REMOVAL OF PHOSPHATE IONS FROM AQUEOUS SOLUTIONS USING

BAUXITE AND KAOLINITE OBTAINED FROM MALAWI

M.Sc. (Applied Chemistry) Thesis

By

MOSES WITNESS KAMIYANGO
B.Ed (Science)-University of Malawi




Submitted to the Department of Chemistry, Faculty of Science, in fulfilment of the
requirements for the degree of Master of Science (Applied Chemistry)

University of Malawi
Chancellor College

September, 2009
DECLARATION

I the undersigned hereby declare that this thesis is my own original work which has
not been submitted to any other institution for similar purposes. Where other peoples
work has been used acknowledgements have been made.


______________________________________________________________
_____________________________________
__Full Legal Name



____________________________________
Signature


____________________________________
Date



CERTIFICATE OF APPROVAL

The undersigned certify that this thesis represents the students own work and effort
and has been submitted with our approval.

Signature: _______________________ Date: ____________________________
Masamba W.R.L., PhD (Professor)
Main Supervisor

Signature: _______________________ Date: ____________________________
Sajidu S.M.I., PhD (Associate Professor)
Member, Supervisory Committee

Signature: _______________________ Date: ____________________________
Fabiano E., PhD (Senior Lecturer)
Member, Supervisory Committee

Signature: _______________________ Date: _____________________________

Mwatseteza J.F., PhD (Senior Lecturer)

Head, Chemistry Department
DEDICATION


I dedicate this thesis to my sister Funny and brother William who supported me
throughout my formal education. I also dedicate this work to my late grandmother
Motsetse and late sister Miriam who through their demise I have found courage and
strength to complete my work.
ACKNOWLEDGEMENTS


I wish to thank all members of my supervisory team: Professor W.R.L. Masamba,
Associate Professor S.M.I. Sajidu and Dr. E. Fabiano for their support. I
acknowledge invaluable support offered by the late Professor E.M.T. Henry during
the early stages of my studies. Acknowledgements should also go to the laboratory
staff in the Chemistry Department of Chancellor College for their assistance.

I also thank the International Science Programme (ISP) through the International
Programme in the Chemical Sciences (IPICS) at Uppsala University for the
fellowship. Finally, I thank God for the good health during my study, and for allowing
me to complete this work.
vi

ABSTRACT


Surface waters in some parts of Malawi are known to contain high concentrations of
phosphate ions due to, among other reasons, the discharge of incompletely treated
municipal and industrial wastewaters into streams. High levels of phosphates in
surface waters pose a threat to aquatic life as such there is a need for materials that
can bind phosphate ions in wastewater during treatment. This thesis therefore
concerns characterisation of locally sourced kaolinite to determine its point of zero net
proton charge and bench studies on removal of phosphate ions from aqueous solutions
by locally sourced kaolinite and bauxite.

Raw bauxite, raw kaolinite, and treated kaolinite (at a dosage of 10 g/L) reduced
concentration of phosphate ions in solutions by 93.2 0.152, 19.3 0.344, and 50.6
0.436 % respectively. Raw bauxite was more effective than kaolinite because of the
presence of minerals that have high affinities for phosphate ions, namely, goethite and
gibbsite. The phosphate removal capacity of kaolinite increased after acid treatment
arguably due to the release of more calcium ions from calcium carbonates that were
present in the kaolinite samples as impurities.
vii

The percent removal of phosphate ions increased with decreasing pH for bauxite, with
high removal achieved below pH 5. This trend was attributed to a ligand exchange
reaction mechanism involving phosphate ions and reactive hydroxyl groups on the
goethite and gibbsite surfaces. Phosphate removal by both raw and treated kaolinite
was achieved through an ion exchange mechanism between pH 2 and 5, whilst
precipitation of hydroxyapatite dominated above pH 7.

Carbonate ions inhibited precipitation of hydroxyapatite and competed with
phosphate for adsorption sites resulting in reduced capacities for kaolinite and bauxite
respectively. Sulphate ions reduced the phosphate removal capacity of bauxite by
competing for active sites with phosphate ions. Both calcium and magnesium ions
enhanced phosphate precipitation by kaolinite and phosphate adsorption on bauxite.
Calcium and magnesium ions enhanced adsorption of phosphate ions through
electrostatic interactions whereas precipitation was enhanced through an increase in
calcium ions, and reduction of carbonate ions in solution as a result of the formation
of magnesium-carbonate ion pairs. Kaolinite recorded a very low phosphate removal
capacity as such cannot be used as a phosphate removing agent during wastewater
treatment. This is in contrast to bauxite which recorded a higher phosphate removal
capacity.






viii


TABLE OF CONTENTS
ABSTRACT .................................................................................................................. vi
LIST OF FIGURES ...................................................................................................... xii
LIST OF TABLES .......................................................................................................xiv
LIST OF ACRONYMS AND ABBREVIATIONS ........................................................ xv
CHAPTER ONE: INTRODUCTION ...........................................................................1
1.1 Background ..................................................................................................1
1.2 Problem statement ........................................................................................2
1.3 Aim and objectives of the study and thesis outline ........................................4
1.3.1 Aim of the study .....................................................................................4
1.3.2 Specific objectives ..................................................................................4
1.3.3 Thesis outline .........................................................................................4
CHAPTER TWO: LITERATURE REVIEW ..............................................................5
2.1 Chemical forms of phosphorus in water ........................................................5
2.2 Sources of phosphates in wastewater ............................................................8
2.3 Effects of excess phosphorus on aquatic ecosystems: Eutrophication .......... 10
2.4 Levels of phosphate pollution in Malawi .................................................... 11
2.5 Methods for phosphate removal during wastewater treatment ..................... 12
2.5.1 Chemical precipitation .................................................................. 12
2.5.2 Biological phosphorus removal ..................................................... 14
2.5.3 Crystallisation technology ............................................................. 15
ix

2.5.4 Ion exchange ................................................................................. 16
2.5.5 Magnetic attraction ........................................................................ 17
2.5.6 Low cost materials for phosphate removal from wastewater .......... 17
2.5.6.1 Phosphate precipitating low cost materials ............................... 17
2.5.6.2 Low cost phosphate adsorbents ............................................... 18
2.6 Chemical composition and acid-base properties of Linthipe kaolinite and
Mulanje bauxite. ......................................................................................... 19

2.6.1 Gibbsite ......................................................................................... 20
2.6.2 Goethite ........................................................................................ 22
2.6.3 Kaolinite ....................................................................................... 23
2.7 Complexation and adsorption ..................................................................... 26
2.7.1 Development of reactive functional groups at the metal oxide-
solution interface ........................................................................... 27

2.7.2 Adsorption of ions at the metal (hydr)oxide-solution interface :
The electrostatic double layer model ............................................. 28

2.8 Protonation of surface functional groups and charge balance ...................... 30
2.9 Patch-wise surface charge heterogeneity on kaolinite and the point of
zero charge ................................................................................................. 34

2.10 Modeling phosphate adsorption on kaolinite and bauxite ............................ 34
2.11 Precipitation of calcium phosphates ............................................................ 35
CHAPTER THREE: MATERIALS AND METHODS ............................................. 38
3.1 Materials .................................................................................................... 38
3.1.1 Adsorbents .................................................................................... 38
3.1.2 Chemicals, reagents and instruments ............................................. 38
3.2 Methods ..................................................................................................... 39
x

3.2.1 Preparation of kaolinite and bauxite samples ................................. 39
3.2.2 Preparation of solutions ................................................................. 39
3.2.2.1 Reagents .................................................................................... 39
3.2.2.1.1 Sodium Carbonate (0.01 mol/L and 1000 mg/L) .................. 39
3.2.2.1.2 Nitric acid (0.02359 mol/L and 1+1) ................................... 40
3.2.2.1.3 Lanthanum solution ............................................................. 40
3.2.2.1.4 Sodium hydroxide (0.020 mol/L) ........................................ 40
3.2.2.1.5 Hydrochloric acid (0.3 mol/L) ............................................. 40
3.2.2.1.6 Sodium nitrate (1.0 mol/L) .................................................. 41
3.2.2.1.7 Sulphate solution (1000 mg/L) ............................................ 41
3.2.2.1.8 Magnesium solution (1000 mg/L) ....................................... 41
3.2.2.1.9 Calcium solution (1000 mg/L) ............................................. 41
3.2.2.2 Standard solutions ...................................................................... 42
3.2.2.2.1 Standard phosphate solution ................................................ 42
3.2.2.2.2 Standard calcium solution (for AAS determination of
calcium). ............................................................................ 42

3.2.3 Determination of phosphate ions in solution using ion chromatography ...... 42

3.2.4 Determination of Calcium ions in solution..................................... 43
3.2.5 Potentiometric titration of raw kaolinite samples ........................... 44
3.2.6 Effect of suspension pH on the amount of phosphate ions
removed by bauxite and kaolinite .................................................. 45

3.2.7 Determination of calcium ions released into solution from
kaolinite ........................................................................................ 45

3.2.8 Effect of bauxite and kaolinite dosage on amount of phosphate
ions removed ................................................................................. 46

3.2.9 Effect of contact time on the amount of phosphate ions removed
by bauxite and kaolinite................................................................. 47

xi

3.2.10 Effect of initial phosphate concentration on the phosphate
removal capacity of bauxite and kaolinite ...................................... 48

3.2.11 Effect of magnesium, calcium, sulphate and
carbonate/bicarbonate ions on amount of phosphate ions
removed by bauxite and kaolinite .................................................. 49

3.2.12 Multi-interactive effect of magnesium, calcium, sulphate, and
bicarbonate ions on the phosphate removal efficiency of bauxite. .. 50
CHAPTER FOUR: RESULTS AND DISCUSSION .................................................. 51
4.1 Point of zero net proton charge for kaolinite ............................................... 51
4.2 Effect of suspension pH on the amount of phosphate ions removed by
kaolinite and bauxite .................................................................................. 53

4.2.1 Description of reactions resulting in removal of phosphate ions
by kaolinite and bauxite................................................................. 56

4.3 Effect of kaolinite and bauxite dosage on the amount of phosphate ions
removed ..................................................................................................... 61

4.4 Effect of contact time on the amount of phosphate ions removed by
bauxite and kaolinite .................................................................................. 64

4.5 Effect of initial phosphate concentration on the phosphate removal
capacity of bauxite and kaolinite ................................................................ 71

4.6 Effect of calcium, magnesium, sulphate and carbonate ions on phosphate
uptake by raw and treated clay.................................................................... 75

4.7 Effect of magnesium, calcium, sulphate and carbonate/bicarbonate ions
on phosphate uptake by bauxite .................................................................. 79

4.8 Multi-interactive effect of magnesium, calcium, sulphate, and bicarbonate
ions on the phosphate removal efficiency of bauxite. .................................. 83
CHAPTER FIVE: CONCLUSIONS AND RECOMMENDATIONS ....................... 85
5.1 Conclusions ................................................................................................ 85
5.2 Recommendations ...................................................................................... 88
REFERENCES ............................................................................................................ 90
xii

LIST OF FIGURES

Figure 1: Structures of cyclotriphosphate and cyclotetraphosphate ...................................6
Figure 2: A schematic representation of a polyphosphate .................................................7
Figure 3: A schematic representation of a branched inorganic phosphate .........................7
Figure 4: Chemical phosphate precipitation. .................................................................. 12
Figure 5: Biological phosphorus removal ....................................................................... 14
Figure 6: The DHV Crystalactor .................................................................................... 15
Figure 7: Sphere model of gibbsite structure.. ................................................................ 21
Figure 8: Ring structure model for gibbsite .................................................................... 21
Figure 9: Ball and stick model for goethite. ................................................................... 22
Figure 10: Silica tetrahedron and layer. .......................................................................... 24
Figure 11: Clay octahedron and octahedral sheet. .......................................................... 24
Figure 12: Kaolinite structure ........................................................................................ 25
Figure 13: The electrostatic double layer model ............................................................. 29
Figure 14: Experimental net proton surface density curve for Na-kaolinite .................... 52
Figure 15: Plot of % phosphate removal against pH for raw and treated kaolinite .......... 53
Figure 16: Plot of % phosphate bound against pH for bauxite ........................................ 55
Figure 17: Fractional composition of phosphate species in solution at different pH ........ 57
Figure 18: Plot of % phosphate removal against dosage for kaolinite ............................. 61
Figure 19: Plot of % phosphate removal against dosage for bauxite ............................... 62
Figure 20: Plot of phosphate uptake against time for kaolinite ....................................... 65
Figure 21: Plot of phosphate uptake against time for phosphate adsorption on bauxite ... 66
Figure 22: Second order fits for phosphate precipitation by kaolinite ............................. 68
Figure 23: First order fits for phosphate precipitation by kaolinite ................................. 68
Figure 24: Second order fits for phosphate adsorption on bauxite .................................. 69
xiii

Figure 25: First order fits for phosphate adsorption on bauxite ....................................... 69
Figure 26: Plot of phosphate uptake against initial phosphate concentration for
phosphate removal by kaolinite ..................................................................... 71

Figure 27: Phosphate uptake by bauxite fitted to the Freundlich equation using non-
linear regression ............................................................................................ 73

Figure 28: Vant Hoff plot for phosphate adsorption on bauxite ..................................... 75
Figure 29: Effect of competing ions on phosphate removal by raw kaolinite .................. 76
Figure 30: Effect of competing ions on phosphate uptake by treated kaolinite ................ 76
Figure 31: Effect of competing ions on adsorption of phosphate ions on bauxite ............ 80
Figure 32: Multi-competitive effect of calcium, magnesium, carbonate, and sulphate
ions on phosphate adsorption on bauxite ........................................................ 83




















xiv

LIST OF TABLES

Table 1: Chemical composition of Linthipe kaolinite .................................................... 20
Table 2: Solution combinations for the effect of initial phosphate concentration ........... 48
Table 3: Solution combinations for the effect of competing ions ................................... 49
Table 4: Ca
2+
concentration in solution and calculated saturation index values .............. 60
Table 5: Equilibrium constants and Gibb's free energy change values ........................... 74
























xv

LIST OF ACRONYMS AND ABBREVIATIONS

ACP Amorphous Calcium Phosphate
APHA American Public Health Association
CSC Correctional Service of Canada
DDL Diffuse Double Layer
GSoM Geological Survey department of Malawi
HAP Hydroxyapatite
PZC Point of Zero Charge
PZNPC Point of Zero Net Proton Charge
TCP Tricalcium Phosphate
UNEP United Nations Environmental Programme

















1

CHAPTER ONE: INTRODUCTION

1.1 Background

Freshwater availability and use, as well as the conservation of aquatic resources, are
key to human well-being, but water quality degradation from human activities
continues to harm human and ecosystem health (UNEP, 2007). According to the
fourth global environmental outlook assessment (UNEP, 2007), the most ubiquitous
freshwater quality problem is high concentrations of nutrients (mainly phosphorus and
nitrogen) resulting in eutrophication, and significantly affecting human water use.
Nutrient pollution from municipal wastewater treatment plants and from agricultural
and urban non-point source run-off remains a major global problem, with many health
implications.

The Malawi State of the Environment Report indicates that the country faces
contamination of its water resources arising mainly from poor sanitation and improper
disposal of wastes, agro-chemicals and effluent from industries, hospitals and other
institutions (Malawi Government, 2002). According to Ferguson and Mulwafu
(2004), the release of untreated sewage directly into rivers and streams is one of the
major causes of water pollution in Malawi. It is partly through the discharge of
untreated or inadequately treated sewage, that some rivers and streams are becoming
loaded with phosphate ions, resulting in degradation of water quality.
2

The Malawi Government, however, seeks to promote effective water pollution
monitoring and prevention programmes based on enforceable water quality guidelines
and standards (Malawi Government, 2002).

1.2 Problem statement

Presence of excess phosphorus (mainly in the orthophosphate form) in stagnant and
flowing water bodies pose a threat to aquatic life. This is due to stimulated growth of
aquatic plants that result in depleted oxygen levels when they decompose, as well as a
bloom of blue-green algae some of which produce cyanotoxins (such as saxitoxins
and anatoxin-a) that are more toxic than cobra venom (Skulberg et al., 1984;
Carpenter et al., 1998).
Sources of phosphate ions in flowing and stagnant water bodies include domestic and
industrial effluents (mainly as a result of use or manufacturing of products containing
phosphate formulations) and excessive fertilizer application to soils (Carpenter et al.,
1998; Smith et al., 1999). A study by Chipofya and Matapa (2003) revealed that the
Mudi reservoir, which is a raw water source for Blantyre water board, is infested with
blue green algae as a result of nutrient inflow from various catchments. Removing
phosphate ions during wastewater treatment is thus essential to minimize phosphorus
loading of receiving rivers and dams (Hammer and Hammer Jr., 2001).
Studies in Blantyre and Zomba, Malawi, have revealed presence of excessive
phosphate ions in effluent from wastewater treatment plants (Kwanjana, 2003; Sajidu
et al., 2007). This indicates inefficiency of the conventional biological filter plants
that are in use, in removing phosphate ions from wastewater. The established methods
3

for wastewater phosphate removal, i.e. biological uptake and chemical precipitation,
are either expensive to run in developing countries or have poor operation stability
(Morse et al., 1998; Akhurst et al., 2006; Li et al., 2006; Karageorgiou et al., 2007;
Huang et al., 2008). As a result of high operational costs and poor operation stability
associated with the established methods, there is a growing interest in search for
cheap materials that can remove phosphate ions either through adsorption or
precipitation of phosphate salts (Pradhan et al., 1998; Johansson and Gustafsson,
2000; Zeng et al., 2004; Kostura et al., 2005; Akhurst et al., 2006; Ozacar, 2006;
Chen et al., 2007; Huang et al., 2008).
Locally sourced kaolinite and bauxite were chosen to be tested for their phosphate
removal capacities following reports that indicated adsorption of phosphate on
kaolinite, gibbsite and goethite minerals obtained from other parts of the world (Chen
et al., 1973; Persson et al., 1996; Kubicki et al., 2007; Stachowicz et al., 2008). Huge
reserves of kaolinite (800 million metric tonnes) and bauxite (26 million metric
tonnes) are available in Malawi (Yager, 2006). The natural abundance of kaolinite and
bauxite provides the possibility of long term or sustainable use if the materials can be
recycled. Waste materials from digestion of bauxite for alumina (red mud) can be an
alternative for application in wastewater treatment systems if the available bauxite
resources are mined and digested for alumina (Pradhan et al., 1998; Huang et al.,
2008).
Use of locally sourced kaolinite and bauxite in wastewater treatment systems would
require well designed preliminary bench studies to obtain information on the
interaction of the adsorbents with phosphate ions in terms of kinetics, reaction
mechanisms, and effects of solution physical and chemical composition. This study
4

was therefore carried out to obtain information on the chemical interactions between
the adsorbents and phosphate ions.

1.3 Aim and objectives of the study and thesis outline

1.3.1 Aim of the study


The study was undertaken to investigate the use of locally sourced kaolinite and
bauxite as adsorbents for phosphate ions in aqueous solutions.

1.3.2 Specific objectives


(i) To determine the point of zero net proton charge for the locally sourced kaolinite
(ii) To determine the effects of conditions such as dosage, initial suspension pH,
contact time, initial phosphate concentration, presence of competing ions
(calcium, magnesium, carbonate and sulphate), and temperature on phosphate
removal using kaolinite and bauxite.

1.3.3 Thesis outline

The outline of the thesis is as follows: Chapter 2 provides the literature review on
chemical forms of phosphorus in water, phosphate sources and effects, phosphate
removal technologies, and chemical properties of kaolinite and bauxite ; chapter 3
presents the materials and methods whereas results and discussions are presented in
chapter 4. Chapter 5 presents conclusions and recommendations for further study.


5

CHAPTER TWO: LITERATURE REVIEW

This chapter starts with a description of chemical forms of phosphorus found in water
and also the various sources of the abundant phosphorus chemical form (phosphate) in
wastewater. Building up on the sources of phosphates, this chapter then presents the
effects of excess phosphates on aquatic ecosystems, and later provides a review of
phosphate levels in Malawi and conventional methods being used to remove
phosphate from wastewater worldwide. This is followed by a review of low cost
materials that have been studied for their phosphate removal capacity in other
countries and later a presentation of the chemical composition of materials that will be
tested for their phosphate removal capacity (locally sourced kaolinite and bauxite)
along with a comprehensive review of their chemical behaviour in aqueous solutions.

2.1 Chemical forms of phosphorus in water

Phosphorus exists in water in either a particulate phase or a dissolved phase.
Particulate matter includes living and dead plankton, precipitates of phosphorus,
phosphorus adsorbed to particulates, and amorphous phosphorus. Dissolved
phosphorus occurs in natural waters and in wastewaters almost solely as phosphates
(APHA, 1989). Phosphates occur in various forms in water and are classified as
monophosphates, condensed phosphates (pyro-, meta-, and polyphosphates), and
organically bound phosphates (APHA, 1989).
6

Monophosphates (orthophosphates) are compounds whose anionic entity, [PO
4
]
3-
, is
composed by an almost regular tetrahedral arrangement of four oxygen atoms centred
by a phosphorus atom. Among the various categories of phosphates, monophosphates
are the most abundant mainly because they are the most stable. (Averbuch-Pouchot
and Durif, 1996). Additionally, all polyphosphates gradually hydrolyze in water to the
stable ortho form (Hammer and Hammer Jr., 2001).

The term condensed phosphates is applied to salts containing polymerized phosphoric
anions. Condensed phosphates are further classified as cyclophosphates,
polyphosphates and branched inorganic phosphates (or ultraphosphates).
Cyclophosphates (metaphosphates) are built up from cyclic anions and have the
composition MPO
3
where M is hydrogen or a monovalent metal. Representatives of
this group are cyclotriphosphate, M
3
P
3
O
9
, and cyclotetraphosphate, M
4
P
4
O
12
, shown
in Figure 1.
Figure 1 (a) Cyclotriphosphate (b) Cyclotetraphosphate

Highly polymerized cyclic phosphates containing as many as 10 to 15
orthophosphoric acid residues have been observed in some samples of condensed
phosphates (Kulaev et al., 2004).
7

Polyphosphates in contrast to cyclophosphates have anions that are composed of
chains in which each phosphorus atom is linked to its neighbours through two oxygen
atoms, thus forming a linear, unbranched structure that may be represented
schematically by Figure 2.

M O P O P O P O
...
P O M
O O
O O
OM OM OM OM


Figure 2 A schematic representation of a polyphosphate

Branched inorganic phosphates are high molecular weight condensed phosphates,
which unlike the linear polyphosphates contain branching points, i.e. phosphorus
atoms that are linked to three rather than two neighbouring phosphorus atoms (Figure
3).

Figure 3 A schematic representation of a branched inorganic phosphate

P O
P O
P
O
... ...
O O O
OM O
OM
P
O
O
O
8

Branched inorganic phosphates undergo unusually rapid hydrolysis in aqueous
solutions irrespective of pH as such they have not been found in living organisms
(Kulaev et al., 2004).

Organic phosphates are phosphates that are bound to plant or animal tissue formed
primarily through biological processes. They include nucleic acids, phospholipids,
inositol phosphates, phosphoamides, phosphoproteins, sugar phosphates, phosphoric
acids, organophosphate pesticides, humic associated organic phosphorus compounds
and organic condensed phosphates in dissolved, colloidal and particle-associated
forms (McKelvie, 2005).

2.2 Sources of phosphates in wastewater

The principle sources of phosphates are point sources such as domestic and industrial
wastewater treatment plant effluents and natural runoff (non-point) from surrounding
uses such as land application of fertilizers and farming operations. Orthophosphates
and certain polyphosphates are major constituents of many commercial cleaning
agents. For example, many synthetic detergents contain 25-45% sodium
tripolyphosphate (Na
5
P
3
O
10
) which acts mainly as a water softener, by chelating and
sequestering Mg
2+
and Ca
2+
in hard water (Greenwood and Earnshaw, 1984).

Trisodium phosphate (Na
3
PO
4
) has scouring, bleaching, and bacteria killing
properties as such it is available in formulations of automatic dish washing powders
(Greenwood and Earnshaw, 1984). Domestic use of phosphate containing synthetic
detergents contributes towards high levels of ortho and polyphosphates in domestic
wastewaters.
9

Various monophosphates are widely used at the industrial level. Sodium hydrogen
phosphate (Na
2
HPO
4
) is widely used in the food industry as an emulsifier in the
manufacture of pasteurized processed cheese and as a starch modifier. Sodium
dihydrogen phosphate (Na
2
H
2
PO
4
) is a solid, water-soluble acid which finds its use
(with NaHCO
3
) in effervescent laxative tablets and in the pH adjustment of boiler
waters. It is also used as a mild phosphatising agent for steel surfaces and as a
constituent in the undercoat for metal paints. Potassium hydrogen phosphate has
buffering properties as such it is added to car-radiator coolants as a corrosion inhibitor
(Greenwood and Earnshaw, 1984). Calcium phosphates also have a broad range of
applications both in the food industry and as bulk fertilizers. The wide application of
phosphate compounds in various industrial processes result in higher phosphate levels
in industrial wastewaters that are in some cases discharged onto surface water bodies
without proper treatment.

Organically bound phosphates are contributed to sewage through body waste and food
residues, and may also be formed from orthophosphates in biological treatment
processes or by receiving water biota. Organic phosphates may occur as a result of the
breakdown of organic pesticides which contain phosphates and they may exist in
solution, as loose fragments, or in the bodies of aquatic organisms (Smith et al.,
1999).

Phosphorus loading of surface water bodies through non-point sources is greatly
linked to excessive application of phosphate containing fertilizers and manure to farm
lands. In many areas, phosphate inputs from fertilizers and manures greatly exceed
phosphorus outputs in farm produce resulting in yearly phosphorus accumulation in
10

the soil. This trend has important implications for phosphate levels in surface waters
because the total amount of phosphates transported in runoff from the landscape to
surface waters increases linearly with the soil phosphorus content (Smith et al., 1999).

2.3 Effects of excess phosphorus on aquatic ecosystems: Eutrophication

Of the many mineral resources required for plant growth, inorganic nitrogen and
phosphorus are the two principle nutrients that limit the growth of terrestrial plants as
well as algae and vascular plants in freshwater and marine ecosystems (Smith et al.,
1999). Waters having relatively large supplies of nutrients are termed eutrophic (well
nourished), and those having poor nutrient supplies are termed oligotrophic (poorly
nourished). Waters having intermediate nutrient supplies are termed mesotrophic and
those receiving greatly excessive nutrient inputs are termed hypertrophic.
Eutrophication is defined as the process by which water bodies become well
nourished through an increase in their nutrient supply (Van den Brandt and Smit,
1998).

The most common effects of increased nitrogen and phosphorus supplies on aquatic
ecosystems are perceived as increases in the abundance of algae and aquatic plants.
The environmental consequences of excessive nutrient enrichment are more serious
and far-reaching than nuisance increases in plant growth alone. Decomposition of
dead nuisance plants can result in oxygen depletion in the water causing death of
aquatic animals such as fish (Carpenter et al., 1998). Microorganisms use much
oxygen during decomposition of dead plants, thus resulting in lowered oxygen levels
in the water. Eutrophication also brings about a shift in phytoplankton species towards
11

dominance of the phytoplankton by blue-green algae (cyanobacteria), some of which
produce cyanotoxins that are more toxic than cobra venom (Skulberg et al., 1984).
Other effects of eutrophication include, reduced water clarity, bad odour and taste of
the water, and a general decrease in perceived aesthetic value of the water body
(Smith et al., 1999). Eutrophication is not limited to lakes and reservoirs, it can also
occur in rivers and streams (Smith et al., 1999). Both phosphorus and nitrogen limit
plant growth, but phosphorus is the primary limiting nutrient in most lakes and
reservoirs consequently most eutrophication management frameworks focus primarily
on phosphorus loading (Hecky and Kilham, 1988).

2.4 Levels of phosphate pollution in Malawi

Reported studies on effluent quality from wastewater treatment plants in Zomba and
Blantyre indicate phosphate levels above the minimum limit of 1.0 mg/L (CSC,
2000). The Soche wastewater treatment plant in Blantyre is a conventional biological
filter plant that receives wastewater from industries such as cloth making and food
processing as well as latrine and septic tank emptyings. A study by Sajidu et al.,
(2007) reported a phosphate influent concentration of 5.39 0.66 mg/L, and an
effluent concentration of 3.86 0.76 mg/L for the plant. In a separate study, a
phosphate effluent concentration of 5.18 0.00 mg/L was reported for the Zomba
wastewater treatment plant (Kwanjana, 2002). Lower phosphate concentrations were
reported for the Limbe wastewater treatment plant with an influent and effluent
concentration of 0.79 0.93 mg/L and 0.63 0.23 mg/L respectively (Sajidu et al.,
2007). Fluctuations in composition of the wastewater influent may result in lower or
higher effluent phosphate concentrations than those reported. Besides the wastewater
12

treatment plants, high phosphate concentrations were reported for Nasolo river (3.20
0.69 mg/L), Limbe stream (3.42 0.00 mg/L), and Mudi river (5.50 3.20 mg/L)
in Blantyre (Sajidu et al., 2007).

2.5 Methods for phosphate removal during wastewater treatment

2.5.1 Chemical precipitation

Chemical precipitation is a physico-chemical process, comprising the addition of a
divalent or trivalent metal salt to wastewater, causing precipitation of an insoluble
metal phosphate that is settled out by sedimentation. Iron and aluminium are the most
suitable metals and are added as chloride or sulphate salts. Lime may also be used to
precipitate calcium phosphate. Chemical precipitation is a flexible technology
allowing for application of the metal salts at several stages during wastewater
treatment (Figure 4).



Influent
Metal salt
Rapid
mix
Flocculant Aid
Primary
clarifier
Aeration
Basin
Seco-
ndary
clarifier
Alternative metal salt
addition points
Sludge to processing
Figure 4: Chemical phosphate precipitation (Sourced from Morse et al., 1998).
13

Removal of phosphates from wastewater using metal salts can be described by simple
chemical reactions involving direct combination of phosphate and the metal ions to
form precipitates, for example a reaction between aluminium sulphate and
orthophosphate can be given by Equation 1 (Hammer and Hammer Jr., 2001).

O H 3 . 14 SO 3 AlPO 2 2PO O .14.3H ) (SO Al
2
- 2
4 4
- 3
4 2 3 4 2
+ + +
(1)

Recent evidence on phosphate removal from aqueous solutions using aluminium
sulphate and aluminium hydroxide (Georgantas and Grigoropoulou, 2007) has shown
that orthophosphate and metaphosphate ions are also removed through a ligand
exchange mechanism in which surface hydroxyl groups on the surface of the
precipitated aluminium hydroxide are exchanged for phosphate ions.

Chemical precipitation typically produces phosphorus bound as a metal salt within the
wasted sludge. The wasted sludge has the potential value of being used as fertilizer
although research on bioavailability of the bound phosphorus is inconclusive (Morse
et al., 1998). Chemical precipitation is an established technology that is easy to
install and operate and can achieve high phosphate removal, however, it requires high
doses of chemicals, there is an increase in sludge production and phosphorus
recyclability is variable (Morse et al., 1998).







14

2.5.2 Biological phosphorus removal

The development of biological phosphorus removal was based on evidence that under
certain conditions, some heterotrophic bacteria in activated sludge could take up
phosphorus in considerable excess to that required for normal biomass growth (luxury
uptake) (Sidat et al., 1999). Biological phosphorus removal is achieved in the
activated sludge process by introducing an anaerobic zone ahead of an aerobic stage
(Figure 5).




In the anaerobic zone, sufficient readily degradable chemical oxygen demand (COD)
must be available, typically as volatile fatty acids provided by pre-fermenting the
sludge using storage or thickeners, or from the addition of acetic acid or sodium
acetate (Sidat et al., 1999). In the absence of oxygen and nitrates, bacteria, such as
Acinetobacter take up the acids and release phosphorus into solution, but in the
aerobic stage luxury uptake occurs, increasing overall phosphorus removal rates to as
much as 80-90%. However, phosphorus removal is variable and, in practice, the
achievement of a low and consistent effluent standard may require complementary
Influent
Aerobic
Clarifier
Return Sludge
Effluent
Anaerobic zone
Figure 5 Biological phosphorus removal (Adapted from Morse et al.,
1998)
15

chemical simultaneous precipitation (Morse et al., 1998). This technology does not
require use of high doses of chemicals, removal of phosphorus and nitrate can be
achieved simultaneously, and the phosphorus is more recyclable (Sidat et al., 1999).
However, biological phosphorus removal has poor operational stability and handling
of huge volumes of sludge maybe more difficult (Morse et al., 1998).

2.5.3 Crystallisation technology

Phosphorus removal is achieved through crystallisation of a phosphate mineral on a
seeding grain in a reactor. The DHV crystalactor
TM
process is an example of
phosphate crystallisation technologies. The DHV Crystalactor
TM
process is based on
the crystallisation of calcium phosphate on a seeding grain, typically sand, within a
fluidized reactor, as shown in Figure 6 (Scholler, undated).



Chemicals
Influent
Periodic injection of seeding grains (0.2 0.6 mm)

Periodic removal of pellets (1 2 mm)
Fluidised bed Grains:
0.2 2 mm
Effluent
Injection
nozzles
Height:
6m
Diameter: 0.5 4 m
Figure 6 The DHV Crystalactor for phosphorus crystallization (Sourced from
Scholler, undated)
16

Process conditions are adjusted to promote calcium phosphate crystallization by
adding either sodium hydroxide or calcium hydroxide. Pellets formed during
crystallization are periodically removed and replaced by small diameter seed grains
and the removed pellets can be recycled by the phosphate industry. A major
advantage of the Crystalactor
TM
technology is that phosphorus removal produces no
additional sludge, only a quantity of water-free pellets consisting almost exclusively
of calcium phosphate (40 - 50%) and seed material (30 - 40%), with other materials
present in small amounts (Scholler, undated). Disadvantages of this technology
include use of chemicals and requirement for operation skills (Morse et al., 1998).

2.5.4 Ion exchange

The RIM-NUT ion exchange-precipitation process is widely used to remove
phosphates and ammonia from wastewater through formation of struvite [magnesium
ammonium phosphate hexahydrate, O H . MgPO ) NH (
2 4 4
6 )]. The process uses a
cationic resin to remove the ammonium ions and a basic resin to remove phosphate
ions. Regeneration of the ion exchange resins releases the ammonium and phosphate
ions that are then precipitated as struvite. Struvite is a good slow release fertilizer, as
such the technology has a high phosphorus recycling potential for agriculture. Despite
having a high potential for phosphorus recycling, the technology is complex, it
requires use of chemicals and disposal of waste eluate is a problem (Morse et al.,
1998).


17

2.5.5 Magnetic attraction

Magnetic attraction systems such as the Smit-Nymegen process uses calcium oxide to
precipitate calcium phosphate attached to magnetite that is later separated using an
induced magnetic field. After isolation, the magnetite is uncoupled from the
phosphate in a separator unit by shear forces and a drum separator. The separated
suspension of calcium phosphate or carbonate in water is further processed depending
on the use of the final product. Magnetic attraction systems can achieve high
phosphate removal; however they are unnecessarily complex and require use of
chemicals (Morse et al., 1998).


2.5.6 Low cost materials for phosphate removal from wastewater

The potential low cost materials for removing phosphate ions during wastewater
treatment can be categorized into two groups, those that involve phosphate
precipitation, and those that adsorb phosphate ions.

2.5.6.1 Phosphate precipitating low cost materials

Materials that have been extensively studied in this group include fly ash, calcite, and
blast furnace slag. Fly ash is a waste product from coal-fired power plants composed
of various metal oxides. Presence of calcium allows for precipitation of calcium
phosphates at high pH levels, whereas iron oxides bind phosphate at low pH levels
through ligand exchange reaction mechanisms (Li et al., 2006; Chen et al., 2007).
18

Generally, favourable conditions for phosphate precipitation are adequate calcium
ions, and high equilibrium pH (>9) (Can and Yildiz, 2006; Chen et al., 2007; Moon et
al., 2007). Phosphate removal by calcite involves a combination of precipitation of
hydroxyapatite and adsorption of phosphate ions on the calcite surface (Karageorgiou
et al., 2007).

Blast furnace slag is an industrial by-product derived from the separation of iron from
iron ore. It is a complex CaO-MgO-Al
2
O
3
-SiO
2
system that also incorporates a
number of minor components that can concentrate on the slag surface during
crystallization or transition to a glassy state (Kostura et al., 2005). Phosphate is
precipitated as hydroxyapatite under strongly alkaline conditions (pH > 9) and large
amounts of soluble calcium ions (Johansson and Gustafsson, 2000).
2.5.6.2 Low cost phosphate adsorbents

Low cost phosphate adsorbents that have been studied include iron oxide tailings, red
mud, and alunite. Iron oxide tailing is an industrial waste derived from iron ore
processing that contain significant amounts of iron oxides. This adsorbent has a high
affinity for phosphate ions and adsorption is high under low pH conditions (Zeng et
al., 2004).

Red mud is a waste material formed during the production of alumina when the
bauxite ore is subjected to caustic leaching (Pradhan et al., 1998). It is a brick
coloured highly alkaline (pH 10 12) sludge containing mostly oxides of iron,
aluminium, titanium, and silica. Red mud has a high phosphate adsorption capacity
because of the presence of iron oxides, but acid treatment is required to lower the pH.
19

For example, red mud can be neutralized by seawater to pH 9.0 0.5 to produce a
commercial adsorbent known as Bauxsol
TM
(Akhurst et al., 2006). Bauxsol
TM
can be
activated by an acid to improve its phosphate adsorption capacity. Initial
neutralization by seawater entails requirement for small amounts of acid during acid
treatment and lower preparation costs as compared to acid activation of raw red mud
(Akhurst et al., 2006). Huang et al., (2008) confirmed low phosphate removal
efficiency for raw red mud as compared to acid activated red mud.
Alunite, KAl
3
(SO
4
)
2
(OH)
6
, is one of the minerals of the jarosite group and is not
soluble in water in its original form (Ozacar, 2006). Alunite gives thermal
decomposition reaction products such as Al
2
O
3
, Al
2
(SO
4
)
3
and K
2
SO
4
when it is
calcined at 973 1023 K (Ozacar and Sengil, 2003). Phosphate ions are adsorbed
onto the resultant metal oxide surfaces via ligand exchange mechanisms. Despite
having a high phosphate removal capacity, requirement for high temperature
calcinations can be a drawback for use of alunite in countries where it is found.

2.6 Chemical composition and acid-base properties of Linthipe kaolinite and
Mulanje bauxite.

Bauxite from Mulanje mountain and kaolinite from Linthipe, Dedza, were
investigated for their potential use as phosphate adsorbents. According to the
Geological Survey Department of Malawi (GSoM), the bauxite is mainly a trihydrate
gibbsite which lies over kaolinite and has free quartz and geothite as the main
contaminants. The kaolinite from Linthipe contains iron oxides and calcium
carbonates (reported as calcium oxide) as the main contaminants (Table 1).

20

Table 1 Chemical composition of kaolinite obtained from Linthipe, as
determined by the GSoM

Parameter Composition (wt.%)
SiO
2
46.7
Al
2
O
3
33.8
Fe
2
O
3
2.0
CaO 1.1
MgO 0.26
K
2
O + Na
2
O 0.28

Gibbsite, kaolinite and iron oxides are the major minerals present in the locally
sourced bauxite as such a detailed understanding of their structural composition and
acid-base properties in aqueous environment is necessary for elucidation of phosphate
binding mechanisms.

2.6.1 Gibbsite

The gibbsite ( ( )
3
OH Al ) structure consists of double layers (AB) of close packed
OH groups with Al atoms occupying two thirds of the octahedral interstices within the
layers. Each Al atom is octahedrally bonded to three O atoms of layer A and three
atoms of layer B. The AB layers are stacked in the sequence AB-BA-AB-BA-
(Figure 7).


21






The structure of gibbsite can also be viewed from the perspective of the Al
3+
ions that
are arranged in a pattern of coalesced hexagonal rings. A single ring of six Al
3+
ions,
joined above and below by six pairs of bridging OH
-
ions, is the smallest recognizable
unit of the gibbsite structure (Goldberg et al., 1996). The
6
) OH ( Al unit can be
represented schematically as an octahedron with each apex being at the centre of a
hydroxyl ion (Figure 8). Two octahedra are joined along one edge by sharing a pair of
OH
-
ions, and six octahedra each sharing two edges yields a
+ 6
12 6
) OH ( Al ring.

Figure 8 Ring structure model for gibbsite
Layer A
Layer B
Figure 7 Sphere model of gibbsite structure. Large spheres
represent OH
-
ions; smaller spheres represent Al
3+
ions in octahedral coordination sites. Part of the
upper OH layer (layer A) is removed to show
arrangement of Al
3+
.

22

Surface functional groups on gibbsite are located at the basal planes and on the edges.
At the basal planes of gibbsite, all OH
-
groups are coordinated to two Al
3+
ions i.e.
OH Al
2
. Both singly coordinated hydroxyl groups ( AlOH ) and doubly coordinated
hydroxyl groups are found in equal amounts on the gibbsite edges.
The singly coordinated hydroxyl groups are considered to be the most reactive and
they occur in pairs on the edge of an
6
AlO octahedron (Goldberg et al., 1996).

2.6.2 Goethite

The structure of goethite is based on the hexagonal close packing of oxygen atoms
with 6-fold coordinated Fe atoms occupying octahedral position (Frost et al., 2003).
Each oxygen atom or hydroxyl group coordinates with three Fe
3+
ions. The Fe atoms
are arranged in a double row to form what can be described as double chains of
octahedra, which run the length of the c- axis (Figure 9).



b c
O
Fe
H
a
Figure 9 Ball and stick model for goethite (Sourced from Frost et
al., 2003).
23

Within the double chains in the b-c plane, all bonds are covalent with each octahedron
sharing four of its edges with neighbouring octahedra. In contrast, bonding between
double chains consists of relatively weak hydrogen bonding directed through apical
oxygen ions directed along the a- axis. In this case, stacking of double chains along
the a- axis can be easily disrupted and this consequently induces structural defects,
such as non-stoichiometric hydroxyl units incorporated into the goethite structure
during crystal growth (Frost et al., 2003).

In the bulk of the goethite mineral two types of triply coordinated oxygen groups are
found, one protonated ( OH Fe
3
) and the other non protonated ( O Fe
3
). Different
types of surface groups are present at the crystal faces that are singly-( (OH) FeOH ),
doubly- ( OH Fe
2
), and triply-coordinated ( O(H) Fe
3
) hydroxyl groups
(Stachowicz et al., 2008). The primary charging behaviour and adsorption reactions
between goethite and ions in solution is attributed to the singly and triply coordinated
surface oxygens located on the dominant 110 crystal face (Hiemstra and van
Riemsdijk, 1996).

2.6.3 Kaolinite

Clays are finely divided aluminosilicates. The principal building elements of the clay
minerals are the two-dimensional arrays of silicon-oxygen tetrahedral (tetrahedral
silica sheet) and that of aluminium- or magnesium-oxygen-hydroxyl octahedral
(octahedral, alumina or magnesia sheet) (Grim, 1995). The tetrahedral layer is
constituted by the coordination of several silica tetrahedrons. In a silica tetrahedron a
silicon atom is at the centre of a structure where the corners are filled with oxygen
24

atoms (Figure 10). Each tetrahedron is formed by one atom of Si
4+
and 4 atoms of O
2-
;
so that its chemical formula is
4
4
SiO : the arrangement in a sheet leads to a general
averaged formula of the type (Si
4
O
10
)
4-
n


Figure 10 Silica tetrahedron and layer (Grim, 1995).

The coordination of aluminium, magnesium, or iron atoms with oxydrils or hydroxyls
gives rise to structural units with an octahedral shape. Octahedrons also have the
capacity to join in sheets (Figure 11).



Figure 11 Octahedron and octahedral sheet (Grim, 1995).

The central cations in the octahedrons have to balance the negative electrical charge
of the octahedral arrangement of oxydrils and hydroxyls, equal to two electrons per
unit. Therefore while the magnesium ion needs to be present in every unit, the
25

aluminium ion will be required just in two cells over three. Sharing of oxygen atoms
between silica and alumina sheets results in two- or three-layer minerals such as 1:1
kaolinite built up from one silica and one alumina sheet (TO), or 2:1 type
montmorillonite, in which an octahedral sheet shares oxygen atoms with two silica
sheets (TOT) (Van Olphen, 1963; Schulze, 2002).

Kaolinite, [ ]
8 10 4 4
) OH ( O ) Al ( Si , is a dioctahedral 1:1 layer aluminosilicate. A kaolinite
unit cell consist of a layer of silica tetrahedral bound to an octahedral alumina layer
whose structure is very similar to that of gibbsite except that some hydroxyls are
replaced by oxygens (White, 2007). Kaolinite particles are formed by the repetition of
the basic layer by overposition, where bonding between subsequent layers is provided
by hydrogen bonds and van der Waals forces (Grim, 1995).




The kaolinite surface consists of three morphologically different planes with different
chemical compositions: a gibbsite type basal plane, a silica type basal plane and edge
G
G
G
7.2
Silica tetrahedral
sheet
Alumina octahedral
sheet
Figure 12 Kaolinite structure
26

planes represented by a complex oxide of the two constituents
3
) OH ( Al and
2
SiO
(Rosenqvist, 2002).

Surface functional groups on kaolinite are located on the octahedral and tetrahedral
basal planes as well as along the edge of the sheets. These functional groups include
doubly coordinated hydroxyl groups ( OH Al
2
) on the octahedral basal plane, the
siloxane group ( O Si
2
) on the tetrahedral basal plane as well as the
aluminol, OH Al , silanol ( OH Si ) and the
2
Al O Si groups located along
the edge of the sheets (Rosenqvist, 2002). The siloxane group , O Si
2
, which is the
only group present at the basal surface of a 2:1 clay or at one basal surface of
kaolinite is unreactive (Avena et al., 2003). For the
2
Al O Si sites, the charge
on the oxygen is fully neutralized and the group is therefore probably also not reactive
(Rosenqvist, 2002). The silanol and aluminol functional groups are therefore
considered reactive and hence contribute towards acid-base behaviour of kaolinte and
complexation reactions with solution speciation.

2.7 Complexation and adsorption

A complex is a unit in which an ion, atom, or molecule binds to other ions, atoms, or
molecules. The binding species is termed a central group and a bound species is
termed a ligand (Goldberg et al., 1996). Adsorption is described in terms of a set of
complex formation reactions between dissolved solutes and surface functional groups.
Ligands can be associated with the surface in different ways. In the formation of
inner-sphere complexes, a chemical (largely covalent) bond between the central atom
27

and the ligand is formed (White, 2007). In outer-sphere complexes on the other hand,
one or more water molecules remain between the ligand and the central atom and no
direct bond is formed. Outer-sphere complexes are held together mainly by
electrostatic forces. Inner-sphere complexes can be classified by the ligands mode of
binding to the surface. If the ligand is attached to only one surface functional group,
the complex is termed monodentate, whereas a ligand connected to two surface
functional groups forms a bidentate complex. Bidentate complexes can in turn be
further classified by considering the number of central atoms (in the solid material)
included in the complex. If a bidentate complex involves two central atoms, the
complex is generally referred to as a bridging complex, whereas a bidentate complex
involving only one central atom is referred to as a mononuclear chelate (Goldberg et
al., 1996).

2.7.1 Development of reactive functional groups at the metal oxide-solution
interface

Oxygen and metal atoms at an oxide surface are incompletely coordinated; i.e., they
are not surrounded by oppositely charged ions as they would be in the interior of a
crystal (White, 2007). Consequently, mineral surfaces immersed in water attract and
bind water molecules that can dissociate leaving a hydroxyl group bound to the
surface metal ion as indicated by Equation 2:

+ +
+ + H MOH O H M
2
(2)
where M denotes a surface metal ion. In a similar fashion, incompletely coordinated
oxygens at the surface can also bind water molecules, which can then dissociate,
again creating a surface hydroxyl group (Equation 3):
28


+ + OH OH O H O
2
(3)

Thus the surface on an oxide immersed in water very quickly becomes covered with
hydroxyl groups, which are considered to constitute part of the surface rather than the
solution. Different surface hydroxyl groups can be identified based on the number of
metal atoms to which they are coordinated. OH groups coordinated to only one metal
atom are called singly coordinated or terminal hydroxyls, whereas OH groups
coordinated to more than one metal are called bridging hydroxyls (Rosenqvist, 2002).
Bridging hydroxyls might be coordinated to two, three, or four metal atoms and are
therefore called doubly, triply and quadrupely coordinated, respectively.

2.7.2 Adsorption of ions at the metal (hydr)oxide-solution interface : The
electrostatic double layer model

As noted previously, oxygen atoms on the surface of a metal (hydr)oxide are
neutralized by both metal ions belonging to the solid and a variable number of
adsorbed protons. Depending on the solution pH, an excess or deficiency of protons
can occur at the interface resulting in a positively or negatively charged surface
respectively (Hiemstra and Riemsdijk, 1999). This surface charge,
0
, is
compensated by electrolyte ions in a double layer, normally assumed to be a diffuse
double layer (DDL) (Hiemstra and Riemsdijk, 1996). The ions present in the DDL are
hydrated and have a finite size, thereby preventing charge neutralization starting
directly from the close-packed surface. This result in counter and co-ions having a
minimum distance of approach to the surface and hence formulation of a charge free
layer called the Stern layer. This double layer picture has been described as the basic
Stern (BS) model (Hiemstra and Riemsdijk, 1996). Hiemstra and Riemsdijk (1996)
29

argued that outer sphere complexes have a minimum distance of approach to the
surface as counter and co-ions. This implies that outer sphere complexes are adsorbed
in an electrostatic plane positioned on the solution side of the Stern layer near the
head end of the diffuse double layer (Figure 13). The innersphere complexes on the
other hand, are closer to the surface, penetrating the stern layer.




In the presence of specific adsorption, a hypothetical electrostatic plane (1-plane)
emerges in the Stern layer in which solution oriented ligands of the specifically
adsorbed complexes are located. This double layer picture, consisting of three
electrostatic planes (0-, 1-, and 2-or d-plane) is called the three plane (TP) model
(Hiemstra and Riemsdijk, 1996). In the absence of specifically adsorbing ions, the TP
model simplifies to the BS model.


Figure 13 The electrostatic double layer model (Sourced
from Hiemstra and Riemsdijk, 2006)
30

2.8 Protonation of surface functional groups and charge balance

Adsorption reactions on metal (hydr)oxide surfaces are pH dependent (Manning and
Goldberg, 1996) as such an understanding of acid-base reactions on the metal
(hydr)oxide surfaces is essential for the description of the effect of pH on the amount
of ions adsorbed. It is assumed that protonating oxygen atoms on an oxide surface can
bind two protons in a pH range. In such approach, surface groups are considered to be
diprotic reacting according to the following scheme presented by Equations 4 and 5:

SOH H SO +
+
(4)

+ +
+
2
SOH H SOH (5)

where

SO , SOH and
+

2
SOH are deprotonated, monoprotonated and diprotonated
surface groups respectively (Kraepiel et al., 1998; Avena et al., 2003). To describe
these two consecutive protonation/deprotonation steps, two pKa values are required
and this conceptual model is therefore known as the two pKa model. Recent
theoretical and experimental evidences however, indicate that two protonation steps
as indicated by Equations 4 and 5 seldom occur at oxygen atoms in aqueous media
either at the surface of solids or in true solutions (Borkovec et al., 2001). In most Al
and Fe containing hydroxides in nature, the metal (M
3+
) ions are most often
octahedrally coordinated to six oxygen atoms. This means that each oxygen atom will
neutralize one sixth of the charge on the metal resulting in 0.5 charge units. If the
oxygen atoms are coordinated to only one metal ion, the half unit charge from the
metal means that the OH group cannot be neutral; it will have either +0.5
31

) MOH (
. +

5 0
2
or 0.5 (

5 . 0
MOH ) charge. Any other protonation steps are unlikely to
occur within the normal pH range, and therefore the protonation of the surface can be
described using a single protonation step and one pKa value. For an oxygen atom that
is singly coordinated to the metal ion (Me), the protonation step can be written as
+ +
+
1/2
2 3
- 1/2
OH Me H MeOH whereas for a triply coordinated surface oxygen one
finds
+ +
+
1/2
3
- 1/2
3
OH Me H O Me (Hiemstra and Riemsdijk, 1999). Generally, a
protonation reaction in the one-pKa approach is given by (Equation 6):

1 x x
AH H A
+ +
= + (6)

where
x
A denotes a functional surface group carrying a charge x (fractional or
integer, negative or positive) and
1 x
AH
+
is the protonated group.

The balance of surface charge on an aluminium oxide mineral in aqueous solution is
given by Equation 7,

0
d os is H
= + + +
(7)


where
H
is the net proton charge, defined by
H
= ) ( F
OH H
, where is a
surface excess concentration;
is
is the inner-sphere complex charge resulting from
the formation of inner-sphere complexes between adsorbing ions (other than H
+
and
OH
-
) and surface aluminol groups;
os
is the outer-sphere complex charge resulting
from the formation of outer-sphere complexes between adsorbing ions and surface
aluminol groups or ions in inner-sphere complexes;
d
is the dissociated charge,
32

equal to minus the surface charge neutralized by electrolyte ions in solution that have
not formed adsorbed complexes with surface aluminol groups (Goldberg et al., 1996).
Consideration of the surface charge balance as a function of pH leads to the definition
of the point of zero charge. The point of zero charge, p.z.c, is the solution pH value
where total net particle charge is zero: 0
d os is H
= = + + . The adsorbent surface
develops a net positive charge below the p.z.c., and a net negative charge above it.
Adsorption of anions is generally higher below the p.z.c. whilst that of cations is
higher above the p.z.c. Determination of the p.z.c. is therefore necessary in
discussions of ion adsorption on metal hydroxide surfaces. The . c . z . p can be
measured directly using electrokinetic measurements or colloidal stability
experiments (Goldberg et al., 1996). When the p.z.c. is measured using electrokinetics
it is often called the isoelectric point (i.e.p.). The point of zero net proton charge,
p.z.n.p.c, is the solution pH value where the net proton charge is zero. The p.z.n.p.c.
can be measured using a potentiometric titration if only selective aluminol groups are
titrated (Goldberg et al., 1996). The point of zero salt effect (p.z.s.e.) is the solution
pH value where the net proton charge is independent of solution ionic strength,
0
I
H
=

. The p.z.s.e. can also be measured by potentiometric titration using either

batch or continuous titrations.

The p.z.n.p.c. is determined as the point where a plot of the apparent proton surface
charge density,
titr , H
, against suspension pH crosses the x-axis, i.e. pH value where
titr , H
= 0. Negative and positive proton surface charge density indicates net coverage
of the surface by hydroxyl groups and protons, respectively. Schroth and Sposito
(1997) described equations used to calculate the apparent proton surface charge
33

density for kaolinite that corrects for proton consumption by aluminium species in
solution, introduced through dissolution of aluminium at low pH values (generally
below pH 4).
The apparent proton surface charge density,
titr , H
, is given by Equation 8:

|
|

\
|
=
+ +
+ +
S
] [H
K
] [H
K
) ] H [ ] H ([ M
w
b
w
S b Soln titr H,
(8)

where M
soln
is the mass of electrolyte solution (per unit dry mass) equilibrated with
kaolinite, [H
+
] is the solution proton concentration, K
w
is the dissociation product of
water, and the subscripts s and b refer to sample and blank solutions, respectively. For
highly acidic samples in which there is significant proton release caused by
dissolution of kaolinite, the apparent surface charge density is corrected for Al in
solution by Equation 9:

[ ] [ ] ( ) [ ] { }
+ + +
+ + =
2
2 3
ln so titr , H Al , titr , H
OH Al AlOH 2 Al 3 M (9)

where the concentrations of the 3 Al species are calculated using the total Al
concentration in solution. This correction accounts for protons that would be
consumed in the release of Al and those that would be generated by Al hydrolysis at
pH 6. (Schroth and Sposito, 1997).



34

2.9 Patch-wise surface charge heterogeneity on kaolinite and the point of zero
charge

Clay mineral particles hold both permanent negative charges on the faces and pH
dependent either negative or positive charges developing mainly on OH Al active
sites at the broken edges and exposed hydroxyl-terminated planes (Tombcz and
Szekeres, 2006). The permanent negative charge sites on the basal planes (faces) are a
result of substitution of the Si- and Al-ions in the crystal lattice for lower positive
valence ions. Since these two types of sites are situated on the given parts of the
particle surface, different charge patches exist on the basal planes and edges of clay
particles (Koopal, 1996). Development of the different surface charge patches on clay
particles affect reactivity of the functional groups on the basal planes and edges
towards adsorption of various ions as well as colloidal behaviour of the particles
(Tombcz and Szekeres, 2006). The size of the surface charge patches and the lateral
interactions of the surface sites are affected simultaneously by suspension pH and
ionic strength of the background electrolyte. Determination of the p.z.c. for clay
particles is with respect to pH-dependent functional groups located on the edges of the
particles as such the point of zero charge is referred to as p.z.c.
edge
.

2.10 Modeling phosphate adsorption on kaolinite and bauxite

Phosphate adsorption on kaolinite and bauxite will be modelled using the Freundlich
isotherm equation (Equation 10). The Freundlich equation is a semi-empirical model
used to describe heterogeneous systems (Milonjic, 2007):

n
1
e f e
c K q =
(10)

35


Where K
f
is the Freundlich constant (dm
3
g
-1
), q
e
is phosphate uptake (mg/g), C
e
is the
equilibrium phosphate concentration (mg/L) and 1/n is the heterogeneity factor. The
Freundlich equation is consistent with an exponential distribution of electrical
potentials or of binding constants on metal hydroxide surfaces (Barrow et al., 2005).
The empirical form of the Freundlich isotherm equation is applicable to both
monolayer adsorption (chemisorption) and multilayer adsorption (van der Waals
adsorption) (Yang, 1998). It is always inappropriate to use the Langmuir equation as a
simple equation to describe sorption by soil because soils do not comprise one
uniform surface; adsorption always induces a change in the properties of the surface
and this is inconsistent with this equation; and the equation usually describes sorption
poorly (Mead, 1981; Barrow, 2000; Barrow et al., 2005; Barrow, 2008). The kaolinite
and bauxite samples used in this study are heterogeneous materials just like soil hence
the use of only the Freundlich equation is justified.

2.11 Precipitation of calcium phosphates


Kaolinite samples used in this study had a significant amount of calcium impurities
(1.1% CaO), giving the possibility of precipitation of calcium phosphates during
treatment if high dosages are used. Under proper physical and chemical environment,
different kinds of calcium phosphates, such as O H CaHPO
2 4
2 . (dicalcium
phosphate dihydrate, DCPD), O H PO H Ca
2 3 4 4
5 . 2 . ) ( (octacalcium phosphate, OCP),
2 4 3
) (PO Ca (tricalcium phosphate) and OH PO Ca
3 4 5
) ( (hydroxyapatite, HAP) may
precipitate from saturated solutions, among which HAP is thermodynamically the
most stable one (Koutsoukos et al., 1980; Van Kemenade and De Bruyn, 1987). The
36

thermodynamic driving force to a chemical reaction is the Gibbs free energy G, and
it is the criterion to judge whether a reaction is spontaneous, in equilibrium, or
impossible, corresponding to G < 0, = 0, or > 0, respectively. Considering a calcium
phosphate precipitation reaction, the Gibbs free energy is given by Equation 11:

sp
K
IAP
ln
n
RT
- G =

(11)

where R is the ideal gas constant (8.314 JK
-1
mol
-1
), T is the absolute temperature, IAP
and K
sp
are respectively the free ionic activities product and the thermodynamic
solubility product of the precipitate phase, and n is the number of ions in the
precipitated compound (Song et al., 2002a). Supersaturation is a measure of the
deviation of a dissolved salt from its equilibrium value, for a solution departing from
equilibrium is bound to return to this state by the precipitation of excess solute. The
saturation index, SI, of a solution with respect to a precipitate phase provides a good
measurement of supersaturation of a system and is defined by Equation 12:

Ksp
IAP
log SI = (12)

Gibbs free energy is therefore related to SI by Equation 13:

SI
n
2.303RT
- G= (13)

37

When SI = 0, hence G = 0, the solution is in equilibrium; when SI < 0, G > 0, the
solution is undersaturated and precipitation is impossible; when SI > 0, G < 0, the
solution is supersaturated and precipitation is spontaneous. SI is a good indicator to
show the deviation of a salt from its equilibrium state, i.e. the thermodynamic driving
force for the precipitation of a calcium phosphate phase. But considering precipitation
kinetics, supersaturation does not certainly mean the quick occurrence of a
spontaneous precipitation. Between the undersaturated zone and spontaneous
precipitation zone there is still a metastable zone, where the solution is already
supersaturated but no precipitation occurs over a relatively long period (White, 2007).
The boundary between metastable zone and spontaneous precipitation zone is called
the critical supersaturation (Joko, 1984).

38

CHAPTER THREE: MATERIALS AND METHODS

3.1 Materials

3.1.1 Adsorbents

Kaolinite and bauxite samples used in this study were collected from kaolin deposits
located at Linthipe in Dedza district, and from Mulanje Mountain, Malawi,
respectively. Both the kaolinite and bauxite samples were identified by the Geological
Survey Department of Malawi (GSoM).

3.1.2 Chemicals, reagents and instruments

The following analytical grade chemicals and instruments were used: anhydrous
potassium dihydrogen phosphate, KH
2
PO
4,
(Glassworld, SA); Lanthanum oxide,
La
2
O
3
, (SAARCHEM, SA); Sodium carbonate, Na
2
CO
3,
(Glassworld, SA); Sodium
hydrogen carbonate, NaHCO
3,
(BDH); Sodium Fluoride, NaF, and (SAARCHEM,
SA); anhydrous Sodium sulphate, Na
2
SO
4
, (ACE, SA); Calcium nitrate, Ca(NO
3
)
2,

(SAARCHEM, SA); and Magnesium nitrate, Mg(NO
3
)
2,
(BDH); Sodium nitrate,
NaNO
3,
(BDH); Sodium hydroxide, NaOH, and nitric acid, HNO
3,
(Associated
Chemical Enterprises (Pty) Ltd, RSA); Hydrochloric acid, HCl, (Glassworld, SA), pH
7.00 0.02 and pH 4.01 0.02 buffer solutions (Mettler-Toledo Inc., Switzerland);
Hettich Rotixa/AP centrifuge (Hettich lab technology, Germany); Gallenkamp
Thermostirrer 85 water bath (Weiss Gallenkamp, UK); Stuart mini orbital shaker
SSM1(Bibby Scientific Ltd, UK).
39

3.2 Methods

3.2.1 Preparation of kaolinite and bauxite samples

Prior to use, kaolinite samples were sun-dried for 48 hours and ground by a traditional
pestle and mortar followed by sieving through a 60-mesh size sieve. Samples
prepared in this way are referred to as raw kaolinite. Acid treatment of the kaolinite
was achieved by addition of 100 mL hydrochloric acid (0.3 mol/L) to a 200 g sample,
ensuring thorough mixing in the process, followed by drying of the wet sample in a
sample-drying oven at 50 C for 7 hours. Drying was followed by crushing of the clay
blocks, and sieving through a 60-mesh size sieve. Kaolinite samples prepared as such
are referred to as treated kaolinite.

Preparation of bauxite samples involved sun-drying for 48 hours, grinding using a
traditional mortar and pestle, and sieving through a 60-mesh size sieve. Raw bauxite
samples were used throughout the study.

3.2.2 Preparation of solutions
3.2.2.1 Reagents
3.2.2.1.1 Sodium Carbonate (0.01 mol/L and 1000 mg/L)

The 0.01 mol/L and 1000 mg/L solutions were prepared by dissolving 1.0820 g and
1.7662 g of Na
2
CO
3
(dried at 250C for 4 hours) in distilled water and diluting to
1000 mL in a volumetric flask respectively.

40

3.2.2.1.2 Nitric acid (0.02359 mol/L and 1+1)

This solution was prepared by diluting 1.7 mL of 55% nitric acid (of density 1.34
g/mL) with distilled water to 1000 mL in a volumetric flask. The dilute nitric acid was
standardized with 0.01 mol/L sodium carbonate (Na
2
CO
3
) through an acid-base
titration, using methyl orange indicator to determine the end point.
The 1+1 nitric acid was prepared by mixing equal amounts of concentrated nitric acid
(55 %) and distilled water.
3.2.2.1.3 Lanthanum solution

The lanthanum solution was prepared by dissolving 58.65g of lanthanum oxide,
La
2
O
3
, in 250 mL concentrated HCl followed by dilution to 1000 mL with distilled
water.
3.2.2.1.4 Sodium hydroxide (0.020 mol/L)

This solution was prepared by dissolving 0.8000g of sodium hydroxide pellets in
deionised water. The solution was poured into a 1 L volumetric flask and diluted to 1
L with distilled water.
3.2.2.1.5 Hydrochloric acid (0.3 mol/L)

The dilute hydrochloric acid solution was prepared by diluting 29.5 mL concentrated
HCl (of density 1.16g/mL) to 1000 mL with distilled water in a volumetric flask.



41

3.2.2.1.6 Sodium nitrate (1.0 mol/L)

This electrolyte solution was prepared by dissolving 42.4985g of NaNO
3
(dried in an
oven at 105 C for 12 hours) and making to the mark in a 500 mL volumetric flask
with distilled water. Working electrolyte concentrations were prepared by diluting
calculated volumes of the 1.0 mol/L sodium nitrate solution.
3.2.2.1.7 Sulphate solution (1000 mg/L)

This solution was prepared by dissolving 1.4790g of Na
2
SO
4
(dried in an oven at 105
C for 6 hours) and making to the mark in a 1000 mL volumetric flask with distilled
water.
3.2.2.1.8 Magnesium solution (1000 mg/L)

This solution was prepared by dissolving 6.1024g of Mg(NO
3
)
2
(dried in an oven at
106 C for 6 hours) and making to the mark in a 1000 mL volumetric flask with
distilled water.
3.2.2.1.9 Calcium solution (1000 mg/L)

The calcium solution was prepared by dissolving 4.0943g of Ca(NO
3
)
2
(dried in an
oven at 106 C for 6 hours) and making to the mark in a 1000 mL volumetric flask
with distilled water.



42

3.2.2.2 Standard solutions
3.2.2.2.1 Standard phosphate solution

A standard phosphate stock solution (1000 mg/L) was prepared by dissolving 1.4330g
of analytical grade anhydrous KH
2
PO
4
(dried for 1 hour in an oven at 105C) in
distilled water and making to the mark in a 1000- mL volumetric flask. Intermediate
standard solutions (100 mg/L) were prepared by diluting 25 mL of the stock solution
in a 250 mL volumetric flask. The intermediate standard solutions were used to
obtain working phosphate concentrations.
3.2.2.2.2 Standard calcium solution (for AAS determination of calcium).

A standard calcium solution (100 mg/L) was prepared by suspending 0.2497g of
CaCO
3
(dried at 180C for 1 hour) in water followed by addition of 20 mL of 1+1
HNO
3
to dissolve. 10 mL of concentrated HNO
3
was added to the solution which was
later diluted to 1000 mL with distilled water.

3.2.3 Determination of phosphate ions in solution using ion chromatography

Phosphate ions in solution were determined using an ion chromatography technique.
In principle, a small volume of aqueous sample is injected into an ion chromatograph
to flush and fill a constant-volume sample loop. The sample is then injected into a
flowing stream of carbonate-bicarbonate mobile phase. The sample is pumped
through two different ion exchange columns, then a conductivity suppressor device,
and into a conductivity detector. The two ion exchange columns, a precolumn or
guard column and a separator column, are packed with an anion exchange resin. Ions
are separated into discrete bands based on their affinity for the exchange sites of the
43

resin inside the guard and analytical columns. The guard column extends the life of
the analytical column by trapping organic compounds and other species that could
destroy the analytical column. The guard column also adds about 20 % to the total
separation capacity of the analytical system. The conductivity suppressor is an ion
exchange-based device that reduces the background conductivity of the mobile phase
to a low or negligible level and simultaneously converts the anions in the sample to
their more conductive acid forms. The separated anions in their acid forms are
measured using an electrical conductivity cell. Anion identification is based on the
comparison of analyte signal peak retention times relative to those of known
standards. Quantitation is accomplished by measuring the peak area and comparing it
to a calibration curve generated from known standards. The ion chromatography
system that was used was composed of a Dionex CDM-1 conductivity detector, an
Ionpac AS14 anion exchange column and Data Apex Clarity chromatography
software. A mobile phase of 3.5 mM Na
2
CO
3
/ 1.0 mM NaHCO
3
was used during the
analysis.

3.2.4 Determination of Calcium ions in solution

The concentration of calcium ions in solution was determined using a Buck Scientific
atomic absorption spectrometer (Buck Scientific, USA) at a wavelength of 422.7 nm
and a slit of 0.7 nm with an air-acetylene flame as described by APHA (1989).
Known standards (from 0-10 mg/L) were used to calibrate the instrument.




44

3.2.5 Potentiometric titration of raw kaolinite samples

Raw kaolinite samples (0.1 kg) were saturated with sodium by repeatedly suspending
them in 1 mol/L NaNO
3
/HNO
3
solution at pH 3. After centrifugation of the
suspension, the supernatant was discarded and replaced with fresh solution. This was
done up until when the supernatant solution pH had dropped to pH 3. This was
followed by repeated suspension of the samples in non-acidified NaNO
3
and
centrifugation up until when the supernatant pH value approached 5. At this point, the
samples were repeatedly suspended in 0.1 mol/L NaNO
3
and centrifuged until a stable
pH was registered. Na-saturated kaolinite stock suspensions were stored at pH 5.2
5.8 in sealed centrifuge bottles. Pre-treatment of the raw kaolinite samples assisted in
removing calcium carbonates that could have interfered during the titration, besides
saturating the clay with sodium.

Suspensions of 6.0 g sample/L were used for the
potentiometric titrations. The raw kaolinite suspension (25 mL) was titrated with
0.02359 mol/L nitric acid and 0.020 mol/L NaOH

stepwise using a Metrohm 775
Dosimat titrator equipped with a Metrohm 806 exchange unit (Metrohm,
Switzerland). The pH reading after equilibration of each step (30 minutes) was
recorded using Ross Sure Flow combination electrode Orion 8172 (Thermo Fisher
Scientific, USA) suitable for pH determination in suspensions connected a Mettler-
Toledo SevenMulti meter (Mettler-Toledo Inc., Switzerland). During the titration the
flask containing the suspension was covered with a parafilm to minimize formation of
carbonates at pH greater than 6. A blank titration (excluding the clay) was run
following the same procedure. The net proton surface charge density was calculated
using Equation 8 and later used to obtain the point of zero net proton charge.



45

3.2.6 Effect of suspension pH on the amount of phosphate ions removed by
bauxite and kaolinite

Bauxite (0.05 g) or kaolinite (0.40 g) samples were weighed in triplicate into acid
washed 50 mL centrifuge tubes. An appropriate amount of nitric acid (0.02 mol/L) or
sodium hydroxide (0.02 mol/L) was added to the clay suspension to adjust pH to a
certain value (pH cannot be predetermined for bauxite, kaolinite and generally for soil
samples). After addition of either nitric acid or sodium hydroxide, the suspensions
were mixed using a vortex shaker and left to stand for 20 minutes to allow for
reactions between constituents of the samples and the hydrogen or hydroxide ions.
Sodium nitrate solution (0.1 mL), intermediate standard phosphate solution (1.0 mL),
and a calculated amount of distilled water was added to each centrifuge tube to ensure
a total volume of 10 mL. The added volumes of sodium nitrate and the phosphate
solution ensured concentrations of 0.01 mol/L and 10 mg/L for sodium nitrate and
phosphate ions respectively. Suspensions prepared in this manner were shaken for 90
minutes and later centrifuged at 2500 rpm for 20 minutes. The supernatant was
filtered through a 0.45m membrane and analyzed for phosphate using ion
chromatography.

3.2.7 Determination of calcium ions released into solution from kaolinite

Raw or treated kaolinite (0.8 g) samples were weighed in triplicates into 50 mL
centrifuge tubes followed by addition of 10 mL distilled water. The pH of the
suspensions was adjusted by adding appropriate amounts of nitric acid (0.02 mol/L)
or sodium hydroxide (0.02 mol/L) followed by addition of calculated volumes of
distilled water to obtain total volumes of 20 mL. The suspensions were shaken for 90
46

minutes followed by centrifuging at 2500 rpm to obtain the supernatants. The
supernatants were further filtered through 0.45 m membrane filters to remove
particulate matter followed by determination of calcium concentration in the filtrates.

3.2.8 Effect of bauxite and kaolinite dosage on amount of phosphate ions
removed


Masses of bauxite or kaolinite were weighed in triplicates into acid washed
polystyrene centrifuge tubes. 0.01, 0.03, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, and 0.8 g
masses were weighed. A phosphate solution (10 mg/L) in 0.01 mol/L NaNO
3

electrolyte was prepared by diluting a mixture of 50 mL intermediate phosphate
solution (100 mg/L) and 5 mL sodium nitrate solution (1.0 mol/L) to 500 mL with
distilled water. 10 mL of the phosphate/sodium nitrate solution was pipetted into each
of the centrifuge tubes containing bauxite or clay and shook for 90 minutes. After
shaking, the mixture was centrifuged for 20 minutes at 2500 rpm followed by
filtration of the supernatant through a 0.45m PES syringe filter membrane to remove
some particles. The filtered samples were later analysed for phosphate via ion
chromatography. The percentage of phosphate ions precipitated or bound was
calculated as,
100% x
C
C C
removal Phoshate %
o
e o
|
|

\
|
=
(14)

where C
o
and C
e
are the initial and equilibrium phosphate concentrations respectively,
given in units of mg/L.


47

3.2.9 Effect of contact time on the amount of phosphate ions removed by bauxite
and kaolinite


Bauxite or kaolinite samples were weighed into acid washed glass 250 mL BOD
bottles. Phosphate solution (10 mg/L) in 0.01 mol/L NaNO
3
electrolyte was prepared
as described under effect of dosage. 100 mL samples of the phosphate/sodium
nitrate solution were equilibrated to 20 and 40 C in water baths with temperature
controls. This was followed by mixing of the equilibrated phosphate solutions (100
mL) with the weighed bauxite or kaolinite samples (to obtain dosages of 5 g/L and 40
g/L respectively), and continued shaking at 20 and 40 C in the water baths. 0.5 mL of
the mixture was withdrawn at 5 minute intervals for the first 10 minutes followed by
withdraws at 10 minute intervals for the next 50 minutes. After 60 minutes of contact,
withdraws were made at 20 minute intervals for the first 60 minutes followed by 30
minute interval withdraws for the last 180 minutes. This provided a total contact time
of 300 minutes (5 hours). The mixture withdrawn after each interval (0.5 mL) was
passed through a 0.45 m membrane and injected into an ion chromatograph for
phosphate analysis within 1 minute. Phosphate uptake (mg/g) was calculated as,
(L) volume Sample x
(g) Dosage
C C
(mg/g) uptake Phosphate
e o
|
|

\
|
=
(15)

where C
o
and C
e
are the initial and equilibrium phosphate concentrations respectively,
given in units of mg/L.


48

3.2.10 Effect of initial phosphate concentration on the phosphate removal
capacity of bauxite and kaolinite

Bauxite (0.05 g) or kaolinite (0.4 g) samples were weighed in triplicate into acid
washed centrifuge tubes. Intermediate phosphate solution (100 mg/L), sodium nitrate
(1.0 mol/L) and distilled water were mixed in proportions shown in Table 2 to obtain
10 mL total volume, 0.01 mol/dm
3
sodium nitrate solution and initial phosphate
concentrations indicated in Table 2.

Table 2 Volumes of phosphate and sodium nitrate solutions mixed and later
made up to 10 mL with distilled water

Initial phosphate
concentration
(mg/L)

5

10

15

20

30

40

50

60
100 mg/L
phosphate (mL)
0.5 1.0 1.5 2.0 3.0 4.0 5.0 6.0
1.0 mol/L sodium
nitrate (mL)
0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1

The mixtures prepared in this manner were shaken for 90 minutes and later
centrifuged at 2500 rpm for 20 minutes. The supernatant was filtered through a
0.45m membrane and analyzed for phosphate using ion chromatography. In the case
of bauxite, the sodium nitrate/distilled water/bauxite mixtures corresponding to the
whole phosphate concentration range (Table 2) were separately equilibrated to 10, 20,
30, 40, 50 and 60 C in water baths before introduction of phosphate. This means that
the effect of initial phosphate concentration on phosphate removal by bauxite was
studied at 6 different temperatures.


49

3.2.11 Effect of magnesium, calcium, sulphate and carbonate/bicarbonate ions
on amount of phosphate ions removed by bauxite and kaolinite

Bauxite (0.05 g) or kaolinite (0.4 g) samples were weighed in triplicate into acid
washed 50 mL centrifuge tubes. A combination of solutions was prepared as shown in
Table 3, to obtained 10 mg/L initial phosphate concentration, 0.01 mol/L sodium
nitrate electrolyte and competing ion (Mg
2+
, Ca
2+
, SO
4
2-
or CO
3
2-
/HCO
3
-
)
concentrations shown in the table.

Table 3 Solution combinations for the effect of competing ions that were later
made up to 10 mL with distilled water

Competing
ion (mg/L)
0 10 20 40 60 80 100 200 300 500
1000 mg/L
competing
ion (mL)
0 0.1 0.2 0.4 0.6 0.8 1.0 2.0 3.0 5.0
1.0 mol/L
sodium
nitrate (mL)
0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1
100 mg/L
phosphate
(mL)
1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0

Phosphate was introduced last during the preparation of the mixtures to prevent
reactions between phosphate and bauxite or kaolinite in the absence of the competing
ion. The mixtures prepared in this manner were shaken for 90 minutes and later
centrifuged at 2500 rpm for 20 minutes. The supernatant was filtered through a
0.45m membrane and analyzed for phosphate using ion chromatography




50

3.2.12 Multi-interactive effect of magnesium, calcium, sulphate, and bicarbonate
ions on the phosphate removal efficiency of bauxite.

Triplicate bauxite samples were weighed into acid cleaned centrifuge tubes of masses
0.01, 0.03, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, and 0.8 g. To each tube, 2 mL of each
solution of 1000 mg/L sodium carbonate, calcium nitrate, sodium sulphate, and
magnesium nitrate were added followed by 1.4 mL of distilled water. Finally, 0.1 mL
of sodium nitrate (1.0 mol/L) and 0.5 mL of phosphate solution (1000 mg/L) was
added to each tube to obtain a total volume of 10 mL. The final composition of the
solution was 50 mg/L phosphate, 0.01 mol/dm
3
sodium nitrate electrolyte, and 200
mg/L of sodium carbonate, calcium nitrate, sodium sulphate, and magnesium nitrate.
Concentrations of 200 mg/L for sodium carbonate, calcium nitrate, sodium sulphate,
and magnesium nitrate, and 50 mg/L phosphate were considered as worst case
simulations of influent wastewaters for treatment plants in Blantyre. Another set of
triplicate bauxite samples was weighed into centrifuge tubes. To this set of bauxite
samples was added the phosphate solution and the sodium nitrate electrolyte only with
a final composition of 50 mg/L phosphate and 0.01 mol/dm
3
sodium nitrate
background electrolyte. The second set of bauxite-phosphate mixtures acted as a
control for the first set to which competing ions were added. Suspensions prepared in
this manner were shaken for 90 minutes and later centrifuged at 2500 rpm for 20
minutes. The supernatant was filtered through a 0.45m membrane and analyzed for
phosphate via ion chromatography.



51

CHAPTER FOUR: RESULTS AND DISCUSSION

4.1 Point of zero net proton charge for kaolinite

Potentiometric acid-base titration over the range of pH between 2 and 10 was used to
characterize the pHdependent charge development on the amphoteric surface sites of
the kaolinite. Acid-base properties of kaolinite are attributed to protonation and
deprotonation of the aluminol functional group ( OH Al ) on the edge;
deprotonation of the silanol group ( OH Si ) on the edge; and protonation and
deprotonation of doubly coordinated basal plane hydroxyl groups ( OH Al
2
) (Ganor
et al., 2003). The doubly coordinated basal plane hydroxyl group is less reactive than
the aluminol group on the edge as such much of acid-base and adsorption reactions on
kaolinite take place on the particle edge. The value of the point of zero net proton
charge (p.z.n.p.c.) for the Linthipe kaolinite was estimated to be 5.102 (Figure 14)
using 0.1 mol/L NaNO
3
background electrolyte.


52


-1.5
-1
-0.5
0
0.5
1
1.5
2 3 4 5 6 7 8 9 10
pH
N
e
t

p
r
o
t
o
n

s
u
r
f
a
c
e

c
h
a
r
g
e

d
e
n
s
i
t
y

(
m
m
o
l
/
K
g
)
Net proton surface charge
density

Figure 14 Experimental net proton surface density curve for Na-kaolinite in 0.1
mol/L NaNO
3
electrolyte

Reported p.z.n.p.c values for other kaolinite samples range from 3.0 to 7.5 (Wieland
and Stumm, 1992; Scroth and Sposito, 1997; Tombacz and Szekeres, 2006). Some of
the disagreements in the reported values can be attributed to differences in sample
preparation and to the use of different models to interpret data. The estimated p.z.n.p.c
value of 5.102 for the kaolinite is within the range of values reported in literature for
other kaolinite samples (Wieland and Stumm, 1992; Scroth and Sposito, 1997;
Tombacz and Szekeres, 2006). This implies that the edge surfaces of the kaolinite
particles attain a net positive charge above pH 5.102 and a net negative charge below
the same pH value. The development of pH-dependent charges would affect the
amount of phosphate ions adsorbed owing to electrostatic interactions between the
approaching ions and the active sites. Increased sorption of phosphate ions is expected
below pH 5.102 owing to formation of inner sphere complexes governed by attraction
of phosphate ions to the positively charged aluminol groups on the particle edges.
53

4.2 Effect of suspension pH on the amount of phosphate ions removed by
kaolinite and bauxite

Figure 15 indicates trends on the effect of initial pH of the clay suspension on the
amount of phosphate ions bound.


Figure 15 Plot of % phosphate removal against pH for raw and treated kaolinite

It is noted from the figure that there was a sharp increase in the amount of phosphate
ions bound with increasing pH up to around pH 3 for raw kaolinite and pH 5 for
treated kaolinite. This was followed by a sharp decrease to around pH 7 in both cases.
The amount of phosphate ions bound increased again from pH 7 levelling off at pH 11
and pH 9 for raw and treated kaolinite respectively. Two pH regions of high
phosphate uptake are noted from the trends, one below, and the other above pH 7.
Manning and Goldberg (1996) reported a different trend on variation of amount of
phosphate ions bound on kaolinite (KGa-1) with increasing pH. Phosphate adsorption
reached a maximum at pH 5 and declined abruptly at both higher and lower solution
pH forming an adsorption envelope. Earlier investigations also reported phosphate
adsorption envelopes on kaolinite of a parabolic shape with phosphate adsorption
54

maxima between pH 4 and 6 (Chen et al., 1973; Edzwald et al., 1976). Abrupt
changes in electrostatic attraction between anions and the kaolinite particle edge occur
in the region of the point of zero net proton charge, p.z.n.p.c.
edge,
which translates into
changes in amount of phosphate ions adsorbed (Manning and Goldberg, 1996). The
estimated p.z.n.p.c.
edge,
value of 5.01 for KGa-1 kaolinite (Schroth and Sposito, 1997),
supports the formation of adsorption envelopes with maxima between pH 4 and 6. As
noted earlier, surface oxygen atoms become protonated below the p.z.n.p.c.
edge

thereby inducing electrostatic attraction with the negatively charged phosphate
species. The increased electrostatic attraction coupled with development of elaborate
electric double layers along the particle edges promotes ion exchange reactions
between surface functional groups and anionic species in solution (Tombacz and
Szekeres, 2006). The decline in phosphate adsorption on kaolinite below pH 4 can be
attributed to dissolution of the kaolinite surface below pH 4.5, which releases soluble
Al
3+
into solution (Wieland and Stumm, 1992), subsequent formation of soluble
aluminium-phosphorus complexes; and the preferential adsorption of the phosphate
species, H
2
PO
4
-
, that dominates in solution between pH 2 and 7 (Figure 17).

The silanol ( SiOH ) and aluminol ( AlOH ) functional groups located on the edges
of kaolinite particles become deprotonated above the p.z.n.p.c.
edge,
resulting in
development of negative charges. Electrostatic repulsions are expected between the
developed negative charges and the negatively charged phosphate ions, resulting in
reduced adsorption. It is noted from the declining trends in Figure 15, starting from
pH 3 and 5 that an adsorption envelope was expected to appear with adsorption
maxima between pH 3 and 6. This expectation is supported by the estimated value of
the p.z.n.p.c.
edge,
for the locally sourced kaolinite of 5.102. The sharp increase in
amount of phosphate ions bound starting from pH 7 to pH 11 and hence the absence
55

of an adsorption envelope suggests that phosphate ions were removed by other species
other than clay particles through a different mechanism all together.

A different trend was obtained for the effect of initial pH of the bauxite suspension on
the amount of phosphate ions bound (Figure 16).


Figure 16 Plot of % phosphate bound against pH for bauxite

The amount of phosphate ions bound decreased gradually with increasing pH up to
around pH 9 and decreased sharply from pH 9 to around pH 11. High phosphate
removal was favoured in lower pH conditions for bauxite in contrast to kaolinite
where both low and high pH favoured high phosphate removal. It is noted from
Figure 16 that an abrupt decline in amount of phosphate ions adsorbed on bauxite
occurred between pH 9 and 10. This suggests that phosphate ions were bound by
mineral species having points of zero charge between pH 9 and 10. An argument is
furthered in the discussion that follows, that goethite and gibbsite were the principle
minerals responsible for binding phosphate ions in bauxite. It has been noted so far
56

that removal of phosphate ions by both kaolinite and bauxite was highly pH
dependent.

4.2.1 Description of reactions resulting in removal of phosphate ions by
kaolinite and bauxite

The kaolinite used in this study contained impurities such as iron oxides and calcium
carbonate (Table 1). Removal of phosphate ions in the low pH region can therefore be
attributed to adsorption on kaolinite particles and iron oxide surfaces. This suggestion
is supported by studies that reported adsorption of phosphate ions on kaolinite and
other clay particles (Goldberg et al., 1996; Manning and Goldberg, 1996) as well as
adsorption on iron oxides that include goethite, ferrihydrite, and hematite (Persson et
al., 1996; Pradhan et al., 1998; Hiemstra and Van Riemsdijk, 1999; Arai and Sparks,
2001; Horanyi and Joo, 2002; Altundogan and Tumen, 2003; Elzinga and Sparks,
2007; Kubicki et al., 2007; Rahnemaie et al., 2007; Zhong et al., 2007; Stachowicz et
al., 2008).

Phosphate ions undergo specific adsorption (forms inner-sphere complexes) on most
metal hydroxides through a ligand exchange mechanism involving surface hydroxyl
groups (Hiemstra and Van Riemsdijk, 1996). Surface complexes formed by ligand
exchange, have common ligand(s) present at the same electrostatic position as the
surface oxygens (Hiemstra and Van Riemsdijk, 1996). As previously noted, the most
reactive functional group on kaolinite is the aluminol, ( AlOH ), group located on
the crystal edges. Similarly for bauxite, the singly coordinated hydroxyl group
AlOH) ( located on gibbsite edges is the reactive aluminol group.

57

The Visual Minteq program was used to calculate the fractional composition of
phosphate species available in solution at different pH values (Figure 17). The Visual
Minteq program is a chemical equilibrium program used for calculation of metal
speciation and solubility equilibria in natural waters, as well as for adsorption
modeling. A phosphate species that is dominant in a certain pH range (Figure 17) is
used in the description of reactions involving phosphate ions under pH conditions that
are within that range.


Figure 17 Fractional composition of phosphate species in solution at different
pH, calculated using the Visual Minteq program

The ligand exchange reactions on kaolinite and gibbsite can be simplistically
represented by Equations 16 and 17 representing formation of a monodentate and
bidentate phosphate surface complexes respectively (Manning and Goldberg, 1996).
Ion exchange reactions were dominant under low pH conditions (generally below pH
7) hence the need to use a phosphate species that is dominant in solution below pH 7
(Figure 17).


58

-
4 2
-
4 2
OH PO AlH PO H AlOH + + (16)

-
2 4 2
-
4 2
OH O H HPO Al) ( PO H AlOH 2 + + + (17)

The release of hydroxyl ions into solution during the reaction (Equations 16 and 17),
explains the observed increase in phosphate adsorption with decreasing pH down to
around pH 3. It is further noted that adsorption of phosphate ions on kaolinite and
gibbsite surfaces involves mainly hydroxyl groups that are singly coordinated to
aluminium ions as such differences in phosphate adsorption capacities of kaolinite
and gibbsite would therefore arise from differences in surface densities of the reactive
functional groups and their electrostatic interaction with phosphate species in
solution.

Formation of phosphate complexes on iron oxides has been widely studied using
spectroscopic techniques, radiotracer studies, and quantum chemical calculations
(Persson et al., 1996; Hiemstra and Van Riemsdijk, 1999; Arai and Sparks, 2001;
Horanyi and Joo, 2002; Elzinga and Sparks, 2007; Kubicki et al., 2007; Rahnemaie et
al., 2007; Zhong et al., 2007; Stachowicz et al., 2008). Results from these studies
indicate that the adsorption behaviour of phosphate on goethite can been described
using three surface species, i.e., a monodentate (
3
FeOPO ), a bidentate
) PO (FeO) (
2 2
, and a bidentate protonated ( POOH FeO) (
2
) species. Similar to
kaolinite and gibbsite, ligand exchange reactions on the goethite surface involves
singly coordinated hydroxyl groups resulting in monodentate and bidentate phosphate
surface complexes (Equations 18 and 19).

59

O H OH FeOPO PO H FeOH
2 2
-
4 2
+ + (18)

O H 2 PO FeO) ( PO H FeOH 2
2 2 2
-
4 2
+ + (19)

Based on the evidence that gibbsite and goethite are the major minerals present in
bauxite, and that they have strong affinities for phosphate ions, it can be argued that
these two minerals were mainly responsible for removal of phosphate ions across the
whole pH range under study. Kaolinite, though present in the bauxite samples, has a
low phosphate removal capacity (Figure 18) as such the high percent phosphate
removal recorded for the bauxite (Figure 19) cannot be attributed to kaolinite. Further
more, Figure 16 does not indicate an adsorption envelope typical for adsorption of
phosphate ions on kaolinite and other clay minerals.

The increase in phosphate uptake above pH 7 for both raw and treated kaolinite,
suggest phosphate precipitation by metal ions as the main uptake mechanism.
Analysis of the clay suspensions revealed significant amounts of calcium ions at
various pH values (Table 4), raising a possibility of precipitation of phosphate as
calcium phosphate minerals. The Visual Minteq program was used to calculate
saturation index (SI) values of possible calcium phosphate minerals using pH, calcium
ion concentration, equilibrium phosphate concentration, contribution of CO
2
towards
carbonate equilibria, and solution speciation as input variables.




60

Table 4 Ca
2+
concentration in solution and calculated saturation index values

Hydroxyapatite (HAP) and tricalcium phosphate (TCP) minerals had more positive
saturation index (SI) values as compared to the other possible calcium phosphate
minerals across the pH range studied (Table 4). It is noted from the calculated SI
values that all the solutions were over-saturated with respect to HAP whereas some
were under- saturated with respect to TCP. The increase in phosphate uptake with
increasing pH starting from around pH 7 could arguably be as a result of precipitation
of phosphate as calcium phosphates mainly HAP as represented by Equation 20
(noting that PO
4
3-
is the dominant phosphate species at higher pH Figure 17).

+ +
+
OH ) PO ( Ca OH PO 3 Ca 5
3 4 5
- 3
4
2
(20)

The normal working pH (without adjustments) for the kaolinite was 8.952 0.2106 as
such effects of other parameters on removal of phosphate ions will be explained on
the basis that phosphate ions precipitated as HAP.


SAMPLE

Equilibrium pH

Ca
2+
(mg/L)
Saturation index (SI)
HAP TCP
Raw kaolinite 6.241 440.9 4.634 -0.9730
7.239 130.2 7.723 0.598
7.928 53.48 8.958 1.092
8.952 12.18 9.641 1.101
9.819 6.283 7.513 -0.519
10.758 4.316 1.386 -4.575
Treated
kaolinite
8.403 157.7 12.88 3.237
61

4.3 Effect of kaolinite and bauxite dosage on the amount of phosphate ions
removed

The effect of increasing dosage on the amount of phosphate ions removed, are shown
by Figures 18 and 19 for kaolinite and bauxite respectively. The bauxite suspensions
used were acidic with a normal pH of 6.116 0.1352 (without adjustments). It is
noted from figure 18 that the amount of phosphate ions removed increased with
kaolinite dosage reaching 98.5 0.136% and 69.7 0.1% removal at a dosage of 80
g/L for the treated and raw clay respectively. Bauxite, however, required a dosage of
15 g/L to remove 98.4 0.105 % of phosphate ions given the same initial phosphate
concentration of 10 mg/L (Figure 19).


Figure 18 Plot of % phosphate removal against dosage for kaolinite

62


Figure 19 Plot of % phosphate removal against dosage for bauxite

Studies on removal of phosphate ions by red mud (Pradhan et al., 1998; Huang et al.,
2008), iron oxide tailings (Zeng et al., 2004), bauxsol (Akhurst et al., 2006), and
alunite (Ozacar, 2006), reported effective dosages ranging from 10 to 20 g/L. These
dosages were reported for various initial phosphate concentrations ranging from 10 to
100 mg/L. The phosphate removal capacity of bauxite is comparable to that of red
mud, iron oxide tailings, bauxsol, and alunite. Phosphate ions have a great affinity for
metal oxides (Goldberg et al., 1996), as such adsorbents that contain iron, aluminium,
and magnesium oxides have high phosphate removal capacities. The increase in
phosphate removal with increasing bauxite dosage noted in Figure 19, can be
attributed to an increase in number of active sites.

Blast furnace slag (Kostura et al., 2005; Johansson and Gustafsson, 2000) and fly ash
(Can and Yildiz, 2006; Chen et al., 2007), are some of the materials that remove
phosphate ions from solution through precipitation of calcium phosphates. More than
90% of phosphate was reported removed from solution using dosages ranging from 4
63

to 20 g/L, for initial phosphate concentrations that ranged from 10 to 1000 mg/L. It is
evident from these reported dosages that kaolinite has a very low phosphate removal
capacity as compared to fly ash and blast furnace slag. This low phosphate removal
capacity can be attributed to low calcium content of the kaolinite (Table 4), hence the
requirement of very high dosages to effect significant phosphate removal.
Hydrochloric acid treated kaolinite indicated a higher phosphate removal capacity
(Figure 18) than raw kaolinite possibly as a result of release of more calcium ions
through the reaction represented by Equation 21.

(l) 2 (g) 2 (aq) 2 (aq) (S) 3
O H 2 CO CaCl 2HCl CaCO + + + (21)

Despite showing a low phosphate removal capacity, an increase in phosphate removal
with increasing dosage was noted for both raw and treated kaolinite. The increase in
dosage translates into an increase in Ca
2+
concentration in solution consequently an
increase in the Calcium/Phosphate (Ca/P) ratio of the system. With the increase of the
Ca/P ratio, the saturation index (SI) with respect to hydroxyapatite (HAP) increases as
well (Song et al., 2002a), and HAP precipitation becomes thermodynamically more
favourable. The increased thermodynamic favourability of HAP precipitation
translates into faster precipitation rates and higher phosphate removal efficiencies
(Song et al., 2002b). It follows from this hypothesis that the observed higher
efficiencies for treated clay resulted from higher Ca/P ratios as compared to raw clay.
Results from the study of the effect of pH on phosphate removal indicated higher
removal in the lower pH region that was attributed to adsorption of phosphate ions on
kaolinite surfaces. It was further noted that 88.8% of phosphate was bound by raw
kaolinite at the lower pH maxima, using a very high dosage of 40 g/L. This indicates
64

that kaolinite has a low phosphate removal capacity as compared to iron, magnesium
and other aluminium hydroxides. The low phosphate removal capacity of kaolinite
could be as a result of lower densities of the aluminol functional group as well as
electrostatic hindrance to adsorption of anions by the permanent negative charges on
the basal planes of the kaolinite particles.

4.4 Effect of contact time on the amount of phosphate ions removed by
bauxite and kaolinite

Figure 20 shows the effect of contact time on amount of phosphate ions removed by
kaolinite (raw) at 20 and 40 C and pH 8.952 0.2106. Phosphate uptake approached
equilibrium faster at 40 C and slowly at 20 C but eventually leveled off after 270
minutes at both temperatures. Equilibrium was established faster for the kaolinite
samples used in this study as compared to 8 hours for blast furnace slags (Kostura et
al., 2005) and 24 hours for fly ash (Chen et al., 2007) recorded at 20C. It is noted
later that adsorption of phosphate ions on bauxite was much faster compared to
precipitation of HAP by kaolinite. Applicability of kaolinite for wastewater treatment
will be hampered by its relatively slow reactions and lower phosphate removal
capacity. The low phosphate removal capacity overrides advantages associated with
precipitation of calcium phosphates such as direct use of the precipitates as fertilizer.
65


Figure 20 Plot of phosphate uptake against time for kaolinite

The effect of temperature on the rate of precipitation of HAP is shown as well in
Figure 20. The difference in the reaction rates at 20 and 40C, suggests a strong
dependence of the precipitation reaction on temperature. Song et al. (2002b) noted
that temperature affects not only precipitation of hydroxyapatite, but also the
association/dissociation reactions of the related species and the thermodynamic
solubility of hydroxyapatite. The overall result is that saturation index values for
hydroxyapatite increase with temperature. The increase in saturation index values
translates into increased reaction rates. Assuming that kaolinite is used during
wastewater treatment, it would be necessary to adjust either the contact time or the
dosage when there is a big fluctuation in temperature of the influent, for example after
the transition from the hot to the cold season and vice versa.

The effect of contact time on the amount of phosphate ions adsorbed on the bauxite
surface is shown by Figure 21. The bauxite suspensions were acidic with a pH of
6.116 0.1352. It is noted from the plot that much of the adsorption on bauxite
66

occurred after 30 minutes of contact. Faster reactions were also reported for
adsorption of phosphate ions on iron oxide tailings (Zeng et al., 2004), as well as on
alunite (Ozacar, 2006). Adsorption of phosphate ions on metal oxides proceed
through a fast reaction stage lasting a few hours, followed by a slow reaction that can
last for months to reach equilibrium (Barrow, 1999). Kinetics studies done for a few
hours provide information about the faster stage of the reaction as was the case in this
study. It can be argued that the faster reaction is of primary importance in use of metal
oxides for wastewater treatment considering that shorter contact times are desirable.
The faster reaction on bauxite would minimize contact times between the adsorbent
and wastewater during treatment enabling treatment of larger volumes of wastewater
in a short time.


Figure 21 Plot of phosphate uptake against time for phosphate adsorption on
bauxite

It is noted from Figure 21 that the reaction rate for adsorption of phosphate ions on
bauxite was not significantly affected by temperature. As the adsorption reaction is
fairly rapid, the effects of temperature are mainly seen on the position of its
equilibrium, rather than on its rate (Barrow, 1999). In an adsorption reaction,
67

temperature can affect the electric potential of the surface, activity of ions in solution,
and the affinity of the ions for the surface (Barrow, 1992). The magnitude, and even
the direction of the effects, can vary according to the reactants. The weak dependence
of phosphate adsorption on temperature would minimize the need for adjustment of
operating conditions when temperature fluctuations occur.

The kinetic mechanism of phosphate uptake by both kaolinite and bauxite was
evaluated with two different models: the Largergren pseudo first order and pseudo
second order models (Huang et al., 2008). The pseudo first order equation takes the
form:

2.303
t k
logq ) q log(q
1
e t e
= (22)

where q
t
is the amount of phosphate removed at time t (mg/g), q
e
is the amount of
phosphate bound at equilibrium (mg/g), and k
1
is the equilibrium rate constant for
pseudo first-order kinetics (s
-1
). The pseudo second-order equation is expressed as:

e
2
e 2
t
q
t
q k
1
q
t
+ = (23)

where k
2
is the equilibrium rate constant for the second-order kinetics (g mg
-1
s
-1
).
Data obtained at 20 and 40C fitted better to the pseudo second-order model than to
the pseudo first-order model as indicated by R
2
values shown in Figures 22, 23, 24
and 25.

68


Figure 22 Second order fits for phosphate precipitation by kaolinite


Figure 23 First order fits for phosphate precipitation by kaolinite



69


Figure 24 Second order fits for phosphate adsorption on bauxite



Figure 25 First order fits for phosphate adsorption on bauxite

Equilibrium rate constants for phosphate removal by kaolinite were calculated to be
1.467 x 10
-3
g/mg.s and 1.092 x 10
-2
g/mg.s, at 20 and 40 C respectively. Rate
constants for adsorption of phosphate ions on bauxite were calculated as 2.747 x 10
-3

and 2.982 x 10
-3
g/mg.s at 20 and 40 C respectively. The larger difference in the rate
70

constants for phosphate precipitation by kaolinite indicates the strong dependence of
the process on temperature. As noted previously, phosphate adsorption on bauxite had
a weak dependence on temperature. The small difference in the rate constants for the
phosphate adsorption process further indicates its weak dependence on temperature.
The value of the activation energy of a reaction can be an important clue to the
mechanism of a reaction (Barrow, 1999). Activation energy (E
a
) for phosphate
precipitation by calcium ions, and adsorption on bauxite, was calculated using the
relation (Equation 24);


|
|

\
|
=
|
|

\
|
a b
a
2(b)
2(a)
T
1
T
1
R
E
k
k
ln (24)

where k
2(a)
and k
2(b)
are second order equilibrium constants at 20 and 40 C
respectively, T
a
and T
b
are temperatures in Kelvin and R is the universal gas constant
(8.314 J/K.mol). Lower activation energy of 3.130 kJ/mol was calculated for
phosphate adsorption on bauxite compared to 76.56 kJ/mol for phosphate
precipitation by calcium ions. This is in agreement with the observation that iron
oxides have a strong affinity for phosphate ions, hence the lower activation energy
and a faster reaction with phosphate ions. Despite reporting lower activation energy
for adsorption of phosphate on bauxite, it cannot be generalized that adsorption
reactions require lower activation energies compared to precipitation as some metal
oxides have poor affinities for ionic species in solution. Furthermore, oversaturation
of a solution with respect to a solid phase would affect the rate of the reaction and
hence the value of the activation energy. The difference in values of activation energy
for two reactions cannot be used to draw a conclusion as to the mechanism involved
71

(adsorption or precipitation) unless conditions forming a basis of comparison are set
clearly.
4.5 Effect of initial phosphate concentration on the phosphate removal
capacity of bauxite and kaolinite

Figure 26 shows the effect of initial phosphate concentration on the amount of
phosphate ions removed by kaolinite. It is noted from the plot that phosphate uptake
increased with increasing initial phosphate concentration with different trends
obtained for raw and treated kaolinite. Phosphate uptake increased steadily across the
whole concentration range for the treated kaolinite but remained fairly constant
between 40 and 60 mg/L for the raw kaolinite (p > 0.05).


Figure 26 Plot of phosphate uptake against initial phosphate concentration for
phosphate removal by kaolinite at a suspension pH of 8.952 0.2106

The influence of increasing phosphate concentration can be described in terms of its
effect on the thermodynamic favourability and rate of the precipitation process. The
saturation index for hydroxyapatite precipitation can be described by Equation 25:

72

SI = 5log(C
Ca
2+
) + 3 log(C
PO4
3-
) + log(C
OH
-
) + log(f
2
5
f
3
3
f
1
) logKsp (25)

where C
x
is the concentration of the x ion and f
i
is the activity coefficient of the i-
valent ion (Song et al., 2002a). The equation indicates strong effects of concentrations
of calcium, phosphate and hydroxyl ions on the saturation index with respect to
hydroxyapatite. Previous studies (Song et al., 2002a,b) reported the saturation index
to be a logarithmic function of the phosphate concentration with respect to HAP at
constant pH, ionic strength, and calcium concentration conditions. Phosphate
precipitation becomes more thermodynamically favourable with increasing phosphate
concentration as indicated by increasing saturation index values. Increasing phosphate
concentration therefore influences supersaturation, then the precipitation rate and
efficiency of the precipitating system (Song et al., 2002b). Considering that
precipitation took place at relatively constant calcium concentrations, increasing
phosphate concentration could have decreased the Calcium/Phosphate (Ca/P) ratio of
the precipitating system. Saturation index values are directly proportional to the Ca/P
ratio under constant pH conditions (Song et al., 2002b) as such slower rates and
stagnant efficiencies are expected with increasing phosphate concentration under
constant pH and calcium concentration conditions. The Ca/P ratios for the raw
kaolinite could have approached the minimum limit of 1.67 (for HAP) much faster as
a result of lower calcium concentrations resulting in fairly constant phosphate uptake
starting from an initial phosphate concentration of 40 mg/L.

The steady increase in phosphate uptake for the treated kaolinite can be explained by
higher calcium concentrations that translated into higher Ca/P ratios across the studied
phosphate concentration range. Fairly constant phosphate uptake with increasing
phosphate concentration is predicted for treated kaolinite beyond 60 mg/L. It would
73

be necessary to adjust the Ca/P ratio during wastewater treatment to compensate for
fluctuations in influent phosphate concentration.

The effect of increasing initial phosphate concentration on phosphate uptake by
bauxite was studied at temperatures of 10, 20, 30, 40, 50 and 60 C. Phosphate uptake
increased rapidly with increasing equilibrium phosphate concentration from 0 up to
around 15 mg/L at all temperatures followed by a gradual increase beyond 15 mg/L
(Figure 27). The rapid increase in phosphate uptake below 15 mg/L could be as a
result of availability of excess active sites that later became saturated with further
increase in phosphate concentration resulting in a gradual increase in phosphate
uptake beyond 15mg/L.



Phosphate uptake at the studied temperatures was fitted to the Freundlich equation
(Figure 27) by means of non-linear regression performed using Sigma plot software.
The data fitted well to the Freundlich equation as indicated by a minimum R
2
value of
Figure 27: Phosphate uptake by bauxite (obtained at a suspension pH of 6.116
0.1352) fitted to the Freundlich equation using non-linear regression.
74

0.9132 obtained for the 283 K system (Table 5). The Freundlich constant values K
f
,
obtained at each temperature were converted to dimensionless values by multiplying
them by 1000 (Milonjic, 2007). The dimensionless equilibrium values K
p
, were used
to calculate the Gibbs free energy change of adsorption G
ad
, at each temperature
(Table 5) using the relation (Equation 26):

lnK RT - G = (26)

Table 5 Equilibrium constants and Gibb's free energy change values
Temperature
(K)
K
f

(dm
3
g
-1
)
R
2
Kp G
(kJmol
-1
)
283 0.7698 0.9132 769.8 -15.64
293 0.8864 0.9818 886.4 -16.53
303 1.054 0.9832 1053.
6
-17.53
313 1.057 0.9961 1056.
7
-18.12
323 1.178 0.9894 1177.
7
-18.99
333 1.215 0.9907 1214.
7
-19.66

Negative G values were obtained for phosphate adsorption on bauxite (Table 4)
providing thermodynamic evidence that the adsorption process was spontaneous. The
calculated G values became more negative with increasing temperature indicating an
increase in thermodynamic favourability of the adsorption process. The enthalpy
change of the reaction, H, was estimated using the Vant Hoff equation (Atkins,
1990), given by Equation 27:

75

R
H -
d(1/T)
lnK d

= (27)
where K is an equilibrium constant, T is temperature (K), R is the universal gas
constant (8.314 JK
-1
mol
-1
), and H is the standard reaction enthalpy. The enthalpy
change of a reaction, H, can be calculated from the slope of a plot of lnK
p
against
1/T (Atkins, 1990). H for phosphate adsorption on bauxite was estimated to be
+6.943 kJmol
-1
(Figure 28) indicating that the process is endothermic as such
adsorption of phosphate ions should increase with increasing temperature.


Figure 28 Vant Hoff plot for phosphate adsorption on bauxite

4.6 Effect of calcium, magnesium, sulphate and carbonate ions on phosphate
uptake by raw and treated clay

Figures 29 and 30 shows the effect of calcium, magnesium, sulphate and carbonate
ions on phosphate uptake by raw and treated kaolinite respectively. It is noted from
the diagrams that phosphate uptake increased with increasing calcium and magnesium
ion concentration (p < 0.05), decreased with increasing carbonate concentration (p <
0.05) and was little affected by sulphate ions in both cases (p > 0.05).
76


Figure 29 Effect of competing ions on phosphate removal by raw kaolinite


Figure 30 Effect of competing ions on phosphate uptake by treated kaolinite

From the thermodynamic point of view, an increase in calcium concentration resulted
in an increase in the Ca/P ratio of the system. An increase in the Ca/P ratio increased
the thermodynamic favourability of calcium phosphate precipitation consequently
increasing the reaction rate and phosphate removal efficiency (Song et al., 2002b).
The increase in phosphate removal with increasing calcium concentration further
77

justifies calcium phosphate precipitation as the dominant reaction mechanism under
high pH conditions. Increasing calcium concentration was therefore synonymous to
the effect of increasing kaolinite dosage. It is noted from Figures 29 and 30 that the
increase in phosphate removal with increasing calcium concentration was more
pronounced for the raw kaolinite as compared to the treated kaolinite. This
observation supports the fact that more calcium ions were released into solution from
treated kaolinite as compared to raw kaolinite such that additional calcium ions
resulted in a small increase of the Ca/P ratio and consequently a small increase of the
phosphate removal efficiency.

Magnesium ions have been reported to inhibit precipitation of HAP (Ferguson and
McCarty, 1971, Cao et al., 2007; Cao and Harris, 2008). Due to the chemical
similarity between Mg
2+
and Ca
2+
, it is possible that Mg
2+
co-precipitates with Ca
phosphates, inducing the formation of the less stable amorphous calcium phosphate
(ACP) rather than stable HAP (Ferguson and McCarty, 1971). The presence of Mg
2+

further inhibits the conversion of ACP to the more stable HAP with the net effects of
maintaining higher phosphate concentrations in solution (Cao and Harris, 2008).
Results from this study however indicate that phosphate precipitation was promoted
by increasing magnesium ion concentration. Inhibition of calcium phosphate
precipitation by Mg
2+
ions is mitigated by presence of carbonate ions under reaction
conditions of pH higher than 9 (Cao and Harris, 2008). This mitigation is attributed to
formation of stable aqueous MgCO
3
ion pairs that reduce amount of Mg
2+
and CO
3
2-

ions available for other reactions in solution. This hypothesis was consistent with
speciation model that predicted that, 12% of MgCO
3
would be formed in such a
system (Cao and Harris, 2008). It is noted from the hypothesis that formation of
MgCO
3
pairs would significantly reduce the amount of calcium ions bound to
78

carbonate therefore increasing calcium ion concentration. It is possible that more
calcium ions would be available in a system containing both magnesium and
carbonate ions as compared to one with magnesium absent. The reaction systems used
in this study were open to air as such presence of carbonates and magnesium probably
resulted in an increase in calcium ion concentration and increasing phosphate removal
in the process.

It is noted from Figures 29 and 30 that the amount of phosphate ions precipitated
decreased with increasing carbonate concentration. Carbonate ions can inhibit
precipitation of calcium phosphate through formation of carbonate-substituted apatites
or precipitation of calcium carbonate (Barralet et al., 1997; Kapolos and Koutsoukos,
1999; Plant and House, 2002; Song et al., 2002b, Cao et al., 2007; Cao and Harris,
2008). The carbonate ion can substitute at two sites in the apatite structure, namely the
hydroxyl and the phosphate ion positions, giving A-type and B-type carbonate
hydroxyapatite (CHA) respectively (Barralet et al., 1997). It is the B-type CHA that is
often formed by precipitation with net results of higher equilibrium phosphate
concentration in solution. Formation of B-type CHA is dominant at pH conditions of
lower than 9 as such the observed effects in this study (pH higher than 9) could have
been guided by a different mechanism. In their study on the interactive effect of
magnesium and carbonate ions on calcium phosphate precipitation, Cao and Harris
(2008) observed precipitation of calcite accompanied by precipitation of the less
stable amorphous calcium phosphate (ACP) at pH 9.1. Precipitation of calcite was
promoted by the higher concentration of carbonate relative to bicarbonate ions. As a
result of calcite precipitation, calcium concentration in solution decreased and the
calcium phosphate precipitation rate was reduced. Inhibition of hydroxyapatite
precipitation and reduced rates had the net effect of increasing the Ca/P ratio without
79

the net effect of increased phosphate precipitation owing to the formation of the more
soluble ACP. This mechanism can be applied to the observed trends in Figures 29 and
30 for raw and treated kaolinite respectively.

Similar to results of this study, Cao et al., (2007) observed that sulphate had little
influence on the crystallization of poorly crystalline HAP. Sulphate ions however,
reduced the HAP precipitation rate to a small extent. This was attributed to sulphate
substitution in the HAP lattice within the first hour of contact, followed by
replacement of sulphate by phosphate after longer contact times (Cao et al., 2007).
Replacement of sulphate by phosphate after longer contact hours resulted in
insignificant changes in amount of phosphate ions precipitated with increasing
sulphate concentration.

4.7 Effect of magnesium, calcium, sulphate and carbonate/bicarbonate ions
on phosphate uptake by bauxite

Figure 31 shows the effect of magnesium, calcium, sulphate, and carbonate ions on
the amount of phosphate ions bound on bauxite. It is noted from the diagram that
phosphate uptake increased in the presence of magnesium and calcium ions (p <
0.05), and decreased in the presence of sulphate and carbonate/bicarbonate ions (p <
0.05).
80


Figure 31 Effect of competing ions on adsorption of phosphate ions on bauxite at
an initial bauxite suspension pH of 6.116 0.1352

Interaction between active surface hydroxyl groups on metal hydroxides and calcium
and magnesium ions in solution is essential for the elucidation of the effect of the ions
on adsorption of phosphate ions on the metal hydroxide surfaces. Rahnmaie et al.
(2006) and Stachowicz et al. (2008) elucidated the interaction between calcium and
magnesium ions with goethite using a combination of proton titration procedures and
quantum chemical calculations. Magnesium showed a high affinity for goethite
described by the formation of bidentate innersphere complexes. According to
Rahnmaie et al. (2006), the introduction of much positive charge due to innersphere
complexation of magnesium ions creates a relatively high change in the potential of
the 1- and 2-plane (as depicted in the electrostatic double layer model, Figure 13).
These potential changes result in lower electrostatic attraction or increased repulsion
of other magnesium ions, as such suppressing the formation of outersphere
complexes.

81

For calcium, a much lower affinity for goethite was found. Rahnemaie et al. (2006)
described the calcium adsorption on goethite using the charge distribution model
using a combination of inner and outer sphere complexation. Preliminary modelling
resulted in a surface charge attribution of Ca
2+
that is typical for monodentate
innersphere complexation. The observed increase in phosphate adsorption in the
presence of calcium and magnesium ions can be explained by electrostatic
interactions (Rietra et al., 2001). The negative charge of the adsorbed phosphate ions
stimulates the binding of the positively charged calcium and magnesium ions
(Stachowicz et al., 2008). The binding of the positively charged calcium and
magnesium ions suppresses an increase in negative charge (arising from the
adsorption of phosphate ions) with suggested mutual effects of increased adsorption
of both phosphate and the cations (Stachowicz et al., 2008). It is noted from Figure 31
that the increase in phosphate adsorption in the presence of magnesium is smaller as
compared to the increase in the presence of calcium. Magnesium adds less charge to
the 1-plane (as depicted in the electrostatic double layer model, Figure 13) hence the
small effect on phosphate adsorption since the mutual interaction between adsorbed
cations and anions mainly stems from the charge attribution to this electrostatic plane
(Stachowicz et al., 2008).

It is noted from Figure 31, that carbonate ions strongly reduced the amount of
phosphate ions adsorbed on goethite and gibbsite minerals that are present in bauxite.
Carbonate interactions with goethite can be described using a combination of inner-
and outer-sphere complexation (Rahnemaie et al., 2007). The main inner-sphere
species is a binuclear bidentate species, CO FeO) (
2
, which interact with Na
+
in an
outer-sphere fashion, forming
+
CO.....Na FeO) (
2
(Hiemstra et al., 2004; Rahnemaie
82

et al., 2007). Formation of carbonate outer-sphere complexes is very limited and is
due to interaction of carbonate rather than bicarbonate with protonated surface groups
(
- 2
3
/
2
CO ...... FeOH
2 1 +
and
- 2
3
1/2
3
CO ....... OH Fe
+
) (Rahnemaie et al., 2007). Besides
goethite, Kubicki et al. (2007) described carbonate binding on haematite using a
combination of bidentate inner-sphere and outer-sphere complexation. As a result of
formation of the carbonate complexes, carbonate ions compete with phosphate for
surface sites leading to a decrease in the adsorption of phosphate. Rahnemaie et al.
(2007) noted that the competitive interaction of carbonate decreases as the pH
increases and is very weak above pH 10.5 with maximum adsorption in closed
systems occurring around pH 6-7. The observed strong competitive effect of
carbonate in this study is justified by the pH conditions of approximately 6.1 for
which maximum carbonate adsorption is expected. It can therefore be predicted that
presence of carbonate in wastewater in concentrations relatively higher than
phosphate would greatly reduce the phosphate uptake efficiency of the bauxite.

Sulphate ions weakly competed for reactive sites with phosphate (compared to
carbonate) resulting in reduced phosphate uptake (Figure 31). Sulphate binding on
iron (hydr)oxide surfaces can be described by the formation of inner-sphere either
monodentate or bidentate bridging mechanisms that are possibly dependent upon
mineral surface loading (Kubicki et al., 2007). Sulphate has a weaker affinity for
metal hydroxide surfaces as compared to phosphate such that no sulphate is adsorbed
above the point of zero charge of goethite (Geelhoed et al., 1997). In sulphate-
phosphate ion systems, presence of sulphate causes a small decrease in phosphate
adsorption at relatively low pH (Geelhoed et al., 1997), which explains the weaker
competition observed in this study.
83

4.8 Multi-interactive effect of magnesium, calcium, sulphate, and bicarbonate
ions on the phosphate removal efficiency of bauxite.


Figure 32 shows the multi-interactive effect of calcium, magnesium, sulphate, and
carbonate ions on phosphate uptake by bauxite. It is noted from Figure 32 that
presence of the competing ions enhanced phosphate removal i.e. 98.93 0.4137 %
removal obtained for the dosage of 15 g/L in the multi-component system compared
to 74.43 0.3431 % in the control system.


Figure 32 Multi-competitive effect of calcium, magnesium, carbonate, and
sulphate ions on phosphate adsorption on bauxite


The presence of carbonate and sulphate ions (200 mg/L) was expected to provide a
very high competition for adsorption sites with phosphate ions even in the presence of
calcium and magnesium ions (200 mg/L). This hypothesis is based on the
understanding that enhancement of phosphate adsorption by calcium and magnesium
ions as a result of electrostatic interactions is smaller in magnitude compared to the
competitive effect of sulphate and carbonate ions. In an attempt to explain the
84

observed enhancement of phosphate uptake in the multi-component system, the
Visual Minteq program was used to calculate saturation index values of possible solid
phases that could precipitate. The results indicated that the system was saturated with
respect to calcite (SI = 2.539), dolomite (SI = 5.584), hydroxyapatite (SI = 17.439),
and magnesite (SI = 1.808) at a pH of 8.452 0.5245. Mass distribution calculations
indicated that 99.68 % of calcium, 83.53 % of carbonate, 63.15 % of magnesium,
97.02 % of phosphate, and 0 % of sulphate ions would precipitate in the multi-
component system. With 83.53 % of carbonate and 0 % of sulphate ions precipitated,
there would still be a higher concentration of carbonate and sulphate ions to offer
phosphate ions stiff competition for adsorption sites. Despite the presence of a higher
concentration of competing ions, 97.02 % of phosphate ions would precipitate mainly
as hydroxyapatite greatly reducing phosphate concentration in the process.

Using results from the Visual Minteq calculations, the observed enhancement of
phosphate removal in the multi-component system (Figure 32) is due to precipitation
of phosphate ions as hydroxyapatite. The optimum bauxite dosage for the multi-
component system can be estimated to be in range of 20 to 30 g/L. The optimum
dosage estimated through the bench studies cannot be directly applied for wastewater
treatment systems owing to differences in composition of the water as well as means
of mechanical mixing of the bauxite and the wastewater. Rigorous shaking is involved
during bench studies that might result in ripping of the bauxite particles and exposure
of extra surface area containing reactive functional groups. Estimation of optimum
dosage for use during wastewater treatment would require a pilot study and
knowledge of the average chemical composition of the wastewater influent.
85

CHAPTER FIVE: CONCLUSIONS AND RECOMMENDATIONS

5.1 Conclusions

The study has shown that raw bauxite has a higher phosphate removal capacity than
both raw and treated kaolinite. At a dosage of 10 g/L, bauxite, raw kaolinite, and
treated kaolinite reduced concentration of phosphate ions in solutions by 93.2 0.152,
19.3 0.344, and 50.6 0.436 % respectively. Raw bauxite was more effective than
kaolinite, attributed to presence of goethite and gibbsite that have a high affinity for
phosphate ions. Acid treatment of kaolinite improved its phosphate removal
efficiency, attributed to release of more calcium ions from calcium carbonates present
as impurities (phosphate was removed through precipitation of hydroxyapatite). The
phosphate removal capacity of kaolinite is much lower than that for fly ash and blast
furnace slag, and kaolinite cannot be considered for use in a wastewater treatment
plant.

Phosphate removal showed a strong dependence on initial pH for both bauxite and
kaolinite. For bauxite, the percent removal of phosphate ions increased with
decreasing pH with high removal achieved below pH 5. This was attributed to
removal of phosphate ions through a ligand exchange mechanism involving reactive
hydroxyl groups on the goethite and gibbsite surface. For kaolinite, a region of high
removal was observed between pH 2 and 5, followed by a region of low removal
between pH 5 and 7 (a valley), and one of high removal above pH 7.
86

High percent phosphate removal between pH 2 and 5 was attributed to adsorption of
phosphate ions on kaolinite particles and iron oxide impurities, whereas high removal
above pH 7 was attributed to precipitation of phosphate ions as hydroxyapatite
(HAP). If kaolinite had a high phosphate removal capacity at both low and high pH, it
could have been a versatile material applicable for various wastewater pH conditions.

Kinetics studies have shown that a faster reaction occurred for the adsorption of
phosphate ions on bauxite, with 63.4 % removal achieved after 30 minutes of contact
(for a dosage corresponding to a maximum removal of 69.28 1.427 %). The faster
reaction on bauxite would minimize contact times between the adsorbent and
wastewater during treatment enabling treatment of larger volumes of wastewater in a
short time. For kaolinite, there was a gradual increase in phosphate precipitation with
time at 20 C compared to 40 C. However, equilibrium conditions were achieved
after 270 minutes at both temperatures. Kinetic data obtained for both bauxite and
kaolinite fitted better to the Largergren pseudo second-order equation, and lower
activation energy of 3.130 kJ/mol was calculated for the reaction on bauxite compared
to 76.56 kJ/mol for precipitation of hydroxyapatite. The lower activation energy for
phosphate adsorption on bauxite confirms the high affinity of goethite and gibbsite
minerals for phosphate ions.

The study has as well shown that phosphate uptake increase with increasing initial
phosphate concentration for both kaolinite and bauxite. For kaolinite, the increase in
phosphate uptake was attributed to an increase in the thermodynamic favourability of
the precipitation process and consequently increased reaction rates, whereas the
increase for bauxite was due to availability of excess reactive functional groups. Data
87

obtained for the effect of increasing initial phosphate concentration on phosphate
uptake by bauxite fitted well to the Freundlich equation by using non-linear
regression. Negative Gibbs free energy change (G) values were calculated using the
calculated equilibrium constants as well as an enthalpy change value of +6.943
kJ/mol. The negative G values confirm the spontaneity of phosphate adsorption on
bauxite; whereas the positive enthalpy change value indicates that the adsorption
process was endothermic as such phosphate adsorption should increase with
increasing temperature.

Results indicate that precipitation of phosphate ions increased with increasing calcium
and magnesium ion concentration (for kaolinite), decreased with increasing carbonate
ions concentration, and was not affected by increasing sulphate ion concentration. An
increase in calcium ion concentration resulted in an increase of the calcium/phosphate
(Ca/P) ratio of the system, resulting in increased reaction rates and efficiency. A
higher concentration of magnesium ions decreased the concentration of carbonate ions
in solution through the formation of MgCO
3
ion pairs, thereby increasing amount of
available calcium ions and reducing carbonate substitution in precipitated HAP. On
the other hand, carbonate ions inhibited precipitation of HAP through carbonate
substitution for phosphate in precipitated HAP and precipitation of calcium as calcite.
Sulphate ions however, did not affect precipitation of HAP because they returned into
solution after longer contact time if substitution of sulphate for phosphate in HAP
occurred.

For bauxite, adsorption of phosphate ions increased with increasing calcium and
magnesium concentrations. This was attributed to electrostatic interactions between
88

the adsorbed positively charged ions and phosphate, with net effects of increased
phosphate adsorption. On the other hand, phosphate adsorption decreased with
increasing carbonate and sulphate concentrations. This was attributed to competition
between carbonate and sulphate ions with phosphate for active sites. It has been
shown that the competing effect of carbonate ions can be mitigated in a multi-
component system containing higher concentration of magnesium and calcium ions as
a result of the precipitation of calcite, magnesite, as well as HAP. The reported
optimum dosage for the multi-component system cannot however be directly adopted
for application at a wastewater treatment plant; as such a pilot study is required to
determine optimum conditions applicable to wastewater treatment plants.

5.2 Recommendations

The conventional methods for removing phosphate ions from wastewater are either
expensive to operate in developing countries or have poor operation stability hence
the need to develop alternative methods. The use of bauxite as a phosphate adsorbent
should be exploited further owing to its high affinity for phosphate ions. Further, it is
recommended that,
Locally sourced calcite and gypsum should be tested for their
phosphate precipitating capacities. This recommendation follows from reports
indicating that some calcium-phosphate precipitates can be applied directly to
soil as fertilizer. This would minimize problems associated with handling of
huge sludge volumes and assist in recycling phosphorus.
A pilot study on use of bauxite should be carried out to devise mixing
mechanisms of bauxite and wastewater, to estimate the optimum dosage with
respect to varying wastewater composition and the mixing mechanism, to
89

assess the effect of the bauxite on other effluent parameters such as turbidity
and concentration of other ionic species, and to devise mechanisms of sludge
disposal or possible recycling and phosphorus recovery.










































90

REFERENCES


Akhurst, D.J., Jones, G.B., Clark, M., and McConchie, D. (2006). Phosphate removal
from aqueous solutions using neutralized bauxite refinery residues (Bauxsol).
Environmental chemistry 3: 65-74

Altundogan, H.S., and Tmen, F. (2003). Removal of phosphates from aqueous
solutions. Journal of chemical technology and biotechnology, 78 (7): 824-833

American Public Health Association (APHA) (1989). Standard methods of the
examination of water and wastewater, 16
th
Edition. APHA, AWWA, and
WPCF, New York.

Arai, Y., and Sparks, D.L. (2001). ATR-FTIR spectroscopic investigation on
phosphate adsorption mechanisms at the ferrihydrite-water interface. Journal
of colloid and interface science, 241: 317-326

Atkins, P.W. (1990). Changes of state: chemical reactions. In, Physical chemistry.
Oxford University press, Suffolk, UK. Pp 142-144

Avena, M.J., Mariscal, M.M., and De Pauli, C.P. (2003). Proton binding at clay
surfaces in water. Applied clay science. 24: 3-9

91

Averbuch-pouchot, M., and Durif, A. (1996). Definition, classification and
nomenclature: monophosphates. In, Topics in phosphate chemistry. World
scientific publishing co., Singapore. Pp 29-39

Barralet, J., Best, S., and Bonfield, W. (1997). Carbonate substitution in precipitated
hydroxyapatite: An investigation into the effects of reaction temperature and
bicarbonate ion concentration. Journal of biomedical materials research part
A, 41(1): 79-86

Barrow, N.J. (1992). A brief discussion on the effect of temperature on the reaction of
inorganic ions with soil. Journal of soil science, 43: 37-45

Barrow, N.J. (1999). The four laws of soil chemistry: the Leeper lecture 1998.
Australian journal of soil research 37: 787-829

Barrow, N.J. (2000). Towards a single-point method for measuring phosphate
sorption by soils. Australian journal of soil research, 38: 1099-1113

Barrow, N.J., Cartes, P., and Mora, M.L. (2005). Modifications to the Freundlich
equation to describe anion sorption over a large range and to describe
competition between pairs of ions. European journal of soil science, 56: 601-
606

Barrow, N.J. (2008). The description of sorption curves. European Journal of Soil
Science, 59(5): 900-910
92


Borkovec, W.F., Jonsson, B., and Koper, G.J.M. (2001). Ionization processes and
proton binding in polyprotic systems: small molecules, proteins, interfaces,
and polyelectrolytes. Journal of colloid and interface science, 16: 99-339

Can, M.Y., and Yildiz, E. (2006). Phosphate removal from water by fly ash: Factorial
experimental design. Journal of hazardous materials B, 35: 165-170

Cao, X., Harris, W.G, Josan, M.S., and Nair, V.D. (2007). Inhibition of calcium
phosphate precipitation under environmentally-relevant conditions. Science of
the total environment, 383: 205-215

Cao, X., and Harris, W. (2008). Carbonate and magnesium interactive effect on
calcium phosphate precipitation. Environmental science and technology,
42(2): 436-442

Carpenter, S.R., Caraco, N.F., Correll, D.L., Howarth, R.W., Sharpley. A.N., and
Smith, V.H. (1998). Non-point pollution of surface waters with phosphorus
and nitrogen. Ecological applications, 8: 559-568.

Chen, J., Kong, H., Wu, D., Chen, X., Zhang, D., and Sun, Z. (2007). Phosphate
immobilization from aqueous solution by fly ashes in relation to their
composition. Journal of hazardous materials B 139: 293-300

93

Chen, Y.S. R., Butler, J. N., Stumm, W. (1973). Adsorption of phosphate on alumina
and kaolinite from dilute aqueous solutions, Journal of Colloid and Interface
Science, 43(2): 421-436

Chipofya, V.H., and Matapa, E.J. (2003). Destratification of an impounding reservoir
using compressed aircase of Mudi reservoir, Blantyre, Malawi. Physics and
Chemistry of the earth, 28: 1161-1164

Correctional Service of Canada (CSC) (2000). Environmental guidelines [online].
Available: http://www.csc-scc.gc.ca/text/plcy/doc/318-6-gl_e.pdf [8/12/09].

Eldwald, J.K., Toensing, D.C., and Leung, M.C.Y. (1976). Phosphate adsorption
reactions with clay minerals. Environmental science and technology, 10: 485-
490

Elzinga, J.E., and Sparks, D.L. (2007). Phosphate adsorption onto hematite: An in situ
ATR-FTIR investigation of the effects of pH and loading level on the mode
phosphate surface complexation. Journal of Colloid and Interface Science,
308: 53-70

Ferguson, A.E., and Mulwafu, W.O. (2004). Decentralization, participation and
access to water resources in Malawi. BASIS collaborative research
programme, University of Wisconsin-Madson [Online]. Available:
http://www.uneca.org/awich/Reports/Malawi%20Water%20Resources%20dec
entralization%20and%20participation.pdf. [29/02/09].
94


Ferguson, J.F, and McCarty, P.L. (1971). Effects of carbonate and magnesium on
calcium phosphate precipitation. Environmental science and technology,
5:53440.

Frost, R.L., Ding, Z., and Ruan, H.D. (2003). Thermal analysis of goethite. Relevance
to Australian indigenous art. Journal of Thermal Analysis and Calorimetry,
71(3): 783-797.

Ganor, J., Cama, J., and Metz, V. (2003). Surface protonation data of kaolinite-
reevaluation based on dissolution experiments. Journal of Colloid and
Interface Science, 264: 67-75

Geelhoed, J.S., Hiemstra, T., and Van Riemsdijk, W.H. (1997). Phosphate and
sulphate adsorption on goethite: single anion and competitive adsorption.
Geochimica et Cosmochimica Acta, 61 (12): 2389-2396.

Georgantas, D.A., and Grigoropoulou, H.P. (2007). Orthophosphate and
metaphosphate ion removal from aqueous solution using alum and aluminum
hydroxide. Journal of colloid and interface science, 315: 70-79

Goldberg, S., Davis, J.A., and Hem, J.D. (1996). The surface chemistry of aluminum
oxides and hydroxides. In: The environmental chemistry of aluminum, G.
Sposito (Ed). Lewis publishers, Florida, USA. pp 271-318

95

Greenwood, N.N., and Earnshaw, A. (1984). Uses of orthophosphates. In, Chemistry
of the elements. Pergamon press, Oxford, UK. Pp 603-609.

Grim, R.E. (1995). Clay mineralogy and double layer in clays. In, clay mineralogy.
Read books. Warwickshire, U.K., Pp 19-37.

Hammer, M.J., and Hammer, Jr. (2001). Water and wastewater technology, 4
th

Edition. Prentice Hall, New Jersey, USA.

Helcky, R.E., and Kilham, P. (1988). Nutrient limitation of phytoplankton
infreshwater and marine environments: a review of recent evidence on the
effects of enrichment. Limnology and oceanography, 33: 796-822.

Hiemstra, T., and Van Riemsdijk, W.H. (1996). A surface structural approach to ion
adsorption: The charge distribution (CD) model. Journal of colloid and
interface science, 179: 488-508

Hiemstra, T., and Van Riemsdijk, W.H. (1999). Surface structural ion adsorption
modelling of competitive binding of oxyanions by metal (hydr)oxides. Journal
of Colloid and Interface Science 210: 182-193

Hiemstra, T., Rahnemaie, R., and Van Riemsdijk, W.H. (2004). Surface complexation
of carbonate on goethite: IR spectroscopy, structure and charge distribution.
Journal of Colloid and Interface Science, 278: 282290

96

Hiemstra, T., and Van Riemsdijk, W.H. (2006). On the relationship between charge
distribution, surface hydration, and the structure of the interface of metal
hydroxides. Journal of Colloid and Interface Science, 301: 1-18

Horanyi, G., and Joo, P. (2002). Some peculiarities in the specific adsorption of
phosphate ions on hematite and -Al
2
O
3
as reflected by radiotracer studies.
Journal of Colloid and Interface Science, 247: 12-17

Huang, W., Wang, S., Zhu, Z., Li, L., Yao, X., Rudolph, V., and Haghseresht, F.
(2008). Phosphate removal from wastewater using red mud. Journal of
hazardous materials, 158: 35-42

Johansson, L., and Gustafsson, J.P. (2000). Phosphate removal using blast furnace
slags and opoka-mechanisms. Water research, 34(1): 259-265

Joko, I. (1984). Phosphorus removal from wastewater by the crystallization method.
Water science and technology, 17: 121-132

Kapolos, J, and Koutsoukos, P.G. (1999). Formation of calcium phosphates in
aqueous solutions in the presence of carbonate ions. Langmuir, 15: 6557
6562.

Karageorgiou, K., Paschalis, M., and Anastassakis, G.N. (2007). Removal of
phosphate species from solution by adsorption onto calcite used as natural
adsorbent. Journal of hazardous materials A, 139: 447-452
97

Koopal, L.K. (1996). Ion adsorption on mineral oxide surfaces. In, Adsorption on new
and modified inorganic sorbents in studies in surface science and catalysis,
vol. 99. Elselvier, Amsterdam. pp 757-796.

Kostura, B., Kulveitova, H., and Lesko, J. (2005). Blast furnace slags as sorbents of
phosphate from water solutions. Water research, 39: 1795-1802

Koutsoukos, P., Amjad, Z., Tomson, M.B., and Nancollas, G.H. (1980).
Crystallization of calcium phosphates. A constant composition study. Journal
of the American chemical society, 102: 1553-1557.

Kraepiel, A.M.L., Keller, K., and Morel, F.M.M. (1998). On the acid-base chemistry
of permanently charged minerals. Environmental science and technology, 32:
2829-2838

Kubicki, J.D., Kwon, K.D., Paul, K.W., and Sparks, D.L. (2007). Surface complex
structures modelled with quantum chemical calculations: carbonate,
phosphate, sulphate, arsenate and arsenite. European journal of soil science,
58: 932-944

Kulaev, I.S., Vagabov, V.M., and Kulakovskaya, T.V. (2004). The chemical
structures and properties of condensed inorganic phosphates. In, The
biochemistry of inorganic polyphosphates. John Wiley and sons Ltd., Sussex,
England. Pp 3-14

98

Kwanjana, E.N. (2003). Evaluation of sewage effluent use in horticulture: Zomba
municipality wastewater treatment plant, Malawi. Msc thesis University of
Malawi.

Li, Y., Liu, C., Luan, Z., Peng, X., Zhu, C., Chen, Z., Zhang, Z., Fan, J., and Jia, Z.
(2006). Phosphate removal from aqueous solutions using raw and activated red
mud and flyash. Journal of hazardous materials B, 137: 374-383

Malawi Government (2002). State of Environment report for Malawi 2002. Ministry
of Natural resources and environmental affairs, Lilongwe Malawi. Pp 1-291

Manning, B.A., and Goldberg, S. (1996). Modelling arsenate competitive adsorption
on kaolinite, montmorillonite and illite. Clays and clay minerals, 44(5): 609-
623

McKelvie, I.D. (2005). Separation, preconcentration and speciation of organic
phosphorus in environmental samples. In, Organic phosphorus in the
environment. CAB international, Oxfordshire, UK. Pp 1-20.

Mead, J.A. (1981). A comparison of the Langmuir, Freundlich and Temkin equations
to describe phosphate adsorption properties of soils. Australian journal of soil
research, 19: 333-342

99

Milonji, S.K. (2007). A consideration of the correct calculation of the
thermodynamic parameters of adsorption. Journal of the Serbian chemical
society, 72(12): 1363-1367

Moon, Y.H., Kim, J.G., Ahn, J.S., Lee, G.G., and Moon, H. (2007). Phosphate
removal using sludge from fullers earth production. Journal of hazardous
materials, 143: 41-48

Morse, G.K., Brett, S.W., Guy, J.A., and Lester, J.N. (1998). Review: Phosphorus
removal and recovery technologies. The Science of the total environment, 212:
69-81

Ozacar, M. (2006). Contact time optimization of two-stage batch adsorber design
using second-order kinetic model for the adsorption of phosphate onto alunite.
Journal of hazardous materials B, 137: 218-225

Ozacar, M., and Sengil, I.A. (2003). Adsorption of reactive dyes on calcined alunite
from aqueous solutions. Journal of hazardous materials B, 98: 211-224.

Persson, P., Nilsson, N., and Sjoberg, S. (1996). Structure and bonding of
orthophosphate ions at the iron oxide-aqueous interface. Journal of Colloid
and Interface Science, 177: 263-275

100

Plant, L.J., and House, W.A. (2002). Precipitation of calcite in the presence of
inorganic phosphate. Colloids and surfaces A: physicochemical and
engineering aspects, 203: 143-153

Pradhan, J., Das, J., Das, S., and Thakur, R.S. (1998). Adsorption of phosphate from
aqueous solution using activated red mud. Journal of colloid and interface
science, 204: 169-172

Rahnemaie, R., Hiemstra, T., and Van Riemsdijk, W.H. (2006). Inner- and outer-
sphere complexation of ions at the goethitesolution interface. Journal of
colloid and interface science, 297: 379-388

Rahnemaie, R., Hiemstra, T., and van Riemsdijk, W.H. (2007). Carbonate adsorption
on goethite in competition with phosphate. Journal of colloid and interface
science, 315: 415-425

Rietra, R.P.J.J., Hiemstra, T., and Van Riemsdijk, W.H. (2001). Interaction between
Calcium and Phosphate Adsorption on Goethite. Environmental science and
technology, 35 (16): 3369-3374.

Rosenqvist, J. (2002). Surface chemistry of Al and Si (hydr)oxides, with emphasis on
nano-sized gibbsite (-Al(OH)
3
). Doctoral thesis submitted to the faculty of
science and technology, Umea University, Sweden [online]. Available:
http://umu.diva-portal.org/smash/get/diva2:143930/FULLTEXT01 [11/09/08].

101

Sajidu, S.M.I., Masamba. W.R.L., Henry, E.M.T., and Kuyeli, S.M. (2007). Water
quality assessment in streams and wastewater treatment plants of Blantyre,
Malawi. Physics and chemistry of the earth, 32:1391-1398

Schulze, D.G. (2002). An introduction to soil mineralogy. In, Soil mineralogy with
environmental applications. Soil science society of America, Wisconsin, USA.
Pp 1-35

Scholler, M. (undated). Phosphate crystallisation processes. DHV water BV [Online].
Available: http://www.ceep-phosphates.org/Files/Document/70/p031-035.pdf,
[1/03/09].

Scroth, B.K., and Sposito, G. (1997). Surface charge properties of kaolinite. Clays
and clay minerals, 45 (1): 85-91

Skulberg, O.M., Codd, G.A., and Carmichael, W.W. (1984). Toxic blue-green algal
blooms in Europue: a growing problem. Ambio, 13: 244-247

Sidat, M., Kasan, H.C., and Bux, F. (1999). Laboratory-scale investigation of
biological phosphate removal from municipal wastewater. Water SA, 25(4):
459-462

Song, Y., Hahn, H.H., and Hoffmann, E. (2002a). Effects of solution conditions on
the precipitation of phosphate for recovery: A thermodynamic evaluation.
Chemosphere, 48: 1029-1034
102

Song, Y., Hahn, H.H., and Hoffmann, E. (2002b). The effect of carbonate on the
precipitation of calcium phosphate. Environmental Technology, 23(2): 207-
215

Smith, V.H., Tilman, J.C., and Nekola, J.C. (1999). Eutrophication: impacts of excess
nutrient inputs on freshwater, marine, and terrestrial ecosystems.
Environmental pollution, 100: 179-196

Stachowicz, M., Hiemstra, T., and van Riemsdijk, W.H. (2008). Multi-competitive
interaction of As(III) and As(V) oxyanions with Ca
2+
, Mg
2+
, PO
4
3-
, and CO
3
2-

ions on goethite. Journal of colloid and interface science, 320: 400-414

Tombacz, E., and Szekeres, M. (2006). Surface charge heterogeneity of kaolinite in
aqueous suspension in comparison with montmorillonite. Applied clay science,
34: 105-124

United Nations Environmental Programme (UNEP) (2007). Global environmental
outlook. United Nations environmental programme [Online]. Available:
http://www.unep.org/geo/geo4/media [01/03/09].

Van den Brandit, H.M.P., Smit, H.P. (1998). Mineral accounting: the way to combat
Eutrophication and to achieve the drinking water objective. Environmental
pollution, 102: 705-709

103

Van Kemenade, M.J.J.M., and De Bruyn, P.L. (1987). A kinetic study of precipitation
from supersaturated calcium phosphate solutions. Journal of colloid and
interface science, 118: 564-585

Van Olphen, H. (1963). Kaolinite. In, An introduction to clay colloidal chemistry.
Interscience, New York. Pp 154-187

White, W.M. (2007). Aquatic chemistry. In, Geochemistry [online textbook].
Available: http://www.imwa.info/geochemistry/ [20/10/08].

Wieland, E., and Stumm, W. (1992). Dissolution kinetics of kaolinite in acidic
aqueous solutions at 25 C. Geochimica et Cosmochimica Acta, 56: 3339-3355

Yager, T.R. (2006). The Mineral industry of Malawi. 2006 minerals year book:
Malawi. U.S. Geological survey [online]. Available:
http://minerals.usgs.gov/minerals/pubs/country/2006/myb3-2006-mi.pdf
[20/02/09].

Yang, C. (1998). Statistical mechanical study on the Freundlich isotherm equation.
Journal of colloid and interface science, 208: 379-387

Zheng, L., Xiaomei, L., and Liu, J. (2004). Adsorptive removal of phosphate from
aqueous solutions using iron oxide tailings. Water research, 38: 1381-1326

104

Zhong, B., Stanforth, R., Wu, S., and Chen, J.P. (2007). Proton interaction in
phosphate adsorption onto goethite. Journal of Colloid and Interface Science
308: 40-48