Development of Environmentally Friendly Lithium-Ion Battery

Components

Dissertation zur Erlangung des Doktorgrades Dr. rer. nat.

der Fakultät für Naturwissenschaften an der Universität Ulm

vorgelegt von:

Diogo Vieira Carvalho

aus Assis - 2018 -

Amtierender Dekan: Prof. Dr. Peter Dürre

Erster Gutachter: Prof. Dr. Stefano Passerini

Zweiter Gutachter: Prof. Dr. Arnulf Latz

Tag der Promotion: 7th June 2018

In memory of Mauro Nascimento Vieira

and Maria Emília Vieira

Abstract

Nowadays, it is fully ascertained that the massive use of fossil fuel has a detrimental

impact on the environment and the population health. These adverse aspects of using

fossil fuels, mostly in mobility, have been intensively discussed worldwide resulting in

new regulations to reduce the emissions of harmful exhaust gases. Concerning to this

point, the search for more efficient and low emission energy source has gained

considerable attention. The use of electricity as an energy source instead of combustion

of gasoline and diesel has gained a great interest in the automotive industry. For this

propose, lithium-ion batteries (LIB) have becoming the choice to store electricity and

power vehicles. LIB offer remarkable advantages such as high energy and power

capability with respect to lead acid and Nickel-Metal hydride (Ni-MH) batteries, and

long cycle life. However, LIB components, as electrodes, electrolytes and separator, are

mostly manufactured using expensive and toxic materials. In this context, this thesis

focuses on study new materials and processes for the realization of more

environmentally friendly LIB. In particular, this dissertation deals with the manufacture

of electrodes and separator with reduced environmental impact and improved

performance.

The LIB components studied in this thesis are made using water soluble polymers from

renewable sources. In Chapter 3, a novel porous membrane based on silicon dioxide

(SiO2) filler and hydroxypropyl guar gum (HPG) binder is presented and tested as a

separator for LIB. The HPG binder shows to be thermally stable at 200 °C and 250 °C

in O2 and N2 atmosphere respectively. Therefore, the membrane exhibits high thermal

stability and no dimensional-shrinkage at high temperatures thanks to the integration of

the ceramic filler and the thermally stable bio-derived polymer. Moreover, the separator

offers superior wettability and high electrolyte uptake due to the optimised porosity of

ca. 52% and the favourable affinity of SiO2 and guar gum microstructure towards

commercial organic liquid electrolytes. The electrochemical tests demonstrate the

satisfactory electrochemical stability of the separator in a wide range of potential, as well

as its outstanding cycle performance when coupled with lithium nickel manganese

cobalt oxide (NMC) and lithium titanate (LTO) electrodes.

Moreover, Chapter 4 described the study of aqueous processed LTO electrodes. The use

of phosphoric acid (PA) to control the pH of LTO slurries is investigated. The additional

of PA avoided the corrosion of the Al current collector and enhanced the adhesion and

the electrochemical performance of LTO electrodes. Additionally, LTO electrodes were

manufactured via the aqueous process using sodium carboxymethylcellulose (CMC),

guar gum (GG) or pectin as binders and these electrodes were investigated regarding of

morphology, adhesion strength and electrochemical performance. Finally, the full

lithium-ion cell made utilising only aqueous processing electrodes, namely NMC

cathodes and LTO anodes offers a stable discharge capacity of ~120 mAh·g−1 (based on

cathode active material) with high coulombic efficiencies. GG binder is also used in

graphite and NMC electrodes as the main binder. Full cell using these electrodes

delivered a stable discharge capacity of ~ 110 mAh g-1 (based on cathode active material)

with high coulombic efficiencies. The surface of pristine and cycled electrodes prepared

using GG or CMC were investigated using X-ray Photoelectron Spectroscopy (XPS).

Post-mortem investigation of cycled graphite electrodes from full lithium-ion cells

revealed the formation of a thinner solid electrolyte interface (SEI) when GG is

employed instead of CMC.

This thesis shows that polymers from bio-sources can be used as binders in the

preparation of LIB components. The use of these bio-polymers instead of the state-of-

art fluorinated binder accounts to lower environmental impact in LIB production.

The study here developed gives an important contribution to the improvement of LIB

towards environmentally friendly methods and materials, which are crucial to the

continuous expansion of renewable energy sources is our society.

Table of Contents

1. Introduction ......................................................................................... 1

1.4.1 Anodes ................................................................................................................. 8

1.4.2 Cathodes ............................................................................................................ 11

1.4.3 Electrolytes ........................................................................................................ 14

1.4.4 Separators .......................................................................................................... 17

1.4.5 Binders .............................................................................................................. 20

2. Theoretical background ................................................................... 27

2.2.1 Electrochemical methods .................................................................................. 30

2.2.2 Scanning electron microscopy (SEM) ............................................................... 33

2.2.3 Thermal gravimetric analysis (TGA) ................................................................ 35

2.2.4 pH-measurement ............................................................................................... 35

2.2.5 X-ray photoelectron spectroscopy (XPS) .......................................................... 36

3. Materials and methods ..................................................................... 39

3.2.1 Membrane preparation ...................................................................................... 40

3.2.2 Membrane characterization ............................................................................... 41

3.2.3 Electrode preparation ........................................................................................ 42

3.2.4 Electrode characterization ................................................................................. 43

3.3.1 Cell configuration .............................................................................................. 44

3.3.2 Cell assembly and electrochemical test ............................................................. 45

4. Aqueous Processing Separator for Lithium-Ion Batteries. ........... 49

Membrane characterization ............................................................................... 52

Electrochemical tests ......................................................................................... 59

5. Aqueous Processing Lithium Titanate Electrodes. ........................ 67

5.3.1 Thermal stability ............................................................................................... 70

5.3.2 Electrode surface characterization..................................................................... 71

5.3.3 Adhesion strength .............................................................................................. 74

5.3.4 Electrochemical characterization ...................................................................... 76

6. Evaluation of Guar Gum as Binder in Lithium-Ion Electrodes: Study of Binder Coverage on Active Material Surface. ...................... 81

6.3.1 Graphite half-cells ............................................................................................. 84

6.3.2 XPS analysis of pristine electrodes ................................................................... 87

6.3.3 Lithium-ion cell ................................................................................................. 89

6.3.4 Lithium-ion cells and post-mortem surface investigation ................................. 91

6.3.5 XPS analysis of cycled cathodes ....................................................................... 92

6.3.6 XPS analysis of cycled anodes .......................................................................... 94

7. Conclusion and final considerations .............................................. 101

8. Appendix .......................................................................................... 103

9. Acknowledgement ........................................................................... 119

10. Curriculum Vitae ............................................................................ 121

11. Scientific contribution .................................................................... 123

1

Work motivation

Since the industrial revolution, the lifestyle of our society and the impact in the

environment has drastically changed as it can be seen in the evolution of CO2 emission

in the last 250 years in Figure 1 [1,2]. The industrial expansion had forced the

development of more efficient machines and manufacturing process. Initially, such a

target was reached by the introduction of steam engines and later by the use of fossil

fuel and the electricity. The search for more profitable manufacturing method did not

consider the consequences in the surrounding environment. Evermore, the global

energetic dependency of coal and petrol sources generated political, diplomatic and

economic discussions.

Only in recent years, pollution starts to be a relevant topic for the energy sector. Many

nations have been demonstrating interest to lower down their needs of fossil oils, and

therefore these countries start to invest in alternative energy sources [3,4].

1760 1800 1840 1880 1920 1960 20000

5000

10000

15000

20000

25000

30000

35000

40000

Est

imat

e C

O2 e

mis

sion

/ 10

6 tons

Year

Figure 1 – Estimation of global CO2 emissions per year. Prepared using data from

reference [2].

2

Nuclear power plants are a well-established energy source that offers low or even no

CO2 emissions. However, investments in nuclear energy have been suffering from

considerable popular pressure due to the fear of accidents as occurred in Fukushima in

2011, Japan [5].

Solar panels and wind turbine are alternative energy converters systems with low

environmental impact. Nevertheless, they share as a drawback the intermittent energy

production due to the variation of daylight and wind flow for solar panels and wind

turbines, respectively. These unpredictable sources demand an energy storage system,

for instance, water dams or/and electrochemical devices [6]. Water storage dams are an

inexpensive energy storage system but limited by geographic aspects [7].

Electrochemical devices can store electricity and deliver it through a chemical process

once it is required. Redox flow batteries, fuel cells, electrochemical capacitors

(supercapacitors) and batteries are examples of electrochemical storage devices [8].

Nowadays, pushed by political pressure to reduce the in-house gas emission, the research

in electrochemical storage devices, especially batteries, were boosted. Although new

materials and different configurations had enhanced the performance and decreased the

price, battery technologies are still facing some weakness. Safety is one of the major

concerns for large batteries, for instance, when applied in hybrid electric vehicles (HEV)

and electric vehicles (EV). Other issues are the high price and toxicity of many cell

components [9]. Studies of battery components and process had shown the impact of the

materials choice in the final price of a battery [10]. Thus, this thesis focuses on the use

of environmentally friendly materials, mostly binders, for the production of battery

components, such as separator membranes, anode and cathode. The common

characteristics of the binders herein investigated are the processability in water and been

extracted from plants that accounts to low price. Moreover, the studied binders are

worldwide commercialised and employed for distinct applications as in food and

cosmetic industry. Therefore, this thesis is aimed to contribute in the battery technology

is given by new materials and process aiming at the low cost and low environmental

impact due to battery production.

Chapter 4 unveils the preparation of a porous membrane based on SiO2 particles and

HPG binder by a simple aqueous process. The prepared membrane displayed excellent

3

wettability with commercial LIB electrolyte, high thermal stability, and electrochemical

stability in lithium cells.

Chapter 5 describes a strategy to make lithium titanate Li4Ti5O12 (LTO) electrode by an

aqueous process. The simple introduction of a mild acid in the slurry enhanced the

electrode adhesion and the discharge capacity. Moreover, natural GG and pectin were

also employed as binder for LTO electrodes showing superior performance than CMC.

Finally, Chapter 6 shows a study of GG polymer as binder in LiNi1/3Mn1/3Co1/3O2

(NMC-111) and graphite electrode formulation. In this chapter, the influence of the

binder coverage on the active material surface will be presented. It is unveiled the

correlation between binder structure and electrochemical performance as well as the

composition of species in the solid electrolyte interface (SEI) layer.

Introduction to commercial batteries

The battery is an energy storage device formed by one or more electrochemical cells that

can store and delivered electricity by an internal chemical process, more specifically,

redox reactions. The amount of energy stored and delivered by the cell is determined by

the active material mass in the electrodes [4].

Batteries can be divided into two major groups, primary and secondary. Primary

batteries are non-rechargeable systems, due to non-reversibility of the redox reaction.

They offer high capacities and are produced in charged state (ready to use). They are

commonly used in many small electric devices, for instance, toys, remote controls and

calculators. Among several battery chemistries, Zn-MnO2 is one of the most used [11].

Secondary batteries differ from the primary by the ability to reverse the redox reactions

and therefore be able to have several discharge/charge cycles. Rechargeable batteries are

mostly produced in the discharged state thus an initial charge process is needed prior the

use. Secondary batteries are rather expensive compared to primaries ones. Nevertheless,

the higher buy value of rechargeable batteries is compensated for the long cycle life.

Initially, lead-acid, Ni-Cd and Ni-MH types were the most common rechargeable

batteries [8,12].

In 1991, Sony company introduced a new battery chemistry, the lithium-ion battery [13].

Coupling technologies reported on 70`s and 80`s, this new secondary battery opened the

4

field of this electrochemical device to some diversified applications. Nowadays, LIB

dominate the market for powering electronic devices, power tools, and HEV and EV

[14,15]. Compared to other secondary batteries, LIB offers superior power and energy

density due to the low weight of the lithium, and a long cycle life [1,5,6]. Figure 2 shows

a comparison of energy delivered by the most commercialised rechargeable battery

chemistries.

The energy stored in LIB has been enlarged by the introduction of new active materials,

upgraded battery formats and processes [14]. Nonetheless, alternatives to replace Li with

cheaper and more abundant elements, for instance Na [17] and Mg [18], are under

investigations.

Figure 2 – Energy comparison of different rechargeable batteries chemistries [12].

5

Lithium-ion battery operation

The working principle of a lithium-ion battery consists of reversible redox reactions that

take places on the electrode surface facing the electrolyte. It is defined that the anode

(negative electrode) is where the oxidation happens while in the cathode (positive

electrode) the reduction, both during the discharge process [4]. The commonly active

materials used for LIB are graphitised carbon (C) in anodes and lithium cobalt oxide

(LiCoO2) in cathodes. The reactions of a LIB based on those materials are below:

Anode reaction: Ca + bLi+ + be- ↔ LibCa (i)

Cathode reaction: LiCoO2 ↔ Li(1 - b)CoO2 + bLi+ + be- (ii)

Total reaction: Ca + LiCoO2 ↔ LibCa + Li(1 - b)CoO2 (iii)

In a LIB, during the discharge, the Li ions released from the anode, de-intercalate in the

case of graphite, are solvated by the electrolyte molecules, and transported to the cathode

driven by the difference of potential. Since the electrolyte is an electronic insulator, the

electrons released from the anode need to follow through an external circuit to reach the

cathode, thus balancing the reactions. On the cathode surface, Li-ions are de-solvated

from the electrolyte molecules and inserted into the material structure. The oxidation

reaction on the cathode is complete by the electron from the external circuit. The electric

current in the external circuit, which is in the opposite direction of the electron flow,

defines the energy or electricity delivered by the cell in the discharge process that is

utilised to power electric devices. The process can be reversed by applying a negative

current to force the Li ions to flow toward the cathode to recharge the battery [11]. The

minimum voltage upon discharge, low cut-off, and the maximum voltage upon charge,

upper cut-off, are carefully defined and controlled following different aspects, as the

working potential of the active materials. The abuse of cell working voltage negatively

affects the cycling life, due to the degradation of active materials and electrolyte, which

can also account for safety hazards [19]. Figure 3 shows a schematic representation of

lithium-ion battery operation.

6

Figure 3 - Schematic representation of the working principle of lithium-ion battery

(from Ref. [20]).

Lithium-ion battery components

A lithium-ion cell comprises two electrodes (anode and cathode), electrolyte, and the

separator as shown in Figure 4. Electrodes are composed of the active material,

conductive carbon and a binding agent. Predominantly, a slurry with the latter materials

and a solvent is cast on a metallic substrate, the so-called current collector (usually

aluminium for cathodes and copper for anodes), by slot-die or transfer roll-process.

Copper is used on the anode side, mostly with graphite as active material, due to the Al

forms alloy with lithium at low potential at ca. 0.2 V versus Li/Li+ [21].

7

Figure 4 - Illustration of lithium-ion battery components.

Usually, in industrial scale, the electrode layer is applied on both sides of the current

collector, which maximise the cell gravimetric energy by reducing the number of

current collector foils. Directly after the cast, the electrodes are dried to remove the

solvent. Afterwards, the electrodes are compressed to reach a determined electrode

density and porosity. This process improves the adhesion between the electrode layer

and the current collector and also homogenise the surface. The electrode´s porosity and

density differ on the cell application and the active material used. The battery is

produced by putting together cathode, anode tapes and separator in the jelly roll as in

cylindrical cell configuration our stacking the tapes as in coin cell and pouch cell as

shown is Figure 5. A final drying step is carried out prior the introduction of the liquid

electrolyte. Usually, the cell is under vacuum to maximise the wetting of electrode and

separator by the electrolyte. After the wetting step, the cell is sealed and afterwards the

formation step is carried out followed by the degassing and final sealing. The formation

step is the first charge of the battery. It may also involve distinct other tests to evaluate

the cell characteristics prior commercialisation [22].

The following sections will introduce relevant materials and components used in most

commercial LIB.

8

Figure 5 – Representation of lithium-ion batteries configuration; electrodes jelly roll in

a) cylindrical and c) prismatic cell and stacked electrodes in b) coin and d) pouch

cell [8].

1.4.1 Anodes

The first lithium rechargeable cell employed Li metal as the anode. This battery type

was used to power mobile phones. Nonetheless, this technology faced many problems,

as poor cycle life and safety, hindering the commercial expansion and further

applications of rechargeable Li metal batteries [23]. The safety issues were promoted

by a short circuit inside of the battery, which leads to a local heat generation, and in

more severe cases, fire or explosion [24]. The short circuit was promoted by the non-

uniform distribution of Li on the anode surface during the charging process. Those

formations, also known as Li dendrites, could grow and in more severe cases cross the

separator allowing the direct contact between electrodes [13,25].

Carbonaceous materials were reported to de/intercalate of Li+ ions [25]. Further

evaluations showed that layered and highly ordered graphite is more suitable to be

applied as anode material in lithium cells than amorphous carbon [26]. One benefit of

using graphite rather than hard or soft carbon is the flat voltage profile that leads to a

9

more stable cell voltage. Although graphite-based cells operate with higher cut-off

voltage, 3.0 V, compared to hard carbon, 2.5 V, the cell with graphite anodes can

deliver higher capacities [13]. The low surface area of graphite used as anode active

material, >10 m² g-1, is beneficial to reduce initial irreversible capacity loss.

Nonetheless, the first commercialised cells had hard and soft carbon as anode active

material and employed propylene carbonate (PC)-based electrolytes that were not

compatible with graphite-based anodes [14]. It was found during the lithiation of

graphite, PC molecule cointercalates into the graphene layers, negatively affecting the

active material structure and hindering the battery operation. This issue was solved by

the introduction of ethylene carbonate (EC) in the electrolyte formulation, which

decomposes during the lithiation process and forms a protective layer, called as solid

electrolyte interface (SEI), on the graphite surface. The SEI impedes further electrolyte

decomposition and grants the battery operation. This process will be discussed in the

section about lithium-ion battery electrolytes.

Since the beginning of 90´s, graphite (LixC6) is still the state-of-the-art anode for

lithium-ion batteries. The benefit of graphite is the high theoretical gravimetric capacity

of 372 mAh g-1, and the flat voltage profile at low potential versus Li/Li+, ca. 0.5 V

[16]. All these features lead to high energy lithium-ion batteries. Graphite-based anodes

have been intensively studied, and many improvements were reached [26]. Nowadays,

the first cycle irreversible capacity is usually lower than 10%, and practical capacities

of graphite half-cell exceed 350 mAh g-1 at high current rates (1C), and high efficiency

was achieved. Nonetheless, one drawback of using graphite as anode is the possibility

of having Li plating upon discharge, since the lithiation process is done at low potential

>0.5 V vs. Li/Li+, and low cut-off voltage may reach ca. 0.01 V. Li planting is also

likely to occur during fast lithiation process, at high current rates, as a result of slow

insertion process of Li+ into graphite [27–30]. The volume change of graphite upon

de/insertion of lithium-ions is circa of 10%, which may induce mechanical losses in the

electrode and further electrolyte decomposition to form a new SEI [31].

Lithium titanate (Li4Ti5O12, LTO) is also commercialised as active anode material, but

with a smaller presence in the market compared to graphite. LTO can de/insert Li+

resulting in a theoretical gravimetric capacity of 175 mAh g-1. The operation potential

of ca. 1.55 V vs Li/Li+ avoids electrolyte decomposition and SEI formation [16,26].

10

Although LTO cells offer lower energy than graphite-based cells, the safety is enhanced

since the LTO high operation voltage hinder the process of lithium plating at the

electrode surface, even at high current rates. Thus short circuit due to Li dendrites is

prevented. Further material development as carbon coating and the use of nano-sized

particles enhanced LTO´s performance in terms of cyclability and discharge capacity

at high current rates [14,32]. Therefore, safe and high power batteries can be produced

using LTO materials. The reaction of LTO upon Li+ insertion and de-insertion are

shown below:

Li4Ti5O12 + 3Li+ + 3e- ↔ Li7Ti5O12 (iv)

Another class of anodes is based on conversion and alloying materials. Those materials

can offer higher gravimetric and volumetric capacities with excellent rate capability

than carbon-based materials [33,34]. Nevertheless, one issue of conversion and alloying

materials are the extremely high volume expansion and contraction for lithiated and

delithiated state respectively. For instance, Si particles expand >270 % (during the

formation of Li15S4) than initial state [35]. The continuous expansion-contraction

causes particle fractures and further mechanical losses in the electrode layer. Due to the

break of the particle, a new SEI is formed consuming more electrolyte and Li,

generating low coulombic efficiencies and poor cyclability [14]. Many attempts have

been reported to overcome those issues, as carbon coating [36,37], use of

nanostructured materials [9,24] and proper binder choice [39–42]. Nowadays, LIB

industry is gradually employing a low amount of Si in anode formulation to increase

the final battery energy. The typical characteristic of graphite, LTO and silicon anodes

are shown in Table 1.

11

Table 1 – Basic characteristics of graphite, LTO and Si anodes [23,26,34].

Compound Common

name Structure

Practical specific

capacity / mAh g-1

Average

Potential vs

Li/Li+ / V

LixC6 Graphite Layered 330-365 0.10

Li4Ti5O12 LTO Spinel 170 1.55

Li15Si4 Si * 3579 0.40

* alloying material

1.4.2 Cathodes

Cathode materials for LIB are mostly based on Li+ intercalation and insertion materials

[43]. Since the late 70´s, different studies presented materials in which Li+ can be

inserted and removed from the host network [14,44,45]. The introduction of lithium

cobalt oxide LiCoO2 (LCO) was the breakthrough for the first commercialisation of LIB

[13]. This layered oxide is suitable for LIB application due to the high average voltage

of 3.9 V versus Li/Li+ (Co3+ to Co4+) and low self-discharge [43]. Although LCO has a

high theoretical specific capacity of 274 mA g-1 (full Li extraction), the practical

capacity is limited at ca. 140 mA g-1, which correspond to 0.5 Li extraction at around

4.2 V (x = 0.5 in Li1 - xCoO2), due to the poor stability of the material at low Li content.

Other limitations of LiCoO2 are the high cost and environmental risk, mainly due to Co

price and toxicity, respectively. Poor thermal stability is also a safety issue [14,16].

Nevertheless, LCO is still the most used cathode material in LIB, but with a constant

tendency to be replaced by other materials with a lower price, enhanced safety and

performance [25].

The layered lithium nickel oxide (LiNiO2) is an isostructural compound to LCO

proposed as cathode material for LIB [46]. The LiNiO2 advantages compared to LiCoO2

is the low price and higher specific capacity of 275 mAh g-1 at a cell voltage of ca. 4 V

(Ni3+ to Ni4+) [14]. However, the material was never commercialised due to difficult to

control the stoichiometry during synthesis [47] and safety issues, mostly due to its poor

12

thermal stability [16]. A more stable structure was obtained by the introduction of small

amount of Co (LiNi1-xCoxO2 for x = 0 - 1), which avoids the formation of Ni2+ in the Li+

side. Moreover, enhanced thermal stability and electrochemical properties were

achieved by the addition of Al (LiNi0.95 - xCoxAl0.05O2) that prevents the complete

formation Ni4+ in the fully charged state [42,44]. Consequently, the state-of-the-art of

cathode material is represented by Li[Ni0.8Co0.15Al0.05]O2 (NCA), with a high practical

capacity of ca. 200 mAh g-1 for a working potential of ca. 3.7 V [14].

The Li[Ni1/3Mn1/3Co1/3]O2 (NMC, also describe as NMC-111) is another layered-

transition metal oxide that nowadays is widely employed as a cathode material [25]. The

first 𝟐𝟐 𝟑𝟑� of lithium extracted from the NMC structure corresponds to the oxidation of

Ni2+ to Ni3+ and Ni3+ to Ni4+ while the oxidation of Co3+ to Co4+ gives the final 𝟏𝟏 𝟑𝟑� of

lithium. Although Mn (at Mn4+) do not participate in the electrochemical reactions, its

presence in the NMC is relevant to stabilise the structure. NMC-111 delivers a practical

capacity of 160 mAh g-1 when 4.3 V is reached as upper cut-off, excellent cycle stability

and enhanced safety respect to LCO and NCA due to less oxygen release in the charge

state [43,48]. Capacities above 200 mAh g-1 can be achieved once charge to 4.5 V,

nevertheless hindering the cycling life. The specific capacity NMC can be further

improved by increasing the Ni content (Ni-rich NMCs), as for instance the

LiNi5Mn3Co2O2 deliver a specific capacity of ca. 175 mAh g-1 [49,50]. Simultaneously,

the price and the environmental impact can be lowered, since the Co amount is reduced

[50]. Therefore, distinct NMC materials have been reported with different amounts of

transition metals, for instance, NMC-442, -532, -622 and -811 (in the following order of

Ni, Mn and Co) [43,45,51–53]. Cui and co-workers report a survey of Ni-rich NMCs

taking to account the Li diffusion and electrochemical performance [54].

The spinel-structure of LiMn2O4 (LMO) allows fast Li+ insertion/deinsertion. Therefore

it is a suitable material to be applied to high power batteries [55]. The LMO average

voltage is ca. 4.1 V (versus Li/Li+) with a practical capacity of 120 mAh g-1. Another

great motivation for the commercialisation of LMO is due to high abundance and low

toxicity of Mn [14]. However, Mn solubility into the carbonate-based electrolyte and

Jahn-Teller distortion due to a higher concentration of Mn3+ compared to Mn4+

negatively affect the long-term cycling life [43,56,57]. The progressive development of

13

LMO results in new materials, for instance, the high-voltage spinel LiNi0.5Mn1.5O4

(>4.7 V), which is an attractive Co-free cathode material [46,57,58].

In 1997, the phospho-olivine LiFePO4 (LFP) cathode material was introduced by Padhi

and co-workers [59]. Nowadays, this cathode material is widely studied and

commercialised. LFP offers excellent thermal and cycling stability and is not toxic

[48,60]. Although LFP is attractive for high power applications and delivers a practical

capacity of ca. 165 mAh g-1 (Fe2+ to Fe3+), the low operation voltage (3.4 V) and low

density compared to layered or spinel cathodes give a low energy density cell [25,43].

LFP possess a low electronic conductivity of ca. 10-9 S cm-1 (Table 2). This drawback

can be overcome by introducing a high electric conduct carbon coated layer on the

particles surface [61,62] and by reducing the particles size (nanostructured particles

>500 nm) and therefore facilitate Li-ion diffusion within the particle [63].

Table 2 summarises the characteristic of the most commercialised LIB cathodes

materials.

Table 2 – Properties of commercial cathode materials for lithium-ion

batteries [14,25,64].

Compound Common

name

Crystal

structure

Practical

specific

capacity /

mAh g-1

Average

Potential

vs Li/Li+ /

V

Electronic

conductivity /

S cm-1

LiCoO2 LCO

Layered

140 3.9 10-4

Li[Ni0.8Co0.15Al0.05]O2 NCA 200 3.8 10-4

Li[NixMnxCo1-x-y]O2 NMC 160a 3.8a 10-8a

LiMn2O4 LMO Spinel 120 4.1 10-6

LiFePO4 LFP Olivine 165 3.4 10-9

a - for NMC-111

14

1.4.3 Electrolytes

The primary function of electrolyte in an electrochemical device is to enable the

movement of ions. Therefore, electrolytes are designed to have high ionic conductivity

for a specific ion. Moreover, it is also required the insulator characteristic, i.e. negligible

electronic conductivity, to force the electrons to flow in the external circuit and avoid

any cell self-discharge. Mainly, a LIB electrolyte contains at least one Li-salt and a

mixture of organic solvents, usually carbonates. The solvents mixture is necessary to

reach all requirements for a LIB liquid electrolyte, such as high dielectric constant, low

viscosity, low melting point and film forming ability. Too viscous carbonates based

electrolytes hinder the cell performance of the cell due to low ionic conductivity.

Additionally, high dielectric constant is demand to enable the dissociation of Li salt to

high concentration, usually 1 molar. Moreover, the electrolyte should have a high flash

point to enhance the cell safety and be electrochemically stable within the electrodes

operation voltage (0.01 - 4.5 V) to not influence the cell performance. Nevertheless, the

electrolyte decomposition at low potential versus Li/Li+ (> 0.8 V) is key factor in LIB

that is discussed below [7,65,66].

Initially, propylene carbonate (PC) was the solvent used in most of the research works

dedicated to LIB [44]. The great interest of using PC as electrolyte solvent was its high

dielectric constant (ε = 64.92 at 25 °C) and wide temperature operation due to liquid

phase [66]. PC was also used in the first LIB commercialisation based on petroleum

coke as anode active material. Further reports showed that PC was not suitable for

graphite-based anodes due to the solvent co-intercalation with Li ions into graphene

layers. The decomposition of PC at ca. 0.8 V within the graphite particles generate gas

(mostly propylene), which forces the expansion of the graphene layers, which are

bonded by weak van de Waals forces. This effect damages the graphite anodes and

prohibits the use of PC as the main solvent for graphite-based anodes in LIB [16].

Nowadays, LIB electrolytes are based on the alkyl ethylene carbonate (EC) in

combination with one or more linear carbonates, for instance, dimethyl carbonate

(DMC) and diethyl carbonate (DEC) and lithium hexafluorophosphate (LiPF6) as salt.

EC is an indispensable component in electrolyte solution due to a crucial role in the SEI

formation on graphite electrode surface. At low potential, > 0.8 V, EC is subject to

15

reduction process and the products of this reaction, mostly alkyl carbonates, forms a

protective film (SEI) on the surface of the carbonaceous electrode.

The ability to form a stable SEI on the surface of graphite anode is a crucial parameter

of an electrolyte, preferably occurring only in the first cycle of the cell, also called as

formation process [67]. The SEI is a layer formed by organic and inorganic compounds

on the surface of both electrodes, however more critical in anode side, where the SEI

acts as a protective layer, blocking the direct contact of electrolyte and the carbonaceous

electrode. Therefore, further electrolyte decomposition does not occur. Its formation on

graphite electrodes is done by the solvent and salt decomposition at ca. 0.8 V. Figure 6

displays the reduction mechanism of commonly used solvents in LIB (EC and DEC).

The SEI should be an ionic conductor to allow the Li+ migration and be an electronic

insulator to avoid further electrolyte decomposition [68]. The nature and the role of SEI

in the cell performance have been intensively investigated [66,69]. The composition and

formation of the SEI species are directly related to electrode and electrolyte components

[70,71]. Temperature, cycling rate and cell storage conditions are also important features

that influence the SEI stability and thickness [72–74].

Figure 6 – Single-electron reduction mechanism of EC (top) and DEC (bottom) [66].

Although EC has a high dielectric constant (ε = 89.78 at 25 °C), it cannot be used as the

single solvent is a LIB electrolyte due to its high melting point of 36.4 °C. Therefore, a

linear carbonate, for instance, DEC with low dielectric constant (ε = 2.8 at 25 °C) but

with a low melting point of - 74.3°C is commonly used to have a liquid solvent at room

16

and low temperature [66,75]. A detailed work from Ding and co-workers reports the

liquid/solid phase diagram of many carbonates used as solvents for LIB electrolytes [76].

Lithium hexafluoroarsenate (LiAsF6), lithium tetrafluoroborate (LiBF4) and lithium

perchlorate (LiClO4) were also studied as salt for lithium battery electrolytes.

Nevertheless, LiPF6 is the most appropriate choice [66,77]. The use of LiPF6 grants

reasonable characteristics, for instance, 1 mol L-1 of LiPF6 at 25 °C in EC:DMC

(1:1 w/w) has an ionic conductivity of 10.7 mS cm-1 and is thermally stable till 80 °C

[67]. Moreover, it has been reported the influence of 1 mol of LiPF6 or LiClO4 in

EC:DEC (1:1 w/w) in the corrosion of current collector foils [78]. The electrolyte

containing LiPF6 have shown low or no corrosion than LiClO4 due to the layer formation

of highly electrical stable species as AlF3, especially for pure aluminium foil (99.99%),

which is used as the current collector for cathodes. The stability of current collector is

highly relevant in the cell, as the corrosion of current collector may increase the cell

impedance by forming insulator products leading to poor cyclability.

Commercialized electrolytes include some other components that only the Li salt and

solvents. Distinct additives, maximum 5 % in weight or volume, can be used to enhance

the battery performance and safety [79]. A well-established additive is vinyl carbonate

(VC). The use of VC grants stable SEI formation in graphite anodes and hence enhanced

performance [80]. The research on electrolyte additives is economically appealing due

to the possibility enhance cell performance by addition of small amount of additive in

the electrolyte formulation. For instance, Sharova and co-workers reported the use of

Li-imide salt as electrolyte additive for LFP/Graphite cells [81]. They show that using

2 wt.% lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) as an additive the SEI on

the graphite electrode is thinner and less resistive than when of using VC

and other Li amide salts (lithium bis(fluorosulfonyl)imide LiFSI and lithium

fluorosulfonyl(trifluoromethanesulfonyl)imide LiFTFSI). Moreover, the use of LiTFSI

enhanced the capacity retention of LFP/Graphite cells. Additives can also improve the

battery cycle life by forming a protective layer on the cathode surface in contact with

the electrolyte and consequently reduce the metal dissolution from the active material in

the electrolyte [82].

Other electrolytes in development for LIB and Lithium metal cells are based on ionic

liquids [63], gel [85] or solid, as ionic conductor polymer [84] and ceramic-based

17

electrolytes [86]. The latter received considerable attention in the last years to be utilised

principally in Li metal cells and Li-Sulphur [87].

1.4.4 Separators

The separator has the major function of avoiding physical contact between the

electrodes while allowing the movement of Li+ within the electrolyte. Therefore, the

requirements of a LIB separator are to be an electronic insulator and to have a porous

matrix structure where the Li+ can move. Moreover, it is vital that the separator is

electrochemical stable within the work potential of the cell and chemically stable

toward the electrode materials and aprotic organic electrolytes [88]. Additionally, a

good affinity between separator and electrolyte in terms of wettability is crucial to

reduce the battery production time and enhanced the electrochemical performance. The

separator density and thickness are factors that influence the final cell performance.

Separators with low density are favourable to cell gravimetric capacity, while the

separator thickness is a compromise of power and energy to safety and mechanical

properties [89]. Usually, the thickness of a commercial separator is ca. 16 – 25 µm with

a porosity about 30 – 50 % and porous smaller than 1 µm [90].

Since the first lithium battery reports, polyethylene (PE) and polypropylene (PP) porous

membranes have been employed as separator [25,91]. These materials revealed to have

excellent mechanical strength and to be chemically stable and have an acceptable cost.

The low melting point of PE, 135 °C, acts as a cell safety protection in case of

increasing cell temperature. Indeed, once the PE melting point is reached, the

membrane porous structure collapses and the cell impedance increases impeding the

ionic conduction. Nevertheless, the PE melting also induces the separator shrinkage,

which may promote a not desired direct contact between electrodes, causing an internal

short circuit [13]. By introduction on a PE membrane at least one layer of PP, with the

higher melting point of 165 °C, the shutdown effect is maintained while thermal

separator stability can be enhanced [88]. However, the drying temperature of PE or

PE/PP membranes is limited by the melting point of PE to around 70 – 90 °C. This

temperature is much lower than used for drying electrodes, ca. 150 °C, which affects

the cell production. In fact, electrodes and separators are dried separately, and the cell

18

manufacturing has to be carried out in a dry room as described in Table 3. Finally, the

cells are additionally dried prior the electrolyte introduction. The cell production in the

dry room is expensive, and the final cell drying is slow due to low temperature applied

(< 90 °C). High-temperature stable separators offer the possibility to assemble the cell

in a clean room and use higher drying temperatures [90,92]. In this way, the battery

production is faster and less expensive.

Table 3- Description of cell production condition using low and high melting point

separators.

Separator

Production steps

Components

drying

Cell

manufacturing

Cell

drying

Electrolyte

wetting

Low melting point

< 90 °C Clean room Dry room

High melting point

> 150 °C Clean room Dry room

Another drawback of PE/PP based separators is the poor affinity to Li-ion electrolytes,

which demands long cell wetting times in controlled atmosphere and, as a consequence,

an increase of the battery’s final cost [90].

Ceramic-based separators have been developed to overcome the PE/PP based

separators drawbacks. The introduction of nano and microstructured particles, for

instance, SiO2, Al2O3, CaCO3 or their mixtures as coating layer or filler, enhanced the

wettability of the PE/PP separator due to the hydrophilic characteristics of those

ceramic particles and also is expected to offer higher thermal stability [93]. These

ceramic particles posse high thermal conductivity and, thus, dissipate heat quickly

throughout the separator membrane [94]. Another significant advantage of using

inorganic filler materials is a noticeable improvement in cycling performance of Li-ion

cells due to their scavenging reaction with hydrofluoric acid (HF) traces, which stem

from the reaction of the lithium salt (LiPF6) with residual moisture [85,95]. Nonwoven

membranes coated with ceramic particles with no shrinkage even at elevated

temperatures, > 200 °C for 24 h, are also commercialised as LIB separator [96].

19

Moreover, polyimide nanofibrous membranes prepared by electrospinning were

explored as a separator in lithium metal cells at high temperatures operations, 120 °C

[97]. Kim and colleagues reported a porous membrane made using cellulose and SiO2

with zero shrinkage and enhanced power and cycling stability than a commercial try-

layer PP/PE/PP separator [98]. Obviously, this separator does not have the shutdown

effect as PE/PP membranes, however, for large size LIB, like those used in HEV and

EV, a high thermally stable separator is more suitable [89,90,94]. For an electro-

mobility application, LIB are connected in series and in parallel to increase the battery

pack voltage and capacity. The battery pack includes numerous temperature sensors in

the cell and electronic circuits to ensures safe operation even in the case of unusual

heating caused by high currents or overdischarge. By using a high-temperature stable

separator that can operate at higher temperatures than PE/PP, the battery pack thermal

management can be facilitated [96].

The properties of commercial separators are displayed in Table 4. Ceramic filled

nonwoven is a separator based on a mixture of polymer fibres and the ceramic particles

while the ceramic coated membranes are a porous membrane matrix, for instance, a PE

or a poly(vinylidene difluoride) (PVDF) membrane, coated with the ceramic particles

[96]. Nanofibers membranes can be based on different polymers (PVDF,

Polyacrylonitrile PAN), prepared by electrospinning or phase inversion method

[99,100].

The use of solid electrolytes, and in some cases gel electrolytes, do not demand a

separator since the electrolytes also act as the physical barrier between the electrodes

Currently, LIB power a broad type of electronic devices and this range of applications

can be extended in the near future. One of the reasons is the adjustable battery output

features by choosing the combination of electrode active materials, formulation and cell

format. Consequently, there is no unique separator capable of fulfilling all the newly

LIB requirements, but, separators can definitely be designed for specific applications.

20

Table 4 – Properties of commercial LIB separators [88–90,94].

Membrane

material

Thickness

/ µm

Porosity

/ %

Shutdown

temperature /

°C

Maximum

drying

temperature /

°C

Wetting

PE 16 - 25 40 135

70 - 90 - PP 25 40 165

PP/PE/PP 16 -25 42 135

Ceramic

filled

nonwoven

23-35 50 - 55 X 100 - 170

+ Ceramic

coated

membranes

22 - 28 50 - 55 X 80 - 110

Nanofibers 25 - 35 45 - 55 X 100 - 170

(-) poor (+) improved wetting

1.4.5 Binders

Binders are electrochemical inactive components in an electrode layer. Although they

do not participate in the redox reactions, binders are critical for battery performance

and production. In fact, the primary role of the binder is to bind the active material and

conductive carbon to each-other and to the current collector. The amount of binder in

commercial electrodes formulation is about 2 – 5 wt.%, depending on active material

particle size, carbon content, slurry processing and desired electrode capacity

[101,102]. Although, the low concentration in the electrode formulation, the binder is

a critical factor that impacts the electrode processing, properties and electrochemical

performance.

Since the introduction in 1991, poly(vinylidene difluoride) (PVDF) is the most used

binder in LIB electrodes [13]. Regardless of well-established electrode production

21

using PVDF binder, its use comes with many drawbacks. Indeed, PVDF is dissolved in

N-methyl pyrrolidone (NMP), which is volatile, toxic and an expensive solvent.

Therefore, the electrode drying has to be done in a controlled atmosphere, and the NMP

must be recovered to avoid air contaminations. This process is expensive due to the

energy required to control the drying step conditions and the NMP recovering system

[10,103].

Nowadays, sodium-carboxymethyl cellulose (CMC) and styrene-butadiene rubber

(SBR) are being used as binders for graphite-based anodes in commercialised LIB. This

combination offers long cycling life and appropriate slurry rheology and electrode

flexibility and adhesion [30,104].

In comparison to organic-based compounds (PVDF with NMP), an aqueous processing

electrode offers important advantages as reduced cost and environmental impact.

Indeed, D. Wood and co-workers reported a comparison of electrode production price

using organic and aqueous process. Taking in to account binder and solvent price and

the energy needed for the dry step and to recover the solvent (only need for NMP) they

showed that an aqueous process is 26 times less expensive than the organic based

(PVDF/NMP) [10]. Moreover, another concern for large-scale LIB production is the

battery end-of-life. The recycling process of an organic-based electrode has to be done

again in a controlled atmosphere, either with the thermal process or by dissolving the

electrodes in the appropriate organic solvent (NMP). Differently, the use of aqueous

electrode preparation facilitates the recycling process by not requiring a thermal

treatment to remove the active material from the current collector since the active

material can be recovered by only washing out the electrode in water.

Nevertheless, aqueous processing of cathodes is still not implemented. Firstly, Li+

leaching may occur in many of metal oxides used as active materials, hindering the

electrode efficiency and electrochemical performance [105]. Secondly, the alkaline

characteristic of aqueous slurries with metal oxides (pH >11) leads the corrosion of

aluminium current collector as shown in the reactions below, affecting the battery

performance, slurries rheology and electrode adhesion [106].

22

Al2O3 + 2 OH- + 3 H2O → 2 [Al(OH)4]- (v)

2 Al + 2 OH- + 6 H2O → 2 [Al(OH)4]- + 3 H2 (vi)

These issues were overcome by distinct approaches, such as the use of carbon coated

current collector [28,107], a mixture of CMC and urethane binder [29], and the

introduction of mild acids in cathode slurry to reduce the pH [105,108]. In particular,

the last approach was used to prepare LTO anodes, and enhanced capacity and stable

cycling were reached due to superior electrode morphology and adhesion [32].

Recently, Kuenzel et al. made LiNi0.5Mn1.5O4 (LNMO) cathodes by an aqueous process

using crosslinked CMC binder and phosphoric acid with improved cycling stability and

efficiency [58].

Aqueous slurries demand a more careful casting process as a matter of the higher

surface tension of the water (ca. 72.8 mN m-1 at 25 °C) compared to NMP

(ca. 41 mN m-1 at 25 °C) leading to poor wetting of the current collector. The low

wettability of aqueous slurries may lead to shrinkage of the wet electrode layer and

poor adhesion. Li et al. introduced a Corona plasma treatment to improve the wetting

of the aluminium current collector for LFP aqueous slurries. Enhanced adhesion and

discharge capacity were achieved by electrodes prepared using this technique [109].

The high surface tension of water may also induce the formation of cracks in the

electrode layer during the drying process due to high force of the water meniscus, as

represented in Figure 7. Du et al. report a strategy to enable thick, crack-free NMC

electrodes. They used a combination of water and isopropyl alcohol as a solvent with

low surface tension (32 mN m-1) and CMC binder to produce high loadings electrodes

(> 25 mg cm-2) with comparable performance to PVDF/NMP based ones [110].

Figure 7 – Representation of the crack generation in the drying process. Reprinted from

reference [111].

23

Ideally, the binder acts as dispersant and thickener, by increasing the slurry viscosity

and homogeneity. Therefore, it may demand the use of two distinct polymers to grant

suitable slurry properties and binding. Usually, the electrode slurry is a dispersion of

ca. 45 wt.% of solid particles (active material/carbon) in a binder/solvent solution.

Many groups have investigated the appropriate solid dispersion in aqueous slurries. Li

and co-workers reported the influence of the ratio of polyacrylic acid (PAA-NH4)

binder in LCO cathode slurries properties and the influence in the electrochemical

performance [112]. Porcher and colleagues described a detailed study of aqueous

slurries for LFP cathodes using sodium-carboxymethyl cellulose (CMC) as binder

[113,114]. It has been found that a gel-like slurry generates a more homogeneous

electrode than a liquid-like one. During the drying step, the evaporation of the solvent

on the top surface of the electrode induces the binder migration to the bottom of the

electrode (Figure 8). In a gel-like slurry, the solvent diffuses within the slurry and

therefore the dispersion in less modified. Müller and co-workers investigated the

distribution of PVDF binder in graphite anodes by changing the drying temperatures

[115]. They claim that in all drying temperatures tests there is binder migration to the

electrode bottom, however with a higher binder concentration at fast drying (higher

temperatures). Improved rate capability and cycle stability were found in LFP electrode

using CMC as binder instead to PVDF [116]. The influence of the degree of substitution

(DS) of CMC was also evaluated in graphite electrodes [117]. It was reported that

electrodes prepared using CMC with low DS (0.7) have higher capacity retention.

Urbonaite et al. combined water and ethanol to improve the dispersion of silicon

nanoparticles and the ratio of 3:7 (ethanol: water) was reported to allow enhanced

cycling stability and specific capacity [118].

24

Figure 8 – Schematic representation of drying process of gel and liquid-like slurry

[102,113].

The research activity on binders for LIB was rather low compared to electrolytes and

active materials. Nevertheless, the research on alternative binders, mostly water-

soluble, is in expansion by the economic advantages and low environmental impact of

aqueous processing electrodes compared to organic based (PVDF and NMP).

Furthermore, many reports claim superior results by choosing appropriate binder. For

example, El Ouatani and co-workers report that CMC binder can facilitate the

formation of a stable SEI in graphite electrodes by only the contact with electrolyte

[70], or, like in the case of GG enhance the cycling performance of Li-rich layered

cathodes [119]. In the latter case, the authors claim that guar gum forms a more

homogenous film on the surface of the active material compared to PVDF binder, and

this film protects the active material from side reactions. Lu and co-workers report

superior cycling stability when GG is used as binder compared to PVDF in Li-sulphur

cells [120]. The interaction between GG molecule and sulphur particles was mention

to enhance the electrode mechanical properties. The bonding between the active

material and the numerous hydroxyl groups in the GG molecule has also been reported

to an important factor to support the volume change of conversion and allowing

materials [40,121].

Cuesta et al. investigated the influence of several natural polysaccharides, such as CMC,

sodium alginate, GG, xanthan gum, gum arabic and others, as binder for graphite-based

anodes [30]. In summary, they show that graphite electrodes made with only 5 wt.%

CMC, sodium alginate or GG have comparable electrochemical performance than

electrode with 8 wt.% of PVDF. Natural polymers, like lotus bean gum, were also

applied to silicon-based electrodes to support the undesired, but occurring silicon

volume changes during the electrochemical reactions [122–124]. Pectin was used as

25

binder for silicon anodes, and the electrochemical performance was compared with

CMC and amylose-based electrodes [121]. The authors claim that the electrochemical

performance, in particular, cycle life, was enhanced due the polysaccharide backbone

structure.

Another component in the electrode layer is the conductive additive, its amount in

electrode formulations range is between 1 – 10 wt.% depending on the material

electronic conductivity and battery application. Carbon black is widely used as a

conductive additive, and in some cases in cathodes formulations, graphite and carbon

black are used [21].

Table 5 shows the structure and other information of the binders used in this thesis.

Table 5 – Molecular structure, source and industrial use of natural source binders used

in this thesis.

26

27

Electrochemical concepts

This section describes the thermodynamics concepts behind a redox reaction and

introduces relevant equations to characterise and compare electrochemical devices.

The maximum work done by an electric device 𝑤𝑤𝑚𝑚𝑚𝑚𝑚𝑚 is given by:

𝑤𝑤𝑚𝑚𝑚𝑚𝑚𝑚 = 𝐸𝐸 ∙ 𝑄𝑄 (1)

where 𝐸𝐸 is the maximum potential and 𝑄𝑄 is the total amount of energy produced

calculated as:

𝑄𝑄 = 𝑛𝑛 ∙ 𝐹𝐹 (2)

where 𝑛𝑛 is the number of electrons (in mole) transferred during the reaction and 𝐹𝐹 is

Faradays constant, 9.65∙104 C F-1. Thus, the maximum electric work can be expressed,

𝑤𝑤𝑚𝑚𝑚𝑚𝑚𝑚 = 𝐸𝐸 ∙ 𝑛𝑛 ∙ 𝐹𝐹 (3)

The variation of free energy of an electrochemical reaction, ∆𝐺𝐺, can be expressed by:

∆𝐺𝐺 = ∆𝐺𝐺0 + 𝑅𝑅 ∙ 𝑇𝑇 ∙ 𝑙𝑙𝑛𝑛 𝐴𝐴 (4)

if ∆𝐺𝐺0 is standard free energy, 𝑅𝑅 is the gas constant (8.3145 J mol-1 K-1), 𝑇𝑇 the absolute

temperature [K] and 𝐴𝐴 is the ratio of activity product of the product 𝑎𝑎𝑂𝑂𝑚𝑚𝑂𝑂 and activity

product of the reactant 𝑎𝑎𝑅𝑅𝑅𝑅𝑅𝑅. In an electrochemical cell it can be expressed by:

28

𝐴𝐴 = 𝑚𝑚𝑂𝑂𝑂𝑂𝑂𝑂𝑚𝑚𝑅𝑅𝑅𝑅𝑅𝑅

(5)

Thus,

∆𝐺𝐺 = ∆𝐺𝐺0 + 𝑅𝑅 ∙ 𝑇𝑇 ∙ ln 𝑚𝑚𝑂𝑂𝑂𝑂𝑂𝑂𝑚𝑚𝑅𝑅𝑅𝑅𝑅𝑅

(6)

If a redox reactions is spontaneous, thus ∆𝐺𝐺 < 0. Therefore, the available energy of an

electrochemical reaction can be expressed by:

−∆𝐺𝐺 = 𝑤𝑤𝑚𝑚𝑚𝑚𝑚𝑚 = 𝐸𝐸 ∙ 𝑛𝑛 ∙ 𝐹𝐹 (7)

∆𝐺𝐺 = − 𝐸𝐸 ∙ 𝑛𝑛 ∙ 𝐹𝐹 (8)

and

∆𝐺𝐺0 = − 𝐸𝐸0 ∙ 𝑛𝑛 ∙ 𝐹𝐹 (9)

when 𝐸𝐸0 is the cell standard potential.

Substituting eq (8) and (9) in eq (6) the van´t Hoff isotherm equation for redox reactions,

it results:

𝐸𝐸 = 𝐸𝐸0 − 𝑅𝑅∙𝑇𝑇𝑛𝑛∙𝐹𝐹

∙ ln 𝑚𝑚𝑂𝑂𝑂𝑂𝑂𝑂𝑚𝑚𝑅𝑅𝑅𝑅𝑅𝑅

(10)

The latter equation is the Nernst equation, which correlates the active mass of the

product and reactant materials involved in the reaction and the final half-cell potential

[V].

The total potential of the cell, 𝐸𝐸𝑐𝑐𝑅𝑅𝑐𝑐𝑐𝑐, is given by:

29

𝐸𝐸𝑐𝑐𝑅𝑅𝑐𝑐𝑐𝑐 = 𝐸𝐸𝑐𝑐𝑚𝑚𝑐𝑐ℎ𝑜𝑜𝑅𝑅𝑅𝑅 − 𝐸𝐸𝑚𝑚𝑛𝑛𝑜𝑜𝑅𝑅𝑅𝑅 (11)

where 𝐸𝐸𝑐𝑐𝑚𝑚𝑐𝑐ℎ𝑜𝑜𝑅𝑅𝑅𝑅 and 𝐸𝐸𝑚𝑚𝑛𝑛𝑜𝑜𝑅𝑅𝑅𝑅 are defined by the potential of the cathode and anode,

respectively.

By following the Faradays law, the mass 𝑚𝑚 [g] of a product that provide electrons is

proportional to quantity of electricity produce as follow,

𝑚𝑚 = 𝐼𝐼∙𝑐𝑐∙𝑀𝑀𝑛𝑛∙𝐹𝐹

(12)

where 𝐼𝐼 is electronic current, 𝑡𝑡 is the time of the current flow and 𝑀𝑀 is the molecular or

atomic weight of the reactant. The theoretical capacity of the active material employed

in the electrodes, 𝐶𝐶 given by as [Ah g-1] or [mAh g-1], can be calculated as,

𝐶𝐶 = 𝑛𝑛∙𝐹𝐹∙𝑚𝑚𝑀𝑀

(13)

Alternatively, the theoretical gravimetric specific capacity 𝐶𝐶𝑠𝑠𝑠𝑠𝑅𝑅𝑐𝑐, in [Ah] or [mAh], is

independent of the mass of the material, and is given by:

𝐶𝐶𝑠𝑠𝑠𝑠𝑅𝑅𝑐𝑐 = 𝑛𝑛∙𝐹𝐹𝑀𝑀

(14)

The electric energy given by an electrochemical device, ℰ is given by:

ℰ = ∫ 𝐸𝐸(𝑐𝑐) ∙ 𝑑𝑑𝑐𝑐 = 𝐸𝐸(𝑐𝑐) ∙ 𝐶𝐶𝐶𝐶0 (15)

as 𝐸𝐸(𝑐𝑐) is the cell potential for a specific state of charge c. The proper comparison of

distinct electrochemical devices can be done using the specific energy ℰ𝑚𝑚𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑅𝑅𝑠𝑠

[Wh kg-1] or volumetric energy ℰ𝑉𝑉𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑅𝑅𝑠𝑠

[Wh L-1] where 𝑚𝑚𝑠𝑠𝑠𝑠𝑠𝑠𝑐𝑐𝑅𝑅𝑚𝑚 and 𝑉𝑉𝑠𝑠𝑠𝑠𝑠𝑠𝑐𝑐𝑅𝑅𝑚𝑚 is the mass

30

and volume of the system, respectively. Analogue comparison can be done using the

specific power [W kg-1], where

𝑃𝑃 = 𝐼𝐼∙𝐸𝐸𝑚𝑚𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑅𝑅𝑠𝑠

(16)

In a real system, the voltage of the cell 𝑉𝑉 [V] during the charge or discharge, is affected

by polarization 𝑛𝑛, which is directly proportional to the internal resistance 𝑅𝑅𝑂𝑂𝑛𝑛𝑐𝑐 at a

constant current 𝐼𝐼. Thus, the output voltage 𝑉𝑉𝑅𝑅𝑂𝑂𝑠𝑠𝑐𝑐ℎ𝑚𝑚𝑎𝑎𝑎𝑎𝑅𝑅 𝑎𝑎𝑛𝑛𝑑𝑑 𝑉𝑉𝑐𝑐ℎ𝑚𝑚𝑎𝑎𝑎𝑎𝑅𝑅 can be written as:

𝑉𝑉𝑅𝑅𝑂𝑂𝑠𝑠𝑐𝑐ℎ𝑚𝑚𝑎𝑎𝑎𝑎𝑅𝑅 = 𝑉𝑉𝑂𝑂𝐶𝐶𝑉𝑉 − 𝑛𝑛𝐼𝐼𝑅𝑅𝑂𝑂𝑠𝑠𝑐𝑐ℎ𝑚𝑚𝑎𝑎𝑎𝑎𝑅𝑅 (17)

And

𝑉𝑉𝑐𝑐ℎ𝑚𝑚𝑎𝑎𝑎𝑎𝑅𝑅 = 𝑉𝑉𝑂𝑂𝐶𝐶𝑉𝑉 + 𝑛𝑛𝐼𝐼𝑐𝑐ℎ𝑚𝑚𝑎𝑎𝑎𝑎𝑅𝑅 (18)

Analytics methods

2.2.1 Electrochemical methods

2.2.1.1 Galvanostatic cycling

Galvanostatic charge and discharge test is a chronopotentiometry technique widely used

to characterise LIB. The method is based on applying a constant current (CC) for a

determined interval of time (Figure 9 left panel). The galvanostatic methods can be

carried out in a 2 electrode cell setup (working (WE) and counter electrode (CE)), and

the cell voltage is recorded. In a 3 electrode cell configuration (WE, CE and reference

electrode (RE)), aside from the cell voltage, the potential of WE and CE versus the RE

can be recorded giving information about the performance of each electrode. In a

31

galvanostatic cycling, the battery is under a positive current during the discharge while

to charge the battery an external energy source is needed to provide the negative

(opposite) current. The current necessary to charge/discharge a battery in one hour can

be defined in terms of nominal capacity C (calculated using equation 14), i.e., 1C.

Therefore, 0.1C correspond the current to charge/discharge the battery in 10 hours while

2C is the current required to charge/discharge the battery in 0.5 hours (30 min). In many

cases, the charge current may be lower than discharge current rate, due to the fact that

graphite lithiation process is rather slower than the delithiation. Therefore, at a high

charge current Li ions accumulate on the surface of the anode, and may form dendrites,

which consume the Li available in the cell. Moreover, a constant voltage (CV) step can

be used in the charging process to continuous insert the remain Li+ on the graphite anode.

The CV step is typically used when the battery reaches the maximum operation voltage

(charge cut-off), where the graphite anode may be at low potential ca. 0.01 V.

Figure 9 (right panel) shows a typical discharge voltage profile of a cell under a CC test.

By the analyses of voltage profiles many aspects of the battery can be identified, for

instance, charge and discharge capacity, average voltage and energy. Moreover, a C-rate

test, which consist of varying the charge/discharge rate upon cycling, is a useful to

identify the power capability of materials and of full-cell. Taking into account the

equation 17 and 18, the influence of the current on the voltage slope can be determined.

Low currents will slight change the voltage profile of the cell while high currents may

significantly affect the voltage curve on charge and discharge.

Figure 9 – Schematic representation of a galvanostatic charge/discharge test (Left) and

typical discharge voltage profile (Right).

32

2.2.1.2 Linear sweep voltammetry (LSV)

The LSV is electrochemical technique based on applying a linear difference of potential

between two electrodes (WE and CE) and records the current during a determined

interval of time (Figure 10 left). The scan rate (V s-1) needs to be low enough

(> 1 mV s-1) to avoid or at least reduce the perturbation from the ohmic resistance of the

electrolyte and double layer capacitance on interface between electrolyte and electrode.

For this technique a 3 electrode setup is demand, due to difficult to control the potential

of the WE versus the CE, since the potential of CE may change due to redox reactions.

Therefore, by using a RE electrode (in a 3 electrode cell), the potential of WE can be

precisely controlled while the current can be recorded between WE and CE. Once the

voltage at which the redox reaction starts is reached only a small current is detected. The

further potential change generates a concentration gradient and consumption of active

species in the redox reaction, which decreases the number of electrons transferred.

By analyses of the LSV plot, I versus V (Figure 10 right panel), currents peaks at a

specific voltage can be identified. The plot may give current responses at different

potentials as a result of distinct redox reactions taking place on electrode surface, for

instance binder and electrolyte decomposition or active material oxidation or reductions.

Therefore, this technique is widely used to identify the electrochemical stability window

(ESW) of LIB components, such as electrolytes and binders.

LSV is thus performed from the cell equilibrium potential, open circuit voltage (OCV),

up to a maximum potential and/or to a minimum one which are close (usually slightly

above) to the operational voltage of the electrode active materials which will compose

the real electrode cell. The occurrence of an anodic or a cathodic peak means that the

specimens inside of the cell undergoes an oxidation or a reduction respectively.

33

Figure 10 – Schematic representation of a linear sweep voltammetry test (Left) and a

typical result plot (Right).

2.2.2 Scanning electron microscopy (SEM)

Scanning electron microscopy (SEM) permits to investigate the morphology of materials

and electrodes used in LIB. The work principle is based on inspecting the sample surface

using a small diameter electron beam (> 4 nm) and record linear scans to form an image.

Electrons are generated in a heated tungsten filament and accelerated by high voltages

(1 – 50 kV), and the electron beam is adjusted and focused by magnetic lens. Once an

electron hit the sample, it interacts with atoms in the sample by distinct ways as shown

in Figure 11 (left panel). The image and physicochemical information of the sample is

study by using the scattered electrons and emitted electrochemical radiation. The image

is formed using the radiation of secondary and backscattered electrons. Secondary

electrons with low energy (< 50 eV) are generated in the electron-atom interaction.

Therefore, the mean free path of these electrons is very short, (2 – 20 nm), giving

information about topography of the sample. The backscattered electrons may have

> 50 eV till the energy of the electron beam. These electrons are primary electrons from

the electron beam than undergo a scattering by the nucleus of the atom in the specimen.

The energy correlation between the electron beam and backscattered electrons are used

to form the image contrast, which is sensible to atomic number of the atoms in the

specimen.

Moreover, Energy Dispersive X-ray Spectroscopy (EDX) can also be performed in a

SEM device. In fact, more energetic electron beam can also remove electrons from the

inner shells, consequently exciting the atom and forcing electrons from high energy

34

levels to decay and emit a photon (Figure 11 right panel). This characteristic X-ray gives

the information about the sample elemental composition and a quantitative estimation

of elements concentration [125]. Auger spectrometry can be performed using low energy

electrons (Auger electrons, < 2 keV) to assed sample elemental composition. These

electrons with low energy give information only about the top surface of the samples,

maximum 2 nm deep, due to short mean free path of the generated electrons.

The surface and the cross-sectional morphologies were investigated using a ZEISS

1550VP Field Emission SEM (Carl Zeiss) while energy dispersive X-ray spectroscopy

(EDX) experiments were performed using an EDX X-MaxN (50 mm2), 10 kV (Oxford

Instruments).

Figure 11 – Representation of the sample volume interaction of electron bean in a SEM

measurement (Left panel) and representation of working principle of emission of

characteristic X-ray due to high energetic electron beam (Right panel).

35

2.2.3 Thermal gravimetric analysis (TGA)

The thermal gravimetric analysis is used to evaluate the thermal stability of a material

or composite, for instance, to determine the maximum temperature that can be used to

dry a composite electrode. Essentially, the sample mass is recorded by applying an

isothermal step and/or a temperature ramp as displayed in the Figure 12. The

experiments can be done in different atmospheres, e.g. O2 (oxidising) or N2 (inert).

The TGA experiments of were carried out by using a Q 5000 IR TGA instrument (TA

Instruments). The experiment procedures are distinct depending of the analyzed sample

in N2 or O2 atmospheres were used.

Figure 12 - TGA profile showing the weight loss by increasing the time (isothermal test)

or temperature.

2.2.4 pH-measurement

The pH of a solution can be determined by activity 𝑎𝑎𝑎𝑎3𝑂𝑂+ ion as,

𝑝𝑝𝑎𝑎 = − log[𝑎𝑎𝑎𝑎3𝑂𝑂+] = log 1𝑚𝑚𝑎𝑎3𝑂𝑂+

(19)

being 𝑎𝑎𝑎𝑎3𝑂𝑂+ given by the activity of the hydronium written as

36

𝑎𝑎𝑎𝑎3𝑂𝑂+ = 𝑦𝑦𝑎𝑎3𝑂𝑂+ ∙ 𝑐𝑐𝑚𝑚𝑎𝑎3𝑂𝑂+ (20)

where 𝑦𝑦𝑎𝑎3𝑂𝑂+ is the activity and 𝑐𝑐𝑚𝑚𝑎𝑎3𝑂𝑂+is its concentration.

In fact, the pH value is estimated using the Nernst equation and measuring the solution

electric potential [126].

𝑝𝑝𝑎𝑎 = log 1𝑚𝑚𝑎𝑎3𝑂𝑂+

= 𝑛𝑛𝐹𝐹𝑅𝑅𝑇𝑇∙ (𝐸𝐸 − 𝐸𝐸0) (21)

2.2.5 X-ray photoelectron spectroscopy (XPS)

The XPS is a surface-sensitive method. It measures the kinetic energy 𝐸𝐸𝑘𝑘 [eV] of

electrons extracted from core levels of an atom under a monochromatic X-ray (ℎ𝜈𝜈)

radiation beam (Figure 13). The X-ray energy commonly used are Al Kα line at

1486.6 eV or Mg Kα line at 1253.6 eV. Aside from H and He, all other elements of the

periodic table can be detected if the incident X-ray beam energy is higher than that of

the binding energy of the electrons in the sample elements. Moreover, only sample top

layer surface is analysed (>10 nm), due to the high probability of interaction of the

extracted electron to other atoms of the sample which reduces the number of detectable

electrons. Additionally, the measurement is done at high vacuum conditions to avoid

any further interaction of the extracted electron with atoms in the sample chamber

atmosphere, and also in some cases, to preserve the sample conditions. The law of

conservation of energy in the system gives that the 𝐸𝐸𝑘𝑘 of the extracted electron is directly

correlated to the energy of the X-ray, the binding energy of electron 𝐸𝐸𝑏𝑏 (core level) and

to constant related to the instrument. The 𝐸𝐸𝑘𝑘 values can be calculated using the

photoelectric effect relation in, equation (22) where ℎ is the Planck constant, and 𝜙𝜙𝑆𝑆𝑠𝑠 is

the work function of the spectrometer. Since 𝐸𝐸𝑏𝑏 is defined discrete value of energy in

eV, consequently, 𝐸𝐸𝑘𝑘 values give atomic information of the sample´s top surface.

Moreover, a great advantage of using XPS, aside from atomic identification, the

oxidation state of the atoms can also be evaluated due to the minor change of the binding

energy of an electron in an orbital by the surrounding electrons [125].

37

𝐸𝐸𝑘𝑘 = ℎ𝜈𝜈 − 𝐸𝐸𝑏𝑏 − 𝜙𝜙𝑆𝑆𝑠𝑠 (22)

The X-ray Photoelectron Spectroscopy (XPS) characterization was performed with a

PHI 5800 MultiTechnique ESCA System, using monochromatized Al-K(alpha)

(1486.6 eV) radiation. The measurements were performed with a detection angle of 45°,

using pass energies at the analyzer of 187.85 and 29.35 eV for survey and detail spectra,

respectively. All XP spectra were calibrated to the signal of amorphous carbon (hard

carbon) at 284.8 eV. The peak deconvolution was done using CasaXPS software.

Figure 13 – Representation of emission of a characteristic electron by a monochromatic

X-ray in a XPS experiment.

38

39

Materials

The electrodes active materials, binders, conductive carbon and the ceramic filler used

to prepare the membranes are listed in Table 6. These materials were used without any

pre-treatment.

Two liquid electrolytes were used, the first one consist of 1 M solution of lithium

hexafluorophosphate (LiPF6) in a mixture of ethylene carbonate and dimethyl carbonate

(EC:DMC (1:1 w/w), LP30, BASF), while the second one, denoted as LP30 + VC,

contains 1 wt.% of vinyl carbonate (VC) as additive. The lithium metal was obtained

from Rockwood (battery grade).

40

Table 6 - List of materials used in this thesis.

Binders Material Type Acronym Supplier Application

Sodium carboxymethyl cellulose

CRT 30000 PA 09 CMC09 Dow Wolff Cellulosics

Anode

CRT 2000 PPA 12 CMC

Cathode and Anode

Guar gum Natural GG

Lamberti SpA

Anode and cathode

Hydroxypropyltrimonium chloride guar gum EC3 HPTG Cathode

Hydroxypropyl guar gum C4W HPG Separator HDR Cathode

Pectin From citrus peel - Alfa Aesar Anode Styrene butadiene rubber TRD 102A SBR JSR Micro Anode

Active materials Material Type Acronym Supplier Application

Li[Ni0.33Mn0.33Co0.33]O2 d90 = 10 µm NMC Toda Cathode

C6 SLP 30

d90 = 32 µm Graphite Imerys Anode

Li4Ti5O12

- LTO Süd Chemie AG

Anode Hombitec LTO5

*ps = 250 µm LTO5 Huntsman

Ceramic filler Silicon Dioxide d50 = 1 µm SiO2 Shott AG Separator

Conductive carbon

Amorphous carbon C-NERGY Super C45 *ps : 30 nm

C45 Imerys All electrodes formulations

*ps - particle size

Methods

3.2.1 Membrane preparation

Firstly, a binder solution was obtained by dissolving a predetermined amount of HPG

(viscosity: 3000 – 5000 cps and pH: 5 – 7 (1% solution)) in deionized water by magnetic

stirring. Separately, SiO2 particles were dispersed in deionized water by vigorous

stirring and ultrasonication. This process was applied to destroy agglomerates of

41

ceramic particles, which would affect the morphology of the final membrane. The

binder solution and the SiO2 dispersion were mixed by magnetic stirring until

homogeneous slurry was obtained. The slurry was cast on a Polytetrafluorethylene

(PTFE) tape, using a glass bar, dried at room temperature (20 °C) and subsequently at

80 °C in an atmospheric oven. After drying, the obtained membranes were removed

from the PTFE tape. The final membrane, denoted in this thesis as SiO2 - HPG, was

dried under vacuum at 120 °C for at least 12 h in order to remove the remaining moisture

before further tests.

3.2.2 Membrane characterization

The porosity 𝑃𝑃 (%) of the membrane was estimated by liquid absorption test using

hexadecane (ReagentPlus®) [88]. The porosity measurements were performed on eight

different samples from each membrane. In details, a piece of the membrane was weighed

before and after dipping in hexadecane for 30 min. The excess of hexadecane on the

surface was wiped using filter paper. The porosity was calculated using Equation (23).

𝑃𝑃 = 𝑤𝑤𝑇𝑇− 𝑤𝑤𝑆𝑆𝑅𝑅𝐻𝐻 ∙𝑉𝑉𝑆𝑆

∙ 100 (23)

where 𝑤𝑤𝑆𝑆 is the weight of the pristine separator, 𝑤𝑤𝑇𝑇 is the weight of separator

with the absorbed hexadecane, 𝑑𝑑𝑎𝑎 is the hexadecane density (0.773 g/mL) and 𝑉𝑉𝑆𝑆 is

the volume of the separator [94].

The dimensional shrinkage was evaluated determining the membrane’s areal change

after 12 h storage at 180 °C under N2 [98,127].

For investigation of the wettability and the electrolyte uptake ability of the membranes,

the commercial electrolyte LP30 was used. For the electrolyte uptake, a piece of

membrane was weighted before and after soaking into the electrolyte for 1 h. These

experimental measurements were carried out in a dry-room (R.H. < 0.1%), with five

different samples. The electrolyte uptake 𝑈𝑈 (%) was calculated by using Equation (24).

42

𝑈𝑈 = 𝑤𝑤1− 𝑤𝑤0𝑤𝑤0

∙ 100 (24)

where 𝑤𝑤0 is the weight of the dry separator and 𝑤𝑤1 is the weight of the separator

after absorbing the electrolyte [127]. A visual analysis of the membrane wetting

speed was performed by comparing the pictures of the pristine membrane before and

after a small amount of electrolyte (80 µL) was dropped on the membrane [88,89,98].

3.2.3 Electrode preparation

All electrodes were prepared with the same methodology. At first, the binder was

dissolved in deionized water by magnetic stirring before adding the conductive carbon

Super C45. Once a homogenous dispersion was obtained, the active material and

phosphoric acid (PA) (1 wt.% respect to the weight of the active material, Bernd Kraft

GmbH, Germany) were added and the slurry stirred for 2 additional hours. The addition

of PA is needed to prevent the corrosion of aluminium current collector and was used

for NMC and LTO electrodes. The same procedure was used for the graphite electrodes,

however, SBR rather than phosphoric acid was added.

Finally, the slurries were mechanically mixed at medium-speed (5000 rpm) (4000-4/65,

DREMEL, USA) prior casting them on copper (graphite) or aluminium (NMC and LTO)

foils. The coated electrodes were dried in an open atmosphere oven at 80 °C and,

afterwards, under vacuum at 160 °C for at least 6 hours. Prior to the electrochemical

tests, small discs (area = 1.13 cm2) were punched out of the electrodes tapes and pressed

using a manual press (Atlas manual hydraulic press 15T, Specac, UK) in order to reduce

the electrode porosity, to ensure a homogeneous surface and to enhance electrode

adhesion and density (graphite: 1 t cm-2; NMC and LTO: 10 t cm-2).

The pH measurements were carried out with a Lab 860 pH meter (SI Analytics) using a

Blue Line 18 pH electrode (Schott Instruments) at a temperature of 24 ± 2 °C. The

experiments were conducted using only the active material LTO in water and PA since

the final slurry pH is not influenced by the conductive carbon and binder. The pH

measurements started with only deionized water and subsequently the LTO powder

(40 wt.% of water) was added. A second experiment was done using PA (1 wt.% of

43

LTO) prior the LTO addition. The slurries were stirred for 150 min while the pH

evolution was measured.

3.2.4 Electrode characterization

The electrode adhesion strength was evaluated using a Z2.5 Zwick/Roell machine

(Zwick Roell) at Zentrum für Sonnenenergie- und Wasserstoff-Forschung Baden-

Württemberg (ZSW), Germany.

Figure 14 displays the basic principle of the measurement. Briefly, an electrode of

defined area (6.45 cm2) is fixed between two planar and parallel plates with the help of

double-sided adhesive tape (3M). After a start phase, in which the specimen is

approached before contact is established, the compression phase takes place. Within the

compression phase, the compression stress rises until a defined pressure level (2000 N)

is achieved and then is kept constant during dwell time (120 s), in order to allow the

adhesive to contact the electrode. Afterwards, the pull-off phase (1000 mm min−1) takes

place, and the maximum tensile force is detected. The adhesion strength σ𝑛𝑛 is calculated,

by using Equation (25), from the maximum tensile force |Ft, max| or pull-off force

related to the sample area A.

σ𝑛𝑛 =|𝐹𝐹t,max|𝐴𝐴 (25)

44

Figure 14 – Representation of the adhesion strength measurement.

Electrochemical tests

3.3.1 Cell configuration

Three-electrode cell (Swagelok® cell) is widely used for electrochemical

investigations. This type of cell has a third electrode, which acts as reference (or quasi-

reference) electrode to control the potential of counter and working electrodes. In this

thesis, Swagelok® cell is used in LSV experiments and for the evaluation of the

electrochemical performance of full lithium-ion cell using graphite and NMC as active

materials (using Li metal as quasi-reference electrode).

Two-electrode cells, either coin-cell (CR2032) or pouch-cell, were mostly used in

galvanostatic experiments to test the separator and the electrodes in half-cell (Li metal

as counter electrode) and in full lithium-ion cell using a LTO electrode as anode.

Figure 15 displays representative schemes of cell configurations herein used.

45

Figure 15 – Schematic representation of cell´s configurations and the their electric circuit

used for electrochemical tests; top: Swagelok®; below: pouch and coin-cell. RE –

reference electrode (Li), CE - counter electrode (Li, LTO or graphite electrodes), WE –

working electrode and SE-separator.

3.3.2 Cell assembly and electrochemical test

3.3.2.1 Linear sweep voltammetry

The compatibility of SiO2 – HPG membrane as separator for lithium-ion batteries was

evaluated using linear sweep voltammetry (LSV) in Swagelok® cells. The cells were

assembled in an argon-filled glove box (O2 < 0.1 ppm, H2O < 0.1 ppm) using metallic

lithium (ROCKWOOD LITHIUM, battery grade) as counter (CE) and quasi-reference

46

(RE) electrodes. A disc of the membrane (area ~1.32 cm2) was placed between the

electrodes and soaked with the electrolyte (LP30), while a glass fibre disc (GF/D,

WHATMAN) was used as separator for the reference electrode. The electrochemical

stability window (ESW) of the membranes was tested using aluminium and nickel discs

(area ~1.13 cm2) as working electrode (WE) for the anodic and cathodic sweeps

respectively. Measurements were performed using a VMP3 (BioLogic). A scan rate of

0.1 mV s-1 was applied from the open circuit potential to 6 V and −2 V for the

anodic and cathodic sweeps respectively.

3.3.2.2 Galvanostatic cycling

Pouch-cells were manufactured in a dry-room (R.H. < 0.1%) at 20 °C ± 1 °C in a half-

cell configuration (Li metal anode). The WE (NMC or LTO electrodes with an area

of 1.13 cm-2) was placed on the aluminium tabs, while nickel tabs were used for the

lithium electrode. The separator (SiO2 - HPG or Asahi Kasei, Hipore SV718) was placed

between the electrodes and soaked with the electrolyte (LP30).

Coin-cells and Swagelok® were assembled in an argon-filled glove box (O2 < 0.1 ppm,

H2O < 0.1 ppm). The CE in coin-cell was Li-metal disc and LTO electrodes for half-

and full-cell, respectively. The electrolyte was applied directly in a glass fibre

membrane (GF/D, WHATMAN) used as separator.

Swagelok® cells were used for full-cell tests. The WE was a NMC electrode (area

of 1.13 cm-2) and CE was a graphite electrode and Li-metal as quasi-reference electrode.

The separator was a glass fibre membrane soaked with LP30 + VC electrolyte.

The details of cells used in this thesis are shown in Table 7.

The galvanostatic cycling was performed using a MACCOR Battery tester 4300

(Tulsa) inside climatic chambers (Binder KB 400, 20 °C ± 0.1 °C). The cells were

evaluated using different test procedures, which are reported in details in next chapters.

47

Table 7 - Summary of all cell configurations used in the electrochemical tests.

Separator evaluation

Test Cell type Working electrode Counter electrode Electrolyte Separator

LSV Swagelok® Aluminium or Nickel foil Li metal LP30

SiO2-HPG Asahi

Galvanostatic cycling

HC Pouch-cell NMC or LTO SiO2-HPG

Asahi Binders investigation

Test Cell type Working electrode Counter electrode Electrolyte Separator

Galvanostatic cycling

HC Pouch-cell LTO5 Li metal

LP30

Asahi

FC Coin-cell NMC LTO5 GF/D HC Pouch-

cell NMC Li metal Asahi

HC Coin-cell Graphite Li metal LP30 + VC GF/D Swagelok® NMC Graphite

SiO2 - HPG – membrane reported in this thesis; LP30 - 1 mol solution of lithium

hexafluorophosphate (LiPF6) in a mixture of ethylene carbonate and dimethyl carbonate

(1:1 w/w); LP30 + VC – LP30 + 1 wt.% of vinyl carbonate (VC); Asahi - porous

polyethylene membrane; GF/D – glass felt membrane; HC – half-cell; FC - full-cell.

48

49

The content of this chapter is published in:

Carvalho, D.V.; Loeffler, N.; Kim, G.-T.; Passerini, S. Separator for an electrochemical

device and method for the production thereof. Patent No. EP 3085432 B1, 20 December

2017.

Carvalho, D.V.; Loeffler, N.; Kim, G.-T.; Passerini, S. High Temperature Stable

Separator for Lithium Batteries Based on SiO2 and Hydroxypropyl Guar Gum.

Membranes 2015, 5, 632-645.

50

Introduction

In this chapter an environmentally friendly process is descried to prepare a high

temperature stable membranes, which are then evaluated as separator for lithium-ion

batteries. The membrane is based on SiO2 particles and a low cost water-soluble

polymeric binder. Due to the high abundance in the Earth’s crust, SiO2 is a low cost and

environmentally friendly compound. This material is widely used in the electronic

industry [128,129]. The binder used is the non-ionic hydroxypropyl derivative of guar

gum (HPG) (Figure 16). Guar gums have been proved as a viable aqueous processable

binder for the production of lithium ion batteries electrode, such as LTO [130] and

silicon [122–124] anodes and recently in cathodes [119,120], and gel polymer

electrolytes for super capacitors [131]. The advantages of using HPG compared with

native GG are the enhanced water solubility and thermal stability [132]. Additionally,

HPG is a benign material, mostly used as thickening agent for cosmetics, food and paper

processing and oil industry [133].

The membranes were characterized with respect to porosity, electrolyte wettability,

electrochemical and thermal stabilities, and, finally, tested as separator in half-cells (Li

metal as negative electrode) with NMC and LTO based positive electrodes, since those

two materials are state-of-the-art electrode materials used in commercial lithium-ion

batteries [3,16].

Figure 16 – Structure of hydroxypropyl guar gum.

51

Experimental

The thermal properties of the membranes were evaluated by TGA and dimensional-

shrinkage measurement. The TGA experiments of pure HPG and SiO2, and the

membranes were carried with a temperature ramp of 10 °C min-1 up to 600 °C in both

N2 and O2 atmospheres.

The galvanostatic charge/discharge tests of NMC half-cells were carried out between

3.0 V and 4.3 V vs. Li/Li+, while those of LTO electrodes were performed between

1.0 V and 2.5 V vs. Li/Li+. The galvanostatic tests were performed at different C-rates,

(0.1C, 0.2C, 1C, 2C, and 5C), considering C = 161 mAh g-1 for NMC and

C = 175 mAh g-1 for LTO.

52

Results

Membrane characterization

Several compositions were tested to produce a suitable membrane for electrochemical

energy storage. For instance, Figure 17a shows a SEM image of a membrane with a

composition of 70:30 weight ratio of SiO2 and HPG. A non-homogenous porous surface

was obtained, which leads to an inappropriate flow of lithium ions during charge and

discharge process. Moreover, an as low as possible fraction of binder inside the separator

is desirable, since binder polymers may envelop the filler particles and additionally

reduce the porous structure of the separator. Therefore, the 70:30 ratio was discarded in

this study prior performing electrochemical tests. Membranes were also produced with

low content of binder, 90:10 SiO2:HPG, nevertheless as shown in the Figure 17b, the

membrane is mechanically unstable and cannot be used as a separator.

Figure 17 – a) SEM image of SiO2 – HPG membrane prepared using the ratio of 70:30

and b) picture of SiO2 – HPG membrane prepared using the ratio 90:10.

After the extended exploration of compositions, self-standing homogeneous membranes

were successfully prepared with the 80:20 weight ratio of SiO2 and HPG.

Figure 18 depicts the SiO2 – HPG membrane after drying at room temperature (20 °C).

The final membranes’ thickness ranged between 30 and 50 µm. The membranes appear

quite homogeneous and uniform both in the bulk and at the surfaces, i.e., without binder

53

or filler agglomerations. This is a result of using highly viscous slurries (i.e., having a

gel appearance; HPG:H2O wt. ratio > 0.02), although the drying process takes much

longer than with diluted (i.e., very liquid-like; HPG:H2O wt. ratio < 0.01) slurries. This

is because in the gelled slurries the drying process involve water diffusion through the

membrane rather than binder segregation at the coating’s surfaces and/or filler

sedimentation [134].

Figure 18 - Images of the self-standing SiO2–HPG membrane after drying and removal

from the PTFE tape. The good flexibility and self-integrity of the membrane is well

evidenced in the bottom image.

The homogeneity and uniformity of the membranes may also result from specific HPG–

SiO2 interactions. In fact, HPG is formed by a carbon backbone with numerous

functional hydroxyl groups (Figure 16). These may interact with sites forming on the

surface of the SiO2 particles when dispersed in water. The active bonding, which

54

reaction (see below) was demonstrated with FT-IR studies [122], probably starts during

the slurry-mixing step but takes place mostly during the drying step (needed to remove

all residual water [135]) as it is thermally activated.

𝑅𝑅 − 𝑂𝑂𝑎𝑎 + 𝑆𝑆𝑆𝑆 − 𝑂𝑂𝑎𝑎∆𝑐𝑐→𝑅𝑅 − 𝑂𝑂 − 𝑆𝑆𝑆𝑆 + 𝑎𝑎2𝑂𝑂 (vii)

The bonding improves the physical properties of the membrane granting a good

flexibility of the membrane (see Figure 18) and supporting the homogenous distribution

of particles and binder.

The membrane’s porosity evaluated via the hexadecane infiltration method was found

to be about 52%, which is an appropriate value for separators in lithium-ion batteries

taking into account the commercial the porosity of commercial separator of ca.

30 – 50 % [88,89].

Figure 19 shows the SEM images of the membrane’s top and bottom surfaces as well as

its cross section. The membrane shows very homogeneous top (air contact at drying

step) (Figure 19a) and bottom (PTFE contact at drying step) (Figure 19b) surfaces, which

appear suitable for lithium-ion battery application due to their quite uniform porosity

distribution. In fact, non-homogeneous porosity leads to preferential redox reactions at

specific areas on the electrodes’ surfaces, which may cause local heat evolution and/or

lithium dendritic growth, resulting, e.g., in a cell short circuit [90]. As displayed in

Figure 19a,b, the membrane possess only small pores < 1 µm that can avoid the

migration of particles, such as the active material particles, between the electrodes

[88,89,94]. Figure 19c shows the SEM cross-sectional image of the membrane, showing

as it is formed by a highly porous matrix with pores diameter of ca. 1 µm. Once

impregnated by the electrolyte, the voluminous pores function as an electrolyte

reservoir, improving the lithium ion path through the separator.

55

Figure 19 - SEM pictures of the SiO2 – HPG membrane. (a) Top face (air contact at

drying step); (b) Bottom face (PTFE contact at drying step); (c) Cross section.

Figure 20 depicts a typical test to investigate the wetting ability of separators. The

commercial Polyethylene (PE) single layer separator (20 µm thickness) were used as

reference. A small amount of electrolyte (80 µL, LP30) was dropped on the membranes

surface. While for the PE separator is possible to see a drop of the electrolyte (Figure

20a), the SiO2 – HPG separator was almost fully wetted after 2 s, as can be seen in Figure

20b. This indicates a very good affinity between the SiO2 particles and the organic-based

electrolyte (1 M LiPF6 in EC:DMC = 1:1 w/w) [98]. The combination of excellent

wettability, due to the hydrophilic characteristic of SiO2 particles [89,90,94,136] and the

chemical structure of GG [130], and relatively high internal porosity resulted in the high

absorbed volume of electrolyte by the separator. After 30 and 60 min the weight of the

56

membrane increased by 290% and 370%, respectively while for the PE separator an

increase of only 98% was measured after 60 min.

Figure 20 - Images of membranes subjected to the wettability test. The pictures were

taken 2 s after dropping 80 µL of LP30 electrolyte on the (a) commercial polyethylene

(PE) single layer separator and (b) SiO2 – HPG membrane.

The thermal properties of HPG, SiO2 and the complete membrane were investigated in

N2 and O2 atmospheres up to 600 °C via TGA (Figure 21). The membrane and its

components are thermally stable up to, at least, 200 °C. The binder HPG is the less stable

component as thermally stability up to about 240 °C under N2 (Figure 21a), but only up

to ~200 °C under O2 (Figure 21b), in agreement with already reported data [137]. The

thermal decomposition is attributed to the loss of hydroxyl groups as water molecules

and the breakdown of the backbone chain of the guar molecule [138]. The reduced

thermal stability in O2 is due to the oxidation of the hydroxyl groups. At higher

temperatures (i.e., >240 °C in N2 and >200 °C in O2), both the HPG and the membrane

showed sharp mass losses. However, while under nitrogen a small but continuous weight

loss is observed due to the incomplete binder decomposition (at 600 °C still 10 % of the

HPG binder weight is still present), under O2 atmosphere all HPG binder was lost at 450

°C due to the reaction of carbon (C) and O2 forming carbon dioxide CO2 [139]. In the

case of SiO2, the loss of hydroxyl groups on the particles surface is the reason for the

slight weight loss (below 0.5%) observed in both atmospheres [140]. In fact, the weight

57

loss of the membrane is lower than 75% in both atmospheres, as expected from its SiO2

content (SiO2 - HPG 80:20 wt.%).

Figure 21 - TGA measurements of HPG, SiO2 particles and the SiO2 – HPG membrane

in nitrogen (a) and oxygen (b) atmospheres. Temperature ramp of 10 °C min-1 up to

600 °C in both atmospheres under gas flow (25 mL min−1) using open Al pans.

The thermo-mechanical stability of the separator is crucial for the safety of lithium-ion

batteries. The cathode and anode electrodes must be, in fact, physically separated at any

58

time. The shrinkage and/or decomposition of the separator membrane should not take

place even at relatively high temperatures, since any direct contact of the two electrodes

would induce continuous heat-generation ending in the cell thermal runaway. Therefore,

the thermal behaviour of the SiO2 – HPG membrane was evaluated performing a TGA

isothermal measurement at 180 °C for 12 h. The results of this test are displayed in

Figure 22a. As it can be seen, a negligible weight loss (0.5%) is observed upon the

isothermal test, which, most likely, originated from the release of remaining moisture in

the membrane and the loss of hydroxyl groups from the SiO2 particles surface [140]. To

verify the thermo-mechanical stability of the membrane, i.e., its dimensional stability

upon exposition to high temperatures, a sample was stored at 180 °C for 12 h. Figure 22b compares the size of the same membrane before and after such a storage,

showing that no shrinkage or variation in size took place upon the test. Both results

indicate that the SiO2 – HPG membrane can very well stand the exposition to high

temperatures without any thermal or mechanical degradation. Additionally, the

membrane can stand the high temperatures applied during the drying processes of

lithium-ion batteries, allowing for the all-in-one drying step of the electrodes and

separator assembly.

59

Figure 22 - Thermal stability test of SiO2 – HPG membrane at 180 °C for 12 h (a) TGA

measurement; (b) dimensional-shrinkage test.

Electrochemical tests

The electrochemical stability of the SiO2 – HPG membrane was examined by linear

sweep voltammetry (LSV) at a scan rate of 0.1 mV s−1. The results are compared in

Figure 23 with those of the glass fiber separator commonly used in laboratory-made

lithium-ion cells (GF/D). The two separators showed a rather similar electrochemical

behavior offering wide electrochemical stability windows (ESWs) ranging from the

onset of lithium plating at 0 V vs. Li/Li+ to above 5.0 V vs. Li/Li+. During the anodic

sweep, the SiO2 – HPG membrane shows no current flow up to, at least, 4.0 V vs. Li/Li+.

Above this potential small features are noticeable, however, these currents are never

exceeding the magnitude of 5 µA cm−2 and are always lower than those observed for the

commercial glass fiber membrane. In the cathodic sweep, the behavior until 1.5 V of

both separators is, also, fully comparable. However, for the peak occurring at 1.2 V vs.

Li/Li+ (associated to SEI formation on the Li metal electrode), the current response of

the cell with the SiO2 - HPG membrane is higher. The reason of such a behavior might

be attributed to the higher reactivity of the hydroxyl groups on GG, which can react

60

forming a passive layer on metallic lithium or the higher surface area of the SiO2 particle

with respect to the glass fiber [141]. Nevertheless, both curves follow the same behavior

until 0 V (vs. Li/Li+), where lithium plating takes place. Hence, the SiO2 – HPG

membrane show suitable electrochemical stability for use as a separator in high voltage

lithium-ion batteries [98,99].

Figure 23 - Linear sweep voltammograms of the SiO2 – HPG membrane and glass fiber

separator using 1 mol L−1 (LiPF6) dissolved in EC:DMC (1:1 w/w) as electrolyte. Scan

rate 0.1 mV s−1 at 20 °C. Three electrode cell, WE = Al disc (anodic scan) or Ni discs

(cathodic scan) with Li metal as quasi-reference and CE electrode.

Finally, the SiO2–HPG membranes were tested as separators for lithium cells. Figure 24

shows the first charge and discharge profiles of half-cells comprising NMC or LTO as

active materials at 0.1C and 20 °C.

Figure 24a depicts the typical sloping voltage profile of NMC vs. metallic lithium. At

0.1C, a high discharge capacity of 151 mAh g−1 was delivered by NMC indicating that

the separator allows the full redox reaction of the cathode material. The slight

irreversibility (the charge capacity is higher than the discharge capacity) is always

61

observed with NMC composite electrodes coated in aqueous binders due to leaching of

lithium during the electrode production [28,142,143]. Additionally, Figure 24b shows

the typical flat potential profile of Li+ (de)insertion of in the LTO spinel structure,

associated with the specific discharge capacity of 161 mAh g−1 (at 0.1C). The first cycle

potential profiles indicate as the LTO composite electrode performed very well with the

SiO2 – HPG separator. With both electrodes no anomalous reactions or decompositions

of the separator were observed. Hence, the cells were subjected to long term cycling

tests at different C-rates to further validate the separator stability.

62

Figure 24 - First charge and discharge profile at 0.1C using SiO2 – HPG membrane as

separator in (a) NMC; (b) and LTO half-cells. Galvanostatic charge/discharge in pouch

bag cells using LP30 electrolyte at 20 °C. NMC and LTO loading ~ 4 mg cm-2. NMC

half-cells cycled between 3.0 – 4.3 V for C = 161 mA g-1 and LTO half-cells cycled

between 1.0 - 2.5 V for C = 175 mA g-1.

Figure 25 displays the cycle performance in terms of specific capacity vs. the number of

cycles at different rates of half-cells using NMC (Figure 25a) or LTO (Figure 25b)

composite electrodes.

63

The NMC half-cells show a stable cycle performance during the first 5 cycles at 0.1C,

with specific capacities high than 150 mAh g−1. Upon C-rate increase to 0.5C, the

capacity decreased, however, increasing again upon further cycling at 1C rate, reaching

good performance after 25 cycles. Such a capacity fluctuation is related to the

inhomogeneous wetting of the electrode by the electrolyte solution, which, however,

takes place within a few further cycles. Moreover, the NMC half-cell also shows good

performance at high current density (5C). Finally, very stable performance is observed

in the additional cycles at 1C combined with extremely high coulombic efficiencies,

especially considering the use of lithium metal anode.

Figure 25b shows the cycle performance of the LTO half-cell. Similar to NMC, a

substantial capacity fading is observed during the initial cycles, which can be, once

more, attributed to an incomplete electrode wetting. Nevertheless, after the initial cycles,

the cell showed good performance at 0.5C rate with specific capacity of ~ 140 mAh g−1.

Moreover, at high current density of 1C and 2C, the cell shows a notable performance

stability, and specific capacity of ~ 100 mAh g−1 and 80 mAh g−1 with high coulombic

efficiency.

The stable performance obtained upon long-term cycling of half-cells, containing the

NMC or LTO composite electrodes, indicate that the SiO2 – HPG separator is suitable

for application in lithium-ion cells.

64

Figure 25 - Cycle performance at different C-rates using SiO2–HPG membrane as

separator in (a) NMC; (b) and LTO half-cells. Galvanostatic charge/discharge in pouch

bag cells using LP30 electrolyte at 20 °C. NMC and LTO loading ~ 4 mg cm-2. NMC

half-cells cycled between 3.0 – 4.3 V for C = 161 mA g-1 and LTO half-cells cycled

between 1.0 - 2.5 V for C = 175 mA g-1.

65

Conclusions

In this chapter, an environmentally friendly process to prepare a high temperature stable

separator for lithium-ion batteries was introduced. The self-standing separator was

prepared using low cost materials, such as SiO2 and HPG, through aqueous

processing. The physical characteristics of the separator made with the new process,

such as porosity of 52% and the homogeneous distribution of pores, are suitable for

lithium-ion battery applications. The SEM morphological analysis showed pores

smaller than 1 µm on both surfaces, however, larger pores (> 1 µm) are observed in

the cross-sectional images, which can function as an electrolyte reservoir, in addition

to allowing Li+ ion transport during the charge and discharge. The separator membrane

shows exceptionally good thermal stability at high temperatures without any significant

dimensional-shrinkage. After applying 180 °C for 12 h a weight loss smaller than

0.50% was observed. Linear sweep voltammetry showed that the SiO2 – HPG

membrane separator is electrochemically stable in a wide potential range. More

important, it demonstrates good performance in half-cells (NMC/metallic lithium and

LTO/metallic lithium) with remarkable cycle stability, despite a slightly more marked

capacity fading on the first cycle for both materials due to not completely wetting

of the electrodes.

66

67

The content of this chapter is published in:

Carvalho, D.V.; Loeffler, N.; Kim, G.-T.; Marinaro, M.; Wohlfahrt-Mehrens, M.;

Passerini, S. Study of Water-Based Lithium Titanate Electrode Processing: The Role of

pH and Binder Molecular Structure. Polymers 2016, 8, 276.

68

Introduction

Aside from graphite, spinel-structured lithium titanate Li4Ti5O12 (LTO) is commercially

used as the anode for LIB, providing remarkable advantages, such as stable cycling

performance even at high C-rates, low toxicity and enhanced safety [26]. Those

characteristics render LTO suitable for high power LIB application as those used in

electric and hybrid vehicles and power tools. Moreover, the combination of the LTO

desirable properties with the aqueous electrode manufacturing would lead to LIB anodes

with enhanced safety and performance. Nevertheless, the alkaline nature of aqueous

LTO slurries favors the corrosion of the aluminum current collector upon electrode

coating [108]. To avoid this issue, the use of mild acids, i.e., phosphoric (PA) and formic

(FA) acids, as pH modifiers of Li[Ni0.33Mn0.33Co0.33]O2 (NMC) electrodes has been

earlier investigated [105]. It was reported that both acids can reduce the pH of the

electrode slurry to values of about 9, thereby preventing corrosion. Moreover, better

cyclability was shown for electrodes prepared with the addition of PA due to the

formation and deposition of highly insoluble phosphate compounds on the active

material surface of NMC, which reduce the transition metal leaching and enhance the

electrochemical performance.

This chapter focuses on the development of an aqueous process to prepare LTO

electrodes, using CMC as the binder, but with the addition of phosphoric acid (PA) as

the pH modifier. Additionally, different polysaccharides, i.e., GG (branched

mannose:galactose 2:1 chain) and pectin (α-linking galacturonic acid chain) (see Table

5), were investigated as alternative binders for LTO electrodes, assessing the adhesion

strength, thermal stability, electrode morphology and electrochemical performance.

Finally, the electrochemical performance of full cells comprising NMC and LTO

electrodes, both prepared via aqueous processing using natural polymeric binders, are

discussed.

69

Experimental

The thermal properties of CMC, GG, pectin and LTO were evaluated by TGA. The TGA

experiments were carried out by heating the respective specimen from 30 °C up to 500

°C with a heating rate of 5 °C min−1 under a nitrogen gas flow (25 mL min−1) using open

aluminium pans. The samples (10 – 20 mg) were evaluated without any pre-treatment.

Half-cells, i.e., cells made with the Li metal anode were assembled in pouch bag

configuration in order to evaluate the effect of PA and the different binders (CMC, GG

and pectin) on the LTO electrodes’ performance.

Table 8 displays the acronym of each electrode used in this chapter. The cells were

assembled in a dry-room using LP30 electrolyte. The porous polyethylene membrane

from Asahi was used as the separator. Full-cells (coin-cells) were also assembled to

evaluate the potential of the aqueous processed electrodes in Li-ion cells in. The

galvanostatic charge/discharge tests of LTO half-cells were performed between 1.0 V

and 2.5 V vs. Li/Li+. The galvanostatic tests were performed at different C-rates (0.1C,

0.5C, 1C, 2C, 3C and 5C). The full cell (NMC/LTO) tests were carried out between 1.3

V and 2.8 V.

Table 8 – List of electrodes used in this chapter.

Active material Binder pH-modifier Acronym

LTO5

Sodium-carboxymethyl cellulose - LTO-CMC

Phosphoric acid

LTO-CMC-PA

Guar gum LTO-GG-PA

Pectin LTO-Pectin-PA

NMC Sodium-carboxymethyl cellulose NMC

(-) without pH-modifier in the slurry;

70

Results and discussions

5.3.1 Thermal stability

Figure 26 displays the TGA results in N2 atmosphere of LTO, CMC, GG and pectin.

LTO particles are thermally stable to 500 °C since no material degradation is detected

up to this temperature. The three binders, on the other hand, show comparable weight

loss profiles up to 200 °C. The weight decrease at 200 °C is related to water desorption

from the polymers since they were not pre-dried. Above 200 °C, pectin exhibits a sharp

decomposition, thus showing a lower thermal stability than CMC and GG, in line with

previous reports [144]. In contrast, CMC and GG start to decompose only at

temperatures above 250 °C [28,145]. The decomposition mechanism of all three

polymers is determined by the breakdown of the main polymer chain

[28,30,92,144,145]. Overall, all binders showed thermal stability at least up to 200 °C,

which allows the high temperature (180 °C) drying of the coated electrode without

thermal decomposition.

71

Figure 26 - TGA weight loss profiles of LTO, CMC, GG and pectin with a heating rate

of 5 °C/min in N2 atmosphere. Heating rate of 5 °C min−1 from 30 °C up to 500 °C

under a nitrogen gas flow (25 mL min−1) using open Al pans.

5.3.2 Electrode surface characterization

SEM images of non-pressed electrodes are shown in Figure 27. Panel (a) shows the

micrograph of the LTO-CMC electrode prepared without PA addition. The electrode

surface is dominated by cavities generated by the gas evolution, which originates from

the reaction of the alkaline slurry with the aluminium current collector during the casting

and drying step. In fact, without the addition of PA, the slurry achieves a pH of 11.4 (see

Figure 28), leading to aluminium (current collector) corrosion and H2 evolution (see

reactions (v) and (vi)) [106]. On the other hand, the electrodes prepared by adding

1 wt % PA (Figure 27b; LTO-CMC-PA) show no cavities due to the pH adjustment at

values around ~ 6.7, which avoids Al corrosion (see Figure 28). Moreover, Figure 27c,d

displays the surface of the electrodes prepared using pectin and GG as binders. Cracks

can be observed on the surface of the LTO-pectin-PA electrode. The LTO-GG-PA

72

electrode surface is much more homogeneous than that of LTO-pectin-PA, but small

defects can also be detected. The defects on the electrode are associated with the binder

shrinkage during the final drying step.

Figure 27 - SEM images of unpressed LTO electrodes using: (a) CMC as the binder

(LTO-CMC); (b) CMC as the binder and phosphoric acid (PA) (LTO-CMC-PA); (c)

pectin as the binder and PA (LTO-pectin-PA); (d) and GG as the binder and PA (LTO-

GG-PA).

73

Figure 28 - pH evolution of the aqueous slurry containing only LTO active material

(circles) and LTO and phosphoric acid (squares). Measured at 20 °C with 40 % of LTO

in water.

For a detailed investigation of the Al current collector corrosion, energy dispersive X-

ray spectroscopy EDX experiments were performed on several spots of the LTO-CMC

electrode. Table 9 shows the elemental composition and Figure 29 the SEM micrograph

of the spots evaluated. Spots 1 (Spectrum 1) and 2 (Spectrum 2) are in the depth of the

cavity, while Spot 3 (Spectrum 3) is on the electrode surface. The high fraction of Al

detected on Spots 1 (87.69%) and 2 (66.12%) reveals the exposure of the current

collector at the bottom of the observed cavities. In Spot 3, a small fraction of Al was

detected (1.2%), which is a side product of the Al corrosion, mainly Al2O3, which after

solubilisation from the current collector, is redeposited onto the solid electrode

components after solvent evaporation. Thus, the high pH value of the aqueous LTO

slurry induces the corrosion of the aluminium current collector, and traces of Al can be

detected in the composite electrode layer. The influence of the electrode corrosion will

be discussed in the next sections.

74

Figure 29 - SEM image of the electrode prepared using CMC as binder without PA

(LTO-CMC) showing three spots where EDX experiments were performed.

Table 9 - Element fractions estimated by EDX measurements on electrode surface

prepared without PA (LTO-CMC).

Spectrum 1 Atomic % Spectrum 2 Atomic % Spectrum 3 Atomic %

C 8.80 C 28.19 C 84.09

O 3.50 O 5.31 O 10.09

Al 87.69 Al 66.12 Al 1.20

Na 0 Na 0 Na 0.45

Ti 0 Ti 0.38 Ti 4.16

Total 99.99 Total 100 Total 99.99

5.3.3 Adhesion strength

Electrode adhesion strength is a relevant factor for LIBs’ development and production

[146]. The coated composite electrode must, in fact, stand the mechanical stress upon

75

cutting, winding, cell assembling processes and for battery cycle life. Figure 30 displays

the adhesion strength of the LTO electrodes. The adhesion strength of the electrode

prepared without PA (LTO-CMC) could not be determined, due to cohesion-failure

during the pull-off step. Indeed, parts of the electrode layer remained on both the

adhesive tape and the current collector. This is an extremely negative phenomenon

because the delamination of small areas of the electrode layer during processing would

reduce the electrode active layer, strongly affecting the cell performance. As depicted in

Figure 27 (LTO-CMC), the electrode’s surface morphology is not homogeneous due to

the aluminium corrosion. Even after electrode compression, the solid particles in the

composite electrode do not adhere to each other. On the other hand, the adhesion strength

of electrodes coated from PA-containing slurries was successfully measured. The LTO-

CMC-PA electrodes exhibited the highest value of adhesion strength (> 1100 kPa)

compared to the LTO-pectin-PA (> 600 kPa) and LTO-GG-PA (> 450 kPa) electrodes.

The higher adhesion strength of CMC compared to GG electrodes can be explained by

the linear β-linkage polymer chain geometry. The interchain hydrogen bonds between

the cellulose chains are stronger than between the galactomannan branched chains [145].

Moreover, the linear α-linkage pectin galacturonic acid chain seems to be weaker than

the β-linkage of the CMC molecule (see Table 5). In fact, it has been reported that α-

linkage polysaccharides are flexible, whereas β-linkage polymers are stiff [121].

76

Figure 30 - Adhesion strength of LTO electrodes fabricated using PA as an additive and

CMC, GG and pectin as the binder.

5.3.4 Electrochemical characterization

In order to evaluate the influence of the polymeric binders in LTO electrodes’

performance, half-cells were assembled and tested at several C-rates.

In Figure 31 are depicted the discharge capacities of half-cells assembled with lithium

metal and LTO electrodes, these latter incorporating different binders (CMC, CMC-PA,

GG-PA and pectin-PA). During the first five cycles at low C-rate (0.1C), the electrodes

made from PA-containing slurries showed similar capacity values, while slightly lower

discharge capacities were observed for the electrode made without PA. Once the current

density increases to 0.5C and subsequently to 1C, the difference in cycling performance

between CMC and CMC-PA electrodes becomes very obvious. While both electrodes

show a stable cycling performance, higher discharge capacities are delivered from the

CMC-PA electrode. At high C-rates (2C, 3C and 5C), the difference is even more

pronounced. At 5C (Figure 31a; 30th cycle), 144 mAh g−1 and 124 mAh g−1 were

LTO-CMC-PA LTO-GG-PA LTO-Pectin-PA 0

200

400

600

800

1000

1200

Adhe

sion

stre

nght

(kPa

)

Binder

77

delivered by the CMC-PA and CMC electrodes, respectively. The lower discharge

capacities of the CMC-based electrode are mostly related to the lower conductivity of

the loosely-packed coated layer. Upon the addition of PA in the slurry (Figure 31a), the

electrodes exhibit remarkable electrochemical performance, even at high C-rates.

Additionally, the cells showed excellent capacity retention after five cycles at 5C and

the subsequent cycles at 1C. However, the LTO-CMC-PA electrode shows a lower rate

capability compared to those based on GG-PA and pectin-PA. This could be related to

the higher electrolyte uptake of, e.g., GG compared to CMC [122,145]. In fact, the

motion of the galactomannan and the ether oxygens’ lone-pair electrons of the GG

molecule coordinate Li+ ions in its structure, comparable to polyethylene oxide (PEO)

in solid electrolytes, which accounts for the improvement in electrochemical

performance [123,131]. However, CMC-PA electrodes show the highest value of

adhesion strength, due to the optimal binding of the solid material components resulting

from the homogeneous binder distribution, i.e., the optimal electrode preparation. Thus,

it is reasonable to assume that the better electrochemical performance of LTO electrodes

made with GG and pectin is rather related to increased ionic conductivities than solid

particle adhesion. Additionally, pectin-PA and GG-PA electrodes show comparable

electrochemical performance at low C-rates of 0.1C, 0.5C and 1C. However, at 2C, 3C

and especially 5C, electrodes using pectin as the binder showed the highest capacity. As

the adhesion strength is obviously not directly correlated with the electrochemical

properties of the herein presented electrodes, the best performance of pectin electrodes

has to be related to its different molecular structure and its affinity towards the

electrolyte facilitating lithium ion transport.

Figure 31b displays the discharge capacity and the coulombic efficiency of a cathode-

limited, lithium-ion cell consisting of NMC and LTO electrodes, both made using CMC

as the binder and PA as the pH-modifier. Besides the first cycle at low current density

(0.1C), the test was performed at constant charge/discharge current densities of 1C. The

cells showed an average discharge capacity of ~120 mAh·g−1 in the course of 190

consecutive charge/discharge cycles at a 1C rate. Moreover, high values of coulombic

efficiency (~99.8%) were achieved. These results confirm the validity of making NMC

and LTO electrodes using CMC and PA as the binder and the pH-modifier, respectively,

resulting in fully-aqueous processed LIB with remarkable performance.

78

Figure 31 - Delivered discharge capacity of (a) LTO half-cells using CMC as the binder

and CMC, pectin and GG as the binder and PA as an additive at several current densities;

LTO mass loading: 3.6–4.1 mg cm−2; electrolyte: LP30; and (b) cathode-limited

Li[Ni0.33Mn0.33Co0.33]O2 (NMC)/LTO full-cell using CMC as the binder and PA as an

additive at 1C; NMC mass loading: ~ 4.3 mg cm−2; electrolyte: LP30.

79

Conclusions

The reported results prove the applicability of polymers from renewable sources as

binders for LTO electrodes, using PA as the pH-modifier. A small addition of the latter,

in fact, avoid the corrosion of Al current collector and leads to a great performance

improvement due to optimal composite electrode cohesion and adhesion to the current

collector.

In addition to sodium carboxymethylcellulose, two natural binders, GG and pectin, were

evaluated for making LTO electrodes. Regarding the thermal stability, the

decomposition of pectin was detected near 200 °C, while GG and CMC are stable up to

250 °C. Thus, all three binders can support high temperature drying (180 °C). Electrodes

prepared using GG showed the lowest adhesion strength, mostly due to its branched

mannose polymer chain. CMC-PA electrodes (linear cellulose chain) were more

adhesive than pectin (linear galacturonic acid chain). The effect of the polymeric chain

was also evident in the electrochemical test: GG- and pectin-based electrodes showed a

slightly superior rate capability compared to CMC electrodes. As discussed, the GG

molecule offers a better affinity for the organic electrolyte than CMC. Additionally, the

binder ability to coordinate Li+ affects the overall electrochemical performance. The Li+

coordination by GG and pectin polysaccharides may be higher than that of the CMC

molecule due to the lower motion of the linear cellulose chain. Moreover, the pectin

polymer might also have good affinity to organic electrolytes, such as the GG molecule,

due to the α-linking between the galacturonic acid rings, resulting in higher electrolyte

absorption.

Finally, full lithium-ion cells were manufactured using NMC and LTO electrodes both

prepared via the aqueous process using CMC as the binder and PA as the pH-modifier.

The full-cell delivered a stable and remarkable discharge performance of ~120 mAh·g−1

at 1C over 190 cycles with high coulombic efficiency (99.8%). These results reaffirm

the suitability of making LIB electrodes by simple and inexpensive aqueous processes.

80

81

The content of this chapter is partially published in:

Carvalho, D.V.; Loeffler, N.; Hekmatfar, M.; Moretti, A.; Kim, G.-T.; Passerini, S.

Evaluation of guar gum-based biopolymers as binders for lithium-ion batteries

electrodes. Electrochimica Acta 2018, 265, 89.

82

Introduction

It was reported, in the previous chapter, the use of bio-source polymers, CMC, GG and

pectin as binder in LTO electrodes. In line with Kim co-workers [130], improved

electrode´s rate capability using GG instead of CMC were obtained. Zhang et al. [119]

reported superior cycling stability of lithium-rich cathode by using GG as binder, while

Lu co-workers. [120] claimed improved capacity retention using GG in lithium-sulphur

cells. The superior electrochemical performance achieved by the electrodes containing

GG has been attributed to the higher affinity of this polymer toward the non-aqueous

electrolyte with respect to CMC. Moreover, as reported by other authors [123,131] GG

coordinates Li+ similarly to polyethylene oxide (PEO) in solid polymer electrolytes,

leading to enhanced ionic conduction. Herein this chapter reports the study of natural

GG as binder for graphite electrodes. The binder coverage on the active material surface

is evaluated by XPS analysis in NMC and graphite electrodes. The graphite electrode

morphology and electrochemical behaviour are assessed and compared with those of

CMC based electrodes. Finally, the investigation on full cells comprising both electrodes

prepared via aqueous processing is presented, coupled with the post-mortem analysis by

X-ray Photoelectron Spectroscopy (XPS).

83

Experimental

The graphite half-cells were tested with a combined CC/CCCV protocol consisting of

galvanostatic charge and discharge cycles between 0.01 and 1 V versus Li/Li+, but with

an additional constant voltage step (CV at 0.01 V vs Li/Li+ until the current was lower

than 0.05 C) at the end of the lithiation. Full Li-ion cells were cycled at 1C between

3 - 4.2 V at 20 ± 1 °C using the CC discharge and CCCV charge protocols.

84

Results and discussions

6.3.1 Graphite half-cells

In order to evaluate the impact of the binder in anodes, graphite electrodes were made

using the same ratio (2 wt.%) of CMC and GG as binder. For these electrodes, 2 wt.%

of SBR was employed in the electrode formulation to enhanced the electrodes

mechanical properties.

The first lithiation/delithiation of graphite half-cells are present in Figure 32. Both cells

show similar voltage profile upon lithiation, and the delivered capacity was ca. 401.5

mAh g-1 and 395 mAh g-1 for CMC and GG based electrodes, respectively. Due to

decomposition of electrolyte upon formation of the solid electrolyte interface (SEI), the

first lithiation capacity is are rather high compared to the theoretical capacity of graphite

(372 mAh g-1). Once again, upon delithiation both cells show same voltage slop, and

capacities of 365 mAh g-1 and 361 mAh g-1 for CMC and GG containing electrodes. The

calculated reversible capacity in the first cycle is 91.7 % for CMC and 91.1% for GG

containing electrodes, which are reasonable values for laboratory scale electrodes.

85

Figure 32 - Comparison first cycle voltage profile of graphite half-cells containing GG

or CMC as binder (2 wt.%). Coin cells using GF/D membrane as separator and LP30 +

VC as electrolyte cycled at 20 °C. Graphite loading ~ 6.3 mg cm-2. C = 372 mA g-1.

The performance of graphite cells can be appreciated in the rate capability test and on

the long-term cycling test in Figure 33. From low to high current rates, both cells display

comparable specific capacity as for instance of ca. 350 mAh g-1 at 1C and above

250 mAh g-1 at 3C. Again once the rate decreases to C/2, both cells still performing

identically. Moreover, in the Figure 33b, the long-term cycling of these cells is shown.

The cells delivered equal capacities of ca. 360 mAh g-1 at C/2 with high coulombic

efficiency (~ 99.9 %). Additionally, the capacity retention (30th to 100th cycle) of the cell

containing GG as binder is around 97.4 %, which is equivalent value to the state-of-art

binders for graphite anodes in LIB (CMC/SBR), showing the feasibility of use GG in

combination to SBR as binder in graphite electrodes.

86

Figure 33 – Comparison of delithiation capacity upon cycling of graphite half-cells

containing GG or CMC as binder (2 wt.%). a) rate-capability test and b) long-term

cycling (cycle 30 to 100) and coulombic efficiency. Coin cells using GF/D membrane

as separator and LP30 + VC as electrolyte cycled at 20 °C. Graphite loading

~ 6.3 mg cm-2. C = 372 mA g-1.

87

Figure 34 displays the SEM image of graphite electrodes prepared using GG or CMC

(2 wt.%). From the low and high magnification image, any particle and binder

agglomeration can be noted. Moreover, the distribution of nanosized conductive carbon

is homogenous on the surface of both electrodes, showing that the electrode formulation

based on the combination of GG and SBR is appropriated to make graphite electrodes.

Figure 34 - SEM images of uncalendered graphite electrodes prepared using 2 wt.% GG

(left) 2 wt.% CMC (right).

6.3.2 XPS analysis of pristine electrodes

The surface of pristine graphite anodes containing 2 wt.% either CMC or GG binders

were examined by X-ray photoelectron spectroscopy (XPS) and can be appreciated in

Figure 35. It has been reported that branched polymers such as GG, tara gum [130],

amylopectin and glycogen [147] can form, on particles surface, a thinner and

homogenous layer compared to linear polymers like CMC and PVDF. Nevertheless,

both samples demonstrate identical peak features in C 1s and O 1s regions with

88

comparable intensities indicating no difference of the binder coverage on the graphite

surface.

Figure 35 - XPS analysis of pristine graphite electrodes prepared using 2 wt.% of GG or

CMC. a) C 1s and b) O 1s binding energy. The dashed lines refer to C-C and C-H

bonding at ca. 285 eV and C-O bonding at 287.8 eV.

The low amount of binder in the electrode formulation, only 2 wt.% of GG or CMC with

the addition 2 wt.% SBR, seems to not influence the formation of a binder film on the

surface of graphite particles as appear in the SEM image (Figure 34) and also not

detectable by the XPS (Figure 35). Therefore, the surface of pristine NMC electrodes

was also investigated using XPS. In the cathode formulation 5 wt.% of the binder was

used, and thus it is expected rather higher influence of the binder coverage on the NMC

surface [148].

Figure 36 displays the Li 1s core level spectra of the cathodes containing 5 wt.% binder.

The Mn 3p signal (the strongest contribution at around 49 eV for Mn 3p3/2) is detected

89

near to the Li 1s peak in Figure 36. It can be noticed that this peak has a lower intensity

in the case of CMC respect to any of the guar-based electrodes, which can be related to

the presence of a thicker binder layer as also reported in LTO electrodes by Lee and co-

workers [130].

Figure 36 - XPS results of pristine NMC cathodes containing 5 wt.% of GG or CMC as

binder. The dashed lines refer to the Li 1s peak at ca. 55 eV, the Mn 3p3/2 one at ca.

49 eV and the Mn 3p1/2 one at ca. 51.5 eV.

6.3.3 Lithium-ion cell

In order to evaluate the influence of GG binder on both electrode formulation, Li-ion

cells were assembled using NMC cathode, and graphite anode prepared with GG binder

as well (2 wt.%). The NMC electrodes using only 3 wt.% GG is reported to deliver

higher capacities and enhanced cycle stability than electrodes prepared using 5 wt.% of

GG [148]. Therefore, the NMC made using only 3 wt.% of GG is used for the full-cell

tests. As it can be appreciated in Figure 37, the lithium-ion cell made with these

electrodes delivered 150 mAh g-1 in the initial cycles at 0.1C and a stable discharge

90

capacity of about 110 mAh g-1 at 1C rate. Moreover, a capacity retention of 85 % and a

remarkable coulombic efficiency of 99.8 % over 200 cycles were achieved.

Figure 37 - Galvanostatic cycling performance of the NMC (3 wt.% GG)/Graphite

(2 wt.%) Li-ion cell. Three electrode cell using GF/D membrane as separator and Li

metal as quasi-reference electrode, NMC electrode (loading ~ 7 mg cm-2) as WE and

graphite electrode as CE (loading ~ 2.8 mg cm-2) at 20 °C. Electrolyte LP30 + VC.

C = 161 mA g-1.

The obtained electrochemical performance is comparable with that achieved using NMC

electrodes containing 5 wt.% of CMC which resulted to be the best formulation for NMC

cathodes in previous works [29,105].

In summary, the graphite anode made using GG displayed comparable electrochemical

performance to those obtained using CMC (Figure 32). Additionally, no difference was

appreciated in terms of surface coverage (Figure 35) and morphology (Figure 34)

indicating that GG is a suitable binder for graphite. Moreover, the stable cycling

performance of the Li-ion cell in Figure 37 confirms that GG can also be used in

cathodes electrode formulation.

91

6.3.4 Lithium-ion cells and post-mortem surface investigation

Figure 38 compares the electrochemical performance of the Li-ion cell with 3 wt.% of

GG and 5 wt.% of CMC in the cathode formulation. At low rate, both cells display

similar capacity, but at 1C the cell with CMC-based electrode delivers higher capacity.

The capacity retention was 91% and 89% for CMC and GG respectively.

Figure 38 - Galvanostatic discharge capacity of Li-ion cells (NMC/Graphite) employing

GG or CMC as binder. Cathode limited full cell. Three electrode cell using GF/D

membrane as separator and Li metal as quasi-reference electrode and cycled at 20 °C;

Electrolyte: LP30 + VC. C = 161 mA g-1.

The influence of GG in the SEI formation on both anode and cathode surface was

investigated by XPS on cycled electrodes extracted from Li-ion cells. After 100 cycles

at 1C, the cells were transferred in a glove-box and disassembled in the complete

discharge state (3 V), i.e. with fully lithiated NMC electrode. The electrodes were rinsed

using DMC and dried under vacuum.

92

6.3.5 XPS analysis of cycled cathodes

It is important to recall here that the amount of binder in the cathodes differ (5 wt.% for

CMC and only 3 wt.% for GG-based one). Thus, a precise quantitative analysis of the

SEI species is inappropriate. Nevertheless, the influence of the binder chemistry on the

formation of SEI species is discussed. Figure 39a-d displays the XPS results of cycled

cathodes. Two distinct features are found in the P 2p region (panel a); the peak centred

at 137 eV accounts for LiPF6 remaining salt and LixPFy decomposition product. The

fitting of the peak at lower binding energy results in two doublets, one at 134.1 eV with

equal intensity on both samples and a second peak around 133.3 eV that is related to P-

O bonds such as those in Li3PO4 or LixPOyFz. Phosphates are expected to be present in

both samples as result of the addition of H3PO4 in the cathode slurries. Those species

are mostly water-insoluble and bound with the NMC surface [105].

The C 1s spectra (Figure 39b), similar for both electrodes, display the signal at ca.

289 eV attributed to poly(VC) resulting from the anodic decomposition of vinyl

carbonate (VC) used as an additive in the electrolyte [71].

The O 1s region in panel c displays the C-O-C and C꞊O bonding energy at ca. 531.4 and

533 eV arising respectively from the oxygen in the binder and the carbonates species.

Once can be assumed, that the most exposed NMC surface in GG-based electrodes

(lower binder coverage) induces more electrolyte decomposition. In contrary, the peak

from degradation carbonates solvents (ca. 287 eV) in the C 1s region and 531 eV (O 1s

spectra) are similar in both samples.

Moreover, the cathode containing GG presents a more intense peak at ca. 529.5 eV

reported to be characteristic of metal-oxygen (M-O) bonding [71,82]. The higher signal

of M-O bonding and the phosphates (Figure 39a) is a result of the early discussed

(Figure 36) thinner binder layer when GG is employed and might result in a thinner SEI.

Concerning the F 1s spectra (panel d), a peak at around 687 eV is present in both

electrodes, which is related to LiPxFy species. However, the sample containing CMC

present a more marked contribution at 685 eV, which is characteristic of LiF [70]. This

fact can be a result of the higher amount of binder in the CMC electrode (5 wt.% vs

3 wt.% in GG-based). The hydroxyl groups in the polymers structure can react with

LiPF6 salt resulting in LiF species [70].

93

In summary, the lower coverage by GG binder did not induce high consumption of

electrolyte and lithium salt but seems to acts as a protective layer. This GG coating layer

was visualised by transmission electron microscopy by Zhang et al. [119] in layered

lithium-rich cathode materials.

Figure 39 - XPS results of cycled NMC electrodes containing 3 wt.% of GG and 5 wt.%

CMC as binder of a) P 2p, b) C 1s c) O 1s and d) F 1s binding energy.

94

6.3.6 XPS analysis of cycled anodes

In the case of the graphite anodes, a more quantitative analysis of the XPS signals is

feasible as the amount of binder employed, CMC or GG, was 2 wt.% in both cases and

the amount of SBR is also equal (2 wt.%).

The XPS results for Li 1s, P 2p, C 1s, O 1s and F 1s core levels of the cycled anodes are

shown in Figure 40a-e. For the electrode containing CMC, the peak intensity of lithium

carbonate/alkoxide (i.e., electrolyte decomposition products) in the Li 1s region

surpassed that of the GG-containing one. Similar trends are observed in the P 2p, C 1s,

O 1s and F 1s regions, with the surface of the CMC-based electrode displaying the peaks

with the highest intensity, respectively attributed to P-O (LixPOyFz or phosphates), C=O

and O-C=O (carbonates) and LiF.

The electrolyte decomposition detected by XPS measurements in CMC-based electrodes

not only by the electrochemical process but also via the chemical reaction between the

binder and the electrolyte [70]. In fact, the SEI on graphite electrode can be formed

during not only dis/charge. Li et al. reported the formation of a LiF rich passivation film

onto graphite electrodes after calendar ageing [72].

95

Figure 40 - XPS results of cycled graphite electrodes containing GG and CMC as binder.

a) Li 1s b) P 2p c) C 1s d) O 1s e) F 1s.

XPS is capable of collecting data only from the topmost surface layer (10 nm). To

enables the analysis of the inner part of the SEI the electrodes were etched via Ar+

sputtering for 3 and 10 minutes. Figure 41 displays the XPS spectra recorded after the

etching process.

96

Figure 41 - XPS results of cycled graphite anodes containing GG and CMC as binder

after 3 and 10 minutes of the etching process. a) Li 1s b) P 2p c) C 1s d) O 1s e) F 1s

binding energy.

To facilitate the interpretation of Figure 41, Figure 42 reports the evolution of the

concentration of LiF, LixPFy, and LixPOyFz species with sputtering time. The SEI formed

on the CMC-based electrode contains higher amounts of these fluorinated species in

comparison with the GG-containing electrode. Even after 3 minutes of etching, the

CMC-based electrode still displays higher quantities of LixPFy and LixPOyFz than the

GG-based electrode. However, comparable values are recorded after 10 min sputtering,

indicating that the graphite bulk was probably reached.

97

Figure 42 - Concentration of LiF, LixPFy, LixPOyFz species by the XPS experiments on

cycled graphite electrodes as obtained from the cell and after 3 and 10 min of etching.

The values are normalized concerning the CMC species.

Overall, the XPS results indicate that a thinner SEI is formed in GG-based anodes and

the SEI has a different composition depending on the binder used in cathodes. GG and

CMC have many hydroxyl (-OH) groups in their structure that can react with LiPF6

(Equation viii) resulting in LiF and HF species as suggested by El Ouatani [70].

𝑅𝑅 − 𝑂𝑂𝑎𝑎 + 𝐿𝐿𝑆𝑆𝑃𝑃𝐹𝐹6 → 𝑅𝑅 − 𝑂𝑂 − 𝑃𝑃𝐹𝐹4 + 𝐿𝐿𝑆𝑆𝐹𝐹 + 𝑎𝑎𝐹𝐹 (viii)

98

However, the GG chemical reactivity seems to be lower than that of CMC as the amount

of LiF and other fluorinated species in the SEI, arising from the salt reaction, is smaller.

Also, the cycled GG-based electrode showed the lower concentration of Li-bearing

species compared whit CMC-based one (Figure 39 and Figure 42). This effect might be

due to the exchange of Na+ with Li+ at the carboxymethyl group as reported by El

Ouatani [70] obviously occurring only with CMC.

99

Conclusions

XPS experiments on pristine cathodes unveil that GG (branched polymer) form a thinner

binder layer on active material than CMC (linear polymer). Further post-mortem

analysis of the NMC and Graphite electrodes from Li-ion cells reveals the influence of

binder in the SEI components. The composition of the SEI formed on the cathode is

mostly unaffected by the binder chemistry. On the other hand, higher contents of Li, P

and F species and thicker SEI were detected on the graphite electrode made using CMC

than GG. Notably, only a simple mechanical process is employed to produces GG, which

is an additional advantage from the environmental point of view when compared with

the mechanical and chemical processes needed to produce CMC. In summary, the

natural polymer extracted from the seeds of guar plant demonstrated to be an excellent

aqueous binder for both the positive and the negative Li-ion electrodes’ production.

100

101

This thesis aimed at exploring and evaluating aqueous processes for the preparation of

LIB components. A porous separator membrane and electrodes were developed using

only aqueous processable polymers. Moreover, the binders herein evaluated are obtained

from renewable sources and offer lower costs compared to the state of the art binders.

The electrode formulations and the materials used in the electrodes (conductive carbon,

active materials, current collector) are similar, or even identical, as used in electrode

preparation on a pilot scale, which facilitates the transference of the laboratory scale

development to a further evaluation at industrial scale.

The separator membrane is prepared by a simple mixing process of 80 % of SiO2 with

20 % of water soluble polymer, HPG. It was found that lower amount of binder leads to

a mechanically unstable membrane while higher amounts of binder generate membranes

with low and inhomogeneous porosity. The successfully developed membrane shows

excellent wettability with commercial liquid electrolyte. Moreover, the separator is

thermally stable till ~ 200 °C in air and revealed to be electrochemically stable within a

broad range of potential. Once tested in combination with NMC or LTO electrodes, the

cells assembled with this membrane show acceptable electrochemical performance

demonstrating that both materials, SiO2 and HPG, and the aqueous process are suitable

for LIB applications. Nevertheless, further evaluations are needed, as mechanical tests,

for instance as elongation at break test. This characterization may provide information

how the membrane supports the forces during the cell assembling. Moreover, the

separator could be evaluated using other active materials, as LFP or graphite, and

different types of electrolytes. These investigations might provide a better understanding

of the use of SiO2 - HPG separator for several combination of electrochemical active

materials for LIB.

A strategy to prepare LTO electrodes via aqueous processing was also presented in this

work. The use of a mild acid, PA, to control the pH of the slurries avoids the corrosion

of Al current collector, enhances the adhesion of the electrode with the current collector

and also account to superior capacities and cycle stability. Besides CMC, GG and pectin

were evaluated as binder to prepare LTO electrodes. The influence of the molecular

structure of each binder was correlated to the electrode adhesion, as the linear backbone

102

of CMC and pectin grant higher values of adhesion force than GG branched polymer.

Nevertheless, the electrode adhesion is not apparently a significant influence on the

electrochemical performance of the electrodes. Electrodes prepared using GG and pectin

show slight superior capacities that CMC-based electrodes. As a proof of concept, a full

lithium-ion cell manufactured using only electrodes made via aqueous slurries, NMC

cathodes and LTO anodes using CMC binder and PA was prepared and it displays a

stable discharge capacity of ~ 120 mAh g-1. This is a good indication of the feasibility

of new binders for use in LIB, as an option for PVDF currently used in presence of toxic

NMP solvent.

The formation of binder film on the surface of active materials particles in NMC and

graphite electrodes, was analysed using XPS experiments. The results show that the low

amount of binder, only 2 wt.% of CMC or GG, in the graphite electrodes did not reveal

any difference between the formation of a binder film on the graphite particle surface.

Nevertheless, in NMC cathodes using 5 wt.% of binder in the formulation, the XPS

analysis demonstrated that using GG as binder, a much thinner binder layer covers the

active material compared when CMC is used. This presence of a thicker binder film on

the surface of the active material may hinder the electrochemical performance of the

electrode by reducing the direct contact of the electrolyte and the active material.

Moreover, the influence of using GG as binder in graphite half-cells and in full lithium-

ion cells was evaluated and compared to CMC based electrodes. The post-mortem

analyses by XPS reveals that a thinner SEI is formed on the graphite electrode upon

cycling when GG is used instead of CMC.

Overall, in this thesis the development and the electrochemical and physicochemical

characterization of the battery components prepared only by aqueous processes were

successful completed. Furthermore, this study shows some examples on how LIB can

be improved, in terms of environmentally friendless and use of less toxic materials. Once

more, it confirms the superiority of LIB for high energy density applications, as mobility,

and shows that nontoxic, cheaper material from natural sources can be used without

compromising its performance.

103

8.1. Bibliography

[1] R. Lal, Soil carbon sequestration to mitigate climate change, Geoderma. 123 (2004) 1–22. doi:10.1016/j.geoderma.2004.01.032.

[2] T. Boden, B. Andres, G. Marland, Global, Regional, and National Fossil-Fuel CO2 Emissions. Carbon Dioxide Information Analysis Center, Carbon Dioxide Information Analysis Center (CDIAC), Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States), 2016. doi:10.3334/CDIAC/00001_V2016.

[3] I.. Dresselhaus, M.S.; Thomas, Alternative energy technologies, Nature. 414 (2001) 332–337. doi:10.1038/35104599.

[4] J.B. Goodenough, Evolution of Strategies for Modern Rechargeable Batteries, Acc. Chem. Res. 46 (2013) 1053–1061. doi:10.1021/ar2002705.

[5] B.B.F. Wittneben, The impact of the Fukushima nuclear accident on European energy policy, Environ. Sci. Policy. 15 (2012) 1–3. doi:10.1016/j.envsci.2011.09.002.

[6] B. Scrosati, J. Garche, Lithium batteries: Status, prospects and future, J. Power Sources. 195 (2010) 2419–2430. doi:10.1016/j.jpowsour.2009.11.048.

[7] M. Whittingham, Materials challenges facing electrical energy storage, Mrs Bull. 33 (2008) 411–421. doi:10.1557/mrs2008.82.

[8] J.M. Tarascon, M. Armand, Issues and challenges facing rechargeable lithium batteries., Nature. 414 (2001) 359–67. doi:10.1038/35104644.

[9] M. Armand, J.-M. Tarascon, Building better batteries., Nature. 451 (2008) 652–657. doi:10.1038/451652a.

[10] D.L. Wood, J. Li, C. Daniel, Prospects for reducing the processing cost of lithium ion batteries, J. Power Sources. 275 (2015). doi:10.1016/j.jpowsour.2014.11.019.

[11] M. Winter, R.J. Brodd, What are batteries, fuel cells, and supercapacitors?, Chem. Rev. 104 (2004) 4245–4269. doi:10.1021/cr020730k.

[12] EPEC, Battery Comparison of Energy Density, (2017). http://www.epectec.com/batteries/cell-comparison.html (accessed November 27, 2017).

[13] Y. Nishi, The development of lithium ion secondary batteries, Chem. Rec. 1 (2001) 406–413. doi:10.1002/tcr.1024.

[14] N. Nitta, F. Wu, J.T. Lee, G. Yushin, Li-ion battery materials: Present and future, Mater. Today. 18 (2015) 252–264. doi:10.1016/j.mattod.2014.10.040.

[15] D. Larcher, J.-M. Tarascon, Towards greener and more sustainable batteries for electrical energy storage., Nat. Chem. 7 (2015) 19–29. doi:10.1038/nchem.2085.

[16] B. Scrosati, Recent advances in lithium ion battery materials, Electrochim. Acta. 45 (2000) 2461–2466. doi:10.1016/S0013-4686(00)00333-9.

[17] N. Yabuuchi, K. Kubota, M. Dahbi, S. Komaba, Research development on sodium-ion batteries, Chem. Rev. 114 (2014) 11636–11682. doi:10.1021/cr500192f.

104

[18] E. Levi, YGofer, D. Aurbach, On the way to rechargeable Mg batteries: The challenge of new cathode materials, Chem. Mater. 22 (2010) 860–868. doi:10.1021/cm9016497.

[19] B.G. Pollet, I. Staffell, J.L. Shang, Current status of hybrid, battery and fuel cell electric vehicles: From electrochemistry to market prospects, Electrochim. Acta. 84 (2012) 235–249. doi:10.1016/j.electacta.2012.03.172.

[20] S. Goriparti, E. Miele, F. De Angelis, E. Di Fabrizio, R. Proietti Zaccaria, C. Capiglia, Review on recent progress of nanostructured anode materials for Li-ion batteries, J. Power Sources. 257 (2014) 421–443. doi:10.1016/j.jpowsour.2013.11.103.

[21] B. Lestriez, Functions of polymers in composite electrodes of lithium ion batteries, Comptes Rendus Chim. 13 (2010) 1341–1350. doi:10.1016/j.crci.2010.01.018.

[22] J. Shim, K.A. Striebel, Effect of electrode density on cycle performance and irreversible capacity loss for natural graphite anode in lithium-ion batteries, J. Power Sources. 119–121 (2003) 934–937. doi:10.1016/S0378-7753(03)00235-0.

[23] M. Winter, J.O. Besenhard, Lithiated Carbons, Handb. Batter. Mater. (2007) 383–418. doi:10.1002/9783527611676.ch17.

[24] W. Xu, J. Wang, F. Ding, X. Chen, E. Nasybulin, Y. Zhang, et al., Lithium metal anodes for rechargeable batteries, Energy Environ. Sci. 7 (2014) 513–537. doi:10.1039/C3EE40795K.

[25] G.E. Blomgren, The Development and Future of Lithium Ion Batteries, J. Electrochem. Soc. 164 (2017) A5019–A5025. doi:10.1149/2.0251701jes.

[26] N. Loeffler, D. Bresser, S. Passerini, Secondary Lithium-Ion Battery Anodes : From First Commercial Batteries to Recent Research Activities Addressing the challenges in rechargeable lithium-ion battery technologies, Johnson Matthey Technol. Rev. 59 (2015) 34–44. doi:http://dx.doi.org/10.1595/205651314X685824.

[27] V. Sharova, A. Moretti, G. Giffin, D. Carvalho, S. Passerini, Evaluation of Carbon-Coated Graphite as a Negative Electrode Material for Li-Ion Batteries, C. 3 (2017) 22. doi:10.3390/c3030022.

[28] N. Loeffler, J. Von Zamory, N. Laszczynski, I. Doberdo, G.T. Kim, S. Passerini, Performance of LiNi1/3Mn1/3Co1/3O2/graphite batteries based on aqueous binder, J. Power Sources. 248 (2014) 915–922. doi:10.1016/j.jpowsour.2013.10.018.

[29] N. Loeffler, T. Kopel, G.-T. Kim, S. Passerini, Polyurethane Binder for Aqueous Processing of Li-Ion Battery Electrodes, J. Electrochem. Soc. 162 (2015) A2692–A2698. doi:10.1149/2.0641514jes.

[30] N. Cuesta, A. Ramos, I. Cameán, C. Antuna, A.B. García, Hydrocolloids as binders for graphite anodes of lithium-ion batteries, Electrochim. Acta. 155 (2015) 140–147. doi:10.1016/j.electacta.2014.12.122.

[31] V.A. Agubra, J.W. Fergus, The formation and stability of the solid electrolyte interface on the graphite anode, J. Power Sources. 268 (2014) 153–162. doi:10.1016/j.jpowsour.2014.06.024.

[32] D.V. Carvalho, N. Loeffler, G.T. Kim, M. Marinaro, M. Wohlfahrt-Mehrens, S. Passerini, Study of water-based lithium titanate electrode processing: The role of pH and binder molecular structure, Polymers (Basel). 8 (2016). doi:10.3390/polym8080276.

[33] S. Passerini, B. Scrosati, Lithium and Lithium-Ion Batteries: Challenges and Prospects, Interface Mag. 25 (2016) 85–87. doi:10.1149/2.F09163if.

[34] M.N. Obrovac, V.L. Chevrier, Alloy Negative Electrodes for Li-Ion Batteries, Chem. Rev. 114 (2014) 11444–11502. doi:10.1021/cr500207g.

[35] H. Wu, Y. Cui, Designing nanostructured Si anodes for high energy lithium ion batteries, Nano Today. 7 (2012) 414–429. doi:10.1016/j.nantod.2012.08.004.

105

[36] F. Mueller, D. Bresser, V.S.K. Chakravadhanula, S. Passerini, Fe-doped SnO2 nanoparticles as new high capacity anode material for secondary lithium-ion batteries, J. Power Sources. 299 (2015) 398–402. doi:10.1016/j.jpowsour.2015.08.018.

[37] A. Varzi, D. Bresser, J. von Zamory, F. Müller, S. Passerini, ZnFe2O4-C/LiFePO4-CNT: A Novel High-Power Lithium-Ion Battery with Excellent Cycling Performance, Adv. Energy Mater. 4 (2014) 1400054. doi:10.1002/aenm.201400054.

[38] W.R. Liu, Z.Z. Guo, W.S. Young, D.T. Shieh, H.C. Wu, M.H. Yang, et al., Effect of electrode structure on performance of Si anode in Li-ion batteries: Si particle size and conductive additive, J. Power Sources. 140 (2005) 139–144. doi:10.1016/j.jpowsour.2004.07.032.

[39] Y.K. Jeong, T. Kwon, I. Lee, T.-S. Kim, A. Coskun, J.W. Choi, Millipede-inspired structural design principle for high performance polysaccharide binders in silicon anodes, Energy Environ. Sci. 8 (2015) 1224–1230. doi:10.1039/C5EE00239G.

[40] S. Komaba, K. Shimomura, N. Yabuuchi, T. Ozeki, H. Yui, K. Konno, Study on polymer binders for high-capacity SiO negative electrode of Li-Ion batteries, J. Phys. Chem. C. 115 (2011) 13487–13495. doi:10.1021/jp201691g.

[41] T.D. Hatchard, P. Bissonnette, M.N. Obrovac, Phenolic Resin as an Inexpensive High Performance Binder for Li-Ion Battery Alloy Negative Electrodes, J. Electrochem. Soc. 163 (2016) 2035–2039. doi:10.1149/2.1121609jes.

[42] I. Kovalenko, B. Zdyrko, A. Magasinski, B. Hertzberg, Z. Milicev, R. Burtovyy, et al., A Major Constituent of Brown Algae for Use in High-Capacity Li-Ion Batteries, Science (80-. ). 334 (2011) 75–79. doi:10.1126/science.1209150.

[43] G. Seckmeyer, A. Zittermann, R. Mckenzie, Encyclopedia of Sustainability Science and Technology, Springer New York, New York, NY, 2012. doi:10.1007/978-1-4419-0851-3.

[44] M. Lazzari, B. Scrosati, A Cyclable Lithium Organic Electrolyte Cell Based on Two Intercalation Electrodes, J. Electrochem. Soc. 127 (1980) 773. doi:10.1149/1.2129753.

[45] M.S. Whittingham, Lithium batteries and cathode materials, Chem. Rev. 104 (2004) 4271–4301. doi:10.1021/cr020731c.

[46] C. Daniel, D. Mohanty, J. Li, D.L. Wood, Cathode materials review, AIP Conf. Proc. 1597 (2014) 26–43. doi:10.1063/1.4878478.

[47] A. Rougier, P. Granereau, C. Delmas, Optimization of the Composition of the Li1-zNi1+z02 Electrode Material: Structural, Magnetic, and Electrochemical Studies, J. Electrochem. Soc. 143 (1996) 1168–1175.

[48] C.M. Julien, A. Mauger, K. Zaghib, H. Groult, Comparative Issues of Cathode Materials for Li-Ion Batteries, Inorganics. 2 (2014) 132–154. doi:10.3390/inorganics2020132.

[49] D. Andre, S.-J. Kim, P. Lamp, S.F. Lux, F. Maglia, O. Paschos, et al., Future generations of cathode materials: an automotive industry perspective, J. Mater. Chem. A. 3 (2015) 6709–6732. doi:10.1039/C5TA00361J.

[50] S.-T. Myung, F. Maglia, K.-J. Park, C.S. Yoon, P. Lamp, S.-J. Kim, et al., Nickel-Rich Layered Cathode Materials for Automotive Lithium-Ion Batteries: Achievements and Perspectives, ACS Energy Lett. 2 (2017) 196–223. doi:10.1021/acsenergylett.6b00594.

[51] J.K. Ngala, N.A. Chernova, M. Ma, M. Mamak, P.Y. Zavalij, M.S. Whittingham, The synthesis , characterization and electrochemical behavior of the layered LiNi0.4Mn0.4Co0.2O2 compound, (2004). doi:10.1149/2.1731712jes.

[52] H.J. Noh, S. Youn, C.S. Yoon, Y.K. Sun, Comparison of the structural and electrochemical properties of layered Li[NixCoyMnz]O2 (x = 1/3, 0.5, 0.6, 0.7, 0.8 and 0.85) cathode material for lithium-ion batteries, J. Power Sources. 233 (2013) 121–130. doi:10.1016/j.jpowsour.2013.01.063.

106

[53] Z. Chen, G.-T. Kim, D. Chao, N. Loeffler, M. Copley, J. Lin, et al., Toward greener lithium-ion batteries: Aqueous binder-based LiNi0.4Co0.2Mn0.4O2 cathode material with superior electrochemical performance, J. Power Sources. 372 (2017) 180–187. doi:10.1016/j.jpowsour.2017.10.074.

[54] S. Cui, Y. Wei, T. Liu, W. Deng, Z. Hu, Y. Su, et al., Optimized Temperature Effect of Li-Ion Diffusion with Layer Distance in Li(NixMnyCoz)O2 Cathode Materials for High Performance Li-Ion Battery, Adv. Energy Mater. 6 (2016) 1501309. doi:10.1002/aenm.201501309.

[55] N. Laszczynski, J. von Zamory, N. Loeffler, G.B. Cho, G.T. Kim, S. Passerini, Synthesis of LiMn2O4 with Outstanding Lithium-Insertion Kinetics and Long-Term Stability, ChemElectroChem. 1 (2014) 1537–1542. doi:10.1002/celc.201402057.

[56] X. Li, Y. Xu, C. Wang, Suppression of Jahn-Teller distortion of spinel LiMn2O4 cathode, J. Alloys Compd. 479 (2009) 310–313. doi:10.1016/j.jallcom.2008.12.081.

[57] J.B. Goodenough, K.S. Park, The Li-ion rechargeable battery: A perspective, J. Am. Chem. Soc. 135 (2013) 1167–1176. doi:10.1021/ja3091438.

[58] M. Kuenzel, D. Bresser, T. Diemant, D.V. Carvalho, G.-T. Kim, J. Behm, et al., Complementary strategies towards the aqueous processing of high-voltage LiNi0.5Mn1.5O4 lithium-ion cathodes: Crosslinking CMC and phosphoric acid treatment, ChemSusChem. (2017) 0–32. doi:10.1002/cssc.201702021.

[59] A.K. Padhi, K.S. Nanjundaswamy, J.B. Goodenough, Phospho-olivines as Positive-Electrode Materials for Rechargeable Lithium Batteries, J. Electrochem. Soc. 144 (1997) 1188–1194.

[60] A. Yamada, S.C. Chung, K. Hinokuma, Optimized LiFePO4 for Lithium Battery Cathodes, J. Electrochem. Soc. 148 (2001) A224. doi:10.1149/1.1348257.

[61] Y. Cui, X. Zhao, R. Guo, Enhanced electrochemical properties of LiFePO4 cathode material by CuO and carbon co-coating, J. Alloys Compd. 490 (2010) 236–240. doi:10.1016/j.jallcom.2009.09.165.

[62] Z.-X. Chi, W. Zhang, F.-Q. Cheng, J.-T. Chen, A.-M. Cao, L.-J. Wan, Optimizing the carbon coating on LiFePO4 for improved battery performance, RSC Adv. 4 (2014) 7795. doi:10.1039/c3ra47702a.

[63] S. Yang, X. Zhou, J. Zhang, Z. Liu, Morphology-controlled solvothermal synthesis of LiFePO4 as a cathode material for lithium-ion batteries, J. Mater. Chem. 20 (2010) 8086. doi:10.1039/c0jm01346c.

[64] M. Park, X. Zhang, M. Chung, G.B. Less, A.M. Sastry, A review of conduction phenomena in Li-ion batteries, J. Power Sources. 195 (2010) 7904–7929. doi:10.1016/j.jpowsour.2010.06.060.

[65] A. Ochel, D. Di Lecce, C. Wolff, G.T. Kim, D.V. Carvalho, S. Passerini, Physicochemical and electrochemical investigations of the ionic liquid N-butyl -N-methyl-pyrrolidinium 4,5-dicyano-2-(trifluoromethyl)imidazole, Electrochim. Acta. 232 (2017) 586–595. doi:10.1016/j.electacta.2017.02.141.

[66] K. Xu, Nonaqueous Liquid Electrolytes for Lithium-Based Rechargeable Batteries, Chem. Rev. 104 (2004) 4303–4418. doi:10.1021/cr030203g.

[67] G.E. Blomgren, Electrolytes for advanced batteries, J. Power Sources. 81–82 (1999) 112–118. doi:10.1016/S0378-7753(99)00188-3.

[68] P. Verma, P. Maire, P. Novak, A review of the features and analyses of the solid electrolyte interphase in Li-ion batteries, Electrochim. Acta. 55 (2010) 6332–6341. doi:10.1016/j.electacta.2010.05.072.

[69] R. Fong, Studies of Lithium Intercalation into Carbons Using Nonaqueous Electrochemical Cells, J. Electrochem. Soc. 137 (1990) 2009. doi:10.1149/1.2086855.

107

[70] L. El Ouatani, R. Dedryvère, J.B. Ledeuil, C. Siret, P. Biensan, J. Desbrières, et al., Surface film formation on a carbonaceous electrode: Influence of the binder chemistry, J. Power Sources. 189 (2009) 72–80. doi:10.1016/j.jpowsour.2008.11.031.

[71] L. Madec, J. Xia, R. Petibon, K.J. Nelson, J.-P. Sun, I.G. Hill, et al., Effect of Sulfate Electrolyte Additives on LiNi1/3Mn1/3Co1/3O2/Graphite Pouch Cell Lifetime: Correlation between XPS Surface Studies and Electrochemical Test Results, J. Phys. Chem. C. 118 (2014) 29608–29622. doi:10.1021/jp509731y.

[72] Q. Li, G. Li, C. Fu, D. Luo, J. Fan, J. Zheng, et al., A study on storage characteristics of pristine Li-rich layered oxide Li1.20Mn0.54Co0.13Ni0.13O2: Effect of storage temperature and duration, Electrochim. Acta. 154 (2015) 249–258. doi:10.1016/j.electacta.2014.12.071.

[73] K. Edström, T. Gustafsson, J.O. Thomas, The cathode-electrolyte interface in the Li-ion battery, Electrochim. Acta. 50 (2004) 397–403. doi:10.1016/j.electacta.2004.03.049.

[74] A. Friesen, X. Mönnighoff, M. Börner, J. Haetge, F.M. Schappacher, M. Winter, Influence of temperature on the aging behavior of 18650-type lithium ion cells: A comprehensive approach combining electrochemical characterization and post-mortem analysis, J. Power Sources. 342 (2017) 88–97. doi:10.1016/j.jpowsour.2016.12.040.

[75] G.E. Blomgren, Liquid electrolytes for lithium and lithium-ion batteries, J. Power Sources. 119–121 (2003) 326–329. doi:10.1016/S0378-7753(03)00147-2.

[76] M.S. Ding, K. Xu, S. Zhang, T.R. Jow, Liquid/Solid Phase Diagrams of Binary Carbonates for Lithium Batteries Part II, J. Electrochem. Soc. 148 (2001) A299. doi:10.1149/1.1353568.

[77] D. Aurbach, Y. Talyosef, B. Markovsky, E. Markevich, E. Zinigrad, L. Asraf, et al., Design of electrolyte solutions for Li and Li-ion batteries: A review, Electrochim. Acta. 50 (2004) 247–254. doi:10.1016/j.electacta.2004.01.090.

[78] C. Iwakura, Y. Fukumoto, H. Inoue, S. Ohashi, S. Kobayashi, H. Tada, et al., Electrochemical characterization of various metal foils as a current collector of positive electrode for rechargeable lithium batteries, J. Power Sources. 68 (1997) 301–303. doi:10.1016/S0378-7753(97)02538-X.

[79] S.S. Zhang, A review on electrolyte additives for lithium-ion batteries, J. Power Sources. 162 (2006) 1379–1394. doi:10.1016/j.jpowsour.2006.07.074.

[80] L. Madec, R. Petibon, J. Xia, J.-P. Sun, I.G. Hill, J.R. Dahn, Understanding the Role of Prop-1-ene-1,3-Sultone and Vinylene Carbonate in LiNi1/3Mn1/3Co1/3O2/Graphite Pouch Cells: Electrochemical, GC-MS and XPS Analysis, J. Electrochem. Soc. 162 (2015) A2635–A2645. doi:10.1149/2.0741512jes.

[81] V. Sharova, A. Moretti, T. Diemant, A. Varzi, R.J. Behm, S. Passerini, Comparative study of imide-based Li salts as electrolyte additives for Li-ion batteries, J. Power Sources. 375 (2018) 43–52. doi:10.1016/j.jpowsour.2017.11.045.

[82] A. Birrozzi, N. Laszczynski, M. Hekmatfar, J. von Zamory, G.A. Giffin, S. Passerini, Beneficial effect of propane sultone and tris(trimethylsilyl) borate as electrolyte additives on the cycling stability of the lithium rich nickel manganese cobalt (NMC) oxide, J. Power Sources. 325 (2016) 525–533. doi:10.1016/j.jpowsour.2016.06.054.

[83] E. Simonetti, G. Maresca, G.B. Appetecchi, G.T. Kim, N. Loeffler, S. Passerini, Towards Li(Ni0.33Mn0.33Co0.33)O2/graphite batteries with ionic liquid-based electrolytes. I. Electrodes’ behavior in lithium half-cells, J. Power Sources. 331 (2016) 426–434. doi:10.1016/j.jpowsour.2016.09.078.

[84] G.T. Kim, G.B. Appetecchi, M. Carewska, M. Joost, A. Balducci, M. Winter, et al., UV cross-linked, lithium-conducting ternary polymer electrolytes containing ionic liquids, J. Power Sources. 195 (2010) 6130–6137. doi:10.1016/j.jpowsour.2009.10.079.

108

[85] J.-H. Cho, J.-H. Park, M.-H. Lee, H.-K. Song, S.-Y. Lee, A polymer electrolyte-skinned active material strategy toward high-voltage lithium ion batteries: a polyimide-coated LiNi0.5Mn1.5O4 spinel cathode material case, Energy Environ. Sci. 5 (2012) 7124. doi:10.1039/c2ee03389e.

[86] M. Keller, G.B. Appetecchi, G.T. Kim, V. Sharova, M. Schneider, J. Schuhmacher, et al., Electrochemical performance of a solvent-free hybrid ceramic-polymer electrolyte based on Li7La3Zr2O12 in P(EO)15LiTFSI, J. Power Sources. 353 (2017) 287–297. doi:10.1016/j.jpowsour.2017.04.014.

[87] D. Bresser, S. Passerini, B. Scrosati, Recent progress and remaining challenges in sulfur-based lithium secondary batteries – a review, Chem. Commun. 49 (2013) 10545. doi:10.1039/c3cc46131a.

[88] P. Arora, Z. Zhang, Battery separators, Chem. Rev. 104 (2004) 4419–4462. doi:10.1021/cr020738u.

[89] S.S. Zhang, A review on the separators of liquid electrolyte Li-ion batteries, J. Power Sources. 164 (2007) 351–364. doi:10.1016/j.jpowsour.2006.10.065.

[90] C.J. Weber, S. Geiger, S. Falusi, M. Roth, Material review of Li ion battery separators, AIP Conf. Proc. 66 (2014) 66–81. doi:10.1063/1.4878480.

[91] K.M. Abraham, Directions in secondary lithium battery research and development, Electrochim. Acta. 38 (1993) 1233–1248. doi:10.1016/0013-4686(93)80054-4.

[92] D.V. Carvalho, N. Loeffler, G.T. Kim, S. Passerini, High temperature stable separator for lithium batteries based on SiO2 and hydroxypropyl guar gum, Membranes (Basel). 5 (2015) 632–645. doi:10.3390/membranes5040632.

[93] H. Lee, M. Yanilmaz, O. Toprakci, K. Fu, X. Zhang, A review of recent developments in membrane separators for rechargeable lithium-ion batteries, Energy Environ. Sci. 7 (2014) 3857–3886. doi:10.1039/C4EE01432D.

[94] X. Huang, Separator technologies for lithium-ion batteries, J. Solid State Electrochem. 15 (2011) 649–662. doi:10.1007/s10008-010-1264-9.

[95] J. Chen, S. Wang, D. Cai, H. Wang, Porous SiO2 as a separator to improve the electrochemical performance of spinel LiMn2O4 cathode, J. Memb. Sci. 449 (2013) 169–175. doi:10.1016/j.memsci.2013.08.028.

[96] S. Augustin, V. Hennige, G. Hörpel, C. Hying, Ceramic but flexible: new ceramic membrane foils for fuel cells and batteries, Desalination. 146 (2002) 23–28. doi:10.1016/S0011-9164(02)00465-4.

[97] W. Jiang, Z. Liu, Q. Kong, J. Yao, C. Zhang, P. Han, et al., A high temperature operating nanofibrous polyimide separator in Li-ion battery, Solid State Ionics. 232 (2013) 44–48. doi:10.1016/j.ssi.2012.11.010.

[98] J.H. Kim, J.H. Kim, E.S. Choi, H.K. Yu, J.H. Kim, Q. Wu, et al., Colloidal silica nanoparticle-assisted structural control of cellulose nanofiber paper separators for lithium-ion batteries, J. Power Sources. 242 (2013) 533–540. doi:10.1016/j.jpowsour.2013.05.142.

[99] S.-J. Chun, E.-S. Choi, E.-H. Lee, J.H. Kim, S.-Y. Lee, S.-Y. Lee, Eco-friendly cellulose nanofiber paper-derived separator membranes featuring tunable nanoporous network channels for lithium-ion batteries, J. Mater. Chem. 22 (2012) 16618. doi:10.1039/c2jm32415f.

[100] H. Lee, M. Yanilmaz, O. Toprakci, K. Fu, X. Zhang, A review of recent developments in membrane separators for rechargeable lithium-ion batteries, Energy Environ. Sci. 7 (2014) 3857–3886. doi:10.1039/C4EE01432D.

[101] S. Chou, Y. Pan, J. Wang, H. Liu, S. Dou, Small things make a big difference: binder effects on the performance of Li and Na batteries., Phys. Chem. Chem. Phys. 16 (2014) 20347–20359. doi:10.1039/c4cp02475c.

109

[102] B. Lestriez, Functions of polymers in composite electrodes of lithium ion batteries, Comptes Rendus Chim. 13 (2010) 1341–1350. doi:10.1016/j.crci.2010.01.018.

[103] Q. Wu, S. Ha, J. Prakash, D.W. Dees, W. Lu, Investigations on high energy lithium-ion batteries with aqueous binder, Electrochim. Acta. 114 (2013) 1–6. doi:10.1016/j.electacta.2013.09.068.

[104] Y.S. Park, E.S. Oh, S.M. Lee, Effect of polymeric binder type on the thermal stability and tolerance to roll-pressing of spherical natural graphite anodes for Li-ion batteries, J. Power Sources. 248 (2014) 1191–1196. doi:10.1016/j.jpowsour.2013.10.076.

[105] N. Loeffler, G.-T. Kim, F. Mueller, T. Diemant, J.-K. Kim, R.J. Behm, et al., In Situ Coating of Li[Ni0.33Mn0.33Co0.33]O2 Particles to Enable Aqueous Electrode Processing, ChemSusChem. 9 (2016) tbd. doi:10.1002/cssc.201600353.

[106] C.C. Li, J.T. Lee, Y.L. Tung, C.R. Yang, Effects of pH on the dispersion and cell performance of LiCoO2 cathodes based on the aqueous process, J. Mater. Sci. 42 (2007) 5773–5777. doi:10.1007/s10853-006-1172-7.

[107] I. Doberdo, N. Loeffler, N. Laszczynski, D. Cericola, N. Penazzi, S. Bodoardo, et al., Enabling aqueous binders for lithium battery cathodes - Carbon coating of aluminum current collector, J. Power Sources. 248 (2014) 1000–1006. doi:10.1016/j.jpowsour.2013.10.039.

[108] G.-T. Kim, N. Löffler, I. Doberdo, N. Laszczynski, D. Bresser, S. Passerini, Method for Producing an Electrode for a Lithium-Ion Battery, WO 2015/062985 A1, 2015. WO 2015/062985 A1.

[109] J. Li, C. Rulison, J. Kiggans, C. Daniel, D.L. Wood, Superior Performance of LiFePO4 Aqueous Dispersions via Corona Treatment and Surface Energy Optimization, J. Electrochem. Soc. 159 (2012) A1152–A1157. doi:10.1149/2.018208jes.

[110] Z. Du, K.M. Rollag, J. Li, S.J. An, M. Wood, Y. Sheng, et al., Enabling aqueous processing for crack-free thick electrodes, J. Power Sources. 354 (2017) 200–206. doi:10.1016/j.jpowsour.2017.04.030.

[111] Z. Du, K.M. Rollag, J. Li, S.J. An, M. Wood, Y. Sheng, et al., Enabling aqueous processing for crack-free thick electrodes, J. Power Sources. 354 (2017) 200–206. doi:10.1016/j.jpowsour.2017.04.030.

[112] C.-C. Li, J.-T. Lee, X.-W. Peng, Improvements of Dispersion Homogeneity and Cell Performance of Aqueous-Processed LiCoO2 Cathodes by Using Dispersant of PAA–NH4, J. Electrochem. Soc. 153 (2006) A809. doi:10.1149/1.2177071.

[113] W. Porcher, B. Lestriez, S. Jouanneau, D. Guyomard, Design of Aqueous Processed Thick LiFePO4 Composite Electrodes for High-Energy Lithium Battery, J. Electrochem. Soc. 156 (2009) A133. doi:10.1149/1.3046129.

[114] W. Porcher, B. Lestriez, S. Jouanneau, D. Guyomard, Optimizing the surfactant for the aqueous processing of LiFePO4 composite electrodes, J. Power Sources. 195 (2010) 2835–2843. doi:10.1016/j.jpowsour.2009.11.088.

[115] M. Müller, L. Pfaffmann, S. Jaiser, M. Baunach, V. Trouillet, F. Scheiba, et al., Investigation of binder distribution in graphite anodes for lithium-ion batteries, J. Power Sources. 340 (2017) 1–5. doi:10.1016/j.jpowsour.2016.11.051.

[116] S.F. Lux, F. Schappacher, A. Balducci, S. Passerini, M. Winter, Low Cost, Environmentally Benign Binders for Lithium-Ion Batteries, J. Electrochem. Soc. 157 (2010) A320. doi:10.1149/1.3291976.

[117] J.-H. Lee, U. Paik, V.A. Hackley, Y.-M. Choi, Effect of Carboxymethyl Cellulose on Aqueous Processing of Natural Graphite Negative Electrodes and their Electrochemical Performance for Lithium Batteries, J. Electrochem. Soc. 152 (2005) A1763. doi:10.1149/1.1979214.

110

[118] S. Urbonaite, I. Baglien, D. Ensling, K. Edström, Effect of ethanol-assisted electrode fabrication on the performance of silicon anodes, J. Power Sources. 195 (2010) 5370–5373. doi:10.1016/j.jpowsour.2010.03.042.

[119] T. Zhang, J. Li, J. Liu, Y. Deng, Z. Wu, Z. Yin, et al., Suppressing the voltage-fading of layered lithium-rich cathode materials via an aqueous binder for Li-ion batteries, Chem. Commun. 52 (2016) 4683–4686. doi:10.1039/C5CC10534J.

[120] Y.-Q. Lu, J.-T. Li, X.-X. Peng, T. Zhang, Y.-P. Deng, Z.-Y. Wu, et al., Achieving high capacity retention in lithium-sulfur batteries with an aqueous binder, 2016. doi:10.1016/j.elecom.2016.09.004.

[121] D.-E. Yoon, C. Hwang, N. Kang, U. Lee, D. Ahn, J. Kim, et al., Dependency of electrochemical performances of silicon lithium ion batteries on glycosidic linkages of polysaccharide binders, ACS Appl. Mater. Interfaces. (2016) 4042–4047. doi:10.1021/acsami.5b11408.

[122] M.K. Dufficy, S. a. Khan, P.S. Fedkiw, Galactomannan binding agents for silicon anodes in Li-ion batteries, J. Mater. Chem. A. 3 (2015) 12023–12030. doi:10.1039/C5TA03126E.

[123] J. Liu, Q. Zhang, T. Zhang, J.-T. Li, L. Huang, S.-G. Sun, A Robust Ion-Conductive Biopolymer as a Binder for Si Anodes of Lithium-Ion Batteries, Adv. Funct. Mater. 25 (2015) 3599–3605. doi:10.1002/adfm.201500589.

[124] R. Kuruba, M.K. Datta, K. Damodaran, P.H. Jampani, B. Gattu, P.P. Patel, et al., Guar gum: Structural and electrochemical characterization of natural polymer based binder for silicon-carbon composite rechargeable Li-ion battery anodes, J. Power Sources. 298 (2015) 331–340. doi:10.1016/j.jpowsour.2015.07.102.

[125] T. Waldmann, A. Iturrondobeitia, M. Kasper, N. Ghanbari, E. Bekaert, L. Daniel, et al., Review — Post-Mortem Analysis of Aged Lithium-Ion Batteries : Disassembly Methodology and Physico-Chemical Analysis Techniques, 163 (2016) 50–52. doi:10.1149/2.1211609jes.

[126] A. Mettler-Toledo GmbH, A Guide to pH Measurement, (2013) 1–52.

[127] W.K. Shin, D.W. Kim, High performance ceramic-coated separators prepared with lithium ion-containing SiO2 particles for lithium-ion batteries, J. Power Sources. 226 (2013) 54–60. doi:10.1016/j.jpowsour.2012.10.082.

[128] W.-S. Chang, C.-M. Park, J.-H. Kim, Y.-U. Kim, G. Jeong, H.-J. Sohn, Quartz (SiO2): a new energy storage anode material for Li-ion batteries, Energy Environ. Sci. 5 (2012) 6895. doi:10.1039/c2ee00003b.

[129] J. Tu, Y. Yuan, P. Zhan, H. Jiao, X. Wang, H. Zhu, et al., Straightforward Approach toward SiO2 Nanospheres and Their Superior Lithium Storage Performance, J. Phys. Chem. C. 118 (2014) 7357–7362. doi:10.1021/jp5011023.

[130] B.-R. Lee, S. -j. Kim, E.-S. Oh, Bio-Derivative Galactomannan Gum Binders for Li4Ti5O12 Negative Electrodes in Lithium-Ion Batteries, J. Electrochem. Soc. 161 (2014) A2128–A2132. doi:10.1149/2.0641414jes.

[131] Y.N. Sudhakar, M. Selvakumar, D.K. Bhat, Tubular array, dielectric, conductivity and electrochemical properties of biodegradable gel polymer electrolyte, Mater. Sci. Eng. B Solid-State Mater. Adv. Technol. 180 (2014) 12–19. doi:10.1016/j.mseb.2013.10.013.

[132] Y. Chen, X. Wu, X. Miao, J. Luo, B. Jiang, Determination of the degree of substitution of hydroxypropyl guar gum at C-6 by Pyrolysis-Gas Chromatography spectrometry, Carbohydr. Polym. 82 (2010) 829–832. doi:10.1016/j.carbpol.2010.06.002.

[133] S. Tripathy, M.K. Das, Guar Gum: Present Status and Applications, J. Pharm. Sci. Innov. 4 (2013) 24–28. doi:10.7897/2277-4572.02447.

[134] D. Hotza, P. Greil, Review: aqueous tape casting of ceramic powders, Mater. Sci. Eng. A. 202 (1995) 206–217. doi:10.1016/0921-5093(95)09785-6.

111

[135] N.S. Hochgatterer, M.R. Schweiger, S. Koller, P.R. Raimann, T. Wöhrle, C. Wurm, et al., Silicon/Graphite Composite Electrodes for High-Capacity Anodes: Influence of Binder Chemistry on Cycling Stability, Electrochem. Solid-State Lett. 11 (2008) A76. doi:10.1149/1.2888173.

[136] X. Huang, J. Hitt, Lithium ion battery separators: Development and performance characterization of a composite membrane, J. Memb. Sci. 425–426 (2013) 163–168. doi:10.1016/j.memsci.2012.09.027.

[137] B.R. Nayak, R.P. Singh, Development of graft copolymer flocculating agents based on hydroxypropyl guar gum and acrylamide, J. Appl. Polym. Sci. 81 (2001) 1776–1785. doi:10.1002/app.1610.

[138] K.S. Soppirnath, T.M. Aminabhavi, Water transport and drug release study from cross-linked polyacrylamide grafted guar gum hydrogel microspheres for the controlled release application, Eur. J. Pharm. Biopharm. 53 (2002) 87–98. doi:10.1016/S0939-6411(01)00205-3.

[139] C.L. Beyler, M.M. Hirschler, Chapter 7 - Thermal Decomposition of Polymers, SFPE Handb. Fire Prot. Eng. 3rd Ed. (2005) 110–131.

[140] L.T. Zhuravlev, The surface chemistry of amorphous silica. Zhuravlev model, Colloids Surfaces A Physicochem. Eng. Asp. 173 (2000) 1–38. doi:10.1016/S0927-7757(00)00556-2.

[141] J. Fan, P.S. Fedkiw, Composite Electrolytes Prepared from Fumed Silica, Polyethylene Oxide Oligomers, and Lithium Salts, J. Electrochem. Soc. 144 (1997) 399–408.

[142] X. Zhang, W.J. Jiang, X.P. Zhu, A. Mauger, Qilu, C.M. Julien, Aging of LiNi1/3Mn1/3Co1/3O2 cathode material upon exposure to H2O, J. Power Sources. 196 (2011) 5102–5108. doi:10.1016/j.jpowsour.2011.02.009.

[143] A. Moretti, G.T. Kim, D. Bresser, K. Renger, E. Paillard, R. Marassi, et al., Investigation of different binding agents for nanocrystalline anatase TiO2 anodes and its application in a novel, green lithium-ion battery, J. Power Sources. 221 (2013) 419–426. doi:10.1016/j.jpowsour.2012.07.142.

[144] S. Tripathi, G.K. Mehrotra, P.K. Dutta, Preparation and physicochemical evaluation of chitosan/poly(vinyl alcohol)/pectin ternary film for food-packaging applications, Carbohydr. Polym. 79 (2010) 711–716. doi:10.1016/j.carbpol.2009.09.029.

[145] B.-R. Lee, S. -j. Kim, E.-S. Oh, Bio-Derivative Galactomannan Gum Binders for Li4Ti5O12 Negative Electrodes in Lithium-Ion Batteries, J. Electrochem. Soc. 161 (2014) A2128–A2132. doi:10.1149/2.0641414jes.

[146] W. Haselrieder, B. Westphal, H. Bockholt, A. Diener, S. Höft, A. Kwade, Measuring the coating adhesion strength of electrodes for lithium-ion batteries, Int. J. Adhes. Adhes. 60 (2015) 1–8. doi:10.1016/j.ijadhadh.2015.03.002.

[147] M. Murase, N. Yabuuchi, Z.J. Han, J.Y. Son, Y.T. Cui, H. Oji, et al., Crop-derived polysaccharides as binders for high-capacity silicon/graphite-based electrodes in lithium-ion batteries, ChemSusChem. 5 (2012) 2307–2311. doi:10.1002/cssc.201200650.

[148] D.V. Carvalho, N. Loeffler, M. Hekmatfar, A. Moretti, G.-T. Kim, S. Passerini, Evaluation of guar gum-based biopolymers as binders for lithium-ion batteries electrodes, Electrochim. Acta. 265 (2018) 89–97. doi:10.1016/j.electacta.2018.01.083.

112

8.2. List of abbreviations

1. CO2 – carbon dioxide

2. HEV – hybrid electric vehicle

3. EV – electric vehicle

4. Zn-MnO2 – Zinc-manganese oxide

5. Ni-Cd – Nickel-Cadmium

6. Ni-MH – Nickel-Metal hydride

7. LIB – Lithium-ion batteries

8. Na - Sodium

9. Mg – Magnesium

10. wt.% – percentage by weight

11. Li - Lithium

12. PC – Propylene carbonate

13. Li4Ti5O12 LTO – Lithium titanate

14. SEI – Solid electrolyte interface

15. Si - Silicon

16. LiCoO2 – LCO – Lithium cobalt oxide

17. Co - Cobalt

18. LiNiO2 – Lithium nickel oxide

19. Al - Aluminium

20. Li[Ni0.8Co0.15Al0.05]O2 – NCA – Lithium nickel cobalt aluminium oxide

21. Li[Ni1/3Mn1/3Co1/3]O2 - NMC-111 – Lithium nickel manganese cobalt oxide

22. Ni - Nickel

23. LiMn2O4 – LMO – Lithium manganese oxide

24. LiNi0.5Mn1.5O4 – Lithium nickel manganese oxide

25. LiFePO4 – LFP – Lithium iron phosphate

26. EC – Ethyl carbonate

27. DMC – dimethyl carbonate

28. DEC – diethyl carbonate

29. LiPF6 – Lithium hexafluorophosphate

30. LiAsF6 - Lithium hexafluoroarsenate

31. LiBF4 – Lithium tetrafluoroborate

32. LiClO4 – Lithium perchlorate

33. VC – Vinyl carbonate

34. LiTFISI - Lithium bis(trifluoromethanesulfonyl)imide

35. LiFSI - Lithium bis(fluorosulfonyl)imide

36. LFTFSI - Lithium fluorosulfonyl(trifluoromethanesulfonyl)imide

113

37. w - Weight

38. PE – Polyethylene

39. PP - Polypropylene

40. SiO2 – Silicon dioxide

41. PVDF - Poly(vinylidene difluoride)

42. PAN - Polyacrylonitrile

43. NMP – N-methyl-pyrrolidone

44. CMC – Sodium carboxymethylcellulose

45. SBR – Styrene-butadiene rubber

46. CC – Constant current

47. CV – Constant voltage

48. LSV – Linear sweep voltammetry

49. ESW – Electrochemical stability window

50. SEM – Scanning electron microscopy

51. EDX – Energy dispersive X-ray spectroscopy

52. TGA – Thermal gravimetric analysis

53. XPS – X-ray photoelectron spectroscopy

54. HPG - Hydroxypropyl guar gum

55. PTFE – Polytetrafluorethylene

56. WE – Working electrode

57. CE – Counter electrode

58. RE – Reference electrode

59. SE – Separator

60. ICE – Internal combustion engine

61. CPS – Counts per second

8.3. List of figures

Figure 1 – Estimation of global CO2 emissions per year. Prepared using data from reference [2]. ........... 1

Figure 2 – Energy comparison of different rechargeable batteries chemistries [12]. ................................. 4

Figure 3 - Schematic representation of the working principle of lithium-ion battery (from Ref. [20]). ..... 6

Figure 4 - Illustration of lithium-ion battery components. ......................................................................... 7

Figure 5 – Representation of lithium-ion batteries configuration; electrodes jelly roll in a) cylindrical and c) prismatic cell and stacked electrodes in b) coin and d) pouch cell [8]. ......................................... 8

114

Figure 6 – Single-electron reduction mechanism of EC (top) and DEC (bottom) [66]. .......................... 15

Figure 7 – Representation of the crack generation in the drying process. Reprinted from reference [111]. ......................................................................................................................................................... 22

Figure 8 – Schematic representation of drying process of gel and liquid-like slurry [102,113]. .............. 24

Figure 9 – Schematic representation of a galvanostatic charge/discharge test (Left) and typical discharge voltage profile (Right). .................................................................................................................... 31

Figure 10 – Schematic representation of a linear sweep voltammetry test (Left) and a typical result plot (Right). ............................................................................................................................................. 33

Figure 11 – Representation of the sample volume interaction of electron bean in a SEM measurement (Left panel) and representation of working principle of emission of characteristic X-ray due to high energetic electron beam (Right panel). ............................................................................................ 34

Figure 12 - TGA profile showing the weight loss by increasing the time (isothermal test) or temperature. ......................................................................................................................................................... 35

Figure 13 – Representation of emission of a characteristic electron by a monochromatic X-ray in a XPS experiment. ...................................................................................................................................... 37

Figure 14 – Representation of the adhesion strength measurement.......................................................... 44

Figure 15 – Schematic representation of cell´s configurations and the their electric circuit used for electrochemical tests; top: Swagelok®; below: pouch and coin-cell. RE – reference electrode (Li), CE - counter electrode (Li, LTO or graphite electrodes), WE – working electrode and SE-separator. ......................................................................................................................................................... 45

Figure 16 – Structure of hydroxypropyl guar gum. .................................................................................. 50

Figure 17 – a) SEM image of SiO2 – HPG membrane prepared using the ratio of 70:30 and b) picture of SiO2 – HPG membrane prepared using the ratio 90:10. .................................................................. 52

Figure 18 - Images of the self-standing SiO2–HPG membrane after drying and removal from the PTFE tape. The good flexibility and self-integrity of the membrane is well evidenced in the bottom image. ......................................................................................................................................................... 53

Figure 19 - SEM pictures of the SiO2 – HPG membrane. (a) Top face (air contact at drying step); (b) Bottom face (PTFE contact at drying step); (c) Cross section. ........................................................ 55

115

Figure 20 - Images of membranes subjected to the wettability test. The pictures were taken 2 s after dropping 80 µL of LP30 electrolyte on the (a) commercial polyethylene (PE) single layer separator and (b) SiO2 – HPG membrane. ....................................................................................................... 56

Figure 21 - TGA measurements of HPG, SiO2 particles and the SiO2 – HPG membrane in nitrogen (a) and oxygen (b) atmospheres. Temperature ramp of 10 °C min-1 up to 600 °C in both atmospheres under gas flow (25 mL min−1) using open Al pans. ................................................................................... 57

Figure 22 - Thermal stability test of SiO2 – HPG membrane at 180 °C for 12 h (a) TGA measurement; (b) dimensional-shrinkage test. ............................................................................................................. 59

Figure 23 - Linear sweep voltammograms of the SiO2 – HPG membrane and glass fiber separator using 1 mol L−1 (LiPF6) dissolved in EC:DMC (1:1 w/w) as electrolyte. Scan rate 0.1 mV s−1 at 20 °C. Three electrode cell, WE = Al disc (anodic scan) or Ni discs (cathodic scan) with Li metal as quasi-reference and CE electrode. ............................................................................................................. 60

Figure 24 - First charge and discharge profile at 0.1C using SiO2 – HPG membrane as separator in (a) NMC; (b) and LTO half-cells. Galvanostatic charge/discharge in pouch bag cells using LP30 electrolyte at 20 °C. NMC and LTO loading ~ 4 mg cm-2. NMC half-cells cycled between 3.0 – 4.3 V for C = 161 mA g-1 and LTO half-cells cycled between 1.0 - 2.5 V for C = 175 mA g-1. ..... 62

Figure 25 - Cycle performance at different C-rates using SiO2–HPG membrane as separator in (a) NMC; (b) and LTO half-cells. Galvanostatic charge/discharge in pouch bag cells using LP30 electrolyte at 20 °C. NMC and LTO loading ~ 4 mg cm-2. NMC half-cells cycled between 3.0 – 4.3 V for C = 161 mA g-1 and LTO half-cells cycled between 1.0 - 2.5 V for C = 175 mA g-1. .................................. 64

Figure 26 - TGA weight loss profiles of LTO, CMC, GG and pectin with a heating rate of 5 °C/min in N2 atmosphere. Heating rate of 5 °C min−1 from 30 °C up to 500 °C under a nitrogen gas flow (25 mL min−1) using open Al pans......................................................................................................... 71

Figure 27 - SEM images of unpressed LTO electrodes using: (a) CMC as the binder (LTO-CMC); (b) CMC as the binder and phosphoric acid (PA) (LTO-CMC-PA); (c) pectin as the binder and PA (LTO-pectin-PA); (d) and GG as the binder and PA (LTO-GG-PA). ............................................. 72

Figure 28 - pH evolution of the aqueous slurry containing only LTO active material (circles) and LTO and phosphoric acid (squares). Measured at 20 °C with 40 % of LTO in water. ............................ 73

Figure 29 - SEM image of the electrode prepared using CMC as binder without PA (LTO-CMC) showing three spots where EDX experiments were performed. .................................................................... 74

Figure 30 - Adhesion strength of LTO electrodes fabricated using PA as an additive and CMC, GG and pectin as the binder. ......................................................................................................................... 76

116

Figure 31 - Delivered discharge capacity of (a) LTO half-cells using CMC as the binder and CMC, pectin and GG as the binder and PA as an additive at several current densities; LTO mass loading: 3.6 - 4.1 mg·cm−2; electrolyte: LP30; and (b) cathode-limited Li[Ni0.33Mn0.33Co0.33]O2 (NMC)/LTO full-cell using CMC as the binder and PA as an additive at 1C; NMC mass loading: ~ 4.3 mg cm−2; electrolyte: LP30. ............................................................................................................................. 78

Figure 32 - Comparison first cycle voltage profile of graphite half-cells containing GG or CMC as binder (2 wt.%). Coin cells using GF/D membrane as separator and LP30 + VC as electrolyte cycled at 20 °C. Graphite loading ~ 6.3 mg cm-2. C = 372 mA g-1. ................................................................ 85

Figure 33 – Comparison of delithiation capacity upon cycling of graphite half-cells containing GG or CMC as binder (2 wt.%). a) rate-capability test and b) long-term cycling (cycle 30 to 100) and coulombic efficiency. Coin cells using GF/D membrane as separator and LP30 + VC as electrolyte cycled at 20 °C. Graphite loading ~ 6.3 mg cm-2. C = 372 mA g-1. ................................................. 86

Figure 34 - SEM images of uncalendered graphite electrodes prepared using 2 wt.% GG (left) 2 wt.% CMC (right). .................................................................................................................................... 87

Figure 35 - XPS analysis of pristine graphite electrodes prepared using 2 wt.% of GG or CMC. a) C 1s and b) O 1s binding energy. The dashed lines refer to C-C and C-H bonding at ca. 285 eV and C-O bonding at 287.8 eV. ........................................................................................................................ 88

Figure 36 - XPS results of pristine NMC cathodes containing 5 wt.% of GG or CMC as binder. The dashed lines refer to the Li 1s peak at ca. 55 eV, the Mn 3p3/2 one at ca. 49 eV and the Mn 3p1/2 one at ca. 51.5 eV. ........................................................................................................................................... 89

Figure 37 - Galvanostatic cycling performance of the NMC (3 wt.% GG)/Graphite (2 wt.%) Li-ion cell. Three electrode cell using GF/D membrane as separator and Li metal as quasi-reference electrode, NMC electrode (loading ~ 7 mg cm-2) as WE and graphite electrode as CE (loading ~ 2.8 mg cm-2) at 20 °C. Electrolyte LP30 + VC. C = 161 mA g-1. ......................................................................... 90

Figure 38 - Galvanostatic discharge capacity of Li-ion cells (NMC/Graphite) employing GG or CMC as binder. Cathode limited full cell. Three electrode cell using GF/D membrane as separator and Li metal as quasi-reference electrode and cycled at 20 °C; Electrolyte: LP30 + VC. C = 161 mA g-1. ......................................................................................................................................................... 91

Figure 39 - XPS results of cycled NMC electrodes containing 3 wt.% of GG and 5 wt.% CMC as binder of a) P 2p, b) C 1s c) O 1s and d) F 1s binding energy. ................................................................... 93

Figure 40 - XPS results of cycled graphite electrodes containing GG and CMC as binder. a) Li 1s b) P 2p c) C 1s d) O 1s e) F 1s. .................................................................................................................... 95

Figure 41 - XPS results of cycled graphite anodes containing GG and CMC as binder after 3 and 10 minutes of the etching process. a) Li 1s b) P 2p c) C 1s d) O 1s e) F 1s binding energy. ................ 96

117

Figure 42 - Concentration of LiF, LixPFy, LixPOyFz species by the XPS experiments on cycled graphite electrodes as obtained from the cell and after 3 and 10 min of etching. The values are normalized concerning the CMC species. .......................................................................................................... 97

8.4. List of tables

Table 1 – Basic characteristics of graphite, LTO and Si anodes [23,26,34]............................................. 11 Table 2 – Properties of commercial cathode materials for lithium-ion batteries [14,25,64]. ................... 13 Table 3- Description of cell production condition using low and high melting point separators. ............ 18 Table 4 – Properties of commercial LIB separators [88–90,94]. ............................................................. 20 Table 5 – Molecular structure, source and industrial use of natural source binders used in this thesis. ... 25 Table 6 - List of materials used in this thesis. .......................................................................................... 40 Table 7 - Summary of all cell configurations used in the electrochemical tests. ...................................... 47 Table 8 – List of electrodes used in this chapter. ..................................................................................... 69 Table 9 - Element fractions estimated by EDX measurements on electrode surface prepared without PA

(LTO-CMC). ................................................................................................................................... 74

118

119

Firstly, I would like to thank Prof. Dr. Stefano Passerini for giving me the possibility to

develop my work in his group. I am very thankful to have the chance to work in a very

professional and trustful atmosphere under his supervision. Besides the great knowledge

in the science, he was always friendly and a source of motivation.

Furthermore, I would like to say thank you to Prof. Dr. Arnulf Latz for his willingness

to be my second referee and his time and support for my Ph D admission process.

Moreover, I am grateful to the European Commission within the FP7 Projects Advanced

manufacturing processes for Low Cost Greener Li-Ion batteries (GREENLION Grant

agreement No. 285268) and HORIZON 2020 Project Silicon and polyanionic

chemistries and architectures of Li-ion cell for high energy battery (SPICY, Grant

Agreement No. 653373) for financial support.

I am very thankful to Dr. Guk-Tae Kim that introduces me to most of the experimental

procedures used in this thesis. He never hesitates when I needed his help to discuss and

teach me. His time, experience, methods and friendship were essential to development

of this thesis. I appreciated the time and valuable help of Dr. Nicholas Löffler during his

time at HIU. Dr. Arianna Moretti is kindly acknowledged for her time and dedication to

supervision and discuss experimental approach and results.

I would like to express my sincere thank to all colleagues at HIU, especially to Huang

Zhang, Bingscheng Qin, Dr. Dominic Bresser, Marlou Keller, Tobias Eisenmann, Dr.

Ulderico Ulissi, Matthias Künzel, Dr. Christoph Vaalma, Markus Ding, Xinwei Dou,

Maral Hekmatfar, Arefeh Kazzazi and Lucas Lodovico for making the working place an

enjoyable atmosphere. Moreover, I am grateful for the administrative support of Andrea

Roessler-Versteeg, Susanne Krauße and Annette Vintiska.

A special thanks go to Dr. Varvara Sharova, with whom I worked together in the SPICY

project, for her limitless source of motivation, responsibility and also for the sweets.

I would like to take the opportunity from the deep of my heart to say thank to my

grandfather Mauro Nascimento Vieira and my grandmothers Maria Emília Vieira and

Maria Deolinda Carvalho, with whom I learned that we can always share something,

independently if it is food, knowledge, experience or just a story.

120

Finally, I am grateful for the support, patience and help of my wife and my son Toni.

They have the gift to transform a bad time into something wonderful by simple acts.

Likewise, I am very fortunate to have the endless support of my father José, my mother

Susi and my brother Thiago.

121

Diogo Vieira Carvalho

Birth: Assis-SP (Brazil)

E-mail: dv-carvalho@live.com

Education

03.1994 – 12.1997 Escola Municipal Profa Maria Clelia de Oliveira

Vallim

03.1998 – 12.2000 Escola Estadual Profa Cleophania Galvao da

Silva

03.2001 – 12.2004 Escola Estadual Dr. Clybas Pinto Ferraz

Higher Education

03.2005 – 12.2012 Diploma in Physics

Supervisor: Prof. Dr. Avacir Casanova Andrello Universidade Estadual de Londrina, Londrina, Brazil Scholarship : Fundação Araucária de Apoio ao Desenvolvimento Científico e Tecnológico

Ph.D. Study

04.2004 Ph.D. Studies in Chemistry

Supervisor: Prof. Dr. Stefano Passerini Helmholtz-Institute Ulm (HIU) / Karlsruhe Institute of Technology (KIT) Ulm, Germany

Professional occupation

Since 04.2014 Research Assistant

Helmholtz-Institute Ulm (HIU) / Karlsruhe Institute of Technology (KIT) Ulm, Germany

122

10.2013 – 03.2014 Research Assistant

Institute of Physical Chemistry Westfälische-Wilhelms Universität Münster Münster, Germany

01.2011 – 12.2011 Internship

Instituto Agronômico do Paraná Londrina, Brazil

07.2010 – 12.2010 Teacher in elementary and high school (Math

and Physics)

Secretaria Estadual de Educação do Paraná Londrina, Brazil

Ulm, ________________

123

11.1. Patent

Carvalho, D.V.; Loeffler, N.; Kim, G.-T.; Passerini, S. Separator for an electrochemical

device and method for the production thereof. Patent No. EP 3085432 B1, 20 December

2017.

11.2. Peer-reviewed publications

Carvalho, D.V.; Loeffler, N.; Kim, G.-T.; Passerini, S. High Temperature Stable

Separator for Lithium Batteries Based on SiO2 and Hydroxypropyl Guar Gum.

Membranes 2015, 5, 632-645.

Carvalho, D.V.; Loeffler, N.; Kim, G.-T.; Marinaro, M.; Wohlfahrt-Mehrens, M.;

Passerini, S. Study of Water-Based Lithium Titanate Electrode Processing: The Role of

pH and Binder Molecular Structure. Polymers 2016, 8, 276.

Ochel, A.; Di Lecce, D.; Wolff, C.; Kim, G.-T.; Carvalho, D. V.; Passerini, S.

Physicochemical and electrochemical investigations of the ionic liquid N-butyl -N-

methyl-pyrrolidinium 4,5-dicyano-2-(trifluoromethyl)imidazole. Electrochimica Acta

2017, 232, 586.

Sharova, V.; Moretti, A.; Giffin, G.A.; Carvalho, D.V.; Passerini, S. Evaluation of

Carbon-Coated Graphite as a Negative Electrode Material for Li-Ion Batteries. C 2017,

3, 22.

124

Kuenzel, M.; Bresser, D.; Diemant T.; Carvalho, D. V.; Kim, G.-T.; Behm, R. J.;

Passerini, S. Complementary Strategies Toward the Aqueous Processing of High-

Voltage LiNi0.5Mn1.5O4 Lithium-Ion Cathodes. ChemSusChem 2018, 11, 562.

Carvalho, D.V.; Loeffler, N.; Hekmatfar, M.; Moretti, A.; Kim, G.-T.; Passerini, S.

Evaluation of guar gum-based biopolymers as binders for lithium-ion batteries

electrodes. Electrochimica Acta 2018, 265, 89.

Zhang, H.; Jeong, S.; Qin, B.; Carvalho, D. V.; Buchholz, D.; Passerini, S. Towards

High-Performance Aqueous Na-ion Batteries: Stabilizing the Solid/Liquid Interface for

NASICON-Type Na2VTi(PO4)3 via the Use of Concentrated Electrolytes.

ChemSusChem 2018, 11, 1382.

11.3. Conference Contributions

“SiO2 and hidroxypropyl guar gum based separator for lithium-ion batteries” by D.V.

Carvalho, N. Loeffler, G.T. Kim, S. Passerini. (Oral presentation, 16th Advanced

Batteries, Accumulators and Fuel Cell in Brno (Czech Republic) 2015).

“Guar gum and its derivatives as binders for lithium-ion battery electrodes” by D.V.

Carvalho, N. Loeffler, A. Moretti, G.T. Kim, S. Passerini (Oral presentation, European

MRS meeting in Strasbourg (France) 2017)

“Enabling silicon/graphite composite anodes in Li-ion batteries via fine tuning of

electrode formulation” by D.V. Carvalho, I. de Meatza, A Moretti, S Passerini. (Oral

presentation, XVI Brazilian MRS meeting in Gramado (Brazil) 2017).

125

Declaration

I herewith declare that the present thesis was executed by my own, and the information

sources has been cited in the bibliography.

Ulm, __________

Diogo Vieira Carvalho

126

Parts of this dissertation have already been published in the following patent journal

articles:

Carvalho, D.V.; Loeffler, N.; Kim, G.-T.; Passerini, S. Separator for an electrochemical

device and method for the production thereof. Patent No. EP 3085432 B1, 20 December

2017.

Carvalho, D.V.; Loeffler, N.; Kim, G.-T.; Passerini, S. High Temperature Stable

Separator for Lithium Batteries Based on SiO2 and Hydroxypropyl Guar Gum.

Membranes 2015, 5, 632-645.

Carvalho, D.V.; Loeffler, N.; Kim, G.-T.; Marinaro, M.; Wohlfahrt-Mehrens, M.;

Passerini, S. Study of Water-Based Lithium Titanate Electrode Processing: The Role of

pH and Binder Molecular Structure. Polymers 2016, 8, 276.

Carvalho, D.V.; Loeffler, N.; Hekmatfar, M.; Moretti, A.; Kim, G.-T.; Passerini, S.

Evaluation of guar gum-based biopolymers as binders for lithium-ion batteries

electrodes. Electrochimica Acta 2018, 265, 89.