PREPARATION, MODIFICATION AND CHARACTERIZATION OF
ACTIVATED CARBONS FOR BATCH ADSORPTION STUDIES ON THE
REMOVAL OF SELECTED METAL IONS
MOHAMMAD ADIL
A thesis submitted in fulfilment of the
requirements for the award of the degree of
Master of Science (Chemistry)
Faculty of Science
Universiti Teknologi Malaysia
JULY 2006
iii
This thesis is dedicated to the memories of my beloved late father
iv
ACKNOWLEDGEMENT
I would like to thank my supervisor Professor Dr. Alias M. Yusof for giving
me the opportunity to work on his project as well as for his valuable guidance,
support and untiring patience. I would also acknowledge the financial support by his
research grant (VOT 74029 and 74513).
I am grateful to Associate Professor Dr. Noor Aini Bt. Abdul Rashid for her
constant vigilance and valuable suggestions throughout this study. I would also
express my appreciation to all other faculty members and staffs in the Chemistry
Department of UTM for their enormous help with my study. I would also extend my
sincere appreciation to other faculties and institutions related to my job with them.
I thank all of my friends, colleagues and laboratory personnel who extended
their time, expertise, generous advice, criticism, technical assistance and
encouragement during my research. I like to acknowledge everyone, but I am to be
constrained to a few in mentioning names as Mr. Ayob Jabal, Mr. Hanan Basri, Mr.
Azmi M. Rais, Mrs. Z. Ain Jalil, Mr. Hj. Yasin bin. M. Sirin, Miss. Nurul H.
Sapiren, Mr. M. Nazri Zainal, Mr. Dinda Hairul, Mr. Hamzah, Mr. Abdul Kadir,
Mrs. Mek zum, Mr. Abdur Rahim, Mrs. Mariam Hassan, Mr. Azani b. Ishak of
Chemistry Department and Chia Chai Har, Tan See Hua, Lee Kian Keat and Ng Shi
Hooi who indebted me most for their assistance in pursuing laboratory work.
I am thankful to the authorities of Bangladesh Chemical Industries
Corporation (BCIC) and Chittagong Urea Fertilizer Limited (CUFL), for granting
continual periods of study leave which made this study possible. A deep and sincere
gratitude are due to Mr. Abdul Hafiz (Secretary, BCIC), Mr. M. Shahjahan (General
Manager, Construction), Mr. M. Mostafa Faruque (General Manager), Mr. M.
Salahuddin (General Manager, Operation), Mr. Syed M. Ilias Chowdhury (Plant Incharge, Ammonia), Mr. Abu Jamal Sarkar (Additional Chief Chemist), Mr. AJM
Ataul Karim (Additional Chief Chemist), Mr. Ruhul Amin (Plant In-charge, Urea),
Mr. Shishir K. Dutta, Mr. Mesbah Saklaein and to all of my colleagues and staffs
there, for their support and encouragement during my study at UTM in Malaysia.
I owe thanks to my all family members, my mother, brothers and sisters for
their help and love without that I cannot continue my study here. I would
acknowledge my wife for her patience and tolerance of domestic disorganization,
forbearances and a sincere support throughout this period. Finally, I must have to
thank my daughter, “Parna” who missed her father most of the time during this
study.
v
ABSTRACT
Among various pollutants in drinking and wastewaters, inorganic species are
major, and difficult to remove because of smaller size, low concentration and
competition with nontoxic species. In this work, various activated carbons have been
prepared by phosphoric acid activation from oil palm and coconut shells to study the
effect of preparation variables and metal removal capacity. Characterizations were
done for the observation of surface area, pore development, textural structure,
surface functional groups and its acidity. Thus preparation was customized for
carbon with high surface area and wider micro- and mesopores, suitable for
removing metal ions. Modification was conducted by dispersing hydrated iron-oxide
onto carbon for selective removal of arsenic. Application was tested through batch
adsorption in single, binary and ternary solutes solutions and fitting the equilibrium
data to various equilibrium and kinetic models. Prepared carbons show very high
adsorption capacities for borderline and soft metal ions, copper(II) (25 mg.g-1),
nickel(II) (20 mg.g-1), zinc(II) (10 mg.g-1), and lead(II) (75 mg.g-1), compared to that
of commercial ones. Adsorption in binary solution shows that hard cation,
calcium(II) is not at all competent to borderline cations while the mechanism is
assessed as inner-sphere complexation. For low concentration range, ternary-solute
adsorption shows the following selectivity order, Ni(II) ≤ Pb(II) ≤ Cu(II). Adsorption
capacity of prepared activated carbon for Cr(VI) is saturated at 46 mg.g-1. Although
the adsorption capacity of commercial activated carbon dramatically decreases with
the lowering of initial solute concentration, it shows a higher monolayer adsorption
capacity (71 mg.g-1). The modification significantly enhances the adsorption
capacity for As(III) (1.5 mg.g-1) and As(V) (1.0 mg.g-1) than that of parent carbon.
The developed activated carbons and modified adsorbents are expected to be useful
in removing borderline and soft metal ions as well as ionic or nonionic soft ligands
of metalloids present in drinking and wastewaters.
vi
ABSTRAK
Di antara pelbagai jenis bahan pencemar yang terdapat di dalam air sisa dan
air minuman, spesies-spesies tak organik adalah pencemar utama dan sukar untuk
disingkirkan disebabkan oleh saiznya yang lebih kecil, kepekatan yang rendah dan
persaingan dengan spesies-spesies yang tidak toksik. Dalam kajian ini, pelbagai jenis
karbon teraktif telah disediakan daripada tempurung kelapa sawit dan kelapa melalui
pengaktifan asid fosforik untuk mengkaji kesan bagi pelbagai pembolehubah dalam
penyediaan serta kapasiti penyingkiran logam. Pencirian telah dijalankan dengan
memerhatikan luas permukaan, pembangunan keliangan, struktur tekstur, kumpulan
berfungsi pada permukaan dan keasidannya. Justeru, penyediaan karbon dengan luas
permukaan yang tinggi dan keliangan mikro serta meso yang lebih besar, adalah
sesuai untuk penyingkiran ion-ion logam yang telah dikenalpasti. Pengubahsuaian
dilakukan dengan penyebaran ferum oksida terhidrat ke atas karbon bagi kepilihan
penyingkiran arsenik. Penggunaannya telah dikaji melalui penjerapan berkelompok
di dalam larutan yang mepunyai zat terlarut tunggal, dedua atau tetiga dan data-data
keseimbangan disuaikan terhadap pelbagai model keseimbangan dan kinetik.
Karbon-karbon yang disediakan menunjukkan kapasiti penjerapan yang tinggi
terhadap ion-ion perbatasan dan ion-ion lembut, kuprum(II) (25 mg.g-1), nikel(II) (20
mg.g-1), zink(II) (10 mg.g-1), dan plumbum(II) (75 mg.g-1), apabila dibandingkan
dengan karbon komersial. Penjerapan dalam larutan dedua menunjukkan penjerapan
kation-kation keras, kalsium (II) adalah lebih lemah daripada kation-kation
perbatasan sementara mekanismanya didapati sebagai pengkompleksan sferadalaman. Bagi lingkungan kepekatan rendah, penjerapan zat terlarut tetiga
menunjukkan susunan kepilihan berikut, Ni(II) ≤ Pb(II) ≤ Cu(II). Kapasiti
penjerapan Cr(IV) bagi karbon teraktif yang disediakan mencapai ketepuan pada 46
mg.g-1. Walaupun kapasiti penjerapan bagi karbon komersial merosot secara
dramatik ketika kepekatan zat terlarut awal dikurangkan, namun ia menunjukkan
kapasiti penjerapan lapisan tunggal yang lebih tinggi (71 mg.g-1). Pengubahsuaian
secara jelas meningkatkan kapasiti penjerapan bagi As(III) (1.5 mg.g-1) dan As(V)
(1.0 mg.g-1) berbanding dengan karbon induk. Karbon teraktif yang dibangunkan dan
penjerap yang terubahsuai diharapkan berupaya untuk menyingkirkan ion-ion logam
perbatasan dan lembut serta ligan ionik dan bukan ionik bagi metaloid yang terdapat
dalam air-air sisa dan air minuman.
vii
TABLE OF CONTENTS
CHAPTER
TITLE
PAGE
THESIS STATUS DECLARATION
SUPERVISOR’S DECLARATION
1
TITLE PAGE
i
DECLARATION
ii
DEDICATION
iii
ACKNOWLEDGEMENT
iv
ABSTRACT
v
ABSTRAK
vi
TABLE OF CONTENTS
vii
LIST OF TABLES
xii
LIST OF FIGURES
xvi
LIST OF SYMBOLS
xxii
LIST OF ABBREVIATIONS
xxiii
LIST OF APPENDICIES
xxv
INTRODUCTION
1
1.1
Research Background
1
1.1.1
2
1.2
1.3
Water Purification
Heavy Metals
5
1.2.1
5
Pollution and Remediation Approaches
Activated Carbon
13
1.3.1
Historical Aspect of Carbon
14
1.3.2
Structure of Activated Carbon
15
viii
1.3.2.1
1.3.3
Surface Functional Groups
16
Preparation
20
1.3.3.1
23
Processing of Lignocellulosic
Materials
1.3.3.2
Utilization of Oil Palm and
27
Coconut Shells as Carbon
Precursor
1.4
Adsorption of Metal Ions onto Activated Carbon
30
1.5
Adsorption Equilibrium Models
41
1.5.1
Langmuir Adsorption Model
42
1.5.2
Freundlich Adsorption Model
43
1.6
1.7
2
Aims and Objectives
45
1.6.1
46
Objectives
Scope and Limitations
46
MATERIALS AND METHODS
47
2.1
Chemicals and Reagents
47
2.2
Instruments
48
2.3
Preparation of Activated Carbon
48
2.3.1
Raw Material Preparation
49
2.3.2
Activation and Washing
50
2.4
Modification of Activated Carbon
51
2.5
Characterization of Activated and Composite
53
Carbon
2.5.1
Yield Determination
54
2.5.2
Surface Porosity Characterization (N2
54
Adsorption-Desorption)
2.5.3
Surface Morphology (SEM and EDX)
55
2.5.4
Fourier Transform Infrared Spectroscopy
55
(FTIR)
2.5.5
Boehm’s Titration
55
2.5.6
Determination of pH at Point of Zero
56
Charge (pHZPC)
ix
2.6
2.5.7
Thermogravimetric Analysis (TGA)
57
2.5.8
Leaching Test of Fe
57
Adsorption Studies
58
2.6.1
Test and Standard Metal Ion Solutions
58
2.6.2
Adsorption Equilibrium Studies: Metal
60
Cations
2.6.2.1
Selection of Prepared Activated
60
Carbons for Metal Cations
2.6.2.2
Single Solute Adsorption
61
2.6.2.3
Binary Solute Adsorption
62
2.6.2.4
Ternary Solute Adsorption
62
2.6.2.5
Reuse of Cation-Exhausted
62
Activated Carbon
2.6.3
Adsorption Equilibrium Studies: Cr(VI)
63
2.6.4
Adsorption Equilibrium Studies: As(III) and
64
As(V)
2.6.5
Adsorption Dynamics: Cu(II), Ni(II), Pb(II)
64
and Cr(VI)
2.6.6
Determination and Uptake of Species of
65
Metal/Metalloid
3
RESULTS AND DISCUSSION: ADSORBENT
67
DEVELOPMENT
3.1
Preparation
67
3.1.1
Activated Carbon
67
3.1.2
Carbon-Hydrated Iron Oxide (C-HFO)
68
Composite
3.2 Characterization of Prepared Activated Carbon
3.2.1
69
Surface Porosity and Morphology
70
3.2.1.1
Effect of Precursor Type
72
3.2.1.2
Effect of Pretreatment
72
3.2.1.3
Effect of Impregnation Condition
76
3.2.1.4
Effect of Precursor Size
78
x
3.2.1.5
3.2.2
Effect of Activation Temperature
80
Surface Functional Groups
82
3.2.2.1
83
Fourier Transform Infrared (FTIR)
Spectroscopy
3.2.2.2
Boehm’s Titration
84
3.2.2.3
Determination of pHZPC of
86
activated carbon (pH Drift
Method)
3.2.3
Thermogravimetric Analysis (TGA) of
87
activated carbon
4
3.3
Characterization of C-HFO Composites
87
3.4
Leaching Test of Fe from C-HFO Composites
92
RESULTS AND DISCUSSION: APPLICATION OF
94
ADSORBENTS
4.1
4.2
Application of Prepared Adsorbents
94
4.1.1
95
Selection of Prepared Activated Carbon
Adsorption Equilibria
97
4.2.1
98
Single Solute Adsorption Equilibria: Metal
Cations
4.2.1.1
Single Solute Adsorption
98
Equilibria: Cu(II)
4.2.1.2
Single Solute Adsorption
103
Equilibria: Ni(II)
4.2.1.3
Single Solute Adsorption
107
Equilibria: Zn(II)
4.2.1.4
Single Solute Adsorption
110
Equilibria: Pb(II)
4.2.2
Binary Solute Adsorption Equilibria:
115
Competitive Adsorption of Cu(II) and
Ca(II) (Cu-Ca) and Evaluation of
Mechanism
4.2.3
Ternary Solute Adsorption Equilibria:
122
xi
Competitive Adsorption of Cu(II), Ni(II)
and Pb(II) (Cu-Ni-Pb)
4.2.4
Reuse of Cation-Exhausted Activated
128
Carbon
4.2.5
Adsorption Studies: Cr(VI)
130
4.2.5.1
Selection of Prepared Adsorbents
130
4.2.5.2
The Effect of pH
132
4.2.5.3
Effect of Initial Concentration and
134
Data Fitting to Models
4.2.6
4.3
Adsorption Studies: Arsenic
137
Adsorption Dynamics
142
4.3.1
144
Adsorption Dynamics of Cu(II) on PSW-Pad-500
4.3.2
Adsorption Dynamics of Ni(II), Pb(II) and
146
Cr(VI) on PSW-P-ad-500
4.4
5
Summary of Thesis
150
CONCLUSIONS AND SUGGESTIONS FOR
152
FUTURE STUDY
5.1
Conclusions
152
5.2
Problems Encountered in Present Study
155
5.3
Recommendations for Further Studies
156
REFERENCES
157
APPENDICES A - L
166
xii
LIST OF TABLES
TABLE NO.
1.1
TITLE
A survey of recent literatures on the production of
PAGE
23
activated carbons from various lignocellulosic precursors
1.2
Proximate analysis of solid oil palm residues (Husain,
29
Zainac and Abdullah, 2002)
1.3
Ultimate analysis of solid oil palm residues (Mahila et al.,
30
2001)
1.4
Proximate and ultimate analyses of coconut shell
30
(Banerjee, et al.,1976)
2.1
Preparation variables of various activated carbons activated
51
either in a muffle (M) or in a tube furnace (T)
2.2
The operating parameters of FAAS in the determination of
66
metal cations
3.1
The contents of Fe, oxygen, Na/Ca and C in the prepared
69
composites and in their parent activated carbons
3.2
Yield, BET surface area, average pore width, total pore
71
volume and pore-types of various activated carbons
3.3
Contents of phosphorous (P) in some of the selected
71
activated carbons as determined by EDX
3.4
Base neutralization capacity and quantification of acidic
85
groups on surfaces of some of the prepared activated
carbons by Boehm’s titration method
3.5
The pHZPC of some activated carbons obtained from pH
86
drift method
3.6
The proximate analysis of activated carbons using TGA
87
xiii
technique
3.7
The BET surface area, average pore width, total pore
88
volume and pore-types of C-HFO composites and their
parent activated carbons
3.8
Properties of Virgin GAC and Fe-GAC obtained from
92
Norit Americas Inc. (Reed, Vaughan and Jiang, 2000)
4.1
Cu(II) uptake capacity of various activated carbons in a
96
single solute solution while initial solution (adsorbate)
concentration and pH were kept constant at 292 ± 6 mg.L-1
and 6 respectively with a constant adsorbent concentration
of 2 g.L-1
4.2
Langmuir and Freundlich model parameters for the
101
adsorption of Cu(II) in single-solute solution while initial
concentrations were varied, < 5 - < 120 mg.L-1
4.3
Langmuir and Freundlich model parameters for the
105
adsorption of Ni(II) by various activated carbons at initial
pH 5 while initial concentrations were varied from 6 to 227
mg.L-1 and adsorbent concentration was kept constant at 2
g.L-1
4.4
The initial and equilibrium concentrations, and fractions
107
removal of Ni(II) by various activated carbons in the low
concentration solution at a constant initial pH 5
4.5
Langmuir model parameters for the adsorption of Zn(II) by
110
various activated carbons at initial pH 5 while initial
concentrations were varied from 19 to 310 mg.L-1 and
adsorbent concentration was kept constant at 2 g.L-1
4.6
Langmuir and Freundlich model parameters for the
112
adsorption of Pb(II) at initial pH 5 and at varying
concentrations, < 50 - < 600 mg.L-1
4.7
The initial and equilibrium concentrations, and fractions
removal of Pb(II) by various activated carbons in the low
concentration range at pH 5
113
xiv
4.8
Langmuir and Freundlich parameters for the adsorption of
115
Pb(II) on PSW-P-ad-500 for initial pH 3 at varying
concentrations, 50.6 ± 0.6-583 ± 2 mg.L-1
4.9
The Langmuir model parameters of Cu(II) and Ca(II) for
117
the adsorption in binary solution at an initial pH 5, while
initial concentrations of Cu(II) and Ca(II) were varied as
23.6 ± 0.1 to 117.5 ± 0.7 and 23.8 ± 0.2 to 120 ± 1 mg.L-1
respectively
4.10
Langmuir and Freundlich parameters for the adsorption of
120
Cu(II) on PSW-P-ad-500 at a constant initial pH 4.5 while
initial solution concentrations were varied from < 25
to < 700 mg.L-1 keeping adsorbent concentration 2 g.L-1
4.11
The Langmuir model parameters of Cu(II) and Ca(II) for
122
the adsorption in binary solution at initial pH 3 while
initial concentrations of Cu(II) and Ca(II) were varied as
26-151 and 20-109 mg.L-1 mg.L-1 respectively
4.12
Langmuir and Freundlich model parameters for the
124
isotherms of Ni(II), Pb(II) in ternary and Cu(II) in single
and ternary solutes solutions at initial pH 5
4.13
The initial and equilibrium concentrations, adsorption
125
capacities and fractions removal of Cu(II), Ni(II), Pb(II) in
the single and ternary solutes solutions by the best-selected
carbon, PSW-P-ad-500 at initial pH 5 and at adsorbent
concentration of 2 g.L-1
4.14
Langmuir model parameters for the adsorption-desorption
130
of Cu(II)
4.15
The adsorption capacities of various adsorbents in the
132
removal of Cr(VI) while initial pH was obtained without
adjustment as 4.79 and adsorbent concentration was
maintained as 2 g.L-1
4.16
Langmuir and Freundlich model parameters for the
adsorption of Cr(VI) on PSW-P-ad-500 while initial
concentrations were varied from 40 - < 200 mg.L-1 at an
136
xv
optimum pH 3, keeping adsorbent concentration constant
at 2 g.L-1
4.17
The initial and equilibrium concentrations, and fractions
136
removal of Cr(VI) by various activated carbons in the low
concentration at an optimum pH 3
4.18
Aqueous and solid phase equilibrium concentration of
139
adsorbents (adsorption capacity) while adsorbate
concentration was 1 mg.L-1, pH 7 and adsorbent
concentration 0.2 g.L-1 (Reed, Vaughan and Jiang, 2000)
4.19
Experimental adsorption equilibrium capacity, qe and
146
kinetic model parameters for the adsorption of Cu(II) at
constant initial concentration, < 50 mg.L-1 and at pH 5,
while contact time was varied from 0.5 to 72 h
4.20
Experimental adsorption equilibrium capacity, qe and
kinetic model parameters for the adsorption of Ni(II),
Pb(II) and Cr(VI)
147
xvi
LIST OF FIGURES
FIGURE NO.
1.1
TITLE
The periodic table showing classification of metals based
PAGE
11
on Lewis acidity as: Class A-hard metals (black), Class
B-soft metals (light grey), Class AB-borderline (dark
grey) (IUPAC, 2002)
1.2
Structures of carboxyl, phenolic hydroxyl and quinone
18
type carbonyl groups
1.3
Structures of lactones, acid anhydrides and cyclic
18
peroxide
1.4
A cross section of masocarp of oil palm fruit bunch
28
showing fibre, shell, kernel and pericarp
1.5
Process flow diagram of palm oil extraction showing the
29
production of crude oil and wastages such as shells,
kernel and fibres
2.1
A flow diagram of activated carbon preparation
49
2.2
A flow diagram of modification of activated carbons
52
3.1
N2 adsorption-desorption isotherms of various activated
70
carbons prepared from oil palm and coconut shell
precursors under various process conditions
3.2
N2 adsorption-desorption isotherms of activated carbons
73
derived from oil palm (A) and coconut shell (B)
precursors
3.3
Yield of various activated carbons as a function of
precursor pretreatment such as non-pretreated (NP),
74
xvii
H2SO4 pretreated (SAP) and H3PO4 pretreated (PAP)
conditions
3.4
BET surface area of various activated carbons as a
75
function of precursor pretreatment such as non-pretreated
(NP), H2SO4 pretreated (SAP) and H3PO4 pretreated
(PAP) conditions
3.5
BET surface areas of various activated carbons derived
77
from different charge states such as wet (PSW-P-500),
apparently/semi-dried (PSW-P-ad-500) and completely
dried (PSW-P-ed-450)
3.6
Total pore volumes of activated carbons derived from
77
different charge states, as wet (PSW-P-500), semi-dried
(PSW-P-ad-500) and dried (PSW-P-ed-450)
3.7
SEM micrograph of an activated carbon obtained from
78
semi-dried feedstock of oil palm shell precursor, namely
PSW-P-ad-500
3.8
N2 adsorption-desorption isotherms of two activated
79
carbons, derived from smaller (1.18-0.6 mm) and larger
(2.36-1.18 mm) particle sizes, namely PSWfg-P-450 and
PSW-P-ed-450
3.9
The development of surface areas of various activated
80
carbons prepared under different activation temperatures
3.10
FTIR spectrums of some of the prepared activated
83
carbons, namely, PSW-P-500, CPW-P-500 and PSW-Pad-500
3.11
Graphs of equilibrated pH (pHf) vs. initial pH (pHi) by
86
some of the prepared activated carbons for determining
the pHZPC from the crossover point
3.12
SEM micrographs of composites of C-HFO-1 (A) and C-
89
HFOCa-1 (B) obtained from parent activated carbon
PSW-P-500 using different bases, such as NaOH and
Ca(OH)2
3.13
SEM micrographs of a composite, C-HFO-2 obtained
90
xviii
from parent activated carbon CPW-P-500 while
magnification factor is different as 200X (A) and 800X
(B)
3.14
SEM micrographs of C-HFO-3 obtained from parent
91
activated carbon PSW-P-ad-500 while magnification
factor is different as 250X (A) and 1200X (B)
4.1
Langmuir adsorption isotherms of Cu(II) at initial pH 3
99
(A) and pH 5 (B) by various activated carbons while
initial solution concentrations were varied from < 5 to
<120 mg.L-1
4.2
Langmuir adsorption isotherms of Cu(II) by prepared
100
activated carbons for initial pH 3 (A) and pH 5 (B) and
by a commercial activated carbon for initial pH 3 and pH
5 (C) while initial concentrations were varied, < 5 - < 120
mg.L-1
4.3
Freundlich adsorption isotherms of Cu(II) at pH 3 (A)
101
and at pH 5 (B) while initial solution concentrations of
Cu(II) were varied, < 5 - < 120 mg.L-1
4.4
Langmuir (A, B, C) and Freundlich (D) adsorption
104
isotherms of Ni(II) at initial pH 5 by various activated
carbons while initial solution concentrations were varied
as 6 - 227 mg.L-1
4.5
Langmuir non-linear (A) and linear (B) isotherms of
108
Zn(II) by CAC at initial pH 5 while solution
concentrations were varied, 19 - 250 mg.L-1
4.6
Langmuir non-linear isotherms of Zn(II) at initial pH 5 by
109
prepared activated carbons, PSW-P-500 (A), CPW-P-500
(B) and PSW-P-ad-500 (C) while initial solution
concentrations were varied, 19 – 310 mg.L-1
4.7
Langmuir linear isotherms of Zn(II) at initial pH 5 by
various prepared activated carbons while initial solution
concentrations were varied from 19 to 310 mg.L-1
110
xix
4.8
Langmuir non-linear (A) and linear (B and C) adsorption
111
isotherms of Pb(II) at initial pH 5 by various activated
carbons while initial concentrations were varied as < 50 < 600 mg.L-1
4.9
Freundlich adsorption isotherms of Pb(II) at initial pH 5
112
by various activated carbons while initial concentrations
were varied as < 50 - < 600 mg.L-1
4.10
Langmuir non-linear (A) and linear (B) adsorption
114
isotherms of Pb(II) at initial pH 3 by the best-selected
activated carbon, PSW-P-ad-500 while initial solution
concentrations were varied from 50.6 ± 0.6 to 583 ± 2
mg.L-1
4.11
Freundlich adsorption isotherms of Pb(II) at initial pH 3
115
by activated carbon, PSW-P-ad-500 while initial solution
concentrations were varied from 50.6 ± 0.6 to 583 ± 2
mg.L-1
4.12
Langmuir non-linear and linear isotherms of Cu(II) (A
116
and B) and Ca(II) (A and C) while Cu(II) ions were
adsorbed from single (Cu), binary (Ca-Cu) and ternary
(Cu-Ni-Pb) solutions and Ca(II) from binary solution
(Ca-Cu) by PSW-P-ad-500 at initial pH 5 and at varying
initial concentrations, < 25 - ≤ 120.mg.L-1
4.13
The proposed interactions of Cu(II) and Ca(II) cations
118
(Lewis acids) onto the surfaces of activated carbon where
SO− and SOH (Lewis bases) represent the deprotonated
surface functional groups on activated carbon. The
vertical bars are depicted as solution-solid interface
4.14
Langmuir non-linear (A), linear (B) and Freundlich (C)
119
isotherms of Cu(II) by the carbon PSW-P-ad-500 at
constant initial pH 4.5 while initial solution
concentrations were varied from < 25 to < 700 mg.L-1
keeping adsorbent concentration constant at 2 g.L-1
4.15
Langmuir non-linear (A) and linear (B and C) isotherms
121
xx
of Cu(II) and Ca(II) for the adsorption in binary solute
solution while concentration of Cu(II) and Ca(II) were
varied as 26-151 and 20-109 mg.L-1 respectively keeping
initial pH and adsorbent concentration constant, at pH 3
and 2 g.L-1 respectively
4.16
Langmuir non-linear (A), linear (B) and Freundlich (C)
123
isotherms of ternary solute, Cu(II), Ni(II) and Pb(II), by
the best selected carbon, PSW-P-ad-500 at initial pH 5,
while their initial concentrations were varied as < 25 ≤ 600 mg.L-1. (Ni samples were analysed for initial
concentration range, < 25 to 305 ± 2 mg.L-1)
4.17
Langmuir non-linear and linear adsorption-desorption
129
isotherms of Cu(II), while adsorbed at varying
concentration, < 25 to < 700 mg.L-1
4.18
Adsorption capacity of Cr(VI) by various selected
133
adsorbents while pH were varied from 3 to 8 keeping
initial concentration constant at about 40 mg.L-1
4.19
Langmuir non-linear (A), Linear (B) and Freundlich (C)
135
adsorption isotherms of Cr(VI) by activated carbons,
CAC and PSW-P-ad-500, at initial pH 3 while initial
concentrations were varied from 40 - < 200 mg.L-1
4.20
Adsorption equilibrium capacities of a HFO composite
138
and its counter activated carbon for the species of As(V)
and As(III), while adsorptions were carried out in the
solutions of As(V) and As(III) with constant initial
concentrations, 1.641 and 2.170 mg.L-1 and with constant
initial pH, at 5.5 and 7.6 respectively
4.21
An illustration of selective bond formation of dispersed
141
hydrated iron oxide with As(V) and As(III) species in
aqueous phase predominantly through Lewis acid-base
interaction
4.22
Graph to check kinetics data of Cu(II) with first order
kinetic model
144
xxi
4.23
Graph of pseudo-second order kinetic model for the
145
adsorption of Cu(II) at constant initial concentration, < 50
mg.L-1 and at pH 5
4.24
A graph of adsorption capacity vs. contact time for the
145
adsorption of Cu(II) at constant initial concentration, < 50
mg.L-1 and at pH 5
4.25
Graphs of pseudo-second order kinetic model for the
147
adsorptions of Ni(II), Pb(II) and Cr(VI) at constant initial
concentrations, < 45, 95 ± 1 and 77 ± 1 mg.L-1
respectively, while the initial pH of solutions of Ni(II)
and Pb(II) were adjusted to 5 and that of Cr(VI) to 3
4.26
Graphs of adsorption capacities vs. contact time for the
148
adsorption of Ni(II) (A), Pb(II) (B) and Cr(VI) (C)
4.27
A comparison in the rate constants of various metal ions
149
for the adsorption onto acid activated carbon, PSW-P-ad500
4.28
A comparison in adsorption capacities (qe) of various
metal ions while qe obtained from adsorption equilibrium
data for kinetics studies for the adsorption onto acid
activated carbon, PSW-P-ad-500
150
xxii
LIST OF SYMBOLS
°C
-
Degree Celsius
C0
-
Initial concentration
Ce
-
Equilibrium concentration
cm
-
Centi meter
dm
-
Deci meter
g
-
Gram
L
-
Liter
m
-
Meter
M
-
Molar
mA
-
Mili ampere
meq
-
Mili equivalent
mg
-
Mili gram
min
-
Minute
mL
-
Mili Liter
mm
-
Mili meter
mM
-
Mili Molar
nm
-
Nano meter
q or qt
-
Metal uptake at any time, t
qe
-
Metal uptake at equilibrium
t
-
Tonne
y
-
year
Å
-
Angstrom
µg
-
Micro gram
µm
-
Micro meter
µmol
-
Micro mole
xxiii
LIST OF ABBREVIATIONS
AAS
-
Atomic absorption spectrophotometer
Av
-
Average
B.C.
-
Before Christ
BET
-
Brunauer, Emmett and Teller
CAC
-
Commercial activated carbon
DDW
-
Distilled deionized water
Dia
-
Diameter
DPC
-
1,5-Diphenylcarbazide
EDL
-
Electric double layer
EDX
-
Energy dispersive X-ray analysis
EFB
-
Empty fruit bunches
EPA
-
Environment Protection Agency
Et
-
Ethyl
FAAS
-
Flame atomic absorption spectroscopy
FTIR
-
Fourier transform infrared spectrophotometer
GAC
-
Granular activated carbon
h
-
hour
HFO
-
Hydrated iron oxide
HSAB
-
Hard and Soft Acids and Bases
ICP-MS
-
Inducedly coupled plasma mass spectrophotometer
IUPAC
-
International Union of Pure and Applied Chemistry
Max
-
Maximum
MOX
-
Malaysian Oxygen
NOM
-
Naturally occurring organic matter
PAC
-
Powered activated carbon
xxiv
QC
-
Quality control
rpm
-
Revolution per minute
RSD
-
Relative standard deviation
SCF
-
Surface complex formation
SD
-
Standard deviation
SEM
-
Scanning electron microscope
SOC
-
Synthetic organic chemicals
SOCs
-
Synthetic organic compounds
TGA
-
Thermogravimetric analysis
USA
-
United States of America
USEPA
-
United States Environment Protection Agency
UV
-
Ultra violet
VOCs
-
Volatile organic compounds
WHO
-
World Health Organization
xxv
LIST OF APPENDICES
APPENDIX
A
TITLE
National Drinking Water Quality
PAGE
166
Standards, 2000 of Malaysia for some
species of metals and metalloids
B
EDX Spectrums of Some Selected
167
Activated and Modified Carbons
C
Summary Report of N2 Adsorption-
170
Desorption Data and BJH Pore size
distribution Graph
D
FTIR Spectrums of Some of the Selected
174
Activated Carbons
E
Quantification of Acidic Groups on the
175
Surface of Some Selected Activated
Carbons as meq Base per Gram Carbon
F
TGA Thermograms of Some of the
176
Selected Activated Carbons
G
FAAS Determined Fe Content in Samples
186
of Leached Water Contacted with C-HFO-1
Composite Adsorbent at pH 4.5
H
FAAS Determined Adsorption Equilibrium
Data of Various Metal Cations in Single,
Binary and Ternary Solutes Solutions by
Various Activated Carbons at Varying
Initial Concentration and pH
187
xxvi
I
UV-Vis Spectrophotometer Determined
203
Adsorption Equilibrium Data of Cr(VI) in
Single Solute Solution By Various
Activated Carbons and C-HFO Composite
Adsorbents
J
ICP-MS Determined Adsorption
208
Equilibrium Data of As(III) and As(V) for
the Adsorption in a Single Point Initial
Concentration By a Typical Prepared
Activated Carbon and its modified ones,
namely, PSW-P-500 and C-HFO-1
respectively
K
Adsorption Kinetics Data of Cu(II), Ni(II),
209
Pb(II) and Cr(VI)
L
Presented Papers and Expected
Publications from This Study
212
CHAPTER 1
INTRODUCTION
1.1
Research Background
Ancient people were concerned about the quality of water. This awareness
made the history of water treatment as old as humans. Historically, the scarcity of
aesthetic quality drinking water was mitigated in most places either by digging wells
or by filtering the surface water through sands. Once it was realized that clear and
tasty water due to bad contaminations can cause diseases outbreak. The industrial
revolution of nineteenth century added some other synthetic chemicals and heavy
metals contaminants in the aquatic system.
The presence of a wide range of
synthetic organic chemicals (SOC) was confirmed by the Environmental Protection
Agency (EPA) of USA in finished drinking water, in many locations, even those are
from ground water supply (Cotruvo, Hanson and Thornton, 1983). This survey
breaks the historical concept of viewing ground water as a relatively uncontaminated
resource, unspoiled by the human activities that affect surface waters. The presence
of even trace quantities of SOC in finished drinking water should be encountered as
a major future threat to the supply water for the existing mechanism of
contamination of the source by man-made pollution. It was soon realized that other
than human activities, nature itself can contaminate natural waters. The natural
contamination of highly toxic arsenic species in groundwater is spreading arsenicpoisoning in some part of the world and those are not affected, still in risk. As the
contamination was manifested, without going through a detailed analysis and
2
monitoring around the year, it is very hard to say whether the quality of drinking
water is safe for health.
This is why regulations are imposed on permissible levels of contaminants in
drinking water from the potential health risks exposure to these. To minimize
human contributing pollution to a lower level, laws are also enforced to treat
wastewater before discharge.
Among various pollutants in drinking and
wastewaters, inorganic species of metals and metalloids are major, and difficult to
remove because of their smaller size, very low concentration and competition with
non-toxic inorganic species. The present concern over the presence of inorganic
species necessitates a cost effective, eco-friendly treatment process to remediate
these contaminants from drinking water and wastewater.
Therefore, it is not
surprising that current research trend is focusing on the utilization of local raw
materials such as agricultural wastes, as low cost eco-friendly materials, in treating
drinking and wastewaters preferentially to remove inorganic toxic species.
As
Malaysia is moving rapidly towards the concept of healthy drinking water, her huge
wastes from agricultural sector, such as oil palm kernel and coconut shells, are
needed to be selected as a promising precursor in the removal of ionic- and non-ionic
inorganic toxic species.
1.1.1
Water Purification
Long before the development of civilization, people noticed differences in
water quality. Waters of various sources have characteristic taste whether sweet,
salty, bitter, or, sour. Hippocrates suggested his colleagues to filter and boil water
before drinking it. The ancient people of Egypt, Greece, Assyria, Persia and India
used wells for drinking water (Ellms, 1928). Probably, wells furnished a safer
drinking water than other water sources such as rivers and lakes. The treatment of
waters by solid adsorbents such as sand is probably as ancient as civilization itself.
The ancient people also learnt the technique of utilization of sandstone and unglazed
earthen ware as domestic filtering media for purifying drinking waters. In some of
3
the ancient water purification systems, muddy waters were settled in tanks or settling
basins.
These early methods of water treatment, no doubt, were crude and imperfect,
but these can be regarded as the beginnings of water purification on which modern
art of water treatment technology is built. The necessity of treatment of public
supply water was overlooked until the end of the nineteenth century, when cholera
epidemics in England showed that infected water supplies were responsible for
public disease. Over 53, 000 people died during this cholera epidemic in London in
between 1848-1849 (Stevens, 1974). It was soon realized that drinking water should
be free from disease causing organisms or toxic substances.
The impurities in water can be classified into three major groups such as i)
suspended and colloidal matters, ii) dissolved substances and iii) biological
contaminants.
Suspended matters and colloidal particulates are responsible for
making the water turbid. Coagulation, sedimentation and filtration are the processes
usually employed to decrease the turbidity of water. Dissolved substances and some
time odour and colour are resulted from the presence of inorganic and organic
constituents. Dissolved impurities especially inorganic ones are difficult to remove.
In the early days, lime and soda were used in the reduction of dissolved hard ions
those causes the hardness of water. Activated carbon was used in controlling taste
and odour. Oxidizing agents followed by filters were used to reduce iron and
manganese. Recent technology, in addition to chemical precipitation and adsorption,
uses ion exchange sorption, membrane separation technique, desalination,
electrodialysis etc. for the purification of dissolved contaminants in water.
The presence of minerals and organic constituents as well as the presence of
sunlight and temperature make a luxuriant growth of algae and micro-organisms in
almost all natural waters. A few causes disease, some impact taste, odour, colour or
turbidity to the water, others are beneficial, and the rest are of little interest (Lehr et
al., 1980). Depending on the water source, disinfection may be a simple or complex
matter. Chlorine, a widely accepted oxidizer to the membrane of micro-organisms,
is used extensively in the disinfection of supply water. Ground water from deep may
not need disinfection.
4
Although the ion exchange phenomenon in clay and soil was reported in the
beginning of the nineteenth century, its use for large scale water softening had begun
only in the beginning of the twentieth century.
At the beginning, natural and
synthetic aluminium silicates were used for softening waters. The limitations of
siliceous ion exchangers became more and more evident as commercial exploitation
of these substances was attempted. Substitution was made by sulphonated coal as
cation exchangers. Attachment of sulphonic acid groups on the surface of coal,
lignite and peat by sulphonation reaction was a distinct contribution to the
development of ion exchange resins. In 1935, Adams and Holmes (cited by Kunin,
1958) observed the ion exchange properties in the synthetic resins of phenolformaldehyde. Further studies resulted in the synthesis of cation exchange resins
containing strongly acidic sulphonic acid groups while condensing phenols and
formaldehyde in the presence of sodium sulphite. Anion exchange resins were
prepared in an analogous manner by the condensation polymerization of amines with
formaldehyde.
Since 1940, the synthesis of ion exchange resins has made
spectacular advances having high cation and anion exchange capacity over the entire
pH range and durable physical properties.
Today’s municipal water treatment plant still uses conventional processes
including screening, coagulation, sedimentation, and chlorination. Although modern
water conditioning technology can solve almost all water quality problems, still the
technology is not reachable to the majority of the world population because of high
treatment cost. The modern water treatment plants are economically feasible in large
scale operation and such big plants are usually installed in highly populated urban
areas. As a consequence, majority of people residing in rural areas not only develop
tolerances for drinking water of poor taste, odour or appearance but also belief that
their water supply requires no treatment. On the other hand, tap water may be
contaminated either from the corrosion of the pipelines or from the sources whether
ground or surface. This is why, not only the rural and small community people,
sometime urban people using supply water are still in scarcity of quality drinking
water. In such cases, it is strictly advised to treat the water before drinking it.
Therefore, the present problem is to utilize the existing knowledge of water
conditioning for developing a small scale low cost treatment process so that major
5
individuals of the world population can benefit using this process in home water
conditioning unit. The successful solution to this challenge largely depends upon the
manufacturing of proper adsorbents followed by subsequent utilization in house hold
filtering unit. The later work consists of three steps of investigations: i) bench-scale
adsorption studies in a batch or column in laboratory using synthetic solution of
mg.L-1 level, ii) medium-scale adsorption studies in a pilot plant using actual
drinking water, iii) full-scale treatment evaluation in a modular unit.
1.2
Heavy Metals
The classification of some elements as heavy metals does not have a rigorous
scientific basis or a chemical definition. But they have widespread usage among
professionals and laymen for over the past two decades. They constitute a group
name for the transition and post-transition metals along with metalloids, arsenic and
selenium, that has been associated with contamination and potential toxicity or ecotoxicity (SenGupta, 2002; IUPAC, 2002). No relation can be drawn with specific
gravity and toxicity. Even then, they are so called only because of their significantly
heavier specific gravity, usually greater than five though exceptions remain, in
comparison to sodium, calcium and other light alkali and alkali-earth metals.
1.2.1
Pollution and Remediation Approaches
Today, the cry of “Pollution” is heard from all nooks and corners of the globe
and it has become a major threat to the very existence of mankind on the planet
earth. Heavy metal pollution affects flora, fauna and other abiotic components of the
ecosystem. Metal leads to various metabolic alterations and undesirable changes,
which in many cases may cause severe injury and health hazards. There are several
factors like human population explosion, unplanned urbanization, deforestation,
profit oriented society and technological advancement etc., whose cumulative effect
is responsible for the origin of pollution crisis on the earth. The ecological crisis of
6
environmental pollution has been blamed on many different things and one of the
major things is the pollution due to metals or their species in the environment.
The presence of heavy metals, strictly speaking ‘the species of heavy metals’,
in the environment is of increasing major, worldwide concern for the last few
decades.
The toxicity and threat of heavy metal species to human life and
environment came into the limelight in 1970s especially after the disclosure of
mercury poisoning by the Minamata disaster in Japan (Kadirvelu et al., 2004). Since
then, World Health Organization (WHO) set up various permissible limits on the
discharge of heavy metals into the environment. According to WHO (cited by
Rivera-Utrilla and Sanchez-Polo, 2003), aluminium, chromium, magnesium, iron,
cobalt, nickel, copper, zinc, cadmium, mercury and lead are the most toxic metals.
Toxic metal pollution can be a much more serious and insidious problem, as these
are intrinsic components of the environment. At high concentrations, all the metal
species are toxic to both animals and plants. Heavy metals are continually released
into the aquatic environment from natural processes, such as volcanic activity and
weathering of rocks, as hidden pollutant. Arsenic, another hidden common trace
inorganic contaminant in drinking water, is extremely detrimental to human beings.
Leaching from geological formations is a major natural source of arsenic in the
ground water.
Thus the presence of toxic elements or ionic/nonionic species
particularly in ground water makes the so-called safer water unsafe at least in some
part of the world and those are not affected, still in risk. Metals are also released into
environment from a wide spectrum of anthropogenic sources such as smelting of
metallic ores, industrial fabrication and commercial application of metals, agrochemicals, pesticides as well as burning of fossil fuels.
These metals are
redistributed in the biosphere and dispersed in the air, soil, water and consequently in
human beings through food chain bio-magnification causing chronic ailments.
Our bodies have become virtual dumping grounds for hundreds and hundreds
of these toxic metals and chemicals. Studies indicate that we have between 400 and
800 toxic metal and chemical residues stored in our fat cells, and our tissues and
organs. The Environmental Protection Agency of USA has estimated that we cannot
metabolize most of these chemicals and metals. That means they accumulate in our
fat cells, in our organs such as the liver, kidneys and brain, in our glands such as the
7
thyroid and adrenals, and in the central nervous system. These chemicals and metals
together result in a chronic toxic overload condition in our bodies. As these toxins
build up, they block the receptor sites for essential minerals, so that minerals such as
iron, calcium and magnesium cannot be utilized and absorbed. This causes enzyme
dysfunction, nutritional deficiencies, hormonal imbalances, neurological disorders,
damages brain chemistry, and can even lead to auto-immune disorders, cancer, and
other debilitating chronic conditions. As early as 1974 in Florence, Italy, at the
Congress of the World Health Organization, attention was drawn to the fact that 6080% of the environmental toxins, which includes heavy metals, contribute to many
diseases like neuro-dermitis, autoimmune disorders, allergies etc.
For short- and long-term toxic effects, the maximum permissible
concentrations of heavy metals in drinking water as well as in municipal and
industrial discharges are closely regulated in most of the countries through
legislation. Drinking water quality standards for some species of metals/metalloids,
according to the National Drinking Water Quality Standards, 2000 of Malaysia, are
presented in Appendix A. Ultimately, as heavy metal regulations for both drinking
water and wastewater are becoming stricter, the conventional means of water
treatment becomes either costly or inefficient.
Precipitation, sedimentation, coagulation, oxidation, filtration, sorption and
membrane processes are generally employed to reduce heavy metal ions from the
contaminated water. No doubt, the toxicity effects of these elements are largely a
function of concentration. The widely used treatment method for the removal of
heavy metal ions from wastewater is chemical precipitation using hydroxide,
carbonate or sulphide or some combination of these chemicals. The most common
processes use hydroxide precipitation, because the theoretical minimum solubility of
metal hydroxides is low and the treatment cost is lower when lime is used.
Iron coprecipitation/adsorption is another well accepted process for the
removal of heavy metal ions from water and wastewater. In this process, ferric salt,
such as FeCl3 or Fe2(SO4)3, is added to water to form an amorphous precipitate of
iron
oxyhydroxide,
(Fe2O3.H2O).
The
dissolved
and
suspended
trace
elements/species are adsorbed onto the floc of iron oxyhydroxide and trapped within.
8
Flocs grow until they get heavy enough to settle down. The settled precipitate is
then separated out from water, leaving a purified effluent. The process depends on
reaction pH, adsorbate and adsorbent concentrations as well as the presence of
competing cations and anions. Depending on pH, this process is also capable of
removing oxyanions of arsenic, chromium and selenium from water and wastewater.
The proposed mechanism of coprecipitation/adsorption of heavy metal ions on
hydrous iron oxide surface can be presented by the following reactions (Benjamin,
Hayes and Leckie, 1982), where oxide surface of iron, depicted by
SO , acts
as weak acid or base while gain or lose a proton.
Protonation of surface oxide:
SO
+ H
SOH + H
=
SOH
(1.1)
=
SOH2+
(1.2)
M(OH)mn-m + mH+
(1.3a)
Cation adsorption on protonated surface oxide:
Mn+ + mH2O =
SOH + M(OH)mn-m =
SO.M(OH)mn-m-1 + H+
(1.3b)
(The system pH decreases as the reactions go forward)
Anion adsorption on hydro-oxylated surface:
SOH + mH+ =
m+
SOHm+1 + An- =
m+
SOHm+1
SOHm+1Am-n
(1.4a)
(1.4b)
(The system pH increases as the above reactions go forward)
Here, M and A represent cationic and anionic adsorbate, respectively. The
above equations, (1.3a) and (1.3b), reveal that protons are released into the system
when an uncomplexed hydrated metal cation is adsorbed and equations (1.4a) and
(1.4b) depict that protons are removed from the system during the adsorption of
anions.
These equations thus describe that increasing the concentration of
deprotonated, SOH/SO− and protonated, SOH2+ surface oxide sites, increases the
adsorption capacity of cations and anions respectively. As the pH increases the
adsorption of cation increases, while anion adsorption increases with the decrease of
9
pH. But the limitation of the process of precipitation and coagulation/flocculation is
that these are effective in removing some metal ions in drinking water to mg.L-1
levels but not always to µg.L-1 levels. This process also generates huge sludge. All
of above, the removal of dissolved trace toxic substances or heavy metal ions is
targeted at the point of use. Because it is considered a monetary burden, especially
in the developing countries, to treat water at an early stage, when only 1% of water
supplied to households is consumed by humans.
Therefore, it is rational that
adsorption processes are substituting precipitation methods.
Ion exchange sorption and membrane processes such as reverse osmosis are
capable of reducing metal concentrations to very low levels. But these are expensive
and subject to operation problems such as fouling (Reed, Vaughan and Jiang, 2000).
Thus, to mitigate the present problem of lowering the heavy metal contamination to
the permissible level, an introduction of cost-effective, eco-friendly highly efficient
sorption technology is needed to impose as down stream treatment process
preferably to be used at the point of use or in house hold filtering unit. The term
high efficiency is used to mean that the process would be selective towards the target
species and capable of treating them to a safe level. To meet the cost-effective
criteria, the process should be based on locally available low cost material and unlike
ion exchange synthetic polymeric resin, should be eco-friendly. Thus the majority of
people especially rural people might benefit from the innovative adoption of existing
knowledge of advanced water treatment.
The necessity of selectivity is highly desired in removing heavy metal cations
from water and wastewater streams while using sorption/adsorption based process.
Because, selective removal of minor (0.01-10 mg.L-1) to trace (<0.01 mg.L-1)
concentration of the heavy metals in the presence of other competing non-toxic, nonregulated ions such as Ca2+, Mg2+, Na+ etc certainly makes the process efficient and
cost effective. Otherwise, treatment cost would be very high due to premature
exhaustion of the sorbent by the above mentioned non-regulated ions.
As for
example, in a conventional ion exchanger, Ca2+ ion is preferred over heavy metal
cations such as Zn2+, Co2+, Cu2+ and Ni2+. The heavy metal cations of mercury and
lead are preferred marginally over Ca2+ (Sengupta and SenGupta, 2002). The lower
10
selectivity towards target heavy metal ions makes the commonly used cation
exchangers uneconomical and unfavourable.
Any elemental species with a net positive charge behaves as a Lewis acid
(Shriver, Atkins and Langford, 1991). Most of the heavy metal cations such as Cu2+,
Hg2+, Pb2+, Ni2+, Cd2+ and Zn2+ being transition-metal ions exhibit fairly strong
Lewis-acid characteristics due to the presence of incomplete d-orbital.
Their
electron clouds are readily deformable by the induction effect of electric field of
other species. They favour to form fairly strong complexes with Lewis bases,
organic and inorganic ligands, in the aqueous phase through inner sphere complexes
by coordinated with donor electrons. Heavy metal cations thus can serve as good
electron accepters. In other words, they are relatively ‘soft’ cation.
In contrast, cations of light metals, such as alkali and alkali-earth metals,
having the inert gas configuration are not good electron acceptors and hence poor
Lewis acids. Their electron clouds are not readily deformable by the electric field of
other species. Thus they are less polar and commonly referred to as ‘hard’ cations.
They form only outer sphere complexes with aqueous-phase ligands containing
primarily oxygen donor atoms through columbic force of attraction. Hence, the
simple ion exchange sorption-desorption processes are applicable to remove,
preferentially, these hard ions. Pearson (1968) first showed that most cations are
Lewis acids and anions are Lewis bases, while he expressed their strengths and
affinity to acid-base-complexation in his famous principle of ‘hard and soft acids and
bases’ (HSAB) using the property of ions as ‘soft’ and ‘hard’.
The interaction of metallic species with living systems and adsorbents can be
demonstrated better by the properties of metal ions as Lewis acids. Thus, the
classification of metals by their Lewis acidity would be consistent in the assessment
of toxicity and in the remediation approach. This classification divided metals into
three groups based on their observed affinity for different ligands, as Class (a)-‘hard
metals’, Class (b)-‘soft metals’ or borderline- ‘difficult to distinguish as hard or soft’
(Ahrland, Chatt and Davies 1958). It should be noted that this classification, being
based on observed chemical behaviour, is empirical. This is why, the same metal ion
is found into different classes in literatures. As hard and soft metals are also denoted
11
as Class A and Class B, the term Class AB is conveniently introduced for the
borderline metals throughout this thesis.
Figure 1.1 shows the periodical table
classifying metals as Class A- hard, Class B- soft and Class AB- borderline metals
(IUPAC, 2002).
Class
A
Class
AB
Class
B
Hard
Border
line
Soft
1
H
2
13
14
15
16
17
Li
Be
B
C
N
O
F
Na
Mg
3
4
5
6
7
12
Al
Si
P
S
Cl
K
Ca
Sc
Ti
V
Cr
Mn
Zn
Ga
Ge
As
Se
Br
Rb
Sr
Y
Zr
Nb
Mo
Te
Sn
Sb
Te
I
Cs
Ba
†
Hf
Ta
W
Bi
Po
At
Fr
Ra
#
Rf
Db
Sg
8
9
10
Co
Ni
Ru
Rh
Pd
Ag
Cd
In
Re
Os
Ir
Pt
Au
Hg
Ti
Bh
Hs
Mt
110
Fe(III)
Fe(II)
11
Cu(II)
Cu(I)
Pb(IV)
Pb(II)
†Lanthanide
La
Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
# Actinide
Ac
Th
Pa
U
Np
Pu
Am
Cm
Bk
Cf
Es
Fm
Md
No
Lr
Figure 1.1:
The periodic table showing classification of metals based on Lewis
acidity as: Class A-hard metals (black), Class B-soft metals (light grey), Class ABborderline (dark grey) (IUPAC, 2002).
Although the term ‘heavy metals’ is used by the legal regulations, it has no
sound terminological or scientific basis in the interpretation of toxicity and has never
been defined by authoritative body like IUPAC- International Union of Pure and
Applied Chemistry. Since no relationship can be drawn between density and toxicity
or eco-toxicity, a classification based on their chemical properties is put forward in
this work to abandon the classification of some metal ions as heavy metals and to
12
follow the guideline of IUPAC. A theoretical basis of classification of metals as
Class A and Class B has been introduced by Klopman (cited in IUPAC, 2002)
depending on the calculated orbital electronegativity of cations or anions. Metals
with calculated orbital electronegativities above 1.45 all belong to Class A, while
those have the orbital electronegativity below -1.88 are all Class B.
Hard cations prefer hard (oxygen) donor ligands and form mainly ionically
bound complexes. As the bonding of hard cation and hard ligand is mostly ionic in
nature, metal ions are displaceable and mobile. Soft cations generally exhibit higher
affinity toward nitrogenous and sulphurous species having donor lone-pair electrons.
Thus soft cations are able to form strong inner-sphere complexes with these donor
atoms. The higher atomic number p-block metals show strong affinity for soft
ligands such as sulphide or sulphur donor and form highly covalent complexes from
which they are difficult to displace.
environment.
Thus they are relatively immobile in the
In living organisms, they are not readily excreted and tend to
accumulate with resultant toxicity in protein molecule containing thiol group (-SH).
The borderline metals generally form relatively stable complexes with both hard and
soft donor ligands, but the exact order of stability is not easily determined. First row
d-block transition metal ions fall mainly into this group and show widely variable
coordination chemistry.
The basis of this principle has widely been used in the synthesis of chelating
ion exchangers with various types of covalently attached functional groups for
separating hard and soft cations. It is thus possible to utilize the Lewis acid-base
interaction for the high selective uptake of heavy metal cations by the sorption
process on chelating cation exchangers. As the composition of the functional groups
of simple cation exchange resins changes from hard oxygen donor atoms,
carboxylate groups, to relatively soft nitrogen donor atoms, such as bispicolylamine:
-CH2-N.(CH2-C5H9N)2, the affinity of soft cations is greatly enhanced over the hard
cations. Chelating exchangers with sulphur containing thiol functional groups offer
significantly higher selectivity for soft Hg(II) over Cu(II) and Zn(II).
The
composition of the functional groups in chelating exchangers can be judiciously
tailored to improve specific affinities toward target metal ions.
13
The synthesis of polymeric chelating cation exchange resins involves a major
step of covalent attachment of organic functional groups onto the parent polymer
beads which makes the preparation difficult and costly. Moreover, sorption of metal
ions onto chelating ion exchangers is kinetically slow due to the slower intraparticle
diffusion of metal ions through the rigid structure and tortuous pathways within
exchangers (SenGupta, 2002). Hence the present problem demands the replacement
of the chelating/polymeric cation exchangers with renewable naturally occurring
low-cost materials those contain carboxylate, carbonyl, phenolic groups and those
exhibit moderate to high affinity toward heavy metal ions.
1.3
Activated Carbon
Activated carbon is a processed carbon material. Due to highly developed
internal surface area and porosity, porous carbon materials or activated carbons have
been used for thousands of years and have now become extremely versatile
adsorbents. The adsorptive properties of carbon were well known long before the
terms active and activated had been developed or doped.
The major applications of activated carbon are the removal of species by
adsorption from liquid or gas phase, to effect the purification or recovery of
chemicals (Girgis et al., 2002), storage of natural gas of vehicles, catalysis supporter
and electric double-layer capacitors (Yang et al., 2004). Although activated carbon
is an inexpensive and widely used adsorbent, its commercial application in water
treatment, till now, is limited to remove taste and odour. Granular activated carbon
(GAC) or powered activated carbon (PAC) is used in water treatment especially in
the developed countries. It is also used in high-tech application for the removal of
synthetic organic compounds (SOCs), volatile organic compounds (VOCs), and
naturally occurring organic matter (NOM).
It is expected that through controlled oxidation/activation, concentration of
carboxylate and phenolic groups in activated carbon particles can greatly be
enhanced, resulting in a higher metal ion removal capacity (SenGupta, 2002). It has
14
been shown by many researchers (Dastgheib and Rockstraw, 2001; Toles et al.,
1999; Johns et al., 1998; Toles et al., 1998; Toles et al., 1997) that lignocellulosic
agricultural wastes, particularly nutshells, are very good precursors for the
production of granular activated carbons having suitable surface active sites for the
removal of heavy metal ions from aqueous solutions. Since the ions of heavy
metals, Class B and Class AB, have the stronger affinity toward donor sulphur and
nitrogen atoms, simultaneous/stepwise activation and functionalization of the
precursor or the surface of activated carbons by these donor atoms can improve
largely the removal capacity of these metal ions (Gomez-Serrano et al., 1998;
Krishnan and Anirudhan, 2002; Krishnan and Anirudhan, 2003; Yantasee et al.,
2004). The utilization of such modified activated carbons would be very promising
for the removal of particular, especially Class B, metal/metalloid species.
1.3.1
Historical Aspect of Carbon
The application of powdered charcoal goes back to prehistoric times. The
use of charcoal in medicine was reported in an Egyptian papyrus as early as 1550
B.C. (Hassler, 1974). A Swedish chemist Karl Wilhelm Scheele, an apothecary by
profession, was the first to discover the adsorptive properties of charcoal in 1773
when an experiment, consisted of the adsorption of gases, was conducted. The
power of charcoal in removing colour from solution was first observed by the
Russian academician Lovits in 1785. A few years later in 1794, wood char was
employed to purify cane sugar in England. During the nineteenth century, much
research had done to develop decolourizing carbons from various source materials
such as bone, blood, cocoanut, flour, tar etc. with a special preference to be used in
sugar industry. The chemicals generally used in these preparations were potash,
magnesium carbonate, phosphate etc. In 1862, Lipscombe manufactured a carbon for
the treatment of drinking water.
In 1900 and 1901, two activation processes were developed and patented by
Ostrejko to manufacture activated carbon from vegetable material. These patents
make the basis for the development of modern activated commercial carbon. In one
15
process metallic chlorides were used as activating agent whereas in another process
carbon dioxide was used as selective oxidizing agent.
Elevated carbonization
temperature was applied for activation with carbon dioxide.
In 1909, a plant was built for the first time to manufacture powdered
activated carbon on industrial scale following Ostrejko’s patents. In 1911, the same
plant produced new kinds of activated carbon from peat by activation with steam. In
1915 during World War I, preparation of granular activated carbon was developed
by Germany as filtering material in gas masks to protect the soldiers against chlorine
gas used by the German soldiers. Coconut shell was emerged then as promising raw
material for the production of granular activated carbon.
During the past decades, activated carbons have been produced from a
variety of raw materials by employing different processes. Manufactured carbons
have been found numerous applications in the industry. Although many of ancient
preparations make the basis of today’s industrial process, they were not then
developed on a commercial scale. The reason is assumed in the difficulties in
process engineering as well as in the lack of knowledge in diversifying its
application.
1.3.2
Structure of Activated Carbon
Activated carbons are highly porous materials with surface areas of up to
3000 m2.g-1. The surface area of a typical activated carbon is about 1000 m2.g-1.
These high surface areas are the result of development of mainly micro- and
mesopores while a little contribution is also come from macropores. According to
the IUPAC classification, pores having the radii less than 2 nm are called micropores
whereas pores having the size of radii in between 2 to 50 nm are identified as
mesopores. Macropores are defined as pores having the radii size above 50 nm.
Cokes, chars and activated carbons are often termed as amorphous carbon.
X-ray diffraction and electron microscopy have shown that these materials have
16
crystalline characteristics, even though they may not show certain features, such as
crystal angles and faces, usually associated with the crystalline state. The existence
of submicroscopic dimensions in so called amorphous carbons makes the microcrystal structure known as crystallites.
It was evident that amorphous carbon
consists of graphitic flat plates in which the carbon atoms are arranged in a
hexagonal lattice, each atom, except those at the edge, is held by covalent linkages to
three other neighbouring carbon atoms. The crystallites formation is the result of the
stacking of two or more of these plate-structures with an interlayer distance of 3.6 Å
(Hassler, 1974).
The diamagnetic anisotropy is the highest in activated carbon among known
materials. The structure of delocalized π-electrons is similar to that of aromatic
hydrocarbons. Each basal layer can be considered a unique macromolecule. The
various basal layers in the lattice, having delocalized electron structures, are
markedly separated from one another. The relatively large distance between layers
and the weak forces that hold the layers together make it possible for atoms, ions or
molecules to enter the interlayer spaces. The adsorbates make room for themselves
if necessary by forcing the sheets farther apart (Cini, Pantani and Sorace, 1980).
1.3.2.1 Surface Functional Groups
The adsorption of acids and bases on carbon was found to be controlled by
the concentration of surface-oxygen. This finding led to study the effect of oxygen
in contact with carbon at different temperatures and pressures. It was observed that
graphite, carbon black, and activated carbon can adsorb molecular oxygen. Elevated
temperature is needed to remove once chemisorbed oxygen.
Wheeler (cited by Faust and Aly, 1987) reported that when oxygen contacts
carbon surfaces some sort of oxygen-carbon complex is formed. It is not known
what surface functional groups are formed during the carbon activation process. In
general, acidic oxides are predominantly formed in carbon when it is prepared under
moist air at 300 to 500°C, and basic oxides in those carbons prepared at 800 to
17
900°C in air, steam, or carbon dioxide. Amphoteric properties are found in carbons
prepared between 500 to 800°C. The activated carbons prepared at low activation
temperature, below 500-600°C, adsorb OH- ions primarily are called L-carbons.
Those activated carbons are prepared at above 500-600°C and adsorb H+ ions are
called H-carbons.
Almost every type of functional group in organic chemistry has been
suggested to be present on activated carbons surface. The ones suggested most often
acidic functional groups are: carboxyl, phenolic-hydroxyl, and quinone-type
carbonyl groups (Figure 1.2) while other suggested groups are ether, peroxide and
ester groups in the forms of normal lactones and fluorescein-like lactones, carboxylic
acid anhydrides and the cyclic peroxide (Figure 1.3). On the other hand, proposed
basic surface groups include chromene and pyrone.
Typical identification reactions of organic chemistry were employed by
Boehm (2002) to characterize oxygen chemisorbed on carbon as comprising four
types of acidic surface groups:
1) a strongly acidic carboxyl group which is neutralized by NaHCO3, Na2CO3,
NaOH, and NaOC2H5,
2) a more weakly acidic carbonyl group which is neutralized by Na2CO3 or
stronger bases such as NaOH and NaOC2H5; but not by NaHCO3,
3) a phenolic hydroxyl group which is neutralized by NaOH and NaOC2H5,
4) a carbonyl group which is neutralized by NaOC2H5 only.
The type and number of surface groups on activated carbon will influence the
extent and rate at which organic and inorganic compounds/species are adsorbed. It
had been observed that acidic surface oxides reduce the adsorption capacity of
phenol and nitrobenzene on activated carbon. There are a couple of explanations for
the role of acidic surface oxygen groups in their inhibition of adsorption of phenol
and nitrobenzene molecules. A major portion of these oxygen bearing functional
groups is located on the edges of the layer planes of carbon and they are not
expected to be interfered by the adsorbed organic molecules on the basal planes from
the view of steric sense.
18
O
O
C
OH
Phenolic hydroxyl groups
Carboxyl groups
Figure 1.2
O
H
Quinones
Structures of carboxyl, phenolic hydroxyl and quinone type carbonyl
groups.
OH
O
O
O
C
Normal lactones
O
Fluorescein type lactones
O
C
C
O
O
C
O
O
C
Cyclic peroxide
Carboxylic acid anhydrides
Figure 1.3
Structures of lactones, acid anhydrides and cyclic peroxide.
Chemically bound oxygen on the edges localizes electrons in surface states
and withdraws them from the π electron system of the basal planes. Depletion of
electrons by oxygen creates an increase in the population of positive holes in the
conduction band of the π electron system and thus reduces the non-polar nature with
an ultimate reduction in adsorption capacity of aromatic compounds.
This is
consistent with the concept of dispersion forces.
The motion of electrons in an atom or molecule would lead to a rapidly
oscillating dipole moment while at any instant experiences the lack of symmetry of
the electron distribution about the nuclei. When this atom or molecule approaches in
close proximity to a non-polar solid surface, each instantaneous dipole induces an
19
appropriately oriented (so that attraction could occur) dipole moment in a molecule
on the surface. The interaction of these instantaneous dipoles with their counter
induced ones produces an instantaneous attraction between adsorbate and adsorbent.
The dispersion forces between the phenol’s π electron system and the π band of the
graphitic planes of the carbon are responsible for adsorption. Removal of electrons
from the π band of the carbon by chemisorbed oxygen may interfere with and
weaken the adsorptive forces resulted from the permanent polarity in the surface.
Since the polarity of carbon arises with the introduction of surface oxygen,
the adsorption of metal ions on carbon surface seems to be favoured accordingly.
The acid-base and metal removal behaviour of several commercial activated carbons
have been modelled successfully by the Surface Complex Formation (SCF) model
(Corapcioglu and Huang, 1987; Reed and Matsumoto, 1991). This model basically
uses the law of mass action to describe ion interactions with hydrous solid surface.
According to this model (SCF), the surface of the hydrous solid acquires a
surface charge due to various surface groups or sites and thus develops an electric
double layer (EDL) around the charged particle. It has been assumed that all surface
sites are capable of binding and releasing protons equally for which the solid can be
modeled as a single, weak diprotic acid.
Using this assumption, the surface
functional groups can be described by the following surface reactions:
SOH2+
SOH +
SOH
SO-
where, the symbol (
+
Hs+
Hs+
(1.5)
(1.6)
SO-) represents the active surface site attached on
interface.
The amphoteric behaviour of activated carbon surface as shown by the above
equations (equations 1.5 and 1.6) was interpreted by Snoeyink and Weber (cited by
Reed, 2002) assuming phenol and lactone functional groups may be amphoteric
while Mattson and Mark (1971) suggested carboxyl and quinone groups. Later on,
the carbon surface, regardless of the specific functional groups, was assumed as a
20
number of weak monoprotic acids rather than a single diprotic acid by many
researchers. There are two types of monoprotic acid sites to consider:
A positively charged site,
Pi OH2+
Pi OH0 + Hs+
(1.7)
N i O- +
(1.8)
And a negatively charged site,
Ni OH0
Hs +
where, ‘i’ is an index to differentiate between sites and Hs+ is the activity of proton at
the solid surface.
This model enables to measure the acidity constants for the above two types
of equations (equations 1.7 and 1.8). It is assumed that the sites that release and bind
protons are also responsible for removing metal cations. We have seen that the
surface acidity of activated carbons play an important role in adsorbing positively
charged species. In addition, many researchers established the ability of activated
carbons in removing heavy metal ions. Therefore, it is hoped that the use of granular
activated carbon in removing heavy metal ions (class AB and class B) will become
more common in near future.
1.3.3 Preparation
Almost any carbonaceous material can be converted into activated carbon
(Hsisheng and Sheng-Chi, 2000; Girgis et al., 2002), whether it is occurring
naturally such as coal (e.g. peat, lignite and all forms of coal) and lignocellulosic
materials (e.g. wood, coconut shell, fruit stones, nut kernels etc) or prepared
synthetically such as organic synthetic polymers (e.g. synthetic resins, rayon,
polyacrylonitrile, polyvinylidene chloride, waste resins, textile wastes, rubber,
dumped gangue etc). Bansal, Donnet and Stoeckil (1988) summarized the principal
commercial carbon feedstocks as: coal, lignite and peat 52%; wood 35%; coconut
shell 10% and others 3%.
21
The development of pores and their size-distributions mainly depends upon
the precursor type and the process of preparation. Acid treatment can be performed
after activation to control the ash content in the resulted carbon. The process,
activation of carbonaceous material, comprised of either physical or chemical
activation method. Based on the required properties of the activated carbon, the
proper raw material and most importantly the activation process are utilized. Low
temperature activation processes are adopted in presence of air or oxidizing materials
to prepare carbon having acidic characteristic to remove preferably metal cations.
Physical activation consists of controlled or partial gasification of
carbonaceous raw material. This uses gaseous flow, at high temperature, of mild
oxidizing agents such as steam and CO2 or a combination of these.
Partial
gasification eliminates unorganized carbon and non-carbon volatile matters as
respective oxides.
The process usually carried out in two steps: thermal
carbonization of carbonaceous precursor in presence of an inert gas (such as N2)
atmosphere followed by activation of the resulting char in a mild oxidizing gaseous
(CO2 and steam) atmosphere. The use of catalyst (e.g. K-, Na-salts) for the partial
gasification of carbon is a variation of physical activation.
Chemical activation process consists of impregnation followed by
carbonization. The impregnation process is applied in order to achieve a good
contact of a solid with a solution containing activating chemical, such as ZnCl2,
H3PO4, or KOH etc. to be deposited on the surface. Carbonization is carried out by
heating at a relatively low temperature under the atmosphere of nitrogen. Chemical
activation is generally applied to uncarbonized feedstock containing cellulosic
materials while activation is based on the dehydration of the feedstock. The more
extensive impregnation produces weaker cellular structure and a large development
of mesoporosity. A combination of both, physical and chemical, processes may also
be practiced.
Carbonization is the pyrolysis or thermal decomposition process to remove
most of the non-carbon elements (hydrogen, oxygen, traces of sulphur and nitrogen)
from a carbonaceous material in gaseous form in the absence of air and without the
addition of chemical agents. This enriches the carbon content and creates an initial
22
porosity in the resulting char. The free valancies of released carbons are satisfied to
group them into organized crystallographic formations known as elementary
graphitic crystallites. The mutual arrangement of these crystallites is irregular, so
that free interstices remain between them and, apparently, as the result of deposition
and decomposition of tarry substances, these become filled or blocked by
disorganized carbon (Wigmans, 1989). Therefore, the resulting carbonized product
needs activation to enhance its adsorption capacity.
Activation process further
develops the porosity and creates some ordering of the structure to generate a highly
porous solid as the final product. Chemical impregnation of lignocellulosic carbon
precursor results degradation of the cellulosic material while carbonization produces
dehydration. The former results in charring and aromatization of the carbon skeleton
whereas the latter creates the pore structure (Smisek and Cerny 1970).
The common feature of all substances used in the chemical activation process
is that they are dehydrating agents that influence pyrolytic decomposition and inhibit
formation of tar. As a result, higher yield is obtained with a high bulk density
suitable for use in column. It is believed that dehydration, due to impregnation,
proceeds by the partial dissolution of biopolymers together with the cleavage of
either bonds between the lignin and cellulose, followed by recombination reactions
in which layer structural units are formed, which ultimately derived a rigid
crosslinked solid (Molina-Sabio et al., 1995).
Then, in the next step, thermal
treatment of the impregnated feedstock modifies the morphology and dimensions of
the cellular units existing before impregnation. There is first an expansion of the
cellular structure at impregnation state, then a subsequent contraction during
carbonization at a temperature above 450°C (Caturla et al., 1991).
Coals are widely used as conventional feed stock for the manufacture of
activated carbons and the process usually followed is physical activation:
carbonization at about 600-650°C to prepare char, followed by partial gasification of
carbon, called activation, using steam or carbon dioxide at 800-950°C to develop
pore structure (Jagtoyen et al., 1992).
The use of chemical activation of wood, sawdust, with phosphoric acid has
been known for many decades (Jankowska et al., 1991) and has been applied to
23
different cellulosic and lignocellulosic materials to prepare high surface area
activated carbons. Phosphoric acid is used to provide oxidation environment. The
phosphoric acid activation process was independently developed in America and
Europe by Hudson and Urbain respectively.
Although a good number of works have been found regarding preparation of
activated carbon from oil palm kernel shell, the publications regarding preparation
variables and application in removing metal ions from aqueous solution are still rare.
So it would be our prime focus to utilize oil palm shell as low cost local raw
materials for the preparation and modification of activated carbon and subsequent
adsorption studies for the removal of metal ions. As huge amounts of coconut shells
are also produced as agricultural wastes, focus would also be given to utilize them as
precursor as well as to study their suitability in removing metal ions.
1.3.3.1 Processing of Lignocellulosic Materials
A summary of the processing of lignocellulosic precursors to obtain activated
carbons, reported in recent literatures, is presented in Table 1.1.
Table 1.1 : A survey of recent literatures on the production of activated carbons
from various lignocellulosic precursors.
Precursor
Carbonization and activation
Surface area,
process
2
m /g
Coconut
H3PO4 impregnation; carbonization
BET: 1360
shells
under air at 300-550°C.
Peach
ZnCl2 impregnation:
BET:
Caturla, et
stones
heat treated; carbonized
1000-2000,
al., 1991
in N2 at 500-800°C,
2700-3000
in CO2 at 825 or 875°C.
Reference
Laine et al.
1989
24
Table 1.1 : Continued.
Precursor
Reference
Carbonization and activation
Surface area,
process
m2/g
Coconut
Physical activation:
BET:
Laine and
shell
CO2, CO2 + K-phosphate catalyst;
696-
Yunes,
carbonization at 800°C.
1341
1992
Activated using steam, CO2,
BET: 1000
Asiah,
Oil palm shell
H3PO4.
Olive stones
Physical activation: CO2 and water
1993
-
Rodríguez
vapour; carbonization at 750-
-Reinoso,
825°C.
et al.,
1995
Peach stones
H3PO4 impregnation, sink 2 h at
BET: 338-
Molina-
85°C; carbonization under N2 at
1323
Sabio et
450°C, hold time 4 h.
Oil palm shell
Impregnated with varying amounts
al., 1995
BET: 1500
of ZnCl2 and carbonized at 500°C.
Hussein et
al., 1996
Rice straw,
Carbonized at 750°C in N2 for 1 h;
sugarcane
activated with CO2/steam at
Marshall
bagasse, soybean
850/800°C for 5-15 h followed by
and Toles,
hull, peanut,
oxidative treatment using air + N2
1998
pecan and walnut
at 300°C or chemical oxidation at
shells
23°C.
Almond, pecan,
H3PO4 impregnation, pre-heat
English walnut,
treatment at 170°C, carbonization
Marshall
black walnut and
at 450°C in N2 or air for 1-2 h
and Johns,
macadamia nut
followed by oxidation in air at
1998
shells
300°C those prepared in N2.
85-721
991-1693
Johns,
Toles,
25
Table 1.1 : Continued.
Precursor
Pecan shell
Carbonization and activation
Surface area,
process
m2/g
Char preparation followed by
618-1561
Reference
Johns,
activation using steam, CO2 or
Marshall
H3PO4. Physical activation was
and Toles,
carried out at 850°C in N2
1999
followed by oxidation at 370410°C. Chemical activation was
carried out by impregnating for
24 h in air and then pre-heat
treated and carbonized at 170
and 450°C respectively.
Almond shells
Pyrolysis at 700-800°C;
425-673 and
Toles et
activation with steam/CO2 at
197-560
al., 2000
27-1017
Dastgheib
800°C, hold time 2/4 h.
Pecan shell
Air and H3PO4, pre-heat
treatment 160-215°C,
and
carbonization temperature: 160,
Rockstraw,
300-500°C.
2001
Coconut shells and
Combining ZnCl2 and CO2
palm seeds
activation; carbonized at 800°C,
Srinivasan,
hold time 2-3 h.
2001
Oil palm stone
Oil palm stone
937-2450
Hu and
Impregnation with ZnCl2, H3PO4
ZnCl2: 1837
Guo and
or KOH for 24 h followed by
H3PO4: 1563
Lua, 2000
carbonization in N2 for 2 h at
KOH: 1408
temperature up to 600°C; after
Physical
cooling activated with CO2 at
activation:
700-900°C for 1 h.
892
Activation using CO2 at 650-
942-1410
950°C for 0.3-3 h.
Lua and
Guo, 2000
26
Table 1.1 : Continued.
Precursor
Carbonization and activation
Surface area, m2/g
Reference
process
Oil palm
Char obtained at 500-900°C;
(High temperature
Daud, Ali
shell
activated at 820°C for 20-250
char results higher
and
minutes under a mixture of
micro-porosity)
Sulaiman,
2000
steam and N2.
Corn cob
Chemically activated with
BET:
Tsai et al.,
potassium salts or physically
437 (un-impregnated)
2001
activated with CO2.
490-903 (KCl,
KOH and K2CO3 were found
K3PO4, KNO3, CO2)
effective catalysts for corn cob.
1266/1682
(K2CO3/KOH, CO2)
369/506
(KOH, K2CO3, N2)
Peanut
Powdered precursor was
BET: 97-253 (steam
Girgis et al.,
hulls
activated chemically by
pyrolysis)
2002
H3PO4, ZnCl2, KOH and
228-420 (ZnCl2 and
physically by steam.
KOH activation)
Carbonized at various
240-1177 (H3PO4
temperatures, 300-900°C. The
activation)
ratio of H3PO4 and precursor
was varied from 0.5-1.6.
Rice straw
Carbonized at 700°C and
BET: 2410
activated at 900°C.
Candlenut
CO2 activation at temperature
800°C while hold times were
varied from 0.5-6 h, and at
700-900°C with 1 h hold time
Oh and
Park, 2002.
990-1050
Turmuzi et
al. 2004
27
Table 1.1 : Continued.
Precursor
Carbonization and activation
Surface area, m2/g
Reference
1014-1062
Guo et al.,
process
Palm shells
Precursor was impregnated
with 5-40 % H2SO4 at 25°C for
2005
24 h, and then evaporated at
110°C. Carbonized in N2
atmosphere at 300-700°C. For
comparison, char was prepared
at 300-700°C for 2 h and then
activated using CO2 500900°C for 2 h.
Apricot
Soaked by impregnating agent
(72 h) and then dried at 70°C.
211-728
Youssef et
al., 2005
H3PO4 (25-75 %) activation at
400 and 600°C; ZnCl2
activation at 600 and 700°C
varying the ratio of ZnCl2 and
precursor. Also activated using
steam at 900°C.
1.3.3.2 Utilization of Oil Palm and Coconut Shells as Carbon Precursor
Although coals are being used as principal precursor, the renewable
agricultural solid wastes have potential prospects over conventional materials
because of high carbon but low ash contents (Guo and Lua, 2002). Malaysia is the
largest producer of palm oil product. Figure 1.4 shows the cross section of masocarp
of oil palm fruit bunch. This country produces 8.5 million tonnes of oil per annum.
Since the oil is extracted from fresh fruit bunches it generates large amounts of
wastes such as fibre (5.4 × 10 6 t y −1 ) from the masocarp, shell (2.3 × 10 6 t y −1 ) from
28
around the kernel and the empty fruit bunches (EFB) as residues (Husain, Zainac and
Abdullah, 2002).
Fibre
Shell and kernel
Pericarp
Figure 1.4
A cross section of masocarp of oil palm fruit bunch showing fibre,
shell, kernel and pericarp.
Figure 1.5 presents process flow diagram of palm oil extraction to show the
generation of solid wastages such as shells and fibres. Palm oil mills use the fibre
and shell, in a ratio of 60:40, as fuel to generate steam and electricity (Husain,
Zainac and Abdullah, 2002). The proximate analysis of solid oil palm residues are
shown in Table 1.2 and the ultimate analysis of those are given in Table 1.3.
Lignocellulosic contents in oil palm shell are as follows: cellulose 29.7,
hemicelluloses 47.7 and lignin 53.54 % respectively (Daud and Ali, 2004). The
fibres of oil palm are short and thin. While the majority of fibres are used as fuel,
the huge amount of palm shells are creating environmental problem. As oil palm
shells can be obtained abundantly, the utilization of these shells as carbon precursor
is very promising.
Although this country produces huge amounts of coconut shells as
agricultural wastages, no statistical data are available on the quantity. The proximate
and ultimate analyses of coconut shells are given in Table 1.4. The proximate
analyses of oil palm and coconut shells are not very different except in ash content.
Ultimate analysis shows higher carbon, nitrogen and sulphur contents in oil palm
shell. The higher nitrogen and sulphur contents in oil palm shell may be converted
into functional groups containing donor nitrogen and sulphur atoms during activation
and thus may be functioned more effectively in removing borderline and soft metal
ions. Both precursors contain higher amounts of volatile matter and oxygen suitable
to develop porosity and surface acidic groups containing oxygen depending on
activation process. Identical carbon content in the proximate analyses of oil palm
and coconut shells grouped them in a similar type of lignocellulosic precursor. Thus
29
the utilization of coconut shells as supplementary precursor in the oil palm shell
based carbon manufacturing industry allows more flexibility on the consumption of
raw material.
This is why, oil palm and coconut shells are chosen as carbon
precursors in this study.
Fresh Fruit Bunch
Nuts
Fibre
Sterilization
Depericarping
Nut silo
Stripping
Digestion
Nut cracker
Pressing
Press cake
Crude oil
Kernel
Cracked mixture
Shell
Figure 1.5
Process flow diagram of palm oil extraction showing the production
of crude oil and wastages such as shells, kernel and fibres.
Table 1.2 : Proximate analysis of solid oil palm residues (Husain, Zainac and
Abdullah, 2002).
Composition
Fibre
Shell
EFB
Volatile matter, wt.%
72.8
76.3
75.7
Fixed carbon, wt.%
18.8
20.5
17
Ash, wt.%
8.4
3.2
7.3
30
Table 1.3 : Ultimate analysis of solid oil palm residues (Mahila et al., 2001).
Element
Fibre
Shell
EFB
Hydrogen, wt.%
6
6.3
6.3
Carbon, wt.%
47.2
52.4
48.8
Sulphur, wt.%
0.3
0.2
0.2
Nitrogen, wt.%
1.4
0.6
0.7
Oxygen, wt.%
36.7
37.3
36.7
Ash, wt.%
8.4
3.2
7.3
Table 1.4 : Proximate and ultimate analyses of coconut shell (Banerjee, et al.,1976).
1.4
Composition
Proximate analysis
Ultimate analysis
Moisture, wt.%
8.4
-
Volatile matter, wt.%
72.8
-
Fixed carbon, wt.%
18.7
-
Ash, wt.%
0.1
-
Carbon, wt.%
-
48.4
Hydrogen, wt.%
-
6.3
Oxygen, wt.%
-
44.3
Sulphur, wt.%
-
0.03-0.09
Nitrogen, wt.%
-
0.1-0.15
Adsorption of Metal Ions onto Activated Carbon
In the last two decades, the adsorptions of metal ions from the single- and a
few binary- and multi-solute solutions onto mostly commercial and some noncommercial, laboratory prepared activated carbons were investigated. These studies
were reviewed in the following paragraphs.
31
Corapcioglu and Huang (1987) studied the adsorption characteristics of some
heavy metals, namely Cu(II), Pb(II), Ni(II) and Zn(II) onto the hydrous surface of 14
different activated carbons obtained from various commercial brands.
The
adsorption characteristics were determined for carbon type, pH and surface loading.
They observed that adsorption is the major phenomenon in the removal of metal ions
at least in the pH less than neutral region. The adsorption equilibrium data were
obtained at varying pH, from 2.5 to 10.5, in a single-solute solution. The initial
concentrations of Cu(II), Pb(II), Ni(II) and Zn(II) were kept constant at 6.4, 20.7,
29.3 and 32.7 mg.L-1 respectively. The concentration of adsorbent was kept constant
at 10 g.L-1. The highest adsorption capacity for the above mentioned metal ions in
the acidic region were observed by the activated carbons having acidic properties
(Nuchar SA and Nuchar SN brands). Since the ashes of these acidic carbons contain
higher level of phosphorus, they speculated that heavy metals were removed through
the formation of organometallic complexes with phosphoryl group rather than
hydroxo complexes. Within the pH range from 3 to 5, the acid activated carbon had
removed above 90% of the metal ions of Cu(II) and Pb(II). But the removal of
Ni(II) and Zn(II) by the acid activated carbon was observed only below 70% at pH 3
which increased gradually to above 90% at pH 5. In the alkaline region, the metal
ions removal capacities of all activated carbons were indistinguishable. The effect of
surface loading was studied varying initial concentrations over the pH range, 2.510.5. The results showed that the percentage of heavy metal removal by acid
activated carbon decreases with increasing surface loading. The adsorption edges
for Cu(II) and Pb(II) were found in the pH range from 3 to 5 and 3 to 6 respectively.
Examining various models to describe adsorption reaction, they selected the
surface complex formation model with some variations. To develop the present
version of this model, they include all the hydroxo species, M(OH)y2-y, along with
the free metal ions, M2+ as adsorbate. Although their evaluation indicated that
covalent or hydrogen bonding is responsible for the adsorption, they concluded their
opinion in favour of hydrogen bonding.
Thus they suspected that electrostatic
interaction (outer-sphere complexation) plays a significant role rather than chemical
interaction (inner-sphere complexation) in the adsorption reaction. This speculation
seems to be questionable since hydrogen bonding results in a very weak bonding and
non-selectivity among the cations having similar charge.
This assumption is
32
contradictory to the findings of Strelko and Malik (2002) that activated carbon
adsorbs metal ions selectively.
Tan and Teo (1987) studied the adsorption of Pb(II) and chromium on
activated carbon. They showed that adsorption depended significantly on the pH,
carbon dosage and initial adsorbate concentration.
They proposed empirical
modifications of the expressions of Langmuir and Freundlich equations in terms of
mass ratio of carbon to initial amount of adsorbate in the solution to correlate the
experimental isotherms for the combined influence of initial adsorbate concentration
and carbon dosage.
Johns, Marshall and Toles (1998) studied the single and competitive
adsorption of Cu(II) and of Cd(II), Cu(II), Pb(II), Ni(II) and Zn(II) in single and
multi-solute solutions respectively onto various granular activated carbons. The
adsorbent and initial Cu(II) concentrations were 10 g.L-1 and 635 mg.L-1 respectively
and the solution was buffered at pH 5. For competitive adsorption, unbuffered (pH
5) solution of concentration of each cation as 2.5 mM was used. Some of the
activated carbons used were derived from various lignocellulosic precursors while
others were obtained from commercial sources. They found that physically activated
and controlled air oxidized activated carbons have higher Cu(II) adsorption
capacities as 41-51 mg.g-1. But the limitation of this study is that they determined
the equilibrium adsorption data for the single point concentration which is not
enough to understand the adsorption characteristics. The highest adsorption of the
total solutes by one of the prepared activated carbon was observed as 510 µmol.g-1
and the selectivity order was reported as Pb > Cu > Cd > Zn >Ni. All granular
activated carbons adsorbed Pb(II) and Cu(II) in much higher amounts than other
cations such as Cd, Zn and Ni.
Toles, Marshall and Johns (1998) in another experiment prepared some
activated carbons using phosphoric acid and controlled air from different nut shell
precursors. They observed that various acid activated/oxidized carbons regardless of
precursor type scavenged 13-19 mg-Cu.g-1-adsorbent. The adsorbent and initial
Cu(II) concentrations were 10 g.L-1 and 190 mg.L-1. Since acid activated carbons
altered the equilibrated pH to a value out of the range of optimum metal adsorption,
33
they buffered the solution pH at 4.8. They also studied the adsorption from the
initial concentration of 635 mg.L-1 and observed the highest adsorption capacity by
one of the acid activated carbon derived from almond shell as 59 mg.g-1. The
adsorption behaviour was not well documented by fitting the data to equilibrium and
kinetic models. They successfully studied the feasibility of the reuse of phosphoric
acid used in precursor activation. Activated carbon prepared from 85% concentrated
recycled acid produces lower surface area. They recycled the acid up to 50%
concentration and mixed with fresh acid in a 1:1 ratio. As the activated carbon
prepared from mixed acid has a much higher surface area, this technique brought
success to them.
Seco and co-workers (1999) investigated the single adsorption of Cd(II) and
Cu(II) on commercial activated carbon as a function of pH, metal ion and carbon
concentrations.
Their results showed that activated carbon has the potential to
remove Cd(II) and Cu(II) from aqueous solution. They also observed that the
adsorption increased with the increase of pH and carbon concentration. They found
that an increase of initial metal concentration reduced the metal removal percentage.
They successfully interpreted the stronger affinity of Cu(II) than Cd(II) towards
carbon from their electronegativities, as 2.00 and 1.69 and first hydrolysis
equilibrium constants (pK values), as 8.00 and 10.08 respectively. They applied
mechanistic model, such as Triple Layer SCF model, to predict the adsorption
behaviour of Cd(II) and Cu(II) on carbon.
They also tested Langmuir and
Freundlich models while Langmuir model failed to fit the data indicating multi-layer
adsorption. They also found that while Triple Layer SCF model has the advantage to
use in a wide range of pH and carbon concentration, Freundlich model is applicable
only in one pH, that is, for the whole pH range Freundlich model requires the
conjunctive use of one equation for each pH.
Chen and Wang (2000) studied the adsorption of Cu(II), Zn(II) and Pb(II) in
their single-, binary- and ternary-solutes solution conducting the experiment in a
fixed-bed column. They observed that the breakthrough capacity increases with the
increase of influent pH and the lowering of flow rate. The ionic strength has small
increasing effect only on the removal of Cu(II).
The selectivity order for the
adsorption of metal ions in single solution was showed as: Cu > Pb > Zn. The multi-
34
solute adsorption indicated that activated carbon has higher affinity towards Cu(II)
and least affinity towards Zn(II). The adsorption capacity for Cu(II) was slightly
suppressed by the presence of competing ions Zn(II) or Pb(II) in binary- and by
Pb(II) and Zn(II) in ternary-solute solution. Adsorption of Pb(II) was found similar
to that of Cu(II) and the competitive effect by Cu(II), and by Cu(II) and Zn(II) was
slightly higher than that for Cu(II). The competing effect of Cu(II), and of Cu(II)
and Pb(II) on the suppression of adsorption of Zn(II) was quite significant. They
also observed that addition of Cu(II) reduces the amount of Zn(II) removal and that
was why, they directly assumed that only one functional group affecting the
adsorption of various metal ions. Since it seemed correct, they uphold their opinion
by adding that if several groups are present to bind different metal ions, there would
be no competition. Actually here they failed to generalize their observation and thus
they went against the very well known concept of the presence of various surface
functional groups on activated carbon. However, utilizing the concept of HSAB
(Pearson, 1968) the various active sites on carbon as well as the types of metal ions
can be categorized to hard and soft sites or ions and can be derived to a more
realistic interpretation.
Alfarra, Frackowiak and Francois (2004) applied successfully the HSAB
concept to interpret the adsorption behaviour of metal ions on activated carbon by
considering surface groups of activated carbon as their hard sites and the surface of
basal planes as soft sites. In a recent report of IUPAC (2002) Pb(II) is shown as soft
cation whereas Ahrland, Chatt and Davies (1958) classified Pb(II), Cu(II) and Zn(II)
as borderline cations. Therefore, Chen and Wang’s (2000) assumption that Pb(II),
Cu(II) and Zn(II) were adsorbed on a same functional group may not be correct.
However, Pb(II) might be suspected as to be adsorbed onto both harder and softer
active sites on carbon while Cu(II) and Zn(II) only onto the harder active sites. Thus
the differences in Lewis acidity of Pb(II) and Cu(II) directed them to bind differently
which explains more correctly the less effect of competing ions in the adsorption of
Cu(II) and Pb(II) on carbon. Chen and Wang (2000) did not explain the cause of
higher affinity towards Cu(II) which can be showed from the stability of complexes
of transition metal as a function of electronegativity in Irving-Williams series (cited
by Shriver, Atkins and Langford, 1991).
35
Uzun and Güzel (2000) studied the adsorption of heavy metal ions, Mn(II),
Fe(II), Ni(II) and Cu(II) on commercial activated carbon, chitosan and agar. They
described the adsorption data as percent removal which is inadequate to evaluate the
adsorption behaviour. However, they obtained the order of adsorption of metal ions
on activated carbon in the series: Cu > Ni > Fe > Mn which was consistent with the
stability of complexes of transition metal in Irving-Williams series.
Chen and Lin (2000) investigated the adsorption of metal ions, Cu(II), Zn(II)
and Co(II), on an H-type (basic) commercial activated carbon. They used adsorbent
and initial solution concentration as 10 g.L-1 and <10 mg.L-1. In these experiments
equilibrated solution pH was found to increase from 5 to 7. The removal order of
cations in single-solute solution was observed as follows: Cu > Zn ≈ Co. They also
observed that the competitive effect of Zn(II) and Co(II) on Cu(II) is less prominent
whereas the adsorption of Zn(II) and Co(II) was reduced dramatically with the
addition of Cu(II). The kinetic study showed that most Cu(II) removal occurred in
the first hour and was completed within four hours. They also observed that Cu(II)
adsorption is much faster than that of Zn(II).
Dastgheib and Rockstraw (2001) prepared activated carbon with a high
adsorption capacity for Cu(II) from pecan shell using air and H3PO4. They proposed
that the acidic groups as detected using Boehm titration not only be considered as
oxygen-containing acidic groups, but also as oxygen/phosphorous groups. They
used buffered (pH 4.8) solution of Cu(II) and found the adsorption capacities by
prepared activated carbons to be varied from 33 to about 40 mg.g-1 depending on the
extent of oxidation. They proposed the mechanism as ion-exchange in very low
concentrations and accounted the pH shift by acid activated carbon as the release of
H+ by the equivalent (eq) Cu2+. Since the release of H+ ion by acid activated carbon
depends on pH at zero point charge (pHZPC) one can relate the released H+ ion
concentration to the adsorbed metal ion density on the adsorbent only to a certain
level of adsorption. That was why, while correlating the meq H+ released per gram
carbon by the meq Cu2+ adsorbed per gram carbon, they failed to establish any
relation for higher concentration. Thus they proposed another mechanism for high
concentration as some other forms of ion-exchange and surface complexation of
36
metal ions with oxygen- and phosphorous-containing functional groups. They did
not show any evidence or interpretation of this assumption.
Toles and Marshall (2002) investigated the utility of phosphoric acidactivated/oxidized, steam-activated and steam-activated/oxidized activated carbons
in removing Cu(II) from solution in batch and column. The equilibrium data were
obtained varying initial Cu(II) concentrations from 31 to 1906 mg.L-1 while
adsorbent concentration was maintained 10 g.L-1. The obtained equilibrium data
were compared with commercial activated carbons by fitting to Freundlich model.
The acid activated carbons were found most effective in both, batch and column
mode, for the removal of Cu(II).
Strelko and Malik (2002) investigated the effect of chemical (HNO3)
oxidation on a commercial activated carbon to modify it to a cation-exchanger.
They tested the adsorption behaviour of some transition metal ions such as Cu(II),
Ni(II), Co(II), Zn(II) and Mn(II), on unoxidized and oxidized activated carbons.
Their results showed that Cu(II) uptake was 4 fold enhanced by the oxidized carbon.
They also observed that Cu(II) was the most preferred cation and its non-linear
Langmuir isotherm was considerably steeper at low concentration than that for other
metal ions. Metal ions uptake capacity by the unoxidized activated carbon was
significantly lower than that of oxidized activated carbon while uptake
pattern/sequence by both carbons was detected similar. They also depicted the
affinity series as: Mn2+ < Co2+ < Ni2+ < Cu2+ > Zn2+ from the respective selectivity
coefficients as: 2.66 × 10 −3 , 6.38 × 10 −3 , 9.45 × 10 −3 , 1.95 × 10 −1 and 5.61 × 10 −3 . They
proposed distorted and more stable octahedral complexation of Cu(II) with activated
carbon to understand its stronger affinity towards carbon.
Mohan and Singh (2002) investigated single and multi-component adsorption
of Cd(II) and Zn(II) on H2SO4 impregnated activated carbon derived from bagasse
precursor. They observed that equilibrium time was reached within 10-12 h and the
kinetic data followed Lagergren pseudo-first order model with rate constants of
Cd(II) and Zn(II) as 0.006 and 0.008 min-1 respectively. They fitted their data to the
Langmuir and Freundlich models and evaluated the adsorption behaviour. They
37
found that the prepared activated carbon had a very high capacity to remove Cd(II)
and Zn(II) in single-solute-solution as 38.03 and 31.11 mg.g-1 respectively. They
also showed that the competitive effect of Cu(II) and Zn(II) on Cd(II) and that of
Cu(II) and Cd(II) on Zn(II) in binary and ternary solutions was not much as the ratio
of maximum monolayer adsorption capacities in multi-solute to single-solute
solutions were within 0.78-0.87 for Cd(II) and 0.61-0.84 for Zn(II).
Krishnan and Anirudhan (2002) studied the adsorption of Pb(II), Hg(II),
Cd(II) and Co(II) on the prepared sulphurized steam activated carbon as a function
of time, concentration, pH, and temperature.
Their results showed that while
adsorbent and initial metal ion concentrations were kept at 2 g.L-1 and 100 mg.L-1
respectively and the pH at 6, the maximum sorption had occurred (Pb 49.60, Hg
48.61, Cd 46.55, Co 40.95 mg.g-1) within four hours. The following adsorption
sequence was observed: Pb(II) > Hg(II) > Cd(II) > Co (II). Their obtained kinetic
data were fitted to pseudo-second order kinetic model. They analysed the adsorption
data by Langmuir and Freundlich isotherms. The maximum adsorption capacities
for Pb(II), Hg(II), Cd(II) and Co(II) were as follows: 200, 188.68, 153.85 and 128.70
mg.g-1. The effect of competing ions on the single adsorption capacities for all metal
ions was found suppressive.
An acidic solution of 0.2 M HCl was found as
regenerating agent for the spent activated carbon.
Macías-García and co-workers (2004) investigated the adsorption of Pb(II)
on SO2-treated as well as on untreated activated carbon from the kinetic and
equilibrium standpoints. The kinetic result showed that adsorption on untreated
carbon was much faster (equilibration time 20 h) than that on treated ones
(equilibration time 80 h). Therefore, they proposed that untreated carbon adsorbs
Pb(II) physically whereas the adsorption on SO2-treated activated carbon might be
due to chemisorption which was resulted from the interaction of metal ions with the
sulphur bearing functional groups. They derived the pseudo-first order rate constant
applying Lagergren kinetic model and assumed that chemisorption is the rate
limiting step. The values of rate constants by one untreated and three various treated
activated carbons at pH 5.4 and at temperature 25°C were respectively as follows:
0.903, 0.065, 0.042 and 0.034 h-1.
38
Aggarwal, Goyal and Bansal (1999) studied the adsorption of Cr(III) and
Cr(VI) on various as received and oxidized activated carbon. They constructed the
adsorption isotherms and found that activated carbon fibres having higher portion of
very small micropores adsorbed least Cr(III) due to inaccessible pores than others as
received activated carbons. The adsorption of Cr(III) increased on oxidation and
decreased on degassing of adsorbent. They found that surface acidic groups were
well developed by treatment with HNO3 acid than that by H2O2, air and (NH4)2S2O8.
It was also observed that the amount of Cr(VI) adsorption was much higher than
Cr(III) on as received activated carbons and carbon-fibers. The larger adsorption
was interpreted by the smaller size of Cr(VI). In the adsorption of Cr(VI) they
noticed a reverse behaviour that the removal capacity decreased with the oxidation
while increased with degassing.
That is, acidic surface groups exposing more
negative active sites repulsed anions of Cr(VI) with an ultimate suppressive effect.
Selomulya, Meeyoo and Amal (1999) studied the removal mechanism of
Cr(VI) using different types of activated carbons. In their experiment, one of the
commercial activated carbons was derived from wood and was of L-type while the
other two were of H-type and obtained from coconut shell (prepared in laboratory)
and dust coal (commercial). They observed that the removal capacity of Cr(VI) ions
always decreased with the increase in pH for all types of carbons used in their study.
They proposed the mechanism of total chromium removal by activated carbon as the
direct adsorption of Cr(VI) that could be followed by the reduction of some Cr(VI)
to Cr(III). They also observed that the mechanism was highly dependent on the type
of activated carbon they used, the pH values and the amount of Cr present in the
solution while the optimum pH for the removal was largely depended on the surface
properties and the reduction capacity of carbons. H-type coconut shell based carbon
was found to be excellent in removing Cr in a wide range of concentration and in a
low level of pH while L-type carbon was found good only at low concentration and
at low pH.
Ho and Mckay (1999) studied the kinetics of adsorption of divalent metal
ions such as Cu(II), Pb(II), and Ni(II) onto sphagnum moss peat. They followed
batch- adsorption technique to obtain the equilibrium adsorption and the extent of
metal ion removal. Since their kinetic data did not fit to Lagergren pseudo-first
39
order model, they developed a pseudo-second order rate equation and calculated the
rate constants and the initial adsorption rates. They claimed that chemisorption,
which involves valency forces through sharing or the exchange of electrons between
adsorbent and adsorbate, was the rate-limiting step. They also showed that the
parameter which had the influence on the kinetics of the adsorption reaction was the
adsorption equilibrium capacity which was a function of initial metal ion
concentration, peat dose and the nature of solute ion.
Rivera-Utrilla and Sánchez-Polo (2003) investigated the potential of a series
of ozonised activated carbons to remove Cr(III) from aqueous solution and the
mechanism of adsorption. They studied their data by applying Langmuir model and
observed that maximum monolayer adsorption capacity of unoxidized basic carbon
(H-type) could be increased from 7.33 to 19.23 mg.g-1 by oxidation with ozone. In
that experiment, they altered the nature of a hydrophobic carbon (pHZPC = 8.82) to a
hydrophilic one (pHZPC = 1.82). Thus they showed that the ionized surface groups of
oxidized carbon attracted Cr(III) species by electrostatic interaction. They also
established a linear relationship between the amount of metal adsorbed and the
concentration of acid groups on the carbon surface. While they observed that basic
carbon even with positive surface charge density could remove the Cr(III) species,
they proposed a mechanism of Cл (л-electron clouds of basal plane of activated
carbon) and cations interactions. They also analysed the effect of pH and reported
the result as follows: the maximum adsorption occurred when the charge sign of the
carbon surface dominated as the opposite to that of the Cr(III) species present at the
pH of that experiment.
Demirbas and co-workers (2004) prepared three activated carbons from
cornelian cherry (CC), apricot stone (AS) and almond shell (ACS) using H2SO4
activation to remove aqueous Cr(VI). They determined the optimum pH for the
removal of Cr(VI) as 1 when all activated carbons showed identical capacity as 2021 mg.g-1. But they observed different adsorption capacities at pH 4, as 4.21, 11.44
and 14 mg.g-1 by CC, AS and ACS carbons respectively. They determined the
equilibration time as 72 hours and fitted the kinetic data to the pseudo-second order
equation. They found that the values of pseudo-second order rate constant decreased
with the increase of initial concentration. The rate constants by three activated
40
carbons, CC, AS, and ACS were determined as 0.058, 0.040 and 0.053 mg.g-1.h-1
respectively while they were adsorbed in the initial concentrations of 53, 58 and 21
mg.L-1.
Yantasee and co-workers (2004) developed an adsorbent material by the
functionalization of amine (-NH2) onto fine-grained activated carbon to remove
transition metal ions such as Cu(II) effectively and selectively. The equilibration
time was reached within 1 min., while the maximum monolayer adsorption capacity
was determined as 54.6 mg.g-1 by the developed adsorbent. They also determined
the effect of competing ions such as Cd(II), Ni(II), and Pb(II) on Cu(II) in the series
as follows Cu(II) >> Pb(II) > Ni(II) > Cd(II).
Reed, Vaughan and Jiang (2000) obtained Fe(III)-oxide impregnated
activated carbon (FeAC) and its non-impregnated counterpart from Norit Americas,
Atlanta, Georgia to study the removal of As(III), As(V), Hg(II), and Pb(II). The
detail of impregnating the virgin carbon, which was derived from lignite precursor,
is trade confidential.
They conducted adsorption experiments using adsorbate
concentration 1 mg.L-1 while pH were varied from 3 to 11 and adsorbent
concentrations were varied between 0.015 and 0.3 g.L-1 for virgin carbon and
between 0.2 and 1.2 g.L-1 for FeAC. They observed that the removal of oxyanions
of As(V) decreased with increasing pH, while reverse phenomenon was observed for
cations, Pb(II) and Hg(II). The removal behaviour of As(III) was observed different
from As(V).
The adsorption of As(III) was maximized at pH ≈ 7; whereas
adsorption was independent of pH at pH below 5 and was decreased at pH above 7.
From a single point of concentration at pH 7, they showed that the adsorption
capacities of virgin and FeAC activated carbons for As(III), As(V), Hg(II) and Pb(II)
were as follows: 0.73 and 4.67, 0.09 and 4.50, 2.1 and 4.57, 3.0 and 4.35 mg.g-1
respectively. A substantial increase in the removal of As(III) and As(V) by the Feoxide impregnated carbon was occurred due to the empty d-orbitals which allowed
the
rooms
for
the
complexation
of
heavy
metal
anions
through
oxolation/replacement of the hydroxyl group. Thus they showed that instead of
carbon-oxygen functional groups’ interaction with anions, the effective removal of
heavy metal anions has occurred by the Fe(III)-oxide impregnated carbon.
41
Peräniemi and Ahlgrén (1995) studied the preconcentration of As(V), Se(IV),
Se(VI) and Hg(II) in aqueous solution using zirconium-loaded activated charcoal
(ZrC*) to develop a rapid and accurate determination technique using dispersive Xray fluorescence as well as to provide an effective recovery for these species. They
used slurry of activated charcoal (25 g) in an aqueous solution (250 mL) of zirconyl
nitrate (10.0 g Zr.L-1, pH 1.6) to load Zr onto charcoal. Their results showed that
their developed method was very suitable to be used for industrial wastewaters.
Since the analyte contents in environmental waters are usually too low they
concluded that this method was not directly suitable for the samples from
environment.
Peng and co-workers (2005) developed carbon nanotubes-iron oxides
magnetic composites as effective adsorbent for the removal of Pb(II) and Cu(II).
They studied the effect of pH on the removal percentage as well as on the
constructed isotherms. They obtained the adsorption capacities for Pb(II) and Cu(II)
at pH 5 as 103 and 45 mg.g-1 respectively.
1.5
Adsorption Equilibrium Models
Adsorption equilibrium models (Weber, McGinley and Katz, 1991) are
classified in two major groups, mechanistic and phenomenological models.
Mechanistic models include hydrophobic, ion-exchange, and surface complex
formation (SCF) models. In ion-exchange and SCF models the calculations are
based on the stability constants of the adsorbent surface. These constants can be
obtained from the potentiometric titrations and other experiments. At equilibrium,
the relationship between the concentrations of solute in the liquid or gas phase with
that of solid phase at a constant temperature is expressed by adsorption isotherm
equations. The phenomenological models are based on these adsorption isotherms.
Many phenomenological models are developed for the gas phase adsorption in single
and multi-component systems. Although these equations were developed for gas
phase adsorption, they also have been successfully applied to dilute liquid phase
adsorption. Among them, Langmuir and Freundlich adsorption isotherms are most
42
common models in the adsorption of gas or liquid on a solid phase (Faust and Aly,
1987).
1.5.1
Langmuir Adsorption Model
The term ‘adsorption’ deals with the process in which molecules accumulate
in the interfacial layer, but desorption denotes the converse process. Adsorption can
result either from the universal van der Waals interactions (physical adsorption,
physisorption) or it can have the character of a chemical process (chemical
adsorption or chemisorption).
Since the adsorption isotherm is the equilibrium
relation between the quantity of the adsorbed material and the pressure or
concentration in the bulk fluid phase at constant temperature, it provides the primary
source of information on the adsorption process.
The Langmuir equation initially derived from kinetic studies was based on
the assumption that on the adsorbent surface there is a definite and energetically
equivalent number of adsorption sites, at each of which one molecule of a perfect gas
may be adsorbed. The bonding to the adsorption sites can be either chemical or
physical, but it must be sufficiently strong to prevent displacement of adsorbed
molecules along the surface. Therefore, adsorption occurs only in monolayer and
thus Langmuir isotherm notices the uniformness or homogeneity of a surface.
Langmuir, for the first time, introduced a clear concept of the monomolecular
adsorption on energetically homogeneous surfaces.
The statement proposed by
Langmuir was applied to chemisorption and with some restrictions to physical
adsorption.
The constant parameters of the Langmuir equation have a strictly
defined physical meaning, in contrast to the parameters of the empirical Freundlich
equation. For adsorption from solution by solid adsorbents, the Langmuir adsorption
isotherm is expressed as:
qe =
q max bC e
1 + bC e
(1.9)
43
where, qe = amount of metal ion adsorbed at equilibrium per unit mass activated
carbon (mg.g-1); Ce = equilibrium concentration of metal ion in solution (mg.L-1);
qmax = the maximum monolayer adsorption capacity (mg.g-1); b = affinity or
adsorption constant, related to the heat of adsorption, (dm3.g-1).
The linear form of equation (1.9) is derived as equation (1.10) to determine
the Langmuir parameters. Plotting Ce/qe against Ce gives a straight line with a slope
1/qmax and an intercept 1/bqmax.
Ce
1
1
Ce +
=
q e q max
bq max
(1.10)
The monolayer capacity, qmax, determined from the Langmuir isotherm,
defines the total capacity of the adsorbent for a specific adsorbate. Reliable qmax
values can be obtained only for systems exhibiting Type-1 isotherms of the
Brunauer’s classification. The monolayer capacity may be used to determine the
specific surface area of the adsorbent by utilizing a solute of known molecular area.
1.5.2
Freundlich Adsorption Model
At the end of the 1940s and at the beginning of the 1950s, work on physical
adsorption of gases on energetically heterogeneous solid surfaces became an object
of interest with an ultimate derivation of an adsorption isotherm in the form of the
so-called generalised Freundlich equation. The main source of heterogeneity for
microporous solids is their complex porous structure which contains micropores of
different dimension and shape. The mathematical forms of the overall adsorption
isotherms depend only on the shape of the energy distribution functions, that
characterise the global heterogeneity of the adsorbent surface.
The Freundlich adsorption equation is perhaps the most widely used
mathematical description of adsorption in aqueous systems.
Generally, the
Freundlich isotherm describes adsorption of organic compounds on activated carbon
44
better than the Langmuir-like isotherms. Freundlich isotherm can be obtained from
the Langmuir isotherm using the following assumptions:
i) The adsorbent surface is heterogeneous.
ii) The site energies are distributed exponentially.
iii) For all sites with the same energy, a Langmuir isotherm is applicable.
The Freundlich equation is expressed as:
qe = K F C e1 / n
(1.11)
where, qe and Ce have the same meanings as in equation (1.9), KF (mg.g-1) and “n”
(g.dm-3) are the Freundlich empirical constants reveal the characteristic of adsorbent
related to adsorption capacity and intensity respectively.
The Freundlich constant, KF unlike Langmuir constant, qmax does not predict
the saturation of the solid surface by the monolayer coverage of the adsorbate
(Mohan and Singh, 2002). But it gives a relative measure in adsorption capacity.
The value KF is thus equivalent to qmax in Langmuir equation. Moreover it estimates
the bond strength (Toles and Marshall, 2002).
The value of “n” discloses the adsorption pattern. The favourable adsorption
is understood from the values of 1 < n < 10 while irreversible adsorption is noticed
from n > 10 and unfavourable from n < 1 (Do, 1998). Thus, one can see that the
Freundlich equation is an empirical expression that encompasses the heterogeneity
of the surface and the exponential distribution of sites and their energies.
The simplified linear logarithm form of equation (1.11) is presented in
equation (1.12). Plotting log qe against log Ce gives a straight line with a slope 1/n
and an intercept log KF.
log q e = log K F +
1
log C e
n
(1.12)
45
1.6
Aims and Objectives
Present research trends showed that many lignocellulosic materials are good
precursors especially for scavenging borderline and soft metal ions. Since Malaysia
is rapidly moving towards the concept of safe drinking water, her huge amounts of
agricultural wastages namely, oil palm kernel and coconut shells are needed to be
testified as metal ions scavenger. In the literature review, it is also observed that
H3PO4 activation and low temperature carbonization produces activated carbon with
enhanced metal ions removal capacity. Since the anions of metals/metalloids are
expected to be adsorbed more effectively on the surface of Fe(III)-oxide, prepared
activated carbons are also needed to modify through dispersing hydrated Fe(III)oxide onto its surface.
The processing industries have increasingly been generating heavy metals of
which the production of copper, lead and zinc had increased tremendously with a
tenfold increase by the years between 1850 and 1990 (Cech, 2005). We have seen
borderline metals form relatively stable complexes with both hard and soft donor
ligands. The first row d-block transition metal ions in the periodic table fall mainly
into the group of borderline. Their stability order in complexes can be shown from
Irving-Williams series. Nickel and lead are the highly toxic (toxic to all life forms
even in low concentrations) metals among all borderline and marginal soft metals.
USEPA Safe Drinking Water Act Amendments of 1996 selected the contamination
level of copper as 1.3 mg.L-1 because of its toxicity to liver and kidney (Cech, 2005).
Thus, the divalent cations of lead, copper, nickel and zinc are given priority
for the removal using activated carbon from the toxicity and environmental
standpoint as well as from the underlying adsorption mechanism. The anions of
chromium(VI) and arsenic(V), and the non-ionic species of arsenic(III) are taken
into adsorption experiment to observe the suitability of the prepared activated carbon
and modified carbon-HFO composites in the removal of highly toxic anionic and
non-ionic species.
Among various adsorption experiments bench scale batch
technique, which uses synthetic solution of mg.L-1, is considered an effective method
especially for the preliminary stage of investigation.
46
1.6.1
Objectives
Particularly the development of the activated carbon preparation method is
based on the trial and error approach guided by the background reading of the
previous processes and reviewed articles.
The following objectives have been
addressed to testify the hypothesis:
1. To prepare and characterize activated carbon from oil palm kernel and
coconut shells as local raw materials.
2. To study the effect of preparation variables and to modify the prepared
activated carbon with subsequent characterization.
3. To conduct the batch adsorption of some metal ions on the prepared
activated carbon and on modified carbon-hydrated iron(III)-oxide
composite adsorbent as to study the adsorption equilibria and kinetics
using established models.
1.7
Scope and Limitations
The prepared acid activated carbons are expected to have the potential in
removing borderline and soft metal ions in single as well as in multi solutes solutions
over a wide range of concentrations and pH. The removal of As(III) and As(V) is
attempted to be feasible through the modification of activated carbon using hydrated
Fe(III) oxide. Thus to suit the adsorbent as filtering media in a fixed bed to be used
especially in water treatment, selective adsorption of target metal ion as well as its
desorption studies are designed to conduct. The main limitation in the preparation of
activated carbon is that the atmosphere of muffled furnace cannot be defined. The
leakage of air through the door controls the furnace atmosphere by design default.
CHAPTER 2
MATERIALS AND METHODS
2.1
Chemicals and Reagents
The reagents used to pretreat and impregnate the raw materials were
sulphuric acid, H2SO4 (98%) from Merck (Darmstadt, Germany) and phosphoric
acid, H3PO4 (85%) from Mallinckrodt (Paris, Kentucky).
Sodium hydrogen
carbonate, NaHCO3 (99.7-100.3%); sodium carbonate, Na2CO3 (99.5%); sodium
hydroxide (NaOH) pellet (99%); and sodium ethoxide, NaOC2H5 (95%) were used
in Boehm’s titration and were obtained from Univer (Auburn, Australia), RiedeldeHaën (Seelze) and Merck (Hohenbrunn, Germany) respectively. Activated carbon
was modified using iron (III) chloride 6-hydrate, FeCl3, 6H2O (97-102%, (Hamburg
Chemicals); calcium (II) hydroxide, Ca(OH)2 (GCE); sodium chloride, NaCl (99.5%,
Farmitalia Carloerba) and Ethanol, C2H5OH (95%, Rinting Scientific/Fluka,
Switzerland). Various metal ion solutions and reagents were prepared using the
following chemicals and reagents: copper (II) chloride dihydrate, CuCl2, 2H2O
(GCE); calcium (II) nitrate tetrahydrate, Ca(NO3)2, 4H2O (Riedel-deHaën,
Germany); nickel (II) nitrate hexahydrate, Ni(NO3)2, 6H2O ((Fluka, Switzerland);
lead (II) nitrate, Pb(NO3)2 (Fluka, Switzerland); zinc (II) nitrate hexahydrate,
Zn(NO3)2, 6H2O (Emory); potassium dichromate, K2Cr2O7 (Merck, Darmstadt,
Germany); 1,5-diphenylcarbohydrazide, (C6H5-NH-NH)2CO (Merck); acetone,
(CH3)2CO (J. T. Baker, NJ, USA); di-arsenic trioxide, As2O3 and di-arsenic
pentoxide, As2O5 (BDH, Poole, England); nitric acid, HNO3 (65%, Merck); and
hydrochloric acid, HCl (37%, Mallinckrodt, Paris, Kentucky). Nitrogen gas was
48
supplied in a cylinder by MOX (Malaysian Oxygen), Malaysia.
Commercial
activated carbon, “Aktivkohle”, (abbreviated, throughout this thesis, as CAC) was
supplied by Riedel-deHaën, Germany. All reagents and chemicals were of analytical
grade. Standard solutions (1000 mg.L-1) of metal ions such as Fe(III), Cu(II), Zn(II)
and Ca(II) were obtained from Merck; Ni(II) and As(III) from BDH; and Pb(II) from
Riedel-deHaën. All glasswares and polypropylene tubes were washed by immersing
in 10% HNO3 for 24 h followed by washing with distilled water.
2.2
Instruments
The following instruments were used to prepare, modify and characterize
activated carbons as well as to conduct batch adsorption studies: mechanical
grinding machine; muffle furnace (Carbolite muffle furnace, model: ELF 11/6B,
Barloworld Scientific, England); tube furnace (Barnstead/Thermolyne tube furnace
21100, model no. F21130-33, Barnstead International, USA); surface area analyser,
ASAP 2010 (Micromeritics Instrument Corporation, USA); scanning electron
microscope (SEM) (Philips, model XL 40) incorporated with energy dispersive Xray
analysis
(EDX)
(EDAX
Inc.
USA);
Fourier
transform
infrared
spectrophotometer, FTIR (model FTIR-8300, Shimadzu, Japan); thermogravimetric
analyser (TGA) (Mettler TA 4000); orbital shaker (Protech, model no. 722);
CyberScan pH/Ion 510 pH meter (Eutech Instruments); atomic absorption
spectrophotometer (AAS) (GBC, model Avanta, Australia/ Perkin Elmer, model
AAnalyst 400); ultra violet-visible (UV-Vis) spectrophotometer (Perkin Elmer,
model Lambda 25) and inducedly coupled plasma mass spectrophotometer (ICPMS) (Perkin Elmer, model Elan 6000).
2.3
Preparation of Activated Carbon
Activated carbon preparation consisted of raw material preparation,
activation and washing. Activation of the prepared raw material was performed
49
through chemical impregnation followed by subsequent carbonization.
A flow-
diagram of the preparation of activated carbon is shown in Figure 2.1.
Raw materials as collected
Washed, dried, crushed and sieved to selected particle sizes,
2.36-1.18 and 1.18-0.60 mm
Pretreatment with acids, 30% H2SO4/H3PO4
Washed with distilled water, dried in the sunlight
Impregnation with H3PO4
Semi-dried charge
Wet charge
Dried charge
Carbonization at 400-700°C either in a muffle
furnace or in a tube furnace
Rinsed in a Soxhlet’s apparatus with distilled water
Dried at 110°C in an oven overnight and stored the product
Figure 2.1
2.3.1
A flow diagram of activated carbon preparation.
Raw Material Preparation
Palm kernel (P) and coconut (C) shells were collected from local palm-oil
processing factory and market places respectively and then repeatedly washed with
tap water to remove dirt followed by drying in the sunlight. Coconut shells were
broken into small pieces with a hammer. The dried palm kernel and broken pieces
50
of coconut shells were crushed using a mechanical grinding machine and sieved to
the selected particle sizes of 1.18 - 2.36 and 0.60 - 1.18 mm. The smaller sizes (0.60
- 1.18 mm) were designed as fg (fine grains). After sieving and selecting, crushed
particles were washed with acids and the process was termed as pretreatment.
Pretreatment was conducted by soaking the raw material in aqueous solution of
either 30% sulphuric acid (-SW-) or 30% phosphoric acid (-PW-), for a period of 24
hours. Finally, the particles were thoroughly washed with distilled water in a bucket
to free acid and dried in the sunlight to eliminate moisture.
2.3.2
Activation and Washing
Prior to activation, the prepared raw materials were impregnated with 42.5%
aqueous solution of H3PO4 (-P-) in a wt. % ratio of precursor to undiluted acid 1:1
(100 g shell + 72 mL H2O + 72 mL 85% H3PO4) under varying impregnation
conditions. The physical state of impregnated feedstock, which was either wet by
the impregnating solution or some form of dry (e.g., apparently-dried/semi-dried, ad-, to completely-dried, -ed-) by the evaporation, before charging into the furnace,
were termed as charge state. Although all activations were carried out in a porcelain
casserole placed in a Carbolite muffle furnace, one of the activated carbons was
prepared from the activation in the presence of N2 flow in a quartz tube placed in the
tube furnace (TN). Residence time in the furnace was maintained at one hour for all
preparations. Different activation temperatures (such as 400, 450, 500 and 700°C)
were used to study the effect of temperature on pore development.
Since the products were obtained as lumps after activation, they were
subsequently broken into grains.
The obtained granular products were then
thoroughly rinsed in a Soxhlet’s apparatus for 2 weeks with distilled water to about
neutrality (litmus paper was used for testing).
The removal of all adhered
phosphates from the granule of the product was confirmed by adding a few drops of
30% solution of Pb(NO3)2 to the rinsed water. The presence of phosphate was
recognized from the white precipitate of Pb3(PO4)2. Finally the product was dried in
51
an oven at 110°C overnight and stored for subsequent characterization and
adsorption studies. The preparation variables are shown in Table 2.1.
Table 2.1 : Preparation variables of various activated carbons activated either in a
muffle (M) or in a tube furnace (T).
Product
code
Raw
material
& sizes
(mm)
Pretreatment
Impregnation
CPW-P500
Coconut
1.18-2.36
30%
H3PO4
H3PO4
(wetted)
CSW-P500
Coconut
1.18-2.36
30%
H2SO4
CR-P-500
Coconut
1.18-2.36
PSW-P500
Activation
temperature
°C
500
(M)
Soaking
time
h
20
Ramp rate
°C.min.-1
H3PO4
(wetted)
20
20
500
(M)
-
H3PO4
(wetted)
20
20
500
(M)
Palm
1.18-2.36
30%
H2SO4
H3PO4
(wetted)
02
20
500
(M)
PSW-Pad-500
Palm
1.18-2.36
30%
H2SO4
H3PO4
(semi-dried)
02
20
500
(M)
PSWfg-P450
Palm
0.60-1.18
30%
H2SO4
H3PO4
(wetted)
72
20
450
(M)
PSW-Ped-450
Palm
1.18-2.36
30%
H2SO4
H3PO4
(dried)
66
20
450
(M)
PSWfg-Ped-450TN
Palm
0.60-1.18
30%
H2SO4
H3PO4
(dried)
66
20
450
(T)
PSW-P400
Palm
1.18-2.36
30%
H2SO4
H3PO4
(wetted)
02
20
400
PSW-P700
Palm
1.18-2.36
30%
H2SO4
H3PO4
(wetted)
02
20
700
2.4
Chemical
activation
20
Modification of Activated Carbon
Three activated carbons, PSW-P-500, CPW-P-500 and PSW-P-ad-500, were
selected for subsequent studies as representatives of all prepared activated carbons
through preliminary test on metal cation uptake capacity. Similarly, these three
carbons were also chosen for further modification to the respective composites by
52
dispersing hydrated iron oxide (HFO) onto the porous surface. The dispersion of
HFO consisted of Fe(III) ion loading followed by conversion into HFO. Figure 2.2
shows the flow diagram of preparing various activated carbon-hydrated iron oxide
(C-HFO) composites from various activated carbons as starting materials and from
different alkaline solutions.
PSW-P-500/CPW-P-500/PSW-P-ad-500
2% FeCl3, 6H2O solution,
pH 1.8
Slurry: 2.0 g carbon in
200 mL of Fe3+ solution
Shaking rate/time: 160 rpm/4 days
Temperature: ambient
Saturated Ca(OH)2
solution, 200 mL
Fe3+ ion loaded carbon
5% NaOH + 5% NaCl
200 mL
Precipitation reaction with an aid of
agitation
Shaking rate/time: 100 rpm/3 h
Precipitation reaction with an aid of
agitation
Shaking rate/time: 100 rpm/3 h
Repeated washing with 50% aqueous
ethanol up to neutrality
Repeated washing with 50% aqueous
ethanol up to neutrality
C-HFOCa-1 (Base: PSW-P-500)
Figure 2.2
C-HFO-1 (Base: PSW-P-500)
C-HFO-2 (Base: CPW-P-500)
C-HFO-3 (Base: PSW-P-ad-500)
A flow diagram of modification of activated carbons.
The loading of Fe(III) ion onto the carbon surfaces was carried out in batch
adsorption technique at low pH, below 2, to avoid any precipitation. In each batch,
53
about 2.0 g of activated carbon was added to a 250 mL Erlenmeyer flask containing
200 mL of 2% Fe(III) ion (FeCl3,6H2O) solution, the pH of which was adjusted
earlier to 1.8.
The slurry of carbon in Fe(III) solution was then shaken placing the sealed
flask in an orbital shaker for 4 days at a rate of 160 revolutions per minute (rpm).
After the loading-up of Fe(III) ion onto the surfaces, the residual solution was
discarded followed by addition of an alkali as the precipitating agent. Precipitating
agents used in this modification were either a solution of 5% aqueous NaOH
incorporated with 5% NaCl or a saturated Ca(OH)2 solution. Precipitating solutions
were discarded after a treatment period of 3 h. Adsorbed Fe(III) ions were thus
entrapped and deposited, supposedly in the submicron scale, as Fe(OH)3
agglomerates/precipitates throughout the surface of porous carbon.
Adhered alkali solution was then removed by thoroughly washing with 50%
aqueous ethanol. The removal of alkali was confirmed by testing the washings with
red litmus paper. Then, aging technique was applied to convert the deposited and
agglomerated Fe(OH)3 precipitate into hydrated Fe(III)-oxide (HFO), in the
dispersed phase. The aging was done by heating the composite at 60°C in an oven
for 24 hours. Finally, activated carbon-hydrated Fe(III)-oxide composite (C-HFO)
adsorbents were stored for characterization and subsequent studies in the aspect of
application.
2.5
Characterization of Activated and Composite Carbon
Prepared activated and modified carbons were subjected to physical
characterization. The yield of the activated carbon obtained was also considered as
an important physical characteristic. The porous property of the prepared activated
carbons, composite carbons and a commercial activated carbon was analysed to
determine the surface area and porosity while surface morphology was studied to
observe developed images as replica of their porous surface.
54
Fourier transform infrared (FTIR) spectroscopy and Boehm’s titration
methods were used to identify and quantify the acidic surface groups on prepared
activated carbons. The pH drift method was used to determine the pH at zero point
charge (pHZPC) of the three selected activated carbons using NaCl solution as ionic
background while initial pH of the solution was adjusted to some pre-selected values
from 2 to 10. Proximate analysis of the activated carbon obtained was performed
using thermogravimetric analysis (TGA) technique that measures the mass of
material as a function of temperature in a controlled atmosphere of N2 and air. The
results were then reported as moisture, carbon and ash contents.
2.5.1
Yield Determination
The yield of the produced activated carbon was calculated from the following
equation (2.1).
Yield (%) =
W2
× 100 ,
W1
(2.1)
where, W1 and W2 are weights of the precursor and activated carbon respectively.
2.5.2
Surface Porosity Charaterization (N2 Adsorption-Desorption)
The porous structure of prepared and commercial activated carbons and
carbon-HFO composites was analyzed by N2 adsorption-desorption at 77K with
surface area analyser, ASAP 2010.
The Brunauer, Emmett and Teller (BET)
equation was used to calculate surface area. The single point total pore volume was
measured from the amount of nitrogen adsorbed at the relative pressure of 0.99.
55
2.5.3
Surface Morphology (SEM and EDX)
The surface morphology of prepared activated carbon samples and C-HFO
composites was investigated from the microhraphs of the porous structure by
scanning electron microscopy (SEM).
The SEM was carried out by the
bombardment of electrons of 30 KeV on target sample particle which was spread
earlier over an aluminum stub with the help of a doubled edged tape followed by
coating the surface with gold film. Electrons that are emitted from the specimen
with an energy of less than 50 eV are defined as secondary electrons and are used for
specimen investigation. Other than scanning electron microscopic investigation,
instrument also imparts the detection of scattered X-ray for the characteristic
radiation of a specific element in an energy dispersive system to identify the
element. The contents of P and Fe in the prepared activated carbon and the contents
of Fe and oxygen in C-HFO composites were determined using the energy dispersive
X-ray (EDX) analysis that is incorporated with SEM.
2.5.4
Fourier Transform Infrared Spectroscopy (FTIR)
FTIR spectroscopy in the wave number range of 4000-400 cm-1 and
potassium bromide (KBr) pellet technique were used to detect the surface functional
groups present in the activated carbon. About 1-3 mg of finely ground sample was
well mixed with about 200 mg of KBr powder. The mixture was then pressed
continuously at a pressure of 10 tonnes for 1 min. to form a transparent pellet using a
Perkin Elmer hydraulic press. During pellet preparation, the system was kept under
evacuation. The pellet was analysed immediately after being prepared.
2.5.5
Boehm’s Titration
The concentrations of various surface acidic groups were quantified using
Boehm’s titration method (cited by Strelko and Malik, 2002). As acidic functional
56
groups on carbon surface are varied in strength, bases of varying strength such as
NaHCO3, Na2CO3, NaOH and NaOC2H5 are used in Boehm’s titration. A solution
of concentration of 0.1N of each of the bases such as NaHCO3 (pK 6.37), Na2CO3
(pK 10.25), NaOH (pK 15.74) and NaOC2H5 was prepared by dissolving the
appropriate amount of each salt in 1 L volumetric flask followed by further dilution
to up to the mark using distilled deionized water (DDW).
Amounts of 0.2 g dried activated carbon were weighed into a series of 250
mL Erlenmeyer flasks and termed as sample flasks. Flasks which contained no
adsorbents were considered as blanks. Then 50 mL solution of various bases of
concentration 0.1 N was poured into a series of sample and blank flasks. Sample and
blank flasks were shaken at a rate of 150 rpm by placing in an orbital shaker for 3
days at room temperature. The supernatant solutions were filtered using Whatman
membrane filter of pore size 0.45-µm. Then 10 mL aliquots were titrated with 0.1 N
HCl using methyl orange (for NaHCO3 and Na2CO3) or phenolphthalein indicator
(for NaOH and NaOEt). The difference between HCl consumption by the blank and
samples were calculated and translated to mili-equivalent (meq) to quantify base
consumed by per gram of carbon for neutralization. Thus the consumed base in meq
was directly used in quantifying the acidity of activated carbon while the result was
conveniently expressed as meq H+ per gram of activated carbon.
2.5.6
Determination of pH at Point of Zero Charge (pHZPC)
A sodium chloride solution of 0.01M was prepared by dissolving an
appropriate amount (0.585 g) of NaCl salt in 1 L volumetric flask followed by
dilution to up to the mark using DDW. Then 50 mL of this solution was pipetted
each time and poured into several 250 mL Erlenmeyer flasks. Nitrogen gas was
bubbled through the solutions to prevent the absorption of CO2 from atmosphere and
the pH of the solutions were continuously measured. When the pH was found to be
stable, the pH of the solutions were adjusted to 2, 3, 4, 5, 6, 8 and 10 using diluted
solutions of HNO3 and NaOH and recorded as initial pH, pHi. About 0.1g of
activated carbon was added into each flask and sealed for 3 days with constant
57
agitation using an orbital shaker. The shaking rate was controlled at a rate of 150
revolutions per minute. At the end of experiment, the equilibrated solution was
decanted and the pH was measured as final pHf. A graph was constructed by
plotting the initial pH, pHi against equilibrated pH, pHf while pHZPC was determined
from the crossover point of pHi and pHf in the graph.
2.5.7
Thermogravimetric Analysis (TGA)
The thermogravimetric (TG) analysis was performed using a TG analyser,
Mettler TA 4000. About 10-20 mg of carbon sample was put into the platinum
crucible. Then the crucible was covered and placed in the chamber of the analyser.
After evacuation of the chamber, the analysis process began. The initial and final
temperatures were chosen as 40°C and 900°C while the heating rate was 20°C min-1.
To determine the moisture and ash contents, the samples were run in the
atmospheres of N2 and air while flow rate of each was controlled at 20 mL.min-1.
The moisture content was determined from the mass loss at 150°C in N2 while ash
content was determined from the mass loss in air at around 900°C. The carbon
content was then calculated using the following equation:
C (%) = 100% – [moisture content (%) + ash content (%)]
2.5.8
(2.2)
Leaching Test of Fe
Since the modification of carbon was performed through the dispersion of
HFO, it is obligatory to assure that treated water would not be contaminated from the
leaching of Fe used as modifier. An amount of 0.1 g of C-HFO-1 composite was
subjected to shake with 50 mL of DDW in an Erlenmeyer flask at a rate of 160 rpm
for 3 days while the pH of water was adjusted to slightly acidic as 4.5 using 0.005 M
HCl. Then the contacted water was separated from the adsorbent by decanting and
tested for any dissolute Fe content using AAS.
58
2.6
Adsorption Studies
All adsorption experiments were carried out in batch and were triplicated to
obtain an average result of adsorption studies. A preliminary adsorption screening
test was conducted using a solution of Cu(II) to limit the number of adsorbents, to be
studied in adsorption experiments, to a few, as many activated carbons were
prepared (see Table 2.1) from different precursors under different conditions. The
single point equilibrium data were obtained for this purpose using an initial
concentration of < 300 mg.L-1 while the initial pH was adjusted to about 6.
Adsorptions of single solute of Cu(II), Ni(II), Zn(II) and Pb(II), onto
prepared and commercial activated carbons were carried out and the obtained
equilibrium data were fitted to various equilibrium models. Adsorption of single
solute of Cr(VI) was carried out using various activated and composite carbons to
evaluate the relative suitability of the prepared adsorbents and to optimize the pH in
removing the anionic species. Then, the performance of the best selected activated
carbon was evaluated and compared with a commercial one at an optimum pH.
Adsorptions of binary solute of Cu(II) and Ca(II), and ternary solute of Cu(II), Ni(II)
and Pb(II) were carried out by the best selected activated carbon to predict the
mechanism and correlate the relative affinity for developing a selectivity order over
a wide range of concentration. Adsorptions of As(III) and As(V) were conducted
onto the samples of a carbon-HFO composite and its non-modified counter activated
carbon. The performance of the adsorbent was evaluated from the single point
equilibrium data. The adsorption dynamics data of Cu(II), Ni(II), Pb(II) and Cr(VI)
were obtained by the best selected activated carbon.
2.6.1
Test and Standard Metal Ion Solutions
Stock solution of Cu(II), about 1000 mg.L-1, was prepared by dissolving
about 2.683 g of copper (II) chloride dihydrate in a 1000 mL volumetric flask with a
few hundreds mL of DDW followed by further dilution to up to the mark. Stock
solution of concentration of about 1000 mg.L-1 of each of the Ni(II), Zn(II), Pb(II)
59
and Cr(VI) was prepared by dissolving appropriate amount of ones salt, namely
nickel (II) nitrate hexahydrate (actually transferred 4.980 g instead of 4.953 g),
zinc(II) nitrate hexahydrate (4.55 g), lead (II) nitrate (1.606 g) and potassium
dichromate (2.83 g) respectively, followed by dilution to 1000 mL using DDW.
Various concentrations of test solutions of single Cu(II), Ni(II), Zn(II), Pb(II) and
Cr(VI) were prepared by subsequent dilution of the respective stock solution using
DDW.
Stock solution of binary solute of Cu(II) and Ca(II), concentration of each
solute is of about 1000 mg.L-1, was prepared by dissolving a mixture of 2.683 g of
copper (II) chloride dihydrate and 5.892 g of calcium (II) nitrate tetrahydrate in a
1000 mL volumetric flask followed by dilution to up to the mark. Test solutions of
binary solute of different concentrations were prepared by subsequent dilution of the
binary stock solution.
Stock solution of ternary solute of Cu(II), Ni(II), and Pb(II) were prepared by
dissolving a mixture of 2.683 g copper (II) chloride dihydrate, 4.95 g nickel (II)
nitrate hexahydrate and 1.606 g lead (II) nitrate in a 1000 mL volumetric flask
followed by further dilution to up to the mark while ternary test solutions were
prepared through proper dilution of the stock solution. The concentration of each
solute in the stock solution was to be 1000 mg.L-1.
The initial pH of test solutions were adjusted to selected values using 0.005M
HNO3 and 0.01M NaOH. All standard solutions of Cu(II), Ca(II), Ni(II), Zn(II) and
Pb(II) were prepared from the dilution of the respective standard solution (1000
mg.L-1) using DDW acidified earlier with 0.2% nitric acid. The standard solution of
Cr(VI) was prepared from the oven-dried (110°C) salt of potassium dichromate. 1,5Diphenylcarbazide (DPC) solution was prepared, prior to analysis of Cr(VI) by UVVis spectrophotometer, by dissolving 0.25 g in 50 mL acetone.
60
2.6.2
Adsorption Equilibrium Studies: Metal Cations
All batch adsorption experiments were carried out at room temperature using
0.1 g dried adsorbent of various activated carbon added to 50 mL metal ion solution
in a polypropylene centrifuge tube.
That is, adsorbent concentration was kept
constant at 2 g.L-1. Shaking was applied placing the sealed tubes in an orbital shaker
with a constant shaking rate of 160 rpm. At the end of adsorption experiment,
centrifuge tubes were removed from the shaker and the solutions were separated
from the adsorbent by decanting. Then proper dilutions were done, prior to analysis
of the samples by FAAS, using DDW acidified earlier with 0.2% HNO3. The
adsorption equilibrium data were obtained by varying initial metal ion
concentrations while the mass of activated carbon, contact time, shaking rate and
initial pH were kept constant. Adsorption was equilibrated in 3 days contact time.
The effect of initial pH on adsorption was studied varying the pH from 3 to 5.
2.6.2.1 Selection of Prepared Activated Carbons for Metal Cations
All prepared activated carbons were grouped into five series depending on
the developed surface area (> 1000 m2.g-1) and preparation variables. Then six
activated carbons (series I: PSW-P-500; series II: CPW-P-500; series III: PSW-P-ad500, PSW-P-ed-450; series IV: PSWfg-P-ed-450-TN; series V: PSWfg-P-450) were
selected as representatives of all prepared activated carbons. Six activated carbons
were scrutinized from a preliminary adsorption screening test to limit the number of
adsorbents, to be studied in adsorption experiments, to three, as many activated
carbons were prepared from different precursors under different conditions. In the
adsorption screening test, Cu(II) was used as adsorbate while the potentials of
activated carbons were compared from the Cu(II) uptake capacities.
The equilibrium data were obtained for the adsorption in a single point initial
concentration of Cu(II) as < 300 mg.L-1 while the initial pH was adjusted to about 6.
That is, all adsorption experimental parameters, namely adsorbate concentration,
adsorbent concentration (2 g.L-1), shaking rate (160 rpm), contact time (72 h) and
61
initial pH were kept constant in order to determine the uptake capacities of various
activated carbons.
2.6.2.2 Single Solute Adsorption
Adsorption isotherms of Cu(II) by various activated carbons including
commercial ones were determined from single solute solution at two different initial
pH, 3 ± 0.1 and 5 ± 0.1, while changing the initial concentrations from < 5 to < 125
mg.L-1.
Adsorption isotherm of Cu(II) was also constructed for changing the initial
concentrations from < 25 to < 700 mg.L-1 by the best selected activated carbon at an
initial pH 4.5 ± 0.1. The initial pH was adjusted to 4.5 instead of 5 to avoid any
precipitation in very high concentration of Cu(II) like 700 mg.L-1.
Adsorption isotherms of Ni(II) by various activated carbon were determined
at an initial pH 5 ± 0.1 while the initial concentrations of Ni(II) were varied from 6.1
± 0.1 to 227 ± 3 mg.L-1.
Adsorption equilibrium data of Zn(II) by various activated carbons were
obtained by varying the initial concentrations from 19.0 ± 0.2 to 310 ± 1 mg.L-1
while initial pH was adjusted to 5 ± 0.1.
Adsorption isotherms of Pb(II) by various activated carbons were determined
for the initial pH 5 ± 0.1 and by the best selected activated carbon at an initial pH 3 ±
0.1, while initial concentrations were varied as ≤ 50 to < 600 mg.L-1.
62
2.6.2.3 Binary Solute Adsorption
The effect of the competing ion, Ca(II) on the adsorption of Cu(II) was
studied from the adsorption in binary solute solutions of Cu(II) and Ca(II). The
initial concentrations of each of the cations in solution was varied from < 25 to < 120
mg.L-1 (Cu(II): 23.6 ± 0.1 – 117.5 ± 0.7 mg.L-1; Ca(II): 23.8 ± 0.2 – 120 ± 1 mg.L-1)
while initial pH was adjusted to 5 ± 0.1. To study the effect of pH on binary
adsorption, adsorption equilibrium data were also obtained for the initial pH 3 ± 0.1,
while initial concentrations were varied as, Cu(II): 26 ± 1 to 151 ± 0 and Ca(II): 20.4
± 0.7 to 109 ± 3 mg.L-1.
2.6.2.4 Ternary Solute Adsorption
The competitive effect among borderline and soft cations was studied from
the adsorption in a solution containing ternary solute of Cu(II), Ni(II) and Pb(II).
The initial solution concentrations of each of the cations were varied in the ternary
solution from < 25 to ≤ 600 mgL-1 (Cu(II): 23.0 ± 0.2 to 568 ± 2; Ni(II): 22.2 ± 0.1
to < 600; Pb(II): 22.3 ± 0.1 to 600 ± 5 mg.L-1) and initial pH were kept constant at 5
± 0.1. Here it is important to note that, observing a decreasing tendency of Ni(II)
uptake capacity with the increase of initial concentration, adsorption equilibrium
data of Ni(II) were analysed for the initial concentration range, 22.2 ± 0.1 to 305 ± 2
mg.L-1 instead of 22.2 ± 0.1 to < 600.
2.6.2.5 Reuse of Cation-Exhausted Activated Carbon
Once adsorbed Cu(II) on the best selected carbon, PSW-P-ad-500 in the
initial concentration range of < 25 to < 700 mg.L-1 at an initial pH 4.5 was subjected
to regeneration for reuse. After decanted off the equilibrated solution, the exhausted
carbon in each batch was shaken with 25 mL of DDW for a few minutes followed by
separating the adsorbent from the washings. This washing experiment was repeated
63
twice. Thereafter, the adsorbent was dried in an oven at 110°C and cooled at room
temperature.
Then, 50 mL of 10% HCl was added to each of the adsorbent
contained in a centrifuge tube. The centrifuge tubes were caped and placed in a
shaker. Shaking was applied at a rate of 160 rpm with a contact time of 3 days.
After removing the tubes from shaker, solutions were decanted and diluted to the
analysis concentration range of Cu(II) by FAAS and analysed.
2.6.3
Adsorption Equilibrium Studies: Cr(VI)
Batch adsorption experiments of Cr(VI) were carried out at room temperature
using 0.05 g dried adsorbent added to 25 mL metal ion solution in a polypropylene
centrifuge tube. The adsorbent concentration was kept constant at 2 g.L-1. Shaking
was controlled at a constant rate of 160 rpm. For all equilibrium studies, contact
time was maintained 72 h.
The suitability of the various adsorbents namely, prepared, commercial and
modified (C-HFO composite) carbons were compared from equilibrium data
obtained at single point initial concentration (< 10 mg.L-1), while the initial pH was
obtained without adjustment of pH as 4.79.
The effect of pH on the adsorption of oxyanions of Cr(VI) onto the selected
adsorbents was studied by varying pH from 3 to 8 while initial concentrations were
kept constant at about 40 mg.L-1. This experiment seeks to optimize the adsorption
pH.
Finally, the adsorption equilibrium data of Cr(VI) were obtained, by CAC
and one of the acid activated carbons (PSW-P-ad-500) as selecting best, varying
initial concentrations as ≤ 40 to 194 ± 2 mg.L-1 and keeping initial pH constant at 3 ±
0.1.
64
2.6.4
Adsorption Equilibrium Studies: As(III) and As(V)
The stock solution of As(III) of concentration of about 500 mg.L-1 was
prepared from As2O3 by dissolving the amount of 0.132 g with 10 mL of 2 M NaOH
in 100 mL volumetric flask followed by further dilution to up to the mark. The pH
of the solution was adjusted to 8 using 0.5 M HNO3 and 0.005 M NaOH solution just
before dilution to up to the mark.
Test solution of As(III) was prepared by
subsequent dilution of the stock solution to about 2 mg.L-1 while pH was adjusted to
about neutral (pH 7.6) using diluted (0.005 M) HNO3 or NaOH. The stock solution
of As(V) of concentration of about 500 mg.L-1 was prepared by dissolving 0.1534 g
of As2O5 in a 100 mL volumetric flask with 10 mL of 2 M NaOH followed by
further dilution to up to the mark. The pH of the stock solution was adjusted to 5
using 0.5 M HNO3 just before dilution to up to the mark. Test solution of As(V) was
prepared by subsequent dilution of the stock solution to about 2 mg.L-1 while pH of
the solution was adjusted to 5.5 using diluted (0.005 M) HNO3 and NaOH just
before dilution to up to the mark.
Adsorbents, C-HFO-1 and PSW-P-500 were used in the preliminary arsenic,
As(III) and As(V), adsorption studies, considering them as typical of prepared
composites and activated carbons respectively. Adsorption experiments were carried
out by the composite carbon, C-HFO-1 using 0.2 g of the adsorbent added to 200 mL
of respective As (III or V) solution in 250 mL Erlenmeyer flask. Adsorptions of
As(III) and As(V) by the activated carbon, PSW-P-500 were carried out by adding
0.1 g adsorbent to 50 mL of respective As solution in 100 mL conical flasks. That
is, adsorbent concentrations were maintained as 1 g.L-1 by C-HFO-1 and 2 g.L-1 by
the activated carbon. Shaking was applied by an orbital shaker at a rate of 160 rpm.
Adsorption was equilibrated in 3 days contact time and after the end of experiment,
the solution was separated from the adsorbent by decanting into another flask. Then
proper dilutions were done using DDW acidified earlier with 2% HNO3 prior to
analysis the sample with ICP-MS. Standard solutions of As were prepared from the
proper dilution of the standard solution of As(III) (1000 mg.L-1) using DDW. Before
dilution of the standard solution to 100 mL in a volumetric flask, 2 mL of
concentrated HNO3 was added so that the diluted As solution contained 2% HNO3.
65
2.6.5
Adsorption Dynamics: Cu(II), Ni(II), Pb(II) and Cr(VI)
For adsorption dynamic/kinetic study, the adsorption was determined as a
function of time while adsorbent concentration, adsorbate concentration, shaking rate
and the initial solution pH were kept constant. Kinetic data were obtained by
varying contact time from 0.5 to 72 h while initial concentrations of Cu(II), Ni(II),
Pb(II) and Cr(VI) were maintained as < 49, < 45, 95 and 77 mg.L-1. The initial pH
was kept constant at 5 for Cu(II), Ni(II) and Pb(II) and at 3 for Cr(VI). Adsorbent
concentration was kept constant at 2 g.L-1. That is alike adsorption equilibrium
studies, 0.1 g of adsorbent was added to 50 mL solution of Cu(II), Ni(II) and Pb(II)
while 0.05 g of adsorbent to 25 mL of Cr(VI) solution.
2.6.6 Determination and Uptake of Species of Metal/Metalloid
The concentration of cations of Cu(II), Ni(II), Zn(II), Pb(II), and Ca(II) in the
solution before and after adsorption was determined in air-acetylene flame using
flame atomic absorption spectrophotometer (FAAS) of either of GBC, model Avanta
or Perkin Elmer, model AAnalyst 400. The concentration of any Fe content in the
leached solution of C-HFO composite was also examined using FAAS.
The
operating parameters for the determination of metal ions by FAAS are shown in
Table 2.2. The concentrations of the diluted solution of As(III) and As(V), before
and after adsorption, were determined using ICP-MS.
The solution of Cr(VI) before and after adsorption was diluted using 1 M
H2SO4 to obtain the pH of the diluted solution 1 (concentrated H2SO4 was used to
adjust the pH of undiluted samples) and then added appropriate amount of 1,5diphenylcarbazide solution (0.5 mL DPC solution per 25 mL diluted metal ion
solution) (Ahmed 2002). After development of colour, due to the formation of
diphenylcarbazone-chromium(II) complex, within 10 minutes, the solution was
analyzed and determined the concentration using UV-Vis spectrophotometer.
66
Table 2.2 : The operating parameters of FAAS in the determination of metal
cations.
Analyte
Operating parameters of instrument (GBC-Avanta)
Hollow Cathode Lamp
Wavelength
Slit width
Current, mA
Energy
nm
nm
Cu
3.0
-
324.8
0.5
Ni
5.0
-
232.0
0.2
Ca
5.0
-
422.7
0.5
Fe
6.0
-
248.4
0.2
Zn
4.0
-
213.9
0.5
Analyte
Operating parameters of instrument (Perkin Elmer)
Hollow Cathode Lamp
Wavelength
Slit width
Current, mA
Energy
nm
nm
Cu
15
88
324.75
2.7/0.8
Cu
15
53
216.51
1.8/1.35
Ca
10
57
422.67
2.7/0.6
Ni
25
65
232.0
1.8/1.35
Pb
10
75
283.31
2.7/1.05
The uptake of metal ion on activated carbon was calculated from the
concentration difference before and after adsorption. The following equation (2.3)
was used to calculate the metal uptake in mg by the per unit mass of adsorbent.
q=
(C 0 − C t )V
1000m
(2.3)
where, q = metal uptake mg.g-1-adsorbent; C0 = initial concentration, mg.L-1; Ct =
concentration, at any time (t), mg.L-1; V= volume of solution in a batch, mL; and m
= mass of adsorbent used in a batch, g.
CHAPTER 3
RESULTS AND DISCUSSION : ADSORBENT DEVELOPMENT
3.1
Preparation
Various activated carbons had been successfully prepared with the
optimization of the preparation variables through studying the effect of precursor
type and size, and of process variables such as pretreatment, impregnation condition,
and activation temperature.
Modification of the prepared activated carbon into
carbon-HFO composite had also been successfully done which was evident from its
high iron content in the dispersed phase.
3.1.1
Activated Carbon
In this work, a 30 % solution of H2SO4 was used in the pretreatment to
enhance surface acidity and porosity. To maximize the development of internal
surface area as well as to enhance the development of various acidic surface groups,
pretreatment was also explored with 30 % H3PO4. Guo and Lua (1999) studied the
effect of pretreatment on oil-palm stones using various concentrations, 5-30 %, of
H2SO4 and KOH. Their results showed that acidic groups such as phenolic and
carboxylic groups were well developed on the surface of carbon obtained from the
precursors pretreated with H2SO4. On the contrary, KOH pretreatment caused the
development of alkaline groups. They further observed that the pore volume also
68
increased with the increase of pretreatment acid concentration. Some literatures
showed that activated carbons prepared from lignocellulosic precursor with H3PO4
and air have very high metal ion adsorption capacity owing to the presence of
oxygen and phosphorus bearing acidic groups (Toles and Marshall, 2002; Dastgheib
and Rockstraw, 2001; Toles, Marshall and Johns, 1999; Johns, Marshall and Toles,
1998). The chamber of the Carbolite muffle furnace is not fully air tight according
to the product specification and thus furnishes some air inside, by design default,
through leakage. To enhance metal ion uptake capacity by concurrent
activation/oxidation, acid pretreated precursors were impregnated with H3PO4 acid
and activated in the controlled oxygenated atmosphere of Carbolite muffle furnace.
3.1.2
Carbon-Hydrated Iron Oxide (C-HFO) Composite
Various modified carbons, C-HFO composites, depending on different
starting materials (acid activated carbons) and metal hydroxide precipitating agents,
were successfully prepared. The dispersion of hydrated iron-oxide onto the activated
carbon was confirmed from the higher level of iron and oxygen content in the
composite. The contents of Fe, oxygen, Na or Ca and C in the composites and Fe
and C in the parent activated carbon were determined using EDX (in weight %), and
the results are shown in Table 3.1. The EDX spectrums are shown in Appendix B.
It is known that both forms of HFO, namely amorphous (FeOOH) and
crystalline (goethite, α-FeOOH and hematite, α-Fe2O3), have strong adsorption
affinity toward both As(III) and As(V) oxyacids and oxyanions through ligand
exchange in the coordination spheres of structural Fe atoms (DeMarco et al., 2003).
This is why, simple aging technique was used without caring for the formation of
crystalline phase in converting and dispersing deposited Fe(OH)3 into HFO. Since
NaOH is a strong base, NaCl is added to the precipitating agent of NaOH to control
the ionic dissociation through common ion effect and thus to oppress the formation
of Fe(III)-hydroxide precipitate. Otherwise uncontrolled precipitation might result
in a severe destruction of the porous structure by clogging the pores.
69
Table 3.1 : The contents of Fe, oxygen, Na/Ca and C in the prepared composites
and in their parent activated carbons.
Adsorbent
Elements (wt. %)
Fe
Oxygen
Na
Ca
Carbon
C-HFO-1
6.9 ± 0.8
17 ± 2
7±2
-
68 ± 2
C-HFOCa-1
13 ± 1
19 ± 2
-
8±1
57 ± 7
PSW-P-500
0.27 ± 0.06
12 ± 2
-
-
82.8 ± 0.9
C-HFO-2
14.1 ± 0.9
18 ± 1
10 ± 1
-
59 ± 3
CPW-P-500
0.18 ± 0.09
15 ± 3
-
-
79 ± 4
C-HFO-3
19 ± 1
15 ± 2
7±1
-
61 ± 3
PSW-P-ad-500
0.5 ± 0.3
10 ± 1
-
85.3 ± 0.3
Saturated solution of Ca(OH)2 was used as the precipitating agent to observe
the effect of mild base in the formation of Fe(OH)3 without much destruction of
pores. In addition the commercial aspect was also considered from its cheaper price
than that of NaOH. Ethanol was used in washing solution (50 % aqueous) to reduce
the dielectric constant of water and thus supposedly to enhance the agglomeration of
HFO particles through suppression of surface charge (DeMarco et al., 2003). Since
the agglomeration occurred inside the pores of carbon, it was supposed that the
particles of HFO would remain in the submicron size. The formation of HFO from
the mixing of Fe(III)-salt and alkali in aqueous phase can be shown as:
Fe3+ + 3OH
Fe(OH)3
3.2
Fe(OH)3
60°C
FeOOH + H2O
(3.1)
(3.2)
Characterization of Prepared Activated Carbon
Prepared activated carbons were characterized and evaluated for the
development of surface area, porosity, surface functional groups and acidity using
various methods as described in Chapter 2.
The contents of C and P were
70
determined from EDX while surface morphology was analysed using SEM. The
yield was determined as a process characteristic. A proximate analysis was also
performed to estimate the moisture, carbon and ash contents in the product carbon.
3.2.1
Surface Porosity and Morphology
Surface porosity and morphology of the prepared activated carbons were
studied to optimize the preparation variables for high surface area activated carbon.
The BET surface area, average pore width, total pore volume and pore-types of
various activated carbons including a commercial one (CAC) were obtained from N2
adsorption-desorption data and isotherm’s characteristic (see Appendix C and Figure
3.1). The results including yields are then shown in Table 3.2.
PSW-P-ad-500
PSWfg-P-450
PSW-P-500
PSW-P-400
700
3
Volume adsorbed, cm /g
600
CPW-P-500
PSW-P-ed-450
CSW-P-500
PSW-P-700
500
400
300
200
100
0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Relative pressure, p/p0
Figure 3.1
N2 adsorption-desorption isotherms of various activated carbons
prepared from oil palm and coconut shell precursors under various process
conditions.
71
Table 3.2 : Yield, BET surface area, average pore width, total pore volume and
pore-types of various activated carbons.
Activated
Yield
SBET
Wp
Vt
Pore-types
carbon
%
m2.g-1
nm
cm3.g-1
(Qualitative assessment)
micro
meso
macro
CR-P-500
37
1049
2.39
0.63
major
minor
-
CSW-P-500
45
1053
2.32
0.61
major
minor
-
CPW-P-500
41
1491
2.52
0.94
major
minor
-
PSW-P-500
44
1049
2.32
0.61
major
minor
-
PSW-P-ad-500
46
1476
2.17
0.80
major
minor
-
PSW-P-ed-450
50
1366
2.19
0.75
major
minor
-
PSWfg-P-450
45
1472
1.96
0.720
all
-
-
PSW-P-400
48
593
2.07
0.31
major
minor
PSW-P-700
20
734
1.95
0.36
all
-
-
CAC
-
1320
2.04
0.67
major
minor
-
SBET = BET surface area; Wp = average pore width; Vt = total pore volume.
It is observed from Table 3.2 that some of the prepared activated carbons,
especially those were obtained from preheat treated feedstock (apparently/semidried, “-ad-” to dried charge, “-ed-”), have higher BET surface area and total pore
volume than that of commercial ones. Since H3PO4 was used in the activation step,
all prepared activated carbons contained a certain amount of phosphorous. The
content of P in some selected activated carbons was determined using EDX and
presented in Table 3.3 (see spectrums in Appendix B).
Table 3.3 : Contents of phosphorous (P) in some selected activated carbons as
determined by EDX.
Activated carbon
P (wt. %)
PSW-P-500
4.5 ± 0.6
CPW-P-500
6±1
PSW-P-ad-500
3.8 ± 0.5
72
3.2.1.1 Effect of Precursor Type
Oil palm and coconut shells derived activated carbons, PSW-P-500 and
CSW-P-500 prepared under similar preparation conditions, showed similar N2
adsorption-desorption isotherms (Figure 3.2). Since both isotherms are of type-1
with sharp knee, they reveal the presence of major microporosity while the presence
of a narrow hysteresis loop within the relative pressure range of 0.4 to 0.98 (type IV
isotherm) discloses the coexistence of mesopores in minor amount.
The BET
surface area and total pore volume of both of activated carbons are nearly the same
(ca.1050 m2.g-1 and 0.61 cm3.g-1 respectively). Therefore, it was speculated that
H3PO4 has similar activating power towards oil palm and coconut shell precursors.
This finding is in agreement with that of other researchers namely Toles, Marshall
and Johns (2000). They developed a single process of preparing activated carbon
using H3PO4 from a group of lignocellulosic nutshell materials which consisted of
almond, pecan, walnut, hazelnut, macadamia nut, coconut and pistachio precursors.
But they did not mention the effect of specific precursor on the development of
various pores and surface area.
3.2.1.2 Effect of Pretreatment
Effects of acid pretreatment on the development of surface area and porosity
were studied from the activated carbons of the non-pretreated and acid pretreated
(either of H2SO4 or H3PO4) precursors. Acid pretreatment enhances processes which
prevent the evolution of large quantities of volatile substances in the activation
process, as a result of which active carbon granules of high mechanical strength are
obtained (Jankowska, et al., 1991). As oil palm and coconut shells contain higher
level of volatile matter, 76.3 and 72.8 wt. % respectively, pretreatment plays an
important role especially in controlling the mechanical strength, yield achievement
and pore development. The effect of pretreatment on the evolution of volatile matter
is shown from the yield of the product in Figure 3.3. It is perceived that H2SO4 acid
controls volatile evolution most yielding a very high yield. Non-pretreated precursor
restricts volatile evolution in a lesser extent resulting relatively a lower yield. All
73
three products, CR-P-500 (non-pretreated), CSW-P-500 (H2SO4 pretreated) and
CPW-P-500 (H3PO4 pretreated) are highly microporous with a small amount of
mesopores (see Table 3.2) as are evident from the isotherms in Figure 3.1.
A
400
3
Volume adsorbed, cm /g
500
300
PSW-P-500
200
100
0
0
0.2
0.4
0.6
0.8
1
0.8
1
Relative pressure, p/p0
B
400
3
Volume adsorbed, cm /g
500
300
CSW-P-500
200
100
0
0
0.2
0.4
0.6
Relative pressure, p/p0
Figure 3.2
N2 adsorption-desorption isotherms of activated carbons derived from
oil palm (A) and coconut shell (B) precursors.
74
45
40
Yield %
35
30
25
20
15
10
5
0
CR-P-500
(NP)
CSW-P-500
(SAP)
CPW-P-500
(PAP)
Activated carbons
Figure 3.3
Yield of various activated carbons as a function of precursor
pretreatment such as non-pretreated (NP), H2SO4 pretreated (SAP) and H3PO4
pretreated (PAP) conditions.
The BET surface areas of these three activated carbons, (CR-P-500, CSW-P500 and CPW-P-500), are shown in Figure 3.4. Various acids such as H2SO4 and
H3PO4 were used in the pretreatment to study the effect of a particular acid on pore
development in the resulting carbons. The use of H3PO4 in pretreatment was a
variation in the commonly used pre-treating agent, H2SO4. Phosphoric acid restricts
the process of volatile evolution to a lesser extent than that by H2SO4. It was evident
from the relatively lower yield, very high BET surface area and wider pore width of
a carbon, CPW-P-500, obtained from H3PO4 pretreatment (see Figures 3.3 and 3.4,
and Table 3.2). It was thus indirectly assumed that H2SO4 pretreatment enhances the
mechanical strength of the particles and yield of the resulting carbon.
Pretreatment of raw material with H3PO4 might extend the impregnation step
of H3PO4 activation resulting in high BET surface area, nearly 1500 m2.g-1 and high
total pore volume (0.94 cm3.g-1). Nearly identical BET surface area (ca 1050 m2.g-1)
and total pore-volume (0.6 cm3.g-1) of carbon samples obtained from non-pretreated
75
and H2SO4 pretreated precursors reveal that H2SO4 has no effect in the development
of pores and surface area which is contradictory to the findings of Guo and Lua
(1999). The reason of contradiction might be due to the fact that they prepared
activated carbon using physical activation and as a consequence, acid used in
pretreatment also took part in activation reaction with the precursor particles.
2
BET surface area, m /g
1500
1200
900
600
300
0
CR-P-500
(NP)
CSW-P-500 CPW-P-500
(SAP)
(PAP)
Activated carbons
Figure 3.4
BET surface area of various activated carbons as a function of
precursor pretreatment such as non-pretreated (NP), H2SO4 pretreated (SAP) and
H3PO4 pretreated (PAP) conditions.
In this study, since H3PO4 was used in activation, the activation effect of
H2SO4 on precursor in the pretreatment step was negligible in comparison to that of
H3PO4 in the impregnation step. That was why when the powerful activating agent
like H3PO4 was used in the pretreatment, it activated the particles to a remarkable
extent resulting relatively very high surface area of an activated carbon, CPW-P-500
(H3PO4 pretreated) than that of H2SO4 pretreated one (CSW-P-500). On the other
hand, H2SO4 pretreatment showed no effect in the development of pores and surface
area especially while the precursor was activated with a powerful activating agent,
such as H3PO4.
76
3.2.1.3 Effect of Impregnation Condition
Activated carbons (CSW-P-500 and PSW-P-500) prepared from wet
feedstock of coconut and oil palm shells at activation temperature of 500°C have the
BET surface area of around 1050 m2.g-1. Activated carbons, PSW-P-500 and CSWP-500 were prepared by impregnating (contacting with H3PO4) the precursor at room
temperature followed by activation of the wet charge i.e., wet-feedstock.
As
impregnation temperature and time can improve the activation process, attempts
were made to study the effect of impregnation conditions and charge state on the
development of surface area in palm shell derived carbon. Thus activated carbon
PSW-P-ad-500 was prepared from the apparently dried charge that was impregnated
with H3PO4 for 2 h in wet condition at room temperature before evaporation.
Another activated carbon, PSW-P-ed-450 was prepared from the completely dried
feedstock. Prior to evaporation of aqueous solvent, impregnated precursor was kept
at room temperature in the wet condition for 66 h.
All activated carbons prepared from preheat treated (evaporated to some
forms of dryness) feedstock have relatively very high surface area than that prepared
from the wet feedstock (see Table 3.2).
It was speculated that heating the
impregnation solution to evaporation/boiling, activated all the particles uniformly
and rapidly which reduced the impregnation hour. On the other hand, wet charge
might experience lack of activation.
A very high surface area (1476 m2.g-1) was observed in one of the carbons,
PSW-P-ad-500, derived from the apparently-dried feedstock (evaporated at 110°C
for 4 h). Relatively lower BET surface area (1366 m-2.g-1) of PSW-P-ed-450, than
that of PSW-P-ad-500, might be resulted from its lower activation temperature rather
than the effect of charge state. Molina-Sabio et al. (1995) pointed out that complete
dryness intensifies the hydrolysis of particles by converting more phosphoric acid to
polyphosphoric acids which are stronger acids. Further study is needed to conclude
whether apparently dried charge produces higher BET surface area than completely
dried charge. Development of BET surface area and total pore volume by oil palm
shell activated carbons obtained from wet, apparently-dried and completely dried
feedstock are shown in Figures 3.5 and 3.6. The SEM micrograph of PSW-P-ad-500
77
is shown in Figure 3.7, which also shows well development of pores inside external
pores.
BETsurface area, m 2/g
1500
1200
900
600
300
0
Wet
Semi-dried
Dried
Charge states of activated carbons
Figure 3.5
BET surface areas of various activated carbons derived from different
charge states such as wet (PSW-P-500), apparently/semi-dried (PSW-P-ad-500) and
completely dried (PSW-P-ed-450).
Total pore volume, cm 3/g
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Wet
Semi-dried
Dried
Charge states of activated carbons
Figure 3.6
Total pore volumes of activated carbons derived from different charge
states, as wet (PSW-P-500), semi-dried (PSW-P-ad-500) and dried (PSW-P-ed-450).
78
Figure 3.7
SEM micrograph of an activated carbon obtained from semi-dried
feedstock of oil palm shell precursor, namely PSW-P-ad-500.
3.2.1.4 Effect of Precursor Size
To study the effect of precursor’s particle size on the development of surface
area and pores, an activated carbon, PSWfg-P-450, was prepared from smaller
particle sizes, 1.18-0.60 mm and compared with another carbon, PSW-P-ed-450
prepared from the same precursor (oil palm shells) but different in particle sizes as
2.36-1.18 mm. Nitrogen adsorption-desorption isotherms of both carbons, PSWfgP-450 and PSW-P-ed-450, are shown in Figure 3.8.
The isotherm of PSWfg-P-450 showed very high N2 adsorption at relatively
low pressure and a single line of adsorption-desorption isotherms over the entire
pressure range.
This pattern of isotherm thus indicates a total contribution of
micropores while the other, PSW-P-ed-450, (derived from the larger particle sizes
precursor) contains major amounts of micropores with little mesopores.
The
79
contribution of BET surface area (1472 m2.g-1) was also observed higher by the
smaller particle sizes precursor (see Table 3.2).
The preparation condition of
PSWfg-P-450 and PSW-P-ed-450 were similar except in charge state. The former
was obtained from the wet charge while the later from the dried charge. Although
the higher activation was expected from the dried charge, even then the high BET
surface areas of PSWfg-P-450 (wet charge derived carbon), evident the significant
contribution of particle sizes in pore/micropore development. That means a better
contact with H3PO4 had occurred while particle sizes were smaller.
This is
consistent with the fact that lowering of particle sizes increased its surface area and
thus ultimately favoured a better contact of H3PO4 with more exposed area, which in
turn contributed to a higher microporosity development.
550
500
400
3
Volume adsorbed, cm /g
450
350
300
PSWfg-P-450
250
PSW-P-ed-450
200
150
100
50
0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Relative pressure, p/p0
Figure 3.8
N2 adsorption-desorption isotherms of two activated carbons, derived
from smaller (1.18-0.6 mm) and larger (2.36-1.18 mm) particle sizes, namely
PSWfg-P-450 and PSW-P-ed-450.
80
3.2.1.5 Effect of Activation Temperature
The highly developed specific surface areas (ca.1500 m2.g-1) are observed in
activated carbons obtained from preheat treated feedstock and at the same time,
those obtained from the activation at temperatures of 450-500°C (see Table 3.2).
This finding is noticeably described in Figure 3.9, which shows the development of
surface area as a function of activation temperature.
BET surface area
1500
1250
1000
750
o
2
Temperarure, C/Surface area, m /g
activation temperature
500
250
PSW-P-700
PSWfg-P-450
PSW-P-ed-450
PSW-P-ad-500
PSW-P-500
CPW-P-500
CSW-P-500
CR-P-500
PSW-P-400
0
Various activated carbons
Figure 3.9
The development of surface areas of various activated carbons
prepared under different activation temperatures.
Laine et al. (1989) also showed the optimum activation temperature as 450°C
while preparing activated carbon in a muffled furnace from coconut shell precursor
using H3PO4 as activating agent. Activated carbons prepared from wet feedstock
and from the activation at 450-500°C have the BET surface area of ≥ 1000 m2.g-1.
Activated carbons prepared at activation temperatures of 400 and 700°C showed
BET surface areas of 593 and 734 m2.g-1 with a total pore volumes of 0.31 and 0.36
cm3.g-1 respectively. The activation at 400°C was observed too low to develop
81
surface area and pores because of lower thermal energy for the evolution of volatile
matter. This was evident from its larger yield (68%). On the other hand, high
temperature (700°C) caused shrinkage of pores for which total pore volume was
reduced largely to 0.36 cm3.g-1 (see Table 3.2).
The N2 adsorption isotherms of carbons (see Figure 3.1) prepared at low
(400°C) and high (700°C) temperatures (PSW-P-400 and PSW-P-700) have similar
adsorption characteristic as very low N2 adsorption throughout the pressure range
indicating lower surface area and pore development. The volume of adsorbed N2 by
oil palm and coconut shell derived carbons obtained from wet feedstock (PSW-P500 and CSW-P-500) at activation temperature of 500°C were moderate among the
experimental carbons. The isotherm pattern of a carbon obtained from completely
dried feedstock, PSW-P-ed-450, is similar to that of oil palm and coconut shells
derived carbons obtained from wet feedstock (PSW-P-500 and CSW-P-500), but the
former has relatively higher potential to adsorb N2 (see Figure 3.1). An activated
carbon obtained from smaller particles, PSWfg-P-450 has very high N2 adsorption at
low pressure and unlike other activated carbons (those have moderate to high N2
adsorption capacity) did not show any hysteresis loop in the isotherm. Thus it
indicates that while all activated carbons prepared from the larger particles and from
the activation temperatures in between 450-500°C contained mixtures of major
micro- and minor mesopores, the activated carbon prepared from smaller particles
contained solely micropores. The very high N2 adsorptions were observed by two
activated carbons, CPW-P-500 (H3PO4 pretreated one) and PSW-P-ad-500 (preheat
treated ones) which indicate their comparatively very high surface area and pore
development among experimental carbons.
Although the activation temperature is the major factor that determines total
surface area and pore-types; pretreatment, impregnation condition and precursor size
also play a vital role. Pretreatment with acid, especially with H2SO4, prevents
uncontrolled evolution and thus restricts deposition of carbon particles onto the
pores.
Pretreatment with H3PO4 acid intensifies the impregnation treatment of
particle activation that contributes wider pore width (2.52 nm) with high total pore
volume (0.94 cm3.g-1) and very large surface area (1491 m2.g-1) in the resulting
carbon. Evaporation of impregnation solution to some form of dryness accelerates
82
and intensifies the activation processes resulting activated carbon with large BET
surface area (1366-1476 m2.g-1) and high total pore volume (0.75-0.80 cm3.g-1).
Raw materials of smaller sized particles also contribute high BET surface area (1472
m2.g-1), with absolute development of microporosity in the resulted carbon. Thus, an
appropriate adoption of preparation variables would result in a customized product
being formed.
3.2.2
Surface Functional Groups
Surface functional groups on various prepared activated carbons were
determined directly using FTIR and quantified indirectly using Boehm’s titration.
Both techniques have some limitations.
The absorbance peaks obtained from
activated carbons are generally very weak while in Boehm’s titration the same
functional group may response differently with different bases.
So, Boehm’s
assumption that only different acidic groups on activated carbon would response
differently with different bases is also questionable (Puri, 1980).
At a pH where protonic surface charge is equal to zero is called pHZPC. The
highly developed surface acidic groups on carbon contribute cation exchange
property over a wide range of pH while that range can be determined from a pH as
said above as pHZPC. Since hydrous surface of activated carbon behaves as acid and
base according to surface reactions (1.6) and (1.5) respectively (see in Chapter 1), its
pK value at a point of surface’s electro-neutrality (when net surface acidity becomes
zero) can be calculated as pHZPC. The pK value at the surface neutrality point can be
calculated from the acidity constants for the surface reactions of (1.5) and (1.6). For
experimental simplicity, pH drift method is widely used in the determination of
pHZPC where batch equilibrium method is used to determine a crossover point of
equilibrated and initial pH as pHZPC (Babić et al., 1999). The crossover point is a
point or a range as horizontal line in the middle of the two extreme vertical/rising
trend lines while equilibrated pH are plotted against initial pH. That means surface
charge may equal to zero over a range of pH (as denoted by the horizontal line in the
graph) rather than at a single point of a specific pH.
83
3.2.2.1 Fourier Transform Infrared (FTIR) Spectroscopy
The FTIR spectra of some of the prepared activated carbons, namely, PSWP-500, CPW-P-500 and PSW-P-ad-500 are shown in Figure 3.10 (see Appendix D).
The activated carbon, PSW-P-500, prepared from wet feedstock of H2SO4 pretreated
palm shell precursor displayed the following bands: i) 3413 cm-1: O−H stretching in
hydroxyl groups; ii) 2972 cm-1: =C−H stretching in unsaturated hydrocarbon; iii)
2917 cm-1: −C−H vibration stretching in saturated hydrocarbon; iv) 2342 cm-1: a
sharp characteristic peak of CO2 while the absorbance arose from its resonance
structure of carbon triply bonded oxygen stretching (Here it was not clear whether
the peak ascribed some form carbon-oxygen groups); v) 1701 cm-1: C=O stretching
in ketones and carboxylic acids; vi) 1653 cm-1: C=C stretching in aromatic rings
(skeletal carbon); C=O stretching in quinones; vii) 1384 cm-1: C−O stretching in
ethers, C−H stretching in methyl groups; viii) 1092 cm-1: P−O stretching in
phosphonic acid groups; P−O−C asymmetric stretching.
Figure 3.10
FTIR spectrums of some of the prepared activated carbons, namely,
PSW-P-500, CPW-P-500 and PSW-P-ad-500.
The activated carbon, CPW-P-500, prepared from H3PO4 pretreated coconut
shell precursor showed the following absorption bands: i) 3450cm-1: O−H stretching
vibrations in hydroxyl functional groups; ii) 1690cm-1: C=O stretching in ketones
and/or in carboxylic acids; iii) 1624cm-1: C=C stretching in aromatic rings (skeletal
84
carbon); C=O stretching in quinones; iv) 1559cm-1: C=P stretching; v) 1016cm-1:
P−O stretching in phosphonic acid groups; P−O−C asymmetric stretching. The main
acid groups present on the surface of CPW-P-500 were assumed as phenolic,
carbonyl and phosphorous acid groups.
The activated carbon, PSW-P-ad-500, obtained from the semi-dried
feedstock of H2SO4 pretreated palm shell precursor displayed the following bands: i)
3486cm-1: O−H stretching in hydroxyl groups; ii) 2884cm-1: −C−H stretching in
saturated hydrocarbon; iii) 2342cm-1: a sharp characteristic peak of CO2 (Here it was
not clear whether the peak ascribed some form of carbon-oxygen groups); iv)
1705cm-1: C=O stretching in ketones and carboxylic acids; v) 1623cm-1: C=C
stretching in aromatic rings (skeletal carbon); C=O stretching in quinones; vi)
1559cm-1: C=C stretching in aromatic rings (skeletal carbon); vii) 1400cm-1: C−O
stretching in ethers, C−H stretching in methyl groups; viii) 1168cm-1: C=O
stretching in ketones and carboxylic acids; ix) 995cm-1: P−O stretching in
phosphonic acid groups; P−O−C asymmetric stretching.
The probable main acidic groups present on the surface of PSW-P-500 and
PSW-P-ad-500 were phenols, carbonyl groups (either isolated or arranged in
quinone-like structure), carboxylic and phosphorous acid groups.
Here, it is
supposed that H2SO4 pretreatment enhanced the development of carboxylic groups
in PSW-P-500 and PSW-P-ad-500. The hydroxyl groups either in phenols and/or in
phosphorous acid groups were well developed from all H3PO4 activated carbons.
3.2.2.2 Boehm’s Titration
Base neutralization capacity and quantitative distributions of acidic groups on
surfaces of some selected prepared activated carbons, CPW-P-500, PSW-P-500 and
PSW-P-ad-500, were determined by Boehm’s titration and the results are tabulated
in Table 3.4. The titration data and subsequent calculations are shown in Appendix
E. In the literature, strong acidic group detected by NaHCO3 is assumed to be only
the carboxylic group. But activated carbon prepared using H3PO4 might contain
85
strongly acidic hydroxyl (-OH) groups due to the presence of phosphorous
containing acids attached to the surface (cited by Dastgheib and Rockstraw, 2001).
Thus the consumption of NaHCO3 by the H3PO4 activated carbon quantified the
carboxylic as well as the phosphorous acid groups as strong acidic groups. Sodium
carbonate, in addition to strong acidic groups (Group I in Table 3.4), is capable of
detecting lactones.
Other than lactones, phosphorous containing acids of
intermediate strength attached on carbon surface may also take part in the
neutralization reaction with Na2CO3.
Sodium hydroxide, in addition to the
previously described groups, can detect phenols and phosphonic acids whereas
sodium ehoxide (NaOEt), in addition to the all other described groups, can detect
carbonyl groups. Results from Boehm’s titration indicated that the total acidity of
oil palm shell derived activated carbon (4.5-4.8) is higher than that of coconut shell
derived activated carbon (3.5).
Table 3.4 : Base neutralization capacity and quantification of acidic groups on
surfaces of some of the prepared activated carbons by Boehm’s titration method.
Group III
Group IV
Total
ad-500
Group II
PSW-P-
Group I
P-500
NaOEt
PSW-
NaOH
P-500
Acidic group quantification, meq.g-1
Na2CO3
CPW-
Base uptake, meq.g-1
NaHCO3
Sample
0.5
1.5
3.0
3.5
0.5
1.0
1.5
0.5
3.5
1.0
3.0
4.5
4.5
1.0
2.0
1.5
-
4.5
1.0
3.0
4.5
4.8
1.0
2.0
1.5
0.3
4.8
Group I = strong acidic groups such as HO-PO2H2, OH-PO3H2, OH-PR3 , -COOH;
Group II = phosphorous acid of intermediate strength, lactones etc.; Group III =
phenols, Phosphonic acids; Group IV = carbonyl groups and NaOEt = NaOC2H5.
86
3.2.2.3 Determination of pHZPC of Activated Carbon (pH Drift Method)
Experimental results of pHZPC determination, using pH drift method, are
shown in Figure 3.11. Here, equilibrated pH (pHf) by activated carbon was plotted
against initial pH (pHi) of the solution having constant ionic strength (0.01M NaCl).
A pH at a point where the initial pH of the solution crossover the pH equilibrated by
a certain amount of an adsorbent/carbon is the pHZPC. The obtained pHZPC of PSWP-500, PSW-P-ad-500 and CSW-P-500 are tabulated in Table 3.5.
5
Equilibrated pH (pH f )
CPW-P-500
4.5
PSW-P-500
PSW-P-ad-500
4
3.5
3
2.5
2
2
3
4
5
6
7
8
9
10
Initial pH (pHi)
Figure 3.11
Graphs of equilibrated pH (pHf) vs. initial pH (pHi) by some of the
prepared activated carbons for determining the pHZPC from the crossover point.
Table 3.5 : The pHZPC of some activated carbons obtained from pH drift method.
Activated carbons
pHZPC
PSW-P-ad-500
3.06
PSW-P-500
3.26
CPW-P-500
3.34
87
3.2.3
Thermogravimetric Analysis (TGA) of Activated Carbon
TG analyses of various activated carbons were carried out and estimated the
moisture, carbon and ash contents. The results of the proximate analysis from TGA
(see Appendix F) of some selected prepared activated carbons are shown in Table
3.6. Moisture content was calculated from the mass lost at 150°C while analysis was
performed in N2 atmosphere. On the other hand, ash content was determined from
the left residue at 900°C in an atmosphere of air while carbon content was calculated
in percent deducting the moisture and ash (see equation (2.2) in Chapter 2). No
relation can be drawn on the ash and hence on the carbon contents of palm shell and
coconut shell derived carbons prepared under different conditions and activation
temperatures. Since only small amounts of carbon sample (10 mg) was used in the
TGA technique, the distribution of ash content might not be uniform throughout the
carbon samples.
Table 3.6 : The proximate analysis of activated carbons using TGA technique.
Sample
moisture
Ash
Carbon
%
%
%
PSW-P-400
13
14.99
72.01
PSW-P-500
10.5
0.79
88.71
PSW-P-ad-500
13
9.94
77.06
PSW-P-600
13
2.59
84.41
CSW-P-500
9
1.71
89.29
CPW-P-500
17.5
8.72
73.78
3.3
Characterization of C-HFO Composites
The porous structure and surface morphology of C-HFO composites were
characterized following the same procedure as that of activated carbon, namely N2
adsorption-desorption and scanning electron microscopy. The BET surface area,
88
average pore width, total pore volume and pore-types of some of the composites and
their parent activated carbons were determined and shown in Table 3.7. The poretypes were assigned from the N2 isotherm plot and BJH desorption pore distribution
graphs. It was observed that the dispersion of HFO reduces the surface area and
pore volume. It was also observed from the N2 adsorption data that the precipitating
agent NaOH, reduced the surface area and pore volume to a greater extent in CHFO-1 and C-HFO-3 while the reduction effect was less dominant in a composite,
C-HFOCa-1, obtained using Ca(OH)2. The SEM micrographs of various C-HFO
composites, namely C-HFO-1 and C-HFOCa-1; C-HFO-2 and C-HFO-3 are shown
in Figures 3.12-3.14 respectively.
Table 3.7 : The BET surface area, average pore width, total pore volume and poretypes of C-HFO composites and their parent activated carbons.
Activated
SBET
Wp
Vt
Pore-types
carbons
m2.g-1
nm
cm3.g-1
(Qualitative assessment)
micro
meso
macro
C-HFO-1
436
2.06
0.224
major
minor
-
C-HFOCa-1
671
2.22
0.372
major
minor
-
PSW-P-500
1049
2.32
0.609
major
minor
-
C-HFO-3
671
2.08
0.350
major
minor
-
PSW-P-ad500
1476
2.17
0.800
major
minor
-
SBET = BET surface area; Wp = average pore width; Vt = total pore volume.
From the SEM micrographs (Figure 3.12-3.14), it is clear that the dispersion
of HFO was also occurred into the external pores as well. Figure 3.12 shows that the
use of Ca(OH)2 as a precipitating agent had widen some of the external pores (Figure
3.12B) unlike NaOH (Figure 3.12A). Figure 3.12 also shows that relatively higher
amounts of HFO-agglomerates had occupied inside these widened external pores.
The similar interactions are speculated in the microporous surface of C-HFO-1 and
C-HFOCa-1 as it can be evidenced from the less destruction (< 40 %) of pore
volume in C-HFOCa-1 and relatively higher destruction (ca 60 % of pore volume) in
C-HFO-1 (see Table 3.7).
89
A
B
Figure 3.12
SEM micrographs of composites of C-HFO-1 (A) and C-HFOCa-1
(B) obtained from parent activated carbon PSW-P-500 using different bases, such as
NaOH and Ca(OH)2.
90
A
B
Figure 3.13
SEM micrographs of a composite, C-HFO-2 obtained from parent
activated carbon CPW-P-500 while magnification factor is different as 200X (A) and
800X (B).
91
A
B
Figure 3.14
SEM micrographs of C-HFO-3 obtained from parent activated carbon
PSW-P-ad-500 while magnification factor is different as 250X (A) and 1200X (B).
The porous structure of CPW-P-500 and PSW-P-ad-500 through their
modification into the composites of C-HFO-2 and C-HFO-3 respectively, had
92
partially filled out by the dispersed HFO agglomerates (see Figures 3.13-3.14). This
phenomenon is much clearer in Figure 3.14 (B), where the size of an external pore
had reduced to half due to the dispersion of agglomerated HFO. This observation of
reduction of external pores is consistent with N2 adsorption data that showed about
50 % of reduction of BET surface area and total pore volume of parent carbons
through dispersion of HFO in C-HFO-3.
In order to compare the preparation of composites with a commercial one,
namely, Fe-GAC, an Fe(III)-oxide impregnated activated carbon, and with its base
carbon, non-impregnated counterpart (virgin GAC), the physical properties of FeGAC and virgin GAC are given in Table 3.8 (Reed, Vaughan and Jiang, 2000). The
Fe content of C-HFO-1 is comparable with that of Fe-GAC as 7 % (see Tables 3.1
and 3.8), while other prepared composites contained higher level of Fe content as 1319% (see Table 3.1). The surface area and pore volumes were less destructed ( 20%)
in the commercial one (Fe-GAC).
Table 3.8 : Properties of Virgin GAC and Fe-GAC obtained from Norit Americas
Inc. (Reed, Vaughan and Jiang, 2000).
Carbon type
Fe content
SBET
Vt
%
m2.g-1
cm3.g-1
Virgin GAC
0.62
1065
0.734
Fe-GAC
7.0
840
0.597
SBET = BET surface area; Vt = total pore volume.
3.4
Leaching Test of Fe from C-HFO Composites
Since the modification of the activated carbon was carried out by dispersing
HFO inside the porous structure of carbon, it is needed to check whether the Fe
content is tightly bonded or easily replaceable with other cations present in water.
No detectable Fe content was found in the leached samples while prepared
93
composites were contacted with water at pH 4.5 and analyzed by FAAS (see
Appendix G). Thus it could be concluded that the dispersed Fe content onto the
porous surface of activated carbon were tightly bound which might not dissolute at
pH above 4.5. Since the adsorbent would be used in the removal of anions/neutral
species of arsenic in the drinking water purification, pH of treated water is not
expected to be lower than 5. However, as Fe(III) was loaded at pH 1.8, it is expected
that its dissolution-pH might reside around/below pH 2.
CHAPTER 4
RESULTS AND DISCUSSION : APPLICATION OF ADSORBENTS
4.1
Application of Prepared Adsorbents
Batch adsorption experiments were carried out to find the application of
prepared adsorbents in the removal of ionic or non-ionic species of some
metals/metalloids. The adsorption performances were also compared with a typical
commercial activated carbon abbreviated as CAC. Since the surface of activated
carbon is considered as relatively hard Lewis base, the adsorptions of borderline
metal cations (class AB in Figure 1.1 in Chapter 1) onto the prepared acid activated
carbon were studied extensively. Thus the metal ions of Cu(II), Ni(II) and Zn(II)
were preferentially chosen as adsorbate. It was discussed in Chapter 1 (section 1.6)
that the removal of Cu(II), Ni(II) and Zn(II) is important from the toxicity and
environmental standpoint. However, the adsorption studies with these metal ions
were also important to understand the mechanism and to predict the adsorption
selectivity among transition metal ions as Irving-Williams (Shriver et al., 1991) had
determined an order of transition metals (first row d-elements such as Mn, Fe, Co,
Ni, Cu and Zn in the periodic table) depending on their stability in a complex.
On the other hand, as the basal structural units or graphitic layers of activated
carbon are considered as soft Lewis base (Alfarra, et al., 2004), the adsorption of
soft ion on the carbon is also interesting. It was also discussed in Chapter 1 (section
1.6) that the remediation of Pb(II) is similarly essential from the toxicity and
environmental standpoint. Thus, the study of the adsorption of Pb(II) as a soft ion
95
(class B in Figure 1.1 in Chapter 1) on the prepared acid activated carbon was also
considered important. While the single solute adsorption was studied to find the
suitability of the prepared acid activated carbon in the removal of a particular metal
ion, adsorptions in binary and ternary solutes solutions were conducted to understand
the mechanism as well as to determine the feasibility in a competitive fashion.
The removal of oxyanions of moderately toxic Cr(VI) and highly toxic
anionic and non-ionic species of As(V) and As(III) (Cech, 2005) were taken into
investigation for their increasing threat to environment and human life.
These
studies also have an additional importance in predicting the suitability of the acid
activated carbons and modified carbon-HFO composites for the removal of other
oxy-anionic and non-ionic species present in the aquatic environment.
4.1.1
Selection of Prepared Activated Carbon
A preliminary adsorption screening test was conducted using a solution of
Cu(II) to limit the number of adsorbents, to be studied in adsorption experiments, to
a few, as many activated carbons were prepared from different precursors under
different conditions. In this experiment (see Chapter 2), the adsorption equilibrium
capacities of various activated carbons were compared from the adsorption in a
single point concentration.
Activated carbons having BET surface area of < 1000 m2.g-1 were not
considered for adsorption studies. Activated carbons having high BET surface area
(> 1000 m2.g-1) were grouped into five series: as obtained from i) wet feedstock
(PSW-P-500, CSW-P-500), ii) H3PO4 pretreated precursor (CPW-P-500), iii) preheat treated feedstock (PSW-P-ad-500, PSW-P-ed-450, iv) tube furnace activation
under N2 flow (PSWfg-P-ed-450-TN) and v) smaller particle sized precursor
(PSWfg-P-450). The carbons, PSW-P-500 and CSW-P-500, prepared from wet
feedstock of palm and coconut shells have identical surface area and porosity. Thus,
the former one (PSW-P-500), considering a representative of both, was investigated
in the screening test. The results of adsorption screening test of various carbons
96
having different BET surface areas (SBET) and pore widths (Wp) are shown in Table
4.1.
Table 4.1 : Cu(II) uptake capacity of various activated carbons in a single solute
solution while initial solution (adsorbate) concentration and pH were kept constant at
292 ± 6 mg.L-1 and 6 respectively with a constant adsorbent concentration of 2 g.L-1.
Adsorbent
Cu(II) uptake
-1
RSD
SBET
2
-1
Wp
capacity (mg.g )
%
m .g
nm
PSW-P-500
60 ± 6
10.57
1049
2.32
CPW-P-500
70 ± 3
3.68
1491
2.52
PSW-P-ad-500
72 ± 3
3.88
1476
2.17
PSW-P-ed-450
62 ± 2
3.38
1366
2.19
PSWfg-P-ed-450-TN
13 ± 0.4
2.94
-
-
PSWfg-P-450
60 ± 3
4.96
1472
1.96
RSD = Relative standard deviation.
All activated carbons prepared under the atmosphere in the muffled furnace
have higher uptake capacity (60-70 mg.g-1) than that of an activated carbon prepared
under the flow of N2 in the tube-furnace (13 mg.g-1). The highest uptake was
observed on PSW-P-ad-500 and CPW-P-500. The BET surface area of PSWfg-P450 is much higher than that of PSW-P-500 and is identical to that of PSW-P-ad500. Even so, the adsorption capacity of PSWfg-P-450 is comparable to that of
PSW-P-500 which is considerably lower than that of PSW-P-ad-500. The relatively
narrow average pore width (< 2 nm) of PSWfg-P-450 indicates the presence of a
higher portion of narrow micropores (ca 1 nm) which might be responsible for its
lower uptake. This is also consistent with the observations of other researchers
(Dastgheib and Rockstraw, 2001; Strelko and Malik, 2002) that a portion of
micropores were not accessible to the hydrated copper ions having ionic radii 1.2
nm. Consequently three activated carbons, one from wet feedstock, the other one
from H3PO4 pretreated precursor and the third one from preheat-treated feedstock
such as PSW-P-500, CPW-P-500 and PSW-P-ad-500 respectively, were selected for
subsequent adsorption studies. It is important to note that the surface precipitation
97
and adsorption are supposed to be responsible for the removal of Cu(II). The reason
for such speculation lies in the fact that adsorption was conducted at high pH, 6
while initial Cu(II) concentration was also reasonably high as < 300 mg.L-1.
4.2
Adsorption Equilibria
Langmuir and Freundlich adsorption isotherms (Faust and Aly, 1987) are
widely employed to evaluate and compare the adsorption performance of adsorbents.
The non-linear and linear equations of Langmuir, as described in equations (1.9) and
(1.10) respectively (see Chapter 1), were used to evaluate adsorption behaviour by
fitting the adsorption equilibrium data. The adsorption behaviour was also predicted
by fitting the equilibrium data to the Freundlich model as described in equation
(1.12) in Chapter 1. The significance of Langmuir and Freundlich model parameters
were also described in the sections 1.5.1 and 1.5.2 respectively in Chapter 1.
According to the Gouy-Chapman-Stern-Grahame Model, the free energy of
adsorption can be defined, by the equation (4.1), as the sum of the free energy of the
chemical interaction and that of an electrostatic force of attraction/repulsion
(Corapcioglu and Huang, 1987; Reed, Vaughan and Jiang, 2000).
∆Gtotal = ∆Gchemical + ∆Gelectrostatic
(4.1)
The pH of solution plays an important role in the adsorption of cations and
anions as illustrated in SCF model. The description of this model and the effect of
pH on surface functional groups of activated carbon including reactions can be found
in the equations of surface reactions, (1.5) and (1.6) in Chapter 1.
The effect of competing ion on the adsorption of a particular ion can be
determined by the ratio of maximum monolayer adsorption capacities for the
adsorption in the single- (qmax-single) and the multi-solute solution (qmax-mix) (Mohan
and Chander, 2001) as described by the following equations (4.2-4.4).
98
q max − mix
>1
q max −sin gle
(4.2)
Indicating, the adsorption is promoted by the presence of other ions.
q max − mix
=1
q max −sin gle
(4.3)
There is no observable net interaction.
q max − mix
<1
q max −sin gle
(4.4)
Adsorption is suppressed by the presence of other ions.
4.2.1
Single Solute Adsorption Equilibria : Metal Cations
In a single solute adsorption equilibrium experiments, adsorption behaviour
of a single, specific ion was predicted applying the adsorption equilibrium models.
The performances of selected three prepared (see section 4.1.1) and one commercial
activated carbon were evaluated and compared for the removal of Cu(II), Ni(II),
Zn(II) and Pb(II) in single-solute solutions. Since the surface precipitation of these
cations may occur at > pH 5 and since the pHZPC of the prepared activated carbon
was determined as circa 3, the initial pH of the experimental solution was varied, in
this study, within 3 to 5 to ensure true and favourable adsorption. All analytical data
of metal cations, before and after adsorption, to determine the adsorption capacity as
well as to calculate the model parameters are presented in Appendix H.
4.2.1.1 Single Solute Adsorption Equilibria : Cu(II)
(a) Effect of initial solution concentrations : Figure 4.1 shows the non-linear
Langmuir adsorption isotherms of Cu(II) for initial pH 3 and 5, by various activated
carbons. Prepared activated carbons have very high affinity to bind Cu(II) as it is
99
seen from the H-type shape in the isotherms according to the classification of Giles
et al. (cited by Jankowska et al., 1991). The H-type shape also reveals complete
adsorption at very low concentration. Linear Langmuir and Freundlich adsorption
isotherms are depicted in Figures 4.2 and 4.3 followed by tabulation of the
corresponding parameters in Tables 4.2.
The values of correlation coefficients (r2) of Langmuir isotherms indicate
that this model fits all the experimental data very well throughout the experimental
range of study. The three prepared activated carbons show almost similar qmax (1923 mg.g-1) which is 4-7 fold higher than that of commercial ones (3-5.6 mg.g-1)
under the experimental conditions of this study.
The adsorption data by three
prepared activated carbons also fitted to the Freundlich model very well (refer to r2
values in Table 4.2). The adsorption data by commercial activated carbon are better
fitted to the Langmuir model.
Adsorption
capacity, q e mg/g
25
A
20
15
PSW-P-ad-500
CPW-P-500
PSW-P-500
CAC
10
5
0
0
Adsorption
capacity, q e mg/g
25
40
80
120
Equilibrium concentration, Ce mg/L
B
20
PSW-P-ad-500
CPW-P-500
PSW-P-500
CAC
15
10
5
0
0
40
80
120
Equilibrium concentration, Ce mg/L
Figure 4.1
Langmuir adsorption isotherms of Cu(II) at initial pH 3 (A) and pH 5
(B) while initial concentrations were varied from < 5 to < 120 mg.L-1.
100
6
A
Ce/qe
4
PSW-P-500
CPW-P-500
PSW-P-ad-500
2
0
0
25
50
75
100
Equilibrium concentration, Ce mg/L
6
B
Ce/qe
4
PSW-P-500
CPW-P-500
PSW-P-ad-500
2
0
0
25
50
75
100
Equilibrium concentration, Ce mg/L
45
C
CAC-pH 3
CAC-pH 5
Ce/qe
30
15
0
0
40
80
120
Equilibrium concentration, Ce mg/L
Figure 4.2
Langmuir adsorption isotherms of Cu(II) by prepared activated
carbons for initial pH 3 (A) and pH 5 (B) and by a commercial activated carbon for
initial pH at 3 and 5 (C) while initial concentrations were varied, < 5 - < 120 mg.L-1.
101
1.5
A
log q e
1
PSW-P-ad-500
CPW-P-500
PSW-P-500
CAC
0.5
0
-0.5
1.5
log Ce
1.5
2.5
1.5
2.5
B
1
log q e
0.5
PSW-P-ad-500
CPW-P-500
PSW-P-500
CAC
0.5
0
-0.5
Figure 4.3
0.5
log Ce
Freundlich adsorption isotherms of Cu(II) at pH 3 (A) and at pH 5 (B)
while initial solution concentrations of Cu(II) were varied, < 5 - < 120 mg.L-1.
Table 4.2 : Langmuir and Freundlich model parameters for the adsorption of Cu(II)
in single-solute solution while initial concentrations were varied, < 5 - < 120 mg.L-1.
Adsorbent
pH
Langmuir parameters
qmax
b
mg.g-1
L.mg-1
21.23
0.55
CPW-P-500
19.57
PSW-P-500
CAC
r2
Freundlich parameters
r2
KF
n
mg.g-1
g.L-1
0.997
11.25
6.51
0.976
0.16
0.991
7.54
4.89
0.977
18.62
0.17
0.997
6.35
4.15
0.972
2.91
0.15
0.967
1.02
4.45
0.748
23.09
0.59
0.999
13.16
7.38
0.989
CPW-P-500
19.76
0.22
0.995
9.21
6.00
0.997
PSW-P-500
19.46
0.23
0.989
9.64
6.42
0.976
CAC
5.56
0.09
0.976
1.50
3.75
0.928
PSW-P-ad-500
PSW-P-ad-500
3
5
102
The higher values of “n” in Freundlich model indicate favourable adsorption.
Although the differences in qmax of three prepared activated carbons are not very
significant, the value of Freundlich constant, KF by one of the carbons, PSW-P-ad500, derived from palm shells’ semi-dried feed-stock, is much higher indicating its
significantly higher affinity towards Cu(II). The non-linear Langmuir isotherm by
PSW-P-ad-500 steeps more in low concentration which is also an indication of its
relatively stronger affinity towards Cu(II).
Throughout the experimental concentration range, oil palm shell activated
carbon obtained from semi-dried feed-stock (PSW-P-ad-500) shows the highest
adsorption capacity as it is seen from the shape of Langmuir and Freundlich
isotherms (refer to tabulated data in Tables H-1.1 and H-1.3 in Appendix H). This
finding is in agreement with the model parameters such as higher values of qmax, KF
and “n” obtained by this carbon. Thus PSW-P-ad-500 is considered as the best
Cu(II) scavenger followed by CPW-P-500, PSW-P-500.
The higher adsorption
capacity of PSW-P-ad-500 can be justified from its well developed acidic groups as
shown in FTIR spectrum (Figure 3.10) as well as from its lower pHZPC ( Figure 3.11)
and higher level of Boehm quantified acidic groups (4.8 meq. g-1) (see Table 3.4).
(b) Effect of initial pH : The effect of pH on Cu(II) as a borderline metal ion
was observed from the constructed Langmuir (Figures 4.1 and 4.2) and Freundlich
(Figure 4.3) isotherms for pH 3 and 5. Since the pK value of the first hydrolysis
product of Cu(II) is 7.73, the dominant species present under the experimental pH
range are [Cu(H2O)6]2+ and [CuOH(H2O)5]+. Removal of Cu(II) depends upon the
nature and distribution of its hydroxo species in solution and their interaction with
hydrous surface oxide of carbon (Corapcioglu and Huang, 1987; Panday, Prasad and
Singh, 1985), thus eventually depends on pH. The pHZPC is the pH at which the
adsorbent has a net zero surface charge while at pH below or above this, has a net
positive or negative charge. With the increase of pH, at > 3 (pHZPC), the prepared
activated carbon, through negatively charged sites (SO−), binds more positively
charged species of copper by chemical and electrostatic interaction (see the values of
qmax in Table 4.2). As the negatively charged site on carbon surface increases with
the increase of pH (see the surface reactions (1.5) and (1.6) in Chapter 1), chemical
and electrostatic interactions between carbon and metal ion increase consistently
103
until surface precipitation commences. Surface precipitation reactions are predicted
in surface precipitation model (Farley, Dzombak and Morel, 1985) from the abruptly
increased adsorption capacity of cations at a specific pH, called pHabr. At pHabr,
uptake of cation occurs on multi-layer resulting very high uptake (see Table 4.1).
At pH ≤ 3, the protonated surface sites of activated carbon exist mainly as
neutral or positively charged species (SOH/SOH2+). The moderately high adsorption
capacity of the prepared activated carbon even at pH 3 (18.6-21.2 mg.g-1) implies
that the surface oxides function as ligands for metal ions while adsorption is
attributed solely to chemical interaction namely, inner-sphere complexation. This
assumption will be verified later. The pH of the equilibrated solution by prepared
activated carbon was found circa 3 for the initial pH whether 3 or 5.
This
phenomenon of pH drift justifies the small increment in qmax for the increase of pH
from 3 to 5 (see Table 4.2). As surface precipitation of Cu(II) may occur at high pH,
many researchers used buffered solutions to compensate the pH drift by acid
activated carbon, especially for the removal of Cu(II) (Toles and Marshall, 2002;
Dastgheib and Rockstraw, 2001; Johns, Marshall, and Toles, 1998; Toles, Marshall
and Johns,1998; Toles, Marshall and Johns, 1997). A buffering of solution pH at 5
maximizes the exposure of negative sites of the adsorbent through dissolution of
more protons and thus enhances the adsorption capacity.
4.2.1.2 Single Solute Adsorption Equilibria : Ni(II)
(a) Effect of initial solution concentrations : Figure 4.4 shows the Langmuir
and Freundlich adsorption isotherms of Ni(II) for an initial pH 5, by various
activated carbons.
The physical parameters for the adsorption of Ni(II) are
determined from the linear Langmuir (Figures 4.4B and 4.4C) and Freundlich
(Figure 4.4D) isotherms and are tabulated in Table 4.3.
One of the prepared
activated carbons, PSW-P-ad-500 (derived from oil palm shells’ semi-dried
feedstock), shows significantly higher Ni(II) adsorption capacity (see Figures 4.4A
and 4.4D). However, all prepared activated carbons have higher affinity towards
Ni(II) as it is evident from the H-type class of their non-linear Langmuir isotherms
104
(Figure 4.4A) according to the classification by Giles et al. (cited by Jankowska et
al., 1991). The higher affinity of prepared activated carbons towards Ni(II) than that
of commercial ones implies that phosphoric acid activation enhances borderline
metal ion removal capacity.
Adsorption
capacity,q e mg/g
20
A
15
PSW-P-ad-500
CPW-P-500
PSW-P-500
CAC
10
5
0
0
50
100
150
200
250
Equilibrium concentration, Ce mg/L
20
B
Ce/qe
15
PSW-P-500
CPW-P-500
10
PSW-P-ad-500
5
0
0
80
50
100
150
Equilibrium concentration, Ce mg/L
200
250
C
Ce/qe
60
CAC
40
20
0
0
Figure 4.4
50
100
150
Equilibrium concentration, Ce mg/L
200
250
Langmuir (A, B, C) and Freundlich (D) adsorption isotherms of Ni(II)
at initial pH 5 by various activated carbons while initial solution concentrations were
varied as 6 - 227 mg.L-1.
105
log q e
1.4
D
0.7
PSW-P-ad-500
CPW-P-500
PSW-P-500
CAC
0
0.5
1
1.5
2
2.5
log Ce
Figure 4.4
Continued. (Langmuir (A, B, C) and Freundlich (D) adsorption
isotherms of Ni(II) at initial pH 5 by various activated carbons while initial solution
concentrations were varied as 6 - 227 mg.L-1).
Table 4.3 : Langmuir and Freundlich model parameters for the adsorption of Ni(II)
by various activated carbons at initial pH 5 while initial concentrations were varied
from 6 to 227 mg.L-1 and adsorbent concentration was kept constant at 2 g.L-1.
Adsorbent
Langmuir parameters
qmax
b
mg.g-1
L.mg-1
PSW-P-ad-500
19.61
0.15
CPW-P-500
12.18
PSW-P-500
CAC
r
2
Freundlich parameters
r2
KF
n
mg.g-1
g.L-1
0.996
9.28
6.88
0.937
0.42
0.999
9.59
21.98
0.835
10.83
0.29
0.996
5.68
8.00
0.936
3.18
0.13
0.970
0.64
3.06
0.864
The values of correlation coefficients (r2) of Langmuir isotherms by all
carbons indicate that this model fits all the adsorption equilibrium data very well
throughout the experimental range of this study.
Therefore, it is assumed that
adsorption involves a direct contact of metal ions onto the surface of acid activated
carbon and thus proceeds up to monolayer coverage.
The performances of adsorbents for the removal of Ni(II) in aqueous solution
are attributed to the maximum monolayer adsorption capacity, qmax (see Table 4.3) in
106
an ascending series as CAC<< PSW-P-500< CPW-P-500<< PSW-P-ad-500. Their
adsorption equilibrium capacities (qe), as shown in Figure 4.4A, over the entire
experimental concentration range also in agreement with this series (also refer to
Table H-2.1 in Appendix H).
The adsorption data by all activated carbons are also fitted to Freundlich
model (see r2 values in Table 4.3), but the better fitting is observed with the
Langmuir model. The very low affinity of commercial activated carbon towards
Ni(II) is also recognized from the significantly lower values of Freundlich constants,
KF and “n”. According to Freundlich model parameters CPW-P-500 shows higher
affinity towards Ni(II) than that by PSW-P-ad-500. The lower value of correlation
co-efficient (r2 = 0.83) makes questionable the values of model parameters and
hence the comparison based on them as well.
The effectiveness of adsorbents to be used in drinking water purification
depends upon its complete removal efficiency of metals ions at low concentration.
The model parameters (considerably higher values of qmax, KF and n), the shape (Htype) of the isotherms (Figures 4.4A and 4.4D) and adsorption equilibrium capacities
(qe) indicate that the prepared activated carbon PSW-P-ad-500 is the best in this
experiment in removing borderline metal ions such as Ni(II) at the entire range of
experimental solution concentration.
The equilibrium concentrations and fractions removal of Ni(II) in the low
concentration by various carbons, are shown in Table 4.4, which verifies the
suitability of acid activated carbons over commercial activated carbon in low
concentration. This Table (Table 4.4) also shows the most feasible acid activated
carbon as PSW-P-ad-500 with the especial aspect in drinking water purification.
The effect of pH was not studied as Ni(II) belongs to the same category of borderline
cation like Cu(II).
107
Table 4.4 : The initial and equilibrium concentrations, and fractions removal of
Ni(II) by various activated carbons in the low concentration solution at a constant
initial pH 5.
Adsorbent
CAC
PSW-P-500
CPW-P-500
PSW-P-ad-500
Concentrations of Ni(II), mg.L-1
Removal of Ni(II)
Initial
Equilibrium
%
6.0 ± 0.05
4.14 ± 0.02
31.7 ± 0.5
30.6 ± 0.3
27.21 ± 0.21
11 ± 1
6.0 ± 0.05
ND
100
30.6 ± 0.3
15.05 ± 0.11
50.9 ± 0.4
6.0 ± 0.05
ND
100
30.6 ± 0.3
9.76 ± 0.02
68.2 ± 0.2
6.0 ± 0.05
ND
100
30.6 ± 0.3
6.71 ± 0.11
78.1 ± 0.6
ND = Not detectable [the calibration curve in FAAS analysis was constructed for the
concentration range of standard Ni(II) solution as 0.5-5.0 mg.L-1 (see Appendix H)].
4.2.1.3 Single Solute Adsorption Equilibria : Zn(II)
(a) Effect of initial solution concentrations : Figures 4.5 shows the non-linear
(A) and linear (B) Langmuir adsorption isotherms of Zn(II) for an initial pH 5, by a
commercial activated carbon, CAC. Figure 4.6 shows the non-linear Langmuir
isotherms of Zn(II) by various prepared activated carbons. The linear Langmuir
isotherms of Zn(II) by various prepared activated carbons are shown in Figure 4.7.
The physical parameters for the adsorption of Zn(II) are tabulated in Table 4.5. One
of the prepared activated carbons, PSW-P-ad-500 shows relatively higher Zn(II)
adsorption capacity at low concentration (see Figure 4.6C) indicating relatively
higher affinity. However, all prepared activated carbons show poor affinity towards
Zn(II) as it is evident from the low adsorption capacity (qmax = 8 to 10 mg.g-1). The
affinity of commercial activated carbon, CAC towards Zn(II) is extremely low as the
qmax is determined as low as 2.7 mg.g-1.
108
The values of correlation coefficients, r2 (although relatively lower than that
of Cu(II) and Ni(II); see Tables 4.2, 4.3 and 4.5) of Langmuir isotherms by all
activated carbons indicate that this model fits all the adsorption equilibrium data
throughout the experimental range of study.
Attempts to fit the adsorption
equilibrium data (see Table H-3.1 in Appendix H) to Freundlich model were not
successful. The performances of adsorbents for the removal of Zn(II) in aqueous
solution can be expressed in an order from the maximum monolayer adsorption
capacity, qmax as: CAC << PSW-P-500 < PSW-P-ad-500 ≤ CPW-P-500. The very
low affinity of acid activated carbons towards Zn(II) is attributed from the
completely filled d-orbital of Zn2+ (d-10). The inconsistency of the adsorption
equilibrium data as shown in Figures 4.5-4.7 results from poor affinity. Since Zn(II)
is also classified as borderline ion alike Cu(II), the effect of pH was not studied here.
Adsorption
capacity q e mg/g
3
A
2
CAC
1
0
0
50
100
150
200
250
300
Equilibrium concentration, Ce mg/L
100
B
Ce/qe
80
60
CAC
40
20
0
0
50
100
150
200
250
300
Equilibrium concentration, Ce mg/L
Figure 4.5
Langmuir non-linear (A) and linear (B) isotherms of Zn(II) by CAC
at initial pH 5 while solution concentrations were varied, 19 - 250 mg.L-1.
109
Adsorption capacity,
qe mg/g
10
A
8
6
4
PSW-P-500
2
0
0
50
100
150
200
250
300
350
Equilibrium concentration, Ce mg/L
Adsorption capacity,
qe mg/g
12
B
10
8
6
4
CPW-P-500
2
0
0
50
100
150
200
250
300
350
Equilibrium concentration, Ce mg/L
Adsorption capacity,
qe mg/g
12
C
10
8
6
4
PSW-P-ad-500
2
0
0
50
100
150
200
250
300
350
Equilibrium concentration, Ce mg/L
Figure 4.6
Langmuir non-linear isotherms of Zn(II) at initial pH 5 by prepared
activated carbons, PSW-P-500 (A), CPW-P-500 (B) and PSW-P-ad-500 (C) while
initial solution concentrations were varied, 19 – 310 mg.L-1.
110
40
PSW-P-500
CPW-P-500
PSW-P-ad-500
Ce/qe
30
20
10
0
0
50
100
150
200
250
300
350
Equilibrium concentration, Ce mg/L
Langmuir linear isotherms of Zn(II) at initial pH 5 by various
Figure 4.7
prepared activated carbons while initial solution concentrations were varied from 19
to 310 mg.L-1.
Table 4.5 : Langmuir model parameters for the adsorption of Zn(II) by various
activated carbons at initial pH 5 while initial concentrations were varied from 19 to
310 mg.L-1 and adsorbent concentration was kept constant at 2 g.L-1.
Adsorbent
Langmuir parameters
qmax (mg.g-1)
b (L.mg-1)
r2
PSW-P-ad-500
9.48
0.204
0.954
CPW-P-500
9.90
0.061
0.958
PSW-P-500
8.14
0.089
0.970
CAC
2.70
0.238
0.993
4.2.1.4 Single Solute Adsorption Equilibria : Pb(II)
(a) Effect of initial solution concentrations : Figures 4.8 and 4.9 show the
Langmuir and Freundlich adsorption isotherms of Pb(II), for an initial pH 5, by
various activated carbons. The physical parameters for the adsorption of Pb(II) are
111
determined from the linear Langmuir (Figures 4.8B and 4.8C) and Freundlich
(Figure 4.9) isotherms and are tabulated in Table 4.6.
Adsorption capacity,
qe mg/g
80
A
PSW-P-500
CPW-P-500
PSW-P-ad-500
CAC
60
40
20
0
0
200
400
600
Equilibrium concentration, Ce mg/L
B
8
Ce/qe
6
4
PSW-P-ad-500
CPW-P-500
2
PSW-P-500
0
0
100
200
300
400
500
Equilibrium concentration, Ce mg/L
25
C
Ce/qe
20
15
CAC
10
5
0
0
200
400
600
Equilibrium concentration, Ce mg/L
Figure 4.8
Langmuir non-linear (A) and linear (B and C) adsorption isotherms of
Pb(II) at initial pH 5 by various activated carbons while initial concentrations were
varied as < 50 - < 600 mg.L-1.
112
2
log q e
1.5
PSW-P-500
CPW-P-500
PSW-P-ad-500
CAC
1
0.5
0
0.5
1
1.5
2
2.5
3
log Ce
Figure 4.9
Freundlich adsorption isotherms of Pb(II) at initial pH 5 by various
activated carbons while initial concentrations were varied as < 50 - < 600 mg.L-1.
Table 4.6 : Langmuir and Freundlich model parameters for the adsorption of Pb(II)
at initial pH 5 and at varying concentrations, < 50 - < 600 mg.L-1.
Adsorbent
Langmuir parameters
qmax
b
mg.g-1
L.mg-1
PSW-P-ad-500
63.69
0.07
CPW-P-500
73.53
PSW-P-500
CAC
Freundlich parameters
2
r2
KF
n
mg.g-1
g.L-1
0.989
22.69
5.73
0.937
0.05
0.997
18.19
4.20
0.943
74.63
0.09
0.998
19.74
4.17
0.929
23.42
0.02
0.977
6.62
5.49
0.775
r
The values of correlation coefficients (see Table 4.6) of all Langmuir
adsorption isotherms indicate that this model fits all the adsorption equilibrium data
very well throughout the experimental range of study.
All prepared activated
carbons show higher adsorption capacity as well as stronger affinity towards Pb(II)
which is also evident from the H-type isotherms (Figure 4.8A). The commercial
activated carbon has comparatively lower affinity towards Pb(II). A comparison in
the removal capacity of Pb(II) by prepared and commercial carbons, from the
observed qmax, determines an order of adsorbents according to their performance,
under the experimental range of study, as CAC < PSW-P-ad-500 < CPW-P-500
113
≤PSW-P-500.
The adsorption data by all activated carbons are also fitted to
Freundlich model (see the values of r2 in Table 4.6), but the better fitting is observed
with the Langmuir model.
The highest adsorption capacity (qmax) of Pb(II) is
observed by one of the activated carbons (PSW-P-500) prepared from “wetfeedstock”. Since two of the activated carbons, one from palm shell precursor as
PSW-P-500 and another from coconut shell as CPW-P-500, were obtained from wetfeedstock, they showed identical adsorption capacity, qmax for Pb(II) (see Table 4.6).
The comparatively higher values of KF and “n” (see Table 4.6) by PSW-P-ad-500
indicate that Pb(II) is more strongly bound onto the surface of PSW-P-ad-500. The
contradiction to the values of qmax and KF may arise from the different mechanism of
adsorption of Pb(II) as soft cation. Unlike borderline cation, it is also expected to be
adsorbed in the basal plane. The equilibrium concentrations and fractions removal
of Pb(II), in the low concentration, are shown in Table 4.7. The results reveal the
suitability of acid activated carbons in the removal of Pb(II) (above 90%) in lower
concentration solution (49-72 mg.L-1) and thus it is expected to be useful for the
removal of Pb(II) in drinking water purification.
On the contrary, commercial
activated carbon removed only 53% of Pb(II) from a solution having a concentration
of 49 mg.L-1.
Therefore low concentration purification directs the following
suitability order of adsorbents: CAC < CPW-P-500 < PSW-P-500 < PSW-P-ad-500.
Table 4.7 : The initial and equilibrium concentrations, and fractions removal of
Pb(II) by various activated carbons in the low concentration range at pH 5.
Adsorbent
CAC
PSW-P-500
CPW-P-500
PSW-P-ad-500
Concentrations of Pb(II), mg.L-1
Removal of Pb(II)
Initial
Equilibrium
%
48.7 ± 0.2
22.4 ± 0.9
53 ± 2
71.6 ± 0.6
46 ± 1
34 ± 2
48.7 ± 0.2
3.8 ± 0.9
92 ± 2
71.6 ± 0.6
6±1
92 ± 1
48.7 ± 0.2
5±1
90 ± 2
71.6 ± 0.6
7±2
91 ± 2
48.7 ± 0.2
2.2 ± 0.7
96 ± 2
71.6 ± 0.6
6.5 ± 0.2
91 ± 0.3
114
(b) Effect of initial pH : To observe the effect of pH, adsorption was also
conducted at initial pH 3 using the best selected activated carbon, PSW-P-ad-500
while initial concentrations were varied as 50.6 ± 0.6 – 583 ± 2 mg.L-1. The
Langmuir and Freundlich adsorption isotherms are depicted in Figures 4.10 and 4.11
while the determined model parameters are presented in Table 4.8. The values of
correlation coefficients of the linear Langmuir and Freundlich isotherms (Table 4.8)
indicate that all the data are fitted to these models very well. The model parameters
reveal that adsorption depends on pH, while qmax decreases from 63.7 to 49.5 mg.g-1
with the reduction of pH, from 5 to 3. It is also observed that PSW-P-ad-500 has
high adsorption capacity even at pH 3. The higher values of Freundlich constants,
KF and n also indicate higher affinity as well as favourable adsorption.
Adsorption capacity,
qe mg/g
50
A
40
30
20
10
0
0
100
200
300
400
500
600
500
600
Equilibrium concentration, Ce mg/L
12
B
Ce/qe
10
8
6
4
2
0
0
100
200
300
400
Equilibrium concentration, Ce mg/L
Figure 4.10
Langmuir non-linear (A) and linear (B) adsorption isotherms of Pb(II)
at initial pH 3 by the best-selected activated carbon, PSW-P-ad-500 while initial
solution concentrations were varied from 50.6 ± 0.6 to 583 ± 2 mg.L-1.
115
log q e
2
1
0
0.5
Figure 4.11
1
1.5
log Ce
2
2.5
3
Freundlich adsorption isotherms of Pb(II) at initial pH 3 by activated
carbon, PSW-P-ad-500 while initial solution concentrations were varied from 50.6 ±
0.6 to 583 ± 2 mg.L-1.
Table 4.8 : Langmuir and Freundlich parameters for the adsorption of Pb(II) on
PSW-P-ad-500 for initial pH 3 at varying concentrations, 50.6 ± 0.6-583 ± 2 mg.L-1.
Adsorbent
PSW-P-ad-500
4.2.2
Langmuir parameters
qmax
(mg.g-1)
49.50
b
(L.mg-1)
0.03
Freundlich parameters
r2
0.991
KF
(mg.g-1)
12.98
n
(g.L-1)
4.69
r2
0.987
Binary Solute Adsorption Equilibria : Competitive Adsorption of Cu(II)
and Ca(II) (Cu-Ca) and Evaluation of Mechanism
The necessity of selectivity is highly desired in removing target metal cations
while using adsorption based process.
Since Ca(II) is the most common and
dominant ion present in natural water, the feasibility of any adsorbent to be used in
drinking water and wastewater treatment depends upon its preferential adsorption
capacity of target ion over Ca(II). In a conventional ion exchanger, Ca(II) ion is
preferred over borderline metal cations such as Zn(II), Co(II), Cu(II) and Ni(II)
(Sengupta and SenGupta, 2002).
Figure 4.12 shows the Langmuir isotherms of Cu(II) for the adsorption in
single (Cu), binary (Ca-Cu), ternary (Cu-Ni-Pb) and that of Ca(II) in binary solutes
116
solutions by the best scavenger, PSW-P-ad-500 at initial pH 5. The Langmuir
parameters of Cu(II) and Ca(II) for the binary adsorptions are given in Table 4.9.
Adsorption capacity,
qe mg/g
25
Single-Cu
Binary-Cu
Ternary-Cu
Ca-Binary
A
20
15
10
5
0
0
20
40
60
80
100
120
140
Equilibrium concentration, Ce mg/L
8
B
Ternary-Cu
Binary-Cu
Single-Cu
Ce/qe
6
4
2
0
0
20
40
60
80
Ce mg/L
100
C
100
Ce/qe
80
60
40
Ca-Binary
20
0
-20
0
20
40
60
80
100
120
140
Equilibrium concentration, Ce mg/L
Figure 4.12
Langmuir non-linear and linear isotherms of Cu(II) (A and B) and
Ca(II) (A and C) while Cu(II) ions were adsorbed from single (Cu), binary (Ca-Cu)
and ternary (Cu-Ni-Pb) solutions and Ca(II) from binary solution (Ca-Cu) by PSWP-ad-500 at initial pH 5 and at varying initial concentrations, < 25 - ≤ 120.mg.L-1.
117
Table 4.9 : The Langmuir model parameters of Cu(II) and Ca(II) for the adsorption
in binary solution at an initial pH 5, while initial concentrations of Cu(II) and Ca(II)
were varied as 23.6 ± 0.1 to 117.5 ± 0.7 and 23.8 ± 0.2 to 120±1 mg.L-1 respectively.
Metal ion
Langmuir model parameters
(in binary)
qmax (mg.g-1)
b (L.mg-1)
r2
Cu(II)
23.15
0.205
0.990
Ca(II)
1.17
-0.061
0.873
It is evident from the corresponding correlation coefficients (r2) of isotherms
of Cu(II) and Ca(II) that the Langmuir model fits the data well. The qmax-single and
qmax-binary (at pH 5) of Cu(II) are obtained as 23.09 and 23.15 mg.g-1 respectively (see
Tables 4.2 and 4.9) and thus the ratio of qmax-binary to qmax-single as 1.00; which means,
according to equation (4.2), the adsorption of Cu(II) was not at all interfered by
Ca(II).
The negative value of Langmuir constant, b (-0.061) reveals negative
intercept of Ca(II) isotherm (see Figure 4.12C) indicating an unfavourable
adsorption with the increasingly selective adsorption of Cu(II) over Ca(II). The qmax
of Ca(II) was determined from the slope as only 1.17 mg.g-1. Because Ca(II) is a
poor Lewis acid and Cu(II) being a transition metal ion exhibits fairly strong Lewisacid characteristics, the very high selective adsorption of Cu(II) over Ca(II)
illuminates the mechanism as inner sphere complexation. Thus, the mechanism of
bond formation of Cu(II) and Ca(II) cations on the surface of activated carbon can be
depicted as a result of major chemical, Lewis acid-base interaction, and minor
electrostatic force of attraction (Figure 4.13).
Here, each deprotonated surface
functional group such as SOH or SO− (see deprotonation of surface functional
groups of activated carbon in equations (1.5) and (1.6), in Chapter 1, as a function of
solution pH in the adsorption of cation) localizes electrons on the surface of carbon
and removes them from the π electron system of the basal planes. Thus providing a
pair of donor electrons, each surface functional group functions as a monodentate
ligand, Lewis base, while the adsorbent as a whole is a polydentate ligand.
Corapcioglu and Huang (1987) reported the adsorption of a number of heavy metal
ions on activated carbon as surface complexation either through inner-sphere or
outer-sphere complexation. Ho and Mckay (2002) assumed the adsorption of Cu(II)
118
onto the surface of sphagnum moss peat is due to the bonding through sharing or
exchanging electrons while Chen and Lin (2001) considered the inner-sphere- as
well as the outer-sphere complexation of metal ion with activated carbon.
2
SO
Ca
SO
H
Cu
SO
H
Ca
2
H
2
SO
2
SO
SO
Cu
SO
SO
H
Coulombic force of attraction
Lewis acid-base interaction
Figure 4.13
The proposed interactions of Cu(II) and Ca(II) cations (Lewis acids)
onto the surfaces of activated carbon where SO− and SOH (Lewis bases) represent
the deprotonated surface functional groups on activated carbon. The vertical bars are
depicted as solution-solid interface.
Therefore, it could be concluded that the adsorption of Cu(II) on acid
activated carbon resulted predominantly from the chemisorption. The very good
fitting of adsorption equilibrium data of Cu(II), those obtained by varying initial
solution concentrations from < 5 to < 120 mg.L-1, to the Langmuir model (see Tables
4.2 and 4.9) also confirms the phenomenon of chemisorption.
The adsorption
experiment was further extended for obtaining the equilibrium data by the best
selected activated carbon, PSW-P-ad-500 in the solution of high Cu(II)
concentration, ranging from < 25 to < 700 mg.L-1. It was planned to adjust the initial
pH of all experimental solutions at 5. But observing a fine precipitation of Cu(OH)2,
at pH 5, in a solution of < 700 mg.L-1, the initial pH of all solutions were, then,
adjusted to 4.5 to confirm true adsorption. The purpose of this study was to observe
whether monolayer adsorption was prevailed even in very high Cu(II) concentration
and thus to furnish more evidence in favour of chemisorption. The constructed
Langmuir and Freundlich isotherms of Cu(II) for the adsorption in single solute
solution in a concentration range of 24.9 ± 0.2 to 693 ± 4 mg.L-1, at initial pH 4.5
and at a constant adsorbent concentration of 2 g.L-1 are shown in Figure 4.14. The
respective model parameters are determined and presented in Table 4.10.
119
Adsorption capacity, qe mg/g
30
A
25
20
15
10
5
0
0
100
200
300
400
500
600
700
600
700
Equilibrium concentration, Ce mg/L
Ce/qe
30
B
20
10
0
0
100
200
300
400
500
Equilibrium concentration, Ce mg/L
1.6
C
log q e
1.2
0.8
0.4
0
0
0.5
1
1.5
2
2.5
3
log Ce
Figure 4.14
Langmuir non-linear (A), linear (B) and Freundlich (C) isotherms of
Cu(II) by the carbon PSW-P-ad-500 at constant initial pH 4.5 while initial solution
concentrations were varied from < 25 to < 700 mg.L-1 keeping adsorbent
concentration constant at 2 g.L-1.
120
Table 4.10 : Langmuir and Freundlich parameters for the adsorption of Cu(II) on
PSW-P-ad-500 at a constant initial pH 4.5 while initial solution concentrations were
varied from < 25 to < 700 mg.L-1 keeping adsorbent concentration at 2 g.L-1.
Adsorbent
PSW-P-ad-500
Langmuir parameters
Freundlich parameters
qmax (mg.g-1)
b (L.mg-1)
r2
KF (mg.g-1)
n (g.L-1)
r2
25.19
0.04
0.999
10.32
7.58
0.964
It is evident from the corresponding correlation coefficients (r2) that the
Langmuir model fits the data much better than Freundlich model. With the increase
of concentration range from < 5-< 120 mg.L-1 to < 25-< 700 mg.L-1, the changes in
qmax was observed as from 23.09 mg.g-1 to 25.19 mg.g-1. The values of “n” in
Freundlich model being identical (see Table 4.2 and 4.10) indicating similar
adsorption pattern. Thus it was reconfirmed that in true adsorption condition (at pH
below surface precipitation as described earlier as pHabr), the adsorption of Cu(II) on
carbon was resulted from the direct contact with surface functional groups. The
necessity of direct contact restricts the adsorption up to monolayer coverage. The
chemisorption of metal ions on the acidic surface of activated carbon was well
demonstrated in SCF model. In the present study, chemisorption is supported from
the better fitting of the data to Langmuir model as well as from the evidence of
surface complexation, namely inner-sphere complexation (see Figure 4.13).
Dastgheib and Rockstraw (2001) indirectly proved that adsorption of Cu(II) on their
prepared H3PO4 activated carbon was resulted from chemisorption. They tried to
leach out the adsorbed Cu(II) through extensive shaking of the Cu(II) loaded carbon
for 2 days in deionized water and found no copper species to be desorbed.
The effect of pH on the binary adsorptions of Cu(II) and Ca(II) were tested
conducting the binary adsorption also at pH 3 followed by comparing the adsorption
behaviour with the constructed isotherms (see Figure 4.12 ) of Cu(II) and Ca(II) for
binary adsorption at pH 5. The Langmuir adsorption isotherms of Cu(II) and Ca(II)
were determined for the adsorption in binary solute solution while initial
concentrations of Cu(II) and Ca(II) were varied as 26 ± 1 to 150.9 and 20.4 ± 0.7 to
109 ± 3 mg.L-1 respectively at a constant initial pH of 3. The best-selected activated
121
carbon, PSW-P-ad-500 was used as an adsorbent keeping its concentration constant
at 2 g.L-1. The constructed Langmuir adsorption isotherms of Cu(II) and Ca(II) for
pH 3 are presented in Figure 4.15 and the model parameters in Table 4.11.
Adsorption
capacity, q e mg/g
25
A
20
15
Cu-binary-pH 3
10
Ca-binary-pH 3
5
0
0
20
40
60
80
100
120
100
120
Equilibrium concentration, Ce mg/L
Ce/qe
6
B
4
2
Cu-binary-pH 3
0
0
20
40
60
80
Equilibrium concentration, Ce mg/L
Ce/qe
60
C
40
Ca-binary-pH 3
20
0
0
20
40
60
80
100
120
Equilibrium concentration, Ce mg/L
Figure 4.15
Langmuir non-linear (A) and linear (B and C) isotherms of Cu(II) and
Ca(II) for the adsorption in binary solute solution while concentrations of Cu(II) and
Ca(II) were varied as 26-151 and 20-109 mg.L-1 respectively keeping initial pH and
adsorbent concentration constant, at pH 3 and 2 g.L-1 respectively.
122
Table 4.11 : The Langmuir model parameters of Cu(II) and Ca(II) for the
adsorption in binary solution at initial pH 3 while initial concentrations of Cu(II) and
Ca(II) were varied as 26-151 and 20-109 mg.L-1 respectively.
Metal ion
Langmuir model parameters
(in binary)
qmax (mg.g-1)
b (L.mg-1)
r2
Cu(II)
22.78
0.111
0.991
Ca(II)
2.50
0.027
0.742
The values of correlation coefficients of the Langmuir isotherms indicate that
the model fits only the data of Cu(II) very well. The qmax-binary of Cu(II) at pH 5 was
obtained as 23.15 mg.g-1 (see Table 4.9) while that of Cu(II) at pH 3 as 22.78 mg.g-1.
That is, the change of pH has virtually no/little effect in the adsorption of Cu(II) in
presence of Ca(II). The ratio of qmax-binary to qmax-single of Cu(II) (Tables 4.11 and 4.2)
at pH 3 being 1.07 indicating slight promotion in Cu(II) adsorption by the competing
ion according to equation (4.2). With the increase of concentration or pH (from 3 to
5), affinity towards Cu(II) increased resulting an unfavourable adsorption of Ca(II).
4.2.3
Ternary Solute Adsorption Equilibria : Competitive Adsorption of
Cu(II), Ni(II) and Pb(II) (Cu-Ni-Pb)
The interaction of metallic species with living systems and adsorbents can be
demonstrated by the properties of metal ions as Lewis acids. Thus the classification
of metals by their Lewis acidity would be consistent in their remediation approach.
A detailed description of this type of classification of metal ions that included a
periodic table (see Figure 1.1) would be found in Chapter 1. The competitive effect
among borderline (class AB) and soft (class B) cations was studied from the
adsorption in a solution containing ternary solute of Cu(II), Ni(II) and Pb(II). Figure
4.16 shows the Langmuir (A and B) and Freundlich (C) adsorption isotherms of
Cu(II), Ni(II) and Pb(II) while their initial concentrations were varied in the ternary
solution as < 25 - ≤ 600 mg.L-1 and the initial pH was kept constant at 5.
123
Adsorption capacity,
qe mg/g
25
A
20
15
Pb(Cu-Ni-Pb)
10
Cu(Cu-Ni-Pb)
Ni(Cu-Ni-Pb)
5
0
0
100
200
300
400
500
600
Equilibrium concentration, Ce mg/L
120
B
Ni(Cu-Ni-Pb)
90
Ce/qe
Cu(Cu-Ni-Pb)
Pb(Cu-Ni-Pb)
60
30
0
0
100
200
300
400
500
600
Equilibrium concentration, Ce mg/L
1.6
C
log q e
1.2
0.8
Pb(Cu-Ni-Pb)
0.4
Cu(Cu-Ni-Pb)
Ni(Cu-Ni-Pb)
0
0.5
1
1.5
2
2.5
3
log Ce
Figure 4.16
Langmuir non-linear (A), linear (B) and Freundlich (C) isotherms of
ternary solute, Cu(II), Ni(II) and Pb(II), by the best selected carbon, PSW-P-ad-500
at initial pH 5, while their initial concentrations were varied as < 25-≤ 600 mg.L-1.
(Ni samples were analysed for initial concentration range, < 25 to 305 ± 2 mg.L-1).
124
In the single solute adsorption of Cu(II), initial concentrations were varied
from < 5 to < 120 mg.L-1 while various activated carbons were used as adsorbents
(see Figure 4.1). For the initial concentration of ca 5 mg.L-1, no detectable residual
concentration was found in the equilibrated solution by prepared activated carbons.
Thus, Cu(II) adsorptions in single solute solutions by prepared carbons had
determined the equilibrium data only for the variations in initial concentrations from
< 25 to < 120 mg.L-1. This is why, the effect of competing ions, such as Ni(II) and
Pb(II), on Cu(II) was also studied by constructing ‘ternary-Cu’ Langmuir isotherms
(see Figures 4.12A-B) for the same concentration range as mentioned above as < 25
to < 120 mg.L-1. The model parameters of the isotherms of Ni(II), Pb(II) in ternary
(see Figure 4.16) and Cu(II) in single as well as in ternary (see Figures 4.12, 4.14
and 4.16) solutes solution are presented in Table 4.12. To compare the removal
efficiencies in the low concentration, the equilibrium concentrations, adsorption
capacities and fractions removal of cations (Cu, Ni, Pb) in ternary and single solutes
solution for the initial concentration of < 25 mg.L-1 are presented in Table 4.13.
Table 4.12 : Langmuir and Freundlich model parameters for the isotherms of Ni(II),
Pb(II) in ternary and Cu(II) in single and ternary solutes solutions at initial pH 5.
Metal ion
Langmuir parameters
(concentration
mg.L-1)
Cu-single
(<25-<120)
Cu-single
(<25-<700, pH 4.5)
Cu-ternary
(<25-<120)
Cu-ternary
(<25-≤600 )
Ni-ternary
(<25-≤600)
Pb-ternary
(<25-≤600)
qmax
b
Freundlich parameters
r2
mg.g-1 mmol.g-1 L.mg-1
KF
n
mg.g-1
g.L-1
r2
23.09
0.363
0.59
0.999
13.16
7.38
0.989
25.19
0.396
0.04
0.999
10.32
7.58
0.964
15.85
0.249
0.12
0.993
5.50
4.55
0.994
18.66
0.294
0.05
0.997
5.87
5.04
0.988
2.68
0.046
-0.13
0.968
11.68
-3.70
0.904
21.74
0.105
0.05
0.993
5.70
4.40
0.875
125
Table 4.13 : The initial and equilibrium concentrations, adsorption capacities and
fractions removal of Cu(II), Ni(II), Pb(II) in the single and ternary solutes solutions
by the best-selected carbon, PSW-P-ad-500 at initial pH 5 and at adsorbent
concentration of 2 g.L-1.
Metal ion
Concentrations (mg.L-1)
Adsorption
Removal
Initial
Equilibrium
capacity (mg.g-1)
(%)
Cu-single
23.5 ± 0.1
0.43 ± 0.05
11.53 ± 0.04
98.2 ± 0.4
Cu-ternary
23.0 ± 0.2
6.5 ± 0.5
8.3 ± 0.2
72 ± 2
Ni-single
30.6 ± 0.3
6.7 ± 0.1
12.0 ± 0.2
78.1 ± 0.6
Ni-ternary
22.2 ± 0.1
7.3 ± 0.3
7.5 ± 0.2
67 ± 1
Pb-single
23.06 ± 0.1
ND
11.80 ± 0.05
100
48.7 ± 0.2
2.2 ± 0.7
23.3 ± 0.3
96 ± 2
22.26 ± 0.07
6.7 ± 0.6
7.8 ± 0.5
70 ± 4
Pb-ternary
The favourable adsorptions of Cu(II) and Pb(II) in ternary solutions are
realized from the considerably high qmax and the higher values of “n” (see Table
4.12). But the competitive effect of Cu(II) on the adsorption of Pb(II) is much more
suppressive as the ratio of qmax-ternary (21.74 mg.g-1) to qmax-single (63.69 mg.g-1) is
determined as 0.34. Therefore it is speculated that a larger extent of Pb(II) was
adsorbed onto the carbon surfaces while the lesser amounts into the basal surfaces
(inter-layers in basal planes). The presence of Cu(II) as competing ion, impeded the
adsorption of softer Pb(II) greatly onto the carbon surfaces. Thus the adsorption of
Pb(II) in ternary solution is assumed to be resulted solely from the adsorption into
the basal surfaces. The total adsorption capacities of multi solute in ternary solution
(0.445 mmol.g-1) exceeded that of Cu(II) in single solute solution (0.393 mmol.g-1)
(see Table 4.12). This finding also supports the assumption of the adsorption of
Pb(II) into the basal structural layers. The obtained total multi solutes adsorption
capacity in this study is comparable to that of one of the best acid activated carbons
(0.5 mmol.g-1) prepared by John, Marshall and Toles (1998).
According to the concept of HSAB (Pearson, 1968), soft Lewis base such as
basal structural unit (soft site) of carbon prefers soft Lewis acid such as Pb(II) while
126
harder site (surface oxide group- Lewis base) prefers relatively harder ion, Cu(II)
(Lewis acid). The adsorption of Pb(II) on both harder and softer sites of carbon
reveals that Pb(II) is a marginally soft ion. This is why, while it is found in some
literature (Ahrland, Chatt and Davies, 1958) as borderline, other literature (IUPAC,
2002) classified it as soft cation (Class B). The interferences of Pb(II) and Ni(II) on
Cu(II) adsorption is little as the ratio of qmax-mix (concentration range <25-<120
mg.L-1) to qmax-single (<25-<120 mg.L-1) has the value 0.69. This finding is consistent
with other researchers (Johns, Marshall and Toles, 1998; Chen and Wang, 2000).
With the increase of initial concentration range of Cu(II), the ratio of qmax-ternary (<25
to ~ 600 mg.L-1) to qmax-single (<25 to <700 mg.L-1) increases to 0.74 that reveals an
increasing of affinity towards Cu(II) in accordance with the rise up of concentration.
The adsorption capacities of Cu(II), Ni(II) and Pb(II), in single and ternary
solutes solutions, are identical in the low initial concentrations (see Table 4.13). For
the adsorption in single solute solution, complete adsorptions of Cu(II) (98%) and
Pb(II) (100%) were observed even at a considerably high initial concentration (<25
mg.L-1). The removal of Ni(II) in single solute solution was observed comparatively
low as 78%, which might be due to adsorption in a relatively higher initial
concentration, such as 31 mg.L-1. The removal efficiencies of Cu(II), Ni(II) and
Pb(II) in the low concentration solution obtained from their ternary mixture are quite
similar as 72, 67 and 70% respectively. The concentration of total solutes of Cu(II),
Ni(II) and Pb(II) in the low concentration ternary solution can be expressed as 0.842
mmol.L-1 (see Tale 4.13), while their adsorption capacity as 0.131, 0.128 and 0.038
mmol.g-1 respectively. Since adsorption occurred by the capturing of ionic species
up to the monolayer coverage or to the fraction, comparison in adsorption capacities
are needed to be expressed in moles as mmol.g-1.
-1
adsorption capacity is 0.297 mmol.g
That is total multi solute
which is comparable to the single Cu(II)
adsorption capacity as 0.297 mmol.g-1 for the adsorption in an initial adsorbate
concentration of 0.758 mmol.L-1 (48.2 mg.L-1) (see the data in Table H-1.3 in
Appendix H).
The identical removal capacity of Cu(II) and Ni(II) indicates a
coverage of similar order whereas a very small coverage by Pb(II) results higher
removal fraction because of its very high atomic mass. The higher adsorption
capacities (mmol.g-1) for Cu(II) and Ni(II) illustrates the stronger affinity of acid
activated carbon towards borderline metal ions in an order Cu(II) > Ni(II) >> Pb(II).
127
Here it is seemed that the aspect of complete adsorption has obviously been
disturbed by the presence of competing ions. But if we consider the fact that the
adsorbent concentrations for both single and multi solutes solutions were maintained
constant as 2 g.L-1, then these removal efficiencies (67 – 72 %), no doubt, can be
considered as excellent. It is known that, in real practice, the concept of complete
removal (or removal to a safe level) of target species is achieved through the
adjustment of adsorbent concentration with the influent water quality and quantity.
The theoretical minimum adsorbent concentration for the complete adsorption of
Cu(II), Ni(II) and Pb(II) in their ternary solute solution by PSW-P-ad-500 (best
adsorbent) is expected (according to the obtained data for single solute adsorption) to
be 6 g.L-1 for an initial adsorbate concentration of < 25 mg.L-1.
Here it is also important to note that while the maximum contaminant level
(MCL) of Cu(II) in drinking water lies in the order of mg.L-1 level, that of Ni(II) and
Pb(II) are fall in the µg.L-1 level (see Appendix A). Therefore further study is
needed to optimize the adsorbent concentration, for the treatment of Ni(II) and Pb(II)
to a safe level, through the determination of the strength of the equilibrated solution
in µg.L-1 level. A gradual runaway in the adsorption capacity of Ni(II) (see Figure
4.16A) with the increase of equilibrium concentration, which is a function of initial
concentration, indicates the very selective adsorption of Cu(II) over Ni(II).
Although it seemed a problem, the solution of this phenomenon is also laid in
previous discussion which suggests the computed tailoring in the quantity of
adsorbent to be used for developing a proper adsorbent concentration according to
the contents of target ions and operating flow/volume of treating water. As long as
the adsorbent remains in an unsaturation level (which depends on the amount of
adsorbent and on the quality of water) to the target adsorbates, no observable
interferences are expected.
This fact has already been illustrated through the
adsorption data in the low concentration solution (see Table 4.13).
In actual
operation, the maximum adsorption capacity cannot be achieved due to breakthrough
of the least favoured target adsorbate while treating water with adsorption based
process.
Thus the adsorption selectivity order, for the adsorption in ternary solution,
can be expressed in two series: one is by the fraction/percent removal onto the
128
adsorbent up to an unsaturated level (fraction coverage) by dint of low initial
concentration of target ions (see Table 4.13) as: Ni(II) ≤ Pb(II) ≤ Cu(II), while
another is by the respective qmax (Table 4.12), which was obtained through complete
coverage of adsorbent by adsorbates, as: Ni(II) << Pb(II) < Cu(II). The comparison
in qmax was made in the unit of mmol.g-1 instead of mg.g-1. This finding is in
agreement with Irving-Williams order (cited by Shriver, Atkins and Langford, 1991)
which describes the stability of inner-sphere complexes of transition metal cations
as: Mn2+ < Fe2+ < Co2+ < Ni2+ < Cu2+ > Zn2+.
A relative affinity series can also be drawn in an ascending order from the
observed qmax (in mmol.g-1) of Cu(II), Ni(II), Zn(II) and Pb(II) for their adsorption in
single solute solution as: Zn(II) << Pb(II) < Ni(II) < Cu(II). Yantasee et al. (2004)
obtained two affinity series, as: Pb >> Cu > Cd > Ni on commercial activated carbon
and as: Cu >> Pb > Ni > Cd on amine functionalized carbon (NH2-AC). In this
study, the observed qmax of Cu(II), Ni(II), Zn(II) and Pb(II), for the adsorption in
single solute solution, was calculated in the unit of mmol.g-1 as 0.40, 0.33, 0.15 and
0.31 respectively.
4.2.4 Reuse of Cation-Exhausted Activated Carbon
The capability of reusing of cation-exhausted activated carbon was studied
through regeneration with acid solution.
According to surface reactions, the
lowering of pH, at below pHZPC, protonates the hydrous surface of carbon resulting a
repulsive force towards cation. This is the basic concept of using acid solutions as
eluents in the regeneration of cation-exhausted adsorbent. Since the adsorption of
experimental metal cations follows Irving-Williams series that is based on the
stability of transition metal cations complexes, Cu(II) is to be bonded onto the
surface of acid activated carbon most tightly. Therefore, a sample of spent carbon
exhausted by Cu(II) was taken as representative of borderline and soft metal cations
exhausted carbon. In this study, 10 % solution of HCl was used in the desorption of
once adsorbed Cu(II). Copper adsorbed (see Table H-6.1 in Appendix H) on carbon,
in the concentration range from < 25 to < 700 mg.L-1, was subjected to desorb using
129
50 mL of HCl solution as adsorption was carried out using 50 mL of Cu(II) solution.
The Cu(II) loaded carbon samples were thoroughly washed with DDW, then dried in
an oven at 110°C and cooled prior to adding acid solution as eluent. The washing
with DDW removes all adhered metal ions on the surface of adsorbent and vessel.
The Langmuir adsorption-desorption isotherms of Cu(II) are depicted in Figure 4.17,
while the model parameters are tabulated in Table 4.14. The adsorption-desorption
equilibrium data for the adsorption in initial concentrations of 191 ± 1 and 693 ± 4
were not plotted in the isotherms, because the amount of desorbed Cu(II) is not
consistent with its adsorbed amount (see Table H-8.2 in Appendix H).
Adsorption/desorption
capacity,
qe mg/g
25
20
15
Adsorption
10
Desorption
5
0
0
100
200
300
400
500
Equilibrium concentration Ce or (C0-Cdesorb), mg/L
Ce/qe or (C0-Cdesorb)/qe
20
16
12
8
Adsorption
4
Desorption
0
0
100
200
300
400
500
Ce or (C0-Cdesorb) mg/L
Figure 4.17
Langmuir non-linear and linear adsorption-desorption isotherms of
Cu(II), while adsorbed at varying concentration, < 25 to < 700 mg.L-1.
130
Table 4.14 : Langmuir model parameters for the adsorption-desorption of Cu(II).
Adsorptiondesorption
Recovery
Langmuir parameters
r2
qmax
b
mg.g-1
L.mg-1
Adsorption
24.21
0.09
0.999
Desorption
23.58
0.10
0.999
(from qmax )
%
97.4
The qmax obtained from adsorption-desorption isotherms are identical while
recovery of adsorbed Cu is determined as 97.4 %. This finding is comparable with
other researcher. In comparison, Dastgheib and Rockstraw (2001) recovered above
98% of adsorbed Cu(II) through regeneration of the exhausted carbon using 10%
HCl.
4.2.5
Adsorption Studies : Cr(VI)
Adsorption studies of Cr(VI) were divided into three parts as i) selection of
adsorbents through suitability test, ii) the effect of pH and iii) the effect of initial
concentration. All the analytical data of Cr(VI) before and after adsorption, to
determine the equilibrium adsorption capacity as well as to calculate the model
equations/parameters, are presented in Appendix I.
4.2.5.1 Selection of Prepared Adsorbents
Since pHZPC of prepared acid activated carbons were obtained as ca. 3 (see
Table 3.5 in section 3.2.2.3 in Chapter 3), the surface functional groups deprotonate
at above this pH and thus expose negatively charged surface sites (see equations
(1.5) and (1.6) in Chapter 1) over a wide range of pH. This is why acid activated
carbons were further modified through the dispersion of HFO that has stronger
131
affinity towards Cr(VI) and arsenic, As(III) and As(V), and that has the capability to
increase pHZPC suitable for anion adsorption. Pierce and Moore (1982) showed the
high affinity of amorphous Fe(OH)3 toward arsenite and arsenate while the reason
was explained for its visualizing a loose and highly hydrated structure, which is
permeable to hydrated ions. The pHZPC of amorphous ferric hydroxide was reported
as 8.5, which causes its surface charge positive (through protonation at any pH
below 8.5) suitable for attracting HCrO4− or other anions through anion exchange
mechanism (Aoki and Munemori, 1982).
Reed, Vaughan and Jiang (2000)
hypothesized that the surface of hydrous ferric oxide impregnated carbon would
behave similarly to a ferric oxide. Since the pHZPC of hydrous ferric oxide was
reported in the literature as between 8.0 - 8.3, they proved their hypothesis by
determining the pHZPC of a Fe(III)-oxide impregnated activated carbon, obtained
from a company- Norit Americas Inc., as 8.2 - 8.7.
In the earlier studies (see section 4.1.1), three activated carbons, PSW-P-500,
CPW-P-500 and PSW-P-ad-500, were selected as representatives of all prepared
activated carbons for the adsorption of metal cations. Since the modification (see
section 2.4 in Chapter 2) consisted of Fe(III) loading first, the above mentioned three
activated carbons were similarly chosen as base materials. The potentials of four
prepared C-HFO composites, their three counterparts (parent/base activated carbon)
and a commercial activated carbon, CAC were compared for the removal of
oxyanions of Cr(VI) from their adsorption equilibrium capacities in a single point
initial concentration, such as < 10 mg.L-1.
The results are presented in
fraction/percent removal (Table 4.15) shows that the modification of carbon to
enhance its potential in removing Cr(VI) was totally unsuccessful as parent carbons
show higher removal efficiency.
Actually hydrated iron oxide supported materials are found in literature for
the selective removal of As(III) and As(V) (Vaishya and Gupta, 2003; DeMarco,
SenGupta and Greenleaf, 2003; Reed, Vaughan and Jiang, 2000). Among modified
carbons, namely C-HFO composites, only one (C-HFOCa-1) shows considerably
high adsorption capacity towards Cr(VI) while all base carbons, PSW-P-500, CPWP-500 and PSW-P-ad-500 show the highest efficiency.
In fact, the maximum
capacity of the latter three prepared activated carbons was not reached, because the
132
undiluted equilibrated solution (10 mL) by these carbons, the pH of which was
adjusted to 1 using 0.5 mL of concentrated H2SO4, gave no colour with 1,5diphenylcarbazide solution.
Thus, three prepared acid activated carbons, one
composite, C-HFOCa-1 and the typical commercial activated carbon, CAC were
selected for subsequent adsorption studies. Although C-HFOCa-1 and C-HFO-3
have similar BET surface area (671 m2/g), they show very different in removing
Cr(VI).
Table 4.15 : The adsorption capacities of various adsorbents in the removal of
Cr(VI) while initial pH was obtained without adjustment as 4.79 and adsorbent
concentration was maintained as 2 g.L-1.
Adsorbent
Concentrations (mg.L-1)
Removal (%)
Initial
Final
C-HFO-1
8.13 ± 0.04
8.03 ± 0.08
1.7 ± 1
C-HFO-2
8.13 ± 0.04
7.87 ± 0.07
3±1
C-HFO-3
8.13 ± 0.04
7.96 ± 0.05
2 ± 0.4
C-HFOCa-1
8.13 ± 0.04
1.81 ± 00
78 ± 0.1
CAC
8.13 ± 0.04
1.87 ± 0.01
77 ± 0.03
PSW-P-500
8.13 ± 0.04
ND
100
CPW-P-500
8.13 ± 0.04
ND
100
PSW-P-ad-500
8.13 ± 0.04
ND
100
ND= Non detectable
4.2.5.2 The Effect of pH
It was demonstrated in the surface reactions, (1.5) and (1.6) in Chapter 1, that
an increase in solution pH releases protons from the surface exposing more negative
sites (SO-) to bind more cations. On the contrary, a decrease in pH protonates the
surface exposing more positively charged sites on carbon to bind more anions. The
effect of pH on the adsorption of oxyanions of Cr(VI) onto the various selected
133
adsorbents was studied at varying pH from 3 to 8 while initial concentrations were
kept constant at about 40 mg.L-1. The results are shown by plotting the adsorption
capacity against pH (Figure 4.18).
PSW-P-500
CPW-P-500
PSW-P-ad-500
C-HFOCa-1
CAC
Adsorption capacity,
qe mg/g
20
17
14
11
8
5
2
3
4
5
6
7
8
pH
Figure 4.18
Adsorption capacity of Cr(VI) by various selected adsorbents while
pH were varied from 3 to 8 keeping initial concentration constant at about 40 mg.L-1.
All adsorbents but modified ones show identical adsorption capacity at pH 3
while the capacities decrease with the increase of pH. The adsorption capacities of
three prepared acid activated carbons remain almost similar over the range of pH 3
to 5. Here it is important to note that the residual Cr(VI) concentrations in the
equilibrated solutions by the three prepared carbons for pH 3 were not detectable.
The adsorption capacity of the commercial activated carbon was greatly decreased
with the increase of pH from 6 to 8. Although the modification of acid activated
carbons into C-HFO composites was made to bind more oxyanions such as Cr(VI)
on the protonated surface of hydrated iron oxide, its capacity is found to be lower.
All prepared adsorbents under the experiments have significantly higher
adsorption capacity than that of commercial ones over the pH range, 6 - 8. The
adsorption capacities of one of the activated carbons, PSW-P-ad-500 and of a
modified adsorbent, C-HFOCa-1, remained constant with the increase of pH from 6
to 8. This consistency in adsorption capacity, especially in the near neutral to slight
134
alkaline region, reveals stronger bond strength to bind Cr(VI) on their protonated
surface sites. Thus it is supposed that the dispersion of HFO enhanced the bond
strength towards Cr(VI) with the modified carbon while the lower capacity might be
resulted from the lower number of active HFO sites (see Table 3.7). Ultimately, due
to lower adsorption capacity of modified carbons for Cr(VI), prepared acid activated
carbons are considered better adsorbents over modified one and no more adsorption
experiments were carried out with modified carbons/C-HFO composites for Cr(VI).
The adsorption capacity of an activated carbon, PSW-P-500 is significantly higher
than that of all other adsorbents.
The performance of prepared adsorbents in
removing Cr(VI) can be expressed in a series for the entire experimental pH range
as: C-HFOCa-1 << PSW-P-ad-500 < CPW-P-500 < PSW-P-500 (see Figure 4.18).
This order is the reverse of a series obtained for removing cations, namely Ni(II) as:
CAC << PSW-P-500 < CPW-P-500 << PSW-P-ad-500. Thus it is observed that
while the removal of cations increases with the surface acidity (consistent with SCF
model), the adsorption of oxyanion of Cr(VI) decreases with acidity. The acid
activated carbon, PSW-P-ad-500 being the most acidic one (lowest pHZPC), shows
the lower adsorption capacity. This finding is in agreement with Aggarwal, Goyal
and Bansal (1999) who showed acidic/oxidized surface groups reduced the
adsorption of Cr(VI). However, the lower adsorption capacity of CAC than acid
activated carbons, at least at higher pH, is not clear. Although PSW-P-500 is the
best, even then, the activated carbon, PSW-P-ad-500 was selected as a better choice
for its uniqueness in removing metal cations as well as for its considerably higher
Cr(VI) adsorption capacity. The optimum pH was determined as 3 (see Figure 4.18).
4.2.5.3 Effect of Initial Concentration and Data Fitting to Models
The adsorption behaviour of a selected acid activated carbon, PSW-P-ad-500
was compared with commercial ones by fitting the adsorption equilibrium data to
Langmuir and Freundlich models. Figure 4.19 shows the non-linear and linear
Langmuir, and Freundlich adsorption isotherms.
The isotherms reveal that the
adsorption capacity of acid activated carbon is higher only in the low concentration
135
region while that of commercial activated carbon is higher in the high concentration
region. The model parameters are shown in Table 4.16.
Adsorption capacity,
qe mg/g
70
A
60
50
40
30
CAC
20
PSW-P-ad-500
10
0
0
1.6
15
30
45
60
Equilibrium concentration, Ce mg/L
75
90
B
Ce/qe
1.2
0.8
PSW-P-ad-500
CAC
0.4
0
0
15
30
45
60
75
90
Equilibrium concentration, Ce mg/L
2
C
log q e
1.8
1.6
1.4
CAC
PSW-P-ad-500
1.2
1
-0.2
0.4
1
1.6
2.2
log Ce
Figure 4.19
Langmuir non-linear (A), Linear (B) and Freundlich (C) adsorption
isotherms of Cr(VI) by activated carbons, CAC and PSW-P-ad-500, at initial pH 3
while initial concentrations were varied from 40 - < 200 mg.L-1.
136
All the data are fitted very well to Langmuir and Freundlich models. The
higher adsorption capacity in the low concentration indicates stronger bond strength
to bind Cr(VI) as well as higher suitability of the adsorbent to be used in the drinking
water purification purposes. The equilibrium concentrations and fractions in the
removal of Cr(VI) in the low concentration are shown in Table 4.17 to illustrate the
suitability of acid activated carbon over commercial ones especially while to be used
in drinking water purification.
Table 4.16 : Langmuir and Freundlich model parameters for the adsorption of
Cr(VI) on PSW-P-ad-500 while initial concentrations were varied from 40 - < 200
mg.L-1 at an optimum pH 3, keeping adsorbent concentration constant at 2 g.L-1.
Adsorbent
Langmuir parameters
qmax
b
mg.g-1
L.mg-1
PSW-P-ad-500
46.30
0.334
CAC
70.92
0.092
r
Freundlich parameters
2
r2
KF
n
mg.g-1
g.L-1
0.990
17.01
3.22
0.996
0.984
28.57
8.91
0.807
Table 4.17 : The initial and equilibrium concentrations, and fractions removal of
Cr(VI) by various activated carbons in the low concentration at an optimum pH 3.
Adsorbent
PSW-P-ad-500
CAC
Concentrations (mg.L-1)
Removal (%)
Initial
Equilibrium
39.9 ± 0.2
ND
100
59.4 ± 0.2
0.7 ± 0.0
98.9 ± 0.0
39.9 ± 0.2
2.5 ± 0.3
93.7 ± 0.8
59.4 ± 0.2
4.4 ± 0.1
92.6 ± 0.2
137
4.2.6
Adsorption Studies : Arsenic
The suitability of a HFO supported composite material, C-HFO-1 in
removing oxyanions of As(V) and non-ionic species of As(III) was testified followed
by comparison with its parent material, PSW-P-500. The activated carbon PSW-P500 and its modified ones, C-HFO-1, considered as typical of other prepared
activated carbons and C-HFO composites respectively, were chosen as adsorbents to
be studied. The solution prepared from As(V) salt was considered to be contained
mostly As(V) species while As(III) solution to be contained mainly As(III) species.
The adsorption equilibrium data were obtained for the respective adsorptions of
As(V) and As(III) in a single point initial concentration, such as < 2 mg.L-1.
SenGupta (2002) showed that adsorption of As(V) decreased with the
increase of pH above 6. Other literature also showed that adsorption of As(V)
decreased with the increase of pH while that of As(III) maximized at around pH 7
(Reed, Vaughan and Jiang, 2000).
In this study, the adsorption experiments were conducted at constant initial
pH and to optimize arsenic removal capacity, the initial pH of As(V) and As(III)
solutions were adjusted to 5.5 and 7.6 respectively. The concentration of adsorbent
of C-HFO-1 was maintained at 1 g.L-1 while that of parent carbon, 2 g.L-1.
Adsorption equilibrium capacities of the adsorbents are determined (see Appendix J)
and the results are presented in Figure 4.20.
The composite, C-HFO-1, shows considerably high adsorption capacity of
As(III) and As(V) than that of unmodified ones. This preliminary study thus proves
the successful modification of activated carbon. Therefore a detailed study is needed
to testify the suitability of other C-HFO composites as well as to predict the
adsorption behaviour by obtaining equilibrium data at varying pH and at varying
concentration either of adsorbate or adsorbent.
While the modification was found unsuccessful for Cr(VI), it is found
considerably very good for As(III) and As(V). The fact needs to examine. Since
activated carbon showed very high affinity towards Cr(VI), the reduction of surface
138
area by the dispersed HFO (see Table 3.7 in Chapter 3) ultimately reduced the active
sites of carbon for Cr(VI). Although, instead of lowering of these active sites, some
more reactive sites of HFO were developed (see Figure 3.15 in Chapter 3), they were
poor in number while compared with highly porous activated carbon. As a result,
overall adsorption capacity for Cr(VI) was greatly reduced in modified carbon.
In case of arsenic adsorption, the reduction of surface area in carbon has
minor effect because of its very low affinity toward As(V) and As(III). Thus, lesser
amount of HFO sites was found enough to exceed the adsorption capacity of parent
carbon.
1.6
Adsorption capacity, mg/g
1.4
1.2
As(V)
As(III)
1
0.8
0.6
0.4
0.2
0
C-HFO-1
PSW-P-500
Adsorbents
Figure 4.20
Adsorption equilibrium capacities of a HFO composite and its counter
activated carbon for the species of As(V) and As(III), while adsorptions were carried
out in the solutions of As(V) and As(III) with constant initial concentrations, 1.667
and 2.158 mg.L-1 and with constant initial pH, at 5.5 and 7.6 respectively.
To compare the results presented in Figure 4.20, the adsorption capacities,
for As(III) and As(V), of a Fe(III) impregnated carbon (Fe-GAC) and its counter
carbon (virgin GAC) (adsorbents were obtained by researchers from Norit Americas
139
Inc.) are tabulated in Table 4.18 (Reed, Vaughan and Jiang, 2000). It is observed
that the adsorption of As(III) is more favourable than that of As(V) on modified and
unmodified carbon.
The findings of Reed, Vaughan and Jiang (2000) is consistent with the results
shown in Figure 4.20. The adsorption capacity of C-HFO-1 is much lower than that
of Norit’s modified one. Prepared carbon, PSW-P-500 (base material) is better in
removing As(V) (0.4 mg.g-1) than that of virgin GAC (base material) (0.09 mg.g-1),
while virgin GAC is superior (0.7 mg.g-1) than that of prepared carbon (0.5 mg.g-1)
in removing As(III).
Table 4.18 : Aqueous and solid phase equilibrium concentration of adsorbents
(adsorption capacity) while adsorbate concentration was 1 mg.L-1, pH 7 and
adsorbent concentration 0.2 g.L-1 (Reed, Vaughan and Jiang, 2000).
Adsorbent
Virgin-GAC
Fe-GAC
Virgin-GAC
Fe-GAC
Adsorbate
As(III)
As(V)
Equilibrium concentration
Adsorption capacity
Ce (mg.L-1)
qe (mg.g-1)
0.83
0.73
0.22
4.67
0.98
0.09
0.10
4.50
Daus, Wennrich and Weiss (2004) studied the adsorption capacities of
various material for the removal of arsenic and observed the sorption capacity of
zirconium-loaded activated carbon and iron hydroxide granulates (GIH), from
column adsorption experiments, as 2.8 and 2.3 mg As.g-1 adsorbent. They also
observed a disadvantage of using zirconium-loaded activated carbon since it released
traces of zirconyl ions which are classified as toxic.
Thus they advised that
zirconium-loaded activated carbon is not suitable for use with drinking water in spite
of its high efficiency and fast removal capacity of arsenate than that of GIH.
Badruzzaman, Westerhoff and Knappe (2004) determined the adsorption
capacity of granular ferric hydroxide for arsenate as 8 mg.g-1. But the important
140
limitation of hydrated iron oxides is that they are synthesized as fine particles which
are physically very poor in mechanical strength and practically unusable in a fixed
bed. On the contrary, dispersed hydrated iron oxides are capable to run in a fixed
bed column.
Recent investigations using extended X-ray absorption fine structure
(EXAFS) spectroscopy confirmed that As(III) and As(V) species are selectively
bound to the Fe(III) oxide surface through formation of inner sphere complexes
(Manning, Fendorf and Goldberg, 1998).
DeMarco et al. (2003) demonstrated the ligand characteristics of As(III) and
As(V) and also illustrated the selective adsorption capability of protonated and
electrically neutral iron oxides toward As(III) and As(V) species through Lewis acidbase interaction.
Thus the mechanism of adsorption of As(V) and As(III) species as
ligands/Lewis base onto the surface of hydrated ferric oxide (HFO) in C-HFO-1 that
serves as Lewis acid is presented in Figure 4.21 that shows inner-spherecomplexation. It is evident from mechanism as shown in Figure 4.21 that the
adsorption of As(V) is more preferable than that of As(III).
But the results (see Figure 4.20) that showed relatively high adsorption
capacity of activated carbon or its modified ones for As(III) is contradictory to the
described mechanism. Therefore it is speculated that a repulsive interaction between
negatively charged surface sites of carbon and oxyanions of As(V) might exist that
prevent the latter namely As(V) from the proximate contact onto HFO sites. An
increase in solution pH, exposing more negative sites on carbon, intensifies this
repulsive force resulting a gradual decrease in adsorption capacity for As(V). Since
As(III) is neutral, no such interaction is expected and thus adsorption is maximized
at about neutral pH. This is consistent with the reports of literatures (SenGupta,
2002; Reed, Vaughan and Jiang, 2000).
141
Solid
Interface
Water
As(V), anionic ligand with an ability to
donate two pairs of electrons and thus
able to form bidentate inner sphere
complex (Lewis acid-base interaction).
Outer-sphere complex is due to
Coulombic electrostatic interaction.
O
Fe
OH2
O
O
As
OH2
OH
O
Fe
O
O
OH
Fe
OH2
As
O
As(V), anionic ligand with an ability to
donate a pair of electrons and thus able
to form monodentate inner sphere
complex (Lewis acid-base interaction).
Outer sphere complex is due to
coulombic electrostatic interaction.
O
OH
As(III), non-onic ligand with an ability
to donate a pair of electrons and thus
able to form monodentate inner sphere
complex (Lewis acid-base interaction).
O
OH
Fe
OH2
As
O
O
Fe
OH2
Cl-
O
Cl-, SO42- anions are poor ligands
and able to form only outer-sphere
complexes
by
Coulombic
electrostatic interaction.
Fe
OH2
SO42Coulombic attraction
Lewis acid-base interaction
Figure 4.21
An illustration of selective bond formation of dispersed hydrated iron
oxide with As(V) and As(III) species in aqueous phase predominantly through Lewis
acid-base interaction.
142
4.3
Adsorption Dynamics
The kinetics of adsorption is recently doped as adsorption dynamics. The
adsorption process is generally represented by,
A (adsorbate) + surface site
A-adsorbed
(4.5)
where, “A-adsorbed” is the product.
The rate of adsorption, being elemental process, is first order with respect to
the concentration of “A” in the solution, [A] and is also proportional to the amount
of active surface sites available for adsorption (Mortimer, 1993). If θ is the fraction
of coverage of the surface by the adsorbate “A”, the rate of adsorption can write as
Rate of adsorption, rad = k ad [ A].(1 − θ )
(4.6)
where, kad = adsorption rate constant.
In batch adsorption experiment, generally, the concentration of solute, [A], is
used in large excess while the adsorbent is in insufficient amount so that the equation
(4.6) takes the form (because rate is then independent towards [A]),
Rate of adsorption, rad = k ad (1 − θ )
(4.7)
If the adsorption proceeds up to the maximum monolayer coverage and
reaches the equilibrium, then we can write the equation (4.7) as
Rate of adsorption, rad =
dq t
= k1ad (q e − qt )
dt
(4.8)
where, qt , is the amount of adsorbed adsorbate mg.g-1, at any time t, and is
equivalent to the fraction of coverage, θ on adsorbent by the adsorbate, and qe is the
same at equilibrium which is equivalent to unit coverage.
Here, the order of adsorption reaction is forced to 1 with respect to (qe − qt )
and unknown to solute and solvent by their large excess presence. Since the order of
143
adsorption is approximated to the first order it is called pseudo-first order. If the
adsorption reaction is approximated to the second order with respect to (qe − qt ) ,
then the equation (4.8) can be written as:
dqt
= k 2 ad (qe − qt ) 2
dt
(4.9)
The constants k1ad and k2ad are the adsorption rate constants for the pseudofirst- and pseudo-second order adsorptions respectively. The differential rate law
equations, (4.8) and (4.9), can be solved to the integrated rate law equations, (4.10)
and (4.11) respectively, by separating the variables followed by an integration within
a limit, t = 0 to t = t for which qt has the values from qt = 0 to qt = qt. Thus, the
derived equations, (4.10) and (4.11) describe the adsorption capacity as a function of
time.
log(q e − q t ) = log q e −
qt =
t
1 / k 2 ad q e2 + t / q e
k1ad
t
2.303
(4.10)
(4.11)
The equation (4.11) is rearranged to obtain the linear form in the following
equation (4.12) as:
t
1
1
=
+
t
2
qt k 2 ad q e q e
(4.12)
Lagergren (1898) determined the rate of adsorption first time using the
pseudo-first-order equation (4.10) and hence the equation is known as Lagergren
pseudo-first-order model. On the other hand, equation (4.12) was reported by Ho
and Mckay (2002) for the adsorption of divalent cations and known as pseudosecond order model. For pseudo-first order adsorption, plotting log(q e − q t ) against t
gives a straight with a slope, -k1ad /2.303 and intercept log qe whereas for pseudo-
144
second order, plotting t/qt against t gives a straight line with a slope,
intercept,
1
k 2 ad qe2
1
and an
qe
.
4.3.1 Adsorption Dynamics of Cu(II) on PSW-P-ad-500
The kinetics data of Cu(II) were obtained as a function of contact time, while
the initial concentration and pH were kept constant at < 50 mg.L-1 and 5 ± 0.1
respectively, by PSW-P-ad-500 (see Table K-1.1 in Appendix K). The kinetics data
are checked with Lagergren pseudo-first order equation in Figure 4.22 that shows the
data are not fitted from the point of linearity. However, they fitted very well to Ho
and Mckay’s (2000) reported pseudo-second order model (Figure 4.23).
To
determine the experimental value of adsorption equilibrium capacity, qe, a graph, by
plotting qt against t, is also obtained and showed in Figure 4.24.
1.2
1
log q e -qt
0.8
0.6
0.4
0.2
0
-0.2
0
20
40
60
80
Time, h
Figure 4.22
Graph to check kinetics data of Cu(II) with first order kinetic model.
145
5
4
t/qt
3
2
1
0
0
20
40
60
80
Time, h
Graph of pseudo-second order kinetic model for the adsorption of
Figure 4.23
Cu(II) at constant initial concentration, < 50 mg.L-1 and at pH 5.
25
qe mg/g
20
15
10
5
0
0
10
20
30
40
50
60
70
80
Time, h
Figure 4.24
A graph of adsorption capacity vs. contact time for the adsorption of
Cu(II) at constant initial concentration, < 50 mg.L-1 and at pH 5.
146
It is seen from Figure 4.24 that the adsorption approached equilibrium within
20 h while the experimental value of qe (adsorption equilibrium capacity) was
obtained as 20.33 mg.g-1. The correlation coefficient of Lagergren pseudo-first order
equation was low as 0.79. A straight line with a very high value of r2 (> 0.999) was
obtained while plotting t/qt against t. The values of qe and k2ad were determined
respectively from the slope and intercept of the linear graph of pseudo-second order
model (Figure 4.23). The very high value of corresponding correlation coefficient
(r2) as well as the identical values of calculated and experimental qe (19.38 and 20.33
mg.g-1 respectively), indicate that the adsorption of Cu(II) on acid activated carbon
follows pseudo-second order model very well. The kinetic parameters are presented
in Table 4.19.
Table 4.19 : Experimental adsorption equilibrium capacity, qe and kinetic model
parameters for the adsorption of Cu(II) at constant initial concentration, < 50 mg.L-1
and at pH 5, while contact time was varied from 0.5 to 72 h.
Order of
Experimental
Calculated
k2ad
Adsorption
qe (mg.g-1)
qe (mg.g-1)
g.mg-1.h-1
Second
20.33
19.38
0.024
4.3.2
r2
1.000
Adsorption Dynamics of Ni(II), Pb(II) and Cr(VI) on PSW-P-ad-500
The adsorption kinetics data of Ni(II), Pb(II) and Cr(VI) by PSW-P-ad-500
(see Tables K-2.1, K-3.1 and K-4.1 in Appendix K) are best fitted to the pseudo
second-order model (Figure 4.25).
The experimental values of adsorption
equilibrium capacities, qe, are determined from the graphs obtained by plotting qt
against t (Figure 4.26).
The rate constants (k2ad) and calculated adsorption
equilibrium capacities (qe) are determined from the model equation. The kinetic
parameters are shown in Table 4.20. The higher values of corresponding correlation
coefficients and the identical values of calculated and experimental qe indicate that
pseudo second-order model fits all these kinetics data very well.
147
5
4
Ni(II)
Cr(VI)
Pb(II)
t/qt
3
2
1
0
0
20
40
60
80
Time, h
Figure 4.25
Graphs of pseudo-second order kinetic model for the adsorptions of
Ni(II), Pb(II) and Cr(VI) at constant initial concentrations, < 45, 95 ± 1 and 77 ± 1
mg.L-1 respectively, while the initial pH of solutions of Ni(II) and Pb(II) were
adjusted to 5 and that of Cr(VI) to 3.
Table 4.20 : Experimental adsorption equilibrium capacity, qe and kinetic model
parameters for the adsorption of Ni(II), Pb(II) and Cr(VI).
Metal ions
r2
Experimental
Calculated
k2ad
qe (mg.g-1)
qe (mg.g-1)
g.mg-1.h-1
Ni(II)
14.51
15.97
0.052
0.998
Pb(II)
42.71
39.5
0.090
1.000
Cr(VI)
32.76
33.78
0.010
0.997
148
A
Adsorption capacity
qt mg/g
20
16
12
8
4
0
0
20
40
60
80
60
80
60
80
Time, h
B
Adsorption capacity ,
qt mg/g
50
40
30
20
10
0
0
20
40
Time, h
C
Adsorption capacity ,
qt mg/g
40
30
20
10
0
0
20
40
Time, h
Figure 4.26
Graphs of adsorption capacities vs. contact time for the adsorptions of
Ni(II) (A), Pb(II) (B) and Cr(VI) (C).
149
A comparison in the rate constants of various metal ions adsorbed onto acid
activated carbon (PSW-P-ad-500) can be depicted in Figure 4.27 that shows
adsorption of Pb(II) is faster followed by Ni(II), Cu(II) and Cr(VI). Since adsorption
follows pseudo-second order reaction, further investigation is needed to determine
the rate-limiting step. Only after then it could be correlated whether diffusion
through porosity or chemisorption controls the rate.
nd
2 order rate const., K2 ad (g/mg-h)
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
Pb
Ni
Cu
Cr
Metal ions
Figure 4.27
A comparison in the rate constants of various metal ions for the
adsorption onto acid activated carbon, PSW-P-ad-500.
However, Ho and Mckay (2000) described adsorption of various divalent
metal cations onto peat using pseudo-second order model based on the assumption
that the rate-limiting step is the chemisorption. Based on this assumption, they
showed the influence of adsorption equilibrium capacity (qe) on the rate of
adsorption reaction while adsorption equilibrium capacity itself is a function of
adsorbent dose and the nature of adsorbate ion.
Assuming that chemisorption is the rate-limiting step in this experiment, we
can directly relate metal cations of higher qe (Figure 4.28) with slower rate constant
150
(Figure 4.27). Other researchers (Macías-García et al., 2004) also observed similar
correlation as while sulphurization of commercial activated carbon increases the
adsorption capacity it slows down the pseudo-first order rate constant of parent
carbon from 0.903 h-1 to 0.065, 0.042 and 0.034 h-1 by three modified carbons.
0.7
Adsorption equilibrium
capacity, q e mmol/g
0.6
0.5
0.4
0.3
0.2
0.1
0
Cr
Cu
Ni
Pb
Metal ions
Figure 4.28
A comparison in adsorption capacities (qe) of various metal ions
while qe obtained from adsorption equilibrium data for kinetics studies for the
adsorption onto acid activated carbon, PSW-P-ad-500.
4.4
Summary of Thesis
Today’s concern over the presence of inorganic species necessitates a cost
effective and eco-friendly treatment process to remediate these contaminants from
drinking water and wastewater.
Present research trend demonstrates local
agricultural raw materials as promising precursors in the removal of borderline and
soft metal ions. Since Malaysia is moving rapidly towards the concept of safe
drinking water, her huge wastes from agricultural sector, such as oil palm kernel and
151
coconut shells, are needed to be testified for the same. Thus the present work
consisted of adsorbent preparation and modification, followed by subsequent
adsorption studies that determined the suitability of prepared adsorbents as
scavengers for borderline and soft ionic or non-ionic species. The findings of this
work are described in this thesis that contains five chapters including a concise
introduction with literature review and objectives in Chapter 1.
Chapter 2 provides the description of processing of oil palm and coconut
shells to obtain activated carbons suitable for removing borderline and soft metal
ions. This chapter also includes a method of dispersion of hydrated Fe(III) oxide
into activated carbons to modify them into composites of carbon-hydrated iron oxide
(C-HFO) for the selective removal of As(III) and As(V). Here it is important to no
note that the dispersion of hydrated iron oxide into carbon is still proprietary secret
and thus the method is unavailable in the literature.
The effect of preparation variables on the characteristics of activated carbons
were studied extensively while physical characterizations were conducted to
determine porous structure, surface morphology, acidity and proximate analysis.
The porous structure, surface morphology and Fe content of composite adsorbents
were determined to observe the changes occurred led by dispersion of HFO. The
findings of these studies of adsorbents development are presented in Chapter 3.
Chapter 4 shows the suitability of prepared activated carbons for the removal
of Cu(II), Ni(II), Zn(II), Pb(II) and Cr(VI) through adsorption equilibria and kinetics
studies. The adsorption mechanism and selectivity series were ascertained from
multi-solute adsorption. The derivation of an alternative technique, that uses binary
adsorption of Cu(II) and Ca(II), to verify the inner-sphere complexation as cation
adsorption mechanism is certainly a new contribution in this field. Although the
technique is simple and the basis of the idea is originated from the very old concepts
of Lewis acid-base and Pearson’s hard and soft acids and bases, no one found, so far
in literature, to utilize this view before. This chapter also includes the adsorption
studies of As(III) and As(V) on a hydrated iron oxide composite carbon. The
adsorption experimental parameters are well described in Chapter 2.
CHAPTER 5
CONCLUSIONS AND SUGGESTIONS FOR FUTURE STUDY
5.1
Conclusions
Like other nutshell precursors, oil palm shells are found to be a very
promising precursor, while activated with H3PO4, in the remediation of toxic species
of metals. Coconut shells are found as alternative precursors to oil palm shells and
thus grouped in the same category of raw materials. It is observed that acid pretreatment controls volatile evolution. Sulphuric acid pretreatment controls volatile
evolution most yielding a very high yield while non-pretreated precursor restricts it
in a lesser extent resulting relatively a lower yield. Pretreatment of raw material
with H3PO4 extends the impregnation step of H3PO4 activation resulting in the
highest BET surface area, nearly 1500 m2.g-1 and very high total pore volume (0.94
cm3.g-1) in the resulting carbon (CPW-P-500). The BET surface area and total pore
volume of non-pretreated and H2SO4 pretreated activated carbons are found in the
same order.
Activated carbons (CSW-P-500 and PSW-P-500) prepared from wet
feedstock of coconut and oil palm shells at activation temperature of 500°C have the
BET surface area of around 1050 m2.g-1.
All activated carbons prepared from
preheat treated (evaporated to some forms of dryness) feedstock have relatively very
high surface area (> 1300 ≈ 1500 m2.g-1) than that prepared from the wet feedstock
(≈ 1050 m2.g-1). Moderately high BET surface area (1366 m-2.g-1) was obtained in
an activated carbon derived from a feedstock that was evaporated to complete
153
dryness (PSW-P-ed-450) prior to activation in a muffled furnace. Very high BET
surface areas (≈ 1500 m2.g-1) were observed in three activated carbons, one derived
from the apparently-dried feedstock (evaporated the H3PO4 impregnated feedstock at
110°C for 4 h) namely PSW-P-ad-500 (1476 m2.g-1) while the other two were from
H3PO4 pretreated feedstock (1491 m2.g-1) namely CPW-P-500 and from smaller
particle sized precursor (1472 m2.g-1), namely PSWfg-P-450.
The lowering of particle sizes from 1.18 – 2.36 mm to 0.60 – 1.18 mm
increased the BET surface area largely from 1049 m2.g-1 to 1472 m2.g-1 and thus
ultimately indicated a better contact of H3PO4 with more exposed area in smaller
particles, which in turn contributed to a higher microporosity development. This is
consistent with the pattern of N2 adsorption isotherm that shows no hysteresis loop.
An oxygenated atmosphere in a muffled furnace and an activating agent,
such as H3PO4, were found to be very good in the development of surface acidity
that was revealed from FTIR spectrums, Boehm’s titration and pHZPC. The very low
pHZPC of prepared acid activated carbon indicates its feasibility in the adsorption of
metal cations even at low pH. Thus prepared activated carbons (PSW-P-500, CPWP-500 and PSW-P-ad-500) were successfully modified through the introduction of
Fe(III) ion at pH < 2, followed by conversion and dispersion as HFO using
precipitation and thermal treatment methods.
The modified activated carbons
contained 7 – 19% of Fe content in the dispersed phase of HFO while surface areas
were reduced from 41 – 45%.
Since the surface of activated carbon is considered as relatively hard Lewis
base, the metal ions of Cu(II), Ni(II) and Zn(II) as borderline metal ions were
correctly chosen as adsorbate.
Thus the adsorption mechanism and selectivity
among transition metal ions were well demonstrated applying Irving-Williams
(Shriver et al., 1991) order of transition metals (first row d-elements such as Mn, Fe,
Co, Ni, Cu and Zn in the periodic table) for their stability in a complex. On the other
hand, as the basal structural units or graphitic layers of activated carbon are
considered as soft Lewis base (Alfarra, et al., 2004), the adsorption of soft ion such
as Pb(II) on the carbon is also interesting in this study.
154
Three activated carbons (PSW-P-500, CPW-P-500 and PSW-P-ad-500) were
selected, as representatives of all prepared activated carbons having highly
developed surface area (> 1000 m2.g-1), to conduct various adsorption studies.
Single solute adsorptions of various metal cations at varying initial concentrations
and at varying pH indicate that the best adsorption occurred onto an oil palm shell
derived carbon that was obtained from H2SO4 pretreated shell and preheat treated
feedstock (PSW-P-ad-500). The higher adsorption capacity of PSW-P-ad-500 can
be justified from its highly
developed surface area as well as from its well
developed acidic groups as shown in FTIR spectrum as well as from its lower pHZPC
( 3.06) and higher level of Boehm quantified acidic groups (4.8 meq. g-1). A relative
affinity series of experimental cations towards the prepared acid activated carbon is
obtained in an ascending order as: Zn(II) << Pb(II) < Ni(II) < Cu(II). All acid
activated carbons show 4 – 7 fold high adsorption capacities for the experimental
metal cations than that of commercial ones under the condition of this study.
The very selective adsorption of Cu(II) over Ca(II) and moderately high
adsorption capacity of Cu(II) at pH 3, illuminates the mechanism of adsorption as
inner-sphere complexation. The adsorption selectivity order for the adsorption in
ternary solute solution is expressed in two series. One is by the fraction/percent
removal onto the adsorbent up to an unsaturated level, due to low initial
concentration of target ions, as: Ni(II) ≤ Pb(II) ≤ Cu(II), while another is by the
respective maximum monolayer adsorption capacity, as: Ni(II) << Pb(II) < Cu(II).
A study in the regeneration of spent carbon (exhausted by the divalent cation of
copper) indicates the capability of reusing of adsorbents through regeneration with
acid solution.
The adsorption of oxyanions of Cr(VI) is highly pH dependable, that
optimizes at pH 3 with an decreasing trend with the increase of pH. The adsorption
capacity of prepared activated carbon for Cr(VI) exceeds that of commercial ones
only in the low concentration range. This finding thus indicates higher suitability of
the prepared activated carbon over commercial ones while the adsorbent is supposed
to be used in drinking water purification purposes. The modification of activated
carbon through the dispersion of hydrated Fe(III) oxide significantly enhances the
adsorption capacity for As(III) and As(V) than that of parent carbon.
155
The adsorption kinetics data shows that adsorption of Cu(II), Ni(II), Pb(II)
and Cr(VI) onto PSW-P-ad-500 follows pseudo second-order reaction. Since order
of adsorption reaction is determined assuming chemisorption and since according to
this assumption, the rate of adsorption is directly proportioal to unoccupied site (1 θ) which is equivalent to (qe – qt), it is possible to correlate the rate constant with the
maximum monolayer adsorption capacity. A relatively faster adsorption is observed
by Pb(II) (k2ad = 0.090 g.mg-1.h-1) followed Ni(II) (k2ad = 0.052 g.mg-1.h-1), Cu(II)
(k2ad = 0.024 g.mg-1.h-1) and Cr(VI) (k2ad = 0.010 g.mg-1.h-1) onto PSW-P-ad-500.
That is the order of the rate constants of experimental metal ions (Pb > Ni > Cu > Cr)
is opposite to that of their maximum adsorption capacities, Cr(VI) > Cu(II) > Ni(II)
> Pb(II).
The influence of adsorption equilibrium capacity (qe) on the rate of
adsorption reaction thus can be obtained as a function of adsorbent dose and the
nature of adsorbate ion.
5.2
Problems Encountered in Present Study
The prepared activated carbons show the potential in removing borderline
and soft metal ions in single as well as in multi solutes solutions over a wide range of
concentrations and pH.
The capability of selective adsorption as well as
regeneration of spent carbon suits the adsorbent as filtering media in a fixed bed to
be used especially in water treatment for the removal of borderline and soft metal
ions.
The removal of As(III) and As(V) was also found feasible through the
modification of activated carbon using hydrated Fe(III) oxide.
The main limitation in the preparation of activated carbon in this study is that
the atmosphere of muffled furnace was not defined. The leakage of air through the
door controls the furnace atmosphere by design default. The adsorption capacity of
Zn(II) is found lower, therefore, it is questionable whether the acid activated carbon
is suitable for its remediation. The kinetics of adsorption of metal ions seem slow
while adjustment of an optimum adsorbent concentration is needed to solve this
problem. Due to the limitation of analysis at µg.L-1 level by FAAS, further study is
needed that will use graphite furnace or ICP-MS as the analysing tool especially for
156
the determination of residual concentration of trace toxins such as Ni(II), Pb(II) and
Cr(VI).
5.3
Recommendations for Future Study
Since preliminary investigations show the feasibility of H3PO4 activated
carbon as metal ion scavenger, further studies are needed with actual water in a pilot
plant and in a modular unit column to evaluate full-scale drinking or wastewater
treatment. The adsorption studies are also needed to be extended for the remediation
of some other metal ions, namely Fe(II), Mn(II), Hg(II) and Cd(II) etc. as well as for
some organic compounds, namely benzene, toluene, dioxane, acetonitrile, acetone
and methanol.
The removal capacities for As(III) and As(V) were determined by one of the
C-HFO composites. Further studies are needed to predict the adsorption behaviour
as well as to scrutinize the best composite adsorbent. Studies are also needed to
optimize the initial solution pH for As(III) and As(V) adsorption and to observe the
competitive effect of competing ions such as, Cl-, SO42- and PO43-.
Therefore, still some works have to do to develop a modular filtering unit to
be used in a home water condition unit or in a fixed bed in water treatment plant for
a complete solution to the contaminated, minor to trace level, drinking water supply
or source or to the wastewater stream.
REFERENCES
Aggarwal, D., Goyal, M., and Bansal, R. C. (1999). Adsorption of Chromium by
Activated Carbon from Aqueous Solution. Carbon. 37: 1989-1997.
Ahmed, B. U. (2002). Standard Methods of Chemical Analysis for Water and
Wastewater. Dhaka, Bangladesh: Panjeree Publications Ltd.
Ahrland, S., Chatt, J., and Davies, N. R. (1958). The Relative Affinities of Ligand
Atoms for Acceptor Molecules and Ions. Quarterly Reviews of the Chem. Soc.
XII (3): 265-276.
Alfarra, A., Frackowiak, E., and Béguin, F. (2004). The HSAB Concept as a Means
to Interpret the Adsorption of Metal Ions Onto Activated Carbons. Appl. Surf.
Sci. 228: 84-92.
Aoki, T., and Munemori, M. (1982). Recovery of Chromium(VI) from Wastewaters
with Iron(III) Hydroxide-Adsorption Mechanism of Chromium(VI) on Iron(III)
Hydroxide. Water Res. 16(6): 793-796.
Asiah, H. (1993). Kesan Teknik Penyediaan Karbon Teraktif Daripada Tempurung
Kelapa Sawit Terhadap Keliangan dan Keupayaan Penjerapan. J. Teknologi. 22:
51-70.
Babić, B. M., Milonjić, S. K., Polovina, M. J., and Kaludierović, B. V. (1999). Point
of Zero Charge and Intrinsic Equilibrium Constants of Activated Carbon Cloth.
Carbon. 37: 477-481.
Badruzzaman, M., Westerhoff, P., and Knappe, D. R. U. (2004). Intraparticle
Diffusion and Adsorption of Arsenic onto Granular Ferric Hydroxide (GFH).
Water Res. 38: 4002-4012.
Banerjee, S., Majumdar, S., Dutta, A. C., Roy, A. K., Banerjee, S. C., and Banerjee,
D. K. (1976). Active Carbon from Coconut Shell. Indian J. Technol. 14: 45-49.
Bansal, R. C., Donnet J. B., and Stoeckil, F. (1988). Active Carbon. New York:
Marcel Dekker.
158
Benjamin, M. M., Hayes, K. F., and Leckie, J. O. (1982). Removal of Toxic Metals
from Power Generation Waste Stream by Adsorption and Coprecipitation. J.
Water Poll. Control Fed. 54: 1472-1481.
Boehm, H. P. (2002). Surface Oxides on Carbon and Their Analysis: A Critical
Assessment. Carbon. 40: 145-149.
Caturla, F., Molina-Sabio, M., and Rodríguez-Reinoso, F. (1991). Preparation of
Activated Carbon by Chemical Activation with ZnCl2. Carbon. 29: 999-1007.
Cech, T. V. (2005). Principles of Water Resources History, Development,
Management, and Policy. 2nd ed. Hoboken, NJ: John Wiley & Sons, Inc.
Chen, J. P., and Lin, M. (2000). Comprehensive Investigation of Important Factors
Governing Metal-Ion Adsorption by an H-Type Granular Activated Carbon. Sep.
Sci. Technol. 35: 2063-2081.
Chen, J. P., and Lin, M. (2001). Surface Charge and Metal Ion Adsorption on an HType Activated Carbon: Experimental Observation and Modeling Simulation by
the Surface Complex Formation Approach. Carbon. 39: 1491-1504.
Chen, J. P., and Wang, X. (2000). Removing Copper, Zinc, and Lead Ion by
Granular Activated Carbon in Pretreated Fixed-Bed Columns. Sep. Purif.
Technol. 19: 157-167.
Cini, R., Pantani, F., and Sorace, G. (1980). Physicochemical Aspects of the Use of
Activated Carbon in Drinking Water Treatment. In: Suffet, I. H., and McGuire,
M. J., ed. Activated Carbon Adsorption of Organics from the Aqueous Phase
Volume I. Ann Arbor, Michigan: Ann Arbor Science Publishers Inc. 425-446.
Corapcioglu, M. O., and Huang, C. P., (1987). The Adsorption of Heavy Metals onto
Hydrous Activated Carbon. Water Res. 21: 1031-1044.
Cotruvo, J. A., Hanson, H. F., and Thornton, T. P. (1983). Overview: Role of
Activated Carbon in EPS’s Regulatory Program. In: McGuire, M. J. and Suffet,
I. H., ed. Treatment of Water by Granular Activated Carbon. Washington, D.C.:
American Chemical Society. 1-6.
Dastgheib, S. A., and Rockstraw, D. A. (2001). Pecan Shell Activated Carbon:
Synthesis, Characterization, and Application for the Removal of Copper from
Aqueous Solution. Carbon. 39: 1849-1855.
Daud, W. M. A. W., Ali, W. S. W., and Sulaiman, M. Z. (2000). The Effects of
Carbonization Temperature on Pore Development in Palm-Shell-Based Activated
Carbon. Carbon. 38: 1925-1932.
159
Daud, W. M. A. W., and Ali, W. S. W. (2004). Comparison on Pore Development of
Activated Carbon Produced from Palm Shell and Coconut Shell. Biosource
Technology. 93: 63-69.
Daus, B., Wennrich, R., and Weiss, H. (2004). Sorption Materials for Arsenic
Removal from Water: A Comparative Study. Water Res. 38: 2948-2954.
Davranche, M., Lacour, S., Bordas, F., and Bollinger, J. C. (2003). An Easy
Determination of the Surface Chemical Properties of Simple and Natural Solids.
J. Chem. Edu. 80: 76-78.
DeMarco, M. J., SenGupta, A. K., and Greenleaf, J. E. (2003). Arsenic Removal
Using a Polymeric/Inorganic Hybrid Sorbent. Water Res. 37: 164-176.
Demirbas, E., Kobya, M., Senturk, E., and Ozkan, T. (2004). Adsorption Kinetics for
the Removal of Chromium(VI) from Aqueous Solutions on the Activated
Carbons Prepared from Agricultural Wastes. Water SA. 30: 533-539.
Do D. D. (1998). Adsorption analysis : equilibria and kinetics. London: Imperial
College Press.
Ellms, J. W. (1928). Water Purification. 2nd ed. New York, USA: McGraw-Hill
Book Company, Inc.
Farley, K. J., Dzombak, D. A., and Morel, F. M. M. (1985). A Surface Precipitation
Model for the Sorption of Cations on Metal Oxides. J. Colloid Interface Sci. 106:
226-242.
Faust, S. D., and Aly, O. M. (1987). Adsorption Processes for Water Treatment.
Stoneham, MA, USA: Butterworth Publishers.
Girgis, B. S., Yunis, S. S., and Soliman, A. M. (2002). Characteristics of Activated
Carbon from Peanut Hulls in Relation to Conditions of Preparation. Mater. Lett.
57: 164-172.
Gomez-Serrano, V., Macias-Garcia, A., Espinosa-Mansilla, A. and ValenzuelaCalahorro, C. (1998). Adsorption of Mercury, Cadmium and Lead from aqueous
solution on the Heat-Treated and Sulphurized Activated Carbon. Water Res. 32
(1): 1-4.
Guo, J., and Lua, A. C. (1999). Textural and Chemical Characterisations of
Activated Carbon Prepared from Oil-Palm Stone with H2SO4 and KOH
Impregnation. Micropor. Mesopor. Mater. 32: 111-117.
160
Guo, J., and Lua, A. C. (2000). Effect of Surface Chemistry on Gas-Phase
Adsorption by Activated Carbon Prepared from Oil-Palm Stone with PreImpregnation. Sep. Purif. Technol. 18: 47-55.
Guo, J., and Lua, A. C. (2002). Characterization of Adsorbent Prepared from OilPalm Shell by CO2 Activation for Removal of Gaseous Pollutants. Mater. Lett.
55: 334-339.
Guo, J., Xu, W. S., Chen, Y. L., and Lua, A. C. (2005). Adsorption of NH3 onto
Activated Carbon Prepared from Palm Shells Impregnated with H2SO4. J.
Colloid Interface Sci. 281: 285-290.
Hassler, J. W. (1974). Purification with Activated Carbon: Industrial Commercial
Environmental. New York, USA: Chemical Publishing Co., Inc.
Ho, Y. S., and Mckay, G. (2000). The kinetics of Sorption of Divalent Metal Ions
into Sphagnum Moss Peat. Water Res. 34: 735-742.
Hsisheng, T., and Sheng-Chi, W. (2000). Preparation of Porous Carbons from
Phenol-Formaldehyde Resins with Chemical and Physical Activation. Carbon.
38: 817-824.
Hu, Z., and Srinivasan, M. P. (2001). Mesoporous High-Surface-Area Activated
Carbon. Micropor. Mesopor. Mater. 43: 267-275.
Husain, Z., Zainac, Z., and Abdullah, Z. (2002). Briquetting of Palm Fibre and Shell
from the Processing of Palm Nuts to Palm Oil. Biomass Bioenergy. 22: 505-509.
Hussein, M. Z., Tarmizi, R. S. H.., Zainal, Z., Ibrahim, R., and Badri, M. (1996).
Preparation and Characterization of Active Carbons from Oil Palm Shell.
Carbon. 34: 1447-1449.
IUPAC Technical Report. (2002). “Heavy Metals”-A Meaningless Term?. Pure
Appl. Chem. 74 (5):793-807.
Jagtoyen, M., Thwaites, M., Stencel, J., McEnaney, B., and Derbyshire, F (1992).
Adsorbent Carbon Synthesis from Coals by Phosphoric Acid Activation. Carbon.
30: 1089-1096.
Jankowska, H., Swiatkowski, A., and Choma, J. (1991). Active Carbon. West
Sussex, England: Ellis Horwood Limited.
John, M. M., Marshall, W. E., and Toles, C. A. (1999). The Effect of Activation
Method on the Properties of Pecan Shell-Activated Carbons. J. Chem. Technol.
Biotechnol. 74: 1037-1044.
161
Johns, M. M., Marshall, W. E., and Toles, C. A. (1998). Agricultural by-Products as
Granular Activated Carbons for Adsorbing Dissolved Metals and Organics. J.
Chem. Technol. Biotechnol. 71: 131-141.
Kadirvelu, K., Kavipriya, M., Karthika, C., Vennilamani, N., and Pattabhi, S. (2004).
Mercury(II) Adsorption by Activated Carbon Made from Sago Waste. Carbon.
42: 745-752.
Krishnan, K. A., and Anirudhan, T. S. (2002). Uptake of Heavy Metals in Batch
Systems by Sulfurized Steam Activated Carbon Prepared from Sugarcane
Bagasse Pith. Ind. Eng. Chem. Res. 41: 5085-5093.
Krishnan, K. A., and Anirudhan, T. S. (2003). Removal of Cadmium(II) from
Aqueous Solutions by Steam Activated Sulphurized Carbon Prepared from
Sugar-cane Bagasse Pith: Kinetics and Equilibrium Studies. Water SA. 29 (2):
147-156.
Kunin, R. (1958). Ion Exchange Resins. New York, USA: John Wiley & Sons, Inc.
Lagergren, S. (1898). About the Theory of So-called Adsorption of Soluble
Substances. Kungliga Svenska Vetenskapsakademiens Handlinger. 24: 1-39.
Laine, J., and Yunes, S. (1992). Effect of the Preparation Method on the Pore Size
Distribution of Activated Carbon from Coconut Shell. Carbon. 30: 601-604.
Laine, J., Calafat, A., and Labady, M. (1989). Preparation and Characterization of
Activated Carbons from Coconut Shell Impregnated with Phosphoric Acid.
Carbon. 27: 191-195.
Lehr, J. H., Gass, T. E., Pettyjohn, W. A., and DeMarre, J. (1980). Domestic Water
Treatment. New York, USA: McGraw-Hill, Inc.
Lua, A. C., and Guo, J. (2000). Activated Carbon Prepared from Oil Palm Stone by
One-Step CO2 Activation for Gaseous Pollutant Removal. Carbon. 38: 10891097.
Macías-García, A., Valenzuela-Calahorro, C., Espinosa-Mansilla, A., BernalteGarcía, A., and Gómez-Serrano, V. (2004). Adsorption of Pb2+ in Aqueous
Solution by SO2-Treated Activated Carbon. Carbon. 42: 1755-1764.
Mahila, T. M. I., Abdulmuin, M. Z., Alamsyah, T. M. I., and Mukhlishien, D.
(2001). An Alternative Energy Source from Palm Wastes Industry for Malaysia
and Indonesia. Energy Convers. Manage. 42: 2109-2118.
162
Manning, B. A., Fendorf, S. E., and Goldberg, S. (1998). Surface Structures and
Stability of As(III) on Goethite: Spectroscopic Evidence for Inner-Sphere
Complexes. Environ Sci Technol. 32(16): 2383-2388.
Mattson, J. S., and Mark, H. B. (1971). Activated Carbon: Surface Chemistry and
Adsorption from Solution. New York: Marcel Dekker, Inc.
Mohan, D., and Chander, S. (2001). Single Component and Multi-Component Metal
Ions Adsorption by Activated Carbons. Colloid. Surf. A. 177: 183-196.
Mohan, D., and Singh, K. P. (2002). Single- and Multi-Component Adsorption of
Cadmium and Zinc Using Activated Carbon Derived from Bagasse-An
Agricultural Waste. Water Res. 36: 2304-2318.
Molina-Sabio, M., Rodríguez-Reinoso, F., Caturla, F., and Sellés, M. J. (1995).
Porosity in Granular Carbons Activated with Phosphoric Acid. Carbon. 33:
1105-1113.
Mortimer, R. G. (1993). Physical Chemistry. Bridge Parkway, California: The
Benjamin/Cummings Publishing Company.
Oh, G. H., and Park, C. R. (2002). Preparation and Characteristics of Rice-straw
Based Porous Carbons with Adsorption Capacity. Fuel. 81: 327-336.
Panday, K. K., Prasad, G., and Singh, V. N. (1985). Copper(II) Removal from
Aqueous Solutions by Fly Ash. Water Res. 19: 869-873.
Pearson, R. G. (1968). Hard and Soft Acids and Bases, HSAB, Part I. J. Chem. Edu.
45: 581-587.
Peng, X., Luan, Z., Di, Z., Zhang, Z., and Zhu, C. (2005). Carbon Nanotubes-Iron
Oxides Magnetic Composites as Adsorbent for Removal of Pb(II) and Cu(II)
from Water. Carbon. 43: 855-894.
Peräniemi, S., and Ahlgrén, M. (1995). Optimized Arsenic, Selenium and Mercury
Determinations in Aqueous Solutions by Energy Dispersive X-ray Fluorescence
After Preconcentration onto Zirconium-Loaded Activated Charcoal. Anal. Chim.
Acta. 302: 89-95.
Pierce, M. L., and Moore, C. B. (1982). Adsorption of Arsenite and Arsenate on
Amorphous Iron Hydroxide. Water Res. 16(7): 1247-1253.
Puri, B. R. (1980). Discussion IV Surface Chemistry and Physical Properties of
Carbon Affecting Adsorption. In: Suffet, I. H., and McGuire, M. J., ed. Activated
Carbon Adsorption of Organics from the Aqueous Phase Volume I. Ann Arbor,
Michigan: Ann Arbor Science Publishers Inc. 483-492.
163
Reed, B. E. (2002). Removal of Heavy Metals by Activated Carbon. In: SenGupta,
A. K., ed. Environmental Separation of Heavy Metals: Engineering Processes.
New York: Lewis Publishers. 205-264.
Reed, B. E., and Matsumoto, M. R. (1991). Modeling Surface Acidity of Two
Powdered Activated Carbons: Comparison of Diprotic and Monoprotic Surface
Representations. Carbon. 29: 1191-1201.
Reed, B. E., Vaughan, R., and Jiang, L. (2000). As(III), As(V), Hg, and Pb Removal
by Fe-oxide Impregnated Activated Carbon. J. Environ. Eng. 126: 869-873.
Rivera-Utrilla, J., and Sánchez-Polo, M. (2003). Adsorption of Cr(III) on Ozonised
Activated Carbon: Importance of Cл-Cation Interactions. Water Res. 37: 33353340.
Rodríguez-Reinoso, F, Molina-Sabio, M., and González, M. T. (1995). The Use of
Steam and CO2 as Activating Agents in the Preparation of Activated Carbons.
Carbon. 33: 15-23.
Seco, A., Gabaldón, C., Marzal, P., and Aucejo, A. (1999). Effect of pH, Cation
Concentration and Sorbent Concentration on Cadmium and Copper Removal by
a Granular Activated Carbon. J. Chem. Technol. Biotechnol. 74: 911-918.
Selomulya, C., Meeyoo, V., and Amal, R. (1999). Mechanisms of Cr(VI) Removal
from Water by Various Types of Activated Carbons. J. Chem. Technol.
Biotechnol. 74: 111-122.
SenGupta, A. K. (2002). Principles of Heavy Metals Separation: An Introduction. In:
SenGupta, A. K., ed. Environmental Separation of Heavy Metals: Engineering
Processes. New York: Lewis Publishers. 1-14.
Sengupta, S., and SenGupta, A. K. (2002). Trace Heavy Metal Separation by
Chelating Ion Exchangers. In: SenGupta, A. K., ed. Environmental Separation of
Heavy Metals: Engineering Processes. New York: Lewis Publishers. 45-96.
Shriver, D. F., Atkins, P. W., and Langford, C. H. (1991). Inorganic Chemistry.
Oxford: Oxford University Press.
Smisek, M., and Cerny, S. (1970). Active Carbon: Manufacture, Properties and
Applications. Amsterdam: Elsevier.
Stevens, L. A. (1974). Clean Water-Nature’s Way to Stop Pollution. New York:
Dutton & Co.
Strelko, V. Jr., and Malik, D. J. (2002). Characterization and Metal Sorptive
Properties of Oxidized Active Carbon. J. Colloid Interface Sci. 250: 213-220.
164
Tan, T. C., and Teo, W. K. (1987). Combined Effect of Carbon Dosage and Initial
Adsorbate Concentration on the Adsorption Isotherm of Heavy Metals on
Activated Carbon. Water Res. 21: 1183-1188.
Toles, C. A., and Marshall, W. E. (2002). Copper Ion Removal by Almond Shell
Carbons and Commercial Carbons: Batch and Column Studies. Sep. Sci.
Technol. 37: 2369-2383.
Toles, C. A., Marshall, W. E., and Johns, M. M. (1997). Granular Activated Carbons
from Nutshells for the Uptake of Metals and Organic Compounds. Carbon. 35:
1407-1414.
Toles, C. A., Marshall, W. E., and Johns, M. M. (1998). Phosphoric Acid Activation
of Nutshells for Metals and Organic Remediation: Process optimization. J.
Chem. Technol. Biotechnol. 72: 255-263.
Toles, C. A., Marshall, W. E., and Johns, M. M. (1999). Surface Functional Groups
on Acid Activated Nutshell Carbons. Carbon. 37: 1207-1214.
Toles, C. A., Marshall, W. E., and Johns, M. M. (2000). Activated Nutshell Carbons
from Agricultural Waste. (U.S. Patent 6,033,573).
Toles, C. A., Marshall, W. E., Wartelle, L. H., and McAloon, A. (2000). Steam- or
Carbon dioxide-Activated Carbons from Almond Shells: Physical, Chemical and
Adsorptive Properties and Estimated Cost of Production. Bioresource
Technology. 75: 197-203.
Tsai, W. T., Chang, C. Y., Wang, S. Y., Chang, C. F., Chien, S. F., and Sun, H. F.
(2001). Preparation of Activated Carbons from Corn Cob Catalyzed by
Potassium Salts and Subsequent Gasification with CO2. Bioresource Technology.
78: 203-208.
Turmuzi, M., Daud, W. R. W., Tasirin, S. M., Takriff, M. S., and Iyuke, S. E.
(2004). Production of Activated Carbon from Cadlenut shell by CO2 Activation.
Carbon. 42: 453-455.
Uzun, Í., and Güzel, F. (2000). Adsorption of Some Heavy Metal Ions from Aqueous
Solution by Activated Carbon and Comparison of Percent Adsorption Results of
Activated Carbon with Those of Some Other Adsorbents. Turk. J. Chem. 24:
291-297.
Vaishya, R. C., and Gupta, S. K. (2003). Arsenic Removal from Ground Water by
Iron Impregnated Sand. J. Environ. Eng. 129: 89-92.
165
Weber, Jr. W. J., McGinley, P. M., and Katz, L. E. (1991). Sorption Phenomena in
Subsurface Systems: Concepts, Models and Effects on Contaminant Fate and
Transport. Water Res. 25(5): 499-528.
Wigmans, T. (1989). Industrial Aspects of Production and Use of Activated Carbons.
Carbon. 27: 13-22.
Yang, J., Shen, Z., and Hao, Z. (2004). Preparation of Highly Microporous and
Mesoporous Carbon from the Mesophase Pitch and Its Carbon Foams with KOH.
Carbon. 42: 1872-1875.
Yantasee, W., Lin, Y., Fryxell, G. E., Alford, K. L., Busche, B. J., and Johnson, C.
D. (2004). Selective Removal of Copper(II) from Aqueous Solutions Using FineGrained Activated Carbon Functionalized with Amine. Ind. Eng. Chem. Res. 43:
2759-2764.
Youssef, A. M., Radwan, N. R. E., Abdel-Gawad, I., and Singer, G. A. A. (2005).
Textural Properties of Activated Carbons from Apricot Stones. Colloid Surf. A.
252: 143-151.
APPENDIX A
National Drinking Water Quality Standards, 2000 of Malaysia for
some species of metals and metalloids
Parameters
Aluminium
Arsenic
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Mercury
Nickel
Selenium
Zinc
Max. = maximum
Max. Acceptable Value, (mg.L-1)
0.2
0.01
0.003
0.05
1.0
0.3
0.01
0.1
0.001
0.02
0.01
3
167
APPENDIX B
EDX Spectrums of Some Selected Activated and Modified Carbons
PSW-P-500
C-HFO-1
C-HFOCa-1
Figure B-1 EDX spectrums of base activated carbon PSW-P-500 and its modified
C-HFO composites, namely C-HFO-1 and C-HFOCa-1.
168
CPW-P-500
C-HFO-2
Figure B-2 EDX spectrums of base activated carbon CPW-P-500 and its
modified composite, namely C-HFO-2.
169
PSW-P-ad-500
C-HFO-3
Figure B-3 EDX spectrums of base activated carbon PSW-P-ad-500 and its
modified composite, namely C-HFO-3.
170
APPENDIX C
Summary Report of N2 Adsorption-Desorption Data and BJH Pore size distribution Graph
Table C-1 : Summary Report of N2 Adsorption-Desorption Data (Analysis Bath: 77 K).
Surface areas (m2.g-1)
Sample
Pore volume (cm3.g-1)
Average pore width (Å)
Single point
BET
Langmuir
Total
by BET
CPW-P-500
1463.17
1490.93
2039.38
0.937639
25.1558
CSW-P-500
1029.36
1053.24
1447.54
0.610459
23.1842
CR-P-500
1044.42
1048.78
1424.73
0.626008
23.8758
PSW-P-500
1034.37
1048.68
1431.74
0.608583
23.2133
PSW-P-ad-500
1429.16
1475.55
2031.06
0.799982
21.6864
PSWfg-P-450
1508.28
1471.73
1965.91
0.720479
19.5817
PSW-P-ed-450
1337.66
1366.36
1873.81
0.749715
21.9478
PSW-P-400
598.15
593.00
798.07
0.306482
20.6733
PSW-P-700
739.92
734.03
991.86
0.358366
19.5286
CAC
1314.71
1320.09
1792.24
0.673259
20.4004
171
Figure C-1 Pore size distribution in PSW-P-500 from BJH desorption dV/dlog(D)
pore volume.
172
Figure C-2 Pore size distribution in CPW-P-500 from BJH desorption dV/dlog(D)
pore volume.
173
Figure C-3 Pore size distribution in PSW-P-ad-500 from BJH desorption
dV/dlog(D) pore volume.
174
APPENDIX D
FTIR Spectrums of Some of the Selected Activated Carbons
Figure D
FTIR spectrums of activated carbons, PSW-P-500, CPW-P-500 and
PSW-P-ad-500.
175
APPENDIX E
Quantification of Acidic Groups on the Surface of Some Selected
Activated Carbons as meq Base per Gram carbon
Table E : Boehm’s titration data and base uptake capacity of some of the prepared
activated carbons in meq.g-1.
NaOH
NaOEt
1.5
3.0
3.5
PSW-P-500
0.4
1.2
1.8
1.8
1.0
3.0
4.5
4.5
PSW-P-ad-500
0.4
1.2
1.8
1.9
1.0
3.0
4.5
4.75
10 mL
0.5
NaOEt
1.4
10mL
1.2
NaOH
0.6
10 mL
0.2
Na2CO3
CPW-P-500
10 mL
Na2CO3
Base Uptake meq.g-1
(Vol. of Titer x 5x 0.1)/0.2
NaHCO3
Volume of Titer (0.1 N HCl) mL
NaHCO3
Sample
N.B. 0.2 g of carbon was shaken with 50 mL of various bases of strength 0.1 N while
10 mL of aliquot samples of various bases were titrated with 0.1N of HCl.
176
APPENDIX F
TGA Thermograms of Some of the Selected Activated Carbons
Figure F-1
TGA thermogram of an activated carbon, PSW-P-400 (in N2).
177
Figure F-2
TGA thermogram of an activated carbon, PSW-P-400 (in air).
178
Figure F-3
TGA thermogram of an activated carbon, PSW-P-500 (in N2).
179
Figure F-4
TGA thermogram of an activated carbon, PSW-P-500 (in air).
180
Figure F-5
TGA thermogram of an activated carbon, PSW-P-ad-500 (in N2).
181
Figure F-6
TGA thermogram of an activated carbon, PSW-P-ad-500 (in air).
182
Figure F-7
TGA thermogram of an activated carbon, CSW-P-500 (in N2).
183
Figure F-8
TGA thermogram of an activated carbon, CSW-P-500 (in air).
184
Figure F-9
TGA thermogram of an activated carbon, CPW-P-500 (N2).
185
Figure F-10 TGA thermogram of an activated carbon, CPW-P-500 (air).
186
APPENDIX G
FAAS Determined Fe Content in Samples of Leached Water
Contacted with C-HFO-1 Composite Adsorbent at pH 4.5
0.25
r 2 = 0.997
Absorbance
0.2
0.15
0.1
0.05
0
0
0.5
1
1.5
2
2.5
3
3.5
4
Concentration of Fe, mg/L
Figure G
Calibration curve of Fe.
Table G : Fe content in leached samples of water after contacting with a typical CHFO composite, namely C-HFO-1.
Samples
Leached water
Fe Concentration (mg.L-1)
Fail
0.026
Fail
Fail
Mean Absorbance
-0.001
0.002
-0.000
-0.000
187
APPENDIX H
FAAS Determined Adsorption Equilibrium Data of Various Metal
Cations in Single, Binary and Ternary Solutes Solutions by Various
Activated Carbons at Varying Initial Concentration and pH
H-1 : Adsorption Equilibrium Data of Cu(II) for the Adsorption in Single
Solute Solution at Varying Initial Concentrations (< 5 - < 120 mg.L-1) and at
Different pH (e.g. 3 and 5). (AAS- GBC, Model- Avanta).
Table H-1.1 : Adsorption equilibrium data of Cu(II) in single solute solution by
various activated carbons at pH 3 and adsorption capacities.
Solution Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ce
SD
4.5
0.1
2.3
0.3
20.5
0.9
15.2
0.5
45.1
0.3
41.1
0.9
67
1
62
1
95.5
0.7
90.5
0.2
116
3
110
2
Adsorption capacity
mg.g-1
qe
SD
RSD %
1.1
0.09
8
2.6
0.4
15
2
0.6
30
2.7
0.7
26
2.5
0.4
16
3
0.6
20
PSW-P-500
4.5
20.5
45.1
67
95.5
116
0.1
0.9
0.3
1
0.7
3
ND
3.6
19
35
62
80.8
0.4
2
3
5
0.8
8.4
13
16
16.6
17.7
0.6
1
0.9
2.5
0.9
7
8
6
15
5
CPW-P-500
4.5
20.5
45.1
67
95.5
116
0.1
0.9
0.3
1
0.7
3
ND
2.3
19
34
61
79
0.3
2
4
4
4
9.1
12.9
16
18
19
0.4
0.9
3
2
2
4
7
19
11
11
PSW-P-ad-500
4.5
20.5
45.1
67
95.5
116
0.1
0.9
0.3
1
0.7
3
ND
0.53
10.8
29.5
53
75
0.22
0.5
0.5
2
3
10
17.2
18.7
21.4
21
0.4
0.4
0.4
0.7
2
4
2
2
3
10
Sample
CAC
188
Table H-1.2 : QC checking while determining Cu concentration for the adsorption
in single solute solution at initial pH 3.
Concentrations, mg.L-1
Spiked
Observed
1.0
0.85
3.5
3.13
3.5
3.25
3.5
3.46
3.5
3.40
3.5
3.44
1.0
0.99
Sample
QC
Recovery %
Average
Recovery %
84.5
89.5
93.0
98.7
97.0
98.3
99.3
94 ± 6
0.8
Absorbance
r 2 = 1.000
0.6
0.4
0.2
0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
Concentration of Cu, mg/L
Figure H-1.1 Calibration curve of Cu for the adsorption equilibrium data in single
solute solution by various activated carbons at an initial pH 3.
0.60
r 2 = 0.999
Absorbance
0.50
0.40
0.30
0.20
0.10
0.00
0.0
1.0
2.0
3.0
4.0
5.0
6.0
Concentration of Cu, mg/L
Figure H-1.2 Calibration curve of Cu for the adsorption equilibrium data in single
solute solution by various activated carbons at an initial pH 5.
189
Table H-1.3 : Adsorption equilibrium data of Cu(II) in single solute solution by
various activated carbons at an initial pH 5 and adsorption capacities.
Solution Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ce
SD
4.9
0.1
1.5
0.1
23.5
0.1
18
0.8
48.2
0.3
39
0.8
71.7
0.4
63
1
96
0.3
86
2
118
1
107
0.9
Adsorption capacity
mg.g-1
qe
SD
RSD %
1.75
0.03
1.71
2.7
0.5
18.52
4.6
0.6
13.04
4.5
0.7
15.56
5
0.8
16.00
5
1
20.00
PSW-P-500
4.9
23.5
48.2
71.7
96
118
0.1
0.1
0.3
0.4
0.3
1
ND
2.1
17
39
60
78
0.6
2
2
1
3
10.7
15.5
16.3
18
20
0.2
0.9
0.9
0.6
2
1.89
5.81
5.52
3.33
10.00
CPW-P-500
4.9
23.5
48.2
71.7
96
118
0.1
0.1
0.3
0.4
0.3
1
ND
2.3
18
39
59
79
0.26
1
1
2
4
10.6
15.1
16.6
18.3
19.3
0.16
0.8
0.6
0.9
1.2
1.51
5.30
3.61
4.92
6.22
PSW-P-ad-500
4.9
23.5
48.2
71.7
96
118
0.1
0.1
0.3
0.4
0.3
1
ND
0.4
10
30
51
72
0.1
1
2
1
1
11.5
18.9
21
22.6
22.7
0.04
0.4
1
0.6
0.4
0.35
2.12
4.76
2.65
1.76
Sample
CAC
Table H-1.4 : QC analysis while determining Cu concentration for the adsorption in
single solute solution at an initial pH 5.
Sample
QC
Concentrations, mg.L-1
Spiked
Observed
3.5
3.50
1.0
1.05
3.5
3.48
3.5
3.46
3.5
3.46
3.5
3.45
3.5
3.45
3.5
3.50
Recovery %
100
105
100
99
99
99
99
100
Average
Recovery %
100 ± 2
190
H-2 : Adsorption Equilibrium Data of Ni(II) for the Adsorption in Single
Solute Solution at Varying Initial Concentrations (6 - 227 mg.L-1) and at a
Constant Initial pH 5. (AAS- GBC, Avanta).
Table H-2.1 : Adsorption equilibrium data of Ni(II) in single solute solution by
various activated carbon at pH 5 and adsorption capacities.
Sample
CAC
Solution Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ce
SD
6.1
0.1
4.1
0.0
30.6
0.3
27.2
0.2
61.3
0.1
55.9
0.5
91.5
0.5
84
1
122.0
1.0
116
1
150
2
143
1
227
3
222
3
Adsorption capacity
mg.g-1
qe
SD
RSD %
1.0
0.0
2.5
1.7
0.2
11.8
2.7
0.2
8.9
3.7
0.6
16.0
3
1
34.2
3
1
17.4
4
0
0.8
PSW-P-500
6.1
30.6
61.3
91.5
122
150
227
0.1
0.3
0.1
0.5
1
2
3
ND
15.1
42.8
68.4
101
129
206
0.1
0.3
0.3
0
3
2
7.8
9.3
11.5
10.3
10.7
11
0.1
0.2
0.4
0.5
0.6
2
1.5
1.8
3.7
4.8
5.9
18.1
CPW-P-500
6.1
30.6
61.3
91.5
122
150
227
0.1
0.3
0.1
0.5
1
2
3
ND
9.8
38
68.3
99
117
203
0.0
1
0.2
3
2
2
10.4
11.8
11.6
11.7
16.4
12.1
0.1
0.7
0.3
0.9
2
2
1.3
5.6
2.4
7.8
10.0
17.7
PSW-P-ad-500
6.1
30.6
61.3
91.5
122
150
227
0.1
0.3
0.1
0.5
1
2
3
ND
6.7
30.4
57
87.8
111
190
0.1
0.2
1
0.6
1
3
12.0
15.5
17.4
17.1
19
19
0.2
0.1
0.3
0.4
1
1
1.6
0.8
1.5
2.3
6.9
6.9
191
Table H-2.2 : QC analysis while determining Ni concentration for the adsorption in
single solute solution at an initial pH 5.
Concentrations, mg.L-1
Spiked
Observed
1.0
0.98
1.0
0.97
1.0
0.99
1.0
0.97
1.0
0.98
1.0
0.98
1.0
0.96
1.0
0.91
4.0
3.85
Sample
QC
Recovery %
Average
Recovery %
98.4
97.3
99.2
97.1
98.2
98
95.7
90.5
96.35
97 ± 3
0.35
Absorbance
0.30
r 2 = 0.9998
0.25
0.20
0.15
0.10
0.05
0.00
0.0
1.0
2.0
3.0
4.0
5.0
6.0
Concentration of Ni, mg/L
Figure H-2.1 Calibration curve of Ni for the adsorption equilibrium data in single
solute solution by various activated carbons at an initial pH 5.
192
H-3 : Adsorption Equilibrium Data of Zn(II) for the Adsorption in Single
Solute Solution at Varying Initial Concentrations (19 - 310 mg.L-1).
Table H-3.1 : Adsorption equilibrium data of Zn(II) in single solute solution by
various activated carbon at pH 5 and adsorption capacities while analyte was
analysed for the initial concentration range, 19 – 310 mg.L-1. (AAS-GBC, Avanta).
Solution Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ce
SD
19.0
0.2
14.9
0.3
78.5
0.3
74.0
0.2
97.6
0.2
92.2
2.1
127.0
0.4
121.4
0.6
250.5
0.6
245.2
0.8
Adsorption capacity
mg.g-1
qe
SD
RSD %
2.0
0.2
11.1
2.3
0.2
9.5
2.7
1.0
36.6
2.8
0.1
3.9
2.6
0.1
5.1
PSW-P-500
19.0
36.8
62.4
78.5
97.6
127.0
166.3
186.7
250.5
310.2
0.2
1.6
0.4
0.3
0.2
0.4
1.0
1.4
0.6
1.3
7.4
27.0
47.1
64.8
82.9
109.1
153.3
173.2
234.8
293.5
0.0
0.1
1.0
0.1
0.1
0.5
0.2
2.1
0.4
0.9
5.8
4.9
7.7
6.9
7.4
9.0
6.5
6.7
7.8
8.4
0.1
0.7
0.4
0.1
0.1
0.4
0.5
0.8
0.5
0.8
2.0
14.8
4.7
1.2
1.2
4.7
7.8
11.6
6.7
9.3
CPW-P-500
19.0
36.8
62.4
78.5
97.6
127.0
166.3
186.7
200.4
250.5
310.2
0.2
1.6
0.4
0.3
0.2
0.4
1.0
1.4
0.6
0.6
1.3
5.4
25.8
47.4
59.7
83.0
107.5
149.8
171.8
181.8
232.7
289.5
0.1
0.0
0.4
0.1
0.6
2.1
0.7
0.2
0.9
0.2
1.0
6.8
5.5
7.5
9.4
7.3
9.7
8.3
7.4
9.3
8.9
10.3
0.1
0.8
0.4
0.2
0.3
1.0
0.2
0.7
0.6
0.2
1.0
1.7
14.2
4.9
1.8
4.3
10.0
2.1
9.7
6.6
2.2
10.0
PSW-P-ad-500
19.0
36.8
62.4
78.5
97.6
127.0
166.3
186.7
200.4
250.5
310.2
0.2
1.6
0.4
0.3
0.2
0.4
1.0
1.4
0.6
0.6
1.3
4.2
20.5
41.4
59.4
77.5
106.1
150.5
171.0
184.6
230.9
289.4
0.1
0.1
0.1
1.5
0.3
0.3
0.5
0.2
0.8
1.4
0.3
7.4
8.2
10.5
9.6
10.0
10.4
7.9
7.9
7.9
9.8
10.4
0.1
0.8
0.2
0.6
0.1
0.3
0.6
0.7
0.6
0.7
0.5
1.2
10.2
2.2
6.6
1.2
2.7
7.9
9.4
7.6
7.1
4.6
Sample
CAC
193
Table H-3.2 : QC analysis while determining Zn concentration for the adsorption in
single solute solution at an initial pH 5.
Concentrations, mg.L-1
Spiked
Observed
0.5
0.597
1.5
1.540
2.0
1.937
1.5
1.546
2.0
1.938
1.5
1.530
2.0
1.970
2.0
1.970
2.0
1.988
2.0
1.966
2.0
1.878
1.5
1.533
Sample
QC
Recovery %
119
103
97
103
97
102
98
99
99
98
94
102
Average
Recovery %
101 ± 6
0.50
0.45
r 2 = 0.996
Absorbance
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
0.0
0.5
1.0
1.5
2.0
2.5
Concentration of Zn, mg/L
Figure H-3.1 Calibration curve of Zn for the adsorption equilibrium data in single
solute solution by various activated carbons at an initial pH 5.
194
H-4 : Adsorption Equilibrium Data of Pb(II) for the Adsorption in Single
Solute Solution at Varying Initial Concentrations (< 50 - < 600 mg.L-1) and at
Different pH (e.g. 3 and 5). (AAS- PerkinElmer, AAnalyst 400).
Table H-4.1 : Adsorption equilibrium data of Pb(II) in single solute solution by
various activated carbon at pH 5 and adsorption capacities.
Solution Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ce
SD
48.7
0.2
22.4
0.9
71.6
0.6
46.2
1.1
94.1
0.9
62.5
0.5
118.1
1.1
89.6
3.4
190.4
0.7
163.6
3.9
285.4
0.0
252.0
3.5
387.0
5.6
344.8
1.7
569.6
0.3
525.9
0.9
Adsorption capacity
mg.g-1
qe
SD
RSD %
12.9
0.5
4.2
12.3
0.8
6.9
16.0
0.2
1.4
13.1
1.2
9.1
14.7
2.3
15.6
17.9
1.8
9.8
20.5
3.7
17.9
22.2
0.6
2.8
PSW-P-500
48.7
71.6
94.1
118.1
190.4
285.4
387.0
569.6
0.2
0.6
0.9
1.1
0.7
0.0
5.6
0.3
3.8
5.7
15.4
25.6
75.6
139.5
240.4
425.5
0.9
1.0
1.6
4.1
3.6
8.0
2.1
10.5
22.5
32.9
39.4
46.3
57.4
72.9
73.3
72.0
0.3
0.8
1.2
1.5
2.2
4.0
1.7
5.4
1.4
2.4
3.2
3.3
3.8
5.5
2.4
7.6
CPW-P-500
48.7
71.6
94.1
118.1
190.4
285.4
387.0
569.6
0.2
0.6
0.9
1.1
0.7
0.0
5.6
0.3
4.7
6.5
21.1
29.5
81.5
161.8
250.8
427.0
1.2
1.6
1.4
5.6
6.3
14.3
1.6
7.7
22.0
32.6
36.5
44.3
54.4
61.8
68.1
71.3
0.7
1.1
1.2
2.3
2.8
7.1
3.6
3.7
3.1
3.3
3.2
5.1
5.1
11.5
5.3
5.2
PSW-Pad-500
48.7
71.6
94.1
118.1
190.4
285.4
387.0
569.6
0.2
0.6
0.9
1.1
0.7
0.0
5.6
0.3
2.2
6.5
17.6
27.0
83.8
185.7
267.1
441.2
0.7
0.2
1.0
4.9
2.9
11.4
8.2
8.0
23.3
32.5
38.3
45.6
53.3
49.8
60.0
64.2
0.3
0.2
1.0
3.0
1.8
5.7
6.9
4.2
1.1
0.6
2.6
6.6
3.4
11.4
11.5
6.5
Sample
CAC
195
0.25
r 2 = 0.999
Absorbance
0.2
0.15
0.1
0.05
0
0.0
2.0
4.0
6.0
8.0
10.0
12.0
Concentration of Pb, mg/L
Figure H-4.1 Calibration curve of Pb for the adsorption equilibrium data in single
solute solution by various activated carbons at an initial pH 5.
Table H-4.2 : QC analysis while determining Pb concentration for the adsorption in
single solute solution at an initial pH 5.
Sample
QC
Concentrations, mg.L-1
Spiked
Observed
4.0
4.03
4.0
4.09
4.0
3.89
4.0
3.99
4.0
3.97
4.0
3.98
4.0
3.97
4.0
3.89
4.0
3.88
4.0
3.90
4.0
3.89
4.0
3.90
Recovery %
101
102
97
100
99
100
99
97
97
97
97
98
Average
Recovery %
99 ± 2
196
Table H-4.3 : Adsorption equilibrium data of Pb(II) in single solute solution by the
best-selected carbon, PSW-P-ad-500 at pH 3 and adsorption capacity.
Sample
PSW-P-ad-500
Solution Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ce
SD
50.6
0.6
9
1
72.6
0.2
23
2
96.4
0.3
40
3
121.0
0.1
58
1
147.8
1.5
80
4
194.7
1.5
126
4
244.5
1.7
171
4
288.9
2.8
203
7
388.8
3.6
302
2
582.9
1.5
487
6
Adsorption capacity
mg.g-1
qe
SD
RSD %
21
1
4.7
25
1
5.0
28
2
6.2
31
1
1.9
34
2
6.3
36
2
4.7
37
2
5.2
43
2
5.3
44
1
3.0
48
3
7.1
Table H-4.4 : QC analysis while determining Pb concentration for the adsorption in
single solute solution at an initial pH 3.
Concentrations, mg.L-1
Spiked
Observed
8.0
7.96
8.0
8.09
8.0
8.09
8.0
8.06
1.0
1.07
8.0
7.98
1.0
1.09
Sample
QC
Recovery %
Average
Recovery %
99
101
101
101
107
100
109
103 ± 4
0.25
r 2 = 0.999
Absorbance
0.2
0.15
0.1
0.05
0
0
2
4
6
8
10
12
Concentration of Pb, mg/L
Figure H-4.2 Calibration curve of Pb for the adsorption equilibrium data in single
solute solution by an activated carbon, PSW-P-ad-500, at an initial pH 3.
197
H-5 : Adsorption Equilibrium Data of Cu(II) and Ca(II) for the Adsorption in
Binary Solute Solution at Varying Initial Concentrations (< 25 - < 150 mg.L-1)
and at Different pH (e.g. 3 and 5).
Table H-5.1 : Adsorption equilibrium data of Cu(II) in binary solute solution of
Cu(II) and Ca(II) (Cu-Ca), at pH 5, by the best-selected carbon, PSW-P-ad-500 and
adsorption capacity, while initial concentrations were varied, < 25 - < 120 mg.L-1.
Solution Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ce
SD
23.6
0.1
1.7
0.3
48.2
0.4
14.8
0.3
PSW-P-ad-500
72.3
0.3
35.0
1.6
97.0
0.1
55.5
1.3
117.5
0.7
72.3
0.8
N.B. AAS- GBC, Avanta.
Sample
Adsorption capacity
mg.g-1
qe
SD
RSD %
10.9
0.1
1.0
16.7
0.1
0.4
18.7
1.0
5.1
20.7
0.6
2.8
22.6
0.8
3.3
Table H-5.2 : Adsorption equilibrium data of Cu(II) in binary solute solution of CuCa, at pH 3, by the best-selected carbon, PSW-P-ad-500 and adsorption capacity,
while initial concentrations were varied, 26 - 151 mg.L-1.
Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ce
SD
26.1
1.1
5.8
2.4
58.9
0.9
24.1
4.0
PSW-Pad-500
90.9
1.4
55.5
2.9
122.8
1.5
80.7
4.7
150.9
0.0
108.3
4.3
N.B. AAS- PerkinElmer, AAnalyst 400.
Sample
Adsorption capacity
mg.g-1
qe
SD
RSD %
10.1
0.7
6.4
17.4
1.5
8.8
17.7
1.2
6.8
21.0
1.9
9.2
21.3
2.2
10.2
Table H-5.3 : QC analysis while determining adsorption equilibrium data of Cu(II)
in binary solute solution at an initial pH 5.
Sample
QC
Concentrations, mg.L-1
Spiked
Observed
3.5
3.5044
3.5
3.4868
3.5
3.4598
3.5
3.4522
3.5
3.5021
3.5
3.4624
Recovery %
100.126
99.6229
98.8514
98.6343
100.06
98.9257
Average
Recovery %
99.4 ± 0.7
198
Table H-5.4 : QC analysis while determining adsorption equilibrium data of Cu(II)
in binary solute solution at an initial pH 3.
Sample
QC
Concentrations, mg.L-1
Spiked
Observed
1.3
1.36
1.3
1.28
1.3
1.27
1.3
1.27
1.3
1.35
Recovery %
104
98
97
98
104
Average
Recovery %
100 ± 3
Table H-5.5 : Adsorption equilibrium data of Ca(II) in binary solute solution of CuCa, at pH 5, by the best-selected carbon, PSW-P-ad-500 and adsorption capacity,
while initial concentrations were varied < 25 - < 120 mg.L-1.
Solution Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ce
SD
23.8
0.2
12.6
0.5
47.6
1.5
39.9
0.4
PSW-Pad-500
70.4
0.5
66.8
0.6
96.6
1.1
94.5
1.2
119.6
1.3
116.4
3.7
N.B. AAS- GBC, Avanta.
Sample
Adsorption capacity
mg.g-1
qe
SD
RSD %
5.6
0.3
5.8
3.9
0.6
16.6
1.8
0.1
7.9
1.1
0.1
8.8
1.6
1.2
8.0
Table H-5.6 : Adsorption equilibrium data of Ca(II) in binary solute solution of CuCa, at pH 3, by the best-selected carbon, PSW-P-ad-500 and adsorption capacity,
while initial concentrations were varied < 25 - < 120 mg.L-1.
Solution Concentrations, mg.L-1
Initial
Equilibrium
SD
Ce
SD
C0
20.4
0.7
18.1
1.1
44.5
0.9
42.0
1.7
PSW-Pad-500
70.1
3.9
67.4
1.7
90.4
1.8
87.2
1.9
109.4
2.6
104.8
2.1
N.B. AAS- PerkinElmer, AAnalyst 400.
Sample
Adsorption capacity
mg.g-1
qe
SD
RSD %
1.2
0.8
72
1.3
1.1
86
1.3
2.6
191
1.6
1.8
117
2.3
2.3
102
Table H-5.7 : QC analysis during determination of Ca(II) concentration in binary
solute solution, Cu-Ca, at pH 5.
Sample
QC
Concentrations, mg.L-1
Spiked
Observed
4.0
3.82
4.0
3.78
Recovery %
95.5
94.4
Average
Recovery %
94.9 ± 0.8
199
Table H-5.8 : QC analysis during determination of Ca(II) concentration in binary
solute solution, Cu-Ca, at pH 3.
Concentrations, mg.L-1
Spiked
Observed
3.79
3.73
3.71
4.0
3.67
3.70
3.69
Sample
QC
0.4
Recovery %
Average
Recovery %
95
93
93
92
92
92
93 ± 1
r 2 = 0.9995
Absorbance
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
0
1
2
3
4
5
Concentration of Ca, mg/L
Figure H-5.1 A calibration curve for the determination of Ca in binary solute
solution of Cu(II) and Ca(II).
6
200
H-6 : Adsorption Equilibrium Data of Cu(II) by PSW-P-ad-500 for the
Adsorption in Single Solute Solution at Varying Initial Concentrations (< 25 - <
700 mg.L-1) and at a Constant Initial pH 5.
Table H-6.1 : Adsorption equilibrium data of Cu(II) by PSW-P-ad-500 and
adsorption capacity for Cu(II) for the adsorption in single solute solution at a
constant initial pH 4.5, while initial concentrations were varied widely, from < 25 to
< 700 mg.L-1.
Solution Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ce
SD
24.9
0.05
1.9
0.01
49.9
0.6
19.9
0.06
96
0.4
60
0.4
PSW-Pad-500
191
1
155.9
0.2
366
3
320.7
0.4
455
0.3
409
2
693
4
644
8
N.B. AAS- GBC, Avanta.
Sample
Adsorption capacity
mg.g-1
qe
SD
RSD %
11.5
0.02
0.17
15
0.2
1.33
18.2
0.3
1.65
18
0.7
3.89
23
1
4.35
23
1
4.35
25
6
24.00
H-7 : Adsorption Equilibrium Data of Cu(II), Ni(II) and Pb(II) (Cu-Ni-Pb) and
Adsorption Capacity for the Adsorption in Ternary Solute Solution at Varying
Initial Concentration (< 25 - < 600 mg.L-1) and at Different pH (e.g. 3 and 5).
Table H-7.1 : Adsorption equilibrium data of Cu(II) in ternary solute solution, CuNi-Pb, at pH 5, by the best-selected carbon, PSW-P-ad-500 and its adsorption
capacity, while initial concentrations were varied < 25 - < 600 mg.L-1.
Solution Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ce
SD
23.0
0.2
6.5
0.5
47.9
0.3
24.8
1.1
74.0
0.2
48.0
0.7
97.8
0.6
71.3
1.2
118.6
0.8
90.3
3.2
PSW-P-ad-500
190.4
0.9
161.0
1.7
239.3
1.7
205.4
2.3
288.0
1.7
253.4
2.9
334.7
0.7
297.5
1.0
380.6
2.0
344.2
2.2
567.6
2.2
534.7
3.7
N.B. AAS- PerkinElmer, AAnalyst 400.
Sample
Adsorption capacity
mg.g-1
qe
SD
RSD %
8.2
0.2
2.2
11.3
0.2
1.7
13.0
0.4
2.7
13.7
0.5
4.0
15.0
1.4
9.4
15.4
0.3
2.2
16.9
0.8
4.9
17.3
1.4
7.8
18.6
0.5
2.5
18.2
0.7
3.8
17.7
1.8
10.2
201
Table H-7.2 : Adsorption equilibrium data of Ni(II) in ternary solute solution, CuNi-Pb, at pH 5, by the best-selected carbon, PSW-P-ad-500 and adsorption capacity,
while initial concentrations were varied < 25 - < 600 mg.L-1.
Adsorption capacity
Solution Concentrations, mg.L-1
mg.g-1
Initial
Equilibrium
SD
Ce
SD
qe
SD
RSD %
C0
22.2
0.1
7.3
0.3
7.5
0.2
2.6
44.1
0.2
35.9
0.3
4.1
0.1
1.7
(66.4
0.2
63.5
0.6
1.7
0.1
5.2)
90.1
0.9
84.4
0.2
3.1
0.2
6.0
PSW-Pad-500 (108.7
0.4
106.6
0.5
1.1
0.2
14.6)
175
1
169.7
0.6
2.9
0.2
7.2
(217
0
215
2
0.9
0.3
31.2)
261
1
256
1
2.4
0.4
18.3
305
2
301
6
3.0
1.6
52.8
1
N.B. The data put in parenthesis are not consistent and therefore eliminated in the
constructed adsorption isotherms.
N.B.2 AAS- PerkinElmer, AAnalyst 400.
Sample
Table H-7.3 : Adsorption equilibrium data of Pb(II) in ternary solute solution, CuNi-Pb, at pH 5, by the best-selected carbon, PSW-P-ad-500 and adsorption capacity,
while initial concentrations were varied 22-600 mg.L-1.
Solution Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ce
SD
22.3
0.1
6.7
1.0
48.1
0.3
26.3
0.5
73.6
0.5
43.2
0.6
97.6
0.7
67.7
2.6
124.3
0.7
87.8
0.3
PSW-P-ad-500
195
1
155
2
249
2
207
2
343
2
306
2
400
1
360
1
600
5
559
4
N.B. AAS- PerkinElmer, AAnalyst 400.
Sample
Adsorption capacity
mg.g-1
qe
SD
RSD %
7.8
0.5
6.3
10.9
0.4
3.4
15.2
0.1
0.5
15.9
0.6
3.7
18.3
0.2
1.0
20.0
0.9
4.5
21.0
0.1
0.6
18.2
0.1
0.4
20
1
5.2
22
1
6.4
202
H-8 : Regeneration Data of Cu(II) While Once Adsorbed Cu(II) on PSW-P-ad500 was Dissolved in Acid solution to Regenerate the Exhausted Carbon.
Table H-8.1 : Desorption data of Cu(II) and adsorption capacity of carbon, while
dissolution was conducted after adsorption in an initial concentration range, from
< 25 to < 700 mg.L-1 at pH 4.5.
Concentrations, mg.L-1
Initial solution Desorbed solution
C0
SD
Cd
SD
24.9
0.05
22.8
0.3
49.9
0.6
30.5
0.5
96.1
0.4
36.8
0.6
PSW-Pad-500
191
1
38.1
0.1
366
3
46.1
1.7
455
0.3
47.7
0.5
693
4
56.8
0.2
N.B. AAS- PerkinElmer, AAnalyst 400.
Sample
Desorption capacity
mg.g-1
qe
SD
RSD %
11.4
0.1
1.14
15.2
0.3
1.66
18.4
0.3
1.52
19.1
0.0
0.24
23.0
0.8
3.63
23.9
0.2
1.00
28.4
0.1
0.36
Table H-8.2 : Adsorption-desorption data of Cu(II) while dissolution was
conducted after adsorption on PSW-P-ad-500 in an initial concentration range, from
< 25 to < 700 mg.L-1 at pH 4.5.
Initial
concentration
mg.L-1
Amount
adsorbed
mg.g-1
RSD
RSD
%
Recovery
%
%
Amount
desorbed
mg.g-1
24.9
11.5
0.2
11.4
1
99
49.9
15
1
15.2
2
101
96.1
18.2
2
18.4
2
101
191
18
4
19.1
0.2
106
366
23
4
23
4
100
455
23
4
23.9
1
104
693
25
24
28.4
0.4
114
203
APPENDIX I
UV-Vis Spectrophotometer Determined Adsorption Equilibrium
Data of Cr(VI) in Single Solute Solution By Various Activated
Carbons and C-HFO Composite Adsorbents
Instrument Parameters:
Instrument: PerkinElmer Lambda 25, Serial No. 101N3061203
Method: Chromate
Ordinate mode: Single wavelength
Wavelength: 540.0 nm
Slit: UV/Vis: 1.00 nm.
Calibration Curve:
Equation, y = 7.357720e-01.x
Residual error: 0.011391
Correlation coefficient: 0.998780
0.9
Absorbance
r 2 = 0.999
0.6
0.3
0
0
0.4
0.8
Concentration, mg/L
Figure I
Calibration curve for Cr(VI).
1.2
204
I-1 : Adsorption Equilibrium Data of Cr(VI) for the Adsorption in a Single
Point Initial Concentration (< 10 mg.L-1), at an unadjusted solution pH (4.79),
for Adsorbent Scrutinizing Purposes.
Table I-1.1 : Adsorption equilibrium data of Cr(VI) by various adsorbents and
calculated adsorption capacities, while adsorption was conducted at a constant initial
concentration and at an unadjusted pH of 4.79, for adsorbent scrutinizing purposes.
Sample
C-HFO-1
C-HFO-2
C-HFO-3
C-HFOCa-1
CAC
PSW-P-500
CPW-P-500
PSW-P-ad-500
Solution Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ce
SD
8.13
0.04
8.03
0.08
8.13
0.04
7.87
0.07
8.13
0.04
7.96
0.05
8.13
0.04
1.81
0.00
8.13
0.04
1.87
0.01
8.13
0.04
ND
8.13
0.04
ND
8.13
0.04
ND
-
Adsorption capacity
mg.g-1
qe
SD
RSD %
0.05
0.06
125
0.13
0.06
42
0.08
0.02
18
3.17
0.02
0.75
3.13
0.02
0.52
-
Table I-1.2 : QC checking for the adsorption equilibrium data of Cr(VI) by various
adsorbents, while adsorption was conducted at a constant initial concentration and at
an unadjusted pH, for adsorbent scrutinizing purposes.
Sample
QC
Concentrations, mg.L-1
Spiked
Observed
0.5
0.46
0.5
0.43
0.5
0.47
0.5
0.45
Recovery %
Average
Recovery %
92
86
94
90
91 ± 4
205
I-2 : Adsorption Equilibrium Data of Cr(VI) by Various Selected Adsorbents,
at Varying pH (3 - 8), While Initial Concentration was Kept Constant (ca. 40
mg.L-1) to Determine Optimum pH for Adsorption.
Table I-2.1 : Adsorption equilibrium data of various adsorbents and calculated
adsorption capacity, while pH were varied from 3 to 8, keeping other parameters
constant.
Solution Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ce
SD
Sample
pH
CAC
3
4
5
6
7
8
40.1
39.0
40.2
37.9
40.2
37.8
0.6
0.2
0.5
0.2
0.4
0.1
1.2
9.0
16.0
25.7
33.0
32.8
0.0
0.1
0.0
0.5
0.1
0.0
Adsorption
capacity
mg.g-1
qe
19.5
15.0
12.1
6.1
3.6
2.5
PSW-P-500
3
4
5
6
7
8
40.1
39.0
40.2
37.9
40.2
37.8
0.6
0.2
0.5
0.2
0.4
0.1
ND
0.8
5.1
5.5
14.7
15.5
0.0
0.0
0.0
0.3
0.0
20.0
19.1
17.6
16.2
12.8
11.1
CPW-P500
3
4
5
6
7
8
40.1
39.0
40.2
37.9
40.2
37.8
0.6
0.2
0.5
0.2
0.4
0.1
0.1
2.9
4.3
10.9
15.3
17.1
0.0
0.0
0.0
0.1
0.1
0.0
20.0
18.1
18.0
13.5
12.5
10.3
PSW-P-ad
3
4
5
6
7
8
40.1
39.0
40.2
37.9
40.2
37.8
0.6
0.2
0.5
0.2
0.4
0.1
ND
3.8
9.6
14.9
18.3
16.9
0.0
0.0
0.2
0.0
0.0
0.0
20.0
17.6
15.3
11.5
11.0
10.4
C-HFOCa-1
3
4
5
6
7
8
40.1
39.0
40.2
37.9
40.2
37.8
0.6
0.2
0.5
0.2
0.4
0.1
6.8
13.3
19.2
25.0
27.8
25.5
0.1
0.3
0.0
0.0
0.1
0.1
16.6
12.8
10.5
6.5
6.2
6.2
206
Table I-2.2 : QC checking for the adsorption equilibrium data of various
adsorbents, while pH were varied from 3 to 8, keeping other parameters constant.
Sample
QC
Concentrations, mg.L-1
Spiked
Observed
0.5
0.46
0.5
0.45
1.0
0.82
1.0
0.87
0.5
0.48
0.5
0.45
0.5
0.43
Recovery %
Average
Recovery %
92
90
82
87
96
89
85
89 ± 5
I-3 : Adsorption Equilibrium Data of Cr(VI) By the Best-Proposed Prepared
Activated Carbon and Commercial ones, at an Optimum pH (e.g. 3), While
Initial Concentrations were Varied to Predict Adsorption Behaviour by Fitting
the Data to Equilibrium Models.
Table I-3.1 : Adsorption equilibrium data of Cr(VI) by a prepared and a
commercial activated carbon and calculated adsorption capacities, while initial
concentrations were varied, keeping initial pH constant, at 3.
Sample
CAC
PSW-P-ad-500
Solution Concentrations, mg.L-1
Initial
Equilibrium
SD
Ce
SD
C0
39.9
0
2.5
0.3
59.4
0
4.4
0.1
77.2
1
10
1
117.0
0
25.6
0.3
133.0
1
32
3
151.9
2
43
2
193.8
2
67
4
Adsorption capacity
mg.g-1
qe
SD
RSD %
18.7
0.1
0.3
27.5
0.1
0.5
33.6
1.2
3.5
45.7
0.2
0.4
50
1
1.6
55
0
0.2
63
3
4.2
39.9
59.4
77.2
117.0
133.0
151.9
193.8
29.4
33
40
44
46
54
0
0
1
0
1
2
2
ND
0.7
12
37
45
61
86
0.0
3
3
5
3
1
0.1
2
2
2
1
1
0.3
6.4
4.2
4.6
1.7
2.3
207
Table I-3.2 : QC checking while determining Cr(VI) for obtaining adsorption
equilibrium data of Cr(VI) by the best-proposed and one of the commercial activated
carbons, at an optimum pH, 3, while initial solution concentrations were varied from
40 to 194 mg.L-1.
Sample
QC
Concentrations, mg.L-1
Spiked
Observed
0.47
0.47
0.48
0.5
0.47
0.47
0.48
0.46
Recovery %
93.4
93.7
95.0
94.9
94.5
95.4
91.6
Average
Recovery %
94 ± 1
208
APPENDIX J
Intensity
ICP-MS Determined Adsorption Equilibrium Data of As(III) and
As(V) for the Adsorption in a Single Point Initial Concentration By
a Typical Prepared Activated Carbon and its modified ones, namely,
PSW-P-500 and C-HFO-1 respectively.
200000
180000
160000
140000
120000
100000
80000
60000
40000
20000
0
r 2 = 0.999986
0
10
20
30
40
50
60
70
80
90 100 110 120 130 140 150
Concentration, µg/L
Figure J
Calibration curve of Arsenic using standard solution of As(III).
Table J-1 : Adsorption equilibrium data of As(V) and As(III) on a prepared
activated carbon (PSW-P-500) and on its modified ones, C-HFO composite.
Sample
Species
pH
C-HFO-1
PSW-P-500
As(V)
5.5
5.5
C-HFO-1
PSW-P-500
As(III)
7.6
7.6
Solution concentrations
µg.L-1
Initial
Equilibrium
1667
671
1667
886
2158
2158
730
1115
Adsorption capacity
µg.g-1
Average
RSD %
996
2
390
5
1429
522
1
1
Table J-2 : QC checking for the adsorption equilibrium data of As(V) and As(III)
on a prepared activated carbon and on its modified ones.
Sample
QC
Concentrations, µg.L-1
Spiked
Observed
80
78
Recovery %
97.5
209
APPENDIX K
Adsorption Kinetics Data of Cu(II), Ni(II), Pb(II) and Cr(VI)
K-1: FAAS Determined Kinetic Data of Cu(II) While Adsorbed on PSW-P-ad500 at a Constant Initial pH 5.
Table K-1.1 : Kinetic data of Cu(II) for the adsorption onto an selected activated
carbon, PSW-P-ad-500.
Solution concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ct
SD
0.5
48.6
0.4
37.4
1.3
2
48.6
0.4
31.6
1.5
3
48.8
0.3
23.9
0.9
4
48.0
0.1
21.3
1.6
6
48.0
0.1
20.7
0.4
8
48.0
0.1
17.7
1.6
12
48.0
0.1
16.2
1.0
20
48.6
0.4
14.0
2.0
24
48.6
0.4
13.2
0.3
48
48.6
0.4
11.5
0.3
72
48.2
0.3
10.4
1.2
N.B. AAS- GBC, Avanta.
Time, h
Adsorption capacity
mg.g-1
qt
SD
RSD %
5.6
0.7
11.7
8.5
0.7
8.4
12.5
0.3
2.7
13.3
0.8
5.8
13.6
0.2
1.4
15.1
0.8
5.4
15.9
0.5
3.2
17.3
0.8
4.6
17.7
0.3
1.6
18.5
0.2
1.3
18.9
0.4
2.4
Tables K-1.2 : QC checking for the adsorption kinetic data of Cu(II) on a prepared
activated carbon, PSW-P-ad-500.
Sample
QC
Concentrations, mg.L-1
Spiked
Observed
3.499
3.492
3.5
3.482
3.483
3.499
Recovery %
99.983
99.757
99.483
99.517
99.983
Average
Recovery %
99.7 ± 0.2
210
K-2: FAAS Determined Kinetic Data of Ni(II) While Adsorbed on PSW-P-ad500 at a Constant Initial pH 5.
Table K-2.1 : Kinetic data of Ni(II) for the adsorption onto an selected activated
carbon, PSW-P-ad-500.
Solution Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ct
SD
0.5
44
2
32
5
2
44
2
32.4
0.8
4
44
2
23.9
0.4
6
42
1
13.0
0.5
8
44
2
14.1
0.2
12
44
2
13.4
0.4
16
44
2
12.4
0.8
20
44
2
12.4
0.2
24
44
2
12.5
0.2
48
44
2
12.6
0.3
72
44
2
12.6
0.0
N.B. AAS- PerkinElmer, AAnalyst 400.
Time, h
Adsorption capacity
mg.g-1
qt
SD
RSD %
6.1
1.6
25.9
5.6
0.6
10.1
9.9
1.2
11.8
14.4
0.0
0.3
14.8
1.1
7.2
15.2
0.8
5.1
15.6
0.6
3.7
15.6
1.1
6.8
15.6
1.1
7.1
15.6
0.8
5.3
15.5
1.0
6.5
K-3: FAAS Determined Kinetic Data of Pb(II)) While Adsorbed on PSW-P-ad500 at a Constant Initial pH 5.
Table K-3.1 : Kinetic data of Pb(II) for the adsorption onto an selected activated
carbon, PSW-P-ad-500.
Solution Concentrations, mg.L-1
Initial
Equilibrium
SD
Ct
SD
C0
0.5
95
1
49
1
2
95
1
31
3
4
95
1
32
6
6
95
1
12
1
8
95
1
17
1
16
95
1
17
0
20
95
1
18
0
24
95
1
16
2
48
95
1
16
0
72
95
1
17
2
N.B. AAS- PerkinElmer, AAnalyst 400.
Time, h
Adsorption capacity
mg.g-1
qt
SD
RSD %
23
1
3.1
32
1
4.0
32
3
8.6
42
0
0.5
39
0
1.2
39
0
2.0
38
0
0.9
39
1
1.5
40
0
0.0
39
1
2.3
211
K-4: US-Vis Spectrophotometeric Determined Kinetic Data of Cr(VI) While
Adsorbed on PSW-P-ad-500 at a Constant Initial pH 3.
Table K-4.1 : Kinetic data of Cr(VI) for the adsorption onto an selected activated
carbon, PSW-P-ad-500.
Time, h
0.5
1
2
4
6
8
16
24
48
72
Solution Concentrations, mg.L-1
Initial
Equilibrium
C0
SD
Ct
SD
77
1
62
1
77
1
61
0
77
1
52
0
77
1
43
1
77
1
39
1
77
1
40
2
77
1
25
0
77
1
22
0
77
1
16
2
77
1
12
1
Adsorption capacity
mg.g-1
qt
SD
RSD %
8
0
1.4
8
0
2.5
13
0
2.3
17
0
0.5
19
1
3.9
19
0
2.1
26
0
0.4
28
0
1.2
31
0
1.5
33
0
0.5
Tables K-4.2 : QC checking for the adsorption kinetic data of Cr(VI) on a prepared
activated carbon, PSW-P-ad-500.
Sample
QC
Concentrations, mg.L-1
Spiked
Observed
0.45
0.50
0.45
0.5
0.45
0.45
0.38
0.42
Recovery %
89.06
99.36
90.98
89.10
89.92
75.92
84.98
Average
Recovery %
88±7
212
APPENDIX L
Presented Papers and Expected Publications from This Study
1.
M. Adil, M. M. Rahman and A. M. Yusof. “Factors affecting the pore development
in preparing highly porous activated carbon from palm kernel shell and coconut shell
using chemical activation method”. Orally presented at the Conference of SKAM-17
2004, (e-Proceeding of SKAM 17), 24-26 August 2004, Kuantan, Pahang, Malaysia.
Will be submitted to Carbon.
2.
M. Adil and A. M. Yusof. “Removal of heavy metal ions with acid activated carbons
derived from oil palm and coconut shells”. Orally presented at the Conference of
SKAM-18 2005, (e-Proceeding of SKAM 18), 12-14 September 2005, Johor Bahru,
Johor, Malaysia. Will be submitted to Water Research.
3.
A. M. Yusof and M. Adil. “Phosphoric acid activated carbon as borderline and soft
metal ions scavenger”. Poster presented at the Conference of SKAM-18 2005, (eProceeding of SKAM 18), 12-14 September 2005, Johor Bahru, Johor, Malaysia.
Will be submitted to Carbon.
4.
A. M. Yusof and M. Adil. “Dispersion of Hydrated Iron(III) Oxide onto Acid
Activated Carbon for the Removal of Cr(VI), As(III) and As(V)”. Accepted for
presentation at Seventh International Conference on Methods and Applications of
Radioanalytical Chemistry (MARC VII), April 3-7, 2006, Kailua-Kona, Hawaii,
USA.
5.
M. Adil and A. M. Yusof. “Removal of Cu (II), Ni(II) and Pb(II) in Single- and
Ternary- Solutes Solutions using Activated Carbon Derived from Oil Palm and
Coconut Shells”. Will be submitted to Water Research.