architecture and (d-f) their galvanostatic
cycling performance at high pressure. Inset in (b) shows the triple point between the LIC, LPSCl and the electronic
conductive pathway. Parts d-f reproduced from reference126.

o Performances of Li3InCl6-based cathode composites under various pressures

The galvanostatic cycling performance of the previously mentioned cathode composite

was systematically investigated in conjunction with a LiIn/LPSCl composite anode within a
dual-SE separator cell design. ASSB cells were assembled and their cycling performance,

59
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

shown in Figure 2.11, was evaluated under different pressure conditions. Interestingly, both
cells, whether cycled at a high pressure of 100 MPa or at a relatively low pressure of 9 MPa,
exhibit comparable first cycle capacities and retentions, reaching 152 and 145 mAh/g with
initial coulombic efficiencies of 89.4% and 88.2% respectively. This stands in contrast to the
poor capacity observed when decreasing stack pressure in LPSCl-based systems (Figure 2.6).
Moreover, a continuous decay that we ascribe to the SEs incompatibility is present and
examining the discharge capacity variations (Figure 2.11c), both cells show a similar trend with
a decay rate, at this low C/20 cycling rate, of 0.98 and 0.83%/cycle et 100 and 9 MPa,
respectively. Important to note is the low-pressure decay that seems to be more scattered,
we attribute this to the effect of environmental conditions (temperature variation, vibrations
in the glovebox, etc.) which have a stronger impact on performance at low pressure. This
variability is often reflected on low-pressure cells in their capacities, coulombic efficiencies,
decay rates as well as their polarizations.
Discharge Capacity (mAh/g)

200 -5
a b c Average decay rate (%)
Voltage (V vs. Li+/LiIn)

3.5 -4 100 MPa: -0.98%/cycle

C/20
Decay rate (%)
150 9 MPa: -0.83%/cycle
RT -3
3.0
100 -2
Initial pressure Initial pressure -1
2.5 50
100 MPa 100 MPa
0
9 MPa 9 MPa
2.0 0 1
0 50 100 150 200 0 5 10 15 20 0 5 10 15 20
(Q-Qo) (mAh/g) Cycle number Cycle number
Figure 2.11. Galvanostatic performance of the LIC-based cathode composite at both high and low pressure. (a)
First galvanostatic cycle profile, (b) capacity retention and (c) discharge capacity decay rate at 100 MPa and at
9 MPa. Both cells are cycled at C/20 under RT condition.

At this point, we successfully achieved cycling pressures ten times lower than those
previously used, with no discernible impact on cycling performance. Nevertheless, it is
important to note that even though we reached 9 MPa, this pressure level is still relatively
high. For practical applications and the implementation of lithium metal anodes, it is
imperative to operate at pressures below 5 MPa, ideally even lower than 2 MPa. In this
context, the same cell was subjected to further cycling at successively lower pressures by
adjusting the top screw of the frame setup as illustrated in Figure 2.4. First, 50 cycles were
performed at 9 MPa before decreasing to 6, 2.5, 1.5, 0.5, 0.3 and finally 0.1 MPa. The results
are presented in Figure 2.12. While it appears that the decay rate slows down after
approximately 30 cycles, decreasing the pressure does not significantly influence the retention

60
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

or the average decay rate. Most of the capacity drop occurs in the cycles immediately
following pressure changes, as highlighted by the filled circles in Figure 2.12b. This is due to
the necessary cell manipulation required to change pressure, which effect becomes
particularly pronounced at low pressures, where maintaining good contact with the current
collector is crucial. Moreover, similar to the increased variability at low pressures mentioned
above, we also observe an increase in the decay rate variations when decreasing the pressure.
This is illustrated by the increase in standard deviation, as indicated by the filled area in
Figure 2.12b.

-12
a b Discharge decay rate (%)
T°C event

Decay at P change (%)

Discharge Capacity (mAh/g)

150 Average & 2SD

Discharge decay rate (%)

-8

T°C event
100 -4

0
50
9 MPa 6 2.5 1.5 0.5 0.3 0.1
4

0
0 25 50 75 100 0 25 50 75 100
Cycle number Cycle number
Figure 2.12. (a) Capacity retention and (b) decay rate of a LIC-based cathode composite at decreasing pressures
from 9 to 0.1 MPa. Filled circles in the decay rate highlight the larger capacity drop after each pressure change
and the filled area highlighting the larger variations at low pressure corresponds to two standard deviations and
does not include the data points after the pressure changes in its calculation. The temperature event that
occurred at around cycle 35 was due to an unusual important increase in the RT and was therefore remove for
the calculation of the standard deviation.

To ensure that cyclability remains consistent at lower pressures, we conducted

another experiment in which the cell was initially subjected to five formation cycles at 9 MPa,
followed by an immediate reduction of the stack pressure to 1 MPa. The results of this
experiment, depicted in Figure 2.13, indicate a capacity of 144 mAh/g for both cells on the
first discharge at 9 MPa, and show that they exhibit a similar decay rate throughout cycling,
eventually reaching 111 mAh/g after 50 cycles at either 9 or 1 MPa. Collectively, these results
provide clear evidence that cycling LIC-based composites at 100, 10 or 1 MPa does not hinder
the performance.

61
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

200
a b
3.5

Discharge Capacity (mAh/g)

C/20
Voltage (V vs. Li+/LiIn)

RT 150

3.0

Pressure change
100

2.5
Initial pressure 50 Initial pressure
9 MPa 9 MPa
9 then 1 MPa 9 then 1 MPa
2.0 0
0 50 100 150 200 0 10 20 30 40 50
(Q-Qo) (mAh/g) Cycle number
Figure 2.13. (a) First galvanostatic cycle and (b) capacity retention of a LIC-based cathode cell cycled either at
9 MPa or at 1 MPa after 5 cycles at 9 MPa. Both cells were cycled at a C/20 rate under RT conditions with loading
of 5 mgNMC/cm².

o The importance of the composite preparation process

It is important to emphasize the polarization difference observed on first cycle

between cells in Figure 2.11a and Figure 2.13a. This discrepancy can be attributed to the fact
that these cells were constructed using cathode composites from two distinct batches.
Following initial experiments (Appendix Figure A2.3a), a notable decline in performance and
reproducibility of subsequent LIC-based composites was observed (Appendix Figure A2.3b and
c). The issue was traced back to the degradation of the SE, likely caused by solvent vapours in
the glovebox atmosphere hence gaining valuable insights into the fabrication process of such
composites, which can be influenced by factors such as the operator, grinding time, and
applied force. Moreover, it highlights a significant challenge in the production of ASSBs that
rely on reactive SEs like halides. These SEs have a tendency to react with air, moisture, and
many solvents, posing a significant obstacle to their widespread utilisation in industrial
applications.

Altogether, we were able to cycle ASSB cells composed of a halide-based cathode

composite at pressures as low as 0.1 MPa or 1 bar. Such low pressure could therefore enable
their use in larger pouch cells design that would be cycled under atmospheric pressure. In
addition, it shows that halide SEs outperform the sulphide LPSCl SE when cycled under low-
pressure conditions, however, we used Li3InCl6, which is not an ideal choice. Indeed, it is costly
to produce due to the scarcity of indium and has a poor reduction stability that leads to the
necessity of using a dual-solid electrolyte cell architecture to prevent its reduction. In that

62
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

regard, we turned our interest towards other types of halide SE, in particular, the yttrium-
based halide SEs.

2.4 – Introducing Li3YBr2Cl4: a mixed halide SE with enhanced reduction

stability and low hardness
Yttrium-based halides were selected for their appropriate properties. In particular,
Li3YCl6 (LYC) displays a theoretically predicted cathodic limit as low as 0.62 V vs. Li+/Li (as
shown in Figure 2.14), limit much lower than the calculated 2.38 V vs. Li+/Li of the Li3InCl6 used
in the previous Section 2.3. Such low reduction stability falls within the range that allows for
compatibility with a Li-In alloy anode.127 However, Asano et al. reported a lower conductivity
of 0.51 mS/cm for LYC compared to Li3InCl6.45 Later, this conductivity was shown to be
improved through the partial substitution of Cl by Br anions and the mixed halide Li 3YBr3Cl3
was reported to display an ionic conductivity of 1.6 mS/cm which can be further improved to
7.2 mS/cm through a hot-pressing process at 150°C.46 However, as depicted in Figure 2.14, a
higher content in Br results in a decrease in the predicted anodic stability from 4.25 V vs. Li+/Li
for LYC to 3.15 V vs. Li+/Li for Li3YBr6 (LYB).

Figure 2.14. Calculated phase equilibria for (left and right) Li3YCl6 and Li3YBr6 based on first-principle
calculations (DFT, adapted from reference127) and potential stability window of Li3YBr2Cl4 obtained
experimentally through linear sweep voltammetry (see Figure 2.15a). Potential of Li metal, LiIn alloy and high
potential CAMs on the right. Stable potential window of SEs are coloured and electronically conductive species
are in red.

o Physical, chemical and electrochemical properties of Li3YBr2Cl4 SE

In our study, we utilised the mixed Br-Cl Li3YBr2Cl4 SE, denoted LYBC hereafter. This
halide was kindly provided to us by Saint-Gobain Recherche Paris. LYBC crystallises with
particle size ranging from 1 to 4 µm (Figure 2.15a) and has an ionic conductivity of 1.27 mS/cm

63
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

(Figure 2.15c). Its electrochemical stability window was determined by linear sweep
voltammetry (LSV) in oxidation and reduction separately (Figure 2.15b). Similar to the
observations previously46 reported with Li3YBr3Cl3 and compared to LYC and LYB, we denote
an intermediate oxidative current onset starting at 3.6 V vs. Li+/Li, which was associated to Br-
oxidation by calculation.127 In reduction, a small onset at 1.4 V and a larger current starting
from 0.4 V vs. Li+/Li are observed. The reduction stability observed in LYBC and the reversibility
of the electrochemical process between OCV and the LiIn potential (Figure A2.4) suggest that,
unlike Li3InCl6, LYBC may remain stable when in contact with a LiIn counter electrode with its
potential of 0.622 V vs. Li+/Li. Thus considerably simplifying the experimental procedure, as
there would be no need for the protective layer of LPSCl in this case.

OCV

4.2 V vs. Li+/Li

a. 0.0 b. ox
c.
red RT, 100 MPa
0 V vs. Li+/Li

 = 1.27 mS/cm

-Im(Z) (Ohm)
-0.2
0.8 V vs. Li+/LiIn
I (mA)

0 20
-0.4 100
I (µA)

-10 10
-0.6
-20 0

-0.8
0.0 0.5 1.0 1.5 2.5 3.0 3.5
Voltage (V vs. Li+/LiIn) Voltage (V vs. Li+/LiIn)
-1.0 0
-1 0 1 2 3 4 0 100 200
Voltage (V vs. Li+/LiIn)
Re(Z) (Ohm)
Figure 2.15. LYBC characteristics. (a) Micrograph of LYBC, (b) linear sweep voltammetry in oxidation and
reduction of LYBC/VGCF (95/5wt.%) composite and (c) impedance spectroscopy on LYBC pellet at 100 MPa.

The cycling performances of NMC622 with the LYBC SE were then evaluated. In that
regard, a cathode composite was prepared through a ball milling process. The same monolithic
uncoated NMC622 as previously used, LYBC, and VGCF in a ratio of 66.5 : 28.5 : 5 wt.% were
weighted and transferred to a 45 mL zirconia jar. Five 10 mm ZrO2 balls were used to reach a
36/1 powder-to-ball mass ratio and the composite was prepared through planetary ball milling
at 150 RPM for 30 minutes.

Next, to assess the stability of LYBC with our LiIn alloy anode, two cells were
constructed and their galvanostatic performance at a rate of C/20 under a pressure 100 MPa
were assessed and compared. The first consists in a dual-SE architecture
NMC//LYBC//LPSCl//LiIn and the second in a full-halide system NMC//LYBC//LiIn as illustrated
by the schematics in Figure 2.16a.

64
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

4.0 200
a NMC vs. LiIn/LPSCl b
C/20, RT NMC vs. LiIn/LYBC
3.5 100 MPa
Voltage (V vs. Li+/LiIn)

12 mgNMC/cm² 150 NMC vs. LiIn/LYBC

Capacity (mAh/g)
3.0 NMC622 : LYBC : VGCF
NMC vs. LiIn/LPSCl
100
2.5

vs.
50
2.0
Charge capacity (mAh/g)
LiIn : LYBC
Discharge capacity (mAh/g)
LiIn : LPSCl
1.5 0
0 50 100 150 200 0 5 10 15 20
(Q-Qo) (mAh/g) Cycle number
-5
c LiIn/LPSCl composite
Discharge capacity variation (%)

LiIn/LYBC composite
-4
- - - - Average decay rate

-3
Average decay rate:
LPSCl: -1.00% per cycle
-2
LYBC: -1.02% per cycle

-1

1
0 5 10 15 20
Cycle number
Figure 2.16. (a) First galvanostatic cycle, (b) capacity retention and (c) discharge capacity decay rate of both (in
brown, denoted 1) the dual-SE design and (in blue, denoted 2) the full-halide cells cycled at RT at a C/20 rate
and a stack pressure of 100 MPa. Dashed lines in (c) represent the average decay rate over the 20 first cycles.

Initial capacities of 182 and 184 mAh/g in charge, 150 and 155 mAh/g in discharge with
ICEs of 82.2% and 84.2% are obtained for the dual-SE and full-halide cells, respectively. This
demonstrates the compatibility of LYBC with LiIn anodes. Comparing the capacity retention of
the two systems, no difference is observed (Figure 2.16b). They both show a decrease in
capacity of about 1%/cycle at C/20 (Figure 2.16c). For the full LYBC cell (denoted “2” in
Figure 2.16), this observed behaviour is unexpected if the previously mentioned
halide/sulphide SE incompatibility (Figure 2.10) is considered as the sole source of
degradation. To interrogate this aspect, we measured the impedance evolution of a SE bilayer
LYBC//LPSCl under accelerated conditions at 80°C. Nyquist plots (Figure 2.17) were obtained
at 25°C in the pristine state and after various heating durations. Notably, within 103 hours of
heating, the impedance reached a plateau with a final increase in the total resistance of only
25%, result in stark contrast to the four-fold increase our group previously reported in the
Li3InCl6//LPSCl system.125

65
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

50
25°C
100 MPa

40

-Im(Z) (Ohm)
30

20 Time at 80°C
0h
20h
10
83h
103h
0
50 60 70 80 90 100 110 120
Re(Z) (Ohm)
Figure 2.17. Nyquist plots measured at 25°C of a LYBC//LPSCl stack after different 80°C exposure durations.

However, upon further mixing of the powders followed by heating to a temperature of

150°C (a detailed experimental protocol conducted by Maria Platonova from our group is
available in the Appendix Section A2.1), a clear evolution of the XRD is observed. Importantly,
even though the chosen temperature is well below the respective decomposition
temperatures of both SEs, a clear reactivity is evidenced by the disappearance of the LYBC XRD
reflections, to the benefit of growing peaks at 29.7, 34.5, and 49.4°, which can be attributed
to the formation of LiCl (Figure 2.18a) that is most likely responsible for the degradation of the
ionic conductivity of the mixture (Figure 2.18b) after heat treatment. These observations
indicate the existence of a thermodynamic instability between our electrolytes, which appears
to be kinetically limited at the lower temperatures at which we are cycling cells. Thus
explaining the comparable cycling performances observed in Figure 2.16 between the two
systems but also raising the question of the origin of this retention decay.

x103
10 3
a LiCl Be b
8
25°C
Pristine
100 MPa
Annealed
-Im(Z) (kOhm)

2
LYBC
Counts

6
LPSCl

4
1

2
Pristine
Annealed
0
0
10 20 30 40 50 0 1 2 3
2Q (°) Re(Z) (kOhm)
Figure 2.18. (a) Diffractograms and (b) Nyquist plots of the EIS measurement of pristine (in bleu) and annealed
at 150°C for 11 days (in red) mixture of LYBC and LPSCl in a 1:1 mass ratio. In part (a), green bands highlight the
evolution of two new reflection that can be attributed to LiCl.

66
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

Overall, the observed continuous decay upon cycling and the aforementioned SE
oxidation above 3.6 V vs. Li+/Li (Figure 2.15b), suggests a mild incompatibility in the cathode
composite itself, which is likely the cause of the increased polarization upon cycling, as
evidenced in the Appendix Figure A2.5.

o The low pressure performance of Li3YBr2Cl4-based composites

As mentioned previously, cycling at extremely high pressures (> 100 MPa) is not a
scalable option and is completely incompatible with the highly soft and ductile lithium metal
anode. Therefore, we investigated the pressure-dependent cycling performance of this new
system. In that context, a “full-halide” solid-state NMC//LYBC//LiIn cell with a loading of
4.6 mgNMC/cm² was assembled and, similarly to the previous study in pressure of the
Li3InCl6-based composite, was cycled at successively decreasing stack pressure from 10 MPa
to 0.1 MPa (= 1 bar). The results are presented in Figure 2.19a and b. Remarkably, at 10 MPa,
this system displayed capacities of 179.6 mAh/g and 150.2 mAh/g in charge and discharge,
respectively, with a high ICE of 84% and a low polarization of 37 mV. Overall, the first cycles
at 10 and 100 MPa (Figure 2.16a) show very similar performance. The pressure was then
gradually decreased down to 0.1 MPa in subsequent cycles and apart from the continuous
degradation, no significant drop in capacity was observed. Moreover, as shown by the filled
dots in Figure 2.19d, the majority of the capacity decay occurs at the cycle following to
pressure changes where the decay rate is higher than the previously determined standard rate
of -1%/cycle. As mentioned for the Li3InCl6 system in Figure 2.12b, this is due to the cell
manipulation required to change pressure. These findings indicate that it is possible to achieve
pressures below 1 MPa although the observed decay should be addressed by choosing
another SE or introducing coating for example.

67
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

200
a 1st cycle b
3.5

Charge Capacity (mAh/g)

Voltage (V vs. Li+/LiIn)

150

3.0
100
C/20
10 MPa 5 2 1 0.5 0.2 0.1
RT
2.5
50
Cycling pressure (MPa) Charge Capacity (mAh/g)
0.1 10 Discharge Capacity (mAh/g)
2.0 0
0 50 100 150 200 0 5 10 15 20 25
Capacity (mAh/g) Cycle number
-6
c

Discharge capacity decay rate (%)

d Decay rate of each cycle
Decay at pressure change
-4

-2

0
10 MPa 5 2 1 0.5 0.2 0.1
Li-In alloy 0 5 10 15 20 25
counter electrode Cycle number
Figure 2.19. (a) Last galvanostatic cycles at each pressure, (b) capacity retention and (d) decay rate of
NMC//LYBC//LiIn cell as a function of the stack pressure (from 10 to 0.1 MPa), cycled at RT and a C/20 rate. (c)
Schematic of the “full-LYBC” cell architecture tested in this experiment. Filled circles in (d) highlight the larger
capacity drop after each pressure change.

At this stage, we have demonstrated that reducing the pressure has little effect on the
capacity retention of such composite at a low loading of 4.6 mgNMC/cm² but other figures of
merits such as the performance as a function of the pressure, loading, and temperature must
be determined.

First the rate capabilities as function of the pressure was estimated (Figure 2.20a). We
assembled cells with loadings of approximately 5 mgNMC/cm² and conducted cycling
experiments according to the following protocol. The first formation cycle was carried out at
C/20 under a pressure of 10 MPa and for subsequent cycles, the pressure was decreased and
the discharge rate capability was monitored. Rate capability measurements were conducted
following by varying the discharge rate from C/20 to 5C, while the charging rate was
maintained in Constant Current Constant Voltage (CCCV) mode at C/10 during the CC phase
and down to C/50 during the CV phase. Three cycles were performed at each rate, and the
average discharge capacity was used to evaluate the rate capability.

68
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

The reduction in pressure resulted in a slight decrease in reversible capacity,

particularly at higher discharge rates, but showed negligible changes for pressures below
10 MPa. Additionally, after measuring the rate capability, the cells were cycled at C/20, and
the calculated capacity decay rate (Appendix Figure A2.6) displayed no clear trend as a
function of pressure, with an average decay rate of around -1%/cycle. These results align with
previous tests conducted at 100 MPa (Figure 2.16c) and confirm our hypothesis that such
halide SE can be used for very low-pressure cycling.

Subsequently, evaluating the rate capabilities as a function of loading, we employed

the same cycling protocol but with a stack pressure of 100 MPa. The loading conditions were
systematically varied between 5 and 25 mgNMC/cm², with the goal of achieving loadings of
5 mAh/cm² for future industrial applications of ASSBs.6 Figure 2.20b clearly illustrates the
significant influence of loading on the rate capabilities with the capacity decreasing from 125
to 86, 25, and 16 mAh/g at 1C when increasing the loading from 5 to 12, 20, and 25 mgNMC/cm².

Additionally, we examined low-pressure performance of this system as a function of

loading. NMC//LYBC//LiIn solid-state cells with loadings of up to 25 mgNMC/cm² were cycled at
C/20 under a relatively low pressure of around 9 MPa (Figure 2.20d and e). Increasing the
loading results in a linear increase in the polarization from 41 mV at a loading of
4.6 mgNMC/cm² to 158 mV at a loading of 25 mgNMC/cm². Remarkably, there is no significant
drop in the first cycle reversible capacity, which remains stable at around 145 mAh/g at this
low rate.

Lastly, we studied the performance of such cells over the 0 to 70°C temperature range.
Anticipating degradations at high temperature that could affect rate capability (Appendix
Figure A2.7b), we have changed our cycling protocol and adopted the "signature curve"
technique. Briefly, it consists in discharging the cell at progressively lower currents to a fixed
discharge cut-off voltage, so that the resulting cumulative capacities are indicative of the
material's rate capabilities. A more detailed explanation of this procedure can be found in the
Appendix Section A2.1 and Appendix Figure A2.1. The results, displayed in Figure 2.20c,
showed reversible capacities of 165 and 169 mAh/g at a C/20 rate under 50°C and 70°C,
respectively. Even at a high rate of 5C, the reversible capacities reached 135 and 138 mAh/g,
demonstrating remarkable rate capabilities at high (70°C) and even moderate (50°C)

69
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

temperatures. Decreasing the temperature resulted in a drop in reversible capacities, reaching

146, 144, and 133 mAh/g at C/20 under 25°C, 10°C, and 0°C, respectively.

a b c
Discharge Capacity (mAh/g)

150

100 RT RT 5 mgNMC/cm²
5 mgNMC/cm² 100 MPa 100 MPa
100 MPa 5 mgNMC/cm² 70°C - Signature
50 10 MPa 50°C - Signature
12 mgNMC/cm²
5 MPa 25°C - Signature
20 mgNMC/cm²
2 MPa 10°C - Signature
1 MPa 25 mgNMC/cm² 0°C - Signature
0
0.1 1 0.1 1 0.1 1
Discharge rate (nC) Discharge rate (nC) Discharge rate (nC)

d e

Reversible capacity (mAh/g)

3.5 150 150
Voltage (V vs. Li+/LiIn)

Polarisation (mV)

3.0 RT, C/20 100 100

9 MPa
4.5 mgNMC/cm²
2.5 12 mgNMC/cm² 50 50
20 mgNMC/cm²
Polarisation
25 mgNMC/cm² Reversible capacity
2.0 0 0
0 50 100 150 200 0 5 10 15 20 25
(Q-Qo) (mAh/g) Loading (mgNMC/cm²)
Figure 2.20. Rate capabilities of NMC//LYBC//LPSCl//LiIn cells as a function of (a) stack pressure, (b) loading
and (c) temperature. (d) First galvanostatic cycles and (e) first cycle polarization and discharge capacity as a
function of the loading, cycled at RT, a stack pressure of 9 MPa and a C/20 rate.

Overall, the low-pressure performance of our halide-based systems highlight the

potential of halide SEs for real-world applications although being still limited in terms of
achievable loading and temperature due to the relatively low ionic conductivity of 1.27 mS/cm
of the LYBC. Comparing the performance of our system to that of LPSCl-based composites
shown in Figure 2.9, it becomes evident that, in addition to ionic conductivity of the SE, other
mechanisms must be at play in the improved performance under low-pressure conditions.

o Discussion on the reason underlying the enhanced low-pressure cyclability of

halides over Li6PS5Cl solid electrolyte

At this stage, a legitimate question regards why cathode composite based on halide
SEs exhibit enhanced low-pressure cyclability. There are several reasons for this that may be
related to the physical, electrochemical, chemical or mechanical properties of the different
components or even from their interaction within the cathode composite. Among them and

70
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

in light of our early discussions (Section 1.3), the ability of the SE material to mitigate strain
and to accommodate volume changes in the cathode composite while maintaining adequate
percolation even at low pressure seems essential.

To explore the issue of percolation in greater depth, we have performed scanning

electron microscopy (SEM) imaging and Energy dispersive X-ray (EDX) measurements on
composite powder and densified pellets in order to assess the homogeneity and the particle
sizes of the different components. Measurements were performed in the Laboratoire de
Réactivité et Chimie des Solides (LRCS) with the help of Carine Davoisne. The detailed methods
can be found in Appendix Section A2.1. Three different composites prepared by hand grinding
of composition in NMC622, SE and VGCF of 66.5, 28.5 and 5% were imaged through these
techniques. Two of them contained LPSCl electrolytes, more specifically the coarse SS-LPSCl
and the fine LPSCl used in Section 2.2 and the third one incorporated the Li3InCl6 prepared
through a water-based synthesis. Regrettably, due to the high sensitivity to air and moisture
of such SEs, we were compelled to outsource these measurements and due to time
constraints, we were unable to conduct a parallel study on composites based on LYBC.
Analysing the cross-section images presented in Figure 2.21 as well as the EDX mapping of the
single elements in Appendix Figure A2.8, it is evident that the SS-LPSCl composite
(Figure 2.21a and d) displays regions of large SE particles while concurrently displaying regions
devoid of it. This observation correlates with a poor ionic percolation, which provides an
explanation for the inferior performance of SS-LPSCl-based composites under low-pressure
conditions. In contrast, both the fine LPSCl and LIC composites appear to exhibit a higher level
of homogeneity. Based on these results, we concluded that particle sizes and homogeneity of
the composites alone are insufficient to fully account for the decrease performance of fine
LPSCl compared to LIC under low-pressure conditions. Consequently, it is imperative to
explore other contributing factors; namely mechanical properties, as hypothesized earlier.

71
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

Figure 2.21. Cross-section SEM imaging (a-c) and EDX mapping (d-f) of bilayer pellets made of a SE substrate
and the cathode composites. Parts (a) and (d) correspond to SS-Li6PS5Cl, (b) and (e) to fine Li6PS5Cl and (c) and
(f) to Li3InCl6-based cathode composites. Green and yellow coloured particles corresponds to SE elements (P, S,
Cl and In), pink/red coloured particles corresponds to NMC elements (Ni, Mn, Co, O) and blue regions to VGCF.

To interrogate the mechanical properties of the electrolyte, microindentation

measurements were conducted with the help of Saint-Gobain Recherche Paris. The results are
illustrated in Figure 2.22b. Notably, LYBC exhibits a hardness of 0.302 ± 0.017 GPa, value
significantly lower than that previously reported for sulphide or oxide-based SE (Figure 2.22a).
Furthermore, it is a hardness value one order of magnitude lower than that reported by
Bruce’s group with their Li3InCl6 halide prepared by ball-milling, although they used a
nanoindentation technique and a different synthesis route that can make the comparison not
straightforward.121 A likely explanation to this contrast in hardness values could be nested in
the different chemistry of the compounds, in particular in the choice of the Y cation and Br
anion. Their larger ionic radius can lead to a more isotropic charge distribution, which lower
the pinning of defects, hence favouring an increase in plasticity, as shown in Figure 2.22b
where the hardness diminishes with an increase in the proportion of Br content. This low
hardness can be advantageous during different phases of the assembly process. Firstly, since
it is generally associated with lower stiffness,128 a lighter load will have to be used during
composite preparation process to deform the SE particles in contact with AM as illustrated in
Appendix Figure A2.9 . This, in turn, enhances the contact area and promotes the uniformity

72
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

of the composite without causing damage to its other constituents. Secondly, during assembly
at the densification step, the SE is likely to be denser thus decreasing porosity in both the
cathode composite and separator and enabling a good percolation without the need of a
sintering step.

a 0.4
b Pellet densified at 300 MPa
15

LYC

Hardness (GPa)
Hardness (GPa)

0.3
10
LYBC
LYB
0.2

5
0.1

0 0.0
C l O P C 0 1 2 5 6
S T G B
LP -LLZ LA LY x in Li3YBrxCl6-x
HP
Figure 2.22. Hardness of different SE from the literature and from our study. (a) Hardness of typical SE (adapted
from literature87 for LPSCl, HP-LLZO and LATP) and (b) hardness measurement by microindentation on pellet
fabricated at 300 MPa as a function of the bromine content in Li3YBrxCl6-x compounds.

In summary, we believe that the material's low hardness plays a pivotal role in
facilitating low-pressure cycling. Nonetheless, acknowledging the potential impact of other
factors, specifically differences in chemical or electrochemical compatibility between
oxide/halide and oxide/sulphide materials is important as the formation of a passivating CEI
layer can lead to increase stress and/or contact losses at the interface between the CAM and
the SE. In addition, one cannot disregard the potential significance of the adhesion and friction
properties of the SE and CAM. Adequate adhesion could significantly help in accommodating
volume changes. However, determining adhesion properties is not straightforward and a lack
of literature on that subject is clear.

o A first step towards larger pouch cells

We have successfully demonstrated the low-pressure performance of halide-based SE

in cathode composites for ASSB using laboratory-scale equipment with 8 mm diameter cells.
The ability to cycle down to 0.1 MPa now allows us to pursue scalability and explore larger cell
sizes. In this regard, we have undertaken experiments with a split-body cell of 13 mm in

73
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

diameter with a Mylar lining (as illustrated in the Figure 2.23a), which facilitates the
unmoulding of the stacked pellet after densification. However, to achieve uniformity over a
larger surface area, we adopted a film-based approach for the separator. Specifically, we
began by fibrillating LYBC with 0.2 wt.% polytetrafluoroethylene (PTFE) binder by mixing and
then rolling and folding the powders using a glass tube until a self-standing membrane was
formed. Subsequently, from this membrane, a 13 mm diameter disk with a weight of
approximately 150 mg was cut and pressed at 100 MPa within the cell. To address the known
decomposition issue of PTFE when in contact with lithium or LiIn, a thin layer of LPSCl
argyrodite was then spread and pressed at 100 MPa onto the LYBC separator before
introducing the LiIn:LPSCl counter electrode. The working electrode, composed of
NMC622:Li3InCl6:VGCF in a 66.5:28.5:5 wt.% ratio, was then added and the whole stack was
densified at 400 MPa to complete the assembly. Finally, the stack was unmoulded, placed into
a pouch cell with aluminium disk and tabs for electric connection and sealed under vacuum
(Figure 2.23b).

Figure 2.23. (a) Annotated picture of the split cell used for assembly and (b) pouch cell with electric tabs.

Subsequent to the fabrication process, we conducted galvanostatic cycling, and the

pouch cell results are illustrated in Figure 2.24a. Note that cycling under atmospheric
pressure, equivalent to 0.1 MPa, yielded mediocre results with capacities of only 52 and
33 mAh/g on first charge and discharge respectively, with a rapid decay over cycling. We
attributed these poor performances primarily to contact issues with the current collectors

74
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

employed in our setup. As an attempt to mitigate this limitation, we also have used different
cell configurations, namely a large (2 cm) Swagelock-type cell and a 2032 coin cell. The results,
shown in Figure 2.24b and c, respectively, exhibit improved performance, presumably
attributed to the better contacts with the current collectors facilitated by the use of higher
pressure in these configurations. However, in the coin cell test, some instabilities were
observed on the galvanostatic profile, due to pellet damage caused by the crimping pressure
applied in the coin cell assembly process.

a b c
Voltage (V vs. Li+/LiIn)

Voltage (V vs. Li+/LiIn)

Voltage (V vs. Li+/LiIn)

3.5 3.5 3.5

3.0 3.0 3.0

2.5 Pouch cell 2.5 2.5

C/20, RT Coin cell Swagelock cell
4.5 mgNMC/cm² C/20, RT C/20, RT
2.0 2.0 4.5 mgNMC/cm²
2.0 4.5 mgNMC/cm²
Atm. pressure
0 50 100 150 0 50 100 150 0 50 100 150
(Q-Qo) (mAh/g) (Q-Qo) (mAh/g) (Q-Qo) (mAh/g)
Figure 2.24. (a-c) Galvanostatic cycling and (d-f) capacity retention of a LIC-based composite in 13 mm large
cells cycled (left to right) in either a pouch cell under 1 bar, a coin cell or a larger Swagelock-type cell with a
small spring inside.

To enhance performance, possible directions include the utilisation of a taped cathode

and anode film or the adoption of an isostatic pressing process, among other strategies. In
summary, while these results are promising, they underscore the need for further
optimisation of the assembly procedure for such cell types.

2.5 – Attempting stable cycling of “full-halide” ASSB

In this chapter, we have seen two halide-type SEs being used in cathode composites.
However, in the first case, Li3InCl6 (LIC), although demonstrating high oxidation stability,
showed poor reduction stability thus necessitating the addition of an LPSCl protective layer
with the counter electrode. This layer revealed highly detrimental to the cell stability due to
the chemical reactivity of LPSCl and LIC. Secondly, employing a different halide, the LYBC SE,
we obtained similar performance in low pressure however, in this case, although being stable
with LiIn, its poor oxidation stability led to a decay over cycling due to the reactivity in the
cathode. Therefore, we now attempted to solve this issue through different strategies.

75
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

o First strategy: coated NMC particles

The first strategy consists in using a different NMC622 that incorporate nanometric
coatings. Materials and detailed protocol for the preparation of these composite can be found
in Appendix Section A2.1. These coatings have been demonstrated to be highly efficient in
preventing the decay in LPSCl-based cells and may be favourable in mitigating the degradation
observed in LYBC-based composites. We chose two different coatings: a zirconium and a
cobalt-based coating. Results, illustrated in Figure 2.25, clearly show an improvement in the
first cycle reversibility from 82.1% to 86.6% and 87.1% for the Zr and Co-coated materials,
respectively. However, the decay over cycling is still present and is not resolved by the use of
coatings NMC.

a b
3.5
Discharge Capacity (mAh/g)

C/20, RT 150
Voltage (V vs. Li+/LiIn)

12 mgNMC/cm²
100 MPa

3.0 100

2.5 50
Uncoated Uncoated
Zr-coated Zr-coated
Co-coated Co-coated
2.0 0
0 50 100 150 200 0 5 10 15 20
(Q-Qo) (mAh/g) Cycle number
Figure 2.25. Galvanostatic performances of coated NMC in LYBC-based composites. (a) First galvanostatic
profile and (b) capacity retention of the Zr-based (in green) and Co-based (in red) coatings compared to the
uncoated NMC (in blue) in NMC622:LYBC:VGCF (66.5:28.5:5 wt.%) composites.

o Second strategy: LIC and LYBC dual-SE architecture

With LYBC showing poor oxidation stability but compatibility with LiIn counter
electrodes and LIC demonstrating a high oxidation stability limit but instability at LiIn potential,
our second strategy involves creating a dual-SE cell architecture with LIC incorporated into the
cathode and LYBC employed in the anode, hence keeping the best of both worlds. An
NMC:LIC:VGCF//LIC//LYBC//LYBC:LiIn cell stack was therefore assembled and cycled. As
illustrated in Figure 2.26, incorporating such design allowed us to decrease the decay from
17.3% to 10% of decay after 20 cycles at C/20. In addition, employing the same configuration,
we prepared and tested a composite consisting of Zr-coated NMC622 with LIC.

76
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

a b

Discharge Capacity (mAh/g)

3.5 C/20, RT 150 -10%

Voltage (V vs. Li+/LiIn)

NMC : Li3InCl6 : VGCF

12 mgNMC/cm²
100 MPa -17.3%
3.0 100
Li3InCl6 Li3YBr2Cl4
Li0.5In : Li3YBr2Cl4

2.5 50
Full LYBC reference Full LYBC reference
Dual SE architecture Dual SE architecture
2.0 0
0 50 100 150 200 0 5 10 15 20
(Q-Qo) (mAh/g) Cycle number
Figure 2.26. (a) Illustration of this dual-SE architecture with LIC and LYBC SEs, (b) first galvanostatic cycle and
(c) capacity retention comparison of this system (dotted line) with the previous LYBC cells (solid line), both
comprising an uncoated NMC622. Cells were cycled at C/20, at RT under 100 MPa of pressure.

The NMC@Zr:LIC:VGCF composite was prepared through manual hand grinding in an

agate mortar with the same ratios as before. Subsequent testing of this composite yielded
even better performance at high pressure (100 MPa), as shown in Figure 2.27, achieving a first
cycle discharge capacity of 156 mAh/g with an ICE of 90%. The capacity retention was also
significantly enhanced, with a notably reduced decay rate, resulting in an 85% retention of the
initial discharge capacity after 50 cycles at C/20.

a 200
b 100
3.5 C/20, RT

Coulombic efficiency (%)

Charge Capacity (mAh/g)
Voltage (V vs. Li+/LiIn)

12 mgNMC/cm²
100 MPa 150
3.0 95
100

2.5
50 Charge Capacity (mAh/g)
90
Discharge Capacity (mAh/g)
Coulombic efficiency (%)
2.0 0
0 50 100 150 200 0 10 20 30 40 50
(Q-Qo) (mAh/g) Cycle number
Figure 2.27. (a) First galvanostatic cycle and (b) capacity retention of a Zr-coated NMC622 cycled in the
LIC/LYBC dual-SE cell architecture. Tests were performed at C/20, at RT under 100 MPa of pressure.

The improved performance in this configuration at high pressure is promising and

subsequent low-pressure tests should be carried out.

2.6 – Chapter conclusion

In conclusion, Chapter 2 has presented a strategy aiming at alleviating the pressure
limitation observed in the development of ASSBs. The chapter started by establishing a

77
 Chapter 2 – Atmospheric-Pressure Operation of All-Solid-State Batteries enabled by Halide SE

baseline reference for high and low pressure cycling, employing an argyrodite-based NMC
composite that is typically used in the field. This reference not only highlights the challenges
associated with low-pressure cycling but also strengthens the greater significance of
composite formulation under low-pressure conditions in achieving high battery performance.

Subsequently, we show that we could partially overcome these limitations by

switching to halides-based electrolytes that show more favourable electrochemical and
mechanical properties. By investigating the performances of the Li3InCl6 (LIC) SE in cathode
composites, we demonstrated its ability to enable low pressure cycling down to 0.1 MPa.
However, this advantage is negated by its chemical and electrochemical incompatibilities with
LPSCl and LiIn, respectively, hence calling for the use of another member of the halide SE
family having greater stability.

Along that line we selected a mixed halide Li3YBr2Cl4 (LYBC) phase, known for its
enhanced reduction stability and low hardness, and demonstrated its similar ability for low-
pressure cycling down to 0.1 MPa and high loading of up to 25 mgNMC/cm². Moreover, we
rationalise such findings based on materials hardness considerations as well as different
chemistry of the compounds, in relying on Y cation and Br anion. Altogether, this revealed that
further exploring the design space of halides stands as an interesting strategy.

Finally, we demonstrated that stable cycling in "full-halide" ASSBs can be achieved by

introducing the dual-SE architecture concept, not by pairing argyrodite and halide SEs, but by
pairing two different type of halides, LIC in the cathode and LYBC in the anode. Such a cell was
shown to deliver a capacity of 156 mAh/g with a notably improved retention compared to the
LYBC system.

In summary, Chapter 2 has opened up a promising route to low-pressure cycling of

ASSBs using a LiIn alloy as the anode. This achievement provided the impetus to continue our
journey towards the realisation of a complete solid-state battery operating at low pressure in
the presence of a Li metal anode.

78
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

Chapter 3 – Solid-Electrolyte-Free
Cathode Concept for Low-Pressure
Cycling of All-Solid-State Batteries1

1
This chapter includes the following publication: Hennequart, B. et al. Solid-Electrolyte-Free O3-LixTiS2 Cathode
for High-Energy-Density All-Solid-State Lithium-Metal Batteries. ACS Appl. Energy Mater. 6, 8521–8531 (2023).

81
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

3.1 – Chapter Introduction

Reducing the stack pressure in all-solid-state batteries (ASSBs), as discussed in the
preceding chapter, represents a significant challenge in the development, commercialisation,
and widespread adoption of this battery technology. Previously, we introduced a method to
address this challenge, which is centred on the optimisation of electrode components to
enable effective low-pressure cycling. However, the sole fact that the typical high
performance cathode active materials (CAMs), such as LiCoO2 or LiNixMnyCo1-x-yO2 (NMC) have
inherently limited electronic and ionic conductivities leads to the necessity of implementing
them within composite systems. Within these composite structures, the large amount of
interfaces created can increase the impact of chemical and electrochemical instabilities
between the different components. This issue is particularly pronounced in the case of
sulphide-based solid electrolytes (SEs), which are known for their poor stability and their
susceptibility to oxidation and decomposition at potentials exceeding 3 V vs. Li+/Li.43,67,73,129–
132 Moreover, another significant challenge stems from the intrinsic volume changes
experienced by the CAM particles during cycling, which we have seen can lead to poor
performance in low pressure conditions in the case of an unoptimised composite.92,133

In light of all these challenges, advantages of a SE-free electrode design could be

exploited. Composite electrodes (Figure 3.1a) mimic the wetting of a liquid electrolyte in
conventional lithium-ion batteries by providing ionic percolation in the cathode, whereas a
SE-free electrode (Figure 3.1b), predominantly composed of active material, relies solely on
interparticle ionic diffusion through the CAM particles. This approach effectively eliminates
most of SE/CAM interfacial challenges by containing them at the planar interface between the
cathode and SE separator and simultaneously contributes to enhancing the energy density.

Figure 3.1. Illustration of (a) a composite cathode and (b) a SE-free cathode design in an ASSB.

82
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

However, the choice of a suitable active material for a SE-free electrode necessitates
the fulfilment of several prerequisites. The materiel must possess adequate ionic and
electronic conductivities, exhibit chemical stability regarding the SE, and be mechanically
ductile enough to allow densification through a simple cold-pressing process. Typical lamellar
transition metal oxides materials, such as NMC, being oxides, tend to exhibit high hardness
and are therefore unsuitable for such design implementation. However, sulphide materials, in
contrast, are known for their greater ductility. Among them, the lamellar phase of TiS2 meets
all the specified requirements; it is a rapid ionic conductor 134, a semimetallic135 and exhibits
higher ductility and better stability with sulphide-based SEs than conventional oxide CAMs.

In this chapter, we first introduced the performance of the O1-TiS2 in this SE-free
electrode design, and thus confirm the performance reported in an earlier report 118 on this
material. Further exploiting this concept, we identified another titanium sulphide, the
O3-LixTiS2136–138 phase that met the above specifications. Through a milling process, this phase
enabled us to achieve a high energy density comparable to NMC cathodes while facilitating
low-pressure cycling. Additionally, we broaden our scope by extending this concept to a
diverse range of CAMs. We embark on two distinct pathways, that have consisted in exploring
i) other Ti-based sulphide polymorphs as well as ii) halide-based materials, taking advantage
of their low ductility for achieving low pressure operation.

3.2 – The O1-TiS2 case: a literature benchmark

Initially, we focused on the O1-TiS2 phase. This phase was shown, by Kim et
al.120,119,139,118, to display remarkable performance within a SE-free or "diffusion-dependent"
electrode configuration. Notably, it delivered most of its theoretical capacity at a high
electrode loading of 45.6 mg/cm², while operating at 60°C. This remarkable performance was
achieved through a straightforward milling process of the pristine material. It was
hypothesized by the authors that this milling process, by reducing particle size and altering
morphology, could effectively reduce the tortuosity within the electrode and therefore enable
successful cycling within the SE-free electrode design.

O1-TiS2 was obtained commercially from Sigma and is referred to as pristine TiS 2
hereafter. In order to reproduce the reported protocol, the pristine material was subsequently
milled for 90 minutes in a high-energy milling device. Figure 3.2 summarizes the

83
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

morphological, structural and electrochemical performance of the material before and after
grinding. As revealed in the SEM images displayed in Figure 3.2a and b, this process effectively
reduced the particle size from tens of microns to less than 1 µm and changed the morphology
from platelets to a shapeless particle structure. Moreover, the X-ray diffraction (XRD) analysis,
as shown in Figure 3.2c, reveals a reduction in crystallinity and an increase in the stresses
within the particles as evidenced by the disappearance of certain reflections and the
broadening of others. In terms of electrochemical performance, both pristine and ball-milled
materials were subjected to testing within the SE-free electrode design (Figure 3.2d). Although
not reaching its theoretical capacity of 239 mAh/g, the ball-milled TiS2 exhibited enhanced
reversibility, recovering 0.76 lithium (183 mAh/g), compared to the pristine material which
returned only 0.05 Li (11 mAh/g) on the first charge, hence confirming the earlier report.118

2.4
c (001)
Pristine TiS2 d Pristine TiS2
Ball-milled TiS2 Ball-milled TiS2
Voltage (V vs. Li+/LiIn)
Normalized intensity

2.0

(002)
1.6
Be
(101) (102)
0.01
1.2

0.8
Normalized to (101) Be reflection

10 20 30 40 0.00 0.25 0.50 0.75 1.00

2q (°) x
Figure 3.2. Characteristics of the pristine and milled O1-TiS2. (a) SEM micrographs, (b) X-ray diffractograms with
crystallographic structure and (c) first galvanostatic cycle at C/30 in a SE-free cathode of the pristine and ball-
milled O1-TiS2, respectively in blue and orange.

Through galvanostatic intermittent titration technique (GITT, Figure 3.3), we

determined a significantly higher apparent Li+ diffusion coefficient in the milled TiS2 sample
compared to the pristine material. This finding provides additional support for the proposed
"diffusion-dependent" electrode model where a decrease in tortuosity of the electrode
enables the SE-free electrode to cycle in a solid-state cell.

84
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

3.5
a Pristine TiS2 1E-8 b

Diffusion coefficient (cm²/s)

Ball-milled TiS2
3.0
Voltage (V vs. Li+/Li)

1E-9

2.5
1E-10

2.0
1E-11

1.5
1E-12
0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.5 1.0 1.0 0.5 0.0
x x in LixTiS2
Figure 3.3. (a) Voltage profiles and (b) apparent diffusion coefficient of both pristine and ball-milled O1-TiS2
calculated from the GITT measurement at different state-of-charge (SOC) in a SE-free cathode system.

The validation of Kim et al. work has motivated us to expand the range of materials
suitable for such configurations.

3.3 – O3-LixTiS2: A Pre-lithiated lamellar phase enabling low pressure

cycling
We conducted further exploration of members of the Ti-based sulphide chemistry,
focusing primarily on the metastable O3-LixTiS2. This phase was synthesized from a high
temperature (800°C) solid-state synthesis by reacting stoichiometric amounts of thoroughly
mixed Ti and Li2S powders in evacuated quartz tubes. This phase exhibits a comparable
lamellar structure to the previously studied O1-TiS2, as illustrated in Figure 3.4.

Figure 3.4. Crystallographic structure of the O3-LixTiS2 phase.

o Characteristics and performances in typical composite cathodes

The formation of the O3-LixTiS2 phase, hereafter referred as LTS, was confirmed
through XRD, as shown in Figure 3.5a and, similar to the O1 phase, the resulting powder

85
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

consists of large platelets measuring 30 to 50 µm as presented in Figure 3.5b. Its

electrochemical signature is somewhat similar to the findings reported by Colbow et al.138 in
that it exhibits a step voltage change that can be attributed to an O3 to O1 structural phase
transition. Through an operando XRD measurement (Appendix Figure A3.2 and insets in
Figure 3.5c), we also confirmed the irreversible disappearance below x=0.5 in the first charge
of the (104) reflection of the O3 phase at 36° to the benefit of the (101) reflection of the O1
phase at around 34°.

x 104
44 a (003) * Impurities b 3.0
O1 O3 c

Voltage (V vs. Li+/Li)

Intensity (counts)

2.5
(006) (104)

2 2.0
Be
O3 (104)

Counts

Counts
O1 (101)

(101) (009) 1.5

1
** (012) (015)
*
1.0 34 36 34 36
2q (°) 2q (°)
0
10 20 30 40 0.4 0.6 0.8 1.0
2q (°) x in LixTiS2
Figure 3.5. Characteristics of the O3-LixTiS2 phase. (a) XRD pattern, (b) SEM image showing the plate shape of
the pristine O3-LixTiS2, (c) first galvanostatic cycle in a liquid system and inset of operando XRD measurements
showing the structural phase transition

This material also displays a high electronic conductivity of 13.28 S/cm determined
through chronopotentiometry (Figure A3.3a). This high electronic conductivity eliminates the
need to add an electronic conductive additive in the cathode for the subsequent
electrochemical tests. Recognising that structural changes in materials can generally have a
negative impact on cell performance, we conducted our initial electrochemical tests by
limiting the upper cut-off voltage to 2.6 V vs. Li+/Li. The material was tested both in liquid cell,
with LP30 electrolyte and a Li metal anode, and in solid-state cells with a LPSCl-based cathode
composite, a LPSCl separator and a lithium-indium (LiIn) alloy composite anode. The
composite preparation and assembly procedure are detailed in the Appendix Section A3.1.
The initial discharge results, shown in Figure 3.6, revealed similar reversible capacities for both
cells, reaching 134 mAh/g and 142 mAh/g, for the liquid and solid-state cells respectively.
However, the latter exhibited a faster capacity decay, reaching 80% of its initial capacity after
20 cycles, while the other cell maintained a nearly constant reversible capacity with only a 5%
decay observed during the first two cycles. Moreover, taking into account the composite
nature of the solid-state cell cathode, the capacity in relation to the total electrode mass,

86
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

represented by the red dots in part d of Figure 3.6, is greatly reduced due to the
electrochemically inactive weight associated with the 30 wt.% of SE in the composite. To
mitigate this decrease in specific capacity, we naturally opted to leverage the SE-free electrode
concept.

0.4 0.6 0.8 1.0 1.2 0 5 10 15 20

250
2.6 1 10 Cycle 1

Discharge Capacity (mAh/g)

100

Coulombic efficiency (%)

Cycle 2
200
Voltage (V vs. Li+/Li)

Cycle 5
2.4
Cycle 10
150
2.2
Liquid cell 95
100
2.0
50
1.8
a 0
b 90
250
2.0 1 10 Cycle 1
Discharge Capacity (mAh/g)

100

Coulombic efficiency (%)

Cycle 2
Voltage (V vs. Li+/LiIn)

200
Cycle 5
1.8
Cycle 10
mAh/gactive material
150
1.6 Solid-state cell
-30% 95
100
1.4 mAh/gcomposite

50
1.2
c 0
d 90
0.4 0.6 0.8 1.0 1.2 0 5 10 15 20
x in LixTiS2 Cycle number
Figure 3.6. Cycling performance in liquid and composite ASSB. (a-c) Galvanostatic cycling and (b-d) capacity
retention and coulombic efficiency of the pristine O3-LixTiS2 in liquid (top) and solid-state (bottom) cells with
30 wt.% LPSCl SE in the cathode. Red dots in (d) correspond to the capacity normalized to the total electrode
composite mass.

o Enabling SE-free cycling of O3-LixTiS2 through a milling strategy

Utilising the SE-free electrode design requires sufficient ionic diffusion throughout the
entire LTS electrode. In this regard, we investigated the performance of the material within
composites with decreasing amounts of SE, ranging from 30 to 0 wt.%. The results presented
in Figure 3.7 clearly indicate that reducing the SE content to 15 wt.%, while maintaining a
similar first cycle discharge capacity of 145 mAh/g, results in an increased polarization and
reduced retention. In addition, completely removing the SE and cycling with this pristine LTS
in a SE-free electrode configuration leads to a poor reversible capacity of only 37 mAh/g and
high polarization even at this very low rate of C/60 (Figure 3.7d). This demonstrates the limited

87
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

apparent diffusion of Li+ ions, similar to what was observed with O1-TiS2. Consequently, we
then applied the same milling approach to enhance the performance in this SE-free electrode.

a
+ LTS : Li6PS5Cl
LTS (100%) +
Li6PS5Cl
– LiIn : Li6PS5Cl –

30% SE 15% SE 0% SE

1 10
2.0 1 10 2.0 1 10
Voltage (V vs. Li+/LiIn)

Voltage (V vs. Li+/LiIn)

Voltage (V vs. Li+/LiIn)

Cycle 1 Cycle 1 2.0 Cycle 1
Cycle 2 Cycle 2 Cycle 2
1.8 Cycle 5 1.8 Cycle 5 1.8 Cycle 5
Cycle 10 Cycle 10 Cycle 10
1.6 1.6 1.6

1.4 1.4 1.4

1.2 b 1.2 c 1.2 d

0.5 0.7 0.9 1.1 0.5 0.7 0.9 1.1 0.5 0.7 0.9 1.1
x in LixTiS2 x in LixTiS2 x in LixTiS2
Figure 3.7. (a) Illustration of the composite and SE-free electrode design with pristine LTS active material. (b-
d) Galvanostatic performance in ASSB cells of the pristine LTS in electrodes incorporating 30, 15 and 0 wt.% of
SE, respectively. Cycling was performed at RT under a pressure of 100 MPa and a C/60 rate.

In order to assess the evolution of the properties and performance as a function of the
milling time, four samples were subjected to grinding under the same conditions for 10, 20,
60, and 90 minutes, and are hereafter referred to as BM10, BM20, BM60 and BM90-LTS,
respectively. SEM micrographs, presented in Figure 3.8a and b, show a significant change in
particle size and morphology with increasing grinding time, transitioning from a plate shape
of about 50 µm to shapeless sub-micrometric particle. XRD patterns obtained from the four
samples (Figure 3.8c) show an increase in their amorphous character due to milling. This is
evident from the gradual decrease in amplitude and increase in broadening of the (003)
reflection, along with the complete disappearance of both the (006) and (009) peaks after only
10 minutes of grinding. The reduction in particle size and the increase in strain typically
contribute to peak broadening.140 Due to the featureless nature of the XRD powders patterns
for our ground material, we could not estimate by the Williamson-Hall analysis the
compression-extension strains in the material. Therefore, considering that only the crystallite

88
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

size is involved, we can approximate a reduction in size from several hundreds of nanometres
in the pristine sample to 15 nm in the BM90-LTS sample using the Sherrer formula.

In addition, we observe a slight decrease in electronic conductivity from 13.28 to

4.67 S/cm upon milling (Appendix Figure A3.3). This value remains significantly higher than
the ionic conduction values of Li+ ions in the LTS material or the SE itself, which typically
reaches up to 6 · 10-3 S/cm in Li6PS5Cl. Based on these findings, we can conclude that electronic
conductivity is not a rate-limiting factor in our materials and that the use of carbon additives
remains unnecessary even after milling.

Solid-state batteries were then assembled with the different milled materials within a
SE-free cathode configuration, and the cells were cycled over the voltage window 1.8 to 2.6 V
vs. Li+/Li at a low current rate of C/60. The voltage-composition curves obtained are shown in
Figure 3.8d. Interestingly, the discharge capacity increases progressively from 37 mAh/g for
the pristine LTS to 123, 141, 157, and ultimately 163 mAh/g for the BM10, BM20, BM60, and
BM90-LTS samples, respectively. In addition, as grinding time increased, we observed lower
polarization, leading to higher energy efficiency.

10 15 20 25 30 35 40 45 0.4 0.6 0.8 1.0 1.2

2.0
a c 003 006 104 Be d Pristine

009 1.6
Pristine
1.2
2.0 Ball-milled 10 min
Ball-milled 10 min 1.6
Voltage (V vs. Li+/LiIn)
Normalized intensity

1.2
2.0 Ball-milled 20 min
Ball-milled 20 min 1.6
b 1.2
2.0 Ball-milled 60 min
Ball-milled 60 min 1.6

1.2
2.0 Ball-milled 90 min

Ball-milled 90 min 1.6

Normalized to (101) Be reflection 1.2

10 15 20 25 30 35 40 45 0.4 0.6 0.8 1.0 1.2
2q (°) x in LixTiS2
Figure 3.8. Micrographs of (a) the pristine and (b) the BM90-LTS, (c) XRD patterns and (d) first galvanostatic
cycle evolution with milling time of the O3-LixTiS2 as a SE-free electrode in solid-state cells at C/60.

89
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

For comparison, we tested the same samples in a liquid system (Figure 3.9). They
exhibit similar discharge capacities regardless of the milling time, indicating that grinding the
material does not introduce new Li+ sites that would otherwise increase the total capacity of
the material. Moreover, this very different behaviour compared to solid-state cells further
confirms the positive effect and the necessity of the milling process in achieving the full
capacity when using SE-free electrodes based on such layered compounds.

200
a b
2.50

Discharge capacity (mAh/g)

150
Voltage (V vs. Li+/Li)

2.25
100

Pristine LTS Pristine LTS

2.00 BM10-LTS BM10-LTS
50
BM20-LTS BM20-LTS
BM60-LTS BM60-LTS
BM90-LTS BM90-LTS
1.75 0
0.4 0.6 0.8 1.0 1.2 0 5 10 15 20
x in LixTiS2 Cycle number
Figure 3.9. Performance of the pristine and milled samples in liquid-based cell. (a) First galvanostatic cycle and
(b) capacity retention for different milling time with lithium metal anodes and a C/20 rate in the limited
electrochemical window 1.8 to 2.7 V vs. Li+/Li.

Thus far, we have demonstrated the potential of the SE-free cathode configuration.
However, we have limited the potential window to a safe range in order to avoid triggering
the O3-O1 structural change above 2.6 V vs. Li+/Li, which consequently limited the achievable
capacity. Therefore, we then expanded the cycling potential window from 1.8-2.6 to 1.4-3.2 V
vs. Li+/Li. The results, illustrated in Figure 3.10, demonstrate a notable enhancement in
capacity. Notably, the BM90-LTS can now deliver up to 226 mAh/g in a composite and
175 mAh/g in a SE-free positive electrode when operated at C/20. Interestingly, in contrast to
the pristine sample presented in Figure 3.5c, neither of the two systems exhibit the plateau at
higher potential associated with the structural change. This phenomenon can be attributed to
the increased amorphous nature of the sample induced by ball milling, as elucidated by the
cycling behaviour difference of a composite consisting of Pristine LTS:LPSCl (70:30) (Appendix
Figure A3.4), which clearly manifests the presence of the aforementioned second plateau. This
observation aligns with earlier findings in the literature141,142, which have consistently
reported the broadening or complete disappearance of such salient features in the voltage-

90
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

composition curves following a milling process. Moreover, comparison of the galvanostatic

curve between the composite and SE-free cell (Figure 3.10a and c), reveals a significant
distinction in the high-potential response. Specifically, the composite electrode exhibits the
ability to remove 0.4 Li during the initial charge, in contrast to the SE-free electrode, which
only removes 0.2 Li.

0.2 0.4 0.6 0.8 1.0 1.2 1.4 0 5 10 15 20

2.8 300 102
1 10 RT

Discharge Capacity (mAh/g)

Cycle 1

Coulombic efficiency (%)

C/20
Voltage (V vs. Li+/LiIn)

2.4 Cycle 2 100

Cycle 5
2.0 200
Cycle 10
98
1.6 Composite electrode
100 96
1.2
Composite electrode

0.8 94
2.8
a 0
b 102
1 10
Discharge Capacity (mAh/g)

RT Cycle 1

Coulombic efficiency (%)

Voltage (V vs. Li+/LiIn)

2.4 C/20 Cycle 2 100

Cycle 5 200
2.0
Cycle 10
98
1.6
SE-free electrode
100 96
1.2
SE-free electrode

0.8 94
c 0
d
0.2 0.4 0.6 0.8 1.0 1.2 1.4 0 5 10 15 20
x Cycle number
Figure 3.10. (a-c) Galvanostatic cycling and (b-d) capacity retention in the enlarged window of 1.4 to 3.2 V vs.
Li+/Li of BM90-LTS in (top) a composite electrode with 30% SE and (bottom) a SE-free electrode.

In addition, despite demonstrating higher capacities, the composite system

(Figure 3.10a) exhibits a more pronounced voltage slip, which can be linked to the previously
reported67 increasing parasitic activity of the LPSCl SE at these potentials. Such effect is further
exacerbated in this context due to the increased contact area between the SE and the CAM in
comparison to the SE-free system. To test this hypothesis, we conducted experiments by
combining the LTS with another SE in both the composite and the separator. In particular, we
selected the halide-type Li3YBr2Cl4 (LYBC) SE, of which we have demonstrated in Chapter 2
Section 2.4 an electrochemical stability between 3 V and 0.8 V vs. Li+/LiIn. Notably, as shown
in the differential capacity curves in Figure 3.11, using such halide SE lead to no additional
capacity at low potential (inset in Figure 3.11b).

91
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

500 500
a b
250 250

dQ/dV (mAh.g-1.V-1)
dQ/dV (mAh.g-1.V-1)

0 0

-250 Cycle 1 -250 Cycle 1

0 0
Cycle 2 Cycle 2
Cycle 5 Cycle 5
-500 Cycle 10 -500 Cycle 10
-50 Cycle 15 -50 Cycle 15
-750 0.8 0.9 Cycle 20 -750 0.8 0.9 Cycle 20
0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5
Voltage (V vs. Li+/LiIn) Voltage (V vs. Li+/LiIn)
Figure 3.11. Differential capacity plots of a LTS:SE (70:30) composite either with (a) SE=LPSCl or (b) SE=LYBC
solid electrolyte. Showing the parasitic reactivity of the argyrodite SE at low potential (in inset a) that its absence
in the case of the halide SE (inset b).

Altogether, we have demonstrated that the milling of the pristine material allows us
to access a considerably larger capacity in a SE-free electrode. However, we have seen a
discrepancy in the electrochemical response at high potential between the SE-free and
composite systems that may be attributed to a decrease in the apparent Li + diffusion
coefficient.

o Improved performance through lithium ion diffusion coefficient enhancement

To investigate this ionic diffusion hypothesis, we conducted Galvanostatic Intermittent

Titration Technique (GITT) measurements on both the Pristine and BM90-LTS samples. The
GITT results for the BM90-LTS (shown in orange in Figure 3.12b) reveal that, at high lithium
content, the apparent diffusion coefficient averages approximately 10-9 cm²/s. However, a
substantial decrease occurs, plummeting to 10-11 cm²/s when the lithium content drops below
x=0.6 during the charging process. Thus somewhat supporting our previous hypothesis and
explaining the lower reversible capacity observed for the SE-free cathode.

The evolution of the rate capability with increasing ball-milling time was further
examined using the "signature curve" method.143 A detailed procedure is available in the
Appendix Section A3.1. A representative discharge "signature curve" is presented in
Figure 3.12c and the cumulative capacity values at various discharge currents are plotted in
Figure 3.12b. The extension of ball-milling time resulted in an enhancement in performance
from ultimately yielding a peak discharge capacity of 200 mAh/g at a C/50 rate for the
BM90-LTS sample.

92
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

3.5
a Pristine LTS b

Diffusion coefficient (cm²/s)

BM90-LTS
1E-9
3.0
Voltage (V vs. Li+/Li)

2.5 1E-10

2.0
1E-11

1.5
1st charge 1st discharge
1E-12
0.4 0.6 0.8 1.0 1.2 0.6 0.6 0.8 1.0 1.2 1.4
x in LixTiS2 x in LixTiS2

c 5C C/200 200 d BM90

Cumulative capacity (mAh/g)

2.4
Voltage (V vs. Li+/LiIn)

BM10
1.6 100 Q1C

Q2C

Pristine Q5C
0.8
Q5C Q2C Q1C 0

0 50 100 150 200 0.01 0.1 1

Cumulative capacity (mAh/g) Discharge rate (nC)
Figure 3.12. (a) Voltage profiles and (b) diffusion coefficient of both pristine and ball-milled O3-LTS calculated
from the GITT measurement at different state-of-charge (SOC). (c) Typical galvanostatic response of a signature
curve at rates ranging from 5C to C/200 and (d) obtained rate capabilities in a SE-free electrode.

In summary, similarly to the O1-TiS2, the observed performance enhancement

following grinding can be attributed to both the reduction in particle size and morphological
changes that facilitate a more isotropic diffusion of Li+ through the electrode thickness and
lead to a lowered tortuosity, as illustrated in Figure 3.13. This interpretation is supported by
the results of the GITT (Figure 3.12a and b) conducted on both Pristine and BM90-LTS.

Figure 3.13. Schematic of the electrodes before and after ball-milling illustrating the decreased tortuosity.

93
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

However, it is important to note that an increase in the intrinsic diffusion coefficient

of the material with extended milling time cannot be ruled out. Consequently, in an attempt
to decouple the effects of apparent (i.e. at the electrode level) and intrinsic (i.e. at material
level) diffusion in our system, we conducted, with the help of Michaël Deschamps in the
CEMHTI laboratory, solid-state nuclear magnetic resonance (ss-NMR) measurements on both
pristine and ball-milled LTS samples and assessed the evolution of lithium ion mobility.

7Li Magic Angle Spinning NMR spectra were acquired for the pristine, BM20, BM60,
and BM90-LTS samples, as shown in Figure 3.14. Detailed methods and fitting parameters are
provided in Appendix Section A3.1 and Appendix Table A3.1, respectively. In the spectrum of
the pristine material (Figure 3.14i), two distinct components at chemical shifts of 12 ppm
(pink) and 2.1 ppm (blue) are observed, representing 49.3% and 50.7% of the total signal,
respectively. Interestingly, when compared to the narrower signal (pink), the broader signal
(blue) exhibits slightly more intense spinning sidebands, indicating greater anisotropy and
potentially reduced mobility. This discrepancy may signify that the broader signal arises from
Li+ ions with lower mobility, while the narrower signal is associated with more mobile ions.

Figure 3.14. 7Li MAS-NMR spectra of (i) pristine, (ii) BM20, (iii) BM60 and (iv) BM90-LTS. The experimental
spectrum is shown in blue and the model in red. Spinning sidebands are highlighted by stars. Model parameters
can be found in Appendix Table A3.1.

94
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

It should be recalled that similar narrow and broad NMR signals were previously
reported in the study of ball-milled nanocrystalline O1-Li2/3TiS2 samples using static 7Li
NMR.144,145 At the time, these observations were attributed to a rapid mobility of Li ions at the
interface for the narrow band and to a reduced mobility within crystalline domains for the
broader signal. To check the validity of this interface/crystalline ionic conduction hypothesis,
we employed a straightforward cylindrical geometry model (as illustrated in Figure 3.15),
which was previously utilised by Winter et al.144 In this model, calculating the surface-to-bulk
ratio is done using the following Equation 3.1.

(𝑑 + 𝑔)2 (ℎ + 𝑔) − 𝑑 2 ℎ
𝑓(𝑑, 𝑔, ℎ) = with 0.5 ≤ 𝑔 ≤ 1 𝑛𝑚 Equation 3.1
(𝑑 + 𝑔)2 (ℎ + 𝑔)
where g is the interface thickness, d the diameter and h the height of the particle.
Considering a typical interface thickness between 0.5 and 1 nm (corresponding to 2 to 4
atomic planes), a diameter and a height of 200 nm (minimum crystallite size deduced by XRD),
an interface fraction between 0.75 and 1.49% is obtained. Using the same equation, we found
that an interface thickness of 25 nm, rather than the typical 0.5 to 1 nm thick interface144, was
necessary to account for the 49.3% obtained by NMR. Such interface thickness is largely
improbable. Consequently, we believe that we are in presence of a two-phase system, which
could differ in terms of Li stoichiometry, similar to vacancies ordering in LixTiS2146, where the
broad component at 2.1 ppm (in blue) would stem from ordered lithium planes, and the
narrow component at 12 ppm (in pink) from a more mobile lithium phase. The change in 7Li
shift between the two phases may result from a difference in local electronic density and
mobility, which governs the paramagnetic, and Knight shifts.

Figure 3.15. Cylindrical model for surface-to-bulk ratio calculation to determine the interface ratio f calculated
with the Equation 3.1. Adapted from reference144.

95
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

Returning to the 7Li MAS-NMR measurements conducted as a function of the milling

time (Figure 3.14), intriguingly, a significant change in the relative proportions of the two
phases becomes evident following just 20 minutes of milling (as seen in Figure 3.14ii).
Specifically, there is an increase in the signal associated with the mobile lithium phase (pink),
which now contributes to 65% of the total signal, while the signal originating from the ordered
Li phase (blue) diminishes, accounting for only the remaining 35%. This change may imply a
reduction in vacancy ordering (i.e. increasing disorder) of the ordered blue phase component,
which in turn would increase the Li-ion mobility and hint towards an improved diffusion of the
ions after milling.

Concomitantly, longitudinal relaxation times were measured for the BM20 sample as
a function of temperature (Figure 3.16a) to investigate nanoscale dynamics. The spin-lattice
relaxation time (T1) for the BM20-LTS sample is found to decrease from 1.5 s to 350 ms over
the 25°C – 150°C range. Interestingly, increasing the temperature leads to the collapse of the
broad peak, which is attributed to the phase with vacancy ordering. This phenomenon
demonstrates a reversibility with temperature, akin to materials displaying a miscibility gap
behaviour. Assuming that 1/T1 is proportional to the site-to-site jump correlation time c in a
simple Bloembergen, Purcell and Pound approach (BPP)147, we could deduce an activation
energy of site-to-site jumps of about 128 meV. Moreover, through a 2D EXchange
SpectroscopY (EXSY) measurement with an exchange delay of 50 ms (Figure 3.16b), cross-
peaks are observed hence confirming that Li+ ions are diffusing between the two
environments at RT and justifying the observed signal ratio change with increasing
temperature.

96
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

Figure 3.16. (a) 7Li MAS NMR spectrum of BM20-LTS as a function of temperature. (b) 2D 7Li EXSY NMR
spectrum obtained for BM20-LTS with a 50 ms mixing time. The cross (off-diagonal) peaks indicate that lithium
ions are able to diffuse from the ordered environment to the mobile phase in the 50 ms timescale.

Finally, in the case of samples milled for 60 and 90 minutes (Figure 3.14iii and iv,
respectively), a new broad feature (represented in yellow) becomes apparent. This newly
observed broad component is likely a consequence of the amorphisation of O3-LixTiS2, which
intensifies disorder within the material, creating asymmetric lithium environments. Its signal
amplitude is estimated to account for 15.8% and 34.6% of the overall signal for the BM60 and
BM90 samples, respectively. This new component leads to a change in the amplitude of the
remaining mobile (pink) and ordered (blue) environment signals for BM60, which now
accounts for 50.8% and 33.4% of the overall Li signal, respectively. In contrast, BM90-LTS
displays a fundamental transformation, as the mobile lithium environment (pink) appears to
disappear in favour of a new broader signal (green) shifted to 7 ppm. This green signal
accounts for 24.7% of the lithium ions, while the main component (blue), now centred at
2.9 ppm, represents 40.7% of the total signal.

Additionally, the site-to-site jump activation energy for BM90 was retrieved from 7Li
longitudinal relaxation times measurements. These relaxation times were shown to decrease
from 353 ms to 186 ms as the temperature increased from 23°C to 151°C. From that, the

97
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

activation energy was determined to be 51 meV, which is notably lower than the value
measured for BM20-LTS (128 meV) and slightly lower than the one reported for amorphous
Li2/3TiS2 (70 meV) in a prior study.145 The decrease in the T1 value already implies a substantial
reduction in the jump time by a factor of 5 between the BM20 and BM90 samples, assuming
the same quadrupolar coupling constant. This suggests that a significant portion of the BM90
sample has become amorphous, thus favouring 3D diffusion as well as leading to a lower
activation energy for site-to-site jumps (e.g. lowering of diffusion barriers) due to both the
disappearance of vacancy ordering and crystallite boundaries modifications.

Altogether, our findings support the hypothesis that the amorphisation process
coupled with particle size reduction and morphology modification induced by ball milling plays
a pivotal role in facilitating isotropic diffusion at the electrode level. This is particularly crucial
for layered compounds when used in a SE-free system, where otherwise, a predominant 2D
Li+ diffusion pattern is observed. To test this hypothesis we will investigate, in a later
Section 3.4, the potential of employing other materials, such as the cubic spinel Ti 2S4 that
offers a 3D ionic diffusion in a SE-free cathode configuration.

o The low pressure performance of the SE-free electrode configuration

At this stage, a legitimate question concerns the performance of these SE-free ball-
milled electrodes under low-pressure conditions. In that regard, we assembled a solid-state
cell comprising a SE-free BM90-LTS electrode, with a loading of 7.76 mgAM/cm², and a
Li0.8In:LPSCl (60:40 wt.%) composite anode. Subsequently, in the same fixed-gap cell setup
used in the previous Chapter 2 Figure 2.4, the cell was cycled by successively reducing the
cycling pressure, starting from 10 MPa and progressively decreasing to 0.1 MPa.

From Figure 3.17, we observed that the cell exhibits a stable reversible capacity of
165 mAh/g even at 0.1 MPa, which corresponds to approximately 87.5% of the initial
discharge capacity obtained at 10 MPa. Moreover, upon returning to a pressure of 5 MPa, the
capacity remained stable at 172 mAh/g. Thus demonstrating that the impact of such low
pressure is relatively minor, resulting in only a slight 4% decrease in reversible capacity. The
same cell was subjected to subsequent cycling at various pressures and rates, as shown in
Figure 3.17c. Remarkably, it exhibited consistent performance, maintaining a stable capacity
of 129 mAh/g at 5 MPa and 112 mAh/g at 1 MPa when cycled at a C/5 discharge rate.

98
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

1st cycle
Voltage (V vs. Li+/LiIn)
2.5
a 200
b

Capacity (mAh/g)
2.0 150

10 MPa 5 2 1 0.5 0.2 0.1 5 MPa

1.5 100

1.0 50
Charge capacity (mAh/g)
Cycling pressure (MPa)
Discharge capacity (mAh/g)
0.1 10
0.5 0
0 100 200 0 5 10 15 20 25 30 35
Capacity (mAh/g) Cycle number

200 c 5 MPa 1 MPa

Capacity (mAh/g)

C/20
150 D/20 C/10
D/10
C/5 C/10
100 Study in D/5 D/5
C/10
pressure D/5 RT
7.76 mgAM/cm²
50
Charge capacity (mAh/g)
Discharge capacity (mAh/g)
0
0 50 100 150 200 250 300
Cycle number
Figure 3.17. (a) Galvanostatic cycling and (b) capacity retention of BM90-LTS in a SE-free electrode cycled at
successively lower pressure versus a LiIn anode. (c) Capacity retention at different rates and pressure of the
BM90-LTS in a SE-free electrode used for the cycling pressure study from 10 to 0.1 MPa. Blue zone in part
(c).corresponds to the study in pressure displayed in panel (b).

Next, we conducted tests on similar cells but having higher electrode loadings,
specifically up to 31 mgAM/cm² and 12.9 mgAM/cm² at pressures of 100 MPa and 10 MPa,
respectively. When comparing the first cycle of these cells (as shown in Figure 3.18), it is
evident that the cell with a higher loading exhibits lower capacity, which can be attributed to
the limited Li+ conduction. This effect becomes particularly pronounced at high state-of-
charge (SOC; for x<0.6), where an increased loading results in a significant loss of capacity,
which is consistent with the reduced apparent lithium diffusion coefficient at high SOC that
we observed earlier.

99
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

2.5
a 5.4 mg/cm²
2.5
b 4.5 mg/cm²
10.5 mg/cm² 7.76 mg/cm²
20.1 mg/cm² 12.9 mg/cm²
Voltage (V vs. Li+/LiIn)

Voltage (V vs. Li+/LiIn)

31.1 mg/cm²
2.0 2.0

QD1 = 211 mAh/g QD1 = 188 mAh/g

QD1 = 190 mAh/g
1.5 QD1 = 163 mAh/g
1.5 QD1 = 188 mAh/g
Region of low DLi Region of low DLi

1.0 RT QD1 = 134 mAh/g 1.0 RT

C/20 C/20 QD1 = 163 mAh/g
100 MPa 10 MPa
SE-free electrode system SE-free electrode system
0.5 0.5
0.4 0.6 0.8 1.0 1.2 1.4 0.4 0.6 0.8 1.0 1.2
x x
Figure 3.18. First galvanostatic cycle of SE-free cathode in ASSB cells as a function of the loading under (a)
100 MPa and (b) 10 MPa of stack pressure.

In summary, these results underline the feasibility of achieving stable cycling even
under low-pressure conditions, despite the inherent volume changes that occur during
cycling. This achievement can be attributed to the utilisation of the SE-free cathode
configuration, which enables homogeneous volume expansion and contraction across the
entire electrode thickness. Standing in contrast to composite systems, where only the CAM
contracts often resulting in contact losses with the SE.

To assess the contacts in our electrodes, we conducted ex-situ SEM characterisation

(Figure 3.19). We examined a BM90-LTS//LPSCl//LiIn stack in its pristine state, as well as after
charging to 2.6 V and discharging to 0.8 V vs. Li+/LiIn. The findings indicate a coalescence of
particles at the surface of the pellet upon lithiation (as observed in Figure 3.19e-g) and the
appearance of vertical cracks upon delithiation (Figure 3.19h-j) while interparticle contacts
remain preserved. Interestingly, a similar morphological evolution was reported by Meng’s
group117 when investigating the lithiation of an anolyte-free silicon anode for ASSBs at high
pressure. The fact that both LTS and Si exhibit a Li-driven morphological evolution suggests
that the feasibility of cycling LTS even at low pressure and Si only at high pressure is likely
more influenced by the difference in material ductility. Additionally, differences in the
amorphous nature between these materials could also play a role in their distinct behaviours
during lithiation.

100
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

Figure 3.19. (a) Galvanostatic profile and surface and cross section micrographs of pellets (b-c-d) pristine as-
densified, (e-f-g) discharged to 0.8 V and (h-i-j) charged to 2.6 V vs. Li+/LiIn.

For comparative purposes, Figure 3.20 displays the energy densities of solid-state cells
cycled at a C/20 rate, utilising various positive electrodes (sulphides and oxides), while
employing the same electrolyte and counter electrode (LiIn:LPSCl composite). Notably, the
cell utilising NMC622 delivers 530 Wh/kg at the material level, but, due to the necessity of
incorporating 30% of electrochemically inactive SE and carbon additives to ensure proper
operation, this translates to only 375 Wh/kg at the electrode level. This additional SE is not
required in our BM90-LTS SE-free electrode configuration thus leading to an energy at the
electrode level of 305 Wh/kg within the limited potential window of 1.8-2.6 V vs. Li+/Li, and

101
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

up to 366 Wh/kg (as shown in Figure 3.20) by enlarging the potential window to 1.4-3.2 V vs.
Li+/Li. In summary, these values are highly competitive with those obtained using NMC622/SE
electrode composites in ASSBs, underscoring that SE-free electrodes stand as a serious
contender for the next generation of solid-state batteries.

Figure 3.20. Obtained energy densities with a typical NMC622 material (cut-off at 4.2 V vs. Li+/Li) and the
BM90-LTS at different potential cut-offs in ASSBs.

Altogether, we have investigated the use of Li-containing Ti-based sulphides as a

promising cathode material for operating ASSB under low-pressure conditions. For proof of
concept, we have selected the O3-LixTiS2 phase and reported that this phase, when milled, can
be directly used as positive electrode in solid-state batteries as a SE-free cathode, thereby
achieving higher energy densities. Moreover, we have shown exceptional cyclability of such
electrode under pressure as low as 0.1 MPa.

Expanding on these findings, there are several avenues worth exploring. In the
upcoming Section 3.4, we evaluate the performance of various materials, with a specific focus
on compounds that offer 3D ionic diffusion pathways with the objective being the possibility
to eliminate the ball-milling step. Moreover, we explore the feasibility of leveraging the ductile
properties of halide materials discussed in the previous chapter to develop SE-free electrodes
utilising halide CAM, hence reuniting in one the two strategies employed in this thesis to
enable cycling under very low-pressure conditions.

102
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

3.4 – Applying the SE-free concept to other CAMs for low pressure
cycling
From an industrial perspective, although being used is the battery manufacturing, the
process of ball milling remains an additional expensive step. Until now, we used it in order to
improve the ionic diffusion through particle downsizing and morphology change of the layered
compounds. However, one such process could be unnecessary in the case that the material
already provides a 3D ionic diffusion pathway. Identifying such materials could lead to more
efficient and cost-effective production processes for SE-free electrode.

o Exploring other titanium-based sulphide polymorphs

In the same Ti-based sulphide family, several compounds, such as the spinel Ti2S4 or
the rock-salt Li2TiS3 phases (Figure 3.21), having a 3D ionic diffusion could be considered.

Figure 3.21. Crystallographic structure of (a) the spinel Ti2S4 and (b) of the disordered rock-salt Li2TiS3.

We first explored the cubic spinel Ti2S4, which was obtained by removing Cu from
CuTi2S4 using Br2,following the previously published148 protocol detailed in Appendix
Section A3.1. When using this phase, it is important to keep in mind that the activation energy
of diffusion in the cubic phase has been shown to be less favourable than in the layered one147
which could hinder the performance in a SE-free electrode configuration. Similar to the
experiments with the O3-LixTiS2, a solid-state cell was assembled and cycled. However, no
notable improvement was observed in the SE-free system until the sample underwent ball
milling, as shown in Figure 3.22. This suggests either that the diffusion coefficient in this spinel
phase is insufficient or that achieving nanoscale powders, from synthesis or through milling,
may be more crucial than an intrinsic isotropic diffusion in designing usable electrodes free of

103
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

SE additives. One potential improvement of this spinel phase involves focusing on the
synthesis of nanoscale particles that seems to be necessary in this case.

2.5 300
a Pristine Ti2S4 b Pristine Ti2S4
BM90-Ti2S4 250 BM90-Ti2S4

Discharge capacity (mAh/g)

BM135-Ti2S4 BM135-Ti2S4
Voltage (V vs. Li+/LiIn)

2.0
BM180-Ti2S4 200 BM180-Ti2S4

150
1.5

100

1.0
50

0
0.0 0.5 1.0 1.5 2.0 2.5 0 5 10 15 20
x in LixTiS2 Cycle number
Figure 3.22.Performance of the Spinel Ti2S4 as a function of the milling time in a SE-free electrode
configuration. (a) First galvanostatic cycle at C/20 and (b) capacity retention of the Spinel Ti 2S4 with different
milling time (0, 90, 135 and 180 minutes) in SE-free ASSBs with a LiIn anode at RT under a pressure of 100 MPa.

Following the same idea, we explored the potential of several other phases to be used
as SE-free cathode materials, some of which were prelithiated. Specifically, we investigated
the rock-salt Li2TiS3 phase with its high reported capacity of 425 mAh/g, the lithium-rich
layered Li1.2Ti0.44+Ti0.43+S2, the chemically lithiated layered O1-LiTiS2, the selenium-based
layered O1-LiTiSe2. Detailed procedures for their preparation are provided in Appendix
Section A3.1. However, it is worth noting that none of these phases allowed for SE-free cycling,
as indicated by the results displayed in Figure A3.5. The layered phases encountered similar
challenges and required a ball-milling step, while the rock-salt phase did not exhibit sufficient
ionic diffusion to support SE-free cycling even at a high pressure of 100 MPa.

In summary, results from our investigations involving various Ti-based chalcogenide

phases highlight the challenge of identifying materials that exhibit a sufficiently high 3D ionic
diffusion, possess a small particle size, and exhibit the necessary ductility to facilitate effective
interparticle contacts within the SE-free electrode.

o Leveraging the ductility of halide-based materials

In addition to chalcogenides, other compounds are known for their ductility. In the
previous chapter, we have harnessed this property from halide-based SE to enable low

104
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

pressure cycling. In the upcoming section, we intend to explore halide-based cathode active
materials that exhibit high ionic conduction.

Our initial exploration focused on the potential of the cubic Li2FeCl4 (LFC) phase to
serve as a SE-free electrode material, enabling low-pressure cycling. This phase not only is
expected to be ductile but also possess 3D ionic conduction pathways and a higher
electrochemical potential than sulphides. In a previous study by Tanibata et al., this phase was
used in a similar context, where an electrode composed of 95 wt.% LFC and 5 wt.% carbon
was employed in a LPSCl-based cell.149 The authors reported an ionic conduction of
2.1 · 10-5 S/cm for the LFC prepared through a mechano-synthesis process. They
demonstrated a one-electron full charge reaction with a high operating voltage of 3.6 V vs.
Li+/Li in an ASSB at a pressure of 100 MPa.

In our study, we synthesized the material following the same protocol by subjecting a
stoichiometric mixture of LiCl and FeCl2 to high-speed (500 RPM) ball milling for 10 hours. A
more comprehensive protocol is available in Appendix Section A3.1. Our electrochemical
impedance spectroscopy measurements revealed an ionic conductivity of 1.80 · 10-5 S/cm,
while DC polarization measurements indicated an electronic conductivity of 4.48 · 10-8 S/cm.

10 5
a b
RT, 100 MPa RT, 100 MPa
4
ionic = 1.80 · 10-5 S/cm E = 1.5 V
elec = 4.48 · 10-8 S/cm
-Im(Z) (kOhm)

3
I (µA)

5
2

0 0
0 5 10 0 10 20 30
Re(Z) (kOhm) Time (min)
Figure 3.23. Characteristics of the as-synthesized cubic LFC. (a) Diffractogram on LFC powder, (b) EIS and (c) DC
polarization measurement on a LFC pellet.

Following the synthesis and characterisation, we proceeded with galvanostatic cycling

of the material. Initially, we incorporated it within a composite electrode consisting of LFC,
LYBC (Li3YBr2Cl4), and VGCF in a mass ratio of 66.5:28.5:5 wt.% and cycled it at 100 MPa to
assess its nominal performance. The potential window was limited to the range 2.7 to 3.3 V

105
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

vs. Li+/Li corresponding to a one-electron reaction. Such limitation in potential was to avoid
the plateau corresponding to the irreversible second lithium deintercalation upon charge
(Figure A3.6). Results presented in Figure 3.24a and b demonstrate a reversible capacity of
around 115 mAh/g on the first discharge, with a substantial capacity decay upon cycling. Next,
we assessed the material in a SE-free electrode configuration by mixing it with only 5 wt.%
VGCF. Cycling this composite resulted in a relatively good discharge capacity of 104 mAh/g in
the first cycle. However, the capacity decay was also present on subsequent cycles.

3.3 a C/20, RT b C/20, RT

100 MPa 150 100 MPa

Charge Capacity (mAh/g)

Voltage (V vs. Li+/LiIn)

3.1
100

2.9 73.5 mAh/g

50

2.7
0
0.8 1.0 1.2 1.4 1.6 1.8 2.0 0 5 10 15 20
x in LixFeCl4 Cycle number
3.3 c C/20, RT d C/20, RT
100 MPa 150 100 MPa
Charge Capacity (mAh/g)
Voltage (V vs. Li+/LiIn)

3.1
100

2.9 69 mAh/g
50

2.7
0
0.8 1.0 1.2 1.4 1.6 1.8 2.0 0 5 10 15 20
x in LixFeCl4 Cycle number
Figure 3.24. Cycling performance of the cubic LFC against a LiIn anode in high pressure. (a and c) First
galvanostatic cycle and (b and d) capacity retention in (a-b) a LFC:LYBC:VGCF and (c-d) a LFC:VGCF composite in
a solid-state cell cycled at a rate of C/20, at RT and a pressure of 100 MPa.

The performance being assessed at high pressure, we then tested the material under
low-pressure conditions. To this end, similar to tests of the BM-LTS, we assembled cells with
either a composite composed of LFC, LYBC and VGCF or LFC and VGCF and we cycled them in
a fixed gap setup with an initial pressure of 10 MPa. Results shown in Figure 3.25 clearly

106
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

demonstrate a lower capacity of first cycle compared to the high-pressure tests as well as a
larger decay.

a C/20, RT b C/20, RT
3.3 150
10 MPa 10 MPa

Charge Capacity (mAh/g)

Voltage (V vs. Li+/LiIn)

3.1
100

2.9
50 58 mAh/g

2.7

0
0.8 1.0 1.2 1.4 1.6 1.8 2.0 0 5 10 15 20
x in LixFeCl4 Cycle number

c C/20, RT d C/20, RT
3.3 150
10 MPa 10 MPa
Charge Capacity (mAh/g)
Voltage (V vs. Li+/LiIn)

3.1
100

2.9
50

2.7
18 mAh/g
0
0.8 1.0 1.2 1.4 1.6 1.8 2.0 0 5 10 15 20
x in LixFeCl4 Cycle number
Figure 3.25. Cycling performance of the cubic LFC in a composite cathode and against a LiIn anode at low
pressure. (a and c) First galvanostatic cycle and (b and d) capacity retention in (a-b) a LFC:LYBC:VGCF and (c-d) a
LFC:VGCF composite in a solid-state cell cycled at a rate of C/20, at RT and an initial pressure of 10 MPa.

In summary, the performance of this halide-based active material in a SE-free cathode

configuration demonstrates poor performance under this reduced operating pressure.
Building upon these results, one avenue for further exploration could involve investigating the
low-pressure performance of the recently reported Li3TiCl6 halide-based CAM. Although this
material has demonstrated a relatively low reversible capacity of approximately 90 mAh/g, it
showed remarkable cycling retention in a SE-free cathode under high-pressure conditions.
Moreover, owing to its high ionic conductivity of 1.04 mS/cm, it showed high rate
capabilities.116 Finally, exploring the performance of other halide CAMs based on different
metal cations like Mn or V could also be a valuable area of investigation.

107
 Chapter 3 – SE-Free Cathode Concept for Low-Pressure Cycling of All-Solid-State Batteries

3.5 – Chapter conclusion

In conclusion, Chapter 3 presented a detailed exploration of an innovative concept of
a SE-free electrode configuration for achieving low pressure cycling in ASSBs, a strategy that
promises to simplify ASSB design and manufacturing processes.

The chapter began by evaluating the case of the O1-TiS2 as a literature reference to
establish a foundation for our investigation on SE-free cathode systems. We then focus on the
evaluation of the O3-LixTiS2 material, a pre-lithiated lamellar phase similar to the O1 phase.
We conducted a comprehensive examination of the characteristics and performance of this
material in typical composite cathodes. Subsequently, we introduced a milling strategy that
enabled SE-free cycling of O3-LixTiS2 through particle size downsizing and morphology change
that resulted in a decreased tortuosity and an enhanced Li+ ion diffusion through the electrode
thickness.

Continuing on this research trajectory, we have demonstrated that the ball-milled

material possesses the remarkable capability of cycling even under extremely low pressures,
as low as 0.1 MPa. This outstanding performance can be attributed to the material's ability to
expand and contract uniformly during cycling, resulting in minimal contact losses inside the
electrode. This contrasts with typical composite materials that can exhibit performance
degradation under low-pressure conditions due to contact-related issues.

Expanding on the SE-free concept, we explored its applicability to other CAMs for low-
pressure cycling. This included exploring of other Ti-based sulphide polymorphs and
leveraging the ductility of halide-based materials to widen the range of materials that can
benefit from SE-free configurations.

Overall, Chapter 3 has presented a SE-free electrode concept whose experimental

results can collectively contribute to the advancement of ASSB. This concept offers a
promising avenue to address the challenges associated with low-pressure cycling. Moving
forward, having demonstrated stable cycling under low-pressure conditions, we opened new
opportunities among which the use lithium metal anodes (LMA), a necessary step towards
making ASSBs more practical and applicable, which has not been done so far. In that regard,
the next chapter will explore this possibility and will discuss current strategies and limitations
that remain to be solved for enabling the use of LMAs in solid-state batteries.

108
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

Chapter 4 – Lithium Metal Anode:

Strategies, Critiques and
Perspectives on its Implementation

111
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

4.1 – Chapter Introduction

The use of a lithium metal anode (LMA) is essential to achieve high energy levels in
solid-state batteries (SSBs). Traditionally, LMAs have been considered unsafe due to their
tendency to form dendritic Li deposits that can penetrate the cell, resulting in short-circuits
and subsequent thermal runaway events. The use of a solid electrolyte (SE) was seen as a
promising solution to the problem due to its ability to mechanically prevent dendrite growth.
In previous chapters, we have discussed the need to reduce the stack pressure to enable the
implementation of this ductile lithium metal. By employing diverse engineering strategies
within the positive electrode, we have illustrated the achievement of high capacity and stable
cycling of cathode materials at pressures as low as 0.1 MPa. However, it should be noted that
we have not yet used LMA as the counter electrode.

In this chapter, our study begins with the integration of LMAs into the two previously
studied systems: the cathode composite employing a halide-based SE, as discussed in
Chapter 2, and the SE-free cathode strategy detailed in Chapter 3. Both of these systems have
previously demonstrated their ability to sustain cycling under extremely low-pressure
conditions. Consequently, we aim to evaluate their performance when paired with a lithium
counter electrode and to empirically highlight the critical challenges associated with LMAs.
This section will serve as the culmination of our examination of lithium metal all-solid-state
battery (Li-ASSB) systems within this thesis and will allow us to move into a discussion and
critique of the current methodologies employed to assess the viability and performance of
LMAs. To conclude, we will provide an overview of the current and most promising strategies
emerging in the literature for their integration

112
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

4.2 – Implementing Li metal anodes in low pressure systems

In Chapter 2 and 3, although demonstrating cycling of positive electrodes under very
low-pressure conditions, we have been using an alternative anode composed of an alloy of
lithium and indium. As illustrated in the In/Li binary phase diagram in Figure 4.1, this alloy has
several biphasic domains characterised by a voltage plateau, with the one occurring at 0.622 V
vs. Li+/Li that corresponds to the formation of the intermetallic InLi phase from Li and In. With
its long and stable potential, it is an ideal candidate to serve as an alternative counter
electrode for half-cell measurements. Moreover, in contrast to pure lithium or indium, which
are ductile, this alloy shows a brittle nature once formed, making it particularly advantageous
for cycling at high pressures, as it does not lead to mechanical extrusion through the SE.
However, due to the high content of the heavy and electrochemically inactive indium, coupled
with its higher potential, the utilisation of this anode in a full-cell system is impractical, as it
would significantly hamper the cell’s energy density. In that context, returning to the use of
low-potential anodes becomes evident. Consequently, we have opted to employ a pure
lithium anode for low-pressure cycling in both of the previously investigated systems.

Figure 4.1. (a) Binary In−Li phase diagram. (b) Results from a coulometric titration experiment at RT. The graph
shows the electrochemically determined open circuit voltage (OCV) over the lithiation from In to the phase In 2Li3.
Reproduced from reference150.

113
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

o Applicability to the halide-based composite system

Focussing on the halide-based composite of NMC622, studied in Chapter 2, we

investigated the performance of a bi-electrolyte NMC//Li3YBr2Cl4//Li6PS5Cl//Li cell
(Figure 4.2c) with a loading of 4.5 mgNMC/cm². It should be noted that, unlike the previous
study where a “full-halide” NMC//Li3YBr2Cl4//LiIn cell could be constructed, in this case, due
to the low reduction stability of LYBC (Figure 2.15b), the use of the LMA requires the
incorporation of a Li6PS5Cl (LPSCl) protective layer with the Li3YBr2Cl4 (LYBC) material. The cell
was cycled at decreasing stack pressures from 10 to 0.2 MPa. Remarkably, on the first cycle at
10 MPa (Figure 4.2a), this system displayed very similar performance as for the half-cell
experiment (Figure 2.19) with capacities of 177.2 and 150.2 mAh/g in charge and discharge,
respectively, and a high ICE of 84.7%. Gradually decreasing the pressure on subsequent cycles
did not lead to an increased capacity decay maintaining an average of -1%/cycle at C/20
(Figure 4.2d). A pressure of 0.2 MPa was reached but instabilities most likely originating from
the Li//LPSCl interface began to appear below.

200
a 1st cycle b
Charge Capacity (mAh/g)

4.0
Voltage (V vs. Li+/Li)

150

3.5 100 10 MPa 5 2 1 0.5 0.2

3.0 50

Cycling pressure (MPa)

Charge Capacity (mAh/g)
0.2 10 Discharge Capacity (mAh/g)
0
0 50 100 150 200 0 10 20 30
Capacity (mAh/g) Cycle number
-6
Discharge capacity decay rate (%)

c d Decay rate of each cycle

Decay at pressure change
-4

-2

0
10 MPa 5 2 1 0.5 0.2
Li metal 0 10 20 30
counter electrode Cycle number
Figure 4.2. (a) Last galvanostatic cycles at each pressure, (b) capacity retention and (d) decay rate of
NMC//LYBC//LPSCl//Li cell as a function of the stack pressure (from 10 to 0.2 MPa), cycled at RT and a C/20
rate. (c) Schematic of the Li metal dual-SE cell architecture tested in this experiment. Filled circles in (d) highlight
the larger capacity drop after each pressure change.

114
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

Instabilities in systems that incorporate LMA are frequently encountered. Despite

achieving a remarkably low pressure of 0.2 MPa in the previous experiment, various problems
quickly become apparent when attempting to replicate low-pressure cells using LMA. Some
cells exhibit diminished performance as soon as the pressure is reduced (as shown in
Figure 4.3a and b), while others experience short-circuits after just a few cycles (Figure 4.3c).

4.5 200
a b

Charge Capacity (mAh/g)

Voltage (V vs. Li+/Li)

4.0 150

10 MPa
3.5 100

1 MPa
3.0 50 10 MPa 1 MPa
Charge Capacity (mAh/g)
C/20, RT Discharge Capacity (mAh/g)
2.5 0
0 50 100 150 200 0 2 4 6 8
(Q-Qo) (mAh/g) Cycle number

c C/20, RT
Voltage (V vs. Li+/Li)

4
10 MPa
Short circuit

0
10
d
Stack pressure (MPa)

4
0 25 50 75 100 125
Time (h)
Figure 4.3. (a and b) Galvanostatic profiles and capacity retention of an NMC//LYBC//LPSCl//Li cell with
decreasing pressure and (c and d) galvanostatic curve and recorded stack pressure as a function of cycling time.
The sudden drop in potential highlighted by a dashed line in part c corresponds to a short-circuit of the cell.

o Lithium metal anode in the SE-free cathode system

Exploring the second strategy addressed in this thesis, we then focused on the
implementation of this LMA in a SE-free cathode of milled O3-LixTiS2 (referred to as
BM90-LTS). A lithium metal solid-state BM90-LTS//LPSCl//Li cell with a loading of around
4.5 mgAM/cm² was constructed by simply pressing a lithium disk on the surface of the cell’s SE

115
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

separator and was cycled at a relatively low initial pressure of 8 MPa. Lower pressures were
not attempted with this system due to the previously observed instabilities. As shown in
Figure 4.4, this cell displays a first cycle discharge capacity of 195 mAh/g with an average
polarization as low as 85 mV. Impressively, this configuration exhibited a remarkable capacity
retention of 94.2% after 25 cycles at C/20 and an energy density at electrode level of over
400 Wh/kg was reached. Similar to our previous observations, instabilities in the form of
lithium dendrites started appearing at the 28th cycle as indicated in Figure 4.4c.

300 102
1 20
3.2 Cycle 1 Cycle 10

Discharge Capacity (mAh/g)

Cycle 2 Cycle 20

Coulombic efficiency (%)

100
Voltage (V vs. Li+/Li)

2.8 Cycle 5
C/20 200

2.4 RT 94,2% of Qinit 98

2.0
100 96

1.6
94
a b
1.2 0
0.4 0.6 0.8 1.0 1.2 1.4 0 5 10 15 20 25
x in LixTiS2 Cycle number
Pressure (MPa) Voltage (V vs. Li /Li)

3.5
c
+

3.0
Dendrites
2.5

2.0
C/20
1.5 RT

9 d External cycling pressure

0 100 800 900

Time (h)
Figure 4.4. (a) Galvanostatic profile, (b) capacity retention and (c-d) pressure evolution (in red) upon cycling
(in black) of the BM90-LTS in the SE-free electrode system with a lithium metal anode in the 1.4 to 3.2 V vs.
Li+/Li electrochemical window.

Interestingly, using our cell configuration, which included an external force sensor, we
monitored the evolution of the cell’s pressure (Figure 4.4d). Due to lithium plating/stripping,
we found that the cell’s pressure mimics the galvanostatic curve and varies reversibly with
pressures oscillating from 6.6 to 8 MPa between the end of charges and discharges,
respectively. It is worth mentioning that, in this particular cell configuration, efforts to increase
the loading beyond 4.5 mgAM/cm² often result in cell failure. This outcome is not entirely

116
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

unexpected, as higher loadings induce greater pressure fluctuations due to lithium plating
until reaching a critical threshold where the cell experiences a short circuit due to the
mechanical extrusion of Li metal through the SE. This is an intrinsic problem of fixed-gap cell
setups, not fully appreciated by the battery community. Such an issue could be mitigated using
a spring-loaded constant pressure setup, presently under development in our group.

Altogether, results obtained with an LMA from either of our two strategies are
promising although a substantial amount of research and engineering on the Li//SE interface
is still required to effectively stabilise it. The well-documented66,103 issues associated to the
use of lithium in ASSBs, that have been comprehensively addressed in Chapter 1 Section 1.3
and that are summarised in the following Figure 4.5, are often limiting the achievable areal
capacities and power output of the cells to prevent failures. However, the development of
Li-ASSB reaching capacities of more than 5 mAh/cm² and capable of cycling at current of 5 to
10 mA/cm² is imperative for future industrial application and commercialisation of the
technology.16,151 In that context, it is imperative to be able to quantify the performance of the
Li//SE interface and to assess the strategies employed to enhance it.

Figure 4.5. Known anode interface-related issues in ASSBs and potential solutions (highlighted in blue). SEI
stands for solid electrolyte interphase and MCI for mixed-conducting interphase. Adapted from reference66.

117
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

4.3 – How accurate is the evaluation of the SE//Li interface?

An important metric to evaluate the performance of a Li-ASSB is the critical current
density (CCD) that can be defined as the maximum achievable current density at which a cell
can cycle stably without any failure occurring. However, as we will see in this section, its
assessment is not straightforward.

o Evaluating the critical current density with our electrolytes

A common approach for determining the CCD, involves the use of lithium symmetric
cells. As illustrated in Figure 4.6b, this assessment typically consists in a bidirectional
plating/stripping galvanostatic staircase test, which is characterised by a gradual increment in
the current with a fixed hold duration at each step. Through this method, researchers have
explored various systems and methods to enhance the CCD of Li-ASSBs. However, the CCD is
influenced by multiple factors such as cell chemistry, stack pressure or temperature and it is
important to note that significant variations in CCD have been reported, even for similar cell
configurations.152,153

0.8
a b c le
y
0.6
² p er c
A/cm
50 µ
0.4
Current (mA)

0.2

0.0

-0.2

-0.4

-0.6

-0.8
0 10 20 30
Time (h)
Figure 4.6. (a) Schematic of a lithium symmetrical cell stack and (b) typical current staircase test protocol.

To evaluate the performance of our pristine Li systems, meaning a SE separator with a

lithium metal anode, we conducted CCD staircase test measurements using two different
sulphide SEs: the β-Li3PS4 (LPS) and the LPSCl argyrodite. Symmetric cells were assembled
according to the protocol available in the Appendix Section A4.1, and were then cycled with a
starting current and rate increase of 50 µA/cm² under 9 MPa in cells comprising a spring-
loaded current collector to maintain a nearly constant pressure (Figure 4.7).

118
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

Upper current
collector
O-ring
Cell body
Sliding cylinder

Spring-loaded
current collector
Spring
Spring casing
Figure 4.7. Schematic of a constant pressure spring-loaded cell setup. The spring can be changed to modify the
stack pressure. Courtesy of Dr. Romain Dugas.

A typical potential profile resulting from a CCD of 0.75 mA/cm² is presented in

Figure 4.8a. Following a linear increase, the plating/stripping polarization ranges from 3.8 to
43 mV when increasing the current from 50 µA/cm² to the CCD of 750 µA/cm². On average,
the CCD obtained were 0.58 and 0.68 mA/cm² for the LPS and the LPSCl SE, respectively. These
observed values are consistent with previously reported findings for these sulphide SEs.152
However, as evident from the box plot in Figure 4.8b, substantial variations exist between cells
within the same SE system, with CCD spanning the range of 0.3 to 1.3 mA/cm² for the LPSCl
SE. Such variations have also been observed in recent works by both Meng's152 and Janek’s153
groups and in our experience, at first, we attributed these large variations to the challenge of
achieving consistent, flat, and uniform separator surfaces in contact with the Li. However, it is
crucial to note that numerous assembly and cycling parameters can also influence the
measured CCD in Li-ASSBs.
1.50 1.50
Interquartile range (25%~75%)
0.04
a Short circuit 1.00 b Mean ± 1.5 SD
CCD Median Line
RT 0.75 1.25 1.25
Critical current density (mA/cm2)

9 MPa
Current density (mA/cm2)

0.02 0.50
Voltage (V vs. Li+/Li)

1.00 1.00
0.25

0.00 0.00
0.75 0.75
-0.25 0.68

0.58
-0.02 -0.50 0.50 0.50

-0.75

-0.04 -1.00 0.25 0.25

7 cells 15 cells
0 10 20 30 40 b-LPS Argyrodite
Time (h)
Figure 4.8. (a) Typical current staircase test (in red) and resulting potential profile (in blue) of a Li//LPSCl//Li
cell and (b) whisker plot of the obtained CCD values for symmetrical cells with either a β-Li3PS4 or LPSCl
electrolyte. In part (b) the box corresponds to the interquartile 25-75%, the line to the median value, the average
is the square point, the whiskers to 1.5 standard deviation while the diamond points are outliers of the 1.5 SD.

119
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

o Parameters influencing the critical current density

Firstly, in line with observations made in liquid LIBs cells, higher temperatures lead to
substantial improvements in Li-ASSB cell stability. With our LPSCl system, measuring the CCD
at either 25°C or 55°C resulted in an improvement of the average CCD from 0.68 to 1.1 mA/m².
This trend is also depicted in the values agglomerated from the literature in Figure 4.9b, where
most CCD values exceeding 1 mA/cm² are achieved for temperatures above 60°C.

1.2 a b
Average CCD (mA/cm²)

1.0 1.1 mA/cm²

0.8

0.6 0.68 mA/cm²

0.4

0.2

0.0
25°C 55°C

Figure 4.9. CCD variation with temperature. (a) CCD measurement of the Li//LPSCl//Li symmetric cell system at
either 25°C or 55 °C and (b) current densities of Li-ASSB cells reported in the literature in different configurations
and temperatures. Part (b) is adapted from reference152.

Moreover, prior investigations have indicated that relatively small changes in stack
pressures can significantly affect the CCD. It has been reported to improve with increasing
stack pressure from 0.4 to 7 MPa (Figure 4.10a), as adjusting the pressure can help mitigate
void formation upon stripping by deforming the lithium.154 In that regard, different stack
pressure can leads to two different failure mechanism, either through dendrite growth and
subsequent short-circuit, as observed in Figure 4.8a, or void formation and contact loss, as
illustrated in Figure 4.10a with the drastic increase in the polarization at low pressure.

120
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

Figure 4.10. (a) Potential response to a constant current density at varying stack pressures and (b) CCD results
are taken from various reports on Li//LLZO//Li symmetric cells plotted versus the interfacial resistance. The
area of the circles represents the cumulative charge that was cycled until a short circuit occurred. To quantify
this capacity, compare the area of the 2.5 mAh/cm² grey circle on the top left corner. Reproduced from
references153,154.

In addition, the CCD measured in Li symmetrical cells does not often translate to full
cell performance where lower values are typically observed. A significant contributing factor
is the areal capacity, especially considering that CCD measurements in symmetrical cells often
use low capacities compared to the full-cell loading (< 1 mAh/cm² versus > 3 mAh/cm²). This
highlights the importance of the separator thickness, which acts as a buffer for dendrite during
their formation. Therefore, a low areal capacity or a thick separator can thus artificially
increase the CCD by preventing the dendrite from reaching the other electrode. Moreover, a
conditioning step consisting in the application of a relatively low assembly pressure 152 and a
mild heat-treatment (Figure 4.11a and b-c, respectively) can effectively decrease the initial
interfacial resistance which is particularly significant as previous research have indicated that
it is one of the most critical parameters influencing the obtained CCD (Figure 4.10b).153

121
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

6
a Total resistance = R1 + R2 + R3 b Before heat conditioning
5 After heating for 1h at 100°C
3

Ewe (mV)
0
4
Total resistance (kW)

-3
Threshold
pressure -6
3 0 50 100 150 200
Time (s)
2
400 c

-Im(Z) (Ohm)
200
1

Before heat conditioning

0 After heating for 1h at 100°C
0
0 20 40 60 0 200 400 600 800 1000
Pressure (MPa) Re(Z) (Ohm)
Figure 4.11. (a) Total resistance evolution as a function of the conditioning pressure used, (b) symmetric
galvanostatic imposition at 10 µA/cm² and (c) Nyquist plots of a Li//SE//Li symmetric cell before and after a
temperature conditioning at 100°C for 1h.

The cell setup is also important, as a fixed-gap configuration will inevitably lead to a
pressure build-up during charge possibly leading to the mechanical extrusion of lithium
through the SE. In contrast, Meng’s group reported significantly improved CDD in a constant
pressure system.152

Overall, although the use of this technique can be traced back to the development of
sodium ASSB in the 1980’s155, in light of the large variability observed in the literature today,
one might question the effectiveness of this method to measure accurate CCD. Specifically, it
is essential to note that the staircase test can introduce certain biases. The test results may be
influenced by the duration of the current hold, as a longer hold allows for more lithium plating.
Additionally, dendrite formation is significantly influenced by the presence of locally high
current densities. While model systems typically assume a uniform current density across the
electrode surface, real-world factors like surface roughness and electrode edges play a
substantial role. Therefore, the conventional use of symmetrical cells with identical-sized
lithium electrodes may lead to premature dendrite growth due to edge effects. 156

o How to improve the measure of the CCD?

Altogether we have seen that assessing the critical current density of a Li//SE interface,
being influenced by various parameters, is not straightforward. Therefore, the next section
will provide a concise list of recommendations to enhance the accuracy of CCD measurements.

122
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

1. Material Control:
 Ensure stringent control over Li surface contamination and bulk purity or use an
anodeless configuration.
 Maintain a flat, low-roughness surface for the SE separator, with uniform bulk density.
In that regard, if pistons are used for densification, make sure they are flat and
polished.
 Consider employing isostatic pressing to enhance the SE bulk density homogeneity.

2. Setup Considerations:
 Choose a constant pressure setup over typical fixed-gap systems.
 Ensure that the thickness of the SE is coherent with the target for the full cell.
 Design the sizes of the different components (Li electrode, separator) to prevent edge
effects, which depending on the system and protocol used may not be applicable in
symmetric cells.

3. Procedure Enhancements:
 Consider a cell pre-conditioning, involving the application of higher pressure (20 to
25 MPa for a few seconds) at assembly to establish good Li//SE contact and potentially
a mild heat-treatment to facilitate Li flow and improve wetting of the separator's
surface, thereby reducing interfacial resistance.
 Conduct a careful impedance analysis to examine the initial state of the cell, ensuring
that the assembly or pre-conditioning did not result in a mechanical formation of
lithium filaments that may lead to a "fake-stable phenomenon" and yield erroneously
high CCD values.153
 When assessing CCD, if the goal is to compare to full-cell performance, confirm that
the areal capacities, temperature and pressure used align with the targeted full-cell
characteristics to ensure the obtained CCDs accurately translate to full cell.
 Consider alternative procedures in addition to the bidirectional staircase test, such as
the measurement of the short-circuit times (Figure 4.12a and b) which consists in an
unidirectional galvanostatic polarization at constant current until a failure occurs.157,158
Moreover, a unidirectional staircase test implementing resting times can also be
considered. Such test can help deconvolute the effect of cycling and cell failure
mechanisms (Figure 4.12c and d).153

123
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

Figure 4.12. (a) Typical short-circuit observed on a Li symmetric cell polarized at 0.6 mA/cm² and (b) current
density vs. short-circuit time for different systems and corresponding tSC from part (a). (c) Current and (d)
polarization of Li//LLZO//Li cells for a unidirectional staircase CCD measurement, which also highlights the
influence of pressure on the failure mechanisms. Adapted from references153,157,158.

While these recommendations provide a basis for improving the accuracy of CCD
measurements, it is essential to acknowledge that the Li//SE interface is a complex system
with many influencing factors. Further research and exploration are likely to reveal additional
parameters and nuances that affect the CCD. These recommendations provide a starting point
for evaluating the performance of different strategies to ensure stable Li-ASSB, and ongoing
research will continue to expand our understanding of this interface. Several approaches have
been explored in order to increase the stability of Li-ASSB cells. In that regard, the next section
will be devoted to highlight some of these promising strategies.

4.4 – Promising strategies and perspectives to enable the use of lithium

metal anodes
o Design of solid electrolytes

From a chemical perspective, most of the SEs easily react at low potential when
contacted with lithium metal. Such a decomposition results in the formation of a resistive
interphases of which the thickness have been reported by Janek’s group to be often largely
underestimated.159 Approaches to tackle this issue are various and a straightforward solution
would be through the design of a SE that displays a thermodynamic stability at the lithium

124
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

potential. However, these are uncommon and, thus far, are not the most performing in terms
of conductivity or mechanical properties.25,26,160

o Alloy anodes

Alloying lithium with another metal is a technique of significant interest. While it is

commonly employed in laboratory-scale studies of ASSBs using lithium-indium alloy as an
anode, it is not a viable choice for constructing high-energy-density ASSBs, as previously
discussed. However, given the array of lithium alloys available, some alternatives may hold
potential.

In particular, for reasons similar to those of LIBs, the use of silicon as an anode may be
a promising option although it has shown difficulties in cycling under low pressures. 117
Magnesium-lithium alloy is also of particular interest as Mg undergoes a kinetically favourable
solid-solution reaction with Li compared to the sluggish intermetallic reaction of other metals
(Si, Au, Zn, etc.).161 Furthermore, as shown in Figure 4.13, this solid-solution domain extends
over a wide range of compositions both on the Mg-rich and Li-rich sides. Moreover, research
into lithium alloys has been extensive, extending beyond binary systems. Various ternary and
quaternary systems that incorporate lithium have been thoroughly investigated and could be
considered as potential candidates for such alloy anodes.162 In particular, a recent work
showed promising results in the use of a Li-Mg-Bi ternary alloy.163

Figure 4.13. Binary Li-Ag (left) and Li-Mg (right) phase diagrams. Reproduced from references164,165.

Although promising, forming such alloys typically leads to substantial volume changes
with lithium content, which can be detrimental to the cycling performance of ASSBs. In that
regard, one might seek alloys with minimal volume changes. An example of this is the open

125
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

metallic framework of LaNi5, capable of hosting 6 hydrogen atoms per formula unit with just
a 20% volume expansion.164

o Coatings and interlayers

As most solid electrolytes form a passivating SEI or a detrimental MCI, a better

approach involves the utilisation of coatings or interlayers. By tailoring the chemistry at the
interface, this approach offers several advantages, including greater thermodynamic stability
but also improved wettability or adhesion of Li with SE and provides a uniform mechanical
barrier to dendrites.107,166,167 Coatings and interlayers are an effective strategy to improve
uniformity of plating and stripping thus preventing the formation of dendrites and pores,
respectively. The most notable enhancement comes from the incorporation of a 5 µm-thin
Ag-C interlayer that demonstrates remarkable effectiveness in allowing a uniform plating and
stripping at current densities of up to 3.4 mA/cm² at 60°C.101,108,110,168

Figure 4.14. Schematic of Li plating–stripping on the current collector with an Ag–C nanocomposite layer
during charging and discharging processes.

Based on this idea, it would be necessary to replace the silver in the Ag-C layer and
explore the use of other materials known to be soluble in lithium or to form alloys with it. The
use of Zn, Al, Sn, and Ni have been attempted without reaching the performance of the
silver169, however, owing to its wide solid-solution range, magnesium is also a promising
candidate here. The exploration of other multi element systems mentioned before can also
be of interest.162

Moreover, we can recall the use of metal fluoride MFx at the interface that upon
lithiation undergo a conversion reaction into LiF and M with subsequent alloying of Li-M
(Figure 4.15).170 With an AgF interlayer, the stable operation at RT and 20 MPa of a Li-ASSB
cell with a high areal capacity of 9.7 mAh/cm² was reported. Similarly, another group used
CuF2 in a PVDF-HFP matrix that was laminated as an interlayer with a Li foil to form LiF and Cu
to protect the Li.171

126
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

Figure 4.15. Schematic of Li plating on a current collector with a MF interlayer. Reproduced from reference170.

o In-situ formed and self-healing coating

The application of a coating at SE/Li interface can be achieved through various

techniques, such as atomic layer deposition, sputtering, and more. Additionally, these
coatings can be applied to different components, either the SE particles, a densified separator,
or directly onto the Li metal itself. However, such a strategy inherently adds complexity and
raises the cost associated with constructing an ASSB. Similar to the use of vinyl carbonate in
lithium-ion batteries, an alternative strategy involves incorporating an electrolyte or an
additive that, in contact with lithium, decomposes to create a controlled electronically
insulating and ionically conductive stable interphase that would healed itself if ever damaged.

A notable example of this type of design can be found in the Li-Iodine primary cells,
mentioned in Chapter 1 Section 1.2, where Li and I2 were directly placed in contact to create
the LiI SE separator. In addition, halide-based SEs, when partially substituted with fluorine,
create a LiF-rich interphase that stabilises the interface despite their expected decomposition
at Li potential. The use of additive for that purpose is less common but one could potentially
incorporate additives like MFx or others like Cu3N into the SE to control the SEI formation.172

o Lithium reservoir-free cell configuration

From a materials standpoint, the processing of lithium metal poses its own set of
challenges. It readily reacts with various atmospheric components, even in trace amounts,
leading to the formation of a contamination layer on its surface that, in turn, can have
detrimental effects on its interaction with the SE. In addition, the purity of the lithium metal
is also an issues, its bulk composition frequently contains elevated levels of sodium and oxygen
that can significantly change its mechanical and electrochemical properties. Some report have
also indicated the presence of a significant number of defect in the bulk that upon cycling tend
to accumulate at the interface with the SE, thereby impacting the overall performance. 173–175
An adequate solution could involve adopting a Li reservoir-free cell alternative configuration,

127
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

where metallic lithium is not initially incorporated at assembly. This design, commonly
referred to as anode-free or anodeless, leads to higher energy density at cell level and, with
the SE serving as a filter, ensures the deposition of pure lithium during the charging process,
effectively addressing the mentioned issues.175 This configuration can also be coupled to an
interlayer strategy as illustrated for the Ag-C and MFx strategies in Figure 4.14 and Figure 4.15
respectively.

In summary, historical research efforts have focused on several strategies to achieve a

stable cycling behaviour for LMAs, as summarised in Figure 4.16. The most remarkable
improvements have arisen from optimising the uniformity of lithium plating and stripping,
with a significant milestone accomplished through the incorporation of interlayers in
anodeless configurations. Specifically, the utilisation of an Ag-C interlayer101,108,110,168,176
enables to reach currents of up to 3.4 mA/cm², while an AgF conversion-type interlayer
achieved capacities of up to 9.7 mAh/cm² at room temperature.170 Consequently, future
efforts to achieve higher current densities with LMAs should focus on the study of similar
micrometric interlayers M-C, MFx or MNx, with other metal M such as Mg, Si, Al, etc., that can
easily be spray coated168 onto the SE or current collector.

7
Areal capacity
(mAh/cm²)
6 10 Mg16Bi84 2.5 MPa
5
Current density (mA/cm²)

5 Si 100 MPa
Temperature (°C)

4 20 40 60 80 In 150 MPa
Ag-C 2 MPa
3 Ag-C 4 MPa

2 Ag-Gr 2 MPa 250 MPa

In
Mg16Bi84 2.5 MPa
1 5 MPa
AgF 20 MPa Mg 2.5 MPa
9 MPa 5 MPa
9 MPa 6 MPa
0
Pristine lithium Interlayer/Coating Alloy
Figure 4.16. Current densities reported in the literature for the LPSCl-based systems depending on the strategy
adopted for stabilising the Li//SE interface (pristine lithium, interlayer or alloy). Data is agglomerated from our
experimental data and references101,108,110,117,150,163,168,169,177–180.

128
 Chapter 4 – Lithium Metal Anode: Strategies, Critiques and Perspectives on its Implementation

4.5 – Chapter conclusion

In summary, Chapter 4 has provided a comprehensive exploration of strategies and
considerations related to the integration of lithium metal anodes (LMAs) into ASSBs. We have
presented direct evidence of the ability to cycle low pressure ASSB with a Li anode. This was
made possible by the use of cathodes capable of cycling under very low pressures that we
have previously explored in Chapter 2 and Chapter 3. Despite this achievement, such a study
had the merit to highlight the challenges associated with the implementation of LMAs.

Faced with these challenges, based on our own experience, we drawn attention to the
accuracy of critical current density (CCD) measurements in ASSBs, a key metric often used to
assess the improved SE//Li interface performance resulting from various strategies.
Understanding the parameters influencing the CCD, we then provided in all modesty
recommendations to enhance the accuracy of its measurement. Finally, we looked at current
strategies and promising perspectives for enabling the use of LMAs in ASSBs. This
encompassed the design of new solid electrolytes, the use of alloy anodes, and the application
of coatings and interlayers from which the latter showed the most remarkable improvements
in performance, with the use of Ag-based interlayers.

129
 General conclusions and perspectives

General conclusions and perspectives

131
 General conclusions and perspectives

Through this thesis, we investigated strategies to enable low-pressure operation of

positive electrodes and we reported on two different approaches, which enable to
significantly decrease the necessary cycling pressure in all-solid-sate batteries.

Given the susceptibility of composite electrodes to contact losses between the solid
electrolytes and the active material, we begin by incorporating a halide-based solid
electrolyte. This choice was motivated by the collective advantages of its highly ductile nature,
good ionic conductivity, robust potential stability and fine particle size, all of which were
expected to improve interfacial contacts within the electrode during cycling. In particular, our
investigation demonstrated that the incorporation of a fine Li 3YBr2Cl4 halide solid electrolyte
in conjunction with the NMC622 active material in the cathode composite enables the
remarkable achievement of cycling ASSBs at room temperature with a stack pressure as low
as 0.1 MPa. This stands in stark contrast to the poor performance we observed for the typical
sulphide argyrodite electrolyte-based solid-state batteries under low-pressure conditions.

Despite achieving remarkably low cycling pressures, this study clearly showed signs of
electrochemical degradation between the solid electrolyte and the high-potential cathode
materials used. Consequently, future research endeavours may benefit from directing their
efforts toward the development of novel ductile solid electrolytes characterised by wider
potential stability windows.

Inspired by the solid electrolyte-free concept developed in various existing literature

works, we have applied this approach to low-pressure cycling. The notable advantage of this
design stems from the expectation that, in contrast to composite electrodes, a solid
electrolyte-free electrode should exhibit uniform volume changes throughout the entire
electrode thickness, thereby mitigating interparticle contact losses. Necessitating a ductile
active material for sufficient compaction through cold pressing, we settled on a Ti-based
sulphide active material, in particular the O3-LixTiS2 (LTS) phase, which upon milling
transformed from a crystalline platelet-like shape of tens of micron in size to an amorphous
shapeless submicrometric particle morphology. The alteration in morphology was imperative
to reduce the tortuosity of Li+ pathways within the electrode. Employing this modified
material, our study demonstrated that a milling duration of 90 minutes was sufficient to
achieve performance levels comparable to those attained with a composite electrode

132
 General conclusions and perspectives

comprising solid electrolyte and the pristine LTS. Furthermore, when cycled at progressively
decreasing stack pressures, we observed no discernible decrease in capacity retention
associated with pressure changes down to 0.1 MPa.

On the basis of these remarkable results, we subsequently extended our survey to

evaluate this concept with other materials that displayed high ductility. Encouraged by the
low-pressure results obtained with halide solid electrolyte-based composites, we sought to
incorporate halide-based cathode material in a solid electrolyte-free system. As an initial proof
of concept, we explored the utilisation of the cubic Li2FeCl4 phase, which displays an operating
voltage higher than Ti-based sulphides while theoretically having an ionic conductivity
sufficiently high to bypass the use of an ionic conductor in our composites. Although our initial
attempts were unsuccessful, this concept deserves further exploration with other materials
displaying low hardness. In this regard, we propose the consideration of other halide-based
active materials. Notably, one compelling idea is to explore the utilisation of the Li3TiCl6 phase,
which exhibits an even higher ionic conductivity, comparable to that of conventional solid
electrolytes.

Having demonstrated the good cyclability of cathodes under very low-pressure

conditions, we then adventured into the next step in line, which is the incorporation of a high
capacity anode, in particular, the lithium metal anode. Utilising the cathode systems
employed throughout this thesis to facilitate low-pressure cycling, we successfully
incorporated and cycled lithium anodes. Nonetheless, this achievement has also highlighted
remaining challenges associated with cycling lithium metal, namely dendrites growth. This
observation brought back the problem of existing analytical methods for assessing the
performance of the solid electrolyte/lithium metal interface, which traditionally revolve
around the determination of the critical current density. Our experience, coupled with insights
from recent research, brought to light the clear limitations of this technique in assessing the
stability of this interface. As a result, we propose a series of recommendations to mitigate this
issue. We conclude by offering a brief overview of the existing strategies employed to
effectively integrate and consistently cycle lithium metal anodes within all-solid-state
batteries, spanning from bare SE//Li interface to alloys, coatings and interlayers.

Moving forward, it is evident that novel strategies regarding the design of interfaces
will be necessary in the successful implementation of high capacity anodes. As a perspective,

133
 General conclusions and perspectives

it becomes apparent that alloy-based approaches call for further investigations. In particular,
in light of recent literature reports, we suggest the exploration of magnesium, either as an
alloying compound or in small quantities in a carbonaceous interlayer to improve the
uniformity of the plating and stripping.

Finally, the advances made in this PhD have strengthened confidence in the practical
feasibility of building all-solid-state batteries over the next decade. These advancements are
particularly significant at a time when the development of all-solid-state batteries often faces
compromises in favour of intermediate solutions that incorporate either liquids, gels, or
polymers in their design.

134
 Appendix

Appendix

137
 Appendix

A2 – Supplementary information for Chapter 2

A2.1 – Materials and methods

Materials. Li3YBr2Cl4 (LYBC) was kindly provided by Saint-Gobain Recherche Paris. Fine
Li6PS5Cl (LPSCl) solid electrolyte was obtained from NEI Corporation.

SS-LPSCl was prepared through a solid-state synthesis consisting in the annealing of a

stoichiometric mixture of Li2S, P2S5 and LiCl. The powders were first thoroughly mixed in an
agate mortar before being transferred to alumina crucibles. The crucibles were then sealed
under vacuum in a quartz tube before being heated at 550°C (5°C/min heating rate) for 72h.
The resulting product was recovered and crushed by hand in a mortar before being used in
cells.

Li3InCl6 (LIC) was synthesized by a water-assisted synthesis. A stoichiometric quantity

of the precursors (LiCl and InCl3) was first dissolved in water and stirred overnight; the solution
was then natural dried in air at 100°C and subsequently under a dynamic vacuum at 100°C and
then at 200°C for 24h each.

LiNi0.6Mn0.2Co0.2O2 (NMC622 uncoated, monolithic, d10 = 3.21 μm, d50 = 5.28 μm and
d90 = 8.47 μm, BET = 0.5 m²/g) as well as the Co-coated (monolithic, d10 = 2.94 µm, d50 =
4.91 µm and d90 = 7.97 µm, BET = 0.45 m²/g) and Zr-coated NMC622 (monolithic, ~4 µm) were
kindly provided by UMICORE.

Composite preparation. The cathode composites were prepared by mixing NMC622,

SE and VGCF in a 66.5/28.5/5 wt.% ratio using two different techniques. Either by hand
grinding in a mortar or by planetary ball-milling (Pulverisette 7, FRITSCH) in a 45 mL zirconia
jar. In the latter case, the composite was milled for 30 min at 150 rpm with a ball-to-powder
mass ratio of 36/1 (5 ZrO2 balls of 10 mm in diameter for 400 mg of powder). The resulting
composite powder was then recuperated and stored in an argon-filled glovebox until further
use.

Lithium-indium composites were prepared in a two-step process. Firstly, the

appropriate amount of Li and In foils was weighed to achieve a Li0.5In stoichiometry. The foils
were then laminated and folded together using a glass tube until the resulting alloy became

138
 Appendix

brittle. Subsequently, the obtained alloy was mixed with SE (either LPSCl or LYBC) in a 60% LiIn
to 40% SE weight ratio in an agate mortar, until a homogeneous fine powder was obtained.

All-solid-state cell assembly. ASSBs were fabricated within a homemade two-

electrode cell developed in-house consisting in a cylindrical polyetherimide body and two
stainless steel pistons of 8 mm diameter as current collector.

In the case of cells based on solely LPSCl or LYBC, the assembly procedure is as follows.
First, 30 mg of SE was cold-pressed in the cell body at 1 t/cm² for a few seconds. Then, the
adequate amounts of NMC cathode and LiIn anode composites prepared with the same SE
were homogeneously spread on opposite sides of the separator and the stack was densified
at 4 t/cm² for 15 minutes.

For cells made with LIC//LPSCl, LYBC//LPSCl or LIC//LYBC bilayer separators (with
electrolytes designated as SE1//SE2 hereafter), a similar protocol is adopted. First 35 mg of
SE1 was cold-pressed in the cell at 1 t/cm², the inside of cell was then cleaned of any remaining
loose SE1 powder before adding 15 mg of SE2 and pressing at 1 t/cm². Following that, the
anode composite prepared with SE2 was spread and pressed on the SE2 surface layer. The
SE1-based cathode composite was then added on the opposite side of the separator (on the
SE1 surface) and the whole stack was densified at 4 t/cm² for 15 minutes. Loadings are
specified in the main text for each measurement and the typical weight ratio for
cathode/anode composites is 10/35.

Finally, the cells were closed differently depending on the stack pressures desired.
High-pressure cells were closed at 100 MPa by tightening the six top cell screws at a torque of
2.3 N.m. The low-pressure cells were cycled using the homemade frame comprising a force
sensor and a screw allowing us to set the initial stack pressure (Figure 2.4). All assembly
procedures were carried out in an Argon-filled glovebox ([O2] < 1 ppm, [H2O] < 1 ppm).

Ionic conductivity measurement was performed as follow. A pellet of about 100 mg

(~1 mm thick) was densified at 4 t/cm² for 3 min in a die set and the obtained pellet was then
recovered, measured and weighted before being transferred to a cell with carbon paper on
each side to improve contacts. Electrochemical impedance spectroscopy (PEIS) was then
performed with an excitation amplitude of 10 mV in the frequency range 5 MHz – 1 Hz under

139
 Appendix

a pressure of 100 MPa on a Biologic MTZ-35 device. The impedance spectra were fitted with
a simple R/CPE-CPE model and the conductivity was calculated with the following equation.
𝑙
𝜎𝑖𝑜𝑛 = Equation A2.1
𝑅∙𝐴
where l is the pellet thickness, R the total resistance measured by EIS and A the surface
area of the pellet.

Electrochemical stability window determination. To determine the electrochemical

potential window of LYBC, linear sweep voltammetry (LSV) was conducted. 8 mg of LYBC:VGCF
composite, prepared by hand grinding LYBC and VGCF in a 95:5 wt.% ratio using an agate
mortar, was utilised. The measurement was performed in the previously mentioned typical
cell setup, consisting of LYBC:VGCF//LYBC//Li0.5In:LYBC at a pressure of 100 MPa.

Galvanostatic cycling, unless specified, were conducted at room temperature within

the potential range of 2.1-3.6 V vs. Li+/LiIn, employing a C/20 rate (where C represents 1 mole
of Li per mole of active material in 1 hour).

Cell polarization ΔV were obtained through the difference of the charge and discharge
average potentials calculated using the following equation.
𝑄𝑐ℎ𝑎𝑟𝑔𝑒 𝑄
∫0 𝑉𝑑𝑄 ∫0 𝑑𝑖𝑠𝑐ℎ𝑎𝑟𝑔𝑒 𝑉𝑑𝑄
𝛥𝑉 = − Equation A2.2
𝑄𝑐ℎ𝑎𝑟𝑔𝑒 𝑄𝑑𝑖𝑠𝑐ℎ𝑎𝑟𝑔𝑒
where Qcharge and Qdischarge represent charge and discharge capacity respectively.

Capacity decay was calculated from one cycle to the other using the following
equation, for cycle n:
𝑄𝑛−1 − 𝑄𝑛
𝐷𝑒𝑐𝑎𝑦 (%) = ∗ 100 Equation A2.3
𝑄𝑛
where Qn corresponds to the discharge capacity of cycle n.

Standard rate capability measurements were conducted by initially performing a cycle

at C/20, followed by varying the discharge rate from C/20 to 5C, while the charging rate was
kept in CCCV mode (C/10 in CC and down to C/50 in CV). Three cycles were conducted at each
rate, and the average discharge capacity was utilised to assess the rate capability. For the
study in pressure, the first cycle was always performed at 10 MPa (except for pressure
> 10 MPa) to ensure successful cell assembly.

140
 Appendix

“Signature curve” method for rate capability. This method consists in discharging the
material at incrementally lower currents. Then the obtained cumulative capacities correspond
to the rate capabilities of the material. Experimentally, the cell was first charged to 3.6 V vs
Li+/LiIn in CCCV mode (C/10 in CC and down to C/50 in CV). Then the cell was discharged down
to 2.1 V vs Li+/LiIn with incrementally lower currents, starting from 5C to C/20, with 30 seconds
rest between each current pulse.

3.5
a 5C C/20 b

Cumulative capacity (mAh/g)

150
Voltage (V vs. Li+/LiIn)

3.0
Q1C
100 Q2C

Q5C
2.5
50

2.0 Q5C Q2C Q1C

0
0 50 100 150 0.1 1
Cumulative capacity (mAh/g) Discharge rate (nC)
Figure A2.1. (a) Typical galvanostatic response of a signature curve at rates ranging from 5C to C/20 and (b)
corresponding rate capabilities obtained by extracting the cumulative capacities from the signature curve.

SE compatibility. The compatibility of SEs was assessed using two different methods.
On one hand, in a LYBC//LPSCl bilayer through impedance evolution. The bilayer was prepared
by first cold-pressing 30 mg of LYBC followed by the addition of 30 mg of LPSCl. The stack was
then densified at 4 t/cm² for 3 min and transferred into a cell, which was closed at 100 MPa.
Subsequently, the cell underwent aging at 80°C for varying durations, and the impedance
evolution was monitored using PEIS at 25°C, following the aforementioned protocol. On
another hand, a mixture of LPSCl and LYBC was prepared by hand grinding the SEs in a 1:1
weight ratio. The powder was then pressed at 4 t/cm² for 3 minutes and the recovered pellet
was placed in a quartz tube that was subsequently sealed under high vacuum. The tube was
then heated at 150°C for 11 days to reach the thermodynamic equilibrium between the two
SE. Their reactivity was then assessed through XRD and conductivity measurements.

Scanning electron microscopy (SEM) and Energy Dispersive X-ray (EDX) mapping
were performed on pellet cross sections. The pellets were cut by simply braking them.
Micrographs were obtained on a FEI Quanta200F scanning electron microscope equipped with
a CPER AGROBIO 2012 EDX detector for chemical analysis. Samples were prepared and placed

141
 Appendix

on the holder inside an Ar-filled glovebox and subsequently transferred using a special
homemade transfer chamber to avoid exposure to air and moisture.

Powder X-ray diffraction (XRD) analysis was conducted utilizing an airtight holder
fitted with a beryllium (Be) window. The XRD patterns were recorded in reflection mode,
employing the Bragg−Brentano geometry, and data acquisition was carried out with a Bruker
D8 ADVANCE diffractometer. A copper (Cu) Kα X-ray source (λ1 = 1.54056 Å and λ2 = 1.54439
Å) was utilized in conjunction with a LynxEye detector for data collection.

A2.2 – Supplementary figures

200
a 180
b 100
3.5 C/20, RT

Coulombic efficiency (%)

Charge Capacity (mAh/g)
12 mgNMC/cm²
160
Voltage (V vs. Li+/Li)

95
100 MPa 140
120
3.0
90
100
80
60 85
2.5
40
Charge Capacity (mAh/g)
Discharge Capacity (mAh/g)
20 80
Coulombic efficiency (%)
2.0 0
0 50 100 150 200 0 5 10 15 20
(Q-Qo) (mAh/g) Cycle number
Figure A2.2. (a) First galvanostatic cycle and (b) capacity retention of a Zr-coated NMC622 cycled in the LPSCl-
based cell. Tests were performed at C/20, at RT under 100 MPa of pressure.

142
 Appendix

a b
3.5 C/20, RT 3.5 C/20, RT
Voltage (V vs. Li+/Li)

Voltage (V vs. Li+/Li)

3.0 3.0

2.5 2.5
First batch Second batch
Hand-ground Hand-ground
5% VGCF 5% VGCF
2.0 2.0
0 50 100 150 200 0 50 100 150 200
(Q-Qo) (mAh/g) (Q-Qo) (mAh/g)

c d
3.5 C/20, RT 3.5 C/20, RT

Voltage (V vs. Li+/LiIn)

Voltage (V vs. Li+/Li)

3.0 3.0

2.5 Third batch 2.5

Light hand-grinding
Mixed by ball-milling
5% VGCF
5% VGCF
2.0 2.0
0 50 100 150 200 0 50 100 150 200
(Q-Qo) (mAh/g) (Q-Qo) (mAh/g)
Figure A2.3. First galvanostatic cycles of LIC-based composites cells at low pressure of 10 MPa as a function of
the composite batch used. The composites used were: (a) the first, (b) second and (c) third composite batches
prepared by hand grinding using the same procedure and (d) a composite prepared by planetary ball milling.
These results highlight the importance of the composite preparation process on the performances and
repeatability of cells.

15

10

0
I (µA)

-5

-10 3
-15
1
-20

0.0 0.5 1.0 1.5 2.0

Voltage (V vs. Li+/LiIn)

Figure A2.4. Cyclic voltammetry measurement of LYBC/VGCF (95/5wt.%) composite from OCV to 0.5 V (in light
blue) and to 0 V vs. Li+/Li (dark blue).

143
 Appendix

a b
3,5 150

Average polarization (mV)

C/20, RT
Voltage (V vs. Li+/LiIn)

12 mgNMC/cm²
100 MPa
3,0
100

2,5

50

2,0
0 50 100 150 200 0 5 10 15 20
(Q-Qo) (mAh/g) Cycle number
Figure A2.5. (a) Galvanostatic cycling and (b) polarization over cycling of an LYBC-based cathode composite
cycled at C/20 and a pressure of 100 MPa at RT.

C/20, RT 100 MPa 2 MPa

Discharge capacity variation (%)

5 mgNMC/cm² 10 MPa 1 MPa

-4
5 MPa Averages

-2

10 MPa 5 MPa
100 MPa 2 MPa
1 MPa

24 26 28 30 32 34
Cycle number
Figure A2.6. Decay rate of NMC//LYBC//LPSCl//LiIn cells as a function of the pressure. Dash lines represent the
averages over the entire dataset for each cell. Measurements were performed after rate capability
measurements.

144
 Appendix

200 200
a C/20 C/10 C/5 C/2 1C 2C 5C b

Discharge Capacity (mAh/g)

Charge Capacity (mAh/g)

150 F 150
ast
dec
ay
100 MPa
100 100
55°C 5 mgNMC/cm²
100 MPa 70°C - Signature
50 5 mgNMC/cm² 50 50°C - Signature
Charge Capacity (mAh/g)
Discharge Capacity (mAh/g) 55°C - Standard
0 0
0 5 10 15 20 0.1 1
Cycle number Discharge rate (nC)
Figure A2.7. (a) Capacity retention of the rate capability measurement at 55°C using the standard protocol and
(b) associated rate capabilities at 55°C compared to the rate capabilities at 50°C and 70°C measured using the
signature curve protocol. The standard rate capability measurement at 55°C clearly displays a fast decay at C/20,
much larger than the nominal 1%/cycle at RT.

Figure A2.8. EDX mapping of single elements for the cross section pellets imaging. (a-c), (d-f) and (g-i)
correspond to SE (LPSCl or LIC), NMC and VGCF elements, respectively, in composite based on SS-LPSCl (in a,d
and g), fine LPSCl (in b, e and h) and LIC (in c, f and i).

145
 Appendix

Figure A2.9. Schematic of SE (LPSCl at the top and LYBC at the bottom) deformation during composite
preparation or densification depending on the stiffness of the material.

Figure A2.10.SEM and EDX imaging of SS-LPSCl-based cathode composite powders to assess adhesion. Green
and yellow coloured particles corresponds to SE elements (P, S, Cl and In), pink/red coloured particles
corresponds to NMC elements (Ni, Mn, Co, O) and blue regions to C.

146
 Appendix

Figure A2.11. SEM and EDX imaging of LIC-based cathode composite powders to assess adhesion. Green and
yellow coloured particles corresponds to SE elements (P, S, Cl and In), pink/red coloured particles corresponds
to NMC elements (Ni, Mn, Co, O) and blue regions to C.

147
 Appendix

A3 – Supplementary information for Chapter 3

A3.1 – Materials and methods

Materials. O1-TiS2 was purchased from Sigma. O3-LixTiS2 phase was synthesized,
following Colbow et al. methodology,138 from a high temperature (800°C) solid-state synthesis
by reacting stoichiometric amounts of thoroughly mixed Ti and Li 2S powders in evacuated
quartz tubes. Ti2S4 was obtained through the removal of Cu from the CuTi2S4 phase using Br2
in acetonitrile.181

Rock-salt Li2TiS3 was prepared through a mechanochemical synthesis, reacting

stoichiometric amount of Li2S and TiS2 following the previously described protocol.182 In brief,
500 mg of a stoichiometric mixture of commercially available Li2S and TiS2 were milled for 25
hours at 500 RPM in a 45 mL ZrO2 jar with 12 balls (10 mm in diameter) in a Pulverisette 7
planetary ball-mill from FRITSCH.

Lithium-rich layered Li1.2Ti0.44+Ti0.43+S2 was synthesized following the procedure

previously reported183 by reacting stoichiometric amount of Li2S, and TiS2 in a vacuum-sealed
quartz tube at 750 °C for 48 hours with a heating rate of 1°C/min.

O1-LiTiSe2 was prepared through a solid-sate synthesis by reacting a stoichiometric

amount of in-house made LiSe2 and commercial Ti and Se powders in a vacuum-sealed quartz
tube at 750°C for 48 hours with a heating rate of 1°C/min. The homemade LiSe 2 was
synthesized by solid-state synthesis, reacting Li and Se powders at 500°C for 48 hours with a
heating rate of 1°C/min.

Cubic Li2FeCl4 was prepared through a mechano-chemical synthesis. A stoichiometric

amount of LiCl and FeCl2 were weighted and transferred to a 45mL ZrO2 ball-milling jar along
with 72 balls of 5 mm in diameter, achieving a 30/1 ball-to-powder mass ratio. The mixture
was then subjected to ball milling (in a Pulverisette 7, FRITSCH) at a rotation speed of 500 RPM
for a duration of 10 hours.

Li6PS5Cl (LPSCl) was prepared as described previously in the literature 123. Briefly, a
solid-state synthesis was carried out by reacting stoichiometric amounts of Li2S, P2S5 and LiCl
reagent at 550°C for 72h in evacuated quartz tubes. Li3YBr2Cl4 (LYBC) was kindly provided by
Saint-Gobain Recherche Paris.

148
 Appendix

Ball milling of AM. Materials were milled in a stainless steel SPEX grinder for the
adequate amount of time by 200 mg batches using 2 stainless steel balls (8 mm in diameter)
corresponding to a ball-to-powder mass ratio of 20/1.

Composite preparation. The cathode composite of Ti-based sulphides were prepared

by mixing the AM and SE in a 70/30 wt.% ratio by hand grinding for around 10 minutes until
an homogeneous mixture was obtained. Composites of halide-based cathode materials were
composed of either AM, SE and VGCF or AM and VGCF in ratios of 66.5/28.5/5 or 95/5 wt.%
and fabricated by hand grinding. The resulting composites powder was then recuperated and
stored in an argon-filled glovebox until further use.

Lithium-indium composite was prepared in a two-step process. Firstly, the appropriate

amount of Li and In foils was weighed to achieve either a Li0.5In or Li0.8In stoichiometry. The
foils were then laminated and folded together using a glass tube until the resulting alloy
became brittle. Subsequently, the obtained alloy was mixed with SE (either LPSCl or LYBC) in
a 60% LiIn to 40% SE weight ratio in an agate mortar, until a homogeneous powder was
obtained.

Liquid cells were assembled using Swagelock hardware and are made of the O3-LixTiS2
(100%) as positive electrode, either in the pristine or ball-milled state, with a Li metal disk as
counter electrode and both were separated by a glass fiber Whattman separator soaked in an
LP30 electrolyte solution (1M LiPF6 in EC:DMC (1:1 vol)) obtained from Dodochem.

All-solid-state cell assembly. ASSBs were fabricated within a homemade two-

electrode cell developed in-house consisting in a cylindrical polyetherimide body and two
stainless steel pistons of 8 mm diameter as current collector. Fabrication consists in
sandwiching LPSCl SE powder (30 mg) between LiIn/LPSCl (in 60:40 wt.%) composite as
counter electrode and the positive electrode. The layers are successively pressed at 1 t/cm²
before being densified at 4 t/cm² for 15 min. Finally, following the same protocol as for the
cell in Chapter 2, the cells were closed differently depending on the stack pressures desired.
High-pressure cells were closed at 100 MPa by tightening the six top cell screws at a torque of
2.3 N.m. The low-pressure cells were cycled using the homemade frame comprising a force
sensor and a screw allowing us to set the initial stack pressure.

149
 Appendix

Three-electrode solid-state cells were fabricated for GITT measurements. They were
assembled in an homemade 3-electrode setup modified from the 2-electrode setup as
previously published by our group.184 Unlike in ref184, first, lithium metal was rolled down to
100 µm and, as depicted in the Figure A3.1, a ring (13 and 8 mm in outer and inner diameter
respectively) was punched and clamped between the reference current collector and the
plastic body to use as a reference electrode. Then, the cell stack was constructed as described
above for the two-electrode cell setup making sure that enough SE was used to entirely cover
the Li metal ring reference electrode.

Figure A3.1. Schematic of the three-electrode cell inner parts. Adapted from ref184.

All assembly procedures were carried out in an Argon-filled glovebox ([O2] < 1 ppm,
[H2O] < 1 ppm).

Ionic conductivity measurement was performed as follow. A pellet of about 100 mg

(~1 mm thick) was densified at 4 t/cm² for 3 min in a die set and the obtained pellet was then
recovered, measured and weighted before being transferred to a cell with carbon paper on
each side to improve contacts. Electrochemical impedance spectroscopy (PEIS) was then
performed with an excitation amplitude of 10 mV in the frequency range 5 MHz – 1 Hz under
a pressure of 100 MPa on a Biologic MTZ-35 device. The impedance spectra were fitted with
a simple R/CPE-CPE model and the conductivity was calculated with the following equation.
𝑙
𝜎𝑖𝑜𝑛 = Equation A3. 1
𝑅∙𝐴
where l is the pellet thickness, R the total resistance measured by EIS and A the surface
area of the pellet.

150
 Appendix

Electronic conductivity measurements were conducted using either of the specified

technique: the chronopotentiometry or DC polarization technique under a pressure of
100 MPa.

For Ti-based sulphide materials, the chronopotentiometry technique was employed.

Pellets consisting of CAM fibrillated with 1.3 wt.% PTFE to enhance their mechanical strength
were pressed under 4 t/cm². To accurately assess the electronic conductivity, 4-wire
measurements were employed, which effectively eliminated any impedance contribution
from the cable. Additionally, a blank measurement was carried out to eliminate any influence
from the cell itself.

For halide-based CAM, the DC polarization technique was adopted and pellets of
around 100 mg of material were densified for 3 minutes under 4 t/cm².

Galvanostatic cycling was carried out at room temperature and at desired potential
window and C-rates (with C corresponding to 1 mol of Li per mole of active material in 1 h).
Unless specified, cycling was performed with loadings of 12-13 mgAM/cm² and a stack pressure
of 100 MPa.

Galvanostatic intermittent titration technique. Two different pulse durations were

used for the pristine and the ball-milled samples due to the much lower diffusion of the first.
Pulses at a rate of C/20 were applied for 10 or 30 minutes for the pristine or ball-milled
samples respectively, followed by a rest period of 5 hours. The apparent Li + diffusion
coefficient Dapp was then calculated via Equation A3.2.

4 𝑚𝑉𝑚 2 𝛥𝐸𝑠 2
𝐷𝑎𝑝𝑝 = ( ) ( ) Equation A3.2
𝜋𝜏 𝑀𝑆 𝛥𝐸𝑡
where m is the mass, Vm the molar volume and M the molar mass of the active
material, S the contact area between the SE and the cathode layer (in SE-free electrode), ΔEs
the steady-state voltage change, ΔEt the transient voltage change when current is applied and
τ the current pulse duration.

“Signature curve” method for rate capability. Rate capabilities were measured on
typical 2-electrode cells by the “Signature curve” method that consists in discharging the
material at incrementally lower currents. Then the obtained cumulative capacities correspond
to the rate capabilities of the material. Experimentally, the cell was first charged to 2.6 V vs

151
 Appendix

Li+/LiIn in CCCV mode (C/20 in CC and down to C/200 in CV). Then the cell was discharged
down to 0.8 V vs Li+/LiIn with incrementally lower currents, starting from 5C to C/200, with 30
minutes rest between each current pulse.

All electrochemical measurements were conducted on Biologic devices (VMP3

potentiostats or BCS-805 cyclers) devices controlled by EC-Lab or BT-Lab software.

Scanning electron microscopy (SEM) was performed on powders and pellet cross
sections. For ex-situ measurements, cells were cycled to the desired potential, then
disassembled and pellets were finally extracted. The pellets were cut by simply braking them.
Micrographs were obtained on a FEI Quanta200F scanning electron microscope equipped.
Samples were prepared and placed on the holder inside an Ar-filled glovebox and
subsequently transferred into the SEM chamber. The transfer was as fast as possible to limit
the exposure to air and moisture.

Powder X-ray diffraction (XRD) analysis was conducted utilising an airtight holder
fitted with a beryllium (Be) window. The XRD patterns were recorded in reflection mode,
employing the Bragg−Brentano geometry, and data acquisition was carried out with a Bruker
D8 ADVANCE diffractometer. A copper (Cu) Kα X-ray source (λ1 = 1.54056 Å and λ2 =
1.54439 Å) was utilised in conjunction with a LynxEye detector for data collection.

Operando X-ray powder diffraction (XRD) was performed in an airtight

electrochemical cell equipped with a Be window. XRD patterns were recorded using the same
methods as described above.

Solid-state NMR experiments were performed on a 4.7 T (200 MHz for 1H) Bruker
Avance HD spectrometer equipped with 4 mm Magic Angle Spinning (MAS) probehead tuned
to 7Li at 77.8 MHz. The MAS rate was set to 10 kHz, and the bearing and drive gas were 100%
N2, the temperature was regulated and calibrated with the 207Pb shift in lead nitrate. The
radiofrequency field was set to 58.5 kHz at 100 W amplifier output power. The 7Li shift was
calibrated with a 1 M LiCl aqueous solution and the longitudinal relaxation times were
measured with a saturation recovery sequence. The spectra were recorded with 16 transients,
with at least 5 T1 magnetization recovery delays between each. The EXSY spectrum was
recorded with 128 transients separated by a 2s delay, and 256 t1 increments in the indirect

152
 Appendix

dimension. The sweepwidth of the indirect dimension was set to 10 kHz and the exchange
delay was set to 50 ms.

A3.2 – Supplementary figures

O1 O3 O1
50
(101) (104) (102)

40
Intensity (arb. units)

30

Time (h)
20

Discharge
10

Charge
0

14 30 34 36 42 44 46 3.0 2.5 2.0

2Theta (°) Voltage (V vs. Li+/Li)
Figure A3.2. Operando X-ray diffraction operando cycling measurement of a Pristine LTS electrode in liquid
cell.

1.00 3.0
a b
Pristine LTS (+ 1.3 wt.% PTFE) 2.5 BM90-LTS (+ 1.3 wt.% PTFE)
0.75 t = 254 µm t = 345 µm
150 mA
Potential (mV)

Potential (mV)

S = 0.503 cm² 150 mA 2.0 S = 0.503 cm²

Rblank = 1.125 · 10-3 W Rblank = 1.125 · 10-3 W
<E> = 2.337 mV
<E> = 0.714 mV
0.50 1.5
100 mA 100 mA
<E> = 0.482 mV <E> = 1.564 mV
1.0
0.25
50 mA 50 mA
0.5
<E> = 0.240 mV <E> = 0.795 mV
0.00 0.0
0 20 40 60 80 0 20 40 60 80
Time (s) Time (s)
Figure A3.3. Chronopotentiometry measurements for electronic conductivity determination of (a) a pristine
LTS and (b) a BM90-LTS pellet. Tests were performed at RT under 100 MPa of pressure.

153
 Appendix

O1 O3
2.5

Voltage (V vs. Li+/LiIn)

2.0

1.5

1.0

0.2 0.4 0.6 0.8 1.0 1.2

x in LixTiS2
Figure A3.4. First galvanostatic cycle at C/20 of the pristine LTS:LPSCl (70:30) composite against a Li 0.8In:LPSCl
counter electrode in ASSB, revealing two plateaus on the first charge attributed to the O3-O1 phase transition
or the LTS.

154
 Appendix

150
a C/20, RT b 100
2.0

Discharge capacity (mAh/g)

100 MPa

Coulombic efficiency (%)

Voltage (V vs. Li+/LiIn)

95
100
1.5 90

85
50
1.0
80

75
0.5 0
0.4 0.6 0.8 1.0 0 2 4 6 8 10
x in LixTiS2 Cycle number
3 150 210
c C/20, RT d 200
100 MPa Discharge capacity (mAh/g)

Coulombic efficiency (%)

Voltage (V vs. Li+/LiIn)

2 100 100
90
80
1 50 70
60
50
0 0 40
0.6 0.8 1.0 1.2 1.4 0 2 4 6 8 10
x in LixTiSe2 Cycle number
100 115
e C/20, RT f
2.5 110
Discharge capacity (mAh/g)

100 MPa

Coulombic efficiency (%)

Voltage (V vs. Li+/LiIn)

75 105

2.0 100

50 95

1.5 90

25 85

1.0 80

0 75
2.0 2.2 2.4 0 2 4 6 8 10
x in LixTiS3 Cycle number
Figure A3.5. Cycling performances in a SE-free configuration of (a and b) a chemically lithiated TiS 2 after ball
milling, (c and d) a ball-milled LiTiSe2 and (e and f) a rock-salt Li2TiS3. (a-c) Galvanostatic cycling and (d-f) capacity
retention of solid-state cells cycling in a SE-free configuration at 100 MPa, RT and a rate of C/20.

155
 Appendix

C/20, RT
4.0
100 MPa

Voltage (V vs. Li+/LiIn)

3.5

3.0

2.5

0.5 1.0 1.5 2.0

x in LixFeCl4
Figure A3.6. Galvanostatic cycling of the LFC material in a LFC:LYBC:VGCF composite in the range 2.5-4.0 V vs.
Li+/LiIn

A3.3 – Supplementary tables

Table A3.1. NMR fitting parameters for the 7Li NMR spectra of pristine, BM20, BM60 and BM90-LTS recorded
at RT
Fraction of Li Chemical Gaussian /
Environment FWHM
signal shift Lorentzian
Pristine

Mobile (pink) 49.3% 12.0 ppm 5.6 ppm 17% / 83%

Ordered Li (blue) 50.7% 2.1 ppm 9.5 ppm 46% / 54%

Mobile (pink) 65% 9.0 ppm 6.1 ppm 0% / 100%

BM20

Ordered Li (blue) 35% 0.2 ppm 7.4 ppm 40% / 60%

Mobile (pink) 50.8% 9.1 ppm 5.9 ppm 0% / 100%

BM60

Ordered Li (blue) 33.4% 0.3 ppm 8.6 ppm 34% / 66%

“Amorphous“ (yellow) 15.8% 2.8 ppm 22 ppm 100% / 0%

New env. (green) 24.7% 7 ppm 7.1 ppm 0% / 100%

BM90

Ordered Li (blue) 40.7% 2.9 ppm 13 ppm 61% / 39%

“Amorphous” (yellow) 34.6% 2.9 ppm 29.4 ppm 100% / 0%

156
 Appendix

A4 – Supplementary information for Chapter 4

A4.1 – Materials and methods

Materials. LiNi0.6Mn0.2Co0.2O2 (NMC622 uncoated, monolithic, d10 = 3.21 μm, d50 =

5.28 μm and d90 = 8.47 μm, BET = 0.5 m²/g) was kindly provided by UMICORE.

O3-LixTiS2 phase was synthesized, following Colbow et al. methodology,138 from a high
temperature (800°C) solid-state synthesis by reacting stoichiometric amounts of thoroughly
mixed Ti and Li2S (Alfa Aesar, purity 99 w/w %) powders in evacuated quartz tubes.

Li3YBr2Cl4 (LYBC) was kindly provided by Saint-Gobain Recherche Paris. Li6PS5Cl (LPSCl)
solid electrolyte was obtained from NEI Corporation. β-Li3PS4 was prepared through a solvent-
mediated synthesis in tetrahydrofuran (THF). Stoichiometric amounts of Li2S (Alfa Aesar,
purity 99 w/w %) and P2S5 (Acros Organics, purity > 98 w/w %) were dispersed in THF (Sigma
Aldrich, 99.9 % v/v) and then, continuously stirred for 48 h at room temperature. The obtained
powder was later recovered by centrifugation at 600 rpm for 3 min. Next, the powder was
dried in a cylindrical Schlenk flask at 100°C (heating rate 5°C/min) for 24 h and subsequently
at 155°C for 24 h under dynamic vacuum (P < 0.1mbar).

BM90-LTS preparation. Pristine O3-LixTiS2 was milled in a stainless steel SPEX grinder
for 90 minutes by 200 mg batches using 2 stainless steel balls (8 mm in diameter)
corresponding to a ball-to-powder mass ratio of 20/1.

Composite preparation. Composites of NMC and LYBC were composed of AM, SE and
VGCF in ratios of 66.5/28.5/5 wt.% and fabricated by hand grinding.

Li metal all-solid-state cell assembly. Low pressure Li-ASSB cells were fabricated
within a homemade two-electrode cell developed in-house consisting in a cylindrical
polyetherimide body and two stainless steel pistons of 8 mm diameter as current collector.
Fabrication consists in first pressing LPSCl SE powder (30 mg) into a separator then the cathode
was added with a loading of 4.5 to 5 mgAM/cm² on one side and the stack was densified at
4 t/cm² for 15 minutes. Finally, a lithium foil was cleaned and rolled to a thickness of 80-100
µm before being punched at 8 mm diameter and placed in the cell against the LPSCl layer. The
low-pressure cells were cycled using the homemade frame comprising a force sensor and a
screw allowing us to set the initial stack pressure.

157
 Appendix

Li symmetric cell were assembled by first densifying 30 mg of LPSCl or β-LPS into the
sliding cylinder of the spring-loaded cell (illustrated in Figure 4.7) and then adding on both side
of the separator Li disks of 8 mm in diameter punched from a lithium foil that was cleaned and
rolled to a thickness of 80-100 µm. To add the Li disks, they were first stuck onto stainless steel
(304 type) disks of 8mm in diameter and 0.5 mm in thickness by pressing them into each
other’s by hand.

All assembly procedures were carried out in an Argon-filled glovebox ([O2] < 1 ppm,
[H2O] < 1 ppm).

Galvanostatic cycling was carried out at room temperature and at desired potential
window and C-rates (with C corresponding to 1 mol of Li per mole of active material in 1 h).
Unless specified, cycling was performed with loadings of 4.5-5 mgAM/cm² and a stack pressure
of 100 MPa.

CCD measurements on Li symmetrical cell were carried out at room temperature

following a staircase test with pulses of 1 hour from a starting current of 50 µA/cm² and step
increase of 50 µA/cm² per cycle. If not specified, the pressure was applied through a spring
and set to 9 MPa

All electrochemical measurements were conducted on Biologic devices (VMP3

potentiostats or BCS-805 cyclers) devices controlled by EC-Lab or BT-Lab software.

158
 Appendix

A5 – Additional supplementary information

Figure A5.1. Electric circuit of the control boxes used for pressure and temperature control.

159
 Appendix

Figure A5.2. Electric circuit of the pressure controller connection to the analogue output and the force sensor
input.

160
 Appendix

Figure A5.3. Soldering connections diagram for the cables connecting the potentiostat to the cell through the
glovebox.

161
 References

References

163
 References

(1) Calvin, K. IPCC, 2023: Climate Change 2023: Synthesis Report. Contribution of Working
Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on
Climate Change [Core Writing Team, H. Lee and J. Romero (Eds.)]. IPCC, Geneva,
Switzerland., First.; Intergovernmental Panel on Climate Change (IPCC), 2023.
https://doi.org/10.59327/IPCC/AR6-9789291691647.

(2) Thomas, T. Pumped Storage: Proceedings of the Conference Organized by the Institution
of Civil Engineers at Imperial College of Science, Technology and Medicine, London; 1990.

(3) Rehman, S.; Al-Hadhrami, L. M.; Alam, Md. M. Pumped Hydro Energy Storage System: A
Technological Review. Renewable and Sustainable Energy Reviews 2015, 44, 586–598.
https://doi.org/10.1016/j.rser.2014.12.040.

(4) Tarascon, J.-M.; Armand, M. Issues and Challenges Facing Rechargeable Lithium Batteries.
Nature 2001, 414 (6861), 359–367. https://doi.org/10.1038/35104644.

(5) Li, Q.; Yang, Y.; Yu, X.; Li, H. A 700 W⋅h⋅kg−1 Rechargeable Pouch Type Lithium Battery.
Chinese Phys. Lett. 2023, 40 (4), 048201. https://doi.org/10.1088/0256-
307X/40/4/048201.

(6) Randau, S.; Weber, D. A.; Kötz, O.; Koerver, R.; Braun, P.; Weber, A.; Ivers-Tiffée, E.;
Adermann, T.; Kulisch, J.; Zeier, W. G.; Richter, F. H.; Janek, J. Benchmarking the
Performance of All-Solid-State Lithium Batteries. Nat Energy 2020, 5 (3), 259–270.
https://doi.org/10.1038/s41560-020-0565-1.

(7) Pile voltaïque. Wikipédia; 2023.

(8) Goodenough, J. B.; Park, K.-S. The Li-Ion Rechargeable Battery: A Perspective. J. Am. Chem.
Soc. 2013, 135 (4), 1167–1176. https://doi.org/10.1021/ja3091438.

(9) Rouxel, J.; Danot, M.; Bichon, M. Les Composites Intercalaires NaxTiS2. Etude Générale
Des Phases NaxTiS2 et KxTiS2. Bull. Soc. Chim 1971, 11, 3930–3936.

(10) Funke, K. Solid State Ionics: From Michael Faraday to Green Energy—the European
Dimension. Sci Technol Adv Mater 2013, 14 (4), 043502. https://doi.org/10.1088/1468-
6996/14/4/043502.

(11) Fast Ion Transport in Solids; Scrosati, B., Magistris, A., Mari, C. M., Mariotto, G., Eds.;
Springer Netherlands: Dordrecht, 1993. https://doi.org/10.1007/978-94-011-1916-0.

(12) International Energy Agency. Global EV Outlook 2023: Catching up with Climate
Ambitions; Global EV Outlook; OECD, 2023. https://doi.org/10.1787/cbe724e8-en.

(13) Assat, G.; Tarascon, J.-M. Fundamental Understanding and Practical Challenges of
Anionic Redox Activity in Li-Ion Batteries. Nat Energy 2018, 3 (5), 373–386.
https://doi.org/10.1038/s41560-018-0097-0.

(14) Li, Q.; Yang, Y.; Yu, X.; Li, H. A 700 W⋅h⋅kg−1 Rechargeable Pouch Type Lithium Battery.
Chinese Phys. Lett. 2023, 40 (4), 048201. https://doi.org/10.1088/0256-
307X/40/4/048201.

164
 References

(15) Louli, A. J.; Eldesoky, A.; deGooyer, J.; Coon, M.; Aiken, C. P.; Simunovic, Z.; Metzger,
M.; Dahn, J. R. Different Positive Electrodes for Anode-Free Lithium Metal Cells. J.
Electrochem. Soc. 2022, 169 (4), 040517. https://doi.org/10.1149/1945-7111/ac62c4.

(16) Janek, J.; Zeier, W. G. A Solid Future for Battery Development. Nat Energy 2016, 1 (9),
16141. https://doi.org/10.1038/nenergy.2016.141.

(17) Pang, M.-C.; Wei, Y.; Wang, H.; Marinescu, M.; Yan, Y.; Offer, G. J. Large-Format Bipolar
and Parallel Solid-State Lithium-Metal Cell Stacks: A Thermally Coupled Model-Based
Comparative Study. J. Electrochem. Soc. 2021, 167 (16), 160555.
https://doi.org/10.1149/1945-7111/abd493.

(18) Faraday, M. I. Experimental Researches in Electricity. —Fifteenth Series. Phil. Trans. R.

Soc. 1839, 129, 1–12. https://doi.org/10.1098/rstl.1839.0002.

(19) O’Keeffe, M. Phase Transitions and Translational Freedom in Solid Electrolytes. In

Superionic Conductors; Mahan, G. D., Roth, W. L., Eds.; Springer US: Boston, MA, 1976; pp
101–114. https://doi.org/10.1007/978-1-4615-8789-7_9.

(20) Nernst, W.; Dolesalek, F. Über die Gaspolarisation im Bleiakkumulator. Z. Elektrotech.

Elektrochem. 1900, 6 (45), 549–550. https://doi.org/10.1002/bbpc.19000064502.

(21) Wagner, C. Über den Mechanismus der elektrischen Stromleitung im Nernststift.

Naturwissenschaften 1943, 31 (23), 265–268. https://doi.org/10.1007/BF01475685.

(22) Tubandt, C.; Lorenz, E. Molekularzustand und elektrisches Leitvermögen kristallisierter

Salze. Zeitschrift für Physikalische Chemie 1914, 87U (1), 513–542.
https://doi.org/10.1515/zpch-1914-8737.

(23) Owens, B. B.; Argue, G. R. High Conductivity Solid Electrolyte System Rbl ‐ AgI. J.
Electrochem. Soc. 1970, 117 (7), 898. https://doi.org/10.1149/1.2407666.

(24) Whittingham, M. S.; Huggins, R. A. Beta Alumina-Prelude to a Revolution in Solid State

Electrochemistry. NBS Special Publications 1972, 364, 139–154.

(25) Alpen, U. v. Li3N: A Promising Li Ionic Conductor. Journal of Solid State Chemistry 1979,
29 (3), 379–392. https://doi.org/10.1016/0022-4596(79)90195-6.

(26) Ginnings, D. C.; Phipps, T. E. TEMPERATURE-CONDUCTANCE CURVES OF SOLID SALTS.

III. HALIDES OF LITHIUM. J. Am. Chem. Soc. 1930, 52 (4), 1340–1345.
https://doi.org/10.1021/ja01367a006.

(27) Owens, B. B.; Salkind, A. J. Key Events in the Evolution of Implantable Pacemaker
Batteries. In Batteries for Implantable Biomedical Devices; Owens, B. B., Ed.; Springer US:
Boston, MA, 1986; pp 37–49. https://doi.org/10.1007/978-1-4684-9045-9_2.

(28) Owens, B. B. Solid State Electrolytes: Overview of Materials and Applications during
the Last Third of the Twentieth Century. Journal of Power Sources 2000, 90 (1), 2–8.
https://doi.org/10.1016/S0378-7753(00)00436-5.

165
 References

(29) akram, ghalia. Pacemaker!! A Marvel Called Artificial Heart. Envirocivil.com.

https://envirocivil.com/health-and-wellness/pacemaker-a-marvel-called-artificial-heart/
(accessed 2023-08-17).

(30) Root, M. J. Medical Device Batteriesmedical Device Battery. In Encyclopedia of

Sustainability Science and Technology; Meyers, R. A., Ed.; Springer: New York, NY, 2012;
pp 6498–6519. https://doi.org/10.1007/978-1-4419-0851-3_658.

(31) Goodenough, J. B.; Hong, H. Y.-P.; Kafalas, J. A. Fast Na+-Ion Transport in Skeleton
Structures. Materials Research Bulletin 1976, 11 (2), 203–220.
https://doi.org/10.1016/0025-5408(76)90077-5.

(32) Wright, P. V. Electrical Conductivity in Ionic Complexes of Poly(Ethylene Oxide). Brit.

Poly. J. 1975, 7 (5), 319–327. https://doi.org/10.1002/pi.4980070505.

(33) Geng, H.; Mei, A.; Lin, Y.; Nan, C. Effect of Sintering Atmosphere on Ionic Conduction
and Structure of Li0.5La0.5TiO3 Solid Electrolytes. Materials Science and Engineering: B
2009, 164 (2), 91–95. https://doi.org/10.1016/j.mseb.2009.07.011.

(34) Sun, Y.; Guan, P.; Liu, Y.; Xu, H.; Li, S.; Chu, D. Recent Progress in Lithium Lanthanum
Titanate Electrolyte towards All Solid-State Lithium Ion Secondary Battery. Critical Reviews
in Solid State and Materials Sciences 2019, 44 (4), 265–282.
https://doi.org/10.1080/10408436.2018.1485551.

(35) Murugan, R.; Thangadurai, V.; Weppner, W. Fast Lithium Ion Conduction in Garnet-
Type Li7La3Zr2O12. Angewandte Chemie International Edition 2007, 46 (41), 7778–7781.
https://doi.org/10.1002/anie.200701144.

(36) Thangadurai, V.; Narayanan, S.; Pinzaru, D. Garnet-Type Solid-State Fast Li Ion
Conductors for Li Batteries: Critical Review. Chem. Soc. Rev. 2014, 43 (13), 4714–4727.
https://doi.org/10.1039/C4CS00020J.

(37) Fahys, B.; Robert, G. SUPERIONIC CONDUCTION IN Li2S - P2S5 - Lil - GLASSES.

(38) Minami, T.; Hayashi, A.; Tatsumisago, M. Recent Progress of Glass and Glass-Ceramics
as Solid Electrolytes for Lithium Secondary Batteries. Solid State Ionics 2006, 177 (26),
2715–2720. https://doi.org/10.1016/j.ssi.2006.07.017.

(39) Kanno, R.; Hata, T.; Kawamoto, Y.; Irie, M. Synthesis of a New Lithium Ionic Conductor,
Thio-LISICON–Lithium Germanium Sulfide System. Solid State Ionics 2000, 130 (1), 97–
104. https://doi.org/10.1016/S0167-2738(00)00277-0.

(40) Kamaya, N.; Homma, K.; Yamakawa, Y.; Hirayama, M.; Kanno, R.; Yonemura, M.;
Kamiyama, T.; Kato, Y.; Hama, S.; Kawamoto, K.; Mitsui, A. A Lithium Superionic Conductor.
Nature Mater 2011, 10 (9), 682–686. https://doi.org/10.1038/nmat3066.

(41) Kato, Y.; Hori, S.; Saito, T.; Suzuki, K.; Hirayama, M.; Mitsui, A.; Yonemura, M.; Iba, H.;
Kanno, R. High-Power All-Solid-State Batteries Using Sulfide Superionic Conductors. Nat
Energy 2016, 1 (4), 1–7. https://doi.org/10.1038/nenergy.2016.30.

166
 References

(42) Deiseroth, H.-J.; Kong, S.-T.; Eckert, H.; Vannahme, J.; Reiner, C.; Zaiß, T.; Schlosser, M.
Li6PS5X: A Class of Crystalline Li-Rich Solids With an Unusually High Li+ Mobility.
Angewandte Chemie International Edition 2008, 47 (4), 755–758.
https://doi.org/10.1002/anie.200703900.

(43) Walther, F.; Koerver, R.; Fuchs, T.; Ohno, S.; Sann, J.; Rohnke, M.; Zeier, W. G.; Janek,
J. Visualization of the Interfacial Decomposition of Composite Cathodes in Argyrodite-
Based All-Solid-State Batteries Using Time-of-Flight Secondary-Ion Mass Spectrometry.
Chem. Mater. 2019, 31 (10), 3745–3755.
https://doi.org/10.1021/acs.chemmater.9b00770.

(44) Nie, X.; Hu, J.; Li, C. Halide-Based Solid Electrolytes: The History, Progress, and
Challenges. Interdisciplinary Materials 2023, 2 (3), 365–389.
https://doi.org/10.1002/idm2.12090.

(45) Asano, T.; Sakai, A.; Ouchi, S.; Sakaida, M.; Miyazaki, A.; Hasegawa, S. Solid Halide
Electrolytes with High Lithium-Ion Conductivity for Application in 4 V Class Bulk-Type All-
Solid-State Batteries. Advanced Materials 2018, 30 (44), 1803075.
https://doi.org/10.1002/adma.201803075.

(46) Liu, Z.; Ma, S.; Liu, J.; Xiong, S.; Ma, Y.; Chen, H. High Ionic Conductivity Achieved in
Li3Y(Br3Cl3) Mixed Halide Solid Electrolyte via Promoted Diffusion Pathways and
Enhanced Grain Boundary. ACS Energy Lett. 2021, 6 (1), 298–304.
https://doi.org/10.1021/acsenergylett.0c01690.

(47) Li, X.; Liang, J.; Luo, J.; Banis, M. N.; Wang, C.; Li, W.; Deng, S.; Yu, C.; Zhao, F.; Hu, Y.;
Sham, T.-K.; Zhang, L.; Zhao, S.; Lu, S.; Huang, H.; Li, R.; Adair, K. R.; Sun, X. Air-Stable
Li3InCl6 Electrolyte with High Voltage Compatibility for All-Solid-State Batteries. Energy
Environ. Sci. 2019, 12 (9), 2665–2671. https://doi.org/10.1039/C9EE02311A.

(48) Liang, J.; Li, X.; Wang, S.; Adair, K. R.; Li, W.; Zhao, Y.; Wang, C.; Hu, Y.; Zhang, L.; Zhao,
S.; Lu, S.; Huang, H.; Li, R.; Mo, Y.; Sun, X. Site-Occupation-Tuned Superionic
LixScCl3+xHalide Solid Electrolytes for All-Solid-State Batteries. J. Am. Chem. Soc. 2020,
142 (15), 7012–7022. https://doi.org/10.1021/jacs.0c00134.

(49) Zhou, L.; Zuo, T.; Li, C.; Zhang, Q.; Janek, J.; Nazar, L. F. Li3–xZrx(Ho/Lu)1–xCl6 Solid
Electrolytes Enable Ultrahigh-Loading Solid-State Batteries with a Prelithiated Si Anode.
ACS Energy Lett. 2023, 8 (7), 3102–3111. https://doi.org/10.1021/acsenergylett.3c00763.

(50) Yin, Y.-C.; Yang, J.-T.; Luo, J.-D.; Lu, G.-X.; Huang, Z.; Wang, J.-P.; Li, P.; Li, F.; Wu, Y.-C.;
Tian, T.; Meng, Y.-F.; Mo, H.-S.; Song, Y.-H.; Yang, J.-N.; Feng, L.-Z.; Ma, T.; Wen, W.; Gong,
K.; Wang, L.-J.; Ju, H.-X.; Xiao, Y.; Li, Z.; Tao, X.; Yao, H.-B. A LaCl3-Based Lithium Superionic
Conductor Compatible with Lithium Metal. Nature 2023, 616 (7955), 77–83.
https://doi.org/10.1038/s41586-023-05899-8.

(51) Tuo, K.; Sun, C.; López, C. A.; Fernández-Díaz, M. T.; Alonso, J. A. New Superionic Halide
Solid Electrolytes Enabled by Aliovalent Substitution in Li3−xY1−xHfxCl6 for All-Solid-State
Lithium Metal Based Batteries. J. Mater. Chem. A 2023, 11 (29), 15651–15662.
https://doi.org/10.1039/D3TA02781C.

167
 References

(52) Kwak, H.; Han, D.; Lyoo, J.; Park, J.; Jung, S. H.; Han, Y.; Kwon, G.; Kim, H.; Hong, S.;
Nam, K.; Jung, Y. S. New Cost‐Effective Halide Solid Electrolytes for All‐Solid‐State
Batteries: Mechanochemically Prepared Fe3+‐Substituted Li2ZrCl6. Adv. Energy Mater.
2021, 11 (12), 2003190. https://doi.org/10.1002/aenm.202003190.

(53) Zhou, L.; Kwok, C. Y.; Shyamsunder, A.; Zhang, Q.; Wu, X.; Nazar, L. F. A New Halospinel
Superionic Conductor for High-Voltage All Solid State Lithium Batteries. Energy Environ.
Sci. 2020, 13 (7), 2056–2063. https://doi.org/10.1039/D0EE01017K.

(54) Zhou, L.; Zuo, T.-T.; Kwok, C. Y.; Kim, S. Y.; Assoud, A.; Zhang, Q.; Janek, J.; Nazar, L. F.
High Areal Capacity, Long Cycle Life 4 V Ceramic All-Solid-State Li-Ion Batteries Enabled by
Chloride Solid Electrolytes. Nat Energy 2022, 7 (1), 83–93.
https://doi.org/10.1038/s41560-021-00952-0.

(55) Liang, J.; Li, X.; Adair, K. R.; Sun, X. Metal Halide Superionic Conductors for All-Solid-
State Batteries. Acc. Chem. Res. 2021, 54 (4), 1023–1033.
https://doi.org/10.1021/acs.accounts.0c00762.

(56) Kwak, H.; Wang, S.; Park, J.; Liu, Y.; Kim, K. T.; Choi, Y.; Mo, Y.; Jung, Y. S. Emerging
Halide Superionic Conductors for All-Solid-State Batteries: Design, Synthesis, and Practical
Applications. ACS Energy Lett. 2022, 7 (5), 1776–1805.
https://doi.org/10.1021/acsenergylett.2c00438.

(57) Braga, M. H.; Ferreira, J. A.; Stockhausen, V.; Oliveira, J. E.; El-Azab, A. Novel Li3ClO
Based Glasses with Superionic Properties for Lithium Batteries. J. Mater. Chem. A 2014, 2
(15), 5470–5480. https://doi.org/10.1039/C3TA15087A.

(58) Numan-Al-Mobin, A. M.; Schmidt, B.; Lannerd, A.; Viste, M.; Qiao, Q.; Smirnova, A.
Interdigitated Cathode–Electrolyte Architectural Design for Fast-Charging Lithium Metal
Battery with Lithium Oxyhalide Solid-State Electrolyte. Mater. Adv. 2022, 3 (24), 8947–
8957. https://doi.org/10.1039/D2MA00512C.

(59) Zhang, S.; Zhao, F.; Chen, J.; Fu, J.; Luo, J.; Alahakoon, S. H.; Chang, L.-Y.; Feng, R.;
Shakouri, M.; Liang, J.; Zhao, Y.; Li, X.; He, L.; Huang, Y.; Sham, T.-K.; Sun, X. A Family of
Oxychloride Amorphous Solid Electrolytes for Long-Cycling All-Solid-State Lithium
Batteries. Nat Commun 2023, 14 (1), 3780. https://doi.org/10.1038/s41467-023-39197-8.

(60) Tanaka, Y.; Ueno, K.; Mizuno, K.; Takeuchi, K.; Asano, T.; Sakai, A. New Oxyhalide Solid
Electrolytes with High Lithium Ionic Conductivity >10 mS Cm −1 for All‐Solid‐State Batteries.
Angewandte Chemie 2023, 135 (13). https://doi.org/10.1002/ange.202217581.

(61) Hu, L.; Wang, J.; Wang, K.; Gu, Z.; Xi, Z.; Li, H.; Chen, F.; Wang, Y.; Li, Z.; Ma, C. A Cost-
Effective, Ionically Conductive and Compressible Oxychloride Solid-State Electrolyte for
Stable All-Solid-State Lithium-Based Batteries. Nat Commun 2023, 14 (1), 3807.
https://doi.org/10.1038/s41467-023-39522-1.

(62) Landgraf, V.; Famprikis, T.; de Leeuw, J.; Bannenberg, L. J.; Ganapathy, S.; Wagemaker,
M. Li5NCl2: A Fully-Reduced, Highly-Disordered Nitride-Halide Electrolyte for Solid-State

168
 References

Batteries with Lithium-Metal Anodes. ACS Appl. Energy Mater. 2023, 6 (3), 1661–1672.
https://doi.org/10.1021/acsaem.2c03551.

(63) Liu, J.; Zhou, F.; Wang, S.; Zeng, R. Novel Nitride-Based Electrodes for Solid-State
Batteries. In Solid State Batteries Volume 2: Materials and Advanced Devices; ACS
Symposium Series; American Chemical Society, 2022; Vol. 1414, pp 15–38.
https://doi.org/10.1021/bk-2022-1414.ch002.

(64) Sharafi, A.; Meyer, H. M.; Nanda, J.; Wolfenstine, J.; Sakamoto, J. Characterizing the Li–
Li7La3Zr2O12 Interface Stability and Kinetics as a Function of Temperature and Current
Density. Journal of Power Sources 2016, 302, 135–139.
https://doi.org/10.1016/j.jpowsour.2015.10.053.

(65) Yu, T.; Liang, J.; Luo, L.; Wang, L.; Zhao, F.; Xu, G.; Bai, X.; Yang, R.; Zhao, S.; Wang, J.;
Yu, J.; Sun, X. Superionic Fluorinated Halide Solid Electrolytes for Highly Stable Li-Metal in
All-Solid-State Li Batteries. Advanced Energy Materials 2021, 11 (36), 2101915.
https://doi.org/10.1002/aenm.202101915.

(66) Janek, J.; Zeier, W. G. Challenges in Speeding up Solid-State Battery Development. Nat
Energy 2023, 8 (3), 230–240. https://doi.org/10.1038/s41560-023-01208-9.

(67) Tan, D. H. S.; Wu, E. A.; Nguyen, H.; Chen, Z.; Marple, M. A. T.; Doux, J.-M.; Wang, X.;
Yang, H.; Banerjee, A.; Meng, Y. S. Elucidating Reversible Electrochemical Redox of Li 6 PS
5 Cl Solid Electrolyte. ACS Energy Lett. 2019, 4 (10), 2418–2427.
https://doi.org/10.1021/acsenergylett.9b01693.

(68) Wang, C.; Adair, K.; Sun, X. All-Solid-State Lithium Metal Batteries with Sulfide
Electrolytes: Understanding Interfacial Ion and Electron Transport. Acc. Mater. Res. 2022,
3 (1), 21–32. https://doi.org/10.1021/accountsmr.1c00137.

(69) Song, S.; Hori, S.; Li, Y.; Suzuki, K.; Matsui, N.; Hirayama, M.; Saito, T.; Kamiyama, T.;
Kanno, R. Material Search for a Li10GeP2S12-Type Solid Electrolyte in the Li–P–S–X (X =
Br, I) System via Clarification of the Composition–Structure–Property Relationships. Chem.
Mater. 2022. https://doi.org/10.1021/acs.chemmater.2c01608.

(70) Riegger, L. M.; Schlem, R.; Sann, J.; Zeier, W. G.; Janek, J. Lithium‐Metal Anode
Instability of the Superionic Halide Solid Electrolytes and the Implications for Solid‐State
Batteries. Angew. Chem. Int. Ed. 2021, 60 (12), 6718–6723.
https://doi.org/10.1002/anie.202015238.

(71) Wenzel, S.; Leichtweiss, T.; Krüger, D.; Sann, J.; Janek, J. Interphase Formation on
Lithium Solid Electrolytes—An in Situ Approach to Study Interfacial Reactions by
Photoelectron Spectroscopy. Solid State Ionics 2015, 278, 98–105.
https://doi.org/10.1016/j.ssi.2015.06.001.

(72) Zhu, Y.; He, X.; Mo, Y. Origin of Outstanding Stability in the Lithium Solid Electrolyte
Materials: Insights from Thermodynamic Analyses Based on First-Principles Calculations.
ACS Appl. Mater. Interfaces 2015, 7 (42), 23685–23693.
https://doi.org/10.1021/acsami.5b07517.

169
 References

(73) Walther, F.; Randau, S.; Schneider, Y.; Sann, J.; Rohnke, M.; Richter, F. H.; Zeier, W. G.;
Janek, J. Influence of Carbon Additives on the Decomposition Pathways in Cathodes of
Lithium Thiophosphate-Based All-Solid-State Batteries. Chem. Mater. 2020, 32 (14), 6123–
6136. https://doi.org/10.1021/acs.chemmater.0c01825.

(74) Zhang, X.; Li, X.; Weng, S.; Wu, S.; Liu, Q.; Cao, M.; Li, Y.; Wang, Z.; Zhu, L.; Xiao, R.; Su,
D.; Yu, X.; Li, H.; Chen, L.; Wang, Z.; Wang, X. Spontaneous Gas–Solid Reaction on Sulfide
Electrolytes for High-Performance All-Solid-State Batteries. Energy Environ. Sci. 2023,
10.1039.D2EE03358E. https://doi.org/10.1039/D2EE03358E.

(75) Nikodimos, Y.; Su, W.-N.; Hwang, B. J. Halide Solid-State Electrolytes: Stability and
Application for High Voltage All-Solid-State Li Batteries. Advanced Energy Materials 2023,
13 (3), 2202854. https://doi.org/10.1002/aenm.202202854.

(76) Rajendran, S.; Pilli, A.; Omolere, O.; Kelber, J.; Arava, L. M. R. An All-Solid-State Battery
with a Tailored Electrode–Electrolyte Interface Using Surface Chemistry and Interlayer-
Based Approaches. Chem. Mater. 2021, 33 (9), 3401–3412.
https://doi.org/10.1021/acs.chemmater.1c00747.

(77) Walther, F.; Strauss, F.; Wu, X.; Mogwitz, B.; Hertle, J.; Sann, J.; Rohnke, M.; Brezesinski,
T.; Janek, J. The Working Principle of a Li2CO3/LiNbO3 Coating on NCM for Thiophosphate-
Based All-Solid-State Batteries. Chem. Mater. 2021, 33 (6), 2110–2125.
https://doi.org/10.1021/acs.chemmater.0c04660.

(78) Hood, Z. D.; Mane, A. U.; Sundar, A.; Tepavcevic, S.; Zapol, P.; Eze, U. D.; Adhikari, S. P.;
Lee, E.; Sterbinsky, G. E.; Elam, J. W.; Connell, J. G. Multifunctional Coatings on Sulfide-
Based Solid Electrolyte Powders with Enhanced Processability, Stability, and Performance
for Solid-State Batteries. Advanced Materials 2023, 35 (21), 2300673.
https://doi.org/10.1002/adma.202300673.

(79) Wang, S.; Xu, X.; Cui, C.; Zeng, C.; Liang, J.; Fu, J.; Zhang, R.; Zhai, T.; Li, H. Air Sensitivity
and Degradation Evolution of Halide Solid State Electrolytes upon Exposure. Advanced
Functional Materials 2022, 32 (7), 2108805. https://doi.org/10.1002/adfm.202108805.

(80) Shi, T.; Tu, Q.; Tian, Y.; Xiao, Y.; Miara, L. J.; Kononova, O.; Ceder, G. High Active Material
Loading in All-Solid-State Battery Electrode via Particle Size Optimization. Advanced
Energy Materials 2020, 10 (1), 1902881. https://doi.org/10.1002/aenm.201902881.

(81) Doerrer, C.; Capone, I.; Narayanan, S.; Liu, J.; Grovenor, C. R. M.; Pasta, M.; Grant, P. S.
High Energy Density Single-Crystal NMC/Li6PS5Cl Cathodes for All-Solid-State Lithium-
Metal Batteries. ACS Appl. Mater. Interfaces 2021, 13 (31), 37809–37815.
https://doi.org/10.1021/acsami.1c07952.

(82) Kim, K. J.; Rupp, J. L. M. All Ceramic Cathode Composite Design and Manufacturing
towards Low Interfacial Resistance for Garnet-Based Solid-State Lithium Batteries. Energy
Environ. Sci. 2020, 13 (12), 4930–4945. https://doi.org/10.1039/D0EE02062A.

170
 References

(83) Bielefeld, A.; Weber, D. A.; Janek, J. Modeling Effective Ionic Conductivity and Binder
Influence in Composite Cathodes for All-Solid-State Batteries. ACS Appl. Mater. Interfaces
2020, 12 (11), 12821–12833. https://doi.org/10.1021/acsami.9b22788.

(84) Kim, S. Y.; Cha, H.; Kostecki, R.; Chen, G. Composite Cathode Design for High-Energy
All-Solid-State Lithium Batteries with Long Cycle Life. ACS Energy Lett. 2023, 8 (1), 521–
528. https://doi.org/10.1021/acsenergylett.2c02414.

(85) Ke, X.; Wang, Y.; Ren, G.; Yuan, C. Towards Rational Mechanical Design of Inorganic
Solid Electrolytes for All-Solid-State Lithium Ion Batteries. Energy Storage Materials 2020,
26, 313–324. https://doi.org/10.1016/j.ensm.2019.08.029.

(86) Jiang, M.; Mukherjee, S.; Chen, Z. W.; Chen, L. X.; Li, M. L.; Xiao, H. Y.; Gao, C.; Singh, C.
V. Materials Perspective on New Lithium Chlorides and Bromides: Insights into Thermo-
Physical Properties. Phys. Chem. Chem. Phys. 2020, 22 (39), 22758–22767.
https://doi.org/10.1039/D0CP02946G.

(87) Papakyriakou, M.; Lu, M.; Liu, Y.; Liu, Z.; Chen, H.; McDowell, M. T.; Xia, S. Mechanical
Behavior of Inorganic Lithium-Conducting Solid Electrolytes. Journal of Power Sources
2021, 516, 230672. https://doi.org/10.1016/j.jpowsour.2021.230672.

(88) Deng, Z.; Wang, Z.; Chu, I.-H.; Luo, J.; Ong, S. P. Elastic Properties of Alkali Superionic
Conductor Electrolytes from First Principles Calculations. J. Electrochem. Soc. 2015, 163
(2), A67. https://doi.org/10.1149/2.0061602jes.

(89) Bucci, G.; Talamini, B.; Renuka Balakrishna, A.; Chiang, Y.-M.; Carter, W. C. Mechanical
Instability of Electrode-Electrolyte Interfaces in Solid-State Batteries. Phys. Rev. Materials
2018, 2 (10), 105407. https://doi.org/10.1103/PhysRevMaterials.2.105407.

(90) Bucci, G.; Swamy, T.; Chiang, Y.-M.; Carter, W. C. Modeling of Internal Mechanical
Failure of All-Solid-State Batteries during Electrochemical Cycling, and Implications for
Battery Design. J. Mater. Chem. A 2017, 5 (36), 19422–19430.
https://doi.org/10.1039/C7TA03199H.

(91) Doux, J.; Nguyen, H.; Tan, D. H. S.; Banerjee, A.; Wang, X.; Wu, E. A.; Jo, C.; Yang, H.;
Meng, Y. S. Stack Pressure Considerations for Room‐Temperature All‐Solid‐State Lithium
Metal Batteries. Adv. Energy Mater. 2020, 10 (1), 1903253.
https://doi.org/10.1002/aenm.201903253.

(92) Koerver, R.; Zhang, W.; de Biasi, L.; Schweidler, S.; Kondrakov, A. O.; Kolling, S.;
Brezesinski, T.; Hartmann, P.; Zeier, W. G.; Janek, J. Chemo-Mechanical Expansion of
Lithium Electrode Materials – on the Route to Mechanically Optimized All-Solid-State
Batteries. Energy Environ. Sci. 2018, 11 (8), 2142–2158.
https://doi.org/10.1039/C8EE00907D.

(93) Culver, S. P.; Koerver, R.; Zeier, W. G.; Janek, J. On the Functionality of Coatings for
Cathode Active Materials in Thiophosphate-Based All-Solid-State Batteries. Advanced
Energy Materials 2019, 9 (24), 1900626. https://doi.org/10.1002/aenm.201900626.

171
 References

(94) Kim, J.; Kim, M. J.; Kim, J.; Lee, J. W.; Park, J.; Wang, S. E.; Lee, S.; Kang, Y. C.; Paik, U.;
Jung, D. S.; Song, T. High-Performance All-Solid-State Batteries Enabled by Intimate
Interfacial Contact Between the Cathode and Sulfide-Based Solid Electrolytes. Advanced
Functional Materials n/a (n/a), 2211355. https://doi.org/10.1002/adfm.202211355.

(95) Lee, K.; Kim, S.; Park, J.; Park, S. H.; Coskun, A.; Jung, D. S.; Cho, W.; Choi, J. W. Selection
of Binder and Solvent for Solution-Processed All-Solid-State Battery. J. Electrochem. Soc.
2017, 164 (9), A2075. https://doi.org/10.1149/2.1341709jes.

(96) Jing, S.; Shen, H.; Huang, Y.; Kuang, W.; Zhang, Z.; Liu, S.; Yin, S.; Lai, Y.; Liu, F. Toward
the Practical and Scalable Fabrication of Sulfide-Based All-Solid-State Batteries:
Exploration of Slurry Process and Performance Enhancement Via the Addition of LiClO4.
Advanced Functional Materials n/a (n/a), 2214274.
https://doi.org/10.1002/adfm.202214274.

(97) Lee, J.; Lee, T.; Char, K.; Kim, K. J.; Choi, J. W. Issues and Advances in Scaling up Sulfide-
Based All-Solid-State Batteries. Acc. Chem. Res. 2021, 54 (17), 3390–3402.
https://doi.org/10.1021/acs.accounts.1c00333.

(98) Boaretto, N.; Garbayo, I.; Valiyaveettil-SobhanRaj, S.; Quintela, A.; Li, C.; Casas-
Cabanas, M.; Aguesse, F. Lithium Solid-State Batteries: State-of-the-Art and Challenges for
Materials, Interfaces and Processing. Journal of Power Sources 2021, 502, 229919.
https://doi.org/10.1016/j.jpowsour.2021.229919.

(99) Xiao, Y.; Turcheniuk, K.; Narla, A.; Song, A.-Y.; Ren, X.; Magasinski, A.; Jain, A.; Huang,
S.; Lee, H.; Yushin, G. Electrolyte Melt Infiltration for Scalable Manufacturing of Inorganic
All-Solid-State Lithium-Ion Batteries. Nat. Mater. 2021, 20 (7), 984–990.
https://doi.org/10.1038/s41563-021-00943-2.

(100) Kim, D. H.; Oh, D. Y.; Park, K. H.; Choi, Y. E.; Nam, Y. J.; Lee, H. A.; Lee, S.-M.; Jung, Y. S.
Infiltration of Solution-Processable Solid Electrolytes into Conventional Li-Ion-Battery
Electrodes for All-Solid-State Li-Ion Batteries. Nano Lett. 2017, 17 (5), 3013–3020.
https://doi.org/10.1021/acs.nanolett.7b00330.

(101) Lee, Y.-G.; Fujiki, S.; Jung, C.; Suzuki, N.; Yashiro, N.; Omoda, R.; Ko, D.-S.; Shiratsuchi,
T.; Sugimoto, T.; Ryu, S.; Ku, J. H.; Watanabe, T.; Park, Y.; Aihara, Y.; Im, D.; Han, I. T. High-
Energy Long-Cycling All-Solid-State Lithium Metal Batteries Enabled by Silver–Carbon
Composite Anodes. Nat Energy 2020, 5 (4), 299–308. https://doi.org/10.1038/s41560-
020-0575-z.

(102) Tan, D. H. S.; Meng, Y. S.; Jang, J. Scaling up High-Energy-Density Sulfidic Solid-State
Batteries: A Lab-to-Pilot Perspective. Joule 2022, 6 (8), 1755–1769.
https://doi.org/10.1016/j.joule.2022.07.002.

(103) Krauskopf, T.; Richter, F. H.; Zeier, W. G.; Janek, J. Physicochemical Concepts of the
Lithium Metal Anode in Solid-State Batteries. Chem. Rev. 2020, 120 (15), 7745–7794.
https://doi.org/10.1021/acs.chemrev.0c00431.

172
 References

(104) Jeppson, D. W.; Ballif, J. L.; Yuan, W. W.; Chou, B. E. Lithium Literature Review:
Lithium’s Properties and Interactions; HEDL-TME-78-15; Hanford Engineering
Development Lab., Richland, WA (United States), 1978. https://doi.org/10.2172/6885395.

(105) Chang, K.; Hallstedt, B. Thermodynamic Assessment of the Li–O System. Calphad 2011,
35 (2), 160–164. https://doi.org/10.1016/j.calphad.2011.02.003.

(106) Kasemchainan, J.; Zekoll, S.; Spencer Jolly, D.; Ning, Z.; Hartley, G. O.; Marrow, J.; Bruce,
P. G. Critical Stripping Current Leads to Dendrite Formation on Plating in Lithium Anode
Solid Electrolyte Cells. Nat. Mater. 2019, 18 (10), 1105–1111.
https://doi.org/10.1038/s41563-019-0438-9.

(107) Kalnaus, S.; Dudney, N. J.; Westover, A. S.; Herbert, E.; Hackney, S. Solid-State
Batteries: The Critical Role of Mechanics. Science 2023, 381 (6664), eabg5998.
https://doi.org/10.1126/science.abg5998.

(108) Kim, J.-S.; Yoon, G.; Kim, S.; Sugata, S.; Yashiro, N.; Suzuki, S.; Lee, M.-J.; Kim, R.;
Badding, M.; Song, Z.; Chang, J.; Im, D. Surface Engineering of Inorganic Solid-State
Electrolytes via Interlayers Strategy for Developing Long-Cycling Quasi-All-Solid-State
Lithium Batteries. Nat Commun 2023, 14 (1), 782. https://doi.org/10.1038/s41467-023-
36401-7.

(109) Fu, K. (Kelvin); Gong, Y.; Fu, Z.; Xie, H.; Yao, Y.; Liu, B.; Carter, M.; Wachsman, E.; Hu, L.
Transient Behavior of the Metal Interface in Lithium Metal–Garnet Batteries. Angewandte
Chemie International Edition 2017, 56 (47), 14942–14947.
https://doi.org/10.1002/anie.201708637.

(110) Spencer-Jolly, D.; Agarwal, V.; Doerrer, C.; Hu, B.; Zhang, S.; Melvin, D. L. R.; Gao, H.;
Gao, X.; Adamson, P.; Magdysyuk, O. V.; Grant, P. S.; House, R. A.; Bruce, P. G. Structural
Changes in the Silver-Carbon Composite Anode Interlayer of Solid-State Batteries. Joule
2023, 7 (3), 503–514. https://doi.org/10.1016/j.joule.2023.02.001.

(111) Cui, C.; Yang, H.; Zeng, C.; Gui, S.; Liang, J.; Xiao, P.; Wang, S.; Huang, G.; Hu, M.; Zhai,
T.; Li, H. Unlocking the in Situ Li Plating Dynamics and Evolution Mediated by Diverse
Metallic Substrates in All-Solid-State Batteries. Science Advances 2022, 8 (43), eadd2000.
https://doi.org/10.1126/sciadv.add2000.

(112) Adeli, P.; Bazak, J. D.; Park, K. H.; Kochetkov, I.; Huq, A.; Goward, G. R.; Nazar, L. F.
Boosting Solid-State Diffusivity and Conductivity in Lithium Superionic Argyrodites by
Halide Substitution. Angewandte Chemie International Edition 2019, 58 (26), 8681–8686.
https://doi.org/10.1002/anie.201814222.

(113) Peng, L.; Yu, C.; Zhang, Z.; Ren, H.; Zhang, J.; He, Z.; Yu, M.; Zhang, L.; Cheng, S.; Xie, J.
Chlorine-Rich Lithium Argyrodite Enabling Solid-State Batteries with Capabilities of High
Voltage, High Rate, Low-Temperature and Ultralong Cyclability. Chemical Engineering
Journal 2022, 430, 132896. https://doi.org/10.1016/j.cej.2021.132896.

(114) Umeshbabu, E.; Maddukuri, S.; Hu, Y.; Fichtner, M.; Munnangi, A. R. Influence of
Chloride Ion Substitution on Lithium-Ion Conductivity and Electrochemical Stability in a

173
 References

Dual-Halogen Solid-State Electrolyte. ACS Appl. Mater. Interfaces 2022, 14 (22), 25448–
25456. https://doi.org/10.1021/acsami.2c04160.

(115) Ganesan, P.; Soans, M.; Cambaz, M. A.; Zimmermanns, R.; Gond, R.; Fuchs, S.; Hu, Y.;
Baumgart, S.; Sotoudeh, M.; Stepien, D.; Stein, H.; Groß, A.; Bresser, D.; Varzi, A.; Fichtner,
M. Fluorine-Substituted Halide Solid Electrolytes with Enhanced Stability toward the
Lithium Metal. ACS Appl. Mater. Interfaces 2023, 15 (32), 38391–38402.
https://doi.org/10.1021/acsami.3c03513.

(116) Wang, K.; Gu, Z.; Xi, Z.; Hu, L.; Ma, C. Li3TiCl6 as Ionic Conductive and Compressible
Positive Electrode Active Material for All-Solid-State Lithium-Based Batteries. Nat
Commun 2023, 14 (1), 1396. https://doi.org/10.1038/s41467-023-37122-7.

(117) Tan, D. H. S.; Chen, Y.-T.; Yang, H.; Bao, W.; Sreenarayanan, B.; Doux, J.-M.; Li, W.; Lu,
B.; Ham, S.-Y.; Sayahpour, B.; Scharf, J.; Wu, E. A.; Deysher, G.; Han, H. E.; Hah, H. J.; Jeong,
H.; Lee, J. B.; Chen, Z.; Meng, Y. S. Carbon-Free High-Loading Silicon Anodes Enabled by
Sulfide Solid Electrolytes. Science 2021, 373 (6562), 1494–1499.
https://doi.org/10.1126/science.abg7217.

(118) Kim, J. Y.; Park, J.; Kang, S. H.; Jung, S.; Shin, D. O.; Lee, M. J.; Oh, J.; Kim, K. M.; Zausch,
J.; Lee, Y.-G.; Lee, Y. M. Revisiting TiS2 as a Diffusion-Dependent Cathode with Promising
Energy Density for All-Solid-State Lithium Secondary Batteries. Energy Storage Materials
2021, 41, 289–296. https://doi.org/10.1016/j.ensm.2021.06.005.

(119) Kim, J. Y.; Jung, S.; Kang, S. H.; Park, J.; Lee, M. J.; Jin, D.; Shin, D. O.; Lee, Y.-G.; Lee, Y.
M. Graphite–Silicon Diffusion-Dependent Electrode with Short Effective Diffusion Length
for High-Performance All-Solid-State Batteries. Advanced Energy Materials 2022, 12 (3),
2103108. https://doi.org/10.1002/aenm.202103108.

(120) Kim, J. Y.; Park, J.; Lee, M. J.; Kang, S. H.; Shin, D. O.; Oh, J.; Kim, J.; Kim, K. M.; Lee, Y.-
G.; Lee, Y. M. Diffusion-Dependent Graphite Electrode for All-Solid-State Batteries with
Extremely High Energy Density. ACS Energy Lett. 2020, 5 (9), 2995–3004.
https://doi.org/10.1021/acsenergylett.0c01628.

(121) Gao, X.; Liu, B.; Hu, B.; Ning, Z.; Jolly, D. S.; Zhang, S.; Perera, J.; Bu, J.; Liu, J.; Doerrer,
C.; Darnbrough, E.; Armstrong, D.; Grant, P. S.; Bruce, P. G. Solid-State Lithium Battery
Cathodes Operating at Low Pressures. Joule 2022, 6 (3), 636–646.
https://doi.org/10.1016/j.joule.2022.02.008.

(122) Quemin, E.; Dugas, R.; Koç, T.; Hennequart, B.; Chometon, R.; Tarascon, J.-M.
Decoupling Parasitic Reactions at the Positive Electrode Interfaces in Argyrodite-Based
Systems. ACS Appl. Mater. Interfaces 2022. https://doi.org/10.1021/acsami.2c13150.

(123) Koç, T.; Marchini, F.; Rousse, G.; Dugas, R.; Tarascon, J.-M. In Search of the Best Solid
Electrolyte-Layered Oxide Pairing for Assembling Practical All-Solid-State Batteries. ACS
Appl. Energy Mater. 2021, 4 (12), 13575–13585.
https://doi.org/10.1021/acsaem.1c02187.

174
 References

(124) Koç, T. In Search of the Best Solid Electrolyte-Layered Oxide Pair in All-Solid-State
Batteries. PhD Thesis, Sorbonne Université, Laboratoire Chimie du Solide et Energie,
Collège de France, 2022. https://theses.hal.science/tel-04137255 (accessed 2023-09-19).

(125) Koç, T.; Hallot, M.; Quemin, E.; Hennequart, B.; Dugas, R.; Abakumov, A. M.; Lethien,
C.; Tarascon, J.-M. Toward Optimization of the Chemical/Electrochemical Compatibility of
Halide Solid Electrolytes in All-Solid-State Batteries. ACS Energy Lett. 2022, 7 (9), 2979–
2987. https://doi.org/10.1021/acsenergylett.2c01668.

(126) Rosenbach, C.; Walther, F.; Ruhl, J.; Hartmann, M.; Hendriks, T. A.; Ohno, S.; Janek, J.;
Zeier, W. Visualizing the Chemical Incompatibility of Halide and Sulfide-Based Electrolytes
in Solid-State Batteries. Advanced Energy Materials 2023, 13 (6), 2203673.
https://doi.org/10.1002/aenm.202203673.

(127) Wang, S.; Bai, Q.; Nolan, A. M.; Liu, Y.; Gong, S.; Sun, Q.; Mo, Y. Lithium Chlorides and
Bromides as Promising Solid-State Chemistries for Fast Ion Conductors with Good
Electrochemical Stability. Angewandte Chemie International Edition 2019, 58 (24), 8039–
8043. https://doi.org/10.1002/anie.201901938.

(128) Ziegenhain, G.; M. Urbassek, H. Effect of Material Stiffness on Hardness: A

Computational Study Based on Model Potentials. Philosophical Magazine 2009, 89 (26),
2225–2238. https://doi.org/10.1080/14786430903022697.

(129) Auvergniot, J.; Cassel, A.; Ledeuil, J.-B.; Viallet, V.; Seznec, V.; Dedryvère, R. Interface
Stability of Argyrodite Li6PS5Cl toward LiCoO2, LiNi1/3Co1/3Mn1/3O2, and LiMn2O4 in
Bulk All-Solid-State Batteries. Chem. Mater. 2017, 29 (9), 3883–3890.
https://doi.org/10.1021/acs.chemmater.6b04990.

(130) Dewald, G. F.; Ohno, S.; Kraft, M. A.; Koerver, R.; Till, P.; Vargas-Barbosa, N. M.; Janek,
J.; Zeier, W. G. Experimental Assessment of the Practical Oxidative Stability of Lithium
Thiophosphate Solid Electrolytes. Chem. Mater. 2019, 31 (20), 8328–8337.
https://doi.org/10.1021/acs.chemmater.9b01550.

(131) Deng, S.; Sun, Y.; Li, X.; Ren, Z.; Liang, J.; Doyle-Davis, K.; Liang, J.; Li, W.; Norouzi Banis,
M.; Sun, Q.; Li, R.; Hu, Y.; Huang, H.; Zhang, L.; Lu, S.; Luo, J.; Sun, X. Eliminating the
Detrimental Effects of Conductive Agents in Sulfide-Based Solid-State Batteries. ACS
Energy Lett. 2020, 5 (4), 1243–1251. https://doi.org/10.1021/acsenergylett.0c00256.

(132) Jung, S.-K.; Gwon, H.; Lee, S.-S.; Kim, H.; Lee, J. C.; Chung, J. G.; Park, S. Y.; Aihara, Y.;
Im, D. Understanding the Effects of Chemical Reactions at the Cathode–Electrolyte
Interface in Sulfide Based All-Solid-State Batteries. J. Mater. Chem. A 2019, 7 (40), 22967–
22976. https://doi.org/10.1039/C9TA08517C.

(133) Liu, X.; Zheng, B.; Zhao, J.; Zhao, W.; Liang, Z.; Su, Y.; Xie, C.; Zhou, K.; Xiang, Y.; Zhu, J.;
Wang, H.; Zhong, G.; Gong, Z.; Huang, J.; Yang, Y. Electrochemo‐Mechanical Effects on
Structural Integrity of Ni‐Rich Cathodes with Different Microstructures in All Solid‐State
Batteries. Adv. Energy Mater. 2021, 11 (8), 2003583.
https://doi.org/10.1002/aenm.202003583.

175
 References

(134) Whittingham, M. S. Chemistry of Intercalation Compounds: Metal Guests in

Chalcogenide Hosts. Progress in Solid State Chemistry 1978, 12 (1), 41–99.
https://doi.org/10.1016/0079-6786(78)90003-1.

(135) Umrigar, C.; Ellis, D. E.; Wang, D.-S.; Krakauer, H.; Posternak, M. Band Structure,
Intercalation, and Interlayer Interactions of Transition-Metal Dichalcogenides: Ti S 2 and
LiTi S 2. Phys. Rev. B 1982, 26 (9), 4935–4950. https://doi.org/10.1103/PhysRevB.26.4935.

(136) Nakhal, S.; Lerch, M.; Koopman, J.; Islam, M. M.; Bredow, T. Crystal Structure of 3R-
LiTiS2 and Its Stability Compared to Other Polymorphs. Zeitschrift für anorganische und
allgemeine Chemie 2013, 639 (15), 2822–2825. https://doi.org/10.1002/zaac.201300330.

(137) Zhang, Z.; Dong, C.; Guan, C.; Yang, L.; Luo, X.; Li, A. Sealed-Tube Synthesis and Phase
Diagram of LixTiS2 (0≤x≤1). Materials Research Bulletin 2015, 61, 499–503.
https://doi.org/10.1016/j.materresbull.2014.10.045.

(138) Colbow, K. M.; Dahn, J. R.; Haering, R. R. The 3R Phase of LixTiS2. Journal of Power
Sources 1989, 26 (3–4), 301–307. https://doi.org/10.1016/0378-7753(89)80138-7.

(139) Kim, J. Y.; Jung, S.; Kang, S. H.; Lee, M. J.; Jin, D.; Shin, D. O.; Lee, Y.-G.; Lee, Y. M. All-
Solid-State Hybrid Electrode Configuration for High-Performance All-Solid-State Batteries:
Comparative Study with Composite Electrode and Diffusion-Dependent Electrode. Journal
of Power Sources 2022, 518, 230736. https://doi.org/10.1016/j.jpowsour.2021.230736.

(140) Suryanarayana, C. Mechanical Alloying and Milling. Progress in Materials Science 2001,
46 (1), 1–184. https://doi.org/10.1016/S0079-6425(99)00010-9.

(141) Tarascon, J.-M.; Morcrette, M.; Saint, J.; Aymard, L.; Janot, R. On the Benefits of Ball
Milling within the Field of Rechargeable Li-Based Batteries. Comptes Rendus Chimie 2005,
8 (1), 17–26. https://doi.org/10.1016/j.crci.2004.12.006.

(142) Sivakkumar, S. R.; Milev, A. S.; Pandolfo, A. G. Effect of Ball-Milling on the Rate and
Cycle-Life Performance of Graphite as Negative Electrodes in Lithium-Ion Capacitors.
Electrochimica Acta 2011, 56 (27), 9700–9706.
https://doi.org/10.1016/j.electacta.2011.06.060.

(143) Doyle, M.; Newman, J.; Reimers, J. A Quick Method of Measuring the Capacity versus
Discharge Rate for a Dual Lithium-Ion Insertion Cell Undergoing Cycling. Journal of Power
Sources 1994, 52 (2), 211–216. https://doi.org/10.1016/0378-7753(94)02012-4.

(144) Winter, R.; Heitjans, P. Intergranular Structure of Nanocrystalline Layered LixTiS2 as

Derived from 7Li NMR Spectroscopy. Journal of Non-Crystalline Solids 2001, 293–295, 19–
24. https://doi.org/10.1016/S0022-3093(01)00640-8.

(145) Winter, R.; Heitjans, P. Li+ Diffusion and Its Structural Basis in the Nanocrystalline and
Amorphous Forms of Two-Dimensionally Ion-Conducting LixTiS2. J. Phys. Chem. B 2001,
105 (26), 6108–6115. https://doi.org/10.1021/jp011200f.

176
 References

(146) Thompson, A. H. Lithium Ordering in LixTiS2. Phys. Rev. Lett. 1978, 40 (23), 1511–1514.
https://doi.org/10.1103/PhysRevLett.40.1511.

(147) Küchler, W.; Heitjans, P.; Payer, A.; Schöllhorn, R. 7Li NMR Relaxation by Diffusion in
Hexagonal and Cubic LixTiS2. Solid State Ionics 1994, 70–71, 434–438.
https://doi.org/10.1016/0167-2738(94)90350-6.

(148) Bodenez, V.; Dupont, L.; Morcrette, M.; Surcin, C.; Murphy, D. W.; Tarascon, J.-M.
Copper Extrusion/Reinjection in Cu-Based Thiospinels by Electrochemical and Chemical
Routes. Chem. Mater. 2006, 18 (18), 4278–4287. https://doi.org/10.1021/cm060436z.

(149) Tanibata, N.; Kato, M.; Takimoto, S.; Takeda, H.; Nakayama, M.; Sumi, H. High
Formability and Fast Lithium Diffusivity in Metastable Spinel Chloride for Rechargeable All-
Solid-State Lithium-Ion Batteries. Advanced Energy and Sustainability Research 2020, 1
(1), 2000025. https://doi.org/10.1002/aesr.202000025.

(150) Santhosha, A. L.; Medenbach, L.; Buchheim, J. R.; Adelhelm, P. The Indium−Lithium
Electrode in Solid-State Lithium-Ion Batteries: Phase Formation, Redox Potentials, and
Interface Stability. Batteries & Supercaps 2019, 2 (6), 524–529.
https://doi.org/10.1002/batt.201800149.

(151) Lu, Y.; Zhao, C.; Yuan, H.; Cheng, X.; Huang, J.; Zhang, Q. Critical Current Density in
Solid‐State Lithium Metal Batteries: Mechanism, Influences, and Strategies. Adv. Funct.
Mater. 2021, 31 (18), 2009925. https://doi.org/10.1002/adfm.202009925.

(152) Ham, S.-Y.; Yang, H.; Nunez-cuacuas, O.; Tan, D. H. S.; Chen, Y.-T.; Deysher, G.; Cronk,
A.; Ridley, P.; Doux, J.-M.; Wu, E. A.; Jang, J.; Meng, Y. S. Assessing the Critical Current
Density of All-Solid-State Li Metal Symmetric and Full Cells. Energy Storage Materials 2023,
55, 455–462. https://doi.org/10.1016/j.ensm.2022.12.013.

(153) Fuchs, T.; Haslam, C. G.; Richter, F. H.; Sakamoto, J.; Janek, J. Evaluating the Use of
Critical Current Density Tests of Symmetric Lithium Transference Cells with Solid
Electrolytes. Advanced Energy Materials 2023, 2302383.
https://doi.org/10.1002/aenm.202302383.

(154) Wang, M. J.; Choudhury, R.; Sakamoto, J. Characterizing the Li-Solid-Electrolyte

Interface Dynamics as a Function of Stack Pressure and Current Density. Joule 2019, 3 (9),
2165–2178. https://doi.org/10.1016/j.joule.2019.06.017.

(155) Viswanathan, L.; Virkar, A. V. Wetting Characteristics of Sodium on Β″-Alumina and on

Nasicon. J Mater Sci 1982, 17 (3), 753–759. https://doi.org/10.1007/BF00540372.

(156) Lee, H.; Chen, S.; Ren, X.; Martinez, A.; Shutthanandan, V.; Vijayakumar, M.; Han, K. S.;
Li, Q.; Liu, J.; Xu, W.; Zhang, J.-G. Electrode Edge Effects and the Failure Mechanism of
Lithium-Metal Batteries. ChemSusChem 2018, 11 (21), 3821–3828.
https://doi.org/10.1002/cssc.201801445.

177
 References

(157) Sannier, L.; Bouchet, R.; Rosso, M.; Tarascon, J.-M. Evaluation of GPE Performances in
Lithium Metal Battery Technology by Means of Simple Polarization Tests. Journal of Power
Sources 2006, 158 (1), 564–570. https://doi.org/10.1016/j.jpowsour.2005.09.026.

(158) Sannier, L. Mise Au Point d’un Électrolyte Gélifié Pour Accumulateurs Au Lithium
Fonctionnant à Température Ambiante. These de doctorat, Amiens, 2003.
https://www.theses.fr/2003AMIE0311 (accessed 2023-10-12).

(159) Otto, S.-K.; Riegger, L. M.; Fuchs, T.; Kayser, S.; Schweitzer, P.; Burkhardt, S.; Henss, A.;
Janek, J. In Situ Investigation of Lithium Metal–Solid Electrolyte Anode Interfaces with ToF-
SIMS. Advanced Materials Interfaces 2022, 9 (13), 2102387.
https://doi.org/10.1002/admi.202102387.

(160) Pantyukhina, M. I.; Andreev, O. L.; Martem’yanova, Z. S.; Batalov, N. N. Cation

Conductivity of Li 8 ZrO 6 –LiYO 2 Solid Solutions. Inorganic Materials 2004, 40 (4), 404–
406. https://doi.org/10.1023/B:INMA.0000023965.10888.eb.

(161) Oh, J.; Choi, S. H.; Kim, J. Y.; Lee, J.; Lee, T.; Lee, N.; Lee, T.; Sohn, Y.; Chung, W. J.; Bae,
K. Y.; Son, S.; Choi, J. W. Anode-Less All-Solid-State Batteries Operating at Room
Temperature and Low Pressure. Advanced Energy Materials 2023, 13 (38), 2301508.
https://doi.org/10.1002/aenm.202301508.

(162) Belov, N. A.; Eskin, D. G.; Aksenov, A. A. Multicomponent Phase Diagrams: Applications
for Commercial Aluminum Alloys - 1st Edition, 1st ed.; Elsevier Science, 2005.

(163) Wan, H.; Wang, Z.; Zhang, W.; He, X.; Wang, C. Interface Design for All-Solid-State
Lithium Batteries. Nature 2023, 1–6. https://doi.org/10.1038/s41586-023-06653-w.

(164) Larcher, D.; Beattie, S.; Morcrette, M.; Edström, K.; Jumas, J.-C.; Tarascon, J.-M. Recent
Findings and Prospects in the Field of Pure Metals as Negative Electrodes for Li-Ion
Batteries. J. Mater. Chem. 2007, 17 (36), 3759–3772. https://doi.org/10.1039/B705421C.

(165) Okamoto, H. Supplemental Literature Review of Binary Phase Diagrams: Ag-Li, Ag-Sn,
Be-Pu, C-Mn, C-Si, Ca-Li, Cd-Pu, Cr-Ti, Cr-V, Cu-Li, La-Sc, and Li-Sc. J. Phase Equilib. Diffus.
2017, 38 (1), 70–81. https://doi.org/10.1007/s11669-016-0504-9.

(166) Sharafi, A.; Kazyak, E.; Davis, A. L.; Yu, S.; Thompson, T.; Siegel, D. J.; Dasgupta, N. P.;
Sakamoto, J. Surface Chemistry Mechanism of Ultra-Low Interfacial Resistance in the
Solid-State Electrolyte Li7La3Zr2O12. Chem. Mater. 2017, 29 (18), 7961–7968.
https://doi.org/10.1021/acs.chemmater.7b03002.

(167) Han, X.; Gong, Y.; Fu, K. (Kelvin); He, X.; Hitz, G. T.; Dai, J.; Pearse, A.; Liu, B.; Wang, H.;
Rubloff, G.; Mo, Y.; Thangadurai, V.; Wachsman, E. D.; Hu, L. Negating Interfacial
Impedance in Garnet-Based Solid-State Li Metal Batteries. Nature Mater 2017, 16 (5),
572–579. https://doi.org/10.1038/nmat4821.

(168) Wu, C.; Emley, B.; Zhao, L.; Liang, Y.; Ai, Q.; Chen, Z.; Robles Hernández, F. C.; Wang, F.;
Risal, S.; Guo, H.; Lou, J.; Yao, Y.; Fan, Z. Understanding the Chemomechanical Function of

178
 References

the Silver–Carbon Interlayer in Sheet-Type All-Solid-State Lithium–Metal Batteries. Nano

Lett. 2023, 23 (10), 4415–4422. https://doi.org/10.1021/acs.nanolett.3c00720.

(169) Suzuki, N.; Yashiro, N.; Fujiki, S.; Omoda, R.; Shiratsuchi, T.; Watanabe, T.; Aihara, Y.
Highly Cyclable All-Solid-State Battery with Deposition-Type Lithium Metal Anode Based
on Thin Carbon Black Layer. Advanced Energy and Sustainability Research 2021, 2 (11),
2100066. https://doi.org/10.1002/aesr.202100066.

(170) Lee, J.; Choi, S. H.; Im, G.; Lee, K.-J.; Lee, T.; Oh, J.; Lee, N.; Kim, H.; Kim, Y.; Lee, S.; Choi,
J. W. Room-Temperature Anode-Less All-Solid-State Batteries via the Conversion Reaction
of Metal Fluorides. Advanced Materials 2022, 34 (40), 2203580.
https://doi.org/10.1002/adma.202203580.

(171) Gao, Q.; Wu, D.; Zhu, X.; Lu, P.; Ma, T.; Yang, M.; Chen, L.; Li, H.; Wu, F. Dendrite-Free
Lithium-Metal All-Solid-State Batteries by Solid-Phase Passivation. Nano Energy 2023, 117,
108922. https://doi.org/10.1016/j.nanoen.2023.108922.

(172) Huo, H.; Chen, Y.; Li, R.; Zhao, N.; Luo, J.; Pereira Da Silva, J. G.; Mücke, R.; Kaghazchi,
P.; Guo, X.; Sun, X. Design of a Mixed Conductive Garnet/Li Interface for Dendrite-Free
Solid Lithium Metal Batteries. Energy Environ. Sci. 2020, 13 (1), 127–134.
https://doi.org/10.1039/C9EE01903K.

(173) Harry, K. J.; Hallinan, D. T.; Parkinson, D. Y.; MacDowell, A. A.; Balsara, N. P. Detection
of Subsurface Structures underneath Dendrites Formed on Cycled Lithium Metal
Electrodes. Nature Mater 2014, 13 (1), 69–73. https://doi.org/10.1038/nmat3793.

(174) Maslyn, J. A.; Loo, W. S.; McEntush, K. D.; Oh, H. J.; Harry, K. J.; Parkinson, D. Y.; Balsara,
N. P. Growth of Lithium Dendrites and Globules through a Solid Block Copolymer
Electrolyte as a Function of Current Density. J. Phys. Chem. C 2018, 122 (47), 26797–
26804. https://doi.org/10.1021/acs.jpcc.8b06355.

(175) Maslyn, J. A.; Frenck, L.; Loo, W. S.; Parkinson, D. Y.; Balsara, N. P. Extended Cycling
through Rigid Block Copolymer Electrolytes Enabled by Reducing Impurities in Lithium
Metal Electrodes. ACS Appl. Energy Mater. 2019, 2 (11), 8197–8206.
https://doi.org/10.1021/acsaem.9b01685.

(176) Jung, Y.-C.; Hwang, C.; Kwak, M.-J.; Jeon, S.-J.; Lee, Y. J.; Kwak, W.-J.; Kim, H.; Kim, K.;
Cho, W.; Yu, J.-S. On-Site Formation of Silver Decorated Carbon as Anodeless Electrode for
High-Energy Density All-Solid-State Batteries. J. Mater. Chem. A 2023.
https://doi.org/10.1039/D3TA05307E.

(177) Risal, S.; Wu, C.; Wang, F.; Risal, S.; Robles Hernandez, F. C.; Zhu, W.; Yao, Y.; Fan, Z.
Silver-Carbon Interlayers in Anode-Free Solid-State Lithium Metal Batteries: Current
Development, Interfacial Issues, and Instability Challenges. Carbon 2023, 213, 118225.
https://doi.org/10.1016/j.carbon.2023.118225.

(178) Ye, Y.; Xie, H.; Yang, Y.; Xie, Y.; Lu, Y.; Wang, J.; Kong, X.; Jin, S.; Ji, H. Solid-Solution or
Intermetallic Compounds: Phase Dependence of the Li-Alloying Reactions for Li-Metal
Batteries. J. Am. Chem. Soc. 2023, jacs.3c08711. https://doi.org/10.1021/jacs.3c08711.

179
 References

(179) Molaiyan, P.; Abdollahifar, M.; Boz, B.; Beutl, A.; Krammer, M.; Zhang, N.; Tron, A.;
Romio, M.; Ricci, M.; Adelung, R.; Kwade, A.; Lassi, U.; Paolella, A. Optimizing Current
Collector Interfaces for Efficient “Anode-Free” Lithium Metal Batteries. Advanced
Functional Materials n/a (n/a), 2311301. https://doi.org/10.1002/adfm.202311301.

(180) Luo, S.; Wang, Z.; Li, X.; Liu, X.; Wang, H.; Ma, W.; Zhang, L.; Zhu, L.; Zhang, X. Growth
of Lithium-Indium Dendrites in All-Solid-State Lithium-Based Batteries with Sulfide
Electrolytes. Nat Commun 2021, 12 (1), 6968. https://doi.org/10.1038/s41467-021-
27311-7.

(181) Bruce, P. G.; Saidi, M. Y. Variation of the Preexponential Factor and Activation Energy
for Lithium Diffusion in Cubic Titanium Disulfide. Journal of Solid State Chemistry 1990, 88
(2), 411–418. https://doi.org/10.1016/0022-4596(90)90236-Q.

(182) Sakuda, A.; Takeuchi, T.; Okamura, K.; Kobayashi, H.; Sakaebe, H.; Tatsumi, K.; Ogumi,
Z. Rock-Salt-Type Lithium Metal Sulphides as Novel Positive-Electrode Materials. Sci Rep
2014, 4 (1), 4883. https://doi.org/10.1038/srep04883.

(183) Flamary-Mespoulie, F. Synthèse et caractérisation de sulfures de métaux de transition

comme matériaux d’électrode positive à forte capacité pour microbatteries au lithium.
phdthesis, Université de Bordeaux, 2016. https://theses.hal.science/tel-01673794
(accessed 2023-10-04).

(184) Dugas, R.; Dupraz, Y.; Quemin, E.; Koç, T.; Tarascon, J.-M. Engineered Three-Electrode
Cells for Improving Solid State Batteries. J. Electrochem. Soc. 2021, 168 (9), 090508.
https://doi.org/10.1149/1945-7111/ac208d.

180
Hennequart Benjamin – PhD thesis/Thèse de doctorat – 2023

Engineering Strategies to Improve All-Solid-State Battery Performance under Low-

Pressure Conditions
Abstract: As the global shift towards renewable energy sources and electric vehicles gains momentum,
lithium-ion batteries (LIBs) are seen as a building block of a decarbonised future. To meet the growing
need for higher energy density and safety, all-solid state batteries (ASSBs) have emerged as a promising
alternative to traditional liquid-based LIBs. Nonetheless, the implementation of ASSBs faces challenges
in many aspects, notably the high operating pressure required for cycling, which prevents the use of
the high capacity lithium metal anode crucial for achieving the desired energy density. Thus, this
doctoral research is dedicated to addressing the challenge of operating pressure in ASSBs through two
key strategies. Initially, utilising a conventional composite electrode, we capitalised on the enhanced
chemical and electrochemical stability of halide-based solid electrolytes as well as their low hardness
to enable low pressure cycling while accommodating high potential cathode active materials. Secondly,
recognising that interfaces in composite electrodes represent a central issue in ASSBs, we utilised the
concept of the solid-electrolyte-free electrode. This concept involves the development of an electrode
that operates without the need for an additional ionic conductor. The outcome is an increase in energy
density and a reduction in the complexity of electrode interfaces. Altogether, both of these strategies
enabled cycling at pressures as low as atmospheric pressure and therefore enabled us to attempt the
implementation of the lithium metal anodes.

Keywords: battery; all-solid-state; interface; low pressure.

Ingénierie pour Améliorer les Performances des Batteries Tout Solide sous Faible
Pression
Résumé : Avec le développement croissant des énergies renouvelables et des véhicules électriques, les
batteries lithium-ion sont considérées comme un élément clé dans un avenir décarboné. Néanmoins,
pour répondre à ce besoin, des avancées majeures sont encore nécessaires en matière de densité
énergétique et de sécurité. Les batteries tout-solide sont apparus comme une alternative prometteuse
aux batteries traditionnelles contenant des liquides. Néanmoins, la mise en œuvre de cette
technologie rencontre des défis majeur, en particulier la pression élevée nécessaire pour le
fonctionnement qui empêche l'utilisation du lithium métal en tant qu’électrode négative qui est
pourtant essentielle pour atteindre les hautes densités énergétiques souhaitées. Ainsi, cette thèse se
concentre sur le défi associé à la pression de fonctionnement des batteries solides au travers de deux
stratégies. Tout d'abord, en utilisant une électrode composite conventionnelle, nous exploitons la
stabilité chimique et électrochimique accrue et la faible dureté des électrolytes solides à base
d’halogénures pour faciliter le fonctionnement à basse pression tout en permettant l’utilisation des
matériaux d’électrode à haut potentiel. Deuxièmement, comprenant que les interfaces dans les
électrodes composites représentent un problème central, nous mettons ensuite à profit le concept
d'électrode dépourvue d’électrolyte solide. Ce concept implique le développement d'une électrode
qui fonctionne sans nécessiter l’ajout d’un conducteur ionique. Il en résulte une augmentation de la
densité énergétique et une simplification des interfaces dans l'électrode. En somme, ces deux
stratégies permettent un fonctionnement des batteries tout-solide à des pressions aussi basses que la
pression atmosphérique, ouvrant ainsi la voie à la mise en œuvre de l'anode en lithium.

Mots clés : batterie ; tout-solide ; interface ; basse pression.