Development of Cathode Materials for Li-ion

Battery and Megalo-Capacitance Capacitor

(リチウムイオン電池と巨大容量キャパシタ用正極活物質の開発)

September 2007

Division of Energy and Materials Science

Graduate School of Science and Engineering

Saga University

Arjun Kumar THAPA

Approval

Division of Energy and Materials Science Graduate School of Science and Engineering

Saga University 1 Honjo-Machi, Saga 840-8502, Japan

CERTIFICATE OF APPROVAL

Ph. D. Dissertation

This is to certify that the Ph. D. Dissertation of

Arjun Kumar THAPA has been approved by the Examining Committee for the dissertation

requirement for the Doctor of Engineering degree in Chemistry at the September, 2007 graduation.

Dissertation committee:

Supervisor, Prof. Hideyuki NOGUCHI Dept. of Applied Chemistry

Member, Prof. Masaki YOSHIO

Member, Prof. Masamitsu NAGANO

Member, Prof. Takanori WATARI

Member, Prof. Hiroyoshi NAKAMURA

Content

Chapter 1

Introduction

1. Introduction---------------------------------------------------------------------------------------1

1.1 Battery and Electric double layer capacitor (EDLC) ---------------------------------------1

1.2 Li-ion battery ------------------------------------------------------------------------------------3

1.2.1 The history of Li-ion Battery ----------------------------------------------------------------3

1.2.1 General concepts of Li-ion battery----------------------------------------------------------4

1.2.2.1 Lithium metal battery------------------------------------------------------------------4

1.2.2.2 Lithium ion battery---------------------------------------------------------------------5

1.2.3 Components for Lithium ion battery--------------------------------------------------------6

1.2.3.1 Requisite for electrode materials-----------------------------------------------------6

1.2.3.2 Cathode materials for lithium ion battery-------------------------------------------9

1.2.3.2.1 Layered transition metal oxide cathodes------------------------------------9

1.2.3.2.2 Li-Mn-O system--------------------------------------------------------------13

1.2.3.2.3 LiFePO4 with olivine structure---------------------------------------------16

1.2.3.3 Electrolytes-----------------------------------------------------------------------------18

1.2.3.4 Separators------------------------------------------------------------------------------18

1.3 Electric double layer capacitor---------------------------------------------------------------20

1.3.1 The history of EDLC------------------------------------------------------------------------21

1.3.2 General concepts of EDLC-----------------------------------------------------------------22

1.3.2.1 Faradaic and non-faradaic process--------------------------------------------------22

1.3.2.2 Advantages of capacitor over other batteries--------------------------------------24

1.3.2.3 Classification of electrochemical capacitors---------------------------------------24

1.3.2.3.1 Carbon-------------------------------------------------------------------------25

1.3.2.3.2 Metal oxide--------------------------------------------------------------------26

1.3.2.3.3 Polymer------------------------------------------------------------------------27

1.3.2.3.4 Electrolytes---------------------------------------------------------------28

1.3.3 Principle of energy storage system--------------------------------------------------------30

1.3.3.1 Specific energy and power density of capacitor and EDLC-----------------31

1.3.3.2 The gravimetric capacitance dependence of on the specific surface area of

activated carbon----------------------------------------------------------------------32

1.3.3.3 Electric conductivity of PC solution on the concentration of solute

for electrochemical capacitor-------------------------------------------------------34

1.3.4 Classes of advanced capacitor-------------------------------------------------------------35

1.3.4.1 Aqueous electrolyte, carbon double layer capacitor----------------------------36

1.3.4.2 Non-aqueous electrolyte, carbon double layer capacitor-----------------------36

1.3.4.3 Li-ion hybrid capacitor-------------------------------------------------------------37

1.4 Target of this research------------------------------------------------------------------------ 38

References ----------------------------------------------------------------------------------------- 41

Chapter 2

Synthesis and Electrochemical Performance of o-LiMnO2-yXy Cathode

Materials at High Temperature (0.01 ≤ y ≤ 0.05)

2.1 Introduction-------------------------------------------------------------------------------------51

2.2 Experimental-----------------------------------------------------------------------------------52

2.3 Result and discussion-------------------------------------------------------------------------54

2.4 Conclusion-------------------------------------------------------------------------------------69

References-------------------------------------------------------------------------------------------70

Chapter 3

Novel Synthesis of LiNi0.5Mn0.5O2 by Carbonate Co-precipitation, as an

Alternative Cathode for Li-ion Battery

3.1 Introduction-------------------------------------------------------------------------------------73

3.2 Experimental -----------------------------------------------------------------------------------74

3.3 Result and discussion--------------------------------------------------------------------------77

3.4 Conclusion--------------------------------------------------------------------------------------95

References-------------------------------------------------------------------------------------------96

Chapter 4

Development of Cathode Materials for Advanced Megalo-Capacitance

Capacitor

4.1 Introduction-------------------------------------------------------------------------------------98

4.2 Experimental----------------------------------------------------------------------------------101

4.3 Result and discussion-------------------------------------------------------------------------105

4.4 Conclusion-------------------------------------------------------------------------------------140

References------------------------------------------------------------------------------------------142

Chapter 5

General Conclusion

Conclusions----------------------------------------------------------------------------------------143

1

1. Introduction

1.1 Battery and electric double layer capacitor (EDLC)

During the past several decades the electronics devices have been rapidly developed.

In recent year, outstanding growth in portable electronic products, such as notebook

computer, cellular phones, PDA’s and camcorder, is leading an increasing demand for

energy storage system. As the portable products become more advanced and their size

and weight continue to reduce, the demands for energy storage device, such as smaller

size, higher energy density and lighter weight are also continuously increased. Generally,

energy storage devices for the mobile electronics are categorized into two kinds of

battery, i.e. primary and secondary battery. However, the secondary battery is better than

primary battery in the viewpoint of economy and environment.

There are several kinds of secondary battery for the portable electronic device, such

as nickel cadmium (Ni-Cd), nickel metal hydride and lithium ion. Among these, lithium

ion battery has mainly used nowadays as an energy storage device for its special merits,

such as high working voltage of 3.6 V, which is three times higher than those of Ni-Cd

and Ni-MH secondary battery. Lithium ion battery has longer cycle life, lower self

discharge rate and higher energy density than Ni-Cd, and Ni-MH secondary battery.

Nowadays, Li-ion batteries are used for mobile phone, digital camera, and notebook

computer in large scales but they need lower power for long time. High power density

and high energy density are required as a power source for hybrid car and electric

vehicle (EV). For high power sources, there is an electric double layer capacitor

(EDLC) utilizing adsorption and desorption on the interface of solid and liquid. The

EDLC based on a double layer mechanism was developed in 1954 by H. I. Becker using

2

a porous carbon electrode [1]. The relationship between specific energy density and

power density of battery and EDLC is shown in Fig.1.1. Li-ion battery has high energy

density but it has low power density. The energy density of Li-ion battery decreases

with increase in rate capability. Conventional EDLC has high power density but it has

low energy density. For hybrid car and EV, the energy density up to the target as shown

in Fig.1.1 is needed.

Therefore, our target is to develop the cathode material having high power density

and high energy density for Li-ion batteries and EDLC in this research.

Figure1.1 Specific energy density of several batteries and capacitors.

100

1000

10000

0.1 1 10 100 1000Energy Density (Wh/kg)

Pow

er D

ensi

ty (

W/k

g)

Conventional EDLC

Lead Acid

Hydrogen Adsorbed System

LIB

Target portion of Traction System

100

1000

10000

0.1 1 10 100 1000Energy Density (Wh/kg)

Pow

er D

ensi

ty (

W/k

g)

Conventional EDLC

Lead Acid

Ni-MH

Li-ion cell

Targeted EDLC for Car

Target for Li-ion

100

1000

10000

0.1 1 10 100 1000Energy Density (Wh/kg)

Pow

er D

ensi

ty (

W/k

g)

Conventional EDLC

Lead Acid

Hydrogen Adsorbed System

LIB

Target portion of Traction System

100

1000

10000

0.1 1 10 100 1000Energy Density (Wh/kg)

Pow

er D

ensi

ty (

W/k

g)

Conventional EDLC

Lead Acid

Ni-MH

Li-ion cell

Targeted EDLC for Car

Target for Li-ion

3

1.2 Li-ion battery

1.2.1 The history of lithium battery

Since Li element is light and possesses a relatively lower potential, many research

groups tried to apply this element to high power density such as Li-ion, Polymer Li-ion,

and Li-battery as shown in Fig 1.2 [2]. Obviously, lithium, an anode, can give higher

energy density than the other anodes. However, metallic lithium easily reacts with water

and emits hydrogen gas. Therefore, the aqueous electrolyte cannot be used for metallic

lithium anode. The research of lithium battery was activated since it was reported that

the metallic lithium can be stable in some electrolyte such as molten salt, liquid SO2 and

organic electrolyte like PC (propylene carbonate) / LiClO4.The primary lithium battery

had been developed in 1970’s [3].

Figure 1.2 Specific energy densities (per weight and per volume) of several batteries.

4

An attempt to develop rechargeable lithium batteries has been carried out from the

beginning of 1970’s, and extensive researches allow the commercialization of the

lithium secondary batteries. However, the lithium secondary battery has a severe

problem which is dendrite formation of lithium metal during charge process. As a matter

of fact, several cellular phones, which are mounted lithium secondary batteries as power

sources, were recalled at the end of 1980’s with respect to several undesirable accidents,

such as firing or explosion during use. In the early stage of 1980’s, a new concept,

“Rocking chair”, was proposed [4] and demonstrated by some research groups [5-7].

However, cathode material should have high redox potential in order to compensate the

decrease in cell voltage.

In 1990’s, lithium ion battery had commercialized by Sony Energytech based on a

carbon (non-graphitic) as an anode and LiCoO2 as a cathode. The name of “lithium ion

battery” was given by T. Nagaura and K. Tozawa [8], and the concept of “lithium ion

battery” was firstly introduced by Asahi Kasei Co. Ltd and they obtained patents over

the world [9]. Sony has made several improvements on the performance of their

particular rocking chair battery (RCB) system. Today, the lithium ion battery is the

fastest growing and the most promising battery.

1.2.2 General concepts of Li-ion battery

1.2.2.1 Lithium metal battery

The discharge process of this battery involves three main steps in lithium battery; i)

dissolution of lithium ion from lithium metal anode, ii) the migration of lithium ions

across the electrolyte and separator and iii) insertion of lithium ions into the host

structure of the cathode. Common cathode materials are inorganic compounds, such as

5

transition metal dechalcogenides and oxides. They are characterized by layered or

tunneled structure, which can provide channels for transfer of lithium ions reversibly

and faster in the host matrix. In fact, it is well known that Li/V6O13 [10], Li/TiS2 [11]

and Li/MoS2 [12] cells have been proved to have a reversible behavior. The Li/MoS2

cell was commercialized by Moli Energy.

In the view point of reversibility, I believe that the lithium secondary battery with

lithium metal anode is capable of very long cycle life and supplies a fairly stable

operation voltage and capacity. However, practical cycleability of the battery is limited

due to the dendritic growth of lithium metal, which causes short circuit of cell, followed

a drastic increase in volume of the anode [13, 14]. The abrupt changes in the anode

would bring about safety hazard or explosion by thermal abuse.

1.2.2.2 Lithium ion battery

Lithium ion battery (LIB) has emerged from lithium metal battery to eliminate its

un-safety problem. The principle of the Li-ion battery is explained in the Fig. 1.3. The

inorganic compound, LiCoO2, is used as the cathode material. This material has a

rhombohedral structure where Li and Co cations fill alternating layers of edge-sharing

octahedral sites in a close packed oxygen array. Carbonaceous material is used as an

anode material. During charging, lithium is deintercalated from the cathode layers, then

transported and intercalated into the carbonaceous anode. In the discharge process, the

lithium ions are deintercalated from the anode and intercalated again to the empty

octahedral site between layers in the cathode. Thus, this battery is sometimes called as

the “Rocking Chair Battery” or the “Swing Battery” from the view point of lithium

transfer. When a cell is assembled, lithium ions are not doped into carbon. Therefore,

6

the battery has low voltage near 0 V.

Positive electrode:

LiCoO2 Li1-xCoO2 + xLi+ + xe-

Negative electrode:

C + xLi+ +xe- Lix + C

Figure 1.3 Basic structure and redox reaction of lithium ion battery.

1.2.3 Components for lithium ion battery

1.2.3.1 Requisites for electrode materials

High specific energy density (related to both average working voltage and reversible

capacity) and long cyclic life (related to the stability of structure and

electrode-electrolyte interface) are simultaneously required for a high performance

discharge

charge

discharge

7

Li-ion battery. Therefore, some aspects should be considered in the design of new

electrode materials for lithium rechargeable batteries. In general, structure, chemical

stability and the available redox couples of electroactive materials are the primary

points. Ideally, it should fill following conditions:

(a) An open structure to permit reversible lithium migration

In brief, the working principle of lithium rechargeable battery is the reversible

migration of lithium between cathode and anode accompanied by a redox process,

whose reversibility is prerequisite to Li secondary battery. Thus, the structure of the

compound should be open for lithium insertion into its lattice.

(b) Stability of electrode and electrolyte

This is the requirement for a long cycle life. The insertion/extraction reaction has a

topotactic character and both insertion and extraction of guest lithium ions into and

from host compound should ideally keep original host structure. On the other hand, the

oxidation of electrolyte should be avoided during cycling.

(c) A higher specific energy density

The specific energy density (per weight or per volume) is related to both the

working voltage and the reversible capacity. The former depends on the potential of the

redox process and the latter is restricted by the reversible amount of lithium

intercalation. The available redox pair should locate in a higher and suitable potential

range and the structure of material should be stable in wide composition range in order

to obtain a high capacity.

(d) Higher electrode conductivities

The electrochemical lithium insertion/extraction reactions involve both lithium ions

diffusion in the lattice and charge transfer process on the particle surface. Thus,

8

electrode’s conductivity includes both lithium ion conductivity in active material bulk

and electronic conductivity of electrode. Higher electronic conductivity is helpful to

keep the inner resistance low and gives an excellent power density. Beside this, lower

interfacial resistance for electrochemical reaction is desirable for the lower polarization

of the electrode.

(e) A low cost and environment benign.

The cost and environmental impact should be always kept in mind for new battery

design. It is one of the future challenges to develop cheaper and hazardless electrode

materials with excellent battery performances.

Figure 1.4 Relation between potential vs. capacity of some electrode materials.

9

Fig. 1.4 shows the working potential and capacity of the various electrode materials

[16]. Compounds with a working potential of more than 3 V can be used as cathode

materials in lithium ion battery. They should include lithium such as LiCoO2, LiMn2O4,

LiNiO2, LiMnO2 and their derivatives, and polyanionic compounds, for example,

LiFePO4.

1.2.3.2 Cathode materials for lithium ion battery

1.2.3.2.1 Layered transition metal oxide cathodes

Typical cathode material is layered metal oxides with a general formula of LiMeO2

(Me = Transition metal elements such as Co, Ni and Mn), in which the Li ion and Me

ion occupy the alternate (111) planes of the rock salt structure. The structure has an

oxygen stacking sequence of ⋅⋅⋅ABCABC⋅⋅⋅ along the c axis of hexagonal setting and the

Li and Me ions occupy the octahedral sites. There are three MeO2 sheets per unit cell.

This structure can be described as a layered structure with a space group of R3_

m, and the

unit cell parameters are conveniently defined in terms of the hexagonal setting [17]. A

schematic layered structure is presented in the Fig. 1.5.

10

Figure 1.5 The structure of layered LiMeO2 (Me = Co, Ni, and Mn) crystal structure.

Layered LiCoO2

The most of the commercial lithium ion batteries are built with the LiCoO2 cathode

material. The LiCoO2 has a hexagonal structure with cell parameters of a = 2.82 Å and c

= 14.08 Å [17-21]. The cathode synthesized at higher temperature above 800 oC shows

O3 type stacking R3－

m with a regular ordering of Li+ and Co3+ ions in the alternate (111)

planes of the rock salt lattice. On the other hand, synthesis at lower temperatures (~400

oC) results in a considerable disordering in distribution of the Li+ and Co3+ ions, which

lead to the formation of a spinel like LiCoO2 phase with poor electrochemical

Lithium

Ni, Co or Mn

Oxygen

11

properties.

LiCoO2 has a theoretical capacity of 274 mAh g-1, corresponding to extraction of 1

mole of Li+ from LiCoO2 under the condition that electrode is charged until Li0.5CoO2.

The cathode undergoes serious phase transformation between hexagonal and monoclinic

when the cathode charged above 4.2 V [22-24]. Since this phase transformation leads

deterioration of cycling performance, practical capacity is restricted as half of

theoretical capacity. Moreover, the capacity fading can be partly attributed to side

reaction, Co dissolution at higher voltage of more that > 4.2 V, from oxide compound

into electrolyte.

In order to suppress phase transformation, substitution of foreign elements such as

Mg [25-27], Al [28-33], Fe [34-36], Ni [37-44], Cr [45], Mn [46,47] and Li [48-50] for

Co has been extensively studied. Recently, Zou et al reported that small amount of

foreign element can give stable cycling performance due to suppression of phase

transformation at ca. 4.2 V during intercalation/de-intercalation process at higher cut-off

voltage of 4.5 V [51-53]. Nowadays, Mg doped LiCoO2 is used for commercial lithium

ion battery. Moreover, it has been found that the cycleability of the cathode material

could be improved significantly by modifying its surface with metal oxide such as

Al2O3 and ZrO2 [54, 55]. However, drawbacks of LiCoO2, such as higher price and

toxicity of cobalt element, accelerate extensive studies to investigate alternatives to

LiCoO2.

Layered LiNiO2

Another well-known member of LiMeO2 family is a lithium nickel oxide (LiNiO2)

compound [56-60]. It has same crystal structure with LiCoO2 and belongs to hexagonal

12

system (R3_

m) with cell constant of a = 2.88 Å and c = 14.18 Å. Its working voltage is

more than 3.7 V and theoretical capacity is 275 mAh g-1. The LiNiO2 provides

important advantages, such as less toxicity, lower price and higher reversible capacity of

200 mAh g-1 as compared to LiCoO2. However, it suffers from a few problems, i.e.

irreversible phase transformation, difficulty in synthesis and safety concerns.

The cathode material shows topotatic reaction consisting of three single phase

reactions for the 0 ≤ x ≤ 0.75 in Li1-xNiO2 and two phase reaction, in 0.75 < x < 1

[61-63]. The phase transformation could cause capacity fading. Moreover, the Li-Ni-O

system is characterized by the existence of a Li1-zNi1+zO2 (0 < z ≤ 0.2) solid solution.

Although the structure can still be described as a layered structure at low z value (z ≤

0.2), it has cation mixing, i.e. presence of extra Ni2+ cations in the Li layers. The extra

nickel ions in the Li layer hinder the diffusion of lithium ions during electrochemical

cycling and resulting in poor battery performance.

In order to overcome these drawbacks, the LiNi1-xCoxO2 solid solution has been

widely studied because these materials seem to combine the advantages of both LiNiO2

and LiCoO2. For instance, the presence of more than 20 % of Co in the LiNiO2 can

stabilize an ideal layered structure without cation mixing, thus LiCo0.2Ni0.8O2 shows

excellent electrochemical behavior [64-67]. However, Co content should be less than

15 % in order to compensate cost problem. Although the reversible capacity is

decreased, the thermal stability of the cathode can be improved by partly substitution of

small amount of Mn [68, 69] and Fe [70, 71] for Ni. Small amount of Mg doping [72]

leads slight decrease in initial capacity, even though it causes and a strong increase in

electronic conductivity by hole formation. Further suppression of phase transformation

by the doped compound results in excellent cycle stability.

13

1.2.3.2.2 Li -Mn-O system

Spinel LiMn2O4 compound

A spinel compound, LiMexMn1-xO4 (Me = Metal element), is an attractive cathode

material for replacing Co or Ni based layered cathode materials in the next generation of

Li-ion batteries. Especially, it has been extensively studied as positive electrode

materials of large-size lithium ion batteries for power sources of hybrid electric vehicles

(HEV) [73,74] because they have several advantages such as lower cost, high-rate

capability and higher thermal stability compared to those of Co or Ni based layered

materials (e.g. LiCoO2 or LiNiO2).

As shown in the Fig. 1.6, the LiMn2O4 crystallizes in space group, Fd3_

m, with Li and

Mn present in 8a tetrahedral sites and 16d octahedral sites in the cubic close-packed

oxygen array, respectively. All other sites, such as tetrahedral 8b and 48f sites and

octahedral 16c sites, are empty [75]. MnO6 octahedra share edges to build a rigid

three-dimensional framework. Li ions diffuse through 8a-16c-8a path [76].

From the spinel cathode, about 1 mole of lithium can be reversibly extracted in two

steps, at 4.05 V and 4.15 V, leading to the delithiated λ-MnO2 [77]. On the other hand,

the spinel LiMn2O4 is changed to tetragonal Li2Mn2O4 when more lithium ions are

inserted into the 16c site at 2.8 V.

The cathode was firstly used in a commercial Li-ion cell in 1996. However, this

compound shows poorer performance than the layered cathodes. The main problem for

the application of Mn spinel is the capacity fading upon cycling which become serious

at elevated temperature (above 60 oC). Many possible sources had been proposed, such

as structural instability [78-80], Mn dissolution into the electrolyte [81,82], the

cooperative Jahn-Teller effect [83].

14

Recently, M. Yoshio et al. elucidated that main reason for capacity fading of spinel,

Figure 1.6 Structure of cubic spinel LiMn2O4.

LiMexMn1-xO4 (Me = Metal element), is closely related to oxygen deficiency [79,

84-86]. In the charge-discharge profiles, oxygen deficient spinel shows a 3.2 V

discharge plateau together with 4.5 V charge plateau, and the charge-discharge plateaus

delivering the equivalent capacity [84, 85]. They also have reported that oxygen

deficient spinel brings the phase transformation from cubic to orthorhombic (or

tetragonal) phase at around lower than room temperatures [79]. Consequently, preparing

oxygen stoichiometric spinel compound is critical point to overcome capacity fading.

Main drawback in application of oxygen stoichiometric LiMexMn1-xO4 for

LiMexMn1-xO4/graphite cell at elevated temperature (over than 50 oC) is dissolution [87]

of Mn ions because the dissolved Mn ions from cathode material [88, 89] attack

15

graphite anode therefore causing degradation of anode. In order to reduce the solubility

of Mn ions at the elevated temperature, surface encapsulation [90], and reducing

specific surface area [86, 91] have been suggested and investigated.

So, oxygen stoichiometric LiMexMn1-xO4 spinel improved cycle performance at

elevated temperature [85, 86, 90, 91].

LiMnO2 compound

Since this compound shows higher theoretical capacity of 285 mAh g-1 than spinel

LiMn2O4 148 mAh g-1, LiMnO2 as cathode materials is of interest. LiMnO2 is

polymorphous, mainly orthorhombic (zigzag layered structure, Pmmn) and monoclinic

(layered structure, C2/m) structure as shown in Fig. 1.7 (a) and (b), respectively.

Orthorhombic LiMnO2 had been firstly reported by Johnson and Heikes in the study

of the Li-Mn-O system [92]. It is well-known that a LiMnO2 structure with an almost

ideal layered arrangement of the lithium and manganese ions can be prepared by Li+

ion-exchange from α-NaMnO2. Monoclinic LiMnO2 has iso-structural with layered

LiCoO2 has been attributed to a strong Jahn-Teller effect induced by the Mn3+ (d4) ions

[93, 94]. Three polymorphs undergo phase transformation to spinel.

16

Figure 1.7 The structure of (a) orthorhombic and (b) monoclinic LiMnO2.

1.2.3.3.3 LiFePO4 with olivine structure

Padhi first demonstrated that orthorhombic LiFePO4 could be used as a cathode for

rechargeable lithium battery [95]. It has a theoretical capacity of 170 mAh g-1 and is

environmentally benign and inexpensive. Further it shows excellent cycle stability due

to structural similarity between charged/discharged states. Specially, this material

exhibits very good thermal stability [96, 97]. These properties make it an attractive

candidate for large scale batteries, such as power source for an electric vehicle.

17

Figure 1.8 Structure of olivine LiFePO4.

Olivine type LiMPO4 (M = transition metal) is one of polyanionic compounds, such

as NASICON Li3Fe2 (PO4)3 [98] and monoclinic (or rhombohedral) Fe2(SO4)3 [99]. The

polyanionic compounds can be viewed as replacement of O2- anions by larger size

(XO4)m- polyanions such as PO43-, SO4

2-, MoO42-, WO4

2- and so forth. Olivine type

LiFePO4 (orthorhombic structure, Pnma) allows two-dimensional Li+ diffusion pathway

as shown in Fig. 1.8. The problem for this material is low electrical (electronic and

ionic) conductivity, so electrochemical properties depend on charge/discharge rate and

operating temperature [100]. In order to overcome this drawback, many works have

been done and it is found that reduction of particle size and increase in electronic

conductivity by coating of conducting agent are quite effective [101-104]. However,

18

reducing particle size leads to lower electrode density and lower energy density, and

coating process requires too much carbon. Therefore, the drawbacks, such as lower

conductivity, lower operating voltage and poor rate capability, prevent the cathode

material from commercial application.

1.2.3.3 Electrolytes

The most popular electrolytes are the liquid-type ones where carbonates or esters of

simple alcohol and glycol are frequently used as solvents which contain LiPF6 as an

electrolyte. Solvent are mixed solution of ethylene carbonate (EC) of high dielectric

constant and methyl ethyl carbonate (MEC) of low viscosity. Sometimes a combination

of γ-butyl-lactone and LiBF4 is utilized. Propylene carbonates is an excellent solvent,

but it decomposes vigorously on the surface of graphite.

Recently, polymer electrolytes have attracted much attention because they enable us

to be free from electrolyte leakage and to make a very thin battery. Many kinds of

polymer electrolytes have been proposed, but only a few are utilized in practical

batteries. Basically, these are not a true solid polymer, but a polymer gel containing

liquid electrolyte as a plasticizer. The gel contains a lot of liquid, however, it strongly

resists leakage of the liquid electrolyte which is kept stably in the matrix.

1.2.3.4 Separators

The separator used in LIB has two important functions: one is to avoid the direct

contact between the anode and cathode, while it allows a free mass transfer of the

electrolyte, and the other is a shutter action to stop the mass transfer in the case of

accidental heat generation, which cause it melt down, resulting in shut down of a cell.

The material should be soft and flexible enough to act as buffer between the both sheets

19

of cathode and anode. The material should be sufficiently stable for a long time while it

is kept in contact with the electrolyte. Orientated polyolefin film is commonly used.

Stretching of film is an important process for obtaining a high porous film. Examples of

SEM images of this film are shown in Fig. 1. 9.

Figure1. 9 SEM images of the two types of olefin based separator sheets

20

1.3 Electric double layer capacitor The discovery of a so-called condenser, (that electric charges could be stored on the

plate) now referred to as a capacitor, was made in the mid-eighteenth century during the

period when the phenomena associated with ‘static electricity’ were being revealed. The

embodiment known as a condenser is attributed to Musschenbroek [105] in 1746 at

Leyden in the Netherland, hence the name ‘Leyden jar’.

The electrochemical capacitor was supposed to boost hybrid electric vehicle to

provide the high or strong power for acceleration, and additionally allow the recovery of

braking energy. New or novel electrochemical capacitors fill in the gap between

batteries and conventional capacitors such as electrolytic capacitors or metallized film

capacitors. Thus, electrochemical capacitors will overcome in terms of power density

and also improve capacitor performance in terms of energy density. In addition,

electrochemical capacitors have a longer cycle life as compared to batteries because no

or negligibly small chemical charge transfer reactions are involved.

(a) Tantalum Capacitor

The major use for tantalum, as the metal powder, is in the production of electronic

components, mainly capacitors and some high audio-grade resistor. Tantalum

electrolytic capacitor exploit the tendency of tantalum to form a protective oxide surface

layer, using tantalum powder, pressed onto a pellet shape, as one “plate” of the capacitor,

the oxide as the dielectric, and conductive solid as the other “plate”. Because the

dielectric layer is very thin, a high capacitance can be achieved in a small volume.

Because of the size and weight advantages, tantalum capacitors are attractive for

portable telephone, pagers, personal computers and automotive electronics.

(b) Aluminium Electrolytic Capacitor

21

Aluminium electrolytic capacitor is used in wide range of electric circuits.

Application for these types of capacitors includes high performance digital equipment,

such as liquid crystal displays, personal computers.

1.3.1 The history of EDLC

The EDLC based on a double layer mechanism was developed in 1954 by H. I.

Becker using a porous carbon electrode. Capacitors which store the energy within the

electric double layer at the electrode/electrolyte interface are known under various

names such as ‘electric double layer capacitor’, ‘supercapacitors’, ‘ultracapacitors’,

‘power capacitors’, ‘gold capacitors’ and so on. Electric double-layer capacitor (EDLC)

is the name that expressed the fundamental charge storage principle of such capacitors.

Electrical energy can be fundamentally stored in two different ways: (1) in-directly

in batteries as potentially available chemical energy requiring Faradaic oxidation and

reduction of the electrochemically active reagents to release charges that can perform

electrical work when they flow between two electrode having different electrode

potentials (i.e., across the voltage difference between the pole of battery cells); and (2)

directly, in an electrostatic way, as negative and positive charges on the plates of a

capacitor, a process known as non-faradaic electrical energy storage.

22

Fig. 1.10 shows a schematic diagram of an electric double layer capacitor consisting

of a single cell with a high surface-area electrode material, which is loaded with

electrolyte [106]. The electrodes are separated by a porous separator filled with

electrolyte solution.

1.3.2 General concepts of EDLC

1.3.2.1 Faradaic and non-faradaic process

There is a general and fundamental difference between the mechanisms of operation

of electrochemical capacitors and battery: for the double layer type of capacitor, the

charge storage process is non-Faradaic, i.e., ideally no electron transfer takes place

across the electrode interface and the storage of electric charge and energy is

electrostatic. In battery-type processes, the essential process is faradaic, i.e., electron

Figure 1.10 Basic structure of electric double layer capacitor.

23

transfer does take place across the double layers, with a consequent change of oxidation

state, and hence the chemistry of the electroactive materials.

The important of differences in the charge storage process in capacitors are as

follows:

(a) Non-Faradaic

The charge accumulation is achieved electrostatically by positive and negative charge

residing on two interfaces separated by a vacuum or a molecular dielectric (the double

layer or, e.g., a film of mica) a space of air or an oxide film.

(b) Faradaic

The charge storage is achieved by an electron transfer that produce chemical or

oxidation state changes in the electroactive materials according to Faraday’s law related

to electrode potential. Pseudocapacitance can arise in some cases. The energy storage

was indirect and analogous to that in a battery. In a battery cell, every negative charge is

Faradaically neutralized by charge transfer, resulting in a change of oxidation state of

some redox-electroactive reagent like e.g.,

Ni3+ .O2-.OH Ni2+.2OH- + e

in the cathode of Ni-Cd battery [107].

In a capacitor, actual negative and positive charge, are accumulated on the electrode

plates with lateral repulsion. However, in some case of double layer charging, some

partial electron transfer occur, giving rise to pseudocapacitance, like e.g., when

chemisorption of electron donate anions such as Cl-, Br-, I- and CNS-.

M+ A M/A(1-δ)- + δe (in M)

+H+

24

The electrons involved in double-layer charging are delocalized conduction-band

electrons of the metal or carbon electrode, while the electrons involved in

faradaic-battery type processes are transferred to or form valence-electron states

(orbital) of the redox cathode or anode reagent, although they arrive in or depart from

the conduction-band states of electronically conducting support material. In some cases,

the faradaically reactive battery material itself is metal conducting like e.g., PbO2, some

sulfide, RuO2, or some well-conducting semiconductor and a proton conductor, like e.g.,

Ni.O.OH.

1.3.2.2 Advantages of capacitor over other batteries

Compared with those of other secondary batteries and Li-ion battery, it has following

advantages:

Long cycle life, >100,000 cycles; some system up to 106.

Good power density [ under certain conditions, limited by IR or equivalent series

resistance (esr) complexity of equivalent circuit]

Simple principle and mode of construction.

Cheap material ( for aqueous embodiment)

Combines state-of-charge indication, Q= CV

Can be combined for rechargeable battery for hybrid application (electric vehicle)

1.3.2.3 Classification of electrochemical capacitors

Recently, supercapacitors have attracted worldwide research interest because of their

potential application for energy devices in many fields. The main advantages of these

devices are long cycle life as compared to Li-ion batteries and high rate capability.

25

Generally, capacitors are distinguished by several criteria such as the electrode material

utilized, the electrolyte, or the cell design. With respect to electrode materials there are

four main categories; carbon based, metal oxide, polymeric materials and graphitic

carbon.

1.3.2.3.1 Carbon

The element carbon exists in several allotropic forms with variuos properties. With

the exception of diamond and fullerenes, the other modification of carbon are variously

used in construction of electrochemical capacitors and batteries of Leclanche or alkaline

types involving MnO2, and in some Li primary batteries, as well as support material for

fuel cell catalysts and conducting polymer capacitor matrices.

For electrochemical capacitors, the carbon must have (1) high specific surface area,

on the order of 1000 m2 g-1; (2) good intra- and interparticle conductivity in porous

matrices; (3) good electrolyte accessibility to intrapore surface area. The surface

conditioning of powdered or fibrous carbon materials for capacitor fabrication is most

important for achieving the best performance, namely, good specific capacitance,

Figure.1.11 Different types of carbon used in capacitor and Li-ion battery

26

conductivity and minimum self-discharge rate. The carbon materials should be free

from impurities (e.g., Fe species, peroxide, O2, quinine).The several types of carbon are

as shown in Fig 1.11. Carbon with high specific surface area is quite susceptible to

anodic electrolyte decomposition which decreases both stability and conductivity in

electrochemical capacitor [108]. The type of carbon depends on the precursor that was

carbonized. A term to identify carbons that are graphitizable (form graphite-like

structures) or non-graphitizable by heat treatment is divided into two types as follow:

(a) Hard carbon

Hard carbons are non-graphitizable and are mechanically hard. Hard carbons are

obtained by carbonizing precursors such as thermosetting polymers (like, phenol

formaldehyde resins, furfuryl alcohol, and di-vinyl benzene styrene copolymer),

cellulose, charcoal, and coconut shells. Transforming such solid precursor to solid

carbon during the carbonization usually form hard carbon. The inability of hard carbon

to form a graphitic structure by heat treatment is the presence of strong cross linking

bonds which impede movement and re-orientation of the carbon atoms to form the

ordered layer structure of graphite.

(b) Soft carbon

Soft carbons are mechanically soft and can be graphitized. Carbonizing precursors

such as petroleum coke, oil, and coal-tar pitch form soft carbons. In these materials, the

formation of carbon proceeds through an intermediate liquid-like phase named as a

mesophase, which facilitates the three-dimensional ordering that is necessary to create a

graphite-like structure.

1.3.2.3.2 Metal oxide

27

Conducting metal oxide like RuO2 or IrO2 were the favored electrode materials in

early EDLCs used for space or military applications [109]. Very high specific

capacitance of up to 750 F/g was reported for RuO2 prepared at relatively low

temperatures [110]. The cyclic voltammogram of RuO2 and also IrO2 electrode have an

almost rectangular shape and exhibit excellent capacitor behavior [111, 112]. However,

the shape of CV is not a consequence of pure double layer charging, but of a sequence

of redox reactions occurring in the metallic oxide. The valence state of Ru may change

from IV to VI with in a potential window of slightly > 1V. The ratio of surface charging

to bulk processes was nicely separated by Trasatti [111]. In aqueous acid electrolytes the

fundamental charge storage process is proton insertion into the bulk material. The high

specific capacitance in combination with low resistance resulted in very high specific

powers. These capacitors, however, turned out to be too expensive. Capacitor cost

showed that 90% of the cost resides in the electrode material. In addition, these

capacitor materials are only suitable for aqueous electrolytes, thus the cell voltage is

limited to 1V.

Several attempts were made to keep the advantage of the material properties of such

metal oxide at reduced cost. The dilution of the costly noble metal by forming

perovskites was investigated by Guther et al. [113]. Other forms of metal compounds

such as nitrides were investigated by Liu et al. [114]. However, these materials are far

from being commercially used in EDLCs.

1.3.2.3.3 Polymers.

Polymeric materials, such as p- and n-dopable poly (3-arylthiphene), p-doped poly

(pyrrole), poly (3-methylthiophene), or poly (1, 5-diaminoanthraquinone) have been

28

used by several research group [115-117] as electrodes for electrochemical capacitors.

The typical cyclic voltammogram of a polymer, however, is not a rectangular shape, as

it is expected for typical capacitor, but exhibits a current peak at the redox potential of

the polymer. In order to be able to use the same electrode material on both capacitor

electrodes polymers with a cathodic and an anodic redox process were utilized recently

[117].

Use of a polymeric material for electrochemical capacitor electrodes give rise to a

debate as to whether such devices should still be called as capacitors or batteries. In

terms of voltage transient during charge and discharge and with CV, they are batteries.

Compared to metallic oxides, however, the term capacitor is justified. The difference

being only that the metallic oxides exhibit a series of redox potentials giving rise to an

almost rectangular CV while the polymer typically has only one redox peak.

High energy density and high power density have been found in such types of

capacitors [117]. The long-term stability during cycling, however, may be a problem.

Swelling and shrinking of electroactive polymers is well known and may lead to

degradation during cycling.

1.3.2.3.4 Electrolytes

Another criterion to classify various electrochemical capacitors is the electrolyte

used. In addition to aqueous acid i.e., and alkaline solution for low-voltage aqueous

supercapacitors, some nonaqueous electrolytes are used for capacitors which could be

operated at higher voltage like (e.g., 3.5 to 4.0V). Some of the electrolyte solutions or

solvents of Li battery technology with substitution of Li salts by tetra-alkyl ammonium

salts have been used owing to their high solubility in nonaqueous solvent of ester

29

carbonate and moderately good conductivity.

Aqueous electrolyte

Aqueous electrolytes limit the unit cell voltage of EDLC to typically 1 V, thus

reducing the available energy significantly compared to organic electrolytes.

Advantages of the aqueous electrolyte are the higher conductance and the fact that

purification and drying process during production are less stringent. The costs of

aqueous electrolytes are usually much lower than suitable organic electrolytes.

Capacitor built by NEC and ECOND [118] use aqueous electrolyte. Aiming to get high

power density, R. Kotz et al. [119] introduced the glassy carbon based capacitor using

an aqueous electrolyte. It should be further pointed out that the capacitor has to be

developed for one or the other electrolyte, not only because of material aspects but also

the porous structure of the electrode has to be tailored for the size and the properties of

the respective electrolytes.

In order to avoid electrolyte depletion problem during charging of the EC, the

electrolyte concentration has to keep high. If the electrolyte reservoir is too small

compared to huge surface area of the electrodes, performance of the capacitor is reduced.

This kind of problem is particularly important for organic electrolytes where solubility

of the salt may be low. Zheng and Jow et al [120] found, however, that concentration

higher than 0.2 molar are sufficient.

Organic electrolyte

Organic electrolyte for electrochemical capacitors have attract considerable attention

due to their greater electrochemical stability (>3 V) compared to aqueous system (1V).

The wider voltage window offered by nonaqueous electrolyte provides increased energy

30

(E=1/2CV2). However, limitation of the power capability of nonaqueous electrolyte is

their inherent lower conductivity. An important property of nonaqueous system is their

lower conductivity which depends on the viscosity, permittivity, and ion concentration

of the electrolyte. In addition, capacitance is also sensitive to the electrolyte

concentration.

Recently, various number of quaternary onium salt have been investigated for

electrolyte in EDLC [121-126]. Ue and his co-workers investigated several ammonium

and phosphonium ion based salts in several organic solvent. They concluded that

tetraethyl ammonium tetrafluroborate (TEABF4) in propylene carbonate (PC) was a

better electrolyte conductivity and stability.

Some chemists have proposed in the view points of use of new ionic liquid [127,128].

Ionic liquids based on immidazolium salt are solidified or crystallized at below room

temperature and have been applied to numerous electrochemical, photovoltaic, and

synthetic field [129-132].Originally, the EMIPF6 ( Ethyl methyl immidazolium) was

synthesized by Fuller and his co-worker for its relatively low melting point

(mp=58-60°C) in thermal battery applications. EMIPF6 make weak ion pair and have a

little hydrogen bonding in the crystal as suggested by J. Fuller [130].

1.3.3 Principle of energy storage system

Electric double layer capacitors store the electric energy in an electrochemical

double layer (Helmholtz Layer) formed at a solid/electrolyte interface. Positive and

negative ionic charges within the electrolyte accumulate at the surface of the solid

electrode. The quantity of ion removed from the bulk electrolyte equals the charge

developed on the electrode surface. Therefore, the maximum energy stored in a

31

capacitor is limited by the capacitance of the capacitor C, and the maximum operating

voltage, V. From the energy density’s view point, the maximum energy density of a

capacitor depends on the specific capacitance of the electrode, Cp under the constant

operating voltage. This is related to esv.

The energy density, Ee, or total energy, Ue, based on the electrode material is

determined by the specific capacitance of the electrode, respectively, and the operating

voltage. It can be expressed by following formulas, respectively:

Ee = 1/8CpV2 or Ue = 1/2 CV2 [1.3.3-1]

1.3.3.1 Specific energy density and power density of Li-ion capacitor and

conventional EDLC

The relation between power density and energy density has attracted much attention

in evaluation of the performance of electrochemical capacitor, especially since such

devices are perceived as capable of delivering high power on discharge although

intrinsically their energy density is low. Fig. 1.12 shows the energy density and power

density of Li-ion capacitor together with those of conventional EDLC. Li-ion capacitors

achieve energy density of 25Wh/kg and 15Wh/kg at power density of 1000W/kg and

4000W/kg (Cell 3), respectively. Li-ion capacitor shows three or four times larger

energy density than that of EDLC and has appreciable energy density level at high

power density.

32

Figure 1.12 Energy density and power density of Capacitors.

1.3.3.2 The gravimetric capacitance dependence on the specific surface area of

activated carbon.

The activated carbon electrode is the most important component in EDLC. EDLC

must have large capacitance and high electrochemical stability to obtain high energy

density and stable performance. The achievement of the excellent performance in EDLC

requires an electrode material (activated carbons) with high specific surface area, having

suitable surface property and pore geometry.

The activated carbons commercially produced for the electrode material of EDLC

are generally divided into two groups. They are hard carbon type activated carbons and

soft carbon type activated carbons. The hard carbon type activated carbons are obtained

by steam activation of coconut shells or alkali activation of resins, most commonly

phenol resin. The soft carbon type activated carbons are produced by the alkali

33

activation of carbonized mesophase pitch and needle coke. Commercially available soft

carbon type activated carbon gives high density and high volumetric capacitance.

However, their swelling and gas evolution during the use are sometimes problems to be

solved.

Fig.1.13 shows the dependence of the gravimetric capacitance of activated carbon

electrode on the specific surface area of the activated carbons used. Hard carbon type

activated carbons having the specific surface area over 2000 m2 g-1 shows the

capacitance of more than 160F/g as shown in Fig. 1.13. Such activated carbons are

derived from resins and obtained by expensive high temperature alkali activation.

Figure1.13 Dependence of gravimetric capacitance on the specific surface area

of the activated carbons.

34

1.3.3.3 Electric conductivity of PC solution on the concentration of solutes for

electric double layer capacitor

The effects of solute concentration on conductivity were measured by T. Morimoto

[133]. Those are most common supporting salts for the nonaqueous electrolyte solution

of electric double-layer capacitors.

Fig.1.14 Dependence of solute concentration in PC solution on the conductivity.

Fig. 1.14 shows the dependence of solute concentration of propylene carbonate (PC)

on the conductivity. The solutes are tetra-ethyl ammonium fluroborate (TEABF4), ethyl

methyl imidazolium fluroborate (EMIBF4), di-ethyl methyl 2-methoxy ethyl ammonium

fluroborate (DEMEBF4) and spiro-(1,1)-bi-pyrrolidinium fluroborate (SBPBF4). The

electric conductivity of these solutions becomes high as the concentration of the solutes

increases. As TEMABF4, and SBPBF4 are more soluble in PC than TEABF4, the PC

35

solution of TEMABF4, EMIBF4, and SBPBF4 show higher conductivity than that of

TEABF4. EMIBF4 is comparatively unstable in the PC solution against higher charge

voltage and weak against high temperature. On the other hand, the PC solution of

TEMABF4 and SBPBF4 are more stable and can be used up to the cell voltage of 2.7 V.

This is because of their electrochemical stability is determined by the electrochemical

decomposition of PC. Consequently, the PC solutions of TEMABF4 and SBPBF4 are

more commonly used as the electrolyte of the capacitor in large pulse power sources.

1.3.4 Classes of advanced capacitor

Table.1 gives the detail information regarding electrode configuration, electrolyte

and operating voltage of advanced capacitors as well as that of EDLC. During charging

and discharging, Li+ cations intercalate and de-intercalate at the intercalation carbon

anode of Li-ion capacitor and dual carbon capacitor. While, PF6- anions intercalate and

de-intercalate at the intercalation carbon cathode of nano-storage capacitor and dual

carbon capacitor. Considering the mechanism of electrode reaction, the operating cell

EDLC

Megalo Cap.

2.5V

Graphitic C

36

voltage and anode potential of each capacitor, it was speculated that Li-ion capacitor

works at the lowest cathode potential and is more favorable for a long term operation

among the capacitors.

In comparison to EDLC, Li-ion capacitor offer higher capacitance, higher cell

voltage, potentially high energy density and lower cost per watt-hour. Advanced

capacitor such as Li-ion capacitor is a relatively new field of development. Considering

the scope of new potential material and combination, it is expected that industrially

advanced capacitors will soon emerge.

1.3.4.1 Aqueous electrolyte, carbon double layer capacitor.

EDLC, carbon double layer capacitor of unit cell voltage for aqueous electrolyte is

limit at 1V. Advantages of the aqueous electrolyte are the higher conductance and both

purification and drying process during production are less stringent. The costs of

aqueous electrolytes are cheaper than organic electrolytes. Some Japanese company,

NEC and ECOND commercialized this kind of capacitors.

1.3.4.2 Nonaqueous electrolyte, carbon double layer capacitor.

Carbon double layer capacitors with non-aqueous electrolytes have some advantage

over aqueous electrolyte, carbon double layer capacitor. It can operate at high voltage of

3.5 to 4.0V, because of the substantially high decomposition voltage of non aqueous

solutions. High energy density and high power density can be achieved. The use of

nonaqueous solution for electrochemical capacitor in order to maximize energy density

has enhanced interest in the specific dependence of double-layer structure and

capacitance on the type of solvent in recent years. It has several indirect aspects to use

nonaqueous solvent for higher voltage capacitors.

37

1. Nonaqueous electrolyte is very difficult to purify and hence contain small but

significant quantities of impurities that can be either electrochemically reduced

or oxidized, leading to self discharge.

2. It has smaller solvating power for ions, ion adsorption at the electrode interface

could be stronger thus modify the double-layer capacitance behavior.

3. Its dielectric constants are usually lower than that of water, specific double

layer capacitance tends to be lower than that in aqueous solution.

4. Due to low dielectric constant, the specific resistivities of electrolyte solution

are usually less. This tends to raise the value of equivalent series resistance in

nonaqueous solution capacitor lead to diminished power densities.

1.3.4.3 Li- ion hybrid capacitors

Hybrid capacitors and a redox capacitor that utilize carbons as a faradaic type

redox electrode materials as shown in Table 1. These advanced capacitors are what we

called Li-ion capacitor, nano-storage capacitor and dual carbon capacitor. Capacitance

of the intercalation carbon electrode is a few times larger than that of the non-faradaic

EDLC. Thus the depth of charge and discharge of the intercalation carbon electrode are

allowed to be shallow, giving them a long cycle life capability.

38

1.4 Target of this research

LiCoO2 is most suitable cathode material for lithium ion battery but it is highly

toxic and expensive as compared to other cathode materials. So, other research group

has introduced cathode materials like spinel LiMn2O4, LiNiO2, LiMnO2, LiNi1/2Mn1/2O2

and LiNi1/3Mn1/3Co1/3O2 to reduce the price of cobalt.

Chapter 2, since o-LiMnO2, brown powder with the ordered rock salt like structure

with an orthorhombic unit cell was first reported by Johnston and Keikes [134] in the

study of LixMn1-xO system. Bongers et al., also prepared this compound and reported its

magnetic susceptibility [135]. Monoclinic LiMnO2 with layered structure is usually

obtained by soft chemical method due to metastable nature, such as ion-exchanged

method [136].

Until now, there are a very few report concerning high capacity maintainable

orthorhombic LiMnO2 synthesized by calcinations of freeze-dried precursor at 950oC

[137]. Ohzuku et al. [138] discovered that orthorhombic LiMnO2 shows specific

capacity near 200mAh g-1 between 2.5 and 4.3V. Reimer et al. [139] prepared samples

of LiMnO2 at low temperature which was more pure form than those of Ohzuku et al.

and which also shows capacity near 200mAh g-1. Y. S. Lee et al. [140] have also

reported a new synthetic method using quenching process to synthesize a

well-crystallized o-LiMnO2 material at 1000oC by one-step method.

I have introduced new type of orthorhombic LiMnO2-yXy material with fluorine and

chlorine doped compounds at 1050ºC by quenching method, which shows good

cycleability at room temperature as well as at elevated temperature (60ºC). I will

describe the structure and electrochemical properties of fluorine and chlorine doped

orthorhombic LiMnO2-yXy compounds prepared by the quenching method. The

39

electrochemical performance of the doped o-LiMnO2-yXy, especially the capacity fade

rate, has been improved at high temperature (60ºC).

In Chapter 3, I have prepared the layered LiNi0.5Mn0.5O2 by carbonate

co-precipitation method. Layered transition metal oxide with hexagonal structure,

LiNi0.5Mn0.5O2 and LiMn0.33Ni0.33Co0.33O2, were introduced by Ohzuku et al. as a

promising cathode material to replace LiCoO2. The combination of Ni, Mn, and Co can

provide better thermal stability and higher reversible capacity than LiCoO2. Although,

LiNi0.5Mn0.5O2 is an attractive cathode material but it is very difficult to prepare such

cathode materials. Many research groups have applied hydroxide co-precipitation

method to prepare the cathode material. Though the hydroxide co-precipitation is one of

the powerful synthesis methods, the precipitated transition metal hydroxide is easily

oxidized in the basic aqueous solution during precipitation process. For example,

Mn(OH)2 is gradually oxidized to MnOOH (Mn3+) or MnO2 (Mn4+) by oxygen under

the precipitation conditions. Such a reaction decreases the homogeneity of the final

product. Therefore, preparation of homogeneous mixed hydroxide precursor with highly

reproducibility is very difficult even it is prepared under inert atmosphere. The main

advantage of this method is that valence of Mn ion was easily kept at Mn2+under

precipitation condition. Therefore, it can be applicable for industrial application. I have

also studied the effect of synthetic condition on structural and electrochemical

properties of cathode materials.

In Chapter 4, EDLC having high energy density capacitor is carried out. I have

introduced AC/Graphite capacitor. The AC/AC capacitor has low energy density

because the big surface area of activated carbon is quite susceptible to anodic electrolyte

decomposition, which in facts limits the working voltage for sake of safety. Usually

40

graphite is not a capacitor material because of its small ability of adsorption. But, I have

found if activated carbon (AC) is used as a counter electrode, cathodic graphite can

adsorb anion on non-aqueous solution. In AC/AC capacitor, one electrode has enough

ability for adsorption and counter electrode does not need such strong ability of

adsorption. So, I have proposed new idea to replace one of its activated carbon by

graphite carbon as a cathode for capacitor.

In this research, I have applied graphitic carbon named KS-6, KS-44, SFG-44 and

SFG-6 (from Timcal Co. Ltd.) having low crystal structure as a cathode material instead

of activated carbon in the activated carbon Graphite/AC capacitor using organic

electrolytes.

41

References

[1]. H. I. Becker, U.S. Patent, 2800616 (1957).

[2]. J. M. Tararascon, in “Recent Advances in Rechargeable Li Batteries”, special

Issue of Solid State Ionics, eds. M. S. Whittingham, 69, No. 3-4 (1994).

[3]. H. Ikeda, T. Saito, and H. Tamura, in Proc. Manganese Dioxide Symp. Vol.1,

eds. A. Kozawa and R. H. Brodd (IC sample Office, Cleveland. OH.1975).

[4]. M. Armand, “Materials for Advanced Batteries”, eds. D. W. Murphy, J.

Broadhead and B. C. H. Steele (Plenum press, New York, 1980) P. 145.

[5]. B. di Pietro, M. Patriarca, and B. Scrosati, J. Power Sources, 8 (1982) 289.

[6]. M. Lazzari, and B. Scrosati, J. Electrochem. Soc., 127 (1980) 773.

[7]. S. Morzilli, B. Scrosati, and F. Sgarlata, Electrochim. Acta, 30 (1985) 1271.

[8]. T. Nagaura, and K. Tozawa, Prog. Batt. Solar cells, 9 (1990) 209.

[9]. A. Yoshino, K. Sanechika, and T. Nakashima, Japanese Patent 62-90863

(1987).

[10]. A. Hooper, and B. C. Tofield, J. Power Sources, 11 (1984) 33.

[11]. F. Croce, S. Passerini, and B. Scrosati, J. Power Sources, 43 (1993) 481.

[12]. R. R. Haering, J. A. R. Stile, and K. Brandt, US patent (1982) No. 4,281,048.

[13]. Y. Toyoguchi, T. Matsui, T. Yamaura, and T. Iijima, 3rd International Meeting

on Lithium Batteries, 1986, 117.

[14]. J. J. Auborn, and Y. L. Barberio, J. Electrochem. Soc., 134 (1987) 638.

[15]. J. B. Goodenough, and V. Manivannan, DENKI KAGAKU, 66 (1998) 1173.

[16]. J. M. Tarascon, and M. Armand, Nature, 414 (2001) 359.

[17]. K. Mizushima, P. C. Jones, P. J. Wiseman, and J. B. Goodenough, Mater. Res.

Bull., 17 (1980) 783.

42

[18]. J. N. Reimers, and J. R. Dahn, J. Electrochem. Soc., 139 (1992) 2091.

[19]. T. Ohzuku, A. Ueda, M. Nagayama, Y. Inakoshi, and H. Konori, Electrochim.

Acta, 38 (1993) 1159.

[20]. E. Plichta, S. Slane, M. Uchiyama, M. Saloman, D. Chua, W. B. Ebner, and W.

H. Lin, J. Electrochem. Soc., 136 (1989) 1865.

[21]. T. Nohma, H. Kurokawa, M. Uehara, M. Takahashi, K. Nisio, and T. Sato, J

Power Sources, 54 (1995) 522.

[22]. J. N. Reimers, and J. R. Dahn, J. Electrochem. Soc., 139 (1992) 2091.

[23]. M. Zou, M. Yoshio, S. Gopukumar, and J. Yamaki, Electrochem. Solid State

Lett., 7 (2004) A176.

[24]. T. Ohzuku, and A. Ueda, J. Electrochem. Soc., 141 (1994) 2972.

[25]. A. Van der Ven, M. K. Aydinol, and G. Ceder, J. Electrochem. Soc., 145 (1998)

2149.

[26]. H. Tukamoto, and A.R. West, J. Electrochem. Soc., 144 (1997) 3164.

[27]. S. Levasseur, M. Ménétrier, and C. Delmas, J. Power Sources, 112 (2002) 419.

[28]. G. Ceder, Y.-M. Chiang, D. R. Sadoway, M. K. Aydinol, Y. I. Jang, and B.

Huang, Nature, 392 (1998) 694.

[29]. Y. I. Jang, B. Huang, H. Wang, D. R. Sadoway, G. Ceder, Y. M. Chiang, H. Liu,

and H. Tamura, J. Electrochem. Soc., 146 (1999) 862.

[30]. H. Huang, G. V. Subba Rao, and B.V. R. Chowdari, J. Power Sources, 81

(1999) 690.

[31]. W. Yoon, K. K. Lee, and K. B. Kim, J. Electrochem. Soc., 147 (2000) 2023.

[32]. S. T. Myung, N. Kumagai, S. Komaba, and H. T. Chung, Solid State Ionics, 139

(2001) 47.

43

[33]. N. Amdouni, H. Zarrouk, F. Soulette, and C. Julien, Mater. Chem. Phys., 80

(2003) 205.

[34]. R. Alcantara, J. C. Jumas, P. Lavela, J. O. Fourcade, C. P. Vincente, and J. L.

Tirado, J. Power Sources, 81-/82 (1999) 547.

[35]. H. Kobayashi, H. Shigemura, M. Tabuchi, H. Sakaebe, K. Ado, H.

Kageyama, A. Hirano, R. Kanno, M. Wakita, S. Morimoto, and S. Nasu, J.

Electrochem. Soc., 147 (2000) 960.

[36]. M. Holzapfel, R. Schreiner, and A. Ott, Eletrochim. Acta, 46 (2001) 103.

[37]. C. Delmas, and I. Saadoune, Solid State Ionics, 53-56 (1992) 370.

[38]. I. Saadoune, and C. Delmas, J. Mater. Chem., 6 (1996) 193.

[39]. T. Ohzuku, and A. Ueda, J. Electrochem. Soc., 144 (1997) 2780.

[40]. J. Cho, G. Kim, and H. S. Lim, J. Electrochem. Soc., 146 (1999) 3571.

[41]. C. Delmas, M. Menetrier, L. Croguennec, I. Saadoune, A. Rougier, C.

Pouillerie, G. Prado, M. Grune, and L. Fournes, Electrochim. Acta, 45 (1999)

243.

[42]. J. Cho, H. S. Jung, Y. C. Park, G. B. Kim, and H. S. Lim, J. EIectrochem. Soc.,

147 (2000) 15.

[43]. S. Madhavi, G. V. Subba Rao, B. V. R. Chowdari, and S. F. Y. Li, J.

Power Sources, 93 (2001) 156.

[44]. R. J. Gummow, and M. M. Thackeray, Solid State Ionics, 53-56 (1992) 681.

[45]. S. Madhavi , G. V. Subba Rao , B. V. R. Chowdari, and S. F. Y. Li,

Electrochim. Acta, 48 (2002) 219.

[46]. R. Stoyanova, E. Zhecheva, and L. Zarkova, Solid State Ionics, 73 (1994) 233.

[47]. A. R. Armstrong, A. D. Robertson, R. Gitzendanner, and P. G. Bruce, J. Solid

44

State Chem., 145 (1999) 549.

[48]. S. Lavassuer, M. Menetrier, E. Suard, and C. Delmas, Solid State Ionics,

128 (2000)11.

[49]. N. Imanishi, M. Fujii, A. Hirano, and Y. Takeda, J. Power Sources, 97-/98

(2001) 287.

[50]. N. Imanishi, M. Fujii, A. Hirano, Y. Takeda, M. Inaba, and Z. Ogumi, Solid

State Ionics, 140 (2001) 45.

[51]. M. Zou, M. Yoshio, S. Gopukumar, and J. Yamaki, Chem. Mater., 15(25)

(2003) 4699.

[52]. M. Zou, M. Yoshio, S. Gopukumar, and J. Yamaki, Mater. Res. Bull., 40

(2005) 708.

[53]. M. Zou, M. Yoshio, S. Gopukumar, and J. Yamaki, Chem. Mater., 17(6)

(2005) 1284.

[54]. J. Cho, Y. J. Kim, and B. Park, J. Electrochem. Soc., 148 (2001) A1110.

[55]. J. Cho, Y. J. Kim, and B. Park, Chem. Mater., 12 (2000) 3788.

[56]. J. R. Dahn, U. Von Sacken, and C. A. Michal, Solid State Ionics, 44 (1990) 87.

[57]. T. Ohzuku, A. Ueda, and M. Nagayama, J. Electrochem. Soc., 140 (1993) 1862.

[58]. H. Arai, S. Okada, Y. Sakurai, and J. Yamaki, Solid State Ionics, 95 (1997) 275.

[59]. H. Arai, S. Okada, Y. Ohtsuka, M. Ichimura, and J. Yamaki, Solid State Ionics,

80 (1995) 261.

[60]. A. Rougier, I. Saadoune, P. Gravereau, P. Willmann, and C. Delmas, Solid State

Ionics, 90 (1996) 83.

[61]. J. R. Dahn, U. von Sacken, and C. A. Michal, Solid State Ionics, 44 (1990) 87.

[62]. T. Ohzuku, A. Ueda, and M. Nagayama, J. EIectrochem. Soc., 140 (1993) 1862.

45

[63]. W. Li, J. N. Reimers, and J. R. Dahn, Solid State Ionics, 67 (1993) 123.

[64]. A. Rougier, I. Saadoune, P. Gravereau, P. Willmann, and C. Delmas, Solid State

Ionics, 90 (1996) 83.

[65]. E. Zhecheva, and R. Stouanova, Solid State Ionics, 66 (1993) 143.

[66]. A. Ueda, and T. Ohzuku, J. EIectrochem. Soc., 141 (1994) 2010.

[67]. C. Delmas, I. Saadoune, and A. Rougier, J. Power Sources, 43-44 (1993)595.

[68]. M. Yoshio, Y. Todorov, K. Yamato, H. Noguchi, J. Itoh, M. Okada, and T.

Mouri, J. Power Sources, 74 (1998) 46.

[69]. M. Yoshio, H. Noguchi, J. Itoh, M. Okada, and T. Mouri, J. Power Sources 90

(2000) 176.

[70]. J. N. Reimers, E. Rossen, C. D. Jones, and J. R. Dahn, Solid State Ionics, 61

(1993) 335.

[71]. G. Prado, A. Rougier, L. Fournès, and C. Delmas, J. Electrochem. Soc., 147

(2000) 2880.

[72]. C. Pouillerie, F. Perton, P. Biensan, J. P. Pérès, M. Broussely, and C. Delmas, J.

Power Sources, 96 (2001) 293.

[73]. T. Kai, H. Ando, Y. Muranaka, T. Horiba, and K. Hironaka, Shin-Kobe Tech.

Rep., 11 (2001) 9.

[74]. T. Horiba, K. Hironaka, T. Matsumura, T. Kai, M. Kosek, and Y. Muranaka, J.

Power Sources, 119-121 (2003) 893.

[75]. S. Megahed, and B. Scrosati, J. Power Sources, 51 (1994) 79.

[76]. M. M. Thackeray, P. J. Johnson, L. A. de Picciotto, P. G. Bruce, and J. B.

Goodenough, Mater. Res. Bull., 19 (1984) 179.

[77]. J. C. Hunter, J. Solid State Chem., 39 (1981) 142.

46

[78]. Y. Xia, and M. Yoshio, J. Electrochem. Soc., 143 (1996) 893.

[79]. Y. Xia, T. Sakai, T. Fujieda, X. Yang, X. Sun, Z. Ma, J. McBreen, and M.

Yoshio, J. Electrochem. Soc., 148 (2001) A723.

[80]. S. J. Wen, T. J. Richardson, L. Ma, K. A. Striebel, P. N. Ross, and E. J.Cairns, J.

Electrochem. Soc., 143 (1996) L136.

[81]. D.-H. Jang, Y.-J. Shin, and S.-M. Oh, J. Electrochem. Soc., 143 (1996) 2204.

[82]. T. Inoue, and M. Sano, J. Electrochem. Soc., 145 (1998) 3704.

[83]. M. M. Thackeray, Y. Shao-Horn, and A. J. Kahian, Electrochem. Soc. Interface,

1 (1998) 7.

[84]. X. Wang, H. Nakamura, and Masaki Yoshio, J. Power Sources, 110 (2002) 19.

[85]. M. Yoshio, H. Noguchi, H. Wang, and X. Wang, J. Power Sources, 154 (2005)

273.

[86]. B. Deng, H. Nakamura, Q. Zhang, M. Yoshio, and Y. Xia, Electrochim. Acta, 49

(2004) 1823.

[87]. S. Komaba, B. Kaplan, T. Ohtsuka, Y. Kataoka, N. Kumagai, and H. Groult, J.

Power Sources, 119-121 (2003) 378.

[88]. N. Kumagai, S. Komaba, Y. Kataoka, and M. Koyanagi, Chem. Lett., 2000,

1154.

[89]. H. Tunekawa, S. Tanimoto, R. Marubayashi, M. Fujita, K. Kifune, and M. Sano,

J. Electorchem. Soc., 149 (2002) A1326.

[90]. Y.-K. Sun, Y.-S. Lee, M. Yoshio, and K. Amine, Electrochem. Solid-State Lett.,

5 (2002) A99.

[91]. B. Deng, H. Nakamura, and M. Yoshio, Chem. Lett., 32 (2003) 942.

[92]. R. Hoppe, G. Brachtel, M. Jansen, and Z. Anorg, Allg. Chem., 417 (1975) 1.

47

[93]. A. R. Armstrong, and P. G. Bruce, Nature, 381 (1996) 499.

[94]. Y.-I, Jang, B. Huang, Y.-M. Chiang, and D. R. Sadoway, Electrochem.

Solid-State Lett., 1 (1998) 13.

[95]. A. K. Padhi, K. S. Najundaswamy, and J. B. Goodenough, J. Electrochem. Soc.,

144 (1997) 1188.

[96]. D. D. MacNeil, Z. Lu, Z. Chen, J. R. Dahn, J. Power Sources, 108 (2002) 8.

[97]. M. Takahashi, S. Tobishima, K. Takei, and Y. Sakurai, Solid State Ionics, 148

(2002) 283.

[98]. A. K. Padhi, K. S. Nanjundaswamy, C. Masquelier, and J. B. Goodenough, J.

Electrochem. Soc., 144 (1997) 2581.

[99]. A. Manthiram, and J. B. Goodenough, J. Power Sources, 26 (1989) 403.

[100]. S. Andersson, J. O. Thomas, B. Kalska, and L. Häggström, Electrochem.

Solid-State Lett., 3 (2000) 66.

[101]. N. Ravet, J. B. Goodenough, S. Besner, M. Simoneau, P. Hovington, andM.

Armand, Abstract 127, The Electrochemical Society and The Electrochemical

Society of Japan Meeting, Vol. 99-2, Honolulu, HI, Oct 17-22, 1999.

[102]. H. Huang, S. C. Yin, and L. F. Nazar, Electrochem. Solid-State Lett., 4 (2001)

A170.

[103]. Z. Chen, and J. R. Dahn, J. Electrochem. Soc., 149 (2002) A1184.

[104]. A. Yamada, S. C. Chung, and K. Hinokuma, J. Electrochem. Soc., 148 (2001)

A224.

[105]. Encyclopedia Britannica, Vol. 6, COLE to DAMA, 14th edn., 1926, NewYork,

p.216.

[106]. R. Kotz, and M.Carlen, Electrochemica Acta, 45 (2000) 2483.

48

[107]. D. Linden, ed., Handbook of Batteries, 2nd ed., McGraw-Hill, New York (1995).

[108]. A. G. Pandolfo, A. M. Vasallo, and G. L. Paul, Proceeding of the 7th

International Seminar on Double Layer Capacitor and Similar Energy Storage

Devices, Florida Educational Seminar, December 1997.

[109]. S. Trasatti, P. Kurzweil, Platinum Met. Rev., 38 (1994) 46.

[110]. J. P. Zheng, P. J. Cygan, and T. R. Jow, J. Electrochem. Soc., 142 (1995)2699.

[111]. S. Ardizzone, G. Fregonara, and S. Trasatti, Electrochim. Acta, 35 (1990)263.

[112]. R. Kotz, and S. Stucki, Electrochim. Acta, 31(1986) 1311.

[113]. T. J. Guther, R. Oesten, J. Garche, in: F. M. Delnick, D. Ingersoll, X. Andrieu,

and K. Naoi (Eds.) Electrochemical capacitors II, Proceedings, vols. 96-25,

The Electrochemical Society, Pennington, NJ, 1997, p. 16.

[114]. T. C. Liu, W. G. Pell, B. E. Conway, and S.L. Roberson, J. Electrochem.Soc.,

145 (1998) 1882.

[115]. X. Ren, S. Gottesfeld, J. P. Ferrasis in: F. M. Delnick, M. Tomkiewicz (Eds.),

Electrochemical Capacitors, Proceeding, vols. 95-29, The Electrochemical

Society, Pennington, NJ, 1996, p. 138.

[116]. C. Arbizzani, M. Mastragostino, and L. Meneghello, Electrochim. Acta, 41

(1996) 21.

[117]. K. Naoi, Extended Abstracts, 49th Annual Meeting of the International Society

of Electrochemistry, Kitakyushu, Japan, September, pp. 647, 1998

[118]. Y. Kibi, T. Saito, M. Kurata, J. Tabuchi, and A. Oichi, J. Power Sources, 60

(1996) 219.

[119]. R. Kotz, M. Bartsch, A. Braun, A. Haas, Proceeding of The 7th Internati-onal

Seminar on Double Layer Capacitors and Similar Energy Storage Devices,

49

Florida Educational Seminar, December 1997.

[120]. J. P. Zheng, and T. R. Jow, J. Electrochem. Soc., 144 (1997) 2417.

[121]. T. Morimoto, K. Hiratsuka, Y. Sanada, and K. Kurihara, Mat. Res. Soc. Proc.

Vol. 393, p. 397 (1995).

[122]. M. Ue, K Ida, and S. Mori, J. Electrochem. Soc., 141 (1994) 2989.

[123]. M. Ue, J. Electrochem. Soc., 141 (1994) 3336.

[124]. Y. Matsuda, M. Morita, M. Ishikawa, and M. Ihara, J. Electrochem. Soc., 140

(1993) L109.

[125]. M. Ishikawa, M. Morita, M. Ihara, and Y. Matsuda, J. Electrochem. Soc., 141

(1994) 1730.

[126]. C. Arbizzani, M. Mastragostino, L. Meneghello, and R. Paraventi, Adv. Mater.,

8 (1996) 331.

[127]. E. I. Copper, and J. M. O’Sullivan, Proceeding of the 8th International

Symposium on Molten Salts, R. J. Gale, G. Blomgren, and H. Kojima, Editors,

PV 92-16, p. 386, The Electrochemical Society: Pennington, NJ (1992).

[128]. C. L. Hussey, Chemistry of Nonaqueous Solvents, A. Popov, and G. Mamantov,

Editor, Chap. 4, VCH Publishers, NY, (1994). R. T. Carlin and J.S. Wilkes,

Chemistry of Nonaqueous Solvents, A. Popov, and G. Mamantov, Editor, Chap.

5, VCH Publishers, NY, (1994).

[129]. P. Bonhote, A.-P Dias, K. K. Papageorgiou, and M. Gratzel, Inorg. Chem., 35

(1996) 1168.

[130]. J. Fuller, R. T. Carlin, H. C. De Long, and D. Haworth, J. Chem. Soc. Chem.

Commun., (1994) 299.

[131]. R. T. Carlin, H. C. DeLong, J. Fuller, and P. C. Trulove, J. Electrochem. Soc.,

50

141, (1994) L73.

[132]. V. R. Koch, L. A. Dominey, C. Nanjundiah, and M. J. Ondrechen, J.

Electrochem. Soc., 143 (1996) 788.

[133]. T. Morimoto, Proceeding of International Conference on Advanced Capacitors

2007, p. 7, Kyoto, May (2007).

[134]. W. D. Johnston and R. R. Heikes, J. Am. Chem. Soc., 78 (1956) 3255.

[135]. P. F. Bongers, PhD dissertation, The University of Leiden, Leiden, The

Netherland (1957).

[136]. A. R. Armstrong, and P. G. Bruce, Nature, 381 (1996) 499.

[137]. Y. I. Jang, B. Huang, H. Wang, D. R. Sadoway, Y. M. Chiang, J. Electrochem.

Soc., 146 (1999) 3217.

[138]. T. Ohzuku, Chemistry Express, 5 (1990) 733.

[139]. J. N. Reimers, E. W. Fuller, E. Rossen, and J. R. Dahn, J. Electrochem. Soc.,

140 (1993) 3396.

[140]. Y. S. Lee, Y. K. Sun, K. Adachi, and M. Yoshio, Electrochim. Acta, 48

(2003)1031.

51

Chapter 2

Synthesis and electrochemical performance of o-LiMnO2-yXy

(0.01≤ y ≤ 0.05) cathode material at high Temperature

2.1 Introduction

Among the promising candidates of the cathode materials for secondary lithium

batteries, such as LiMn2O4 [1, 2, 3], LiNi0.8Co0.2O2 [4, 5], LiNi1/3Co1/3Mn1/3O2 [6, 7],

LiFePO4 [8], and LiMnO2 [9, 10], o-LiMnO2 has theoretical capacity of 285 mAh g-1

and better cycleability than spinel LiMn2O4 over a wide voltage range of 2.0-4.5V [11,

12].

According to the synthetic conditions, LiMnO2 forms orthorhombic and monoclinic

phases with the space group Pmnm (a = 2.805 Å, b = 5.757 Å, and c = 4.527 Å) and a

monoclinic unit cell with the space group C/2m (a = 5.439 Å, b = 2.809 Å and c = 5.388

Å) [13, 14]. It has been reported that o-LiMnO2 transforms to a phase with a spinel-like

structure during cycling [15]. The dissolution of manganese is the cause for poor

cycling performance at the elevated temperature and leads to defective spinels [16]. To

stabilize the structural instability for elevated-temperature performance, cation

substitution such as o-LiAl1-xMnxO2 has been studied. While capacity retention was

found to be better than that of o-LiMnO2 on cycling test at 55°C, Al substitution

(o-LiAl1-xMnxO2) did not prevent John-Teller distortion [17]. An alternative approach to

improving the structural stability is by coating the o-LiMnO2 surface with metal oxides,

which might minimize the structural instability from HF attack [18, 19].

52

Since o-LiMnO2, brown powder with the ordered rock salt like structure with an

orthorhombic unit cell was first reported by Johnston and Keikes [20] during the study

of LixMn1-xO system. Bongers et al., also prepared this compound and reported its

magnetic susceptibility [21]. o-LiMnO2 materials have generally been prepared by a

solid state reaction method by heating the reaction mixtures of manganese oxides and

lithium salts under inner atmosphere at high temperature of [22, 23, 24] or heating an

equimolar mixture of γ-MnOOH and LiOH at temperature between 300 and 450 ºC

under nitrogen or argon atmosphere [25]. Wet chemical processes of ion exchange,

hydrothermal [26, 27], sol-gel [28], and reverse-micro emulsion [29] methods were also

applied to synthesize o-LiMnO2. However, the electrochemical behavior of the

o-LiMnO2 is greatly dependent on the synthetic route.

Recently, improvements on cycleability through the co-doping of cations (Li, Mg, Al,

etc) and anions (F, S, etc) into LiMn2O4- or LiNiO2- based cathode have been reported

[30, 31, 32]. Dahn et al. [33] obtained dense LiNixCo1-2xMnxO2 material at 900 ºC by

adding LiF as a sintering agent. In this section, I have synthesized o-LiMnO2-yXy (0.01

≤ y ≤ 0.05) at 1050 ºC for 10 h by quenching method. The enhancement of the

electrochemical properties of o-LiMnO2 cathodes by LiF and LiCl doped were

investigated.

2.2 Experimental

2.2.1 Preparation of orthorhombic LiMnO2-yXy

A new type of orthorhombic LiMnO2-yXy material was prepared using LiOH.H2O

(Osaka Kishida Chemical, Japan), γ-MnOOH (Tosoh Chemical, Japan), LiF (Osaka

Kishida Chemical, Japan) or LiCl (Sigma Aldrich, Japan). The mixture of LiOH.H2O,

53

γ-MnOOH and LiF or LiCl (molar ratio Li/Mn = 1) was thoroughly ground in an agate

mortar. The mixture was pre-calcinated at 470 ºC and 550 ºC for 5 h in air. The mixture

was pressed at 1 ton cm-2 pressure into a 25 mm diameter pellet to improve the

reactivity between the particles of the precursor. After that, the mixture was post

calcinated at 1050ºC for 10 h under flowing argon and then removed from the furnace

and rapidly quenched in liquid nitrogen.

2.2.2 XRD, SEM, BET and Chemical analysis

The powder X-ray diffraction (XRD; Rint 1000, Rigaku Ltd., Japan) using CuKα

radiation was employed to identify the crystalline phase of the prepared materials. The

N2 gas absorption isotherm was measured using a Gemini 2375 instrument and the

specific surface area was calculated by the Brunauer, Emmett and Teller (BET) method.

The particle morphologies of the resulting compounds were observed using a scanning

electron microscope (SEM, JSM-5300E, Japan Electron, Ltd., Japan). For the average

oxidation of Mn in the synthesized compound, titration of excess Fe2+ ion by a standard

KMnO4 solution was carried out. To determine the contents of fluorine and chlorine in

the synthesized material, ion chromatography was measured using a DIONEX

ICS-1500 instrument, Japan.

2.2.3 Electrochemical performance

The electrochemical characterizations were carried out using a CR2032 coin-type

cells. A cathode which consisted of 20 mg of accurately weighed active material and 12

mg of conductive binder (Teflonized acetylene black) was pressed onto a 2 cm2 stainless

steel mesh used as the current collector at a pressure of 2 ton cm-2 and dried at 170 ºC

54

for 3 h in a vacuum oven. The tested cell consisted of a cathode with active material

and a lithium metal anode separated by a glass fiber (ADVANTEC, GA-100).The

electrolyte used was a 1 M LiPF6-ethylene carbonate (EC) - dimethyl carbonate (DMC)

(1:2 by volume Ube Chemicals, Japan). All the assembling of the cell was carried out in

an argon-filled dry box. Charge-discharge test of cell was carried out by using the

instrument Nagano, (BTS2004W, Japan). The charge discharge cycling was carried out

at a current density of 0.4 mA cm-2 with a voltage range of 2.0-4.5 V at both room

temperature and high temperature of 60ºC. Cyclic voltammetry study was conducted

using a three-electrode cell where lithium foil was used as both the counter and

reference electrodes. The working electrode consisted of 2.5 mg of the active material

and 1.5 mg of conducting binder (TAB-2), which was pressed onto a stainless steel

mesh. CV measurements were performed using a HSV-100 (Hokuto Denko Ltd., Japan).

CV experiments were carried out at a scan speed of 0.1 mV/min between 2.0 and 4.5 V

vs. Li/Li+ at high temperature (60ºC).

2.3 Result and Discussion

2.3.1 XRD, SEM and Chemical analysis

Fig. 2.1 shows the XRD pattern of the LiMnO2-yXy materials calcined at 1050ºC.

All the samples were prepared at 1050ºC for 10 h by the quenching method. Quenching

treatment is one of the best ways to preserve its original LiMnO2-yXy structure, which

prevents structural transformation into the spinel-like phase during the cooling process

[34]. X-ray diffraction data of the calcined powder were carefully collected in the 2θ

range of 10-80º. The entire sample contains small amount of the Li2MnO3 which is

55

express by an asterisk in figure. It has already confirmed that the mixed material shows

the excellent electrochemical performance.

Figure 2.1 X-ray diffraction patterns of the synthesized o-LiMnO2-yXy cathode materials.

The unit cell parameters for the orthorhombic LiMnO2-yXy phase and average

oxidation state of Mn in mixed phase are reported in Table 2.1. All the halogen doped

o-LiMnO2-yXy shows nearly similar parameter values. Anion substitution does not cause

10 20 30 40 50 60 70 80

(e)LiMnO1.95Cl0.05

*

2θ / deg (CuKα)

(d)LiMnO1.99Cl0.01

*

(c)LiMnO1.95F0.05

*

Inte

nsity

(a.u

.)

(b)LiMnO1.99F0.01

*

(a)

o-LiMnO2* Li2MnO3

*

(042

)

(040

)(1

22)

(130

)(120

)(0

21)

(111

)(002

)(0

21)

(101

)(1

10)(011

)(010

)

56

any particular change in the lattice constants (a, b and c) of o-LiMnO2-yXy. The

substitution of oxygen ion by monovalent halogen causes the decrease of Mn valence.

However, all the prepared samples showed oxidation states of Mn more than 3+. It

would be doubtful weather oxygen ions were substituted by halogen or not. I have

measured the halogen content and the data were shown in Table 2.2. This result

corresponds to the presence of Li2MnO3 as shown Fig 2.1.

Table 2.1 Lattice parameter and oxidation state of Mn of LiMnO2-yXy.

Furthermore, we performed a chemical analysis of the prepared samples using ion

chromatography in order to confirm whether was fluorine and chlorine are doped. It was

revealed that the fluorine-doped sample contains LiMnO1.996F0.004 and LiMnO1.97F0.03

while the chlorine doped one was LiMnO1.997Cl0.003 and LiMnO1.98Cl0.02 as shown in

Table 2.2. During the calcinations, halogen doped compounds get evaporated with

oxygen at high temperature that lead to the crystal growth of o-LiMnO2-yXy. We

observed that the chlorine-doped compound is more severely affected during the

calcinations as compared to the fluorine-doped o-LiMnO2-yXy based on the analysis.

MnO2 + LiCl MnOOLi + 1/2 Cl2

MnOOH + Li+ + Cl- MnOOLi + HCl

Composition a(Å) b(Å) C(Å) (Å)3 Oxidation of Mnn+

o-LiMnO2 2.805 5.757 4.525 73.912 3.038 LiMnO1.99F0.01 2.803 5.745 4.565 73.511 3.062 LiMnO1.95F0.05 2.802 5.742 4.567 73.479 3.040 LiMnO1.99Cl0.01 2.798 5.729 4.577 73.368 3.062 LiMnO1.95Cl0.05 2.808 5.760 4.585 74.158 3.051

57

Table 2.2 Doped y value after chemical of LiMnO2-yXy.

Fig. 2.2 shows the scanning electron micrographs (SEM) of the o-LiMnO2-yXy

compounds calcined at 1050°C for 10 h. These samples consist of particles of ca.

8-10µm in diameter with somewhat round and bar-shaped like small spherical particles

of about 2-3 µm. As the fluorine and chlorine contents increased, the particle size also

increased in diameter. A similar morphological change was previously reported in the

F-doped LiMn2O4 [35]. It is well known that LiF and LiCl were effective mineralizing

agents and the small quantity of LiF and LiCl could accelerate the crystal growth of the

powder with an enlarged particles size. This indicates that the particle shape and size of

the prepared powders could be controlled by the doped fluorine and chlorine contents.

Mixed composition F amount Cl amount Initial composition 0.01 0.05 0.01 0.05

Found 0.004 0.03 0.003 0.02

58

Figure 2.2 Scanning electron micrograph (SEM) image and surface area of

o-LiMnO2-yXy powder (a) o-LiMnO1.996F0.004 (b) o-LiMnO1.997Cl0.003 (c) o-LiMnO2.

0.53 m2g-1

0.49 m2g-1

0.55 m2g-1

59

2.3.2 Electrochemical performance

Fig. 2.3 shows the first, tenth, twentieth, thirtieth and fiftieth discharge curves of

the o-LiMnO2-yXy electrode at room temperature. The data were collected at a current

density of 0.4 mA cm-2 between 4.5 and 2.0 V. However, during cycling, the discharge

capacity further increased and reached 114 mAh g-1 after 50 cycles. Upon cycling, the

profile of the discharge curve increasingly obtained a spinel-like character with the

development of two distinct plateaus, one at 4 V and the other at 3 V, which are in good

agreement with the results of J. M. Tarascon et al [36]. The discharge curves of the

o-LiMnO2-yXy are indication of Li intercalation on different sites, tetrahedral site over 4

V and octahedral site over 3 V into the cycle-induced spinel like LiMn2O4 [37]. After 10

cycle o-LiMnO1.997Cl0.003 shows spinel like discharge curves

0 40 80 120

2.0

2.5

3.0

3.5

4.0

4.5

50

3020101

Cel

l vol

tage

/ V

Capacity / mAh g-1

Figure 2.3 Discharge curves of o-LiMnO1.997Cl0.003 at room temperature.

Fig. 2.4 shows the cyclic performance of the o-LiMnO2-yXy at room temperature. The

doped o-LiMnO2-yXy electrodes, which was cycled at room temperature showed a very

small initial discharge capacity of about 35 mAh g-1 in the first cycle and slowly

60

increased in capacity upon cycling. It produces a higher capacity than that previously

reported by Croguennec et al. [20]. The obtained capacities as well as the cycleability

are better than the o-LiMnO2 previously reported by other groups [15, 17, 34,]. The

excellent cycleability is probably due to the formation of the particle oxide. For both

0 10 20 30 40 500

50

100

150

(a)

o -LiMnO2

Cycle number

0

50

100

150

(b)

LiMnO1.996F0.004

Cap

acity

/ m

Ah

g-1

0

50

100

150

(c)

LiMnO1.997 Cl0.003

Figure 2.4 Cycle performance of o-LiMnO2-yXy at room temperature.

doping o-LiMnO2-yXy, although their initial capacities are different, all samples shows a

good cyclability at room temperature. Compared to o-LiMnO2, every doping element

can improve the electrochemical performance of the o-LiMnO2-yXy cathode.

Furthermore, Cronguennec et al. [40] reported the effect of the grain sizes for o-LiMnO2

on the various discharge capacities as previously described. Therefore, we assumed that

61

the grinding treatment is a very useful way to change the surface area, which can affect

the characteristic between the electrolyte and particle surface of the o-LiMnO2 material.

In order to increase the specific surface area of the o-LiMnO2-yXy compound, it was

thoroughly ground in a ball-milling machine (ORIENTAL MOTOR CO. LTD., Japan).

0 5 10 15 20 25100

150

200

250

300

0

1

2

3

4

5

6

Spec

ific

surf

ace

area

/ m

2 g-1

Capacity

Firs

t disc

harg

e ca

paci

ty /m

Ah

g-1

Grinding time / h

Surface area

Figure 2.5 The relation between dependence of first specific discharge capacity at 60oC and specific surface area of LiMnO1.997Cl0.003 compounds on the grinding time.

Fig. 2.5 shows the relation between the initial discharge capacity and specific surface

area of the o-LiMnO1.997Cl0.003 compound which depends on the grinding time. The

surface area of the compound before grinding was 0.53 m2 g-1. After a 12 h grinding, it

had a surface area of about 3.3 m2 g-1. The o-LiMnO1.997Cl0.003 compound before

grinding showed the initial discharge capacity of about 124 mAh g-1 in the first cycle at

60ºC. The samples, which were ball-milled for a shorter time, shows a significant

increase in capacity upon cycling as reported by J. Kim and A. Manthiram et al. [41, 42],

while those ball-milled for a longer time show a slight increase in capacity. J. Kim and

A. Manthiram et al. suggested that a moderate ball-milling time improves the electrical

conductivity through a better dispersion of the particles and possibly the Li+-ion

62

conductivity through a decrease in the particle size of the active material, and thereby

leads to a better electrochemical utilization of the cathode even during the initial cycles.

Based on this viewpoint, we can say that our result shows the similar behavior after

grinding as reported by J. Kim et al.

Fig. 2.6 shows the scanning electron microscope (SEM) image of the o-LiMnO2-yXy

material after a 12 h grinding. The o-LiMnO2-yXy compound before grinding had a

Figure 2.6 SEM of o-LiMnO2-yXy materials after 12 h grinding (a) o-LiMnO1.996F0.004 (b) o- LiMnO1.997Cl0.003.

63

typical bar shape-like structure as we mentioned earlier. However, after grinding, the

particle size and shape of the o-LiMnO2-yXy suddenly decreased and changed compared

with those of the original o-LiMnO2-yXy. The BET analysis strongly supports the

grinding effect for LiMnO2.

0 50 100 150 200 2502

4

1st

Capacity / mAh g-1

2

4

20th

Vol

tage

/ V

2

4

30th

LiMnO1.997Cl0.003

Figure 2.7 The first charge discharge curve of LiMnO1.997Cl0.003 at 60oC.

The electrochemical performance of the o-LiMnO2-yXy at the elevated temperature

(60ºC) is shown in Fig. 2.7. The initial charge-discharge capacities showed a significant

increase in capacities in comparison with cycling at room temperature. The discharge

profiles follow an analogous shape as reported for λ-Li2Mn2O4 then it shows a first order

64

phase transition occurring in the relatively flat lower plateau near 3 V and a

homogenous phase transition occurring in the more sloping upper plateau near 4V [9].

The hysteresis in the voltage profile is about the same regardless of the voltage limit

upon charge. The lithium removed from the cathode near 3.8 V, on charge, is re-inserted

on discharge in the 2.9 V plateau is shown to be strongly rate dependent but

independent of the rate over the current range [38].

In order to determine the relation between the discharge capacity and the testing

temperature, we tested the Li/LiMnO2-yXy cell at 60ºC. The initial discharge capacity

increased with the increase in temperature as shown in Fig. 2.8. However, after a 12 h

grinding the o-LiMnO1.997Cl0.003 material showed the highest initial discharge capacity

of 232 mAh g-1 in the first cycle and delivered 231 mAh g-1 after 51 cycles at 60°C. The

cycle retention rate was 100%, and the fluorine doped compound o-LiMnO1.996F0.004

showed initial discharge capacities of about 226 mAh g-1 in the first cycle and 222 mAh

g-1 after 50 cycles at 60ºC, with a cycle retention rate of 98.23% after 50 cycles. On the

other hand, undoped o-LiMnO2 had a discharge capacity of 220 mAh g-1 in the second

cycle and 199 mAh g-1 after 50 cycles at 60ºC. The cycle retention rate was 90% after

50 cycles. The capacity fading rate was higher for the undoped o-LiMnO2. It seems

that the o-LiMnO2-yXy material can be cycled both over the 4 and 3 V ranges without

any significant capacity fading upon cycling. The activation at higher temperature

delivered a much higher capacity as well as good cycleability upon cycling without

capacity fading, because the relatively higher temperature activation contributes to the

much faster phase transformation. The cycle-induced spinel-like phase, which was

necessary to showed a high capacity [39].

65

0 10 20 30 40 500

50

100

150

200

250

(a)

Cycle number

0

50

100

150

200

250

(b)

Cap

acity

/mA

h g-1 0

50

100

150

200

250

(c)

Figure 2.8 Cycle performances of (a) o-LiMnO2, (b) o-LiMnO1.996F0.004 and (c) o-LiMnO1.997Cl0.003 at high temperature (60ºC).

To know the effect of grinding, we measured the cyclic voltammograms of

o-LiMnO1.996F0.004 and o-LiMnO1.997Cl0.003 before and after grinding. The

potential was scanned at the rate of 0.1 mV/min between 2.0-4.5 V at high temperature

(60ºC). Fig.2.9a shows the electrochemical reaction of the o-LiMnO1.996F0.004 before

grinding, which clearly indicated three oxidation peaks at 3.050 V, 3.900 V and 4.150 V,

and three reduction peaks at 4.100 V, 3.850 V and 2.800 V. It shows clearly that it takes

part in the electrochemical reaction in the 3 V regions. However, after 12 h grinding, it

exhibits three oxidation and three reduction peaks, with long and sharp peaks in the

66

same voltage range. Specially, the oxidation peak at 3.050 V and reduction peak at 2.80

V had remarkably increased after grinding. But, we did not found such a big difference

between o-LiMnO1.996F0.004 and o-LiMnO1.997Cl0.003 before and after grinding. In Fig.2.9

2.0 2.5 3.0 3.5 4.0 4.5

-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3 (a)

12 H.G.

0 H.G.

Cur

rent

/ m

A

Cell voltage / V

2.0 2.5 3.0 3.5 4.0 4.5

-0.5

-0.4

-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3(b)

0 H.G.

12 H.G.

Cur

rent

/ m

A

Cell voltage / V

Figure 2.9 CV curves of o-LiMnO2-yXy materials before grinding and after 12 h grinding (a) o-LiMnO1.996F0.004 (b) o-LiMnO1.997Cl0.003. The potential was scanned at the rate of 0.1 mV/min between 2.0-4.5 V at 60°C.

(a) and (b), both the 4 V and 3 V reduction peaks, this correlated with the lithium

67

insertion in the tetrahedral and the octahedral sites in the spinel-like structure [17]. Kim

et al. [43] also reported that the transformation of the orthorhombic into the spinel-like

structure becomes faster with the increase in the number of stacking faults. o-LiMnO2

with a low number of stacking faults displays a better cycleability when cycled between

2.0 and 4.5 V.

2.3.4 Ex-situ XRD after charge-discharge

Fig. 2.10 shows the ex-situ XRD results for the sample after the first

charge-discharge of o-LiMnO2 and after the 50 discharge cycles of the

o-LiMnO1.997Cl0.003. After the first charge, undoped o-LiMnO2 shows a distinct (010)

peak start to disappear and (120) peak may appear to be spinel. After discharge of

o-LiMnO2, it shows (112) spinel like peak starts to appear. In the case of

o-LiMnO1.98Cl0.02 after first charge, it retains pure XRD pattern. After discharge of

o-LiMnO1.98Cl0.02, it all the peaks start to disappear and (112) spinel like peak starts to

appear. After discharge of the 50 cycles, the ex-situ XRD profile showed that the (010)

peak totally disappeared but the other peak showed the same pattern and no spinel peak

was clearly observed. Except for 2θ at 55°, after the discharge, the peak corresponds to

that of the spinel tetragonal Li2Mn2O4 at 2 V. This suggests that the phase

transformation is dependent on the crystal/crystallite size, and that a large size results in

unreacted o-LiMnO1.997Cl0.003 after the 50 cycle. It is found that the transformation from

the orthorhombic structure to the spinel-like structure becomes easier or faster with the

increase in the degree of stacking faults [36].

68

10 20 30 40 50 60 70 80

o Spinel

o(0

11)

(010

)

* Graphite*

(112

)

(120

)

(002

)LiMnO1.95Cl0.05

After 50 cycle discharge of LiMnO1.997Cl0.003

After discharge LiMnO1.98Cl0.02

After charge LiMnO1.98Cl0.02

After discharge LiMnO2

After charge LiMnO2

Inte

nsity

(a.u

.)

2θ /deg (CuKα)

Figure 2.10 Ex-situ X-ray data for the first charge discharged and fiftieth discharged sample at high temperature (60ºC) using the test cell.

69

2.4 Conclusion

o-LiMnO2-yXy compounds were easily prepared using LiOH·H2O, γ-MnOOH, LiF

or LiCl as the starting materials at 1050ºC in flowing argon by a quenching method. An

XRD analysis revealed that the compound showed an orthorhombic phase of space

group Pmnm. Halogen doping o-LiMnO2-yXy shows higher crystallity than undoped

o-LiMnO2, it could be clearly seen by XRD and SEM. However, after 12 h grinding,

o- LiMnO1.997Cl0.003 delivered capacities about 232 mAh g-1 in the first cycle and 231

mAh g-1 after 50 cycles at 60°C. Halogen doped o-LiMnO2-yXy shows higher capacity

than the other commercially used LiCoO2 and LiMn2O4 in Li-ion cell. The synthesized

material can be cycled in both the 3 and 4 V regions without any significant capacity

fading upon cycling. The doping element had a significant influence on the crystal

morphology of the o-LiMnO2-yXy. Normally, the samples are harder thereby making it

difficult to grind. The charge /discharge test showed that the capacity fade rate for the

doped o-LiMnO2-yXy compounds is better than that of the non-substituted compounds.

Moreover, the o-LiMnO2-yXy shows an excellent cyclic performance when it was

operated at elevated temperature. Therefore, o-LiMnO2-yXy cathode is an alternative

cathode for lithium ion battery.

70

References

1. M. M. Thackeray, P. J. Johnson, L. A. De Picciotto, P. G. Bruce, and J. B.

Goodenough, Mater. Res. Bull., 19 (1984) 179.

2. T. Ohzuku, M. Kitagawa, and T. Hirai, J. Electrochem. Soc., 137 (1990) 769.

3. J. M. Tarascon, and D. Guyomard, Electrochim. Acta., 38 (1993) 1121.

4. F. Nobili, F. Croce, B. Scrosati, and R. Marassi, Chem. Mater., 13 (2001) 1642.

5. J. Ying, C. Wan, C. Jiang, and Y. Li, J. Power Sources, 99 (2001) 78.

6. N. Yabuuchi, and T. Ohzuku, J. Power Sources, 119-121 (2003) 171.

7. I. Belharouak, Y. K. Sun, J. Liu, and K. Amine, J. Power Sources, 123 (2003) 247.

8. K. Padhi, K. S. Nanjundaswamy, and J. B. Goodenough, J. Electrochem. Soc., 144

(1997) 1188.

9. I. J. Davidson, R.S. McMillan, I. J. Murray, and J. E. Greedan, J. Power Sources,

54 (1995) 232.

10. A. J. Peterson, A. D. Robertson, and P. G. Bruce, Electrochem. Solid-State Lett., 7

(2004) A331.

11. L. Croguennec, P. Deniard, R. Brec, and A. Lecerf, J. Mater. Chem., 7 (1997) 511.

12. Y.-M. Chiang, D. R. Sadoway, Y.-I. Jang, B. Huang, and H. Wang, Electrochem.

Solid-State Lett., 2 (1999) 107.

13. R. J. Gummow, D.C. Liles, and M. M. Thackeray, Mater. Res. Bull., 28 (1993)

1249.

14. A. R. Armstrong, P. G. Bruce, Nature, 381 (1996) 499.

15. J. N. Reimers, E. W. Fuller, E. Rossen, and J. R. Dahn, J. Electrochem. Soc., 140

(1993) 3396.

16. A. Blyr, C. Sigala, G. G. Amatucci, D. Guyomard, Y. Chabre, and J. M. Tarascon, J.

71

Electrochem. Soc., 145 (1998) 194.

17. Y. I. Jang, B. Huang, H. Wang, D. R. Sadoway, and Y. M. Chiang, J. Electrochem.

Soc., 146, (1999) 3217.

18. J. Cho, Y. J. Kim, T.-J. Kim, and B. Park, Chem. Mater., 13 (2001) 18.

19. J. Cho, Y. J. Kim, T.-J. Kim, and B. Park, J. Electrochem. Soc., 149 (2002) A127.

20. W. D. Johnston, and R.R. Heikes, J. Am. Chem. Soc., 78 (1956) 3255.

21. Bongers P.F., PhD dissertation, The University of Leiden, Leiden, The Netherland

(1957).

22. Y. I. Jang, and Y. M. Chiang, Solid State Ionics, 130 (2000) 53.

23. L. Croguennec, P. Deniard, R. Brec, P. Biensan, and M. Broussely, Solid State

Ionics, 89 (1996) 127.

24. T. Ohzuku, A. Ueda, and T. Hirai, Chem. Express, 7 (1992) 193.

25. G. Pistoia, A. Antonini, and D. Zane, Solid State Ionics, 78 (1995) 115.

26. M. Tabuchi, K. Ado, C. Masquelier, I. Matsubara, H. Sakaebe, H. Kageyama, H.

Kobayashi, R. Kanno, and O. Nakamura, Solid State Ionics, 89 (1996) 53.

27. S. T. Myung, S. Komaba, and N. Kumagai, Electrochim. Acta, 47 (2002) 3287.

28. Z. P. Gou, K. Konstantinov, G. X. Wang, H. K. Liu, and S. X. Dou, J. Power

Sources, 119-121 (2003) 221.

29. C. H. Lu, and H. C. Wang, J. Eur. Ceram. Soc., 24 (2004) 717.

30. P. S. Whitfield, and I. J. Davidson, J. Electrochem. Soc., 147 (2000) 4476.

31. W. Xiaomei, Z. Xianfu, Y. Qinghe, J. Jhongkao, and W. Haoqing, J. Fluorine Chem.,

107 (2001) 39.

32. K. Kubo, M. Fujiwara, S. Yamada, S. Arai, and M. Kanda, J. Power Sources, 68

(1997) 553.

72

33. S. Jouanneau, and J. R. Dahn, J. Electrochem. Soc., 151 (2004) 1749.

34. Y. S. Lee, Y. K. Sun, K. Adachi, and M. Yoshio, Electrochim. Acta., 48 (2003)

1031-1039.

35. G. G. Amatucci, N. Pereira, T. Zhang, and J. M. Tarascon, J. Power Sources, 81-82,

(2004) 39-43.

36. J. M. Tarascon, E. Wang, F. K. Shokoohi, W. R. McKinnon, and S. Colson, J.

Electrochem. Soc., 138 (1991) 2859.

37. T. Ohzuku, M. Kitagawa, and T. Hirai, J. Electrochem. Soc., 137, (1990) 769-775.

38. Z. X. Shu, I. J. Davidson, R. S. McMillan, and J. J. Murray, J. Power Sources, 68,

(1997) 618.

39. S. Komaba, S. T. Myung, N. Kumagai, T. Kanouchi, K. Oikawa, and T. Kamiyama ,

Solid State Ionics, 152-153 (2002) 311.

40. L. Croguennec, P. Daniard, R. Brec, P. Biensan, and M. Broussely, Solid States

Ionics, 89 (1996) 127.

41. J. Kim, and A. Manthiram, Electrochem. Solid-State Lett., 2 (1999) 55.

42. J. Kim, and A. Manthiram, Nature, 390 (1997) 265.

43. J. M. Kim, and H.-T. Chung, J. Power Sources, 115 (2003) 125.

73

Chapter 3

Novel Synthesis of LiNi0.5Mn0.5O2 by Carbonate Co-precipitation, as an

Alternative Cathode for Li-ion Batteries

3.1 Introduction

Since the commercialization of LiCoO2 by Sony in 1990, many efforts have been

made to find other potential cathode material to replace it because of its high cost and

toxicity. It was reported that LiNi1/3Mn1/3Co1/3O2, which is solid solution between

LiNi0.5Mn0.5O2 and LiCoO2, shows an excellent battery performance, such as a high rate

capability, milder thermal stability and high reversible capacity. This cathode material is

in the process to replace LiCoO2. Furthermore, in order to decrease the price of the

cathode, the cobalt amount should be reduced. Layered LiNi0.5Mn0.5O2 has been

extensively investigated as a promising cathode material for lithium secondary batteries

because of its many advantages, such as high reversible capacity, good cyclic

performance, and mild thermal stability as well as low cost and toxicity as compared to

the commercially used LiCoO2[1,2]. The layered LiNi0.5Mn0.5O2 does not transform to

the spinel structure under electrochemical conditions. The valence of Ni and Mn are 2+

and 4+ respectively [3]. The electrochemical redox process is based on the two electron

reaction of Ni2+/4+. Thus, the electrochemically inactive Mn4+ is believed to support the

host structure during the redox process which leads to structural stability.

Although LiNi0.5Mn0.5O2 is an attractive cathode material, it is very difficult to

prepare cathode material with an excellent electrochemical performance. A hydroxide

74

co-precipitation method has been used as the major preparation technique to obtain a

homogeneous hydroxide solid solution [4-9]. Though the hydroxide co-precipitation is

one of the powerful synthesis methods, the Mn2+ ion in precipitated transition metal

hydroxide is easily oxidized by oxygen in the basic aqueous solution during

precipitation process. For example, Mn(OH)2 is gradually oxidized to MnOOH (Mn3+)

or MnO2 (Mn4+) under the precipitation conditions by oxygen. Such a reaction leads to

the decomposition of hydroxide solid solution and decreases the homogeneity of the

final product. Therefore, the control of Mn valence in precipitate has been found to be a

critical point to obtain pure product. In order to obtain the excellent electrochemical

performance of LiNi0.5Mn0.5O2, preparation of a homogeneous solid solution is the main

key point.

Therefore, we selected the layered LiNi0.5Mn0.5O2 cathode material prepared by an

improved solid solution reaction using the carbonate co-precipitation technique [10]. In

a solid solution, the valence state of Mn in the carbonate form can be stable as 2+. This

implies that the homogeneous solid solution containing Mn could be readily prepared

by the carbonate co-precipitation method. In this paper, we reported the structural and

electrochemical performance of the LiNi0.5Mn0.5O2 cathode material prepared by the

carbonate co-precipitation.

3.2 Experimental 3.2.1 Preparation of layered LiNi0.5Mn0.5O2

We have used NiSO4·6H2O, MnSO4·4-5H2O, and Na2CO3 as the starting material

for preparing the transition metal carbonate powder Ni0.5Mn0.5CO3. We prepared two

aqueous solution, 1 M transition metal sulfate and 1 M sodium carbonate solution. A

bright green Ni0.5Mn0.5CO3 sediment was obtained by feeding these solutions into the

75

hot reaction bath at 60°C under a CO2 atmosphere for more than one week. During the

precipitation, the amount of an equimolar mixture of Ni and Mn solution is controlled

by a perista pump (Atto Corporation AC-2110, Japan). The pH of the solution was

fixed at 7.5 in the hot reaction bath. After the formation of a precipitate, the sediment

Figure 3.1 A schematic flow-chart of experimental procedure for preparing LiNi0.5Mn0.5O2.

was filtered and washed with distilled water and then dried at room temperature for one

day. The co-precipitated Ni0.5Mn0.5CO3 powder was preheated at 500°C for 5 h in air.

The mixed carbonate powder decomposes into metal oxide during pre-heating process.

After pre-heating, we applied an EDTA titration to determine the exact amount of total

transition metal ions present in the pre-heated powder. An excess amount of lithium

hydroxide was added to the powder (Li/Me= 1.1:1) and calcined at various temperatures

800, 900, 950 and 1000oC for 15h-Air

MeSO4 Solution Na2CO3 Solution

MeCO3 Powder

Me3O4

Mixing

Grinding

TestHeating

Pre-heating- 500oC-5h Air

EDTA

Pre-heating- 500oC-5h Air

XRD, SEM, ICP and XPS

1 drop/5min, 1week for precipitation

LiOH ( Li/Me = 1.1:1)

800, 900, 950 and 1000oC for 15h-Air

MeSO4 Solution Na2CO3 Solution

MeCO3 Powder

Me3O4

Mixing

Grinding

TestHeating

Pre-heating- 500oC-5h Air

EDTA

Pre-heating- 500oC-5h Air

XRD, SEM, ICP and XPS

1 drop/5min, 1week for precipitation

LiOH ( Li/Me = 1.1:1)

76

from 800°C-1000°C for 15 h in air. The experimental procedure was schematically

presented in Fig.3.1.

3.2.2 XRD, SEM, SA, XPS and ICP analysis

An X-ray diffraction analysis for the precursor and synthesized LiNi0.5Mn0.5O2

powder was carried out by using a Rigaku Rint 1000 diffractometer with CuKα

radiation in the scanning range of 10 to 80 degree (2θ). The content of metal ions in the

product was determined by an Inductive coupled plasma method (ICP: SPS 7800, Seiko

instruments, Japan). The specific surface area of the synthesized materials were

determined using a Micromeritics Gemini 2375 (USA) by the B.E.T. method. Scanning

electron microscopy (SEM: JSM-5300E, JEOL, Japan) was carried out to observe the

morphology of the synthesized materials. X-ray photoelectron spectra (XPS) of the

transition metal in the synthesized materials were measured using a PHI 5800

(ULVAC-PHI INC, USA) spectrometer with monochromatic Al-Kα radiation. The

distributions of transition metals of the materials were observed using energy dispersive

X-ray analysis (EDX: S-2300, Hitachi, Japan).

3.2.3 Electrochemical performance

The electrochemical characterizations were carried out using the CR-2032 type coin

cell. A cathode was prepared by pressing an active material film, which consists of 20

mg active material and 12 mg conducting binder (Teflonized acetylene black), on the

stainless steel mesh. The cell was composed of the cathode, the lithium foil as an anode

and 1 M LiPF6-ethylene carbonate (EC)/di-methylene carbonate (DMC) (1:2 in

volumes) as an electrolyte. The electrochemical cycling test was carried out in the range

of 2.8-4.5 V and 2.8-4.6 V at 30°C. Cyclic voltammetry studies were conducted using

three electrode cells with 1 M LiPF6-EC/DMC (1:2 in volumes) as the electrode and

77

lithium foils were used as both the reference and counter electrodes. The cyclic

voltammetry experiment was carried out at a scan rate of 0.1mVs-1 in the voltage range

of 2.0-4.5 V vs. Li/Li+ using a potential galvanostat (HUKUTO DENKO: HVS-100,

Japan).

3.3 Result and Discussion 3.3.1 XRD and EDX of carbonate precursor

The structure of the precursor prepared by the carbonate precipitation method was

investigated by X-ray diffraction experiment. Figure 3.2 shows the XRD patterns of the

precursor and standard MnCO3 (JCPDS 07-0268). We could get single phase of XRD

pattern and all the diffraction patterns can be indexed as transition metal carbonates

with a hexagonal structure (space group 167; R3_

c). From the EDX data, a homogeneous

mixture of Ni and Mn carbonate was successfully synthesized by the carbonate

co-precipitation. The mixture of Ni and Mn contents with dense particles were

aggregated each other to form large particle as shown in EDX images.

As mentioned in introduction, a homogeneous mixed state of starting material is

very important to get the LiMn0.5Ni0.5O2 cathode material with excellent battery

performance.

78

10 20 30 40 50 60 70 2θ / deg. CuKα

Inte

nsity

MnCO3 (R-3c)JCPDS 07-0268

Co-precipitated powder

(116

)

(300

)(2

14)

(211

) (12

2)

(018

)(0

24)

(202

)(1

13)

(110

)

(104

)

(012

)

Figure 3.2 X-ray diffraction patterns of co-precipitated precursor and standard MnCO3 pattern as a reference.

Figure 3.3 (a) EDX image of Mn0.5Ni0.5CO3, (b) Mn and (c) Ni.

The Fig. 3.3 shows the energy dispersive X-ray analysis (EDX) of Mn0.5Ni0.5CO3

5µm

(a) (b) (c)

5µm

(a) (b) (c)

79

powder. The EDX data supported that homogenous mixture of Ni and Mn carbonate

source was successfully synthesized by optimized carbonate co-precipitation. The EDX

image shows homogeneous mixture of Ni and Mn with dense round shaped particles

aggregated each other to form large secondary particles.

3.3.2 Effect of calcinations and surface morphology

A SEM observation was carried out to investigate the morphology of the sample

prepared at different calcination temperatures. Fig.3.4 (a), (b), (c), and (d) show the

morphology of the sample prepared at 800, 900, 950 and 1000°C, respectively. The

secondary particle size and shape of cathode materials were almost same. While the

primary particles size become quite different. The sample prepared at 800°C is

composed of nano-sized (100-200 nm) primary particles with higher specific surface

area of 4.6 m2 g-1. The primary particle sizes become larger with increase in calcination

temperatures resulting decrease in specific surface area. By comparing to previous

work in our lab [6], it was confirmed that the dense and round shaped cathode material

were synthesized by optimized carbonate co-precipitation method. The morphological

properties could increase volumetric energy density of lithium ion battery.

3.3.3 Structural and chemical analysis of LiNi0.5Mn0.5O2 compounds

X-ray diffraction analysis for the synthesized cathode material was carried out in

order to investigate the relation between crystallographic properties and the

calcinations temperature. As shown in Fig. 3.5, the entire sample show typical

diffraction pattern of α-NaFeO2 hexagonal structure (space group: 166, R3－

m) and no

remarkable secondary phase was observed. The hexagonal parameters namely a and c

80

of synthesized LiMn0.5Ni0.5O2 at 900°C were calculated from the least square method.

The calculated lattice parameter and c/a were summarized in Table 3.1. The lattice

parameter a and c of the cathode material gradually increased with the increase in

(a) (b)

(c) (d)

4.62m2g -1 1.7 9 m2g -1

1.16 m2g -1 0.63 m2g -15µm 5µm

5µm 5µm

Figure 3.4 SEM image and surface area of synthesized LiMn0.5Ni0.5O2 (a) 800°C, (b) 900°C, (c) 950°C, and (d) 1000°C.

calcinations temperature till 1000°C. The similar phenomena were reported in the

lithium excess layered cathode material [8, 9]. As can be seen in the table, lithium ion

amount decreased with increase in calcinations due to volatization of lithium at high

temperature. Thus, the increase of lattice parameter could be attributed to reduction of

81

Ni3+ (rNi3+ = 0.56 Å) ions to Ni2+ (rNi

2+ = 0.69 Å) to compensate the charge valence

10 20 30 40 50 60 70 80

(a)

2θ / deg. CuKα

(b)

(c)

Inte

nsity

(d) 11311

010

8

10710

5

104

102

00610

1

003

Figure 3.5 XRD patterns of synthesized LiMn0.5Ni0.5O2 (a) 800°C, (b) 900°C, (c) 950°C and (d) 1000°C.

caused by decrease of lithium in the cathode material [8]. The structural integrity of

layered cathode material can be estimated by the c/a ratio and intensity ratio of

I(003)/I(104). The former is related to hexagonality, i.e., the higher c/a value indicates well

ordered hexagonal structure.

82

The latter is closely related to cation mixing in the layered structure, i.e., the higher

intensity ratio indicates lower degree of cation mixing. The calculated intensity ratio

Table 3.1 Chemical analysis, calculated lattice parameter and specific surface area for the synthesized compound

Lattice

parameter

(°C)

Composition

a(Å) c(Å)

c/a

Volume

(Å)3

S.S.A

(m2g-1)

I(003)/I(104)

800 900 950 1000

Li1.07Ni0.506Mn0.494O2

Li1.04Ni0.499Mn0.501O2 Li0.99Ni0.498Mn0.502O2 Li0.97Ni0.507Mn0.493O2

2.8752.8792.8822.884

14.20514.26514.27214.272

4.9414.9534.9514.948

101.68102.43102.72102.98

4.62 2.79 1.16 0.63

1.01 1.26 1.28 1.23

of the cathode materials synthesized at 800, 900, 950 and 1000°C were 1.01, 1.26, 1.28,

and 1.23, respectively. The XRD analysis revealed that the temperature range of 900 to

950 °C is proper to synthesize cathode materials with well ordered hexagonal structure

as well as low cation mixing.

The hexagonal lattice parameters of the LiNi0.5Mn0.5O2 compounds were plotted as

shown in Fig.3.6. Lattice parameters of samples, a and c, were gradually increases with

increase in calcinations temperature. It is well known that higher peak intensity ratio of

I(003)/I(104) indicates low cation mixing and layered hexagonal structure. The highest

I(003)/I(104) values are obtained for 900-950°C. Further increase in calcinations

temperature leads to decrease in intensity ratio. The c/a value also shows similar trend

with the variation of intensity ratio. Therefore, I conclude that the structural integrity

and hexagonal ordering are changed sensitively by the calcinations temperature.

83

800 850 900 950 10002.8742.8762.8782.8802.8822.884

a (A)

Calcination Temperature / oC

14.2014.2214.2414.2614.28

c (A)

4.940

4.944

4.948

4.952

c/a

0.961.041.121.201.28

I(003) / I(004)

Figure 3.6 The variation in structural parameter depends on the calcination temperature.

3.3.4 XPS analysis

XPS studies are very useful techniques for getting information about the oxidation

states of the metal in the material. There has been much debate on the oxidation state of

Mn and Ni in these synthesized materials [11,13]. Hence, we also studied the

oxidation state of Ni and Mn, in the present LiNi0.5Mn0.5O2 powders. An X-ray

photoelectron spectroscopy analysis was performed and the spectra of Ni 2p3/2 and Mn

2p3/2 are given in the Fig.3.7a and 3.5b. The Ni XPS spectrum of all samples revealed

the characteristic peak of ca. 854.9eV and hence could be assigned to Ni2+. The

presence of the satellite peak has also been observed by other research groups [16]. The

84

reason for such satellite peak has been ascribed to the multiple splitting in the energy

level of the nickel oxide [14, 15].

850 855 860

1000oC 900oC 800oC

(a) Ni2p 3/2854.9 eV

In

tens

ity /

a.u.

Binding energy / eV

635 640 645 650

(b) Mn 2p3/2641.8

1000oC 900oC 800oC

Inte

nsity

/ a.

u.

Binding energy / eV

Figure 3.7 XPS of LiNi0.5Mn0.5O2 synthesized materials at 800°C, 900°C, and 1000°C (a) Ni 2p3/2 (b) Mn 2p 3/2.

85

Fig. 3.7 (b) shows the XPS spectra of Mn2p3/2 located at a binding energy of around

641.8 eV. These values coincide well with the binding energy of Ni2+ and Mn4+ as

reported by S. Gopu Kumar et al. [13, 16, 17]. Therefore, the valence state of Ni and

Mn in the compound should be estimated to be 2+ and 4+, respectively.

3.3.5 Electrochemical performances of LiMn0.5Ni0.5O2

Fig. 3.8 show the charge/discharge curves of LiMn0.5Ni0.5O2 cathode material

synthesized at 800-1000°C with a cutoff voltage of 2.8-4.5V with current density of 0.2

mA cm-2. It is notable that the cathode material synthesized at 800°C shows higher

reversible capacity of about 184 mAh g-1 despite of low calcinations temperature.

Moreover, the cathode material synthesized at 900°C exhibited the highest initial

discharge capacity of nearly 200 mAh g-1.The obtained reversible capacity gradually

decreased with increase in calcination temperature. The cathode material synthesized at

950 and 1000°C delivered 192 and 171 mAh g-1 of discharge capacities, respectively.

The highest reversible capacity for the cathodes synthesized at 900°C can be attributed

to its high structural integrity as confirmed by XRD analysis. The lowest reversible

capacity of the cathode material synthesized at 1000°C can be attribute to deficiency of

lithium ion and larger particle sizes.

86

0 50 100 150 200 250

3

4

800oC

Capacity / mAh g-1

3

4

900oC

Cel

l vol

tage

/ V

3

4

950oC

3

4

1000oC

Figure 3.8 Charge discharge curves of LiMn0.5Ni0.5O2 cathode materials at the voltage range of 2.8-4.5V at 30°C.

Figs. 3.8 (a) and (b) shows the cyclic performance of LiNi0.5Mn0.5O2 in the high

voltage range of 4.5V-2.8V and 4.6V-2.8V at 30°C at the current density of 0.2mA

cm-2. Fig 6(a) charged at 4.5V-2.8V shows stable cycleability as well as an excellent

87

cyclic performance. The cathode material synthesized at 900°C gives initial discharge

capacities of 200 mAh g-1 whereas the cathode materials synthesized at 800°C, 950°C,

and 1000°C give initial discharge capacities about 184 mAh g-1, 192 mAh g-1, and 171

mAh g-1 in the first cycle respectively. The retained discharge capacity of the cathode

material synthesized from 800°C - 1000°C were 98.2%, 99.6%, 99.5%, and 98 %

after30 cycles.

The cathode material synthesized from 800-1000°C when charged at 4.6-2.8V gave

an initial discharge capacity of 208 mAh g-1, 206 mAh g-1, 208 mAh g-1, and 199 mAh

g-1 in the first cycle and still delivered 199 mAh g-1, 205 mAh g-1, 199 mAh g-1, and

182 mAh g-1 after 30 cycles at 30°C. The retained discharge capacities of the cathode

material from 800°C-1000°C were 95.6%, 99.5%, 95.7%, and 92.4% respectively after

30 cycles. On increasing the calcinations temperature, the obtained discharge capacity

gradually decreases except for the sample synthesized at 900°C.

I thought the reason for the decrease in initial discharge capacity loss was due to

the decrease in the lithium content with an increase in the calcination temperature. The

sample prepared at 900°C shows an excellent cycleability as well as high structural

stability, when charged at 4.5V-2.8V and 4.6V-2.8V at 30°C. It is obvious that the

excess lithium on LiNi0.5Mn0.5O2 give rise to a high structural stability resulting in

high capacity, good cycleability, reduced resistance and faster Li+ diffusion based on

the high quality cathode material prepared by the carbonate co-precipitation.

88

0 5 10 15 20 25 30100

150

200

250

(b) 2.8-4.6V

950oC 900oC 1000oC 800oCD

.Cap

acity

/ m

Ah

g-1

Cycle number

Figure 3.9 Effect of cycle performance on the synthesized temperature of LiNi0.5Mn0.5O2 cathode material in the voltage range of (a) 2.8-4.5 V and (b) 2.8-4.6 V.

Cyclic voltammetry is a useful technique to evaluate the cyclic performance of the

LiNi0.5Mn0.5O2 electrode, the cyclic voltammogram of the cells were recorded with

lithium metal as reference electrode, in the voltage range of 2.8-4.5V at the scan rate

0 5 10 15 20 25 300

50

100

150

200

250

(a) (2.8-4.5V)

800oC 900oC 950oC 1000oC

Cap

acity

/ m

Ah

g-1

Cycle number

89

of 0.1 mV s-1. Fig. 3.10 shows a typical cyclic voltammogram of 1-5 cycles for the

LiNi0.5Mn0.5O2 prepared at 800, 900, and 950°C, respectively. As we observed no

peak around 3 V it indicates that no manganese is present in the Mn3+ state in our

synthesized materials [18]. However, the absence of such redox peaks in the 3 V

region of LiNi0.5Mn0.5O2 cathode indicates that Mn ions are not in the 3+ oxidation

states; rather it may presumably be in the electrochemically inactive form of 4+

oxidation state [20]. The layered LiNi0.5Mn0.5O2 samples prepared at 800, 900, and

950°C shows major oxidation and reduction peaks exists at 3.90 V and at 3.72 V may

be ascribed to the Ni2+↔ Ni4+ redox reaction. This observation is similar to the

presence of Ni2+/4+ couple observed in the xLi(Ni0.5Mn0.5)O2.(1-x)Li2TiO3 compound

reported by Kim et al.[21] and LiNi1/3Co1/3Mn1/3O2 system by Shaju et al [3]. These

peaks are also a sign of a hexagonal phase and show perfect reversibility. The

potential difference between the anodic and cathodic peaks (∆V) for the samples

prepared at 800, 900, and 950°C are determined to be 0.20 V. For the 3rd and 5th

cycles, ∆V value is smaller than the previously reported value even at faster scan rate

of 0.1mV s-1 [16, 19]. We find that the major hexagonal peak is centered at the lower

voltage of 3.90 V as compared to 4.03 V in the case of Kang et al [13] but is similar in

position to Lu et al. [11]. Our CV results are in close agreements to that reported in

literature [11, 13]. These figures show a good reversibility.

90

2.0 2.5 3.0 3.5 4.0 4.5-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

0.4

0.5

(a)

3.70V

3.90V800oC

1st

3rd

5th

Cur

rent

/ m

A

Cell voltage / V

2.0 2.5 3.0 3.5 4.0 4.5-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

0.4

0.5

(b)

3.70V

3.90V900oC

1st

2nd

3rd

Cur

rent

/ m

A

Cell voltage / V

91

2.0 2.5 3.0 3.5 4.0 4.5

-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

0.4

0.5

(c) 950oC 1st 3rd 5th

Cur

rent

/ m

A

Cell voltage / V

Figure 3.10 Cyclic voltammograms of LiNi0.5Mn0.5O2 material synthesized at (a) 800°C (b) 900°C (c) 950°C.

Fig.3.11 shows rate capability test for the LiNi0.5Mn0.5O2 sample synthesized at

900°C. The cell was charged to 4.5V and discharged to 2.8V. It is clear that the

discharge capacities decrease with the increase in the rate capability value. The

discharge capacity for the 900°C sample was reduced from 195 mAh g-1 to 167 mAh

g-1 when the discharge current density increased from 0.1 C, 0.2 C, 0.5 C, 1 C, and 2 C,

where the C-rate was determined by using 200 mAh g-1 as a theoretical capacity

indicating that only 85% of the initial discharge capacity was retained. The rate

performance of the 900°C sample is excellent. Specially, the discharge capacity of

92

Figure 3.11 Rate capability for the layered LiNi0.5Mn0.5O2 cathode material of 900°C in the voltage range of 2.8 V-4.5 V for the (a) 0.1C, and (e) 2C.

1000°C sample more rapidly decreased than those of the other samples. I believe that

the reason for the low rate capability of the sample prepared at high temperature was

due to the growth of the primary particle size as well as a smaller specific surface area.

3.3.6 Ex-situ XRD study during charge and after discharge

Usually, the intercalation/de-intercalation of lithium ions into/from the host structure

leads to a change in the lattice volume, and sometime a structural transformation to

other phases. To determine the phase transformation of the layered LiMn0.5Ni0.5O2

during the charge and discharge, we measured the ex-situ XRD of LiMn0.5Ni0.5O2 of

different synthesized sample during charging to 4.5V and discharge after one cycle as

0.0 0.5 1.0 1.5 2.0100

150

200

250

900oC

Cap

acity

/ m

Ah

g-1

C-rate

93

10 20 30 40 50 60 70 80

800oCAfter discharge

2θ / deg. Cukα

800oC

Charged to 4.5V

900oCAfter discharge

900oCCharged to 4.5V

Inte

nsity

After discharge

Si* 1000oC

1000oC

(113

)(1

10)

(108

)

(107

)

(105

)(104

)

(102

)(006

)(1

01)

(003

) Chareged to 4.5V

Figure 3.12 Ex-situ XRD patterns of the layered LiNi0.5Mn0.5O2 cathode materials.

given in Fig. 3.12. The graphite, Silicon, was used as standard in this experiment.

There is no change, but peaks are shifting in terms of the XRD patterns, suggesting no

phase transformation from the hexagonal to monoclinic structure. When the electrode

are charged to 4.5 V and then discharged to 2.8 V, the layered structure of the different

synthesized LiMn0.5Ni0.5O2 remain same and the lattice parameter values of a and c

change as shown in Table 3.2. The volume of lattice expansion at 800, 900, and

94

1000°C was nearly 2.71%, 1.85%, and 1.65% which is very small. This result suggests

Table 3.2 Lattice parameter values of LiMn0.5Ni0.5O2 cathode materials after charged to 4.5 V and discharge at 2.8V after one cycle

that the LiMn0.5Ni0.5O2 cathode materials have an excellent structural stability during

cycling in this voltage range. With an increase in the calcination temperature, the

expanded percentage decreases. Therefore, the cycleability of the layered

LiMn0.5Ni0.5O2 is excellent in the voltage range of 4.5 V-2.8 V.

(°C) Condition a(Å) c(Å) (Å3)

800°C Charged at 4.5V 2.854 14.373 101.43 800°C After discharge at 2.8V 2.889 14.401 104.14 900°C Charged at 4.5V 2.847 14.466 101.61 900°C After discharge at 2.8V 2.880 14.405 103.48 1000°C Charged at 4.5V 2.863 14.428 102.47 1000°C After discharge at 2.8V 2.889 14.400 104.13

95

3.4 Conclusion

We have successfully synthesized round shaped Mn0.5Ni0.5CO3 precursor for the

LiMn0.5Ni0.5O2 cathode material by optimized carbonate co-precipitation method. An

EDX observation revealed that the Mn and Ni were homogeneously dispersed in the

precursor. By using the precursor, we synthesized the well ordered layered

LiMn0.5Ni0.5O2 cathode material even at lower temperature of 800°C. From the

structural analysis by XRD, we found that the temperature range of 900-950°C is

proper to synthesize well ordered the layered cathode as well as lower cation mixing.

The highest reversible capacity of 200 mAh g-1 at 4.5V and 206 mAh g-1 at 4.6V was

obtained from the cathode synthesized at 900°C. The cycle performance of

LiMn0.5Ni0.5O2 cathode material is excellent at high voltage condition. The layered

LiMn0.5Ni0.5O2 cathode materials suppress the decomposition of electrolyte at high

voltage. The discharge capacity of layered LiMn0.5Ni0.5O2 cathode materials is 10%

higher than that of other reported synthesis method and its rate capacity performance is

excellent due to the dense round shaped particle prepared by carbonate co-precipitation

method. Therefore it could be conclude that the improved carbonate co-precipitation is

an outstanding method to synthesize the cathode material with excellent battery

performance, and the 900°C is a proper temperature to synthesize the cathode material.

96

References

1. T. Ohzuku, and Y. Makimura, Chem. Lett., 2001 744.

2. Y. Makimura, and T. Ohzuku, J. Power Sources, 119 (2003) 156.

3. K. M. Shaju, G. V. Subba Rao, and B. V. R. Chowdari, Electrochim. Acta, 48

(2002) 145.

4. T. Ohzuku, and Y. Makimura, Chem. Lett., 2001 642.

5. N. Yabuuchi, and T. Ohzuku, J. Power Sources, 119-121 (2003) 171.

6. I. Belharouak, Y.-K. Sun, J. Liu, and K. Amine, J. Power Sources, 123 (2003)

247.

7. Z. Lu, D. D. MacNeil, and J.R. Dahn, Electrochem. Solid-State Lett., 4 (2001)

A200.

8. D. D. MacNeil, Z. Lu, and J.R. Dahn, J. Electrochem. Soc., 149 (2002) A1332.

9. M.-H. Lee, Y.-J. Kang, S.T. Myung, and Y.-K. Sun, Electrochim. Acta, 48 (2003)

1505.

10. S. M. Park, T. H. Cho, H. Noguchi and M. Yoshio, Chem. Lett., 33 (2004) 748.

11. Z. Lu, L. Y. Beaulieu, R. A. Donaberger, C. L. Thomas and J. R. Dahn, J.

Electrochem. Soc., 149 (2002) A778.

12. W. S. Yoon, Y. Paik, X.-Q. Yang, M. Balasubramanian, J. McBreen, C. P. Grey,

J. Electrochem. Solid-State Lett., 5 (2002) A263.

13. S. H. Kang, J. Kim, M. E. Stoll, D. Abraham, Y. K. Sun, and K. Amine, J. Power

Sources, 112 (2002) 41.

14. A. F. Carley, S. D. Jackson, J. N. Oshea, and M. W. Roberts, Surf. Sci., 440

(1999) L868.

97

15. K. Amine, H. Tukamoto, H. Yasuda, and Y. Fujita, J. Electrochem. Soc., 143

(1996) 1607.

16. S. Gopu Kumar, K. Y. Chung, and K. B. Kim, Electrochim. Acta., 49 (2004) 803.

17. G. E. Muilenburg, C.D. Wager, W. M. O. Riggs, L. E. Davis, and J. F. Moulder

(Eds.) Handbook of X-ray Photoelectron Spectroscopy, Perking-Elmer Corp.,

New York, (1979)

18. J. M. Paulsen, C. L. Thomas, and J. R. Dahn, J. Electrochem. Solid-State Lett.,

147 (2000) 861.

19. J.-M. Kim, and H.-T. Chung, Electrochim. Acta, 49 (2004) 937.

20. S. H. Kang, and K. Amine, J. Power Sources, 119-121 (2003) 150.

21. J. S. Kim, C. S. Johnson, and M. M. Thackeray, Electrochem. Commun., 4

(2002) 205.

98

Chapter 4

Development of Cathode Materials for Advanced

Megalo-Capacitance Capacitor

4.1 Introduction

Recently, supercapacitors have drawn extensive attention because of their potential

applications as electric storage devices in many fields. In the conventional electric

double-layer capacitors (EDLCs), the storage of charge on electrodes is through the

non-faradaic adsorption at the interfaces between the electrodes and electrolyte.

Generally, both the positive and negative electrodes in EDLCs are identical activated

carbon (AC) electrodes. This type of symmetrical capacitors inevitably restricts their

functional ranges such as specific capacitance, working voltage, energy density, etc. [1,

2], although they possess the merits of long cycle life and high power density. Moreover,

during the course of industrial production of activated carbon, the precursor of activated

carbon is mixed with 2-5 times with KOH and heated at a certain high temperature

under 1000 °C in the inert atmosphere. But alkaline activated carbon is several times

costly than the other activation. Alkaline activated carbon is used for memory back up

in capacitor. Most commonly used activated carbon is steam activated carbon, which is

synthesized under water vapor steam. Steam activated carbon is commonly used for all

type of EDLC.

99

Usually graphite is not a capacitor material because of its small ability of adsorption.

But, I have found special cathodic graphite can adsorb anion in non-aqueous solution if

activated carbon (AC) is used as a counter electrode. In AC/AC capacitor, one electrode

has enough ability for adsorption if counter electrode has an ability for adsorb a counter

ion. So, I have tried to find such special graphite and then I have found that such a

special carbonous materials are KS-series graphite. I have proposed new idea to replace

one of its activated carbon by graphite carbon as a cathode for capacitor.

I have named as Megalo-capacitance capacitor using graphitic carbons as the

positive electrode. Organic electrolyte was applied to ensure the high working voltage

up to 3.5V. It shows that graphite can provide satisfactory performance as positive

electrode for electrochemical capacitors.

The typical charge-discharge curves of AC/AC and graphite/AC capacitor are shown

in Fig.4.1. I can see clearly from Fig.4.1 with the increase in voltage, both AC/AC and

graphite/AC capacitors show the increase in capacity. AC/AC (RP-20) capacitor has

capacities of 33, 42, and 58 mAh g(+)-1, ( g(+) means cathode weight) at charging

voltage 2.5, 3.0, and 3.5V, respectively which means that the energy density increase

with increase in charging voltage. In the case of AC/AC capacitor, charge-discharge

curve shows an almost straight line. The charging voltage of AC/AC capacitor is limited

around 2.5V, because of electrolyte decomposition. But in the case of graphite/AC

capacitor, while during charging the potential jump suddenly up to 1V and move slowly

with time. During charge, anions are adsorbed on the surface of graphite electrode

because a cation is adsorbed to the anode activated carbon. With the raise in voltage up

to 3.5V, graphite/AC capacitor shows straight line as capacitor

100

Figure 4.1 Voltage dependence on the charge-discharge curves AC (RP-20)/AC(RP-20)

and graphite/AC(RP-20) capacitor weight ratio 1:1.

curves at voltage 3.5-2.0V with capacity of 40 mAh g-1.The electrolyte decomposition

of graphite/AC capacitor is low because the surface area of KS-44 is small. During

study to increase the energy density of EDLC, I have found that special graphite can

0 20 400

1

2

3

4

KS-44/AC-1 1:1

3.5V

0 20 40 60

0

1

2

3

4

3.5V

AC-1/AC-1 1:1

0 20 400

1

2

3

3.0V

Vol

tage

/ V

0 20 40 600

1

2

3

3.0V

2.5V

0 20 400

1

2

2.5V

Capacity / m Ah g-1 (cathode w eight)

0 20 40 600

1

2

101

adsorb anion if anode (AC) adsorbed a cations and be able to apply higher charging

voltage. One of the special graphite is KS-series graphite from Timcal Co. Ltd.,

Switzerland.

The capacitance (CF) in Farad of EDLC was given by formula as follows:

CF = A × T / V' (4.1.1)

Where A is unit of current in ampere, T is time in second, and V' is voltage.

V' is commonly used as:

V ' = Vmax/2

Vmax = applied maximum voltage

On the contrary energy density is defined as:

Ed =Ah × Vav. (4.1.2)

Where Vav. is average discharge voltage

Then the relation of CF and Ed is given by

Ed = CF/3600 × Vmax /2 × Vav. (4.1.3)

In this study, we investigated the electrochemical behavior of graphite/AC

capacitors by the means of galvanostatic charge–discharge and in situ XRD

measurements, etc. Some factors influence the performance of the capacitors will be

addressed.

4.2 Experimental

Artificial graphite KS6, KS-44, SFG-44 and SFG-6 (Timcal) were selected as the

positive electrode material, whereas activated carbon PW15M13130 formed by steam

activation (Kureha), Langmuir specific surface area of 1050 m2 g−1, activated by steam

102

and RP-20 formed by steam activation (from Kuraray), Langmuir specific surface area

of 1300 m2 g-1 activated by steam were applied as the negative electrode material in this

study. The electrodes material consist of over 10mg were mixed with conductive binder

TAB (teflonized acetylene black) consist of 3 mg and pressed on a stainless steel mesh

as a current collector. The Graphite/AC weight ratio is same and the electrolyte used in

this study were 1.5M TEMAPF6 or 1.5M TEMABF4 dissolved in PC solvent

(Tomiyama) (TEMA stands for triethylmethyl ammonium ion). Coin cells were

fabricated for galvanostatic (constant current) charge–discharge tests. In a coin cell,

glass fiber filters as a separator soaked with the electrolyte were sandwiched by the

positive and negative electrodes (geometric area, 2 cm2). Coin cell assembling was

performed in a glove box filled with Ar atmosphere (dew point, lower than −80 °C).

Charge–discharge tests were carried out to test the electrochemical behavior of the

asymmetric capacitors using the instrument Nagano (BTS-2004W, Japan). Generally,

the constant current density was controlled at 0.5 mA cm−2. The adsorption isotherm

and pore volume of graphite is measured by using the instrument BELSORP 18 PLUS

SP (BEL. INC., Japan).The voltage range was from 0 to 3.5 V for graphite/AC

capacitors. To trace the positive and negative electrode potentials independently,

four-electrode cells were also constructed as shown in Fig.4.2. Graphite positive

electrode, AC negative electrode and the reference electrode (Li metal chip), were

dipped into the electrolyte contained in a glass beaker. During the course of

galvanostatic charge–discharge between the positive and negative electrodes, the

potentials of either positive or negative electrodes with respect to the reference electrode

were measured by an auxiliary potentiometer. Generally, the electrode potentials in the

lithium electrolyte were reported against Li metal. The potential of AC electrode vs.

103

Li/Li+ is assumed to be 3 V. The total voltage of the capacitor actually equals to the

difference between the potentials of positive and negative electrodes.

Figure 4.2 Schematic diagram of four electrode system.

The SEM image of AC is shown in Fig. 4.3. Macropores can be clearly found on the

surface of each AC particle. Therefore, in the conventional EDLC (AC/AC capacitor),

the restriction of energy density might be attributed mainly to the positive electrode side.

Not only the potential, but also the capacitance at the interface between the positive

electrode and the electrolyte is heavily shackled.

W.E. R.E.

Rubber cork

Glass

R.E. C.E.

Electrolyte

W.E. R.E.

Rubber cork

Glass

R.E. C.E.

Electrolyte

104

Figure 4.3 SEM image of activated carbon electrode for negative material.

Powder X-ray diffraction (XRD, Rint-1100, Rigaku, Japan) measurement using Cu

Kα radiation was employed for in situ XRD study of graphite positive electrodes. The in

situ XRD cells used in this study is diagrammed in Fig. 4.4. It consists of two electrodes

separated by glass fiber filters soaked with electrolytes. Thin Al film was applied as the

current collectors for graphite positive electrode as well as X-ray window. The slurry

containing KS6, acetylene black powders together with PVDF/NMP solution was

coated onto an Al foil with a doctor blade to prepare the positive graphite electrode for

in situ XRD measurement.

105

Figure 4.4 Schematic view of in situ XRD cell

4.3 Result and Discussion

4.3.1 XRD and SEM profiles of graphitic materials

The Fig 4.5 is the XRD patterns of several natural graphite sample like Chinese

natural graphite and artificial graphite MAGD-20, MCF, MCMB6-28 and KS-6.

Common graphite shows the (002) peak is located at 26.50-25.88°. From this Fig 4.5,

we can see clearly that Chinese natural graphite flake has high peak intensity value

where as MAGD-20(Hitachi Co. Ltd., Japan), MCF (Petoca Co. Ltd., Japan) and

MCMB6-28 (Osaka Gas Co. Ltd., Japan) has medium peak intensity values. Most of the

Japanese companies were using MAGD graphite in a large scale for the production of

Li-ion battery. But the peak height of (002) plane of other graphite is very low

106

compared to Chinese natural graphite flake because crystal structure of KS-6 is low as

compared to other graphite.

Figure 4.5 X-ray diffraction patterns of Chinese natural graphite flake, MAGD-20, MCF, MCMB 6-28 and KS-6 graphitic carbon.

Fig.4.6 shows the XRD patterns of several sample named KS-44, SFG-44, and

SFG-6 from (Timcal Co. Ltd., Switzerland). This figure shows the XRD patterns of

artificial graphite like KS-44, SFG-44, and SFG-6. The peak height of (002) plane of

KS-44 is very small compared to NG flake and bit larger than other graphite. SFG-44

and SFG-6 graphite shows similar peak height intensity. Graphite sample with low

intensity and low surface area was chosen as cathode electrode for capacitor.

0

50000

100000

150000

200000

250000

300000

350000

27262726272627262726

KS-6

MCMB(6-28)

MCF

MAGD-20

Natural Graphite

Inte

nsity

(a.u

.)

2θ / deg (Cukα)

107

Figure 4.6 X-ray diffraction patterns of KS-6, KS-44, SFG-44 and SFG-6 graphitic carbon.

Fig.4.7 shows the SEM image of different graphite samples. KS-44 shows needle

and pellet like structure with surface area of about 6.8 m2 g-1 where as SFG-44 shows

plate like structure with specific surface area of 3.6 m2 g-1 and SFG-6 sample shows

corn flake like structure with specific surface area of 12 m2 g-1. KS-44 shows crystal

structure with more than 6µm in diameter where as SFG-44 and SFG-6 has less crystal

structure with less than 5µm in diameter

0

10000

20000

30000

40000

50000

60000

70000

2726272627262726

KS-6

SFG-6SFG-44

KS-44

Inte

nsity

(a.u

.)

2θ / deg (Cukα)

108

Figure 4.7 SEM images of (a) KS-44, (b) SFG-44 and (c) SFG-6

(a)

(b)

(c)

6.8 m2g-1

3.6 m2g-1

12.2 m2g-1

109

4.3.2 Adsorption/disorption isotherm study of anode and cathode materials.

The relation between the amount of substance adsorbed by an adsorbent and the

equilibrium pressure or concentration at constant temperature is called an adsorption

isotherm. Six general types of isotherm have been observed in the adsorption of gases

on solid [3]. There are shown in the Fig.4.8. In case of chemisorption only type (I) are

encountered, while in physical adsorption all five types occur.

Type (I):- This types of adsorption isotherm curve is obtained in such cases where

mono molecular layer is formed on the surface of the adsorbent. It means that there no

change in the value with the increase in pressure onwards. This type of isotherm is

Langmuir adsorption isotherm.

Type (II):- This type of isotherm has a transition point which represents the pressure

at which the formation of monomolecular layer is complete and that of the

multi-molecular layer being started.

Type (III):- In this type of isotherm there is no transition point. In this type, the

multi-molecular layer formation start even before the formation of monomolecular layer

is complete.

Type (IV):- In this case there is a tendency for saturation state to be reached in the

multi-molecular layer region as well.

Type (V):- This isotherm indicates multi-molecular layer formation in the begaining.

At high pressure, there is a tendency for a to remain constant. It means that the

saturation state has been reached.

Type (VI):- This type of isotherms is of surfaces with an extremely homogeneous

structure like pyrolytic graphite using, for example, argon and methane as adsorbates

but not nitrogen.

110

Figure 4.8 The IUPAC classification of adsorption isotherm shapes.

The porosities of concern to adsorption process are the micro and mesopore.

Micropore have entrance dimension of w < 2 nm, where ‘w’ is the pore size. Mesopore

have entrance dimension between 2nm < w < 50 nm where as macropores have entrance

dimension of w > 50 nm which was defined by IUPAC. The Va is amount adsorbed in

volume STP (cc g-1), where P is equilibrium partial pressure, Po is saturation vapor

pressure .The nitrogen adsorption isotherm of the graphitic carbons KS-44 prepared in

this study is primarily of type (III), which showed that multilayer adsorption starts

before monolayer adsorption was finished. The adsorbed volume is very small as shown

(P/Po)

Va/

cc g

-1

111

in Fig. 4.9, which shows small surface area. The amount of micropores on the KS-44

graphite was very less shown by less amount of nitrogen gas adsorbed at lower partial

pressure as shown in Fig 4.9. The small hysteresis loop of the adsorption/desorption

isotherm is cylindrical type in pore structure. The increase in amount of nitrogen gas

adsorbed at higher partial pressure indicates the presence of macropores. Inorganic solid

or adsorption to pleateau plane indicates BET theory can be applicable.

Figure 4.9 Adsorption/disorption isotherm curves of KS-44 graphitic carbons.

The nitrogen adsorption isotherm of the Chinese natural graphite flake prepared in

this study is primarily of type (III). Va is very small which is due to low surface area.

The amount of micropores on the natural graphite flake was very less shown by less

amount of nitrogen adsorbed at low partial pressure as shown in Fig.4.10. The increase

0.0 0.2 0.4 0.6 0.8 1.00

20

40 Desorption

Adsorption

KS-44

Va/c

c g-1

P/Po

112

in amount of nitrogen gas adsorbed at higher partial pressure indicates the presence of

macropores too.

Figure 4.10 Adsorption/desoprtion isotherm curve of natural graphite flake.

The nitrogen adsorption isotherm of the activated carbon named PW15M in this

study was type (I) as shown in Fig. 4.11. In the case of chemisorptions only isotherm of

type (I) are encountered, while in physical adsorption all five types occur.

This type of adsorption isotherm curve is obtained in such case where

mono-molecular layer is formed on the surface of adsorbent. This curve shows that a

saturation state is reached. It means that there is no change in the value with the increase

in partial pressure onward. The adsorption/desorption isotherm of PW15M shows only

one curve but initial raise with partial pressure means the adsorption to micropore. The

pore radius of PW15M contains micropore size around 0.7 nm. The value of Va is very

0.0 0.5 1.0

0

5

10

15Chinese NG Flake

Desorption

Adsorption

Va/c

c g-1

P/Po

113

large which is due to high surface area. If I apply Langmuir adsorption in the case of

PW15M then its surface area is 1635 m2 g-1 but from BET adsorption, I have found the

surface area of PW15M is around 1200 m2 g-1.

Figure4.11 Typical adsorption/desorption isotherm curve of activated carbon

(PW15M).

The pore volume of KS-44 and Chinese natural graphite was measured by using

instrument BELSORP 18 PLUS SP (BEL, INC., Japan) as shown in Fig.4.12. The

adsorption/desorption isotherm indicates the presence of small amount of micropore

phase in KS-44. The pore radius of KS-44 was around 1.6 nm observed by

adsorption/desorption isotherm. The existence of micropore in natural graphite is not

clear.

P/Po

Va/c

c g-1

P/Po

Va/c

c g-1

114

Figure 4.12 Variation of pore volume with pore diameter of the graphitic carbon KS-44 and natural graphite flake.

The variation of pore volume with pore radius of activated carbon (PW15M) is

shown in Fig.4.13. The activated carbon (PW15M) shows micropore of 8 nm much

broader distribution of pores extending. The pore radius of activated carbon (PW15M)

is 7-8 nm observed by adsorption/desorption isotherm.

0 2 4 6 8 10

0.0

0.5

1.0

1.5

2.0NG flake

rp /nm

-2

0

2

4

6

8

rp peak : 1.6 nm

KS-44

dV /

dr (c

c g-1

)

115

Figure 4.13 Variation of pore volume with pore diameter of the activated carbon

(PW15M)

dV/d

r(c

c g-

1 )

r/nm

dV/d

r(c

c g-

1 )

r/nm

116

4.3.3 Novel capacitor using graphitic carbon

I have developed a novel electric double capacitor, using graphitic carbons as the

positive electrode, and named as Megalo-capacitance capacitor. Here, I have introduced

some KS-series graphitic carbon as cathode materials instead of activated carbon and

anode materials were activated carbon electrode which serves as negative electrode.

Organic electrolyte was applied to ensure the high working voltage up to 3.5V. It shows

that graphite can provide satisfactory performance as positive electrode for

electrochemical capacitors.

The charge discharge curves AC/AC and graphite/AC capacitor were shown in Fig

4.14(a) to compare the energy density. The charge-discharge curve of AC/AC capacitor

shows almost straight line, whereas charge-discharge curve of graphite/AC capacitor

shows straight line at 2.0-3.5V (discharge line). So the energy density of AC/AC

capacitor is 25 Wh/kg (cathode weight).

(a)(a)

20 Ah/kg × 1.25 V = 25 Wh/kg

117

Figure 4.14 Charge-discharge curves of (a) PW15M/PW15M and (b) graphite/AC capacitor (weight ratio 1:1).

On the other hand, the energy density of graphite/AC capacitor is 100 Wh/kg

(cathode weight) at high working voltage of 3.5V. Comparing these figures I can say

that Graphite/AC capacitor can give at least 4 times high capacity than conventional

EDLC. Advanced megalo-capacitance capacitor shows satisfactory performance in

terms of high capacity and voltage too.

4.3.4 Charge-discharge curves and its cyclic performance

Fig. 4.15 shows the charge-discharge curves of different graphite/AC electrode.

Here we select activated carbon (RP-20) from Kuraray which shows satisfactory

performance in term of high capacity, high energy density. All the graphite shows same

charge-discharge curves nearly around 42-45 mAh/g capacity. KS-44 shows highest

0 10 20 30 40 500

1

2

3

4

(b)

Vol

tage

/ V

Capacity / mAh g-1(cathode weight)

40 Ah/kg × 2.5V = 100 Wh/kg

118

capacity of about 45mAh/g.

Figure 4.15 Charge-discharge curves of several graphites using AC (RP-20) weight ratio 1:1 at the voltage range of 3.5 V- 0V.

Fig.4.16 shows the charge-discharge curves of different graphite sample with

activated carbon named PW15M from Kureha. Using PW15M activated carbon it

shows lower capacity than the (RP-20) activated carbon. All the graphite/PW15M

capacitor like KS-44, SFG-44 and SFG-6 shows initial capacity of about 32-34 mAh g-1.

0 20 4001234

KS-6/RP-20 1:1

Capacity / mAh g-1(cathode weight)

01234

SFG-6/RP-20 1:1

Vol

tage

/ V

01234

SFG-44/RP-20 1:1

01234

KS-44/RP-20 1:1

119

Figure 4.16 Charge-discharge curves of different graphite using AC (PW15M) at the

voltage range of 3.5V- 0V.

The Fig. 4.17(a) shows the cyclic performance of different graphite samples with

RP-20 anode. KS-44/RP-20 show initial discharge capacity of 45mAh/g at the first

cycle and still maintain capacity retention rate around 86.4%. Where as SFG-44 and

SFG-6 shows the discharge capacity of about 44mAh/g and 43 mAh g-1 at the first cycle

and still retain the 82.6% and 84.4% of capacity after 100 cycles. Comparing from

cycleability, KS-44/RP-20 is best but cycle life may not be enough as a capacitor.

On the other hand in Fig.4.17 (b), several graphite samples with PW15M anode

0 20 40

0

1

2

3

4

SFG-6/PW15M 1:1

Capacity / mAh g-1(cathode weight)

0

1

2

3

4

SFG-44/PW15M 1:1

Vol

tage

/ V

0

1

2

3

4

KS-44/PW15M 1:1

120

electrode give low capacity about 28-32 mAh g-1and cycle retention rate was higher

than graphite sample with activated carbon RP-20. It would be suitable to use PW15M

as an anode material, even though it shows lower capacity than RP-20.

Figure 4.17 (a) Cycle performance of several graphite samples with activated carbon

(RP-20) weight ratio 1:1.

0 20 40 60 80 1000

20

40

60

KS-6/RP-20

Cycle number

0

20

40

60

SFG-6/RP-20

0

20

40

60

SFG-44/RP-20

Cap

aict

y / m

Ah

g-1

0

20

40

60

KS-44/RP-20

121

Figure 4.17 (b) Cycle performances of several graphitic samples with activated carbon

(PW15M) weight ratio 1:1.

0 20 40 60 80 1000

20

40

60

SFG-6/PW15M

Cycle number

0

20

40

60

SFG-44/PW15M

Cap

acity

/ m

Ah g

-1 0

20

40

60

KS-44/PW15M

122

4.3.5 Potential profile during charge/discharge of AC/AC and graphite/AC capacitor

Fig.4.18 shows the potential profiles of a conventional EDLC (AC/AC) during

galvanostatic charge–discharge cycles (the positive and negative electrodes have the

Figure 4.18 Potential profiles of a conventional EDLC (AC/AC) during galvanostatic charge–discharge cycles (the positive and negative electrodes have the same weights of active material (9.3 mg), electrolyte: 1.5 M TEMAPF6-PC).

same weights of active material). The potential profiles of both the positive and negative

AC electrodes are almost linear as well as the cell voltage curve. At the point of charge

cutoff in each cycle (capacitor voltage 2.7 V), the potential of the positive electrode can

reach as high as about 4 V vs. Li/Li+ (ceiling potential), while that of the negative

electrode drops down to 1.3 V vs. Li/Li+ (bottom potential). On the other hand, when

123

the capacitor voltage becomes 0 V, both the positive and negative electrodes have the

same potential of about 2.7 V vs. Li/Li+. Therefore, the interface between positive

electrode and electrolyte is burdened with about 1.3 V (4 −2.7 V) from the total working

voltage of 2.7 V. On the other hand, that corresponding to the negative side shares 1.4 V

(2.7 − 1.3V) from the total working voltage of 2.7 V. So the faradaic solvent electrolysis

at the positive electrode may be very mild because PC solvent decomposition starts at

about 5.4 V vs. Li/Li+ [4].

I considered the disadvantages of the AC positive electrode as mentioned above and

attempted to use the graphite positive electrode instead of AC. Fig. 4.19 shows the

potential profiles of the graphite/AC capacitor during galvanostatic charge–discharge

processes (the positive and negative electrodes have the same weights of active

material). The potential profile of graphite positive electrode in the graphite/AC

capacitor is quite different from that of the AC positive electrode in conventional EDLC.

In the initial charge process of the graphite/AC capacitor, the potential of graphite

positive electrode jumps up suddenly to about 4.4 V vs. Li/Li+ at first, and then rises up

slowly until about 4.7 V vs. Li/Li+. On the other hand, the potential of AC negative

electrode drops down almost linearly in the charge process of capacitor, it can reach the

lowest potential of about 1.7-4.8 V vs. Li/Li+. Accordingly, the total voltage curves of

the graphite/AC are straight lines, roughly comprise two linear portions with different

slopes.

In the graphite/AC capacitor, the graphite positive electrode works at rather high

potential ranges, the linear portion from 4.5 to 4.0V vs. Li/Li+ occupies a big part of the

capacity delivered by the positive electrode. Although some electrolyte decomposition

at the graphite positive electrode is inevitable at high potentials, it is not

124

Figure 4.19 Four electrode Potential profiles of the KS-6/RP-20 capacitor (+3.0V=Li/Li+) the positive and negative electrodes have the same weights of active material (5.5 mg), electrolyte: 1.5M TEMAPF6-PC).

so drastic since the surface area of graphite is not as high as activated carbon. During

the charge process of the graphite/AC capacitor, the anion-PC species are gathered at

the interface between graphite positive electrode and the electrolyte at first. Because the

surface area of the graphite positive electrode is not so big to accommodate many

adsorbed anion-PC species (highly polarizable electrode), the potentials of the graphite

electrode rises up quickly until it arrives at certain high value, say, 4 V vs. Li/Li+. The

-2

-1

0

1

2

3

4

0 10 20 30 40 50 60 70 80 90 100 110 120 130

Capacity / mAh/g

Pote

ntia

l / V

125

anions start to insert into the interlayer spaces of graphitic carbon and the potential

profile bend and rise up pace slowly (a nearly non-polarizable electrode).

If a reference electrode (Li metal) was introduced into a graphite/AC capacitor, the

respective potential of positive and negative electrodes can be traced clearly against the

reference electrode during galvanostatic charge-discharge curves. Fig.4.20 shows the

potential profiles for positive material KS-44 and negative (RP-20) electrode vs. Li

metal as well as the total voltage of KS-44/RP-20 capacitors ( KS-44 and RP-20 are

equal in weight). The total capacitor voltage equals to the difference between the

positive and negative electrode potentials. The potential profiles of KS-44 positive

electrodes are bend curves. For instance, in the charging process, the potential of

graphite jumps up to 4.5V vs. Li/Li+ at first and then climbs up slowly with times. It is

some electrolyte decomposition at the graphite positive electrode is inevitable at high

potentials, it is not so drastic since the surface area of graphite is not as high as activated

carbon. During the charge process of graphite/AC capacitor, the anion-PC species are

gathered at the interface between graphite positive electrode and electrolyte at first.

Because the surface area of the graphite positive electrode is not so big to accommodate

many adsorbed anion PC species (highly polarizable electrode), the potential of the

graphite positive electrode rise up quickly until it arrives at certain high value, 4.5V vs.

Li/Li+. Noticed that the charge process in the first cycle show quite similar features

from those in the following cycles. The potential of graphite positive electrode can

reach the voltage around 4.75V vs. Li/Li+, where as the potential of AC (RP-20) drops

down to about 1.29V vs. Li/Li+ at the end of initial charge process. Accordingly, the

total voltage curves of the graphite/AC are bent line, roughly comprise two linear

portions with different slopes. In the graphite/AC capacitor, the graphite positive

126

Figure 4.20 Potential profiles of KS-44/RP-20 ratio (1:1).

electrode works principally at rather high potential ranges, the linear portion from 4.5 to

4.7V against Li metal occupies a big part of the capacity delivered by positive electrode.

0 500 1000 15000.00.51.01.52.02.53.03.54.04.55.0

+ vs Li/Li+

- vs Li/Li+

Total voltage

Pote

ntia

l / V

Time/Second

127

4.3.6 Energy storage mechanism of Novel capacitor by in-situ XRD

I have found a new capacitor. So, in cell no intercalation phenomena are carried out

in charge process. From the comparison between graphite/AC and AC/AC capacitors, it

is clear that the new type capacitor (graphite/AC) has advantages in the terms of high

capacitance, high working voltage and high energy density. However, its potential

shortcomings must be taken into account before practical applications. I performed in

situ XRD measurements on the graphite positive electrode in the graphite/AC capacitor

to test no intercalation. Fig. 4.21 shows the in situ XRD patterns of the graphite (KS6)

positive electrode during the initial charge process of graphite/AC capacitor with weight

ratios of AC to graphite 1:1. At the low weight ratio of 1, the (002) diffraction peak of

KS6 positive electrode remains its shape and position in the XRD patterns in the whole

working voltage range. This means that intercalation does not take place because the

crystal lattice of the host (KS6) does not change apparently between cell voltage

0-3.5V.

128

Figure 4.21 Insitu XRD of KS-6/AC (RP-20) capacitor (1:1)during initial charging.

Beside KS-6, I have measured insitu XRD using the other synthetic graphite such as

KS-44, SFG-44 and SFG-6 at the working voltage of capacitor at 3.5V. Fig. 4.22 (a)

and (b) shows the in situ XRD patterns of KS-44 electrode during initial charge process

of KS-44/RP-20 and KS-44/PW15M capacitors (graphite/AC weight ratio equals to 1).

The (002) diffraction peak of KS-44 never split to lower Bragg angles accompanying

the rise in cell voltage. The (002) diffraction peak of KS-44 positive electrode remain

same in its shape and position in the XRD patterns in the whole working voltage range.

This means that deep intercalation does not take place because the crystal lattice of

KS-44 does not change apparently. The above fact implies that there is no insertion of

23 24 25 26 27 28 29 30 31

0

200400

600800

1000

1200

1400

16001800

2000

22002400

2600

2800

3000

In

tens

ity

2θ / o

2.5 V

3 V

3.1 V

3.2 V

3.3 V

3.4 V

3.5 V

OCV

2 V

2θ / deg (CuKα)

23 24 25 26 27 28 29 30 31

0

200400

600800

1000

1200

1400

16001800

2000

22002400

2600

2800

3000

In

tens

ity

2θ / o

2.5 V

3 V

3.1 V

3.2 V

3.3 V

3.4 V

3.5 V

OCV

2 V

2θ / deg (CuKα)

129

anions into interlayer space of KS-44 cathode. While in the case of KS-44/PW15M

capacitors too, we cannot find any change on the crystal lattice of KS-44 in the in-situ

XRD measurement.

Figure 4.22 (a) In-situ XRD of KS-44/RP-20 Capacitor (1:1) during initial charging

25 30 35

0.337ocv

1.5v 1.9v

2.3v

2.5v

2.7v

2θ / deg (Cukα)

2.9v

Inte

nsity

(a.u

.) 3.1v

3.3v

3.5v

3.7v

130

Figure 4.22 (b) In-situ XRD profile of KS-44/PW15M (1:1) during initial charging.

25 30 35

0.323v

2θ / deg (CuKα)

1.5v

1.9v 2.3v

2.5v

2.7v

2.9v

3.1v

Inte

nsity

(a.u

.)

3.3v

3.5v

3.7v

131

Fig.4.23 shows the in-situ XRD patterns of SFG-6 electrode during initial charge

process of SFG-6/RP-20 capacitor (graphite/AC weight ratio equals to1).

Corresponding to electrolyte TEMAPF6- in PC, the (002) diffraction peak of graphite

split and move to lower Braggs angles accompanying the rise in cell voltage. The fact

Figure 4.23 Insitu XRD profile of SFG-6/RP-20 (1:1) during initial charging.

implies that the insertion of anions into interlayer spaces of SFG-6 cathode. In the case

of SFG-6, the (002) diffraction peak start to split at about 3.1 V. Hahn et al have

observed the intercalation of 1M TEABF4 dissolved in acetonitrile [5]. The graphite

25 30 35

0.321ocv

2θ / deg (CuKα)

1.5V

Inte

nsity

(a.u

.)

1.9V

2.3V 2.5V

2.7V

2.9V

3.1V

3.3V

3.5V

132

SFG-6 cannot be able to use in capacitor cathode material.

The other electrochemical behavior of novel capacitor is such as follows:

High current application.

Long cycle life over than 10,000 cycles.

There are no peaks in CV.

All these phenomena show new capacitor works as pure EDLC but not a hybrid

capacitor or electrochemical capacitors.

4.3.7 Development of high energy density hybrid capacitor using high weight ratio of

anode activated carbon in graphite/AC capacitor

Because of low energy density of EDLC, an alternative choice is proposed as the

hybrid supercapacitor. An asymmetrical configuration in which battery electrode

replaces one activated carbon electrode [6], [7], [8] and [9]. The battery electrodes

accumulate charge through faradaic electrochemical process (redox reaction), which

cannot only increases the specific capacitance of the capacitor, but also extends the

working voltage. As a result, the energy density of capacitor is enlarged considerably.

Generally, there are three categories of redox-reaction materials are used in the hybrid

supercapacitors nowadays, metal oxides, conductive polymers and intercalation

compounds. But there is no commercial use of hybrid supercapacitors till now. These

types of capacitor would not be applicable to high current and temperature over than

60°C operations as like as battery.

The term intercalation literally refers to the reversible insertion of guest species into

a lamellar host structure with maintenance of the structural features of the host. The

133

term can equally be applied to one and three-dimensional solid and to inorganic

materials. The essential features of the intercalation reaction, and that which makes its

study so exciting and profitable, is that the guest and host experiences some degree,

along a spectrum from the subtle to extreme, of perturbation in their geometric,

chemical, electronic, and optical properties. Thus, hosts span the range of electronic

conductivity from insulator (such as MoO3, zeolites, and clays) through semiconductors

(graphite and transition metal dichalcogenides) to metals (LaNi5). This conductivity

behavior can change markedly on intercalation depending on the degree of electron

transfer between guest and host.

The electrochemical intercalation of anions from organic electrolytes into graphite

electrodes has been investigated towards battery utilization from 1970s to 1980s [10],

[11] and [12]. However the battery using graphite as a cathode has never been used

because of electrolyte decomposition at higher voltage.The storage capacity of graphite

is further increased by the means of adjusting the comparative weight ratio of activated

carbon electrode. The charge-discharge curves of KS-6/AC (RP-20) capacitor shows

increase in capacity with the increase in the case of high weight ratio of anode like 4, 8,

12 and 20 as shown in Fig.4.24. I have developed high energy density hybrid capacitor

using high weight ratio of anode (AC) in graphite/AC capacitor. KS-6/AC (RP-20)

capacitor gives capacities of 40, 65, 105, 120, 147 and 168 mAh g-1 as the weight ratio

of anode increased graphite from 1:1, 2:1, 4:1, 8:1, 12:1 and 20:1. It indicates the

formation of intercalation of anion.

134

Figure 4.24 Charge-discharge curves of KS-6/AC (RP-20) capacitor at different weight

ratio of anode to graphite.

Fig.4.25 shows the maximum capacity of several graphite samples with different

weight ratio of anode (RP-20) activated carbon charged with the voltage range of

3.5V-0V with constant current of 1mA. As the weight ratio of anode/ cathode is equal to

1, the entire graphitic sample gives nearly capacity of = 50 mAh g-1 in which no

intercalation occurs. As the weight ratio of anode to cathode increased up to 12,

MAG-D and MCMB6-28 gave highest capacity of over than 160 mAh g-1. But in the

case of Chinese natural graphite flake, it gives similar capacity of 110 mAh g-1 after

weight ratio of anode/cathode increased up to 4. From this figure, I could say that the

graphitic carbon will reach over than 160 mAh g-1 in which energy density of 400

Wh/kg is ca. 20 times larger than activated carbon.

0 50 100 150 200 250 3000

1

2

3

4

KS-6/AC-1(RP-20) weight ratio 1:1 1:2 1:4 1:8 1:16 1:20

Vol

tage

/ V

Capacity / mAh g-1(anode weight)

135

Figure 4.25 Capacity vs. weight ratio of anode/cathode using several graphite/RP-20 (activated carbon) capacitors

In our previous study [13], we show that peak splitting and shift happens for the

(002) peak in the case of another artificial graphite (MAG D-20 from Hitachi Kasei Ltd.

Co.) under similar conditions (graphite/AC weight ratio, 1; electrolyte, 1.5 M

TEMAPF6-PC). What is the main difference between KS6 and MAG D-20? Comparing

the XRD patterns of these graphite samples in the powder form, we found that MAG

D-20 is more crystalline than KS6 because the (002) peak of MAG D-20 is sharper than

that of KS6. Presumably, disorder in graphitic carbon helps depress the intercalation of

anions into interlayer spaces. The above assumption is similar to the case of lithium

cation intercalation into graphitic carbon [14]. On the other hand, the (002) peak of KS6

split and shifts to low diffraction angles at high graphite/AC weight ratios (2 and 4) as

shown in Fig.4.26 (a) and (b). The above fact implies that as the weight ratio of AC to

graphite rises, the anions insert more deeply into graphite lattice. This conclusion

0 2 4 6 8 10 120

40

80

120

160

200

KS-6 MCMB6-28 MCF MAG D-20 China NGFC

apac

ity /

mA

h g-1

Weight ratio of anode/cathode

136

accords with the result, that is, with the rise in weight ratio of AC to graphite, the cycle

performance of graphite/AC capacitor becomes worse. It is likely that bigger volume

changes of graphite matrix due to anion intercalation destroy more severely the crystal

structure of graphite electrode. The fact that the cycleability become worse with the

increase in weight ratio of AC to graphite.

Figure 4.26(a) Insitu XRD of KS-6/AC capacitor (ratio 1:2) using TEMA BF4-PC.

23 24 25 26 27 28 29 30 31

0

1000

2000

3000

4000

5000

Inte

nsity

2θ / o

OCV

3.3V

3.5V

3.4V

3.2V

3.1V

2.9V

2.7V

2.5V

1.9V

23 24 25 26 27 28 29 30 31

0

1000

2000

3000

4000

5000

Inte

nsity

2θ / o

OCV

3.3V

3.5V

3.4V

3.2V

3.1V

2.9V

2.7V

2.5V

1.9V

2θ / deg (CuKα)

23 24 25 26 27 28 29 30 31

0

1000

2000

3000

4000

5000

Inte

nsity

2θ / o

OCV

3.3V

3.5V

3.4V

3.2V

3.1V

2.9V

2.7V

2.5V

1.9V

23 24 25 26 27 28 29 30 31

0

1000

2000

3000

4000

5000

Inte

nsity

2θ / o

OCV

3.3V

3.5V

3.4V

3.2V

3.1V

2.9V

2.7V

2.5V

1.9V

2θ / deg (CuKα)

137

Figure 4.26 (b) Insitu XRD of KS6/AC capacitor (ratio 1:4) using TEMABF4-PC. Figure 4.26 In-situ XRD patterns of the graphite (KS6) positive electrode during the initial charge process of graphite/AC capacitor with different weight ratios of AC to graphite, electrolyte: 1.5M TEMABF4-PC: (a) graphite:AC = 1:2 and (b) graphite:AC =1.4

23 24 25 26 27 28 29 30 31

0

200

400

600

800

1000

1200

1400

1600

1800

In

tens

ity

2θ / o

OCV

1.6 V

2.0 V

2.2 V

2.3 V

2.6 V

2.7 V2.8 V

3.0 V

3.5 V

2θ / deg (CuKα) 23 24 25 26 27 28 29 30 31

0

200

400

600

800

1000

1200

1400

1600

1800

In

tens

ity

2θ / o

OCV

1.6 V

2.0 V

2.2 V

2.3 V

2.6 V

2.7 V2.8 V

3.0 V

3.5 V

2θ / deg (CuKα)

138

23 24 25 26 27 28 29 30 31 32 33 34 35

0

200

400

600

8001000

1200

1400

1600

1800

2000

22002400

2600

2800

3000

In

tens

ity

2θ / deg (CuKα)

OCV 1.9V 2.1V 2.2V 2.3V 2.4V 2.5V 2.6V 2.7V 2.8V 2.9V 3.1V 3.3V 3.5V

Figure 4.27 Insitu XRD of MCMB6-28/RP-20 capacitor (weight ratio 1:1).

Fig. 4.27 shows the in situ XRD of MCMB6-28/RP-20 capacitor (weight ratio 1:1)

during initial charging. The (002) diffraction peaks start to split at 2.5V and moves to

lower Braggs angle with the rise in cell voltage. We also found similar case in the case

of NG, MAG-D, and MCF too.

As measured from the four-electrode cells, the potential of graphite positive

electrode against the reference electrode likely experiences higher voltage load at higher

139

graphite/AC weight ratio. Table 4.1 lists the potentials of graphite and AC electrodes in

the capacitor (vs.Li/Li+) with different weight ratios of AC to graphite when the

capacitor was charged to 3.5 V.

Generally, both the ceiling potential of graphite positive electrode and the bottom

potential of AC negative electrode rise with the increase in the weight ratio of AC to

graphite. As the AC/graphite weight ratio becomes bigger than 2, the potential of

graphite positive electrode is higher than 5 V versus Li while the potential of AC

negative electrode climbs up to over 1.5V versus Li. Thus, the irreversible

decomposition at AC negative electrodes will become milder, whereas the irreversible

decomposition at graphite positive electrodes will get more severe. The heavy voltage

burden on the graphite positive electrode destroys the cycle performance of graphite/AC

capacitors.

Table .1 Four electrode potential of KS6/AC in the initial galvanostatic charge-discharge curves. Weight ratio of AC

to graphite Ceiling potential of graphite positive electrode (Vvs.Li /Li+)

Bottom potential of AC negative electrode(Vvs.Li/Li+)

1:1 3.83 1.34 1:2 5.04 1.55 1:4 4.99 1.50

140

4.4 Conclusion

As compared with the conventional EDLCs (symmetric AC/AC type), the

asymmetric graphite/AC capacitors have the advantages of high capacitance in the high

voltage range, high energy density, etc. Usually graphite is not a capacitor material

because of its small ability of adsorption. But, I have found if activated carbon (AC) is

used as a counter electrode, cathodic graphite can adsorb anion on non-aqueous solution.

In AC/AC capacitor, one electrode has enough ability for adsorption and counter

electrode does not need such strong ability of adsorption. So, I have proposed new idea

to replace one of its activated carbon by graphite carbon as a cathode for capacitor.

Some graphite were anion intercalated. In common graphite, intercalation of anion

takes place at ≥2.5V when graphite/AC weight ratio equals to 1. Common graphite

shows the 002 peak intensity value near 26.50-25.88°. I have selected graphite like

KS-6, KS-44, SFG-44 and SFG-6 from (Timcal Co. Ltd. Switzerland) which has low

crystal with small surface area. Graphite/AC capacitor shows capacity of 45 mAh g-1

when charged to 3.5V-0V. Graphite/AC capacitor shows at least 4 times high energy

density than AC/AC capacitor and show satisfactory performance in terms of high

voltage and high energy density due to the low crystal and small surface area of

graphitic carbon. On the other hand, graphite like KS-6, KS-44 and SFG-44,

intercalation of anion didn’t take place.

I have determined that the condition for the preparation of cell with no intercalation

of anion to the graphitic carbon.

1) Energy density is increased more than 4 times using KS-series graphite.

2) Maximum voltage is less than 3.7 V.

141

3) Weight ratio of AC/KS-6 is less than 1∼1.5 depending on graphitic carbons.

Under this condition, cell is working as capacitor not an electrochemical capacitor.

The weight ratio of negative electrode material (AC) to the positive electrode

material (graphite) has great effects on the electrochemical performance of the

graphite/AC capacitors. In situ XRD measurements prove that at high graphite/AC

weight ratios, anions intercalates into the interlayer spaces deeply during the charge

process of the capacitor. Moreover, high AC/graphite weight ratios cause the graphite

positive electrode to be burdened with more heavy voltage load. These factors lead to

the deterioration of the cycle-ability of the graphite/AC capacitors.

New type of hybrid or electrochemical capacitors have been developed changing the

weight ratio of anode/cathode materials.

142

References

1. Barbieri, M. Hahn, A. Herzog, and R. Kotz, Carbon, 43 (2005) 1303.

2. M. Hahn, A.Wursig, R. Gallay, P. Novak, and R. Kotz, Electrochem. Commun., 7

(2005) 925.

3. H. Marsh and F. R. Reinoso, Activated Carbon, Elsevier, p.155 (2006).

4. M. Yoshio, A. Kozawa, ed., Lithium Ion Battery, Nikka Kogyo, Kogyo (in

Japanese) p. 75 (2000), H. Yoshitake, Functional Electrolyte in M. Yoshio, A.

Kozawa, ed.

5. M. Hahn, O. Barbieri, F. P. Campana, R. Kotz, R. Gallay, Appl. Phys., A82 (2006)

633.

6. H. Li, L. Cheng, and Y. Xia, Electrochem. Solid-State Lett., 8 (2005) A433–A436.

7. G. G. Amatucci, F. Badway, A. DuPasquier, and T. Zheng, J. Electrochem. Soc.,

148 (2001) A930.

8. K. Naoi, S. Suematsu, M. Hanada, and H. Takenouchi, J. Electrochem. Soc., 149

(2002) 472.

9. W. Sugimoto, H. Iwata, Y. Yasunaga, Y. Murakami, and Y. Takasu, Angew. Chem.,

115 (2003) 4226.

10. A. Brenner, J. Electrochem. Soc., 118 (1971) 461.

11. T. Ohzuku, Z. Takehara, and S. Yoshizawa, Denki Kagaku (Electrochem.), 46

(1978) 438.

12. Y. Matsuda, M. Morita, and H. Katsuma, J. Electrochem. Soc., 131 (1984) 104.

13. H. Wang, and M. Yoshio, Electrochem. Commun., 8 (2006) 1481.

14. T. Zheng, J. N. Reimers, and J. R. Dahn, Phys. Rev. B, 51 (1995) 734.

143

Chapter 5

Conclusions

2. By solid state reaction method, fluorine and chlorine doped o-LiMnO2 cathode were

successfully synthesized at high temperature. The doping of fluorine and chlorine

contents was confirmed by ion exchange chromatography. The entire o-LiMnO2

sample shows Li2MnO3 impurity phase in the XRD patterns. The o-LiMnO2 (0.01≤

y ≤ 0.05) with fluorine and chlorine doped compounds show high capacities of 226

mAh g-1, and 232 mAh g-1 respectively, when it was operated at a current density of

0.4 mA cm-2 (40 mA g-1) in the voltage range of 2.0-4.5V at elevated temperature

(60°C). After 50 cycles, capacity fading rate was negligible on doped o-LiMnO2-yXy.

On the other hand, capacity fading rate was higher for undoped o-LiMnO2. The

o-LiMnO2 can be cycled at both 3V and 4V without any significant capacity fading

upon cycling. The doping of fluorine and chlorine had significant influence on the

crystal morphology of o-LiMnO2-yXy. Normally after synthesized, compounds

become harder thereby making it difficult to grind.

3. By applying carbonate co-precipitation, I have successfully synthesized the layered

LiNi0.5Mn0.5O2 by means of XRD, EDX, SEM and XPS spectroscopy. By X-ray

diffraction analysis, it was confirmed that all the material shows hexagonal phase

despite of different calcinations temperature. In the layered LiNi0.5Mn0.5O2 cathode

material, Ni and Mn are homogeneously distributed with round spherical particles.

The valence state of Ni and Mn were found to be +2 and +4, respectively by XPS

144

analysis. From electrochemical experiment, it was concluded that 900°C and 950°C

were an optimum calcination temperature for LiNi0.5Mn0.5O2 cathode material. The

layered LiNi0.5Mn0.5O2 cathode prepared at 900°C shows highest discharge capacity

of about 200 mAh g-1 at the voltage range of 2.8V-4.5V with current density of 0.2

mA cm-2 (20 mA g-1).

4. I have introduced megalo-capacitance capacitor using graphitic carbon as a cathode

material instead of activated carbon, whose energy density is 4 times larger than

common capacitors. Usually graphite is not a capacitor material because of its small

ability of adsorption. But, I have found if activated carbon (AC) is used as an anode,

cathodic graphite can adsorb anion on non-aqueous solution. In AC/AC capacitor,

one electrode has enough ability for adsorption and counter electrode does not need

such strong ability of adsorption. So, I have proposed new idea to replace one of its

activated carbon by graphite carbon as a cathode for capacitor.

Some graphite was anion intercalated. In common graphite, intercalation of

anion takes place at ≥2.5V. Common graphite shows the 002 peak intensity value

near 26.50-25.88°. I have selected graphite like KS-6 from Timcal Co. Ltd. which

has low crystal with small surface area. Graphite/AC capacitor shows capacity of 45

mAh g-1 when charged to 3.5V-0V. Graphite/AC capacitor shows 4 times high

capacity than AC/AC capacitor and show satisfactory performance in terms of high

voltage and high energy density due to the low crystal and small surface area of

graphitic carbon.

I have introduced different kinds of graphite like KS-44, SFG-44, and SFG-6

(Timcal Co. Ltd.) as the cathode material for graphite/AC capacitor. These graphite

samples were bit different in crystal structure and having small specific surface area

145

as compared to KS-6 graphite. SFG-44 and SFG-6 samples have similar peak

intensity value with crystal structure less than 6µm. KS- 44/RP-20 capacitor shows

capacity of 45 mAh g-1 at the voltage range of 3.5 V-0V with a current density of 0.5

mA cm-2. From in situ XRD result, it was confirmed that (002) peak does not split

and never move to lower Bragg angles accompanying the rise in cell voltage up to

3.5V. It indicates anion intercalation of graphite does not take place in the KS-6,

KS-44, and SFG-44 graphitic samples. Graphitic carbon like KS-6, KS-44 and

SFG-44 are most suitable cathodes for hybrid capacitor. But in the case of SFG-6,

during initial charging anion intercalation of graphite start to split at the 3.1V in situ

XRD patterns indicating anion intercalation phenomena occur in SFG-6.

Acknowledgement

I would like to express deep sense of appreciation and admiration to my senior

Professor Masaki Yoshio for his fruitful supervision and his encouragement during my

doctoral course. I express gratitude to Professor Hideyuki Noguchi for his solicitous

care and valuable discussion. Without their patient instruction, this work could not be

fulfilled smoothly.

I am deeply indebted to Professor Hiroyoshi Nakamura and Mr. Kenichi Isono for

their valuable discussion and unselfish help during three years.

I would like to thank all of the degree committee members, Professor Hideyuki

Noguchi, Professor Masamitsu Nagano, Professor Takanori Watari and Professor

Hiroyoshi Nakamura for their kind reviewing and valuable comments.

I would like to express appreciation to Dr. Y. S. Lee, Dr. H. Wang, Dr. N. Dimov, Dr.

D. Li, Dr. T. H. Cho, Dr. P. Ghimire, Dr. Q. Zhang, Dr. M. Zou, Mr. Y. Xia, Mr. G. J.

Park, Mr. Ono, Mr. Kawaguchi, Mr. Iguchi and all other students in this laboratory for

their kind help in this work.

Special thanks are given to Professor Raja Ram Pradhananga and Dr. Kedar Nath

Ghimire at Tribhuvan University, Nepal, for their innovative suggestion and

encouragement during my research periods.

I would like to thank sincerely to Department of Applied Chemistry, Saga

University, for providing financial support during doctoral course.

Last, but not least, I would like to give special thanks to my parents, brother, sisters

and wife, Arpana for their encouragement and moral support.

I would like to dedicate this thesis to my mother, Mrs. Jabuna Devi Thapa

A

PUBLICATIONS

1. Improvement in Electrochemical Performance of an o-LiMnO2-yXy at High

Temperature, ITE Letters, Vol. 8, No. 1 (2007) 22-28.

Arjun Kumar Thapa, Hiroyoshi Nakamura, Hideyuki Noguchi and Masaki Yoshio

2. From Symmetric AC/AC to Asymmetric AC/graphite, a Progress in

Electrochemical Capacitors, Journal of Power Sources, 169 (2007) 375.

Hongyu Wang, Arjun Kumar Thapa, Hiroyoshi Nakamura and Masaki Yoshio

3. Novel Synthesis of LiNi0.5Mn0.5O2 by Carbonate Co-precipitation, as an Alternative

Cathode for Li-ion Battery, ITE Letters, Vol. 8, No. 4 (2007) B11-18.

Arjun Kumar Thapa, Hiroyoshi Nakamura, Hideyuki Noguchi and Masaki Yoshio

4. Enhanced Electrochemical Performance and Preparation of Layered LiNi0.5Mn0.5O2

by Carbonate Co-precipitation, Electrochemistry, under revision.

Arjun Kumar Thapa, Tae-Hyung Cho, Hongyu Wang, Hiroyoshi Nakamura,

Hideyuki Noguchi and Masaki Yoshio

5. Development of Novel Megalo-Capacitance Capacitor, Proceeding of International

Conference on Advanced Capacitor (ICAC2007), Kyoto, p 67-68, May 28, (2007)

Japan

Arjun Kumar THAPA, Hongyu WANG, Hiroyoshi NAKAMURA and Masaki

YOSHIO

B

Academic Activities

1. Improvement in Battery Performance of a new types o-LiMnO2-yXy Cathode for

Li-ion Battery

Arjun Kumar Thapa, Hiroyoshi Nakamura, Masaki Yoshio and Kazuyuki Adachi

The 71st Meeting of Electrochemical Society of Japan, Yokohama, Japan, April

24-26, 2004.

2. Characterization and Synthesis of LiNi0.5Mn0.5O2 by Carbonate Co-precipitation

Method

A. K. Thapa, S. M. Park, T. H. Cho, H. Nakamura, H. Noguchi and M. Yoshio

The 73rd Meeting of Electrochemical Society of Japan, Tokyo, Japan, April 1-3,

2006.

3. Development of Advanced Megalo-Capacitor Using Graphitic Carbon Material (1)

Arjun Kumar Thapa, Hongyu Wang, Hiroyoshi Nakamura, Masaki Yoshio and

Tetsuya Ishihara

The 2006 Meeting of Electrochemical Society of Japan, Kyoto, Japan, September

14-15, 2006.

4. Development of Advanced Megalo-Capacitance Capacitor 1. Effect on Cathode

Material

H. Nakamura, A. K. Thapa, Z. Meijing, H. Wang, M. Yoshio, T. Ishihara and H.

Nakamura

The 47th Battery Symposium of Japan, Tokyo, Japan, November 20-22, 2006.

5. Development of Advanced Megalo-Capacitor Using Graphitic Carbon Material (3)

C

Arjun Kumar Thapa, Hongyu Wang, Hiroyoshi Nakamura and Masaki Yoshio

The 74th Meeting of Electrochemical Society of Japan, Tokyo, Japan, March 28-30,

2007.

6. Development of Novel Megalo-Capacitance Capacitor

Arjun Kumar THAPA, Hongyu WANG, Hiroyoshi NAKAMURA and Masaki

YOSHIO

The 2007 International Conference on Advanced Capacitor (ICAC2007), Kyoto,

Japan, May 28-30, 2007.