SEPARATION OF LITHIUM FROM BRINES

A THESIS SUBMITTED TO
THE GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES
OF
MIDDLE EAST TECHNICAL UNIVERSITY

BY
BARIŞ ERDOĞAN

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

FOR
THE DEGREE OF DOCTOR OF PHILOSOPHY
IN
CHEMICAL ENGINEERING

FEBRUARY 2015
 Approval of the thesis:

SEPARATION OF LITHIUM FROM BRINES

submitted by BARIŞ ERDOĞAN in partial fulfillment of the requirements for the

degree of Doctor of Philosophy in Chemical Engineering Department, Middle
East Technical University by,

Prof. Dr. Gülbin Dural Ünver

Dean, Graduate School of Natural and Applied Sciences

Prof. Dr. Halil Kalıpçılar

Head of Department, Chemical Engineering

Prof. Dr. H. Önder Özbelge

Supervisor, Chemical Engineering Dept., METU

Assoc. Prof. Dr. Yusuf Uludağ

Co-supervisor, Chemical Engineering Dept., METU

Examining Committee Members:

Assoc. Prof. Dr. Naime Aslı Sezgi

Chemical Engineering Dept., METU

Prof. Dr. H. Önder Özbelge

Chemical Engineering Dept., METU

Prof. Dr. Halil Kalıpçılar

Chemical Engineering Dept., METU

Prof. Dr. Niyazi Bıçak

Chemistry Dept., ITU

Assoc. Prof. Dr. Çerağ Dilek Hacıhabiboğlu

Chemical Engineering Dept., METU

Date: 06.02.2015
I hereby declare that all information in this document has been obtained and
presented in accordance with academic rules and ethical conduct. I also declare
that, as required by these rules and conduct, I have fully cited and referenced
all material and results that are not original to this work.

Name, Last name : Barış Erdoğan

Signature :

iv
 ABSTRACT

SEPARATION OF LITHIUM FROM BRINES

Erdoğan, Barış

Ph.D., Department of Chemical Engineering

Supervisor: Prof. Dr. H. Önder Özbelge

Co-supervisor: Assoc. Prof. Dr. Yusuf Uludağ

February 2015, 184 pages

Lithium is one of the valuable and promising metals having extensive potential use
in long lasting rechargeable batteries in automotive and electronics industry.
Regarding scarcity of world lithium reserves, the most efficient method must be
utilized in lithium separation and purification process. The scope of this study is to
increase the efficiency of lithium separation by liquid-liquid extraction and
adsorption methods with various materials synthesized throughout the study.

In liquid-liquid extraction method, N-alkyl formamides (hexyl formamide, octyl

formamide, dibutyl formamide, dihexyl formamide) were synthesized and lithium
separation performances were evaluated. Hexyl formamide was found to have a
distribution coefficient of 0.14 for LiCl in dilute concentrations at 25 oC which is the
largest value cited in the literature for a single solvent so far.
v
In adsorption method, lithium manganese oxides (LiMnO) were synthesized with
solid-solid and hydrothermal reaction procedures under different temperature and
Li/Mn ratios. The performance of the adsorbent with the highest capacity value was
characterized at different pH, lithium and foreign ion (Na, K, Mg) concentrations. It
was found out that, LiMnO adsorbent has a capacity value of 22.8 mg Li/g adsorbent
at pH:10.2 in dilute lithium chloride solutions with a very high selectivity towards
foreign ions such as sodium, potassium and magnesium.

In order to employ LiMnO adsorbents in an adsorption column, micron sized

particles were impregnated into millimeter sized beads made of poly(styrene-maleic
anhydride-glycidyl methacrylate) (PSMA) by precipitation method. PSMA has been
utilized in bead formation process for the first time in this study and gave superior
results compared to previous analogs in terms of its swelling and self-crosslinking
ability. The column performance filled with PSMA-LiMnO beads were evaluated
with artificial brine, concentrated brine taken from Çamaltı Salina (İzmir) and boron
clay extract taken from Bigadiç. The lithium concentration in brines was increased
from 0.43 ppm to 1675.6 ppm while reducing the foreign ion concentrations
significantly in adsorption column. The post treatment of the column product with
LiMnO resulted in a solution with 3200 ppm Li. That concentration is above the
required minimum concentration for Li2CO3 production, which is the raw material
for lithium battery industry.

The adsorption method, utilizing new PSMA-LiMnO beads in a column was found
as a promising candidate for practical use in conventional separation of lithium from
brines.

Keywords: Lithium separation, brines, N-alkyl formamide, lithium manganese

oxides, organic metal oxide support

vi
 ÖZ

TUZLU SULARDAN LİTYUM AYRIŞTIRILMASI

Erdoğan, Barış

Doktora, Kimya Mühendisliği Bölümü

Tez Yöneticisi: Prof. Dr. H. Önder Özbelge

Ortak Tez Yöneticisi: Doç. Dr. Yusuf Uludağ

Şubat 2015, 184 sayfa

Lityum, otomotiv ve elektonik endüstrisi içerisinde uzun ömürlü, şarj edilebilir

pillerde oldukça fazla kullanım alanı bulan değerli ve gelecek vaadeden bir metaldir.
Dünya lityum reservlerinin kısıtlı olması, lityum ayrıştırma ve saflaştırma
işlemlerinde mümkün olan en verimli yöntemin kullanılmasını gerektirmektedir. Bu
çalışmanın amacı, sentezlenen çeşitli malzemeler kullanılarak sıvı-sıvı özütlemesi ve
adsorpsiyon yöntemleri ile lityum ayrıştırma işlemlerinin verimini artırmaktır.

Sıvı-sıvı özütlemesi yönteminde N-alkil formamitler (hekzil formamit, oktil

formamit, dibutil formamit, dihekzil formamit) sentezlenmiş ve lityum ayrıştırma
performansları hesaplanmıştır. Hekzil formamit, seyreltik LiCl çözeltileri için 25
o
C’de 0.14 değeri ile literatürde tek bir çözücü için bahsedilen en yüksek dağılım
katsayısını vermiştir.
vii
Adsorpsiyon yönteminde, lityum mangan oksitler (LiMnO) katı-katı ve hidrotermal
reaksiyon yöntemleri ile farklı sıcaklık ve Li/Mn mol oranlarında sentezlenmiştir. En
yüksek kapasite değerini veren adsorbentin performansı farklı pH, lityum ve rakip
iyon (Na, K, Mg) derişimlerinde karakterize edilmiştir. LiMnO adsorbentlerin
kapasite değeri seyreltik lityum klorür çözeltileri için pH:10.2’de ve sodyum,
potasyum, magnezyum gibi rakip iyonlar varlığında, lityuma karşı yüksek selektivite
ile 22.8 mg Li/g olarak bulunmuştur.

Lityum mangan oksit parçaçıkları poli(stiren-maleik anhidrit-glisidil metakrilat)

(PSMA) içine emdirilerek mikron boyutundan milimetre boyutunda boncuklara
çevrilmiş ve adsorpsiyon kolonunda kullanılmıştır. PSMA ilk defa bu çalışmada
boncuk sentezinde kullanılmış, şişme ve kendi kendine çapraz bağlanma özellikleri
ile önceki benzerlerinden üstün sonuçlar vermiştir. Yapay tuzlu su, Çamaltı
konsantre deniz suyu ve Bigadiç bor kili özütü ile yapılan deneyler sonucunda
PSMA-LiMnO boncukları ile doldurulmuş kolonun performansı ölçülmüştür. Tuzlu
sular içerisinde bulunan lityum konsantrasyonu 0.43 ppm’den 1675.6 ppm’e kadar
çıkartılmış ve rakip iyonların konsantrasyonu kayda değer biçimde düşürülmüştür.
Kolon ürünü çözelti LiMnO ile direk muamele edildiğinde 3200 ppm Li derişimine
ulaşılmıştır. Bu değer lityum pil endüstrisinde hammadde olarak kullanılan Li2CO3
üretimi için gerekli minimum konsatrasyon değerinin üzerindedir.

PSMA-LiMnO boncukları kullanılarak uygulanan adsorpsiyon tekniği, tuzlu

sulardan endüstriyel lityum ayrıştırma yöntemleri arasına girmeye aday, umut
vaadeden bir metot olarak değerlendirilmiştir.

Anahtar kelimeler: Lityum ayrıştırma, tuzlu sular, N-alkil formamit, lityum mangan
oksit, organik metal oksit destek

viii
To my grandmother

ix
 ACKNOWLEDGEMENTS

I wish to express my sincere gratitude to my supervisor Prof. Dr. H. Önder Özbelge for his
support, guidance and help, in all the possible way, throughout my study. Without his
encouragements and practical solutions to problems, I would not have achieved this work
with success. I would also like to express my special thanks to my co-supervisor Assoc.
Prof. Dr. Yusuf Uludağ for his helpful comments and suggestions. I also wish to thank to
Prof. Dr. Niyazi Bıçak for his great leadership in synthesis reactions held on throughout the
study.

I would like to especially thank to Bilal Bayram for his advices and contributions to see the
whole picture during the study. I would also thank to Bilsen Aytekin for her cooperation
during the laboratory studies.

Heartfelt thanks to my colleagues who helped me, opened my mind and shared the hard
times, Emine Kayahan, Hasan Zerze, Saltuk Pirgalıoğlu, Eda Açık, Merve Akkuş, Canan
Yeniova, Özge Mercan, Sevler Gökçe Avcıoğlu.

I should like to acknowledge Faculty Members of Chemical Engineering Department during

my studies at METU, Kerime Güney for her great help in AAS analysis, Turgut Aksakal in
finding all possible chemicals and equipments located in the departmant, Ertuğrul
Özdemir in building the vapor pressure measurement set-up, Yavuz Güngör in XRD
measurements, Gülten Orakçı in GC and HPLC measurements, Gülüstan Görmez
and Naciye Kaya for solving bureaucratic problems throughout the study.

I wish to thank to my parents, Nevin and Mustafa Erdoğan, my sister Nihan

Erdoğan, and my uncle Raşit Ülkü for their love and encouragement. I would like to
thank to my grandmother Nermin Ülkü for her great wisdom of life and endless love
towards me.

x
And lastly, I would like to thank to my dear friends who are always besides me through the
study, Şaziye, Uğur, Barış, Esen, Eylem, Ozan, Ceren, Soner, Bilgen, Başar, Burcu, Mert,
Dominic, Ilgın, Pelin, Burak, Şafak, Aslı, Ulaş Barış, Onur, Ufuk, Salih, Aytaç.

The financial support provided by TÜBİTAK through the projects 109M499 and 113M919
is gratefully acknowledged.

xi
 TABLE OF CONTENTS

ABSTRACT ................................................................................................................. v
ÖZ… .......................................................................................................................... vii
ACKNOWLEDGEMENTS ......................................................................................... x
TABLE OF CONTENTS ........................................................................................... xii
LIST OF TABLES ..................................................................................................... xv
LIST OF FIGURES .................................................................................................. xix
LIST OF SYMBOLS ............................................................................................... xxii
CHAPTERS

1 INTRODUCTION ................................................................................................ 1
1.1 The usage areas of lithium and its compounds ............................................. 1
1.1.1 Ceramics and glass ................................................................................. 2
1.1.2 Electrochemical applications ................................................................. 2
1.1.3 Lubricating oils ...................................................................................... 3
1.1.4 Ventilation systems and air conditions .................................................. 3
1.1.5 Metallurgical use .................................................................................... 3
1.1.6 Organometallic compounds ................................................................... 3
1.1.7 Nuclear reactors ..................................................................................... 3
1.1.8 Medicines ............................................................................................... 4
1.1.9 Hydrogen storage ................................................................................... 4
1.2 Lithium Sources ............................................................................................ 4
1.2.1 Minerals ................................................................................................. 5
1.2.2 Clays ...................................................................................................... 6
1.2.3 Brines ..................................................................................................... 6
1.3 Conventional lithium separation process from minerals and clays ............... 7
1.4 Conventional lithium separation processes from brines ............................... 8
xii
 1.4.1 Clayton Valley, Nevada ......................................................................... 9
1.4.2 Salar de Atacama, Chile ....................................................................... 11
1.4.3 Salar de Hombre Muerto, Argentina .................................................... 13
1.4.4 Searles Lake, California ....................................................................... 14
1.4.5 Lakes in Turkey.................................................................................... 15
1.5 Literature survey on lithium separation methods from brines..................... 15
1.5.1 Liquid-liquid extraction ....................................................................... 16
1.5.2 Adsorption ............................................................................................ 18
2 EXPERIMENTAL METHODS ......................................................................... 27
2.1 Materials ...................................................................................................... 27
2.2 Synthesis and experimental procedures ...................................................... 27
2.2.1 Synthesis of N-alkyl formamides ......................................................... 27
2.2.2 Synthesis of lithium manganese oxide adsorbents ............................... 28
2.2.3 Synthesis of lithium manganese oxide precursors ............................... 30
2.2.4 Synthesis of poly (styrene-maleic anhydride-glycidyl methacrylate)
(PSMA) [90] ....................................................................................................... 30
2.2.5 Liquid-liquid extraction procedure ...................................................... 31
2.2.6 Polymer enhanced ultrafiltration procedure ......................................... 31
2.2.7 Batch adsorption procedure ................................................................. 32
2.2.8 LiMnO bead preparation ...................................................................... 33
2.2.9 Column experiment procedure ............................................................. 34
2.3 Equipment ................................................................................................... 35
2.3.1 Measurement of metal ion concentration in aqueous solutions ........... 35
2.3.2 Measurement of organic content in aqueous solutions ........................ 36
2.3.3 Spectroscopic methods ......................................................................... 36
2.3.4 Thermal analysis .................................................................................. 36
2.3.5 Bead characterization ........................................................................... 37
2.3.6 Physical properties of the synthesized formamide ............................... 37
3 RESULTS AND DISCUSSIONS ....................................................................... 39
3.1 Synthesis and characterization of materials employed in lithium
separation… ............................................................................................................ 39
3.1.1 Synthesis and characterization of N-alkyl formamides ........................... 39
3.1.2 Synthesis and characterization of lithium manganese oxide................ 44
xiii
 3.1.3 Synthesis and characterization of lithium manganese oxide-polymer
beads….. ............................................................................................................. 64
3.1.4 Acquisition of lithium containing brines from nature.......................... 75
3.2 Liquid-liquid extraction by N-alkyl formamides ........................................ 78
3.3 Adsorption via lithium manganese oxide adsorbents .................................. 87
3.3.1 Effect of pH on adsorption ................................................................... 87
3.3.2 Effect of initial lithium concentration on adsorption ........................... 88
3.3.3 Kinetics of the adsorption process ....................................................... 89
3.3.4 Equilibrium Isotherms ......................................................................... 90
3.3.5 Stability experiments ........................................................................... 93
3.3.6 Selectivity experiments ........................................................................ 94
3.4 Separation of lithium by PSMA-LiMnO beads in an adsorption column ... 95
3.4.1 The breakthrough plots for solutions containing lithium chloride
only…… ............................................................................................................. 95
3.4.2 The breakthrough plots for solutions containing foreign ions ........... 102
3.4.3 The breakthrough plots for the artificial and real brine solutions ...... 106
3.4.4 Recovery of lithium and regeneration of the beads ........................... 110
3.4.5 Separation of lithium by a post-treatment with LiMnO particles
following column adsorption ............................................................................ 125
4 CONCLUSIONS .............................................................................................. 129
RECOMMENDATIONS ......................................................................................... 133
REFERENCES ........................................................................................................ 135
APPENDICES

A. ALTERNATIVE MATERIALS EMPLOYED IN LITHIUM

SEPARATION….. ................................................................................................... 143
B. XRD MEASUREMENTS OF SYNTHESIZED LITHIUM MANGANESE
OXIDES ................................................................................................................... 165
C. SEM PHOTOGRAPHS OF SYNTHESIZED PSMA-LiMnO BEADS .......... 173
D. THE FEASIBILITY STUDY TO RECOVER LITHIUM CARBONATE FROM
ÇAMALTI SALINA AND BORON CLAY EXTRACT ........................................ 175

CURRICULUM VITAE .......................................................................................... 183

xiv
 LIST OF TABLES

TABLES

Table 1.1 Areas of usage and consumption ratios for lithium compounds .................. 2
Table 1.2 Lithium production and reserves.................................................................. 5
Table 1.3 Commercially valuable lithium minerals ..................................................... 6
Table 1.4 Commercially important brines ................................................................... 7
Table 1.5 Clayton Valley, Brine Composition ............................................................. 9
Table 1.6 Salar de Atacama, Brine Composition ....................................................... 11
Table 1.7 Salar de Hombre Muerto, Brine Compositon ............................................ 13
Table 1.8 Searles Lake, Brine Composition............................................................... 14
Table 1.9 The compositon of lakes in Turkey............................................................ 15
Table 1.10 Lithium chloride solubility in different alcohols ..................................... 16
Table 1.11 The distribution coefficients of various solvents for lithium, sodium,
potassium and calcium ions [21, p. 137] .................................................................... 17
Table 3.1 Surface tension values for the N-alkyl formamides at 10 oC ..................... 42
Table 3.2 The densities of the N-alkyl formamides at different temperatures........... 43
Table 3.3 LiCl, NaCl and KCl solubilities in hexyl formamide ................................ 44
Table 3.4 Methods employed in synthesis of lithium manganese oxide adsorbents. 45
Table 3.5 Optimization plan constructed by applying factorial design for the upper
and lower limits of four variables versus acquired adsorbent capacity (response) .... 47
Table 3.6 Effect of oven temperature on capacity of adsorbents synthesized by
Hydrothermal-1 procedure ......................................................................................... 48
Table 3.7 Effect of gel/water ratio on capacity of adsorbents synthesized by
Hydrothermal-1 procedure ......................................................................................... 49
Table 3.8 Effect of Li/Mn ratio on capacity of adsorbents synthesized by
Hydrothermal-1 procedure ......................................................................................... 49

xv
Table 3.9 Effect of autoclave temperature on capacity of adsorbents synthesized by
Hydrothermal-1 procedure ......................................................................................... 50
Table 3.10 Finding crystal system and lattice constant of LiMnO ............................ 63
Table 3.11 Different solvent-polymer couples utilized for LiMnO bead synthesis by
precipitation method .................................................................................................. 65
Table 3.12 PSMA beads synthesized with different GMA content at 70 oC (Benzyl
peroxide molar ratio:0.01, monomer/solvent ratio: 0.33) .......................................... 67
Table 3.13 The amount of the components used in bead synthesis ........................... 71
Table 3.14 BET results of PSMA-LiMnO beads ....................................................... 75
Table 3.15 The composition of the concentrated brine in Çamaltı Salina ................. 76
Table 3.16 Leaching of lithium ions from boron clay ............................................... 77
Table 3.17 The composition of the clay extract taken from Bigadiç ......................... 77
Table 3.18 The effect of pH on distribution coefficient of LiCl for the formamides at
25 oC, 10 ml water/10 ml formamide ratio (The initial LiCl concentration was 1.4 M)
.................................................................................................................................... 79
Table 3.19 The effect of temperature on distribution coefficient values for different
formamides at pH:4, 10 ml water/10 ml formamide ratio, initial lithium
concentration of 1.4 M Li. ......................................................................................... 79
Table 3.20 ΔHo and ΔSo values calculated from Van’t Hoff plots ............................ 81
Table 3.21 Equilibrium data between extract and raffinate phase for the formamide-
water-lithium chloride system at 25 oC, pH: 4 and 10 ml water/10 ml hexyl
formamide .................................................................................................................. 82
Table 3.22 Artificial concentrated seawater composition and the resulting
distribution coefficient values in selectivity experiments. ......................................... 84
Table 3.23 The selectivity values of hexyl formamide for lithium near foreign (Na,
K, Mg, Ca) ions in sea water...................................................................................... 85
Table 3.24 Composition of the concentrated sea water taken from Çamaltı Salina and
resulting distribution coefficients (25 oC, 10 ml HeFo/10 ml water, pH:7.4) ........... 85
Table 3.25 The effect of pH on recovery of lithium from hexyl formamide ............. 86
Table 3.26 The effect of pH on recovery of lithium from hexyl formamide (Second
recovery) .................................................................................................................... 86

xvi
Table 3.27 Capacity values of adsorbents at different pHs (Total volume: 100 ml,
Adsorbent amount: 0.02 g) ......................................................................................... 88
Table 3.28 Capacity value of the adsorbents at different initial lithium concentrations
(pH:10.2 and adsorbent amount: 0.02 g ) ................................................................... 88
Table 3.29 The stability of lithium manganese oxide particles in lithium recovery .. 93
Table 3.30 Selectivity of lithium manganese oxides with respect to Na, K and Mg. 94
Table 3.31 The column operating conditions for the pH dependency of adsorption
process ........................................................................................................................ 96
Table 3.32 The column operating conditions for different initial lithium
concentration in adsorption process ........................................................................... 97
Table 3.33 The column operating conditions for different flow rates in adsorption
process ........................................................................................................................ 99
Table 3.34 The column operating conditions for different bead sizes in adsorption
process ...................................................................................................................... 101
Table 3.35 The column operating conditions for feed solutions containing NaCl .. 102
Table 3.36 The column operating conditions for feed solutions containing KCl .... 104
Table 3.37 The column operating conditions for feed solutions containing MgCl . 105
Table 3.38 The column operating conditions for artificial brine solutions .............. 106
Table 3.39 The column operating conditions for concentrated brine solution taken
from Çamaltı Salina ................................................................................................. 108
Table 3.40 The column operating conditions for boron clay extract taken form
Bigadiç ..................................................................................................................... 109
Table 3.41 The initial and final concentrations in the feed tank before and after the
operation ................................................................................................................... 109
Table 3.42 Elution process data of the column after Run 3. .................................... 111
Table 3.43 The elution of the column saturated with artificial brine solution and
swept by tap water .................................................................................................... 114
Table 3.44 The comparison of the artificial brine and elution solution composition
after sweeping by tap water...................................................................................... 114
Table 3.45 The second adsorption-sweeping-elution process composition with the
feed solution having concentration of elution process in Run 18 ............................ 115

xvii
Table 3.46 The elution of the column saturated with artificial brine solution and
swept by acidic water at pH 6.1. .............................................................................. 116
Table 3.47 The comparison of the artificial brine and elution solution composition
after sweeping by acidic solution (pH 6.1) .............................................................. 117
Table 3.48 The second adsorption-sweeping(at pH:6.1)-elution process composition
with the feed solution having concentration of elution process in Run 26 .............. 117
Table 3.49 The elution of the column saturated with concentrated brine solution and
swept by demineralized water .................................................................................. 118
Table 3.50 The comparison of the real concentrated brine and elution solution
composition after sweeping by deionized water ...................................................... 119
Table 3.51 The elution of the column saturated with concentrated brine solution and
swept by acidic water at pH 6.1. .............................................................................. 120
Table 3.52 The comparison of the real concentrated brine and elution solution
composition after sweeping by acidic water at pH 6.1 ............................................ 120
Table 3.53 The elution of the column saturated with clay extract and swept by tap
water ......................................................................................................................... 122
Table 3.54 The comparison of the boron clay extract and elution solution
composition after sweeping with tap water ............................................................. 123
Table 3.55 The elution of the column saturated with clay extract and swept by acidic
water at pH 6.1 ......................................................................................................... 123
Table 3.56 The comparison of the boron clay extract and elution solution
composition after sweeping with acidic water ......................................................... 124
Table 3.57 The initial and final concentrations of the adsorption process done with
column effluent ........................................................................................................ 125
Table 3.58 The amount of eluted ions and concentrations in elution process of post-
treatment .................................................................................................................. 126
Table 3.59 Li concentration, Li/foreign ion ratio, Li loss in each separation step
starting from artificial brine ..................................................................................... 127
Table 3.60 Li concentration, Li/foreign ion ratio, Li loss in each separation step
starting from clay extract ......................................................................................... 128

xviii
 LIST OF FIGURES

FIGURES

Figure 1.1 The process scheme of Clayton Valley .................................................... 10

Figure 1.2 Production scheme of LiCO3 in Salar de Atacama ................................... 12
Figure 1.3 Lithium insertion/extraction mechanisms (a) Lithium extraction (b)
Lithium insertion O: redox side; □: ion exchange side .............................................. 21
Figure 1.4 Breakthrough curve for 150 days operation ............................................. 23
Figure 2.1 Continuous ultrafiltration set-up [92] ....................................................... 32
Figure 2.2 Scheme of the set-up for column experiments ......................................... 34
Figure 2.3 Photos of the set-up for column experiments ........................................... 35
Figure 2.4 Vapor pressure measurement set-up ......................................................... 37
Figure 3.1 Synthesis pathway of N-alkyl formamides ............................................... 40
Figure 3.2 Temperature dependence of viscosity for N-alkyl formamides ............... 42
Figure 3.3 Temperature-vapor pressure relation for hexyl formamide ...................... 43
Figure 3.4 Hydrothermal-1 procedure in ten steps..................................................... 46
Figure 3.5 Heating curve for MnO and LiOHH2O reaction system .......................... 52
Figure 3.6 Heating curve for MnO2 and LiOHH2O reaction system ......................... 52
Figure 3.7 Heating curve for MnCO3 and LiOHH2O reaction system ...................... 53
Figure 3.8 Heating curve for Mn3O4 and LiOHH2O reaction system ........................ 54
Figure 3.9 FTIR spectrum of spinel lithium manganese oxide synthesized by MnCO3
and LiOHH2O (Li/Mn: 1/1) F30, before (LMO) and after (HMO) acid leaching ..... 55
Figure 3.10 FTIR spectrum of spinel lithium manganese oxide synthesized by
MnCO3 and LiOHH2O (Li/Mn: 1/0.8 mole ratio) F31, before (LMO) and after
(HMO) acid leaching.................................................................................................. 56
Figure 3.11 FTIR spectrum of spinel lithium manganese oxide synthesized via
MnCO3 and LiOHH2O (Li/Mn: 0.8/1 mole ratio) F32, before (LMO) and after
(HMO) acid leaching.................................................................................................. 57

xix
Figure 3.12 FTIR spectra for three different adsorbents (after acid leaching) having
different Li/Mn molar ratios(F30-Li/Mn:1, F31-Li/Mn:1/0.8, F32-Li/Mn:0.8) ........ 58
Figure 3.13 Effect of acid leaching process on lattice proton at different times for
adsorbents synthesized by Hydrothermal-1 procedure .............................................. 59
Figure 3.14 The adsorbents synthesized with three different methods ...................... 59
Figure 3.15 XRD spectrum of the spinel lithium manganese oxide LMO (450 oC,
Li/Mn=1) .................................................................................................................... 61
Figure 3.16 XRD spectrum of the spinel lithium manganese oxide HMO (450 oC,
Li/Mn=1) .................................................................................................................... 61
Figure 3.17 Structure of poly maleic anhydride-styrene-glycidy methacrlate (PSMA)
.................................................................................................................................... 67
Figure 3.18 FTIR spectra of poly (styrene-maleic anhydrid-glycidyl methacrylate)
terpolymer (PSMA) ................................................................................................... 69
Figure 3.19 13C-NMR spectra of PSMA .................................................................... 70
Figure 3.20 SEM photographs for Bead T16 ............................................................. 72
Figure 3.21 SEM photographs for Bead T16 after acid leaching .............................. 72
Figure 3.22 SEM photographs for Bead T18 ............................................................. 73
Figure 3.23 SEM photographs for Bead T21 ............................................................. 73
Figure 3.24 SEM photographs for Bead T22 ............................................................. 74
Figure 3.25 Van’t Hoff Plots for hexyl formamide (a), octyl formamide (b), dibutyl
formamide (c)............................................................................................................. 80
Figure 3.26 Effect of initial LiCl concentration on k of hexyl formamide at pH:4.
The plots were drawn for three different water/formamide volume ratios (1:1, 1:2,
1:3) at 10 oC. .............................................................................................................. 81
Figure 3.27 The lithium concentration in extract (formamide) and raffinate (water)
phases at 25 oC and pH:4 for 10 ml water/10 ml hexyl formamide ........................... 83
Figure 3.28 The kinetics of adsorption process ......................................................... 89
Figure 3.29 The equilibrium concentration of lithium in solution vs uptake of lithium
ions ............................................................................................................................. 90
Figure 3.30 Langmuir isotherm curve for LiMnO at 25 oC. ...................................... 91
Figure 3.31 Freundlich isotherm curve for LiMnO at 25 oC. .................................... 92
Figure 3.32 The breakthrough curves at different pH values .................................... 96

xx
Figure 3.33 The breakthrough curves at different initial lithium concentrations. ..... 98
Figure 3.34 The breakthrough curves at different flow rates. .................................. 100
Figure 3.35 The breakthrough curves for different bead sizes................................. 101
Figure 3.36 The breakthrough curve for sodium chloride feed solution ................. 103
Figure 3.37 The breakthrough curve for potassium chloride feed solution ............. 104
Figure 3.38 The breakthrough curve for magnesium chloride feed solution ........... 105
Figure 3.39 The effluent lithium concentration in a column fed with artificially
prepared brine solution ............................................................................................. 107
Figure 3.40 The effluent lithium concentration in a column fed with concentrated
brine solution from Çamaltı Salina .......................................................................... 108
Figure 3.41 The elution curve after Run 3 ............................................................... 110
Figure 3.42 The change of ion concentrations in sweeping process applied after Run
12 .............................................................................................................................. 112
Figure 3.43 The molar ratio of potassium and magnesium with respect to lithium
during sweeping process .......................................................................................... 113
Figure 3.44 The change of ion concentrations in sweeping process applied after Run
14 .............................................................................................................................. 121

xxi
 LIST OF SYMBOLS

Calcite Calcium carbonate

Gypsum Calcium sulfate dihydrate
Slyvinite NaCl and KCl mixture
Glaserite Potassium sulfate
Soda ash Sodium carbonate
PEUF Polymer enhanced ultrafiltration
o
ΔH Standard enthalpy change
ΔSo Standard entropy change
Rg Universal gas constant
T Absolute temperature
k Distribution coefficient
HeFo Hexyl formamide
OcFo Octyl formamide
DbFo Dibutyl formamide
DhFo Dihexyl formamide
EMK Ethyl methyl ketone
AAS Atomic Absorbance Spectrophotometer
CF Cross Flow
UF Ultrafiltration
LiMnO Lithium manganese oxide adsorbent
GMA Glycidyl methacrylate
AIBN Azobisisobutyronitrile
NMP N-methyl pyyrolidone
PSMA Poly (styrene-maleic anhydride-glycidyl methacrylate)
FTIR Fourier Transform Infrared Spectroscopy

xxii
NMR Nuclear Magnetic Resonance
TGA Thermogravimetric Analysis
DTA Differential Thermal Analysis
SEM Scanning Elecron Microscopy
BET Brunauer–Emmett–Teller
LMO Lithium manganese oxides before acid leaching
HMO Lithium manganese oxides after acid leaching
CTAB Cethy tert butyl ammonium bromide
PolyMMA-co-BA Polymethyl methacrylate-co-butyl acrylate
PET polyethylene terephthalate
PE Polyethylene
PS Polystyrene
PS-Br Brominated poly styrene
THF Tetrahydrofuran
DMF Dimethylformamide
DMSO Dimethyl sulfoxide
PVC Polyvinylchloride
S Selectivity
Ce Lithium concentration in solution at equilibrium
Qe Uptake amount of lithium at equilibrium (mg/g)
Kads Emprical equilibrium constant (L/mg)
Qm Theoretical maximum adsorption capacity (mg/g)
MWCO Molecular weight cut-off
Fp Flux of permeate
Cf Feed concentration
Cp Permeate concentration
R Retention
PES Poly ether sulfone
C-PEG Comb like poly ethylene glycol
TLC Thin Liquid Chromatography

xxiii
 CHAPTER 1

INTRODUCTION

1 INTRODUCTION

Lithium is one of the most valuable and promising metals having extensive potential
use especially in design of new generation battery systems. The four main reasons
that make lithium metal attractive as a battery anode material are its light weight,
high voltage, high electrochemical equivalence, and good conductivity [1]. Demand
to this chemical is expected to grow in the future due to its application in long lasting
rechargeable batteries in automotive and electronics industry.

1.1 The usage areas of lithium and its compounds

The lithium consumption in the world first stepped up by the demand of US Atom
Energy Commission (AEC) in 1950 till 1960s. In the following years as the
technology level in the world evolved, new application areas for lithium emerge and
showed great variety from batteries to medicines today [2]. The areas of usage and
consumption ratios are given in Table 1.1 [3].

1
Table 1.1 Areas of usage and consumption ratios for lithium compounds

Area of usage Consumption ratio (%)

Ceramics and glass 31
Batteries 23
Lubricating oils 10
Ventilation systems and air conditions 5
Production of aluminum 3
Other 28

Areas of usage for lithium and its compounds can be grouped into nine sections in
detail.

1.1.1 Ceramics and glass

Small ionic radius resulting in a high ionic charge makes lithium a superb flux
material in production of ceramics. Lithium tetraborate is a good example of this
kind, it fastens the clay and glaze formations in the production of glass [4].

1.1.2 Electrochemical applications

The battery technology was based on cadmium and nickel until 1990’s. After that
lithium based batteries got in charge with their superior performance resulting from
their higher energy/weight ratio (energy density). Besides, lithium-ion batteries do
not have memory effect, and a slow loss of charge when not in use [5]–[7]. The
electronic devices spread away to all over the world, and energy requirements of
concerned developed technologies get increasing. One of the most important
electronic devices are hybrid or electrical cars which are presented as the vehicles of
the future recently. This increasing demand on energy will inevitably bring a very
important role to the lithium metal in the future [8].

2
1.1.3 Lubricating oils

Lithium is used in oils as a cation within the structure of the surfactants. Lithium
containing oils have lower freezing points, improved viscosity, surface tension and
water compatibility [9].

1.1.4 Ventilation systems and air conditions

The concentrated aqueous solution of lithium bromide has low vapor pressure so that
it can be used in the air conditioning systems. Lithium bromide solutions can be
replaced with chloroflorocarbon which is a harmful compound to the environment
[10].

1.1.5 Metallurgical use

Lithium forms alloys with numerous metals. Lead and aluminum alloys are the most
important ones. Lead alloys are used for bearings of railroad cars. Aluminum alloys
have been used in structural components of aircraft since 1957. Low weight and high
stability make the aluminum-lithium alloys unique for aircraft manufacturers [11].

1.1.6 Organometallic compounds

The production of organometallic compounds is one of the major usage areas of

lithium metal such as n-butyl lithium [12]. These compounds are formed by the
reaction of dispersions of lithium metal with organic halogens. N-butyllithium is also
used as an initiator in polymer synthesis [13] and widely used in agriculture and
pharmaceutical chemical synthesis [14].

1.1.7 Nuclear reactors

Lithium-6 is used as a source material in production of tritium which is a fuel in

fusion reactors [15]. Also it is utilized as an absorbent material for neutrons [16].
Besides, the low density, high heat capacity and high latent heat of evaporation
makes lithium one of the best cooling materials in nuclear reactors [17]. As the

3
nuclear technology evolves with the fusion reactors, the importance of lithium metal
will increase.

1.1.8 Medicines

Lithium compounds have been used in medicines since 1900’s. It is used in

treatment of bipolar disorder and for related diagnoses, such as schizoaffective and
cyclic major depression. The healing mechanism of the lithium ion in the body
cannot be understood completely [18].

1.1.9 Hydrogen storage

Hydrogen was stored in the form of lithium hydrate in World War 2. Lithium
hydrate can be reacted by seawater easily and it gives hydrogen gas as a product. In
the literature Ichikawa et al utilized lithium amide and lithium hydride in 1:1 ratio
and found a new promising material for the storage of hydrogen. These materials
absorb or desorb the hydrogen gas depending on the medium temperature [19]. In
another study, Xiong et al. produced a new material by using lithium and sodium
aminoborates which was published in “Nature Materials” [20]. These new generation
materials are promising in storage of hydrogen gas in future.

1.2 Lithium Sources

In nature, lithium cannot be found in pure form, rather it is found in complex form
most of the time with aluminum and silicates. The lithium reserves on earth crust are
found at low concentrations; between 7 and 60 ppm [21]. The lithium reserves and
production amount of lithium compounds in 2012 and 2013, for different countries
are given in Table 1.2 [22]–[24].

4
Table 1.2 Lithium production and reserves

Mine production (tons) Amount of reserve

Country
2012 2013 (tons)
United States * * 38,000
Argentina 2,700 3,000 850,000
Australia 12,800 13,000 1,000,000
Brazil 150 150 46,000
Canada 690 480 180,000
Chile 13,200 13,500 7,500,000
China 4,500 4,000 3,500,000
Portugal 560 570 60,000
Zimbabwe 1,060 1,100 23,000
World (Total) 35,000 35,000 13,000,000
* Commercial secret

Identified lithium resources for Bolivia is 9 million tons and it is not processable due
to lack of infrastructure.

Lithium and its compounds were mainly produced conventionally from three sources
as, minerals, clays and brines. Spent lithium ion batteries (LIB) can also be
considered as secondary resources of lithium, yet there is not an industrial
production of lithium from LIB contemporarily.

1.2.1 Minerals

Geologists state that 145 of minerals on earth contain some amount of lithium (% >
0.002 Li2O) and 25 of these minerals contain significant amount of lithium (% > 2
Li2O). Minerals which are commercially valuable in the production of lithium are
listed in Table 1.3 [25]–[28].

5
Table 1.3 Commercially valuable lithium minerals

Minerals Formula Lithium content (%)

Spodumene LiAlSi2O6 or Li2OAl2O34SiO2 3.73
Lepidolite LiKAl2F2Si3O9 or LiFKFAl2O33SiO2 3.56
Amblygonite LiAlFPO4 or 2LiFAl2O3P2O5 4.74
Triphylite LiFePO4 or Li2O 2FeOP2O5 4.40
Petalite LiAlSi4O10 or Li2OAl2O38SiO2 2.27
Bikitaite LiAlSi2O6H2O 3.28
Eucryptite LiAlSiO4 5.53
Zabuyelite Li2CO3 18.75

The biggest lithium mineral producers were Sons of Gwalia (Australia) with 150,000
tons, Bikita (Zimbabve) with 55,000 tons, and Tanco (Canada) with 21,000 tons in
2002 [29].

1.2.2 Clays

Clays are fine-grained rocks that consist of mostly metal oxides and silicates.
Hectorite is one of the famous clays which is rich is in lithium and magnesium
having 0.3 to 0.6 % Li [21]. The by-product of boron mine in Bigadiç-Balıkesir in
Turkey can be considered as a rich lithium clay which has 0.3 % Li.

1.2.3 Brines

Besides minerals, brines are another important lithium sources. The concentration of
lithium in brines is lower than minerals but the total amount of lithium in brines
makes up to 66 % of the world lithium resources. The brines with a commercial
importance in production of lithium are listed in Table 1.4 [21].

6
Table 1.4 Commercially important brines

Country Brine Li conc (ppm)

Chile Salar de Atamaca 1500
China Zabuye Lake 970
Salar de Hombre
Argentina 600
Muerto
USA Clayton Valley 450
- Seawater 0.17

1.3 Conventional lithium separation process from minerals and clays

Spodumene which is an igneous rock is used chiefly for lithium production among
the commercial lithium minerals. Approximately 250 kg of spodumene is required to
produce 1 kg of lithium.

Minerals are treated by two major methods which are acid, alkaline processes. There
is also chlorination process used barely.

In acid process, minerals are roasted with sulfuric acid at 250-400 oC to dissociate
silicate structure. Spodumene must be first converted from α- phase to β- phase at
1100 oC before treatment with sulfuric acid. The general reaction scheme is shown
below [30]:

Li2 OAl2 O3 4SiO2 + H2 SO4 → Li2 SO4 + Al2 O3 4SiO2 + H2 O

Finally, lithium carbonate is acquired by the addition of soda ash after the required
evaporation and pH adjustment steps.

In alkaline process, spodumene is calcined via limestone at 825-1050 oC. The

resulting product is mixed with water to have lithium hydroxide and further treated
with hydrochloric acid to have lithium chloride. The reaction scheme is shown below
[31];

Li2 OAl2 O3 4SiO2 + CaCO3 → Li2 O + CaOAl2 O3 4SiO2 + H2 O

7
In chlorination process, which is not widely used, the mineral is roasted with
chlorine gas or hydrochloric acid at 880-1100 oC and lithium chloride is produced as
a final product [32].

1.4 Conventional lithium separation processes from brines

There are five important brines in which lithium is produced commercially in all
over the world. These sources can be listed as; Clayton Valley (Nevada), Salton Sea
(California), Searless Lake (California) in North America and Salar da Atacama
(Chile), Salar de Hombre Muerto (Argentina) in South America. Among these
lithium sources Salar de Atacama has the highest lithium production with two plants;
SQM and Chemetall (Rockwood Lithium) [33], [34]. The plant on Humbre Muerto
was partly closed in 1999 due to the price reduction of SQM. Olarez Salina which is
placed near Humbre Muerto currently takes investment by Orocobre, which is the
largest lithium resource developer in Argentina [35]. Among those plants, Simbol
Materials which is on Salton Sea in California is the newest one and it has started its
production in September 2011 [36]. Although Bolivia has the largest lithium
reserves in brines with 9 million tons capacity, there is no production of lithium due
to lack of infrastructure in high hills where those resources are located [37]. China
also declared to produce lithium from the high lakes in Tibet like Zabuye Lake and
continues its research on the field since 2002 [38].

Due to the dilute concentration of lithium in even the most favorable brine, solar
evaporation of the brine to further concentrate it has been a necessary first step in all
of the world’s current lithium brine production. After the evaporation of excess
water in large pools, suitable precipitation agents are employed and unwanted salts
are separated. Thereafter, lithium is acquired in desired purity. Currently all the
conventional processes utilize evaporation and precipitation techniques. The other
techniques like ion-exchange, adsorption or membrane separation does not take
place in the processes. On the other hand, it is known that POSCO which is located
in South Korea will start production of lithium from seawater by 2015 [39]. Due to

8
the extremely dilute concentration of lithium in sea water, it is supposed that
Koreans will employ an additional technique like adsorption or ion exchange in
addition to evaporation and precipitation. POSCO claims that the production time for
lithium carbonate will be reduced to 8 hours from 12 months which also implies the
utilization of additional processes other than evaporation and precipitation [40].

Lithium production processes in four different brines were described and listed
below.

1.4.1 Clayton Valley, Nevada

Production of lithium from this deposit was initiated in 1966 by Foote Chemicals.
The plant changed hands to Chemetall in 1998 and now has a capacity of 6400 tons
of lithium carbonate per year. The valley has a 400 ppm lithium concentration
initially but after years of operation its lithium concentration now reduced to 160
ppm. The concentration of lithium and other major ions in Clayton Valley is listed in
Table 1.5.

Table 1.5 Clayton Valley, Brine Composition

Concentration (ppm) Concentration (ppm)

Ion 1966 1998
[41] [42]
Na 7.5 4.69
K 1 0.4
Li 400 163
Mg 600 190
Ca 500 450
Cl 11.7 7.26
CO3 - 74
B - 67
Br - 23

In Clayton Valley, lithium production starts with the evaporation of water in large
ponds. In the first four ponds, brine gets concentrated while some amount of calcite
(calcium carbonate) and gypsum (calcium sulphate dihydrate) precipitates. After 10
9
months of evaporation period, slaked lime (calcium hydroxide) is added to the brine,
so that gypsum and magnesium hydroxide is precipitated (magnesium concentration
in brine reduces to 2-3 ppm). Meanwhile the brine becomes basic and calcium starts
to precipitate by absorption of carbon dioxide near slyvinite (NaCl, KCl salt).
Slyvinite is sent to a different plant for the recovery of potassium. Afterwards,
glaserite (K3Na[SO4]2), potassium chloride, and remaining gypsum and calcite
precipitates in the last three ponds, then the final brine containing 5000 ppm lithium
is sent to lithium carbonate plant [21, p. 101]. In this plant, concentrated brine is
first treated with slaked lime so that the whole remaining magnesium and some part
of remaining sulfate and borate ions are separated. Next, calcium in the brine coming
from lime is withdrawn by adding small amount of soda ash (sodium carbonate) to
o
the mixture which is followed by heating the brine to 93 C and adding
stoichiometric amount of soda ash to precipitate lithium carbonate as a final product.
During this process some excess amount of water is added to the brine in order to
prevent the precipitation of other salts. The final product is accepted to be in
industrial grade due to its 400 ppm boron content. The whole lithium production
scheme in Clayton Valley is shown in Figure 1.1.

Figure 1.1 The process scheme of Clayton Valley

10
1.4.2 Salar de Atacama, Chile

The most suitable conditions such as high lithium concentration and high
evaporation regimes for the production of lithium from brines are present in this
playa. Salar de Atacama, a few Andean Salars and the Tibetan region of China are
the only places in the world where the humidity is low enough to allow bischoffite
(MgCl26H2O) and lithium salts to be crystallized in solar ponds on a commercial
scale. The prosperous conditions of the lake attracted many investors to the playa, so
that two different plants, SQM and Chemetall were built on this district. The salt
concentration in Salar de Atacama is shown in Table 1.6 [43, p. 483].

Table 1.6 Salar de Atacama, Brine Composition

Ion Na K Mg Li Ca B HCO3
Concentration
91000 23600 9650 1570 450 440 230
(ppm)

The general process scheme that shows the production of lithium compounds from
Salar de Atacama is shown in Figure 1.2 [44].

11
Figure 1.2 Production scheme of LiCO3 in Salar de Atacama

1.4.2.1 Chemetall (Rockwood Lithium)

The process in Chemetall starts with the evaporation of water in large ponds and
precipitation of sodium chloride in first pond, following by calcium chloride addition
and precipitation of slyvinite, carnalite (KCl.MgCl2.6H2O), bischoffite and lithium
carnalite (LiCl.MgCl2.6H20) [45]. The brine which is concentrated up to 4-6 %
lithium concentration in ponds, contains also 8000 ppm boron is sent to boron
extraction column. In this operation, boron is treated with alcohol-kerosene mixture
whose pH is adjusted to 2 by HCl solution. By this way the concentration of boron
ions in brine is reduced below 5 ppm in expense of losing 5-10 % lithium. After that
operation, lime is added to the solution so that magnesium and sulfate ions are
separated. Finally by the addition of soda ash and heating, lithium carbonate is
obtained [46].

12
1.4.2.2 SQM

SQM differs from other plants in terms of their production of by products like
potassium nitrate, slyvinite and other salts which bring them an economical
advantage near competitors. Like other plants, SQM lithium carbonate production
process starts in ponds with evaporation of water. In ponds, sodium chloride settles
down first following by potassium and sulfate salts which are sent to potassium
sulfate and boric acid plants. Slyvinite which settles down in following ponds is sent
to potassium nitrate plant. At that step in the process, brine lithium concentration
reaches up to 1 % and sent to ponds again to reach 6 % lithium. Meanwhile,
magnesium concentration reaches 1.8 % and boron concentration reaches 0.8 %
approximately. After that, the concentrated brine is sent to lithium carbonate plant
and similar to Chemetall after removal of boron, lime is added to the brine and
magnesium, sulfate ions are separated. In the last step lithium carbonate is
precipitated by the addition of soda ash and heating the brine [21, p. 122].

The production capacity of SQM is around 18000 tons of lithium carbonate per year.
After they break into market in 1988, the lithium prices lowered by half and this
reduced price considerably stimulated the lithium market and cause most of the
lithium producers from ores (which are relatively high-cost producers) close their
plants.

1.4.3 Salar de Hombre Muerto, Argentina

Lithium production from this brine was initiated by FMC in 1997. The composition
of lithium and other major ions in the lake is listed in Table 1.7 [43].

Table 1.7 Salar de Hombre Muerto, Brine Compositon

Ion Na K Ca Mg Li B
Concentration
15 76 210 72 2.1 7.8
(ppm)

Different from other lithium production methods, adsorption of lithium by alumina

was firstly utilized in this plant. The recovery of lithium starts with evaporation of
13
water by sunlight in large ponds similar to other plants until the concentration of
lithium reaches up to 600 ppm. Then the brine is sent to polycrystal alumina column
and lithium is captured by alumina particles. After that, the column is washed with
water and 1 % lithium chloride solution is acquired in column effluent which could
then be concentrated in solar ponds to the desired concentration for further
processing with lime and soda ash to produce lithium carbonate. The eluted column
would next be given a saturated sodium chloride wash to recover the entrained
lithium and to raise the ionic concentration in the alumina [47]. In preparing alumina
adsorbents to employ in lithium adsorption columns, lithium chloride is reacted with
alumina in saturated sodium chloride solution and forms LiCl/Al(OH)3 crystals in
which lithium chloride is present in amounts up to a 0.2 mole fraction. The alumina
adsorbent has a strong affinity for the lithium in high ionic solutions; oppositely the
lithium can be eluted in dilute solutions [21, p. 131].

FMC announced that, the process they have with alumina adsorbents has 20 % better
efficiency than other conventional solar evaporation, magnesium and sulfate
precipitation processes. However after SQM break into lithium market, the
decreasing prices of lithium carbonate result in partly closed plant and an agreement
to purchase lithium carbonate from SQM [21, p. 132].

1.4.4 Searles Lake, California

Searles Lake is operated by American Potas and Chemical Co. The brine
composition is given in Table 1.8 [42].

Table 1.8 Searles Lake, Brine Composition

Ion Na K Ca Mg Li B
Concentration
118400 15700 16 - 60 4120
(ppm)

The brine in Searles Lake is evaporated in triple-effect evaporators. In these

evaporators, the tanks are at different pressures and the vapor boiled off in one tank
can be used to heat the next, and only the first tank (at the highest pressure) requires

14
an external source of heat. During the evaporation process, sodium chloride, burkeite
(3Na2SO4.Na2CO3) and dilithium sodium phosphate (licons) are crystallized. Licons
are then converted to phosphoric acid and lithium carbonate by a reaction. In this
reaction, licons are firstly baked in an oven and organic residues are cleaned, then
licons are reacted with sulfuric acid at 115 oC and splitted into phosphoric acid and
lithium, sodium sulfates. Lastly, sulfates react with soda ash and lithium carbonate is
obtained [21, p. 134] .

1.4.5 Lakes in Turkey

There is no commercial production of lithium compounds in Turkey. Tuz Lake has

the highest lithium composition with 325 ppm but high magnesium content in the
lake blocks the production of lithium. The composition of ions in some of the lakes
in Turkey is listed in Table 1.9 [48].

Table 1.9 The compositon of lakes in Turkey

Lake Lithium Magnesium Potassium Sodium

Acı Lake 16 4108 1000 46000
Burdur Lake 3 970 46 5550
Tersakan Lake 34 6300 2000 23500
Boluk Lake 36 8800 2000 50000
Tuz Lake 325 37500 12000 61000

1.5 Literature survey on lithium separation methods from brines

Lithium is currently produced by evaporation and precipitation processes from brines

conventionally. However the high concentration of the foreign ions in brines with
lithium, force the researchers to develop an additional method to be integrated in
conventional processes. Liquid-liquid extraction and adsorption methods have been
the most investigated methods cited in the literature so far.

15
1.5.1 Liquid-liquid extraction

In this method, lithium in water is extracted by a proper solvent into another phase.
The solvability power of the utilized solvents towards lithium and foreign ions like
sodium, potassium and magnesium determine the performance of the extraction
process. In literature, it is seen that various organic solvents are studied in the
extraction of lithium from brines. Most commonly, organic alcohols were studied
and their performance was found to be the best among other solvents [49], [50]. The
solvability power of these alcohols generally increases with the decrease in the
carbon number of the alkyl chains. At the same time, selectivity of those alcohols
towards lithium ion decreases correspondingly. In 1966, Hermann reported that,
lithium, sodium and potassium chlorides and carbonates were dissolved in long
alcohols containing 3-8 carbons. For n-butanol the solubility of lithium is 10.57
grams LiCl per 100 grams of n-butanol [51]. The solubility of lithium chloride in
various alcohols and tetrahydrofuran (THF) reported in different studies are
tabulated in Table 1.10 [21, p. 136].

Table 1.10 Lithium chloride solubility in different alcohols

Solvent Lithium chloride solubility (g LiCl/g solvent)

N-butanol 10.6
Isopropyl alcohol 12.0
Propanol 16.2
Isobutanol 7.3
Pentanol 8.1
2-ethyl hekzanol 9.0
Alyl alcohol 4.4
Tetrahydrofuran 4.6

The lithium concentration in sea water is almost constant at 0.173 mg/L (0.173
ppm). The lithium concentration is very low but the total quantity occurring in sea
water is tremendous and estimated to be 2.4 x 1011 tons. The limited lithium content
on earth brings attention to this huge source in sea water. The low concentration of
lithium in sea water makes the pretreatment process inevitable. In the pretreatment
process, concentration of lithium is increased from 0.17 ppm to 30 ppm [52] by solar

16
energy. Than with a suitable solvent, lithium is extracted from aqueous phase to an
organic phase and finally by evaporating solvent or precipitation of lithium with a
proper anion would achieve the separation. The drawback of this process is the high
concentration of foreign ions in sea water which are magnesium (1290 ppm), sodium
(10770), potassium (380 ppm) and calcium (410 ppm) [53]. Another drawback is the
low distribution coefficient of the extraction solvents utilized in the process, which
makes the 5-10 evaporation and condensation cycles unavoidable. The solvents
utilized in extraction processes and their distribution coefficients towards lithium,
sodium, potassium and calcium ions are given in Table 1.11.

Table 1.11 The distribution coefficients of various solvents for lithium, sodium,
potassium and calcium ions [21, p. 137]

Distribution coefficient Separation factor

Solvent
kLi kNa kK kCa 𝑺𝑳𝒊
𝑵𝒂 𝑺𝑳𝒊
𝑲 𝑺𝑳𝒊
𝑪𝒂

n- butanol 0.058 0.023 0.020 0.017 2.5 2.9 3.4

Sec-butanol 0.044 0.022 0.021 0.020 2.0 2.09 2.1
Isobutanol 0.018 0.009 0.009 0.005 2.0 2.00 3.6
Pentanol 0.011 0.008 0.005 0.006 1.38 2.20 1.8
Isopentanol 0.01 0.006 0.007 0.005 1.66 1.43 2.0
2-ethylhexanol 0.001 0.007 0.004 0.006 0.57 1.00 0.7
2-ethylisohexanol 0.005 0.006 0.005 0.006 0.83 1.00 0.8
Octanol 0.005 0.007 0.006 0.006 0.71 0.83 0.8
2,4-Dimethyl-3-
0.006 0.007 0.008 0.014 0.86 0.75 0.4
pentanol
o-Chlorophenol 0.018 0.0056 0.0020 0.013 3.21 0.90 1.4
p-Chlorophenol 0.0119 0.0049 0.01 0.012 2.43 1.19 1.0
o-Cresol 0.045 0.019 0.03 0.006 2.36 1.50 7.5
m-Cresol 0.007 0.014 0.013 0.0008 0.50 0.53 8.8
p-Cresol 0.011 0.01 0.006 0.0012 1.0 1.83 9.2
Cresol 0.008 0.019 0.0049 0.009 0.42 1.63 0.9
Molar sol. Of
0.011 0.003 0.005 0.011 3.67 2.20 1.0
phenol/benzene
p-sec-
0.005 0.017 0.0013 0.009 0.29 3.85 0.6
Butylphenol/benzene
p-1,1,3,3-Tetramethyl-
0.002 0.003 0.007 0.003 0.67 0.29 0.7
(butyl)phenol/benzene

17
Table 1.11 (continued)
p-tert-
0.012 0.0089 0.008 0.012 1.35 1.50 1.0
Butylphenol/benzene

1.5.2 Adsorption

Separation of lithium via adsorption method is the most convenient technique to be

integrated in conventional lithium production processes which are evaporation and
precipitation. However, there is no industrial application of adsorption method
contemporarily. In this method, simply lithium ions are separated selectively from
aqueous solutions through physical or chemical adsorption. Research on adsorbents
focuses mostly on alumina and manganese oxide up to now.

The usage of alumina was first investigated by Goodenough in 1960 [54]. According
to this study, when aluminum chloride or aluminum hydroxide mixed with neutral or
basic water, it combines with lithium ions and precipitates in the form of a gel as
aluminium hydroxide. After that, lithium was separated from alumina by various
solvents and purified. The patents written by the followers of Goodenough claimed
that brines including high concentration of calcium and magnesium showed the
optimum separation values at 81 oC and pH value of 6.8 [55]. The separation factors
and optimum values change with brine source and physical conditions, also in these
studies separation of lithium from the precipitated gel was done by high temperature
water.

In Salton Sea, lithium was separated via alumina adsorbents and 99 % of lithium was
recovered from the brine but the process was found to be too expensive. The alumina
adsorption was investigated by many scientists with different brines. In 1986,
Rothbaum et al. studied the lithium recovery from waste water of a geothermal
energy plant by alumina in order to minimize the effect of lithium content on the
environment [56]. According to the study, after the separation of silica from water,
sodium aluminate was added to the water and at pH equals to 10, gelation occurs.
After that, the gel was washed with 60 oC water and the remaining gel was recovered
by washing it with sodium hydroxide solution. Unfortunately the purity of the
18
resultant lithium product was low. In the studies on geothermal water in France and
England, lithium separation was explored by aluminium chloride and most of the
lithium was gathered from water but the resultant product could not be acquired in
desired purity [57].

Alumina adsorption to separate lithium was also studied with sea water [52].
According to the study, 50 % mixture of magnesium hydroxide and aluminum
hydroxide was added to the brine and after the gel was formed, it was washed with
0.05 M HCl solution and lithium concentration was increased nearly seven fold. At
the end, approximately 40 % of lithium could be recovered from sea water. In
another study, brine in Dead Sea was concentrated from 30 ppm to 680 ppm by
aluminum chloride in 1978 [58], however to adjust the pH of brine, 73 tons of lime
stone was used for each ton of brine and to wash the precipitated gel, 19000 tons of
water and 92 tons of HCl was used. In 1981, Epstein et al carried a research and used
isoamyl alcohol to dissolve the gel [59]. After that, they extracted lithium from
alcohol by water, but the final product still contains 1.2 % Ca, 0.4 % Mg and large
amount of alcohol was lost in the process. Ryabtsev et al. investigated the structure
of the precipitated gel and they stated that the gel has a chemical formula with
LiCl.2Al(OH)3.mH2O and a layered structure [60]. They noted that gel was formed
in a layered structure which resulted in a molecular-sieve effect that could only be
penetrated by lithium. They further stated that, up to 40 % of the lithium in this
compound could be released by a water wash, and then be replaced (the adsorption
reaction) by contact with a lithium-containing strong brine. During the preparation of
the adsorbent, equimolar amount of LiCl and AlCl3 was used and the capacity was
turned out to be 7 mg Li/g adsorbent. This capacity reduces to 5-6 mg Li when PVC
was used as a binder. In the study, the capacity loss of the adsorbent after each use
and the final lithium product purity were not mentioned.

In another study conducted in 2007, the effect of pH and mixing time on the
efficiency of alumina adsorption was investigated at 25 oC by response surface
methodology (RSM). In this study, alumina gel was dissolved in n-hexanol, 2-ethyl
hexanol and methyl-iso-butyl acetone to separate lithium chloride [61]. The optimum

19
values were found as Al/Li mol ratio: 4.7, pH: 7.2 and 3 hours of mixing time with a
95 % of lithium retention capacity.

By the increase in usage of lithium ion battery technology, the research on finding a
more efficient anode material was also increased. For this reason lots of different
metal oxides (Zr, Ti, Sb, Cs, Rb, Ni, Mg, Na, Al, Co, Fe, V, Sn Mn) were
synthesized and investigated [62]–[70]. These metal oxides were also recognized as
adsorbents to be employed in lithium separation processes and most of them were
studied in terms of their separation efficiency. After it was realized that metal oxides
which include manganese give better separation factor among other adsorbents, the
research on this subject focused on manganese including adsorbents. The first article
about the subject was published by Shen in 1986 [71], where it was investigated the
mechanism of lithium insertion/extraction on manganese oxides and found out that
the insertion/extraction mechanism of lithium ions towards manganese adsorbent
occurs tapotactically which does not disrupt the crystal structure. After that, in
another study carried by Ooi et al., the crystal system of the lithium manganese oxide
was shown to be spinel, besides it was said that, manganese has +3 and +4 oxidation
states in closed cubic packed (ccp) oxygen system [72]. In the continuation study of
the same group, lithium addition and extraction mechanism was described in three
ways as redox type, specific ion-exchange and non-specific ion exchange. The
synthesis conditions determine which type dominates [73]. In another study again by
Ooi et al., the researchers showed that, adsorbents synthesized below 500 oC works
with redox type and adsorbents synthesized above 500 oC works with ion exchange
mechanism. In their conclusion they also stated that, lithium extraction/insertion
mechanism mostly happens to be with ion-exchange mechanism. In Figure 1.3, the
mechanisms of ion-exchange and redox were shown [74].

20
Figure 1.3 Lithium insertion/extraction mechanisms (a) Lithium extraction (b)
Lithium insertion O: redox side; □: ion exchange side

Manganese oxide is known to adsorb most of the metals onto its surface, even it is
known that the heavy metal balance in oceans are controlled by the manganese oxide
nodules in water [75]. In literature, there are many articles about synthesis of
LiMnO2 adsorbents. In these articles, synthesis studies were conducted with different
precursors, temperatures and molar ratios of Li/Mn. The gathered performance
values investigated under different pH values, temperature and initial lithium
concentrations in feed solution. In the studies of Ooi et al where the adsorption
mechanism of LiMn2O4 described, the adsorbents were synthesized by solid-solid
reaction [72]–[74] in 1989-1992. In these syntheses, MnOOH and MnCO3 were
utilized as manganese source while LiCO3 was utilized as lithium source. Reaction
takes place at temperatures between 400-800 oC, for 4 hours. Li/Mn mol ratio was
adjusted to 0.2. 0.5 and 0.8. The resultant product was mixed with 0.5 M HCl for
two days and lithium ions were exchanged with hydrogen ions. After that adsorbents
were dried in oven at 70 oC and put into LiOH solution to investigate the exchange
mechanism between hydrogen and lithium ions. In 1995, Feng and Ooi investigated
the synthesis of LiMn2O4 and NaMn2O4 with a hydrothermal method [76]. In this
method, NaOH or LiOH was mixed with -MnO2 in a desired mol ratio and put into
an autoclave where the reaction takes place at 170 oC. After 24 hours synthesis time,
the product was washed with water and left for drying at room temperature for 3
days. After the synthesis, it was observed that, lithium manganese oxide was in
21
spinel type and sodium manganese oxide was in birnessite type. In the same study,
the selectivity of the adsorbents towards lithium ions with foreign ions like sodium,
potassium and magnesium was investigated. The selectivity order of the ions can be
listed as Li+ < Na+ < Cs = K for sodium manganese oxide and Na+ < Li+ < K+ < Rb+
< Cs+ for lithium manganese oxide at pH=4, which means actually low selectivity
for lithium. Roziere and Burns made a similar study in 1995 but synthesized the
adsorbents with three different reactant groups as lithium carbonate-manganese
carbonate, lithium nitrate--manganese oxide and lithium nitrate-β-manganese oxide
in solid-solid medium [77]. They fixed the molar ratio of Li/Mn as 0.8 and oven
temperature as 400 oC. The synthesized adsorbents were then mixed with 0.2 M HCl
solution for 48 hours and dried in an oven at 60 oC. The lithium extraction ability of
the resultant product was measured by 0.1 M LiCl solution but only the mechanisms
were reported, no capacity value was noted. In 2001, Chitrakar et al. published an
article, and claimed that they found the highest lithium adsorption capacity with their
lithium manganese adsorbent as 40 mg Li/g adsorbent [70]. There are some other
studies which has an adsorbent capacity near 40 mg, however they used concentrated
lithium chloride solutions. The selectivity of the synthesized adsorbent has a formula
of Li1.6Mn1.6O4 with a high selectivity to lithium and decreasing selectivity in order
with Na+ > K+ > Mg2+ > Ca2+ . In Chitrakar’s adsorbent synthesis procedure -
MnOOH and LiOH were autoclaved at 120 oC for 1 day. LiMn2O4 was taken as a
product and dried at 60 oC. After that it was placed into an oven operating at 400 oC
for 4 h, Li1.6Mn1.6O4 was taken as an intermediate final product. Thereafter, 1.5 g of
Li1.6Mn1.6O4 was mixed with 2 L 0.5 M HCl solution and H1.6Mn1.6O4 was taken as a
final product. The same synthesis also described for reflux method with same
procedure except that the reaction took place for 8 hours. Adsorbents which are
synthesized by using an autoclave gave slightly higher capacity values than the
adsorbents synthesized by reflux method.

Another important study about lithium manganese adsorbents were conducted by

Yoshizuka et al. [78]. In their first paper which was published in 2002, they made the
synthesis of the adsorbents by using Mn3O4 and LiOH as reactants. Li/Mn ratio was
adjusted to 0.5 and the synthesis was done in two steps, one of which was at 425 oC,
5 h and the other was 500 oC, 5 h in electrical oven. The resulting product was
22
cooled in oven slowly for 12 h and after that mixed with 1 M HCl to form HMn2O4.
Adsorbent capacity was found as 10.5 mg Li/g adsorbent approximately and it was
observed that as pH of the initial lithium chloride solution increases, the capacity
value also increases. In another study by the same group [79], the adsorbent capacity
was increased by changing the Mn/Li ratio to 1.5 and keeping the other variables
constant during the synthesis. In the following study in 2006, the researchers
investigate the application of the adsorbents in recovery of lithium from sea water by
designing an adsorption column operating at 200 L/h [80]. After the column was
filled with 60 kg of lithium manganese oxide, it was fed with 140 m3 of sea water in
30 days. Finally, the column was washed with HCl solution and after drying, 151.7 g
of precipitated salt was acquired. This salt has a composition of 10.35 % LiCl, 38.3
% NaCl, 0.89 % KCl, 36.9 % CaCl2 and 5.3 % MnCl2. In another run in the same
column for 150 days, the final composition was found as 33.3 % LiCl, 20.4 % NaCl,
3.3 % KCl, 8.2 % MgCl2, 13.4 % CaCl2 and 19.4 % MnCl2. The breakthrough curve
gathered from the column was reported as in Figure 1.4 [80].

Figure 1.4 Breakthrough curve for 150 days operation

Yoshizuka et al. published their last study on lithium adsorption in 2011. In this
article, they integrated additional ion exchangers on lithium chloride separation
process in order to purify lithium chloride more and get rid of foreign ions like
sodium and magnesium [81]. The process started with feeding sea water into the
adsorption column, filled with lithium manganese oxide. Then column was eluted

23
with HCl solution to extract lithium. Afterwards, the elution solution with 160 ppm
lithium concentration was fed to DIAION SK110 ion exchanger to separate Mg, Ca,
Sr and Mn. After that, the brine was fed to an ion exchange column filled with 1-
phenyl-1,3-tetradecanedione to separate manganese and magnesium further and
finally reache 900 ppm Li concentration. Afterwards the water was evaporated and
by addition of ammonium carbonate, lithium carbonate was precipitated. The final
product acquired has a purity of 99.9 % and the total process had an efficiency value
of 56 %.

Kima et al. investigated the structure of lithium manganese oxide in 2003 and tried
to impregnate Ti and Fe into the structure [64]. In this study, solid-solid reaction
system was chosen and lithium carbonate was used as a lithium source while
manganese carbonate was used as an manganese source. For iron and titanium
resource, FeOOH and TiO2 were employed. In this article no separation experiments
were done but the crystal systems of the resulting products were described and
selectivity of the new adsorbents were discussed.

In 2006, Wang et al. studied the parameters affecting the capacity of the lithium
manganese oxide adsorbents synthesized by solid-solid method. They analyze the
effect of Li/Mn ratio, heating time, heating temperature, precursors, leaching
conditions etc. [82]. They used manganese carbonate as single manganese source,
lithium carbonate and lithium hydroxide as lithium sources. In their conclusion, it
was stated that, LiOH and MnCO3 pair gave highest capacity values and as Li/Mn
ratio increases lithium adsorption capacity of the adsorbents increase also. It was
found out that when LiOH and MnCO3 were used and Li/Mn ratio was adjusted to 1,
adsorbents worked most efficiently and give 23 mg/g capacity value.

Wang et al. published another article in 2008 and they synthesized the adsorbents
with MnCl2, LiOH and H2O2 by hydrothermal process [83]. They performed the
adsorption experiments under constant pH by using an ammonium chloride-
ammonia buffer and investigated the effect of ionic strength and temperature on
capacity. It was found out that ionic strength decreases the efficiency of the
adsorption process, which was explained by the formation of an ionic layer on
adsorption surface that increases the diffusion resistance of lithium ions into the
24
adsorbent particle. In their study, they also measured the values of ΔH, ΔG and ΔS
during adsorption and they found out that the lithium adsorption process is
endothermic which means increasing capacity with the increase in temperature. In
another aspect, they explained the increase in capacity by temperature in terms of
requirement of heat during exchange. The exchange of Li+–H+ need first the bond
breaking of O–H and also the sorption of Li+ or the insertion of Li+ for a diffusion
process, which was an endothermic process; i.e., the rise of temperature favors
adsorbate to transport within the pores of the adsorbent.

Wang et al. published another article in 2009 and discussed the effect of pH on the
capacity of adsorbents [84]. They found out that, as pH increases from 8 to 12 the
capacity value increases from 8 mg/g to 35 mg/g.

Insertion of alternative metals into lithium manganese oxide spinels was another
subject studied by the researchers. For instance, Ma et al. investigated the insertion
of antimony and used Sb2O3 as reactant with MnO2 and repeat the synthesis
experiments under different temperature and mole ratios [63]. They found out that
adsorbents having a molar ratio of Sb/Mn=0.05 and Li/(Sb+Mn)=0.5 synthesized at
800 oC gave 20 % better capacity values with respect to ordinary lithium manganese
oxide. In another study which tries to impregnate Mg onto adsorbent structure, it was
found that Mg tends to stabilize the crystal structure of the adsorbent [85]. In this
work, MnCl2·4H2O, Mg(NO3)2·6H2O LiOH ve H2O2 were employed as reactants.

Lithium manganese oxide synthesis was also done with ultrasonic pyrolysis method
by Özgür et al in 2010 by using lithium nitrate and manganese nitrate as reactants
[86]. The surface area of the resultant adsorbents were similar to adsorbents
synthesized with ordinary solid-solid reaction but the morphology of the adsorbents
were found smoother. The total capacity of the adsorbents decrease a little but the
kinetics of the adsorption get increased so that the adsorption equilibrium was
reached in less time with respect to ordinary synthesized adsorbents.

25
26
 CHAPTER 2

EXPERIMENTAL METHODS

2 EXPERIMENTAL METHODS

2.1 Materials

The chemicals used throughout the study were reagent grade chemicals that were
purchased from Sigma Aldrich or Merck, unless otherwise stated. N-alkyl
formamides, lithium manganese oxides, beads used in adsorption experiments and
functional polymers were synthesized and characterized in this study.

2.2 Synthesis and experimental procedures

2.2.1 Synthesis of N-alkyl formamides

The formamides were synthesized by action of dry formic acid on primary and
secondary amines with appropriate alkyl groups. In a typical procedure, one mole of
the amine compound was placed in a conical flask equipped with a pressure
equalized dropping funnel. The flask was mounted in an ice bath on a magnetic
stirrer. While stirring vigorously, one mole of formic acid (57.5 g of 85 % solution)
was added drop wise via the dropping funnel. An ammonium formate salt was
formed as white precipitate or waxy product. Soon after the formic acid addition, the
flask was transferred into silicon oil bath and attached to a Dean-Stark trap and a
27
reflux condenser. To the flask, there was added 60 mL toluene to remove water
azeotropically. The temperature of the oil bath was maintained at 140 oC. In all cases
the water evolution was completed (1.0 mol) within 45-50 min in these conditions.
After removal of toluene the residual liquid was distilled under vacuum. The
products so obtained were optically clear.

2.2.2 Synthesis of lithium manganese oxide adsorbents

2.2.2.1 Solid-solid-1 [79]

5 g (0.022 mol) Mn3O4 and 0.792 g (0.033 mol) LiOH was mixed and grounded for
15 min in a ceramic crucible (setting Mn/Li=2). Than the mixture was sintered at
425 oC for 5 h in oven and cooled at room temperature for 1.5 h. Then, the blend was
mixed and grounded again for 15 min. After that, it was sintered at 500 oC for 5 h in
oven and cooled for 12 h in the oven. Spinel type LiMn2O4 was acquired.

2.2.2.2 Solid-Solid-2 [73], [74], [87]

Mn source (MnCO3, MnO, MnO2) and LiOH.H2O were weighed and put into
ceramic crucibles according to the molar ratio of Li/Mn and grounded for 15
minutes. The crucibles were placed into oven which was already adjusted to a
desired temperature (450 or 550 oC) and kept there for 6 hours. After that, the
crucibles were removed from oven and cooled at room temperature in a dessicator.

2.2.2.3 Hydrothermal-1 [83]

4.8 g of LiOH was dissolved in 100 ml of water and 2 M LiOH solution was
prepared. 3.96 g of MnCl2.4H2O was dissolved in sufficient amount of water
separately. 40 ml of 2 M LiOH solution was added dropwise into MnCl2 solution
with vigorous magnetic stirring (means, 0.02 mol MnCl2.4H2O was mixed with 0.08
mol of LiOH). The color of the solution turns yellow and brown during addition.
Half an hour later, 5 ml of 30 % H2O2 solution was added into above solution

28
dropwise and stirred for 2 h. A black gel was formed and aged for 1 day (The color
of the solution turned to black from brown). The gel was transferred into a teflon
lined stainless steel vessel and deionized water was added until 70-80 % of the vessel
was reached. The vessel was placed into oven at 120 oC for 24 h. The precipitate was
filtered and washed with 1 L of deionized water and dried at 60 oC for minimum 6
hours. The obtained product was placed into 350 oC oven for 5 h and cooled in room
temperature. A leaching solution of 0.5 M HCl acid was prepared and 0.5 g of cooled
product was mixed with 250 ml of leaching solution in batch shaker for 24 hour.
Than the solution was filtered under vacuum and washed with 500 ml pure water.
The acid solution can be analyzed for its lithium content. The precipitate was dried at
60 oC for minimum 6 hours.

2.2.2.4 Hydrothermal-2 [70]

4.8 g of LiOH was dissolved in 100 ml of water and 2 M LiOH solution was
prepared. 1 g of -MnOOH was mixed with 40 ml of 2 M LiOH solution (means
0.0115 mol MnOOH was mixed with 0.08 mol of LiOH) in a teflon lined stainless
steel vessel and autoclaved at 120 oC for 1 day. The precipitate was filtered and
washed with 500 ml of deionized water and dried at 60 oC. The obtained product
(LiMnO2) was ready for use after kept in oven at 400 oC for 4 h.

2.2.2.5 Hydrothermal-3 [88]

Equimolar (0.7-1.5) amounts of H2O2 and LiOH were dissolved in sufficient amount
of water in different beakers and mixed. Menawhile, 0.4 mol Mn(NO3)2 was
dissolved in a teflon lined stainless steel autoclave and LiOH-H2O2 solution was
dripped into it. After the solution was maintained at room temperature for 2 h, the
autoclave was sealed and heated to 70-180 oC for 8 hours. The resulted black solid
products were filtered and washed with plenty of water (1 L), then dried at 70 oC for
12 hours.

29
2.2.3 Synthesis of lithium manganese oxide precursors

2.2.3.1 MnCO3 synthesis procedure [89]

16.9 g of manganese (II) sulfate-1-hydrate (0.1 mole) and 10.56 g of ammonium

carbonate (0.11 mole) were dissolved in sufficient amount of water in separate
beakers. Ammonium carbonate solution was added in small lots to the hot
manganese sulfate solution (60-70 oC). The progress of the reaction could be
observed by the precipitation of manganese carbonate and release of carbon dioxide
bubbles. The manganese carbonate precipitate, which was flesh in color, was
allowed to precipitate and washed with plenty of water several times by a
decantation until the supernatant solution gives no further test for sulfate.

2.2.3.2 Mn3O4 synthesis procedure [89]

Manganese carbonate was taken to an oven at 80 oC for 4 hours. After that, the dried
precipitate was transferred to a ceramic china dish and put into high temperature
oven. The temperature of the oven was raised to 400 oC in steps of 100 oC, the
decomposition of manganese carbonate was observed due to change in color to
black. The material was taken out and cooled in dessicator to room temperature, then
taken again in furnace and heated to 940 oC in steps of 100 oC. The product (Mn3O4)
was kept in 940 oC for 1 hour, (the color of the material turns to brown in this step)
taken out and cooled in desiccator.

2.2.4 Synthesis of poly (styrene-maleic anhydride-glycidyl methacrylate)

(PSMA) [90]

0.1 mol (9.806 g) maleic anhydride, 0.1 mol (10.415 g) styrene, 0.038 mol (5.42 g)
glycidyl methacrylate and 0.00238 mol (0.577 g) benzoyl peroxide were dissolved in
51.3 g ethyl methyl ketone (EMK) in a balloon flask and put into a constant
temperature bath on a temperature controlled-heater magnetic stirrer. The reaction
continued for 4 hours at 70 oC. The resulting mixture was precipitated in cold diethyl

30
ether. The white product was taken to an oven in dry atmosphere and dried at 40 oC
under vacuum for 12 hours.

2.2.5 Liquid-liquid extraction procedure

The method used in the extraction experiments was as follows [91]; 10 ml

formamide and 10 ml water phases were weighed into a separation funnel directly
for the measurements in room temperature or weighed into an erlenmeyer for the
higher temperature experiments. Erlenmeyers were then placed into a constant
temperature oil bath. After thermal equilibrium was reached, the samples were
mixed vigorously at constant temperature for several minutes (3-5 minutes), until the
equilibrium was reached. Than the samples in the erlemeyers were taken into the
separation funnels which also kept at constant temperature by using a drying oven.

The mixing time of the samples can change according to the viscosities, densities
and the interfacial tension between two phases. After phase separation was
established, the water phase (raffinate) and formamide phase (extract) were taken
into separate glass containers (formamides are strong solvents and they can solve the
plastic containers). Then, the raffinate was analyzed by atomic absorption
spectrophotometer to determine lithium concentration and by total nitrogen analyzer
to determine the formamide concentration.

2.2.6 Polymer enhanced ultrafiltration procedure

Desired amount of polymer was dissolved in less than 1 L pure water in a 1 L

volumetric flask. Known concentration of lithium chloride was then added to the
polymer solution to obtain a final desired concentration of lithium chloride in 1 L.
After that, the pH of the solution was adjusted by adding droplets of concentrated (3
M) ammonium hydroxide or hydrochloric acid solution. The solution was mixed by
a magnet for at least 1.5 hours. Then pH was again adjusted if needed. After that,
the solution was taken to a 1 L erlenmeyer and ultra-filtered continuously, in other
words returning the retentate and permeate back to the feed solution in order to keep
the feed concentration constant. Ultrafiltration pressure difference was 200kPa and

31
temperature was kept constant at 25 oC. Samples were taken from both the feed
solution and the permeate stream at different time intervals for lithium analysis in
AAS and for polymer analysis in total organic carbon analyzer.

Polymer purification and fractionation was carried out in batch ultrafilter (Amicon-
Millipore Model 8050). Ultrafiltration experiments were conducted in continuous
ultrafiltration set-up which is shown in Figure 2.1. The set-up mainly consists of
Osmonics Sepa CF Membrane Cell (7), UF membrane (8), pressure gauge (9), pH
meter (3), pH probe (2), feed tank (1), feed pump (4), rotameter (6) and pressurized
air (11).

Figure 2.1 Continuous ultrafiltration set-up [92]

2.2.7 Batch adsorption procedure

Batch adsorption experiments were carried in a 250 ml flask at room temperature.

100 ml lithium chloride solution was added to 0.05 g of adsorbent (LiMnO) and
mixed in a batch shaker at room temperature (25 oC). The pH of the mixture was
adjusted to a desired value by 1 M ammonia and 1 M hydrochloric acid solutions. In
order to attain equilibrium, shaking time was adjusted to 24 h. After mixing, the
supernatant was taken and measured for ion content in AAS.

32
Leaching experiments were carried out by a leaching solution of 0.5 M HCl with a
certain amount of product in a 250 ml erlenmeyer for 24 hour at room temperature,
than the supernatant was taken for desired ion analysis in AAS.

2.2.8 LiMnO bead preparation

In LiMnO bead preparation, two methods were employed, which were suspension
polymerization technique and precipitation technique

- Suspension polymerization technique

First organic and aqueous phases were prepared separately. Organic phase consists
of 0.57 g glycidyl methacrylate (GMA), 1.4 g methyl methacrylate, 0.4 g ethylene
glycol dimethacrylate and 0.03 g azobisisobutyronitrile (AIBN) and 2.37 ml toluene.
In aqueous phase, 0.57 g sodium sulfate, 0.08 g polyvinyl pyrrolidone were
dissolved in 28.4 g of water. Lithium manganese oxide particles were added to
organic phase and formed a slurry. After aqueous phase reached to 65 oC in a
constant temperature bath by a temperature controlled heater, the organic slurry was
dripped onto aqueous phase slowly. The organic drops split into individual droplets
in aqueous phase and the process was completed in 4 hours.

- Precipitation technique

In precipitation technique, 3 grams of poly (styrene-maleic anhydride-glycidyl

methacrylate) (PSMA) was mixed with 1.5 g N-methyl pyrrolidone (NMP) and 1 g
of LiMnO. LiMnO was added in excess amount into the mixture to be sure that
NMP-PSMA solution was saturated with LiMnO content. Tap water was put into
another 1 L beaker and few drops of cetyl tert butyl ammonium bromide was added.
The slurry was dripped into 1 L beaker by a 5 ml volume dropper while the water
was mixing with a magnet on a magnetic stirrer. After dripping process, newly
formed beads continued to be mixed by the stirrer for 24 hours. Then the beads were
filtered by an ordinary filter paper, washed with plenty of deionized water and dried
in room temperature for 24 h.

33
2.2.9 Column experiment procedure

Certain amount of salt (LiCl, NaCl, KCl or MgCl26H2O) was dissolved in 5 L

volumetric flask to have desired final ion concentration. The solution was taken to a
10 L plastic bottle and fed to an adsorption column from the bottom. The feed
solution was prepared regularly until the end of the experiment. The column has 20
mm diameter and 60 cm long which is made of borosilicate glass. It has two glass
filters at the bottom and top ends in order to hold the bead content in column. The
outlet of the column was sent to another 10 L plastic bottle. The flow was derived by
a peristaltic pump and the flow rate was adjusted by the valve below the column and
measured by a stop-watch and a 50 ml graduated cylinder. In every data point, pH
measurements were done and the samples were taken to AAS for concentration
measurement.

The experimental set-up was as shown in Figure 2.2 and Figure 2.3.

Flowrate and pH are

measured at effluent

Outlet
Inlet tank Li, Na, K, Mg
tank
concentrations are
AAS measured

Figure 2.2 Scheme of the set-up for column experiments

34
Figure 2.3 Photos of the set-up for column experiments

The area under breakthrough curve was calculated with numerical integration by
Trapezoidal Rule. The capacity of the beads were found by subtracting the area
from the total amount of lithium fed to the column.

2.3 Equipment

2.3.1 Measurement of metal ion concentration in aqueous solutions

Lithium, sodium, potassium, magnesium and calcium contents in aqueous solutions

were measured by the aid of Shimadzu Atomic Absorbance Spectrophotometer using
acetylene as a fuel. Lithium, sodium and potassium ions were measured by emission
technique while others were measured by absorbance using specialized lamps.

Each individual metal ion concentration was measured by first constructing a

calibration curve with metal ion solutions at known concentrations.

35
2.3.2 Measurement of organic content in aqueous solutions

Analysis of formamide content in water in liquid-liquid extraction experiments was

done by measuring the total nitrogen content by Total Nitrogen Analyzer Shimadzu
TNM-1. The nitrogen content in formamides were used to calculate the total
formamide content in water.

The polymer content in permeate in PEUF experiments was checked by Total

Organic Carbon Analyzer. The carbon content in permeate was directly related with
the polymer content which shows the filtration performance of the PEUF set-up.

2.3.3 Spectroscopic methods

The structure of the synthesized materials throughout the study were identified by
Fourier Transform Infrared Spectroscopy (FTIR) using Shimadzu IR Prestige-21
FTIR spectrophotometer with a Pike Miracle ATR apparatus (KBr was used for a
reference) and Nuclear Magnetic Resonance Spectroscopy using 400 MHz of
Bruker NMR Spectrophotometer using CDCl3 as solvent with Ultrashield
Superconducting magnet.

The crystal structure of lithium manganese oxides were identified by using Rigaku
Ultima-IV X-Ray diffractometer that generates a voltage of 40 kV and current 40
mA from Cu Kα radiation source (λ = 1.5418 Å). The diffraction angle 2θ was
scanned from 10° to 60° with scanning rate of 1°/min and a step size of 0.01°.

2.3.4 Thermal analysis

The synthesis mechanism of lithium manganese oxides were identified by

Thermogravimetric Analysis (TGA) and Differential Thermal Analysis (DTA) which
performed simultaneously in air at a heating rate of 10 °C min−1 in alumina
crucibles.

36
2.3.5 Bead characterization

The characterization of the beads was done by Scanning Electron Microscopy (SEM)
and Brunauer–Emmett–Teller (BET) analysis. Surface area and pore size distribution
measurements were done by Quantachrome Corporation Autosorb-6B, under
nitrogen atmosphere at -197 oC and 1x10-5 - 1.0 atm. The micron sized photos of the
beads were taken in SEM by QUANTA 400F Field Emission SEM machine after
coating the particles with Au-Pd.

2.3.6 Physical properties of the synthesized formamide

2.3.6.1 Temperature-vapor pressure measurements

In order to measure the vapor pressure of hexyl formamide with respect to changes
in temperature, an experimental set-up shown in Figure 2.4 was established.

Figure 2.4 Vapor pressure measurement set-up

The name and functions of the numbered constituents are explained below for Figure
2.4;

37
1: Vacuum pump
2: Pressure gauge, to measure the pressure in the system
3: Needle valve, to adjust the pressure in the system accurately
4: Pressure gauge, to measure the pressure in the tank
5: Ball valve, controls the connection between tank and the vacuum pump
6: Tank
7: Temperature controlled heater
8: Thermometer, to measure the system temperature
9: Condenser, to condense the vapors coming from the system
10: The balloon flask, to hold the liquid to be measured

In this set-up, one can measure the pressure when the liquid in the system starts to
boil by adjusting the pressure on top of the liquid with using a needle valve. The
measured pressure where the first bubble appears actually shows the vapor pressure
of the compound at that temperature. The large tank (6) here is employed in order to
attain a constant pressure value during the measurements.

2.3.6.2 Viscosity measurements

The viscosity measurements were done by Brookfield LVDV-III Ultra rheometer

with using a SC4-34 spindle and small sample adaptor. The device has a 0.01-250
rpm speed range. The data acquired was processed by using Brookfield Rheocalc
computer program. The temperature was kept constant by using constant temperature
circulating bath.

2.3.6.3 Surface tension measurements

The surface tension of the solutions was measured by using a Cole Parmer Surface

Tensiomat 21 with a platinum-iridium ring (circumference: 5.965 cm). The surface

tension values were measured in between 30-70 Dynes/cm.

38
 CHAPTER 3

RESULTS AND DISCUSSIONS

3 RESULTS AND DISCUSSIONS

The results of this thesis study were subdivided into four groups; the synthesis and
characterization of the materials employed in the study, liquid-liquid extraction by
N-alkyl formamides, adsorption results by lithium manganese oxide adsorbents and
lastly adsorption results in a continuous column by PSMA-LiMnO beads. Additional
materials synhthesized and methods used in the study were summarized in
APPENDIX A.

3.1 Synthesis and characterization of materials employed in lithium

separation

3.1.1 Synthesis and characterization of N-alkyl formamides

3.1.1.1 NMR and FTIR

To investigate lithium extraction behavior of the formamides, four formamides

bearing hexyl, octyl, dibutyl and dihexyl groups were synthesized as depicted in
Figure 3.1.

39
Figure 3.1 Synthesis pathway of N-alkyl formamides

The synthesis procedure was quite straightforward in which equimolar quantity of

amine was mixed with formic acid and heated to 140 oC to evolve the reaction water.
The products so obtained were pure enough as inferred from 1H-NMR and FTIR
spectra. It is important to note that both NMR and FTIR spectra of the formamides
show very similar patterns, for instance the hexyl formamide derived from hexyl
amine and formic acid exhibits a strong carbonyl vibration at 1650 cm-1 and a N-H
plane bending band at 1530 cm-1 in the FTIR spectrum. The other peaks, N-H
stretching vibration at 3350 cm-1 and C-N vibration bands at 1220 and 1100 cm-1
support the proposed formamide structure. 1H-NMR spectrum of this product shows
a weak signal of formamide proton at 8.0 ppm. The broad signal centered at 5.3 ppm
must be associated with N-H proton. The proton signals of methylene group linked
to nitrogen appears as downfield signal at 3.3 ppm, while the remaining protons of
aliphatic carbons lie in between 0.9-1.6 ppm range. The residual hexyl amine in the
crude product shows proton signals at 3.2 ppm as a weak triplet.

The products except dihexyl formamide were liquid at room temperature. Dihexyl
formamide is a low melting solid that melting slightly above room temperature (35-
40 oC). The FTIR and 1H-NMR spectra were described in detail as listed below.

Hexyl formamide (distilled product): Boiling point (120 oC/12 mm Hg). 1H –NMR, δ
8 (1H, s, OC-H), 5.3 (1H, b, N-H), 3.3 (2H, t, N-CH2), 1.2-1.6 (8H, m, CH2 of the
linear chain), 0.9 (3H, t, CH3 end group). Significant FTIR peaks: sharp peak at 1640
cm-1 (C=O stretching vibration), the strong band at 1530 cm-1 (N-H plane bending

40
vib.), broad band at 3350 cm-1 (N-H stretching vibration), small peak at 1100 cm-1
(C-N stretching vibration), sharp peak at 1220 cm-1 (OC-N stretching band).

Octyl formamide (distilled product): Boiling point (123 oC/12 mm Hg). 1H –NMR,
δ 7.8 (1H, s), 5.1 (1H, b), 3.2 (2H, t), 1-1.4 (12H, m), 0.9 (3H, t). Significant FTIR
peaks: sharp peak at 1640 cm-1 (C=O stretching vibration), the strong band at 1530
cm-1 (N-H plane bending vib.), broad band at 3350 cm-1 (N-H stretching vibration),
small peak at 1100 cm-1 (C-N stretching vibration), sharp peak at 1220 cm-1 (OC-N
stretching band).

Dibutyl formamide (distilled product): Boiling point (135 oC/12 mm Hg). 1H –NMR,
δ 8 (1H, s), 3.2 (4H, t), 1.4 (4H, quin), 1.2 (4H, sex), 0.9 (6H, t). Significant FTIR
peaks: sharp peak at 1640 cm-1 (C=O stretching vibration), the strong band at 1530
cm-1 (N-H plane bending vib.), broad band at 3350 cm-1 (N-H stretching vibration),
small peak at 1100 cm-1 (C-N stretching vibration), sharp peak at 1220 cm-1 (OC-N
stretching band).

Dihexyl formamide (distilled product): Boiling point (140 oC/12 mm Hg). 1H –

NMR, δ 8 (1H, s), 3.2 (4H, t), 1.2-1.5 (16H, quin), 0.9 (6H, t). Significant FTIR
peaks: sharp peak at 1640 cm-1 (C=O stretching vibration), the strong band at 1530
cm-1 (N-H plane bending vib.), broad band at 3350 cm-1 (N-H stretching vibration),
small peak at 1100 cm-1 (C-N stretching vibration), sharp peak at 1220 cm-1 (OC-N
stretching band).

(s: singlet, b: broad, t: triplet, m: multiplet, quin: quintet, sex: sextet)

3.1.1.2 Viscosity measurements

All liquids obey the Newtonian law and give the same viscosity values at all shear
rates. The viscosity values decrease as the temperature increases. The results can be
seen from Figure 3.2.

41
 35

30

25
Viscosity (cP)

20
DbFo
15
OcFo
10 HeFo

0
0 10 20 30 40 50
Temperature (oC)

Figure 3.2 Temperature dependence of viscosity for N-alkyl formamides

3.1.1.3 Surface tension measurements

The values were measured at 10 oC and shown in Table 3.1.

Table 3.1 Surface tension values for the N-alkyl formamides at 10 oC

Formamide Surface Tension (dynes/cm)

Hexyl formamide 37.3
Octyl formamide 35.1
Dibutyl formamide 33.5

3.1.1.4 Density measurements

Another important physical property of the N-alkyl formamides are their densities.
The density measurements were done at three different temperatures by using a
pyknometer and listed in Table 3.2.

42
Table 3.2 The densities of the N-alkyl formamides at different temperatures

Density (g/cm3)
N-alkyl formamide
10 oC 25 oC 40 oC
Hexyl formamide 0.896 0.890 0.883
Octyl formamide 0.880 0.876 0.874
Dibutyl formamide 0.886 0.883 0.879
Dihexyl formamide - - 0.875

3.1.1.5 Vapor pressure measurements

In order to avoid the decomposition risk of hexyl formamide, the temperature of the
system was not increased above 180 oC in the experiments. The relation between
temperature and vapor pressure was drawn and shown in Figure 3.3.

50
45
40
Vapor pressure (kPa)

35
30
25
20
15
10
5
0
410 420 430 440 450 460
Temperature (K)

Figure 3.3 Temperature-vapor pressure relation for hexyl formamide

In general, vapor pressure of materials changes exponentially with respect to

temperature. Therefore the data represented in Figure 3.3 was fitted to an exponential
curve. The R2 value was found as 0.98. The equation of the curve is given below;

𝑃(𝑘𝑃𝑎) = 3 ∗ 10−9 ∗ exp(0.0515 ∗ 𝑇 (𝐾)) Eqn. 1

43
Equation 1 was used to estimate the vapor pressure of HeFo at room temperature
and it was found as 0.016 kPa at 25 oC. With the same equation the normal boiling
point of HeFo was estimated as 243 oC. This is such a high temperature that hexyl
formamide will probably decompose before it evaporates even at vacuum conditions.
This high boiling temperature also implies a high energy for evaporation of the
solvent, which can be required in recovering the solvent after the extraction.

3.1.1.6 The solubility of lithium chloride and foreign ions in hexyl formamide

The solubilities of lithium chloride and foreign ions in hexyl formamide were
measured in order to estimate and predict the selectivity of HeFo towards lithium.
The concentration of ions in HeFo was found by dissolving and diluting HeFo
mixture in water. The analysis was done by atomic absorption spectrophotometer.
Results were given in Table 3.3.

Table 3.3 LiCl, NaCl and KCl solubilities in hexyl formamide

Salt Solubility (g/100 ml HeFo)

Lithium chloride 6.94
Sodium chloride 0.47
Potassium chloride 0.19

3.1.2 Synthesis and characterization of lithium manganese oxide

Lithium manganese oxide adsorbents were synthesized with five different

procedures having different precursors throughout the study in order to find the
LiMnO with the highest capacity value. The methods used in the study can be
generally grouped into two as hydrothermal and solid-solid synthesis methods. In
hydrothermal method, the procedures of Chitrakar et al, Wang et al, and Zhang et al
were studied [70], [84], [88]. The procedures were slightly modified and
optimization of the resulting capacity values was done. In solid-solid synthesis, the
methods based on Kitajou et al and Ooi et al were studied [79], [87]. Table 3.4
44
shows Li, Mn sources and oxidants employed in synthesis of lithium manganese
oxide adsorbents throughout the study.

Table 3.4 Methods employed in synthesis of lithium manganese oxide adsorbents.

Manganese
Method Lithium source Oxidant Reference
source

Hydrothermal-1 LiOHH2O MnCl2 H2O2 [84]

Hydrothermal-2 LiOH MnO2 H2O2 [70]

Hydrothermal-3 LiOHH2O Mn(NO3)2 H2O2 [88]

Solid-Solid-1 LiOHH2O Mn3O4 - [79]

MnCO3, MnO,
Solid-Solid-2 LiOHH2O, - [87]
MnO2

It is important to note that, there are additional procedures described with reflux
method and spray pyrolysis method. Reflux method is an alternative method for
hydrothermal method and spray pyrolysis method was employed only for once in a
study made in Turkey [86].

3.1.2.1 Adsorbents synthesized by Hydrothermal-1 procedure

Among the hydrothermal methods, Hydrothermal-1 gave the adsorbent with the
highest adsorption values and studied in detail. In this method LiOH and MnCl2
were reacted to form a gel which was then transferred into an autoclave. The
autoclave was taken to high temperature oven for reaction. After that, the precipitate
was filtered and dried in room temperature and taken into high temperature oven to
finalize the reaction. The whole process was divided in ten discrete steps in order to
analyze the effect of each parameter on the resulting product efficiency as shown in
Figure 3.4.

45
Figure 3.4 Hydrothermal-1 procedure in ten steps
3.96 𝑔 𝑀𝑛𝐶𝑙24𝐻2𝑂
([S1]: 2 M LİOH solution, [S2]: aq solution, [S3]: 30 % H2O2 (aq), [S4]: 5 ppm Li
10 𝑚𝑙

includes 4 ml desired pH buffer (aq), [S5]: 0.5 M HCl solution (aq))

Among those ten steps, critical variables were chosen that may affect the resulting
adsorbent efficiency and the remaining variables were fixed. Critical variables were
chosen as,

46
  Li/Mn ratio: 1, 2, 4 (STEP 1)
 Gel/water ratio in autoclave: 0.3, 0.6, 1.0 (STEP 4)
 Autoclave temperature: 120, 150, 180 oC (STEP 4)
 Oven temperature after autoclave: 350, 450, 650 oC (STEP 6)
In order to quantify the effect of each parameter, a factorial experimental design was
done [93]. If three different variables were analyzed in each step, there comes out to
be around 34=81 experiments that should be performed. If repetitions were
considered, there were 162 experiments that should be conducted which would be an
excessive hard work. In order to reduce the number of experiments, the upper and
lower limits for each variable were taken into account and 24=16 experiments were
performed. According to that plan, Table 3.5 was constructed and a set of 16
experiments were conducted.

Table 3.5 Optimization plan constructed by applying factorial design for the upper
and lower limits of four variables versus acquired adsorbent capacity (response)

Sample Li/Mn Gel/water Autoclave temp. Oven temp. Adsorbent

no fraction ratio (oC) (oC) Capacity
0-40 (mg Li/g
Limits 1-4 0.3-1.0 120-180 350-650
ads)
1 1 0.3 120 350 2.1
2 4 0.3 120 350 4.8
3 1 1 120 350 3.2
4 4 1 120 350 6.5
5 1 0.3 180 350 5.1
6 4 0.3 180 350 7.0
7 1 1 180 350 4.8
8 4 1 180 350 8.1
9 1 0.3 120 650 2.5
10 4 0.3 120 650 4.9
11 1 1 120 650 4.4
12 4 1 120 650 13.3
13 1 0.3 180 650 6.9
14 4 0.3 180 650 9.2
15 1 1 180 650 5.8
16 4 1 180 650 7.3

47
The data listed in Table 3.5 enlightens the effect of variables in lithium manganese
oxide synthesis on adsorbent capacity. It can be deduced that the most effective
variables are the oven temperature and Li/Mn ratio. The next effective variable is
gel/water ratio in autoclave. The effect of autoclave temperature has the smallest
effect on capacity values.

After the effect of each variable was roughly determined, a detailed experimental
investigation was held to identify the effect of each individual variable by additional
experiments.

First, the effect of oven temperature was examined as shown in Table 3.6.

Table 3.6 Effect of oven temperature on capacity of adsorbents synthesized by

Hydrothermal-1 procedure

Adsorbent synthesis conditions Adsorption process conditions

Capacity
Oven Autoclave Initial Li Final Li
Gel/water Li/Mn 𝐦𝐠 𝐋𝐢
temperature temperature concentration concentration pH ( )
ratio fraction 𝐠 𝐚𝐝𝐬.
(oC) (oC) (ppm) (ppm)
350 1 120 4 5 3.52 10.1 6.53
450 1 120 4 5 2.59 10.1 11.78
650 1 120 4 5 2.22 10.1 13.34
750 1 120 4 5 3.65 10.1 6.48

Table 3.6 shows that as oven temperature increases, the capacity value of the
adsorbent also increases up to 650 oC. After that a lower capacity value was acquired
at 750 oC. It is known that the thermodynamic equilibrium for spinel manganese
oxide is established around 450-550 oC [94]. Therefore at 750 oC, the crystal
structure may be distorted and the capacity values get decreased.

Second, the effect of gel water ratio in autoclave was investigated and the results
were tabulated in Table 3.7.

48
Table 3.7 Effect of gel/water ratio on capacity of adsorbents synthesized by
Hydrothermal-1 procedure

Adsorbent synthesis conditions Adsorption process conditions

Capacity
Oven Autoclave Initial Li Final Li
Gel/water Li/Mn 𝐦𝐠 𝐋𝐢
temperature temperature concentration concentration pH ( )
ratio fraction 𝐠 𝐚𝐝𝐬.
(oC) (oC) (ppm) (ppm)
0.3 650 120 4 5 4.83 10.1 4.9
0.7 650 120 4 5 3.00 10.1 9.98
1 650 120 4 5 2.22 10.1 13.34

It can be shown from Table 3.7 that, as gel/water ratio in autoclave decreases, the
capacity values of the adsorbents also decrease. The effect of water amount in
autoclave on adsorption capacity of the resulting adsorbents can be explained by the
concentration of the reactants in the medium. More concentrated mixtures give
higher extent of reaction which result in higher amount of desired product.

Third, the effect of Li/Mn ratio was investigated in another set of experiments. The
results were listed in Table 3.8.

Table 3.8 Effect of Li/Mn ratio on capacity of adsorbents synthesized by

Hydrothermal-1 procedure

Adsorbent synthesis conditions Adsorption process conditions

Capacity
Oven Autoclave Initial Li Final Li
Li/Mn Gel/water 𝐦𝐠 𝐋𝐢
temperature temperature concentration concentration pH ( )
fraction ratio 𝐠 𝐚𝐝𝐬.
(oC) (oC) (ppm) (ppm)
4 650 120 1 5 2.22 10.1 13.34
2 650 120 1 5 3.65 10.1 6.48
1 650 120 1 5 4.09 10.1 4.35

It is seen that the Li/Mn ratio has a remarkable effect on the resulting capacity
values. The best result was acquired with Li/Mn=4 in which lithium ion presence in
the crystal lattice has a highest value. It is also important to note that as lithium ion
increases in the crystal lattice, the stability of the absorbents tends to reduce in
consecutive adsorption-desorption cycles.

49
Lastly, the effect of autoclave temperature was investigated and findings were
tabulated in Table 3.9.

Table 3.9 Effect of autoclave temperature on capacity of adsorbents synthesized by

Hydrothermal-1 procedure

Adsorbent synthesis conditions Adsorption process conditions

Capacity
Autoclave Oven Initial Li Final Li
Gel/water Li/Mn 𝐦𝐠 𝐋𝐢
temperature temperature concentration concentration pH ( )
ratio fraction 𝐠 𝐚𝐝𝐬.
(oC) (oC) (ppm) (ppm)
110 1 650 4 5 2.84 10.1 11.78
120 1 650 4 5 2.22 10.1 13.34
130 1 650 4 5 3.00 10.1 9.98
140 1 650 4 5 2.84 10.1 11.78
150 1 650 4 5 3.52 10.1 6.79
180 1 650 4 5 3.48 10.1 7.34

Table 3.9 shows that lower autoclave temperatures result in higher capacity values.
The temperature in an aqueous autoclave system can also affect the pressure of the
media. The pressure of the system reaches to vapor pressure of water at specified
temperature. The capacity loss at high temperature and pressures may be attributed
to the nature of MnCl2 and LiOH reaction but could not be identified explicitly.

Considering the experiments conducted to optimize the capacity of adsorbents

synthesized by Hydrothermal-1 procedure, the highest capacity value of 13.3 mg
Li/g ads. was acquired at gel/water ratio=1, oven temperature=650, Li/Mn fraction=
4 and autoclave temperature=120. This value corresponds to 33 % of the maximum
theoretical adsorption capacity (40 mg Li+/g adsorbent) of the lithium manganese
oxide adsorbents. The value of 13.3 mg Li/g implies that one third of the active sides
are available for lithium adsorption in LiMnO and the remaining sides were not
reached by lithium ions due to mass transfer limitations or the synthesis was failed in
some extent.

In order to get rid of mass transfer limitation in the particles and make the particles
more porous, the lithium manganese oxide adsorbents was synthesized with
Pluronics which was a polyethylene glycol-poly propylene glycol-polyethylene

50
glycol copolymer. The optimum synthesis conditions were chosen and 100 ml of
0.85 % Pluronics solution was added to the reaction mixture in autoclave stage
instead of water. Unfortunately the resulting adsorbent gave a capacity value which
was even smaller than the previous one as 2.53 mg Li/g adsorbent.

3.1.2.2 Adsorbents synthesized by solid-solid procedures

In solid-solid procedures, different manganese and lithium sources were mixed in a

crucible and heated to certain temperatures where the reaction takes place.

In order to find the most efficient solid-solid method, several characterization

techniques were employed. The characterization of the adsorbent synthesis was first
done in DTA-TGA instrument in order to identify the behavior of the reactants
during heating process.

- DTA-TGA measurements

Thermogravimetric analysis (TGA) and differential thermal analysis (DTA) were

performed simultaneously in air at a heating rate of 10 °C min−1. Four different
reactants were used in the synthesis as manganese source which were, MnO, MnO2,
Mn3O4 and MnCO3. The lithium source was fixed as LiOHH2O.

As seen in Figure 3.5, MnO reacts with LiOH.H2O under increasing temperature and
given heat. It is seen that, at 85 oC, the water in medium evaporates. The
endothermic peaks at around 428 oC and 476 oC are attributed to the convertion of β-
MnO2 to a more stable α-Mn2O3 with release of oxygen gas [82].

51
Figure 3.5 Heating curve for MnO and LiOHH2O reaction system

Figure 3.6 shows that, MnO2 reacts with LiOH under increasing temperature and
given heat. It is seen that, the behavior of the curve is very similar to the case in
Figure 3.5. At 96 oC, the water in medium evaporates. The endothermic peak at
around 428 oC and 467 oC are attributed to the convertion of β-MnO2 to a more
stable α-Mn2O3 with release of oxygen gas.

Figure 3.6 Heating curve for MnO2 and LiOHH2O reaction system
52
In Figure 3.7, MnCO3 reacts with LiOHH2O under increasing temperature and given
o
heat, the weight loss around 78 C implies the evaporation of water. The
endothermic peak at 339 oC and weight loss can be ascribed to the decomposition of
manganese carbonate to manganese oxide and carbon dioxide. The peak around 725
o
C was attributed to transformation from Mn2O3 to Mn3O4.

Figure 3.7 Heating curve for MnCO3 and LiOHH2O reaction system

In Figure 3.8, Mn3O4 reacts with LiOHH2O under increasing temperature and given
heat. It is seen that, the behavior of the curve is very similar to Figure 3.7. The
weight loss around 79 oC implies the evaporation of water. The endothermic peak at
around 428 oC and 471 oC are attributed to the convertion of β-MnO2 to a more
stable α-Mn2O3 with release of oxygen gas.

53
Figure 3.8 Heating curve for Mn3O4 and LiOHH2O reaction system

After the synthesis of four different adsorbents by solid-solid synthesis method, it

was seen that adsorbents synthesized by manganese carbonate and lithium hydroxide
gave the highest uptake value which is 22.8 mg Li/g ads. The high performance of
those adsorbents was attributed to the micron scale particle size. The carbonate
group attached to manganese decomposes at around 340 oC and CO2 evolves.
Carbon dioxide evolution during synthesis results in a porous structure and decreases
the particle size of the resulting products. Smaller particle size leads to increase in
surface area which results in high capacity for the synthesized adsorbents.

- FTIR measurements

In order to identify the structure of synthesized adsorbents and investigate the effect
of Li/Mn molar ratio on structure before and after acid leaching, FTIR measurements
were carried out. Acid leaching was the last step in adsorbent preparation for all
methods and function to exchange the hydrogen ions with lithium ions. Exchange of
H+ and Li+ is a reversible process and do not distort the crystal structure. As seen in
Figure 3.9, conservation of the structure after acid leaching is one of the fingerprints
that validate successful synthesis. Acid leaching was done by dilute hydrochloric
acid (0.5 M) solution for all adsorbents. The adsorbents before acid leaching were
named as LMO and after acid leaching as HMO.
54
 72,8

70 HMO

65
910,21

60 1627,01

LMO
55

531,34
1631,00
50 3350,29 619,11
%T

45
3389,38

40

35 515,13

647,39

30

27,9
4000,0 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 600 450,0
cm-1

Figure 3.9 FTIR spectrum of spinel lithium manganese oxide synthesized by MnCO3
and LiOHH2O (Li/Mn: 1/1) F30, before (LMO) and after (HMO) acid leaching

In the spectrum, spinel type lithium manganese oxide peaks can be identified as
fingerprint of the adsorbents. The peaks around 3400 and 1600 cm-1 are belong to
water in samples (stretching vibration of hydroxyl groups). The peaks around 630
and 500 cm-1 are belong to manganese-oxygen bonds. The peak around 910 cm-1 is
ascribed to the coupled lattice vibration of protons which shows the completion of
acid leaching process in the adsorbent. Peaks around 2900 cm-1 probably show the
organic residue present in the FTIR cell or optic pathway in the instrument. The
remaining peaks do not change much which imply that the lattice structure of the
adsorbents is preserved before and after acid leaching which also validates the
tapotactic extraction of lithium ions from crystal lattice.

In Figure 3.10, FTIR pattern of spinel lithium manganese oxide before and after acid
leaching can be seen for Li/Mn:1/0.8 molar ratio.

55
 HMO

799,86

1621,26

%T
LMO 3379,99

537,78

1632,84 623,28

3393,32

515,55

643,20
4000,0 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 600 450,0
cm-1

Figure 3.10 FTIR spectrum of spinel lithium manganese oxide synthesized by

MnCO3 and LiOHH2O (Li/Mn: 1/0.8 mole ratio) F31, before (LMO) and after
(HMO) acid leaching

The spectrum shown in Figure 3.10 depicts the loss of structure after acid leaching.
The water in the structure gives signals at 3400 cm-1 and 1600 cm-1. Similarly, the
characteristic peaks of Mn-O stretching are observed at 640 and 530 cm-1. However
the diversity in the rest of the spectrum implies the loss of crystal structure. The
proton signal expected at 910 cm-1 is also distorted and additional signals appear.
The brown color appears in the mixture of adsorbent and lithium chloride solution
also validates the loss of structure. The brown color emerges by decomposition of
Mn+3 to Mn+4.

56
 HMO

1621,12

909,75

%T
LMO
3338,83 526,92

611,13
1621,41

3368,86

648,16
515,23

4000,0 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 600 450,0
cm-1

Figure 3.11 FTIR spectrum of spinel lithium manganese oxide synthesized via
MnCO3 and LiOHH2O (Li/Mn: 0.8/1 mole ratio) F32, before (LMO) and after
(HMO) acid leaching

In Figure 3.11, Mn-O signals at 640 cm-1 and 530 cm-1 and water signals at 3400
cm-1 and 1600 cm-1 are clearly noticed. Two trends in spectrum before and after acid
leaching are quiet similar. Also proton signal around 910 cm-1 is much more obvious
and sharp with respect to former adsorbents which implies higher hydrogen content.
In fact, the strength of this signal implies a higher lithium capacity for the
synthesized adsorbent. However the higher capacity comes in return for loosening
the structure stability. The adsorbent capacity gets lost after two lithium adsorption-
desorption cycle.

The adsorption experiments with the adsorbents synthesized with three different
molar ratios showed that the highest capacity was acquired by the adsorbent having
Li/Mn ratio of 0.8. However the stability of the adsorbent weakens after first
adsorption-desorption cycle while the adsorbent having Li/Mn ratio of 1 lasts for
several cycles.

The FTIR pattern for three adsorbents having different Li/Mn ratio (after acid
leaching) were summarized in the same plot as shown in Figure 3.12.

57
 F30-HMO

F31-HMO 1624,14 910,82

3341,93 2056,91 531,18

1621,26 619,41
908,14

%T

F32-HMO
3379,99

1621,12
909,75 623,28

526,92
3338,83
611,13

4000,0 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 600 450,0
cm-1

Figure 3.12 FTIR pattern for three different adsorbents (after acid leaching) having
different Li/Mn molar ratios(F30-Li/Mn:1, F31-Li/Mn:1/0.8, F32-Li/Mn:0.8)

During the adsorption experiments, it was found out that, mixing the adsorbents with
the acid leaching solution for a long time leads to an irreversible effect and the
hydrogen atoms in the structure lose the position in lattice structure. This fact was
proved by the FTIR measurements which can be seen in Figure 3.13. The signal at
905 cm-1 slowly grows up to 4 hours. After that it disappears during acid leaching,
which implies the integration of protons into lattice. After 24 hours the signal totally
gets lost and the resulting adsorbent becomes useless to be employed in lithium
separation. Therefore it was understood that, there is an optimum time of acid
leaching process for the adsorbents synthesized by Hydrothermal-1 procedure.

58
 1h

1630,18 905,23

3350,38
2h

4h 906,04
1624,17

3343,95
611,21
%T 535,13
1624,24 905,95

3341,97

535,12
611,27

24 h

1628,98 531,24
611,11
3430,10 535,27

610,97
4000,0 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 600 450,0
cm-1

Figure 3.13 Effect of acid leaching process on lattice proton at different times for
adsorbents synthesized by Hydrothermal-1 procedure

The FTIR pattern of adsorbents synthesized by procedures Solid-Solid-2,

Hydrothermal-1 and Hydrothermal-3 are given in Figure 3.14.

Figure 3.14 The adsorbents synthesized with three different methods

As it was observed from Figure 3.14, the three spectra for three different adsorbents
were very similar. Mn-O signals were observed at the same wavelength. The signal
59
at 910 cm-1, which shows the coupled lattice vibrations of protons in the structure,
Hydrothermal-3 adsorbents differ from others. This also explains why these
adsorbents generally have very low adsorption capacities, around 3-4 mg/g.
Hydrothermal-1 (Capacity: 13.3 mg Li/g ) and Solid-Solid-2 (Capacity: 21.1 mg
Li/g) procedures give nearly exact spectrum which implies the similar structure. At
that point, the difference of the capacity values between two adsorbents was
explained with the particle size and the porosity difference between the adsorbents.
In solid-solid synthesis, manganese carbonate decomposition into carbon dioxide
leads to a more porous structure and smaller particle size in microns which produce
higher surface area and increasing capacity. The detailed capacity values of the
adsorbents synthesized by Solid-Solid-2 method are given in Chapter 3.3.

- XRD measurements

After the analysis of the adsorbent synthesis procedures listed in Table 3.4, Solid-
Solid-2 method was found as the most efficient method which leads to adsorbents
with highest lithium uptake value. In order to increase the performance of the
adsorbents, a set of experiments were conducted under different synthesis
temperatures and Li/Mn molar ratios. 350, 450, 550 and 650 oC were chosen as
reaction temperatures and 0.8, 1 and 1.25 were chosen as Li/Mn ratios. The XRD
pattern for the synthesized adsorbents were taken both in lithium manganese oxide
(LMO) and hydrogen manganese oxide (HMO) forms. It was found out that, Li/Mn
ratio of 1 and 450 oC synthesis temperature gave the most efficient adsorbents.

The XRD pattern of the selected adsorbents are shown in Figure 3.15 and Figure
3.16 for LMO and HMO forms respectively. The permanent signals before and after
acid leaching show the stability of crystal structure which implies tapotactic
exchange of lithium and protons in crystal lattice.

60
 Intensity (cps) Intensity (cps) Intensity (cps) Intensity (cps)

10

Li/Mn=1)
Li/Mn=1)
0

-76
-26
24
74
50
100
150
200
250
300
350
400
450

10
10

-58
-18
22
50
100
150
200
250
300
350

10
[1], d=4.717(5), 2-theta=18.795(19)

20
[1], d=4.668(4), 2-theta=18.996(17) 20

20
20
30
30

30
30
2-theta (deg)

2-theta (deg)
[2], d=2.457(3), 2-theta=36.54(5)
[2], d=2.443(3), 2-theta=36.77(5)
40
40

40
40
[3], d=2.0229(8), 2-theta=44.763(18)
[3], d=2.0242(11), 2-theta=44.73(3)

[4], d=1.868(4), 2-theta=48.70(11)

50
50

[4], d=1.854(4), 2-theta=49.09(11)

50
50
60
60

60
60

Figure 3.16 XRD spectrum of the spinel lithium manganese oxide HMO (450 oC,
Figure 3.15 XRD spectrum of the spinel lithium manganese oxide LMO (450 oC,

61
The crystal system and lattice constant of the selected lithium manganese oxide was
found by applying Sadoway’s method. There are two equations that was used in
calculations.

First equation is Braggs’ Law,

𝜆 = 2𝑑𝑠𝑖𝑛𝜃 Eqn. 2

λ: Wavelength of the cupper target

d: Spacing between the planes in the atomic lattice

θ: Angle between the incident ray and the scattering planes

Second equation relates lattice constant with d,

a
𝑑 = √h2 Eqn. 3
+𝑘 2 +l2

From those two equations, following equation is acquired,

𝜆2 sin2 𝜃
= Eqn. 4
4𝑎2 h2 +𝑘 2 +l2

Here the left side of the equation is constant, which means that the right side of the
equation must be a constant value too.

Wavelength of the cupper target is known as 1.5418 Å.

The calculations to find the crystal system of manganese oxide and lattice constant
are given in Table 3.10.

62
Table 3.10 Finding crystal system and lattice constant of LiMnO

Clear
Compute Compute a
Normalize fractions by
(hkl)
2θ Sin2 θ values with multiplying
𝒔𝒊𝒏𝟐 𝜽 𝝀𝟐 𝒔𝒊𝒏𝟐 𝜽
first value with 3 (for =
𝒉𝟐 + 𝒌𝟐 + 𝒍𝟐 𝟒𝒂𝟐 𝒉𝟐 + 𝒌𝟐 + 𝒍𝟐
this case)
18.7 0.0264 1 3 (111) 0.0088 8.22
36.5 0.098 3.712 11.136 (311) 0.0089 8.17
44.7 0.145 5.492 16.477 (400) 0.0088 8.22
48.7 0.17 6.439 19.318 (331) 0.0088 8.22
58.5 0.239 9.053 27.159 (333) 0.0088 8.22

The planes of (111), (311), (400), (331) and (333) listed in Table 3.10, imply a FCC
crystal system, with a lattice constant of 8.21 Å. These results coincide with the
XRD data given in the literature [74] in which lithium at tetrahedral sites, manganese
(III) and manganese (IV) at octahedral sites of a face centered cubic crystal system
with 8.21 Å lattice constant.

The XRD pattern of adsorbents having Li/Mn ratio different than 1 also gives similar
pattern and it was found that, the molar ratio changes of Li/Mn does not affect the
cyrstal structure significantly. In that manner adsorbents having Li/Mn ratio higher
than 1 can give higher lithium content in the adsorbent and increase the capacity of
lithium uptake. Unfortunately, in that case the stability of the adsorbent was found
insufficient which was indicated by the blurred solutions due to dissolved manganese
ions during adsorption process.

The adsorbents synthesized at 650 and 750 oC give almost the same peaks in pattern.
However peaks have high amount of noise and red colored impurities in crucible
were observed after the synthesis. At temperatures 450 and 550 oC, the synthesis of
manganese oxides give almost the same peaks. However peaks for 550 oC have less
amount of noise. It is concluded that both temperatures are suitable for synthesis. In
terms of lower energy consumption, 450 oC was used in the final adsorbents. The
XRD pattern for the adsorbents synthesized at different temperature and molar ratios
are given in APPENDIX B.

63
3.1.3 Synthesis and characterization of lithium manganese oxide-polymer
beads

In order to employ the synthesized lithium manganese oxide adsorbents in an

adsorption column, the size of particles must be increased from microns to
millimeters to overcome the large pressure drop. Inorganic supporting materials like
silica and alumina are used to address this problem. In general, the active material
was mixed with supporting material in a mixture and impregnated or coated on the
supporting material by various methods [95]. In this study these methods were found
insufficient in impregnating the lithium manganese oxide onto silica or alumina.
Even the synthesis of lithium manganese oxides were done on targeted surfaces, but
the expected adhesion could not be observed.

A suspension polymerization technique was used in order to overcome this problem

and lithium manganese oxide particles are tried to be covered by a polymeric matrix.
In this technique, an organic phase and aqueous phase were prepared separately. In
organic phase, glycidyl methacrylate (GMA), methyl methacrylate, ethylene glycol
dimethacrylate and AIBN were dissolved in toluene. In aqueous phase, water,
sodium sulfate, polyvinyl pyrrolidone were dissolved. Lithium manganese oxide
particles were added to organic phase and then dripped into aqueous phase slowly at
65 oC.

After polymerization reaction was completed, lithium manganese oxides were

weighed and it was found as 0.85 g of polymer covers 1 g of lithium manganese
oxides, which shows that the polymerization reaction works and covers the LiMnO
particles. Unfortunately the size of the manganese oxide-polymer particles at the end
of the reaction was still in micron level. In order to increase the polymer content
around the particles, an additional synthesis was done where the monomer amount
was increased by five folds. But the polymer around manganese oxides was found as
1.5 g per 1 g of MnO2 and the size of the particles were still in microns. Therefore
this method was found insufficient to increase the particle size of the lithium
manganese oxides.

As an alternative, a precipitation method was used. In this method, a polymer binder,

targeted material and solvent were mixed to form a slurry of suspended particles in
64
the mixture. Then, this mixture was dripped into a non-solvent liquid to form the
desired beads. In this method, various couples of solvent-polymer were used to find
out the optimum combination for bead formation. Different solvent-polymer pairs
and the observations on the resulting beads were listed in Table 3.11.

Table 3.11 Different solvent-polymer couples utilized for LiMnO bead synthesis by
precipitation method

Polymer Solvent Result

Poly vinyl alcohol NMP Beads dissolve in water
Poly MMA-co-BA NMP Does not give ordinary spheres
Poly MMA-co-BA THF Does not give ordinary spheres
Poly MMA-co-BA DMF Does not give ordinary spheres
Poly MMA-co-BA Dioxane Does not give ordinary spheres
Poly MMA-co-BA DMSO Does not give ordinary spheres
PET NMP Polymer is not dissolved
PET THF Polymer is not dissolved
NMP-THF-DMF-
PE Polymer is not dissolved
Dioxane-DMSO
PS NMP Beads are formed but does not show stability
PS-Br NMP Beads are formed but does not show stability
Poly urethane NMP Does not give ordinary spheres
Poly vinyl acetate NMP Beads are formed but does not show stability
Beads are formed and show stability in water,
Poly (styrene-maleic
NMP but lose their shape in air.
anhydride)
In water high pH dissolves beads.

Among the polymer-solvent couples listed in Table 3.11, most promising results
were acquired for poly (styrene-maleic anhydride). However, beads lost their form in
air and high pHs (above 8). In order to enhance the bead stability, poly glycidyl
methacrylate polymer was synthesized and added to poly (styrene-maleic anhydride)
before mixing with LiMnO. The idea behind that is to introduce an epoxy group to
polymeric medium and form crosslinkings which will increase the stability of the
beads. The epoxide groups are known to open at high temperature and acidic
medium. Therefore the slurry was dripped into 0.5 M HCl at 90 oC. At first, the
beads changed color from black to white and gain stability in atmosphere. However,

65
they were still unstable in high pH aqueous solution. They swell and lose their
stability. In order to overcome this problem, terpolymerization of maleic anhydride,
styrene and glycidyl methacrylate were done. The logic is to introduce epoxy groups
into the polymeric backbone and finally increase stability and strength.

3.1.3.1 Synthesis of poly (styrene-maleic anhydride-glycidyl methacrylate)-

lithium manganese oxide beads

Poly (styrene-maleic anhydride-glycidyl methacrylate) was first synthesized in 1999

by a visiting professor from Azerbaijan Academy of Sciences in METU Chemistry
department. The product was proposed to be used in films and coatings owing to its
good mechanical and dielectric properties. In the polymerization reaction, all
monomers were dissolved in ethyl methyl ketone and put into a reaction balloon to
polymerize by benzoyl peroxide initiator [90]. Styrene and maleic anhydride
monomers are known to lead to an alternating polymer which is called SMA, and it
is used in paper sizing, powder coating, pigment dispersions etc. [96]. It also shows
strong rubbery behavior and high elasticity. In order to sustain rubbery property of
the polymer, styrene and maleic anhydride were used in equal molar proportions and
GMA ratio was changed from 5 to 20 % in this study. After the polymerization
reaction, poly maleic anhydride-styrene-glycidyl methacrylate (PSMA), MnO2 and
NMP was mixed in certain proportion and spherical beads were acquired by dripping
the slurry into water. Water which was used as a non-solvent was also mixed with
cetyl tert butyl ammonium bromide (CTAB) which was a well-known surfactant, in
order to decrease the surface tension of the water and ease the formation of the
spherical beads. The epoxy groups in resulting beads were known to form crosslinks
with maleic anhydride groups even by the humid in the air at 125 oC or UV light
[90]. The resulting beads were first dried and then exposed to UV light for 24 hours.
Another set of beads were put into an oven at 125 oC for 40 minutes directly. The
beads which were heat treated gave more consistent results; therefore heat treatment
of beads was preferred rather than UV treatment to have a crosslinked product. The
resulting PSMA polymer has a structure shown in Figure 3.17.

66
Figure 3.17 Structure of poly maleic anhydride-styrene-glycidy methacrlate (PSMA)

Optimization of GMA ratio in PSMA to have a more rigid and stable beads, a set of
polymerization reactions were carried out. The molar ratio of the monomers and the
properties of the synthesized beads are given in Table 3.12.

Table 3.12 PSMA beads synthesized with different GMA content at 70 oC (Benzyl
peroxide molar ratio:0.01, monomer/solvent ratio: 0.33)

Maleic Properties of the resulting beads

Styrene GMA
anhydride
(mol. %) (mol. %) Rigidity Swelling at pH:9
(mol. %)
47.5 47.5 5 Low High
45 45 10 Medium High
44 44 12 Medium Medium
42.5 42.5 15 High Low
42 42 16 High Low
41.5 41.5 17 Form crosslinks before bead formation
40 40 20 Form crosslinks before bead formation

Table 3.12 shows that as GMA ratio increases in the polymer, the rigidity of the
resulting bead increases and its swelling amount decreases up to 16 % molar ratio.
At higher GMA content, polymer forms crosslinks during bead preparation which
disrupts the spherical shape of the beads. Therefore 16 percent GMA molar ratio was
chosen to be an optimum value to prepare manganese oxide-PSMA beads. The
volume of the beads increases by 4 % at pH 9 while keeping their form and stability.

67
In bead preparation process, manganese oxide/PSMA ratio, NMP/PSMA ratio and
the mixing time for the beads in water were assumed to be critical variables in order
to get optimum bead strength, lithium adsorption capacity and stability. The effect of
water temperature, in which the beads are dripped into and CTAB concentration in
water have minor contribution to the resulting form of the beads. The mixing time of
24 hours was fixed and used for all bead synthesis. Manganese oxide/PSMA ratio
affects the final manganese oxide content of the beads, which directly affect the
lithium adsorption capacity. In order to have the largest manganese oxide ratio in the
resulting beads, the amount of the manganese oxide in the slurry was kept at
maximum and the slurry was saturated with MnO2. NMP/PSMA ratio affects the
viscosity of the slurry, shape and MnO2 content of the resulting beads. When NMP
amount is high, the beads are formed perfectly in spherical shape but the MnO2
content of the beads decreases. When NMP amount is low, the viscosity becomes too
high and it becomes difficult to form spherically shaped beads while dripping.
Therefore a value of 2 was found as optimum PSMA/NMP ratio after series of trials.
The resulting beads contain 51 % by weight manganese oxide.

PSMA polymer has never been used as a polymeric support material for any metal
oxide bead preparation before. So this is the first study which examines the bead
preparation by PSMA and its usage in lithium separation. In the previous methods,
researchers use PVC [97], chitin [79], polyurethane [98] and water glass [99] as a
binder, but there are two problems with those materials.

First problem was the weak strength (rigidity) of the resulting beads. The polymers
do not crosslink in those studies, which results in weak mechanical properties and
stability. On the other hand, eventual self-crosslinking property of PSMA in water
during bead formation makes it unique as a metal oxide support material.

Second problem is the hydrophobic structure of the beads, which limit the diffusion
of water into the beads and slow down the overall separation. On the contrary,
PSMA beads swell in alkaline medium which increases the diffusion of water and
ions into the beads and enhance the overall separation process. Self-crosslinking and
swelling properties of the PSMA make it superior than any polymer binder
employed up to now.
68
3.1.3.2 Characterization of PSMA content

The characterization of PSMA was done by FTIR and C-NMR techniques. FTIR
spectrum of PSMA polymer is shown in Figure 3.18. At 950 cm-1, O-H bending
signal is observed. At 1770 cm-1, C=O streching and at 1450 cm-1, C-H bending
signals are observed. The results coincide with the literature findings [90]. FTIR
spectrum is not enough to prove the structure of the polymer alone. It designates the
presence of some characteristic bonds in the structure. However, spectrum is
accepted as a fingerprint.

108

103

98
%T

93

88

83

78
0 500 1000 1500 2000 2500 3000 3500 4000 4500
Wavenumber (cm-1)

Figure 3.18 FTIR spectrum of poly (styrene-maleic anhydrid-glycidyl methacrylate)

terpolymer (PSMA)
13
C-NMR spectrum of PSMA is shown in Figure 3.19. The signal at 165 ppm
designates esteric carbonyl carbon in glycidyl methacrylate and maleic anydhride.
The large signal at 125 ppm belongs to aromatic carbons in styrene. Wide signal
between 5 and 60 ppm represents aliphatic carbons. Signal at 235 ppm shows the
remaining ethyl methyl ketone carbonyl carbons.

69
Figure 3.19 13C-NMR spectrum of PSMA

3.1.3.3 Characterization of the PSMA-LiMnO beads

3.1.3.3.1 SEM characterization

In order to observe the effect of contents in PSMA-LiMnO bead morphology, a set

of experiment was performed and the ingredients were changed according to Table
3.13.

70
Table 3.13 The amount of the components used in bead synthesis

% LiMnO in
Bead # PSMA (g) EMK (g) NMP (g) LiMnO (g)
resulting bead

T14 2.62 0.5 3 1.1 52

T15 3 2 1 1 51
T16 3 - 1.5 1 51
T17 3.52 2 3.52 6 81
T18 3.4 2.68 2.26 6 82
T19 3.1 4.12 1.03 6 83
T20 3.38 2.68 2.26 1.5 58
T21 2.67 1.92 1.78 3 74
T22 3.1 4.12 4.12 6 84

After the synthesized beads were analyzed physically, it was seen that T17 and T18
lose their sphere shapes and become lentil shaped. The high manganese oxide
content may induce that result. Bead T19 was precipitated like needles. Thin and
long particles were formed. This behavior was attributed to the lack of NMP in
slurry. Bead T20 was precipitated in leaf form rather than a sphere. The reason for
that behavior was probably the lack of manganese oxide content in mixture. Bead
T21 gave perfect sphere shapes. But unfortunately they coagulate in filter paper and
could not retain their shapes. Bead T22 look similar to lentil shape. But the lentils
were distracted. Bead T16 was the perfect ratio that we acquired in bead synthesis so
far. Well-shaped spheres were acquired with high stability.

The SEM photographs of selected beads are shown below for T16, T18, T21 and
T22. The remaining SEM photos are given in APPENDIX C.

71
Figure 3.20 SEM photographs for Bead T16

Figure 3.21 SEM photographs for Bead T16 after acid leaching

72
Figure 3.22 SEM photographs for Bead T18

Figure 3.23 SEM photographs for Bead T21

73
Figure 3.24 SEM photographs for Bead T22

SEM photographs describe that the beads have highly porous structure. The outer
surface of the beads seemed to be covered by a smooth polymer layer. This layer
holds the LiMnO particles inside the bead and let the diffusion of water and metal
ions inside and outside of the particle. The bead morphology before and after acid
leaching remains nearly same which designates the stability of bead structure
towards acidic and basic conditions. When the beads absorb water especially in basic
conditions by swelling, the porous structure inside the beads provide large surface
area for adsorption.

3.1.3.3.2 BET measurements

BET measurements were done with beads T14, T15, T16, T16 after acid leaching,
T18 and T21. The results were tabulated in Table 3.14

74
Table 3.14 BET results of PSMA-LiMnO beads

Cumulative
Multipoint BET Adsorption Pore
Bead # 2
Adsorption Pore
(m /g) Diameter (Å)
Volume (ml/g)
T 16 22.7 0.030 25.2
T 16 after acid
35.14 0.104 10.7
leaching
T18 27.2 0.057 10.6
T21 21.9 0.037 10.9
T 22 10.95 0.023 34.9
* BJH method was used in calculations of pore volume and diameter

Table 3.14 shows that, the surface area of the beads changes between 10 to 35 m2/g.
The maximum surface area was found as 35.14 m2/g for T16. It was also seen that
after acid leaching the surface area, pore volume and pore diameter increases which
produces additional mass transfer area.

After analyzing the visual observations during bead preparation and the results
acquired in SEM and BET, Bead T16 was chosen to be employed in column
experiments as it provides the perfect spherical shape, highest porosity, surface area,
strength and stability.

3.1.4 Acquisition of lithium containing brines from nature

In addition to artificial lithium solutions prepared, following lithium containing

aqueous solutions were obtained from nature.

3.1.4.1 Concentrated brine from Çamaltı Salina, İzmir

In Çamaltı Salina, 3.5 bome sea water is concentrated to approximately 30 bome in

large ponds by utilizing solar energy. Corresponding plant is harvesting halite
(NaCl) from the ponds for 3 months starting with September in every year. The
samples of concentrated sea water were taken from the last pond where the water is
finally pumped back to sea (it is the water with the highest salt concentration in
Salina). The water samples are pink in color due to a halophilic algea which lives in
75
it. The algea is the dominant species in the water, and does not allow growing any
other organic content. It was deduced that, calcium was absent in concentrated sea
water. It is because of the precipitation of calcium ions in the form of calcium
carbonate and sulfate during the evaporation in the ponds. The composition of the
concentrated brine in Çamaltı Salina is tabulated in Table 3.15.

Table 3.15 The composition of the concentrated brine in Çamaltı Salina

Concentration (ppm)
Metal ion
(mg/L)
Sodium 76750
Magnesium 21000
Calcium -
Potassium 11023
Lithium 1.25

3.1.4.2 Preparation of boron clay extract from Bigadiç

Bigadiç has the largest boron reserves in the world. After boron is produced in mine,
remaining residue (clay) is disposed to nature. The clay is called as ulexite and there
is around 3000 ppm lithium content in the clay, mostly bounded to silicates
(Li2Si2O5). The conventional process to extract lithium from the clay needs hot
temperature (900-1100 oC) roasting of clay with sodium sulfate which makes the
process expensive and brings alternative economical methods into consideration. In
order to take lithium into the solution directly, acid treatment is required. Mostly
hydrochloric acid and sulfuric acid are used. In this study, the acid quantity and type
used for leaching of lithium from the clays were optimized. In order to do that
optimization, different volumes of different acids were mixed with clay under
varying water content and time. The results were summarized in Table 3.16.

76
Table 3.16 Leaching of lithium ions from boron clay

Li Li
Li
Clay Acid/ Rxn. conten content Extracted
Exp Acid content in
amount water Time t in in Li
# type solution
(g) (v/v) (h) clay solution (%)
(g)
(g) (ppm)
1 H2SO4 50 15 /250 0.25 0.15 55 0.0275 18
2 H2SO4 50 30 /300 0.50 0.15 242 0.121 81
3 H2SO4 50 30/300 2 0.15 250 0.125 83
4 H2SO4 50 30/300 43 0.15 279 0.1395 93
5 H2SO4 50 45/250 2 0.15 283 0.1415 94
6 H2SO4 50 60/300 2 0.15 287 0.1435 96
7 HCl 50 100/120 2 0.15 267 0.1335 89
8 HCl 50 100/120 20 0.15 284 0.142 95

After Table 3.16 is analyzed, it is seen that above 90 % of lithium can be extracted
from clay. The remaining acid in the solution was found by 0.1 M NaOH titration
and it was decided that, maximum amount of lithium leached by using minimum
amount of acid was in the case of 30 ml H2SO4 and 300 ml water couple (Exp. 3).

The lithium solution acquired after filtering the solid content in the clay was very
acidic. In order to use the solution in adsorption experiments NH3 was added into the
solution. After addition of ammonia, a white precipitate formed which was filtered
afterwards. In this study the clay was treated with sulfuric acid prior to process in the
column in order to take lithium ions into the aqueous phase. The final concentration
of the clay extract is shown in Table 3.17

Table 3.17 The composition of the clay extract taken from Bigadiç

Metal ion Concentration (ppm)

Sodium 4880
Magnesium 15110
Potassium 530
Lithium 388

77
3.2 Liquid-liquid extraction by N-alkyl formamides

Lithium chloride extraction experiments were carried out using either lithium
chloride solutions in the absence or presence of foreign ions namely Na, K, Mg and
Ca. The distribution factor, k was used for comparison of their extractabilities and
the selectivities. The distribution factor (Nernst distribution coefficient) is defined as
follows,

𝑆𝑜𝑙𝑢𝑡𝑒 𝑎𝑚𝑜𝑢𝑛𝑡 𝑖𝑛 𝑒𝑥𝑡𝑟𝑎𝑐𝑡 𝑝ℎ𝑎𝑠𝑒 (𝐸)

𝐴𝑚𝑜𝑢𝑛𝑡 𝑜𝑓 𝑝ℎ𝑎𝑠𝑒 𝐸
𝐷𝑖𝑠𝑡𝑟𝑖𝑏𝑢𝑡𝑖𝑜𝑛 𝑐𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 (𝑘) =
𝑆𝑜𝑙𝑢𝑡𝑒 𝑎𝑚𝑜𝑢𝑛𝑡 𝑖𝑛 𝑟𝑎𝑓𝑓𝑖𝑛𝑎𝑡𝑒 𝑝ℎ𝑎𝑠𝑒 (𝑅)
𝐴𝑚𝑜𝑢𝑛𝑡 𝑜𝑓 𝑝ℎ𝑎𝑠𝑒 𝑅

Equal volume of water and formamide (10 mL of each) were used in all experiments
for the sake of simplicity.

In order to make precise estimation for the extractability from pure lithium chloride
solutions, the extraction experiments were performed at various pH and temperature
conditions.

Table 3.18 shows LiCl distribution coefficients of the formamides at pH: 4.0, 7.0 and
10.0. The results tabulated, imply that hexyl formamide is the best for the extraction
of LiCl at pH 4 and 10. The results also show that the formamides exhibit very little
changes of LiCl extraction in between pH 4 and 10.

The low pH dependency observed is somewhat surprising, because if the polar

formamide group involves in dipole-dipole interaction with lithium ion, one must
observe pH dependency for the extraction due to blocking of non-bonding electrons
of oxygen with proton. The reverse situation might be due to nearly equal affinity of
Li+ and protons towards formamide group.

The highest extraction capability of hexyl formamide is most probably due to its low
molecular weight which means maximum formamide content per unit mass.

78
Table 3.18 The effect of pH on distribution coefficient of LiCl for the formamides at
25 oC, 10 ml water/10 ml formamide ratio (The initial LiCl concentration was 1.4 M)

Formamide pH Distribution coefficient

4 0.14
Hexyl formamide 7 0.12
10 0.14
4 0.08
Octyl formamide 7 0.05
10 0.06
4 0.04
Dibutyl formamide 7 0.04
10 0.06
(1.4 M Li solution preparation: Dissolve 5.95 g of LiCl in 100 ml volumetric flask)

Table 3.19 represents LiCl distribution coefficients of the formamides at different

temperatures namely 10, 25, 40 oC. The results show that, distribution coefficients
increases with decreasing temperature. The extraction experiments could only
performed with dihexyl formamide at 40 oC.

Table 3.19 The effect of temperature on distribution coefficient values for different
formamides at pH:4, 10 ml water/10 ml formamide ratio, initial lithium
concentration of 1.4 M Li.

Formamide Temperature (oC) Distribution coefficient

10 0.16
Hexyl formamide 25 0.14
40 0.11
10 0.09
Octyl formamide 25 0.08
40 0.07
10 0.04
Dibutyl formamide 25 0.04
40 0.03
Dihexyl formamide 40 0

In order to understand the effect of temperature on extraction process, standard

enthalpy change (ΔHo) and standard entropy change (ΔSo) can be predicted using the
Van’t Hoff equation:
79
 ∆𝐻 0 ∆𝑆 0
𝑙𝑛𝑘 = − + Eqn. 5
𝑅𝑇 𝑅

The values of standard enthalpy change and entropy change were calculated from the
slope and intercept of Von’t Hoff plot respectively (See Figure 3.25). In the equation
k is the distribution coefficient, R is the universal gas constant and T is the absolute
temperature. The R2 values calculated in linear regression are 0.9631, 0.9958 and
0.7244 for hexyl formamide, octyl formamide and dibutyl formamide respectively.

Figure 3.25 Van’t Hoff Plots for hexyl formamide (a), octyl formamide (b), dibutyl
formamide (c)

The calculated values of ΔHo and ΔSo were tabulated in Table 3.20. The change in
enthalpy values in extraction was found negative which implies the process is
exothermic in nature. The increase in distribution coefficients with respect to
decreasing temperature confirms the findings. The negative values of ΔSo indicates
decrease in disorder in the system during extraction. Therefore the extraction of LiCl
with formamides is enthalpy driven process.

80
Table 3.20 ΔHo and ΔSo values calculated from Van’t Hoff plots

Formamide ΔHo (kJ/mol) ΔSo (J/mol.K)

Hexyl formamide -9.16 -47.40

Octyl formamide -6.16 -41.75

Dibutyl formamide -6.94 -50.89

The effects of initial lithium concentration and volume ratio of water to formamide
towards distribution coefficients were represented in Figure 3.26. The extracted
lithium was plotted against initial concentration at different formamide/water volume
ratios as depicted in Figure 3.26. It is clearly seen that the distribution coefficients do
not change with the volume ratios indicating typical Nernst extraction behavior. The
initial LiCl concentration however greatly effects on the distribution coefficient so
that when concentration is increased from 100 ppm to 10000 ppm, the extracted
lithium increases 80 percent. This can be explained in terms of high ionic strength
effect.

0,17
0,16
0,15
Distribution coefficient

0,14
0,13 1:1
0,12 1:2
0,11 1:3
0,10
0,09
0,08
0 2000 4000 6000 8000 10000
Initial lithium chloride concentration (ppm)

Figure 3.26 Effect of initial LiCl concentration on k of hexyl formamide at pH:4.

The plots were drawn for three different water/formamide volume ratios (1:1, 1:2,
1:3) at 10 oC.

The high concentration range (100-10000 ppm Li) covered here was considered to be
insufficient for the extraction of low concentrated solutions. For a better
81
understanding of the extraction from dilute solutions, a series of experiment were
carried out in which initial Li concentration was varied from 10 to 100 ppm. The
results obtained are given in Table 3.21.

Table 3.21 Equilibrium data between extract and raffinate phase for the formamide-
water-lithium chloride system at 25 oC, pH: 4 and 10 ml water/10 ml hexyl
formamide

Water phase Formamide phase

Li Water Formamide Li Water Formamide

(ppm) (%) (%) (ppm) (%) (%)
11.250 98.467 1.533 0.749 15.870 84.129
21.940 98.459 1.541 1.488 15.953 84.047
32.967 98.462 1.538 2.915 15.747 84.253
41.073 98.453 1.547 3.825 15.755 84.245
52.246 98.463 1.537 5.738 15.804 84.195
61.022 98.477 1.523 7.160 16.218 83.781
73.369 98.478 1.522 8.043 14.995 85.005
81.611 98.472 1.528 8.412 15.396 84.603
90.126 98.468 1.532 8.619 15.596 84.403
103.312 98.508 1.492 10.763 16.053 83.946

The data tabulated in Table 3.21 helps to identify the extraction equilibria more
precisely and construct a guide to a possible extraction column design.

Ternary system of hexyl formamide-water-lithium chloride is drawn in dilute lithium

chloride region. It is deduced that the formamide fraction in water phase and water
fraction in formamide phase do not vary much with respect to lithium concentration.
Therefore the resulting data was fitted to a two dimensional chart rather than a
ternary diagram and shown in Figure 3.27.

82
Figure 3.27 The lithium concentration in extract (formamide) and raffinate (water)
phases at 25 oC and pH:4 for 10 ml water/10 ml hexyl formamide

The lithium concentration in extract phase and raffinate phase can be fitted in a
linear equation with an R2 value of 0.978 in below.

𝑦 = 0.11𝑥 − 0.48 Eqn. 6

Where y is lithium concentration in formamide phase and x is the lithium

concentration in water phase, in ppm..

The behavior of the lithium uptake by hexyl formamide remains constant while
lithium concentration changes in dilute regime.

Selectivity:

Selectivity of hexyl formamide towards lithium ion in the presence of sodium,

potassium, calcium and magnesium ions was investigated with an artificially
concentrated sea water. The artificial sea water was prepared by following the
83
procedure in ASTM D1141. In seawater sample, sodium, potassium, magnesium and
calcium ions have the highest concentration with 10770, 380, 1290, 412 ppm
respectively. The lithium ion concentration is so small as around 0.17 ppm. In
Salinas, seawater is concentrated for ten times larger than its initial concentration.
Therefore, the artificial seawater used in selectivity measurements was evaporated at
25 oC until its weight gets 10 % of its initial weight. In the selectivity experiments,
temperature was kept constant at 25 oC and 10 ml hexyl formamide/ 10 ml water was
used. The experiments were repeated for acidic (pH:4) and basic (pH:10) conditions.
The distribution coefficient of HeFo towards lithium and foreign ions (Na, K, Mg,
Ca) with the final composition of artificially concentrated seawater were listed in
Table 3.22.

Table 3.22 Artificial concentrated seawater composition and the resulting

distribution coefficient values in selectivity experiments.

Artificial concentrated
sea water composition Distribution Distribution coefficient
Metal ion
coefficient (pH:4) (pH:10)
(mg/L)

Sodium 62810 0.058 0.048

Magnesium 6940 0.21 0.16

Calcium 3523 0.4 0.43

Potassium 5270 0.06 0.07

Lithium 2.1 0.12 0.11

The selectivity values which were found by dividing distribution coefficients into
each other were listed in Table 3.23.

84
Table 3.23 The selectivity values of hexyl formamide for lithium near foreign (Na,
K, Mg, Ca) ions in sea water.

pH:4 pH:9

SLi/Na SLi/Mg SLi/K SLi/Ca SLi/Na SLi/Mg SLi/K SLi/Ca

Selectivity 2.07 0.57 2.00 0.30 2.29 0.69 1.57 0.26

Table 3.22 and Table 3.23 imply that the selectivity of HeFo towards lithium
compared to potassium and sodium are higher than 1 but lower than 1 for calcium
and magnesium ions. The selectivity values do not change very much with respect to
changes in pH like in the case of distribution coefficients. Following the calculation
of the selectivity values for artificial sea water, selectivity experiments were
conducted by the real concentrated brine samples taken from Çamaltı Salina, İzmir.
The initial solution composition and acquired distribution coefficients were listed in
Table 3.24.

Table 3.24 Composition of the concentrated sea water taken from Çamaltı Salina and
resulting distribution coefficients (25 oC, 10 ml HeFo/10 ml water, pH:7.4)

Concentration (ppm) Distribution Selectivity values

Metal ion
(mg/L) coefficient (klithium/kmetal)
Sodium 76750 0.01 7.8
Magnesium 21000 0.09 0.86
Calcium - - -
Potassium 11023 0.01 7.8
Lithium 1.25 0.078 1

Table 3.24 shows that the distribution coefficient values are lower than the values in
the experiments conducted for the artificial sea water. The reason of decrease in the
values may be additional foreign ions present in sea water, which may compete with
lithium ions and reduce the extraction amount. Eventually the distribution coefficient
value was found as 0.078 for hexyl formamide, still larger than n-butanol, which
shows the solvation power of hexyl formamide.

85
Regeneration of formamides:

The regeneration of formamides was done by using hydrochloric acid solution at

different pH. After the extraction process, the resulting formamides were taken in a
separate flask and mixed with acid solution. Following the vigorous shaking at 25 oC
in a separation funnel, 1 hour was allowed in order to have phase separation
completely. The results were tabulated in Table 3.25.

Table 3.25 The effect of pH on recovery of lithium from hexyl formamide

Hexyl Acid C (b) pH of the

The percent
(a)
formamide Co (ppm) volume recovery of
solution
amount (g) (ml) (ppm) LiCl (%)

10 1.6 10 0.15 1 91

10 1.6 10 0.16 2 90

10 1.6 10 0.19 3 88

10 1.6 10 0.14 4 92

(a) Initial Li concentration in hexyl formamide, (b): Final Li concentration in hexyl formamide.

It is seen that the pH of the acidic solution used to recover lithium from hexyl
formamide does not affect the percent recovery amount which is around 90 %.

The recovery of lithium ions from hexyl formamide can be further repeated by
addition of acidic solutions into remaining hexyl formamides. The recovery amount
is again observed as 90 % as shown in Table 3.26.

Table 3.26 The effect of pH on recovery of lithium from hexyl formamide (Second
recovery)

Hexyl
Co (a) Acid C (b) The recovery
formamide volume pH solution
(ppm) (ppm) of LiCl (%)
(ml)
(g)

10 0.15 10 0.015 1 90

10 0.16 10 0.016 2 90

10 0.19 10 0.02 3 89

10 0.14 10 0.013 4 91

(a): Initial Li concentration in hexyl formamide, (b): Final Li concentration in hexyl formamide

86
The recovery of lithium chloride from hexyl formamide was done also by using 1 M
hydrochloric acid solution. In that case 93 % of lithium can be recovered from hexyl
formamide phase. The recovered lithium can be further precipitated as lithium
carbonate by adding equal molar soda ash.

3.3 Adsorption via lithium manganese oxide adsorbents

Among the various synthesis methods and reactants, lithium manganese oxide with
the maximum capacity value was synthesized by Solid-Solid-2 procedure in which
employs manganese carbonate and lithium hydroxide monohydrate was employed as
precursors. The performance of the adsorbents was evaluated by the experiments
conducted with variable pH and initial lithium chloride concentration.

3.3.1 Effect of pH on adsorption

In order to investigate the effect of pH on capacity values of LiMnO adsorbents, a

series of experiments were conducted under varying pHs.

87
Table 3.27 Capacity values of adsorbents at different pHs (Total volume: 100 ml,
Adsorbent amount: 0.02 g)

Capacity
pH Initial Li conc. (ppm) Final Li conc. (ppm)
(mg Li/g adsorbent)

6.3 9.02 7.98 5.2

7.3 9.02 7.66 6.8
8.2 9.02 7.51 7.6
9.3 9.02 6.12 14.5

10.2 9.02 4.46 22.8

Table 3.27 shows that, pH of the adsorption medium strictly affects the performance
of the adsorbents. As pH increases from 6.3 to 10.2, the capacity values increases
from 5.2 to 22.8 mg Li/g adsorbent. This effect is mainly caused by the competition
between the lithium ions and hydrogen ions due to the nature of ion
exchange/adsorption process. When hydrogen ion concentration is high (i.e, at low
pH), H+ ions have a high chance to replace Li+ ions. At high pH range, H+
concentration is low, then lithium ions have higher chance for attachment to the
adsorbent. Therefore, the capacity of the adsorbents increases with increasing pH
values.

3.3.2 Effect of initial lithium concentration on adsorption

Similar to pH, the effect of initial lithium concentration on capacity values of the
adsorbents were investigated by performing a set of experiments under varying
initial lithium chloride concentrations.

Table 3.28 Capacity value of the adsorbents at different initial lithium concentrations
(pH:10.2 and adsorbent amount: 0.02 g )

Initial Li conc. Final Li conc. Capacity

(ppm) (ppm) (mg Li/g)
4.73 0.14 22.8
8.36 3.99 21.6
44.89 40.4 22.9

88
Table 3.28 shows that the initial lithium concentration does not affect the capacity
value of the adsorbents. In other words the lithium concentration in the medium does
not affect the total lithium amount adsorbed on the particle. This behavior implies
that the mass transfer limitations around the particle and inside the particle are not
significant to affect the thermodynamic limits of the adsorption process within given
time (24 h). The small particle size and high porosity of the adsorbents reduces the
mass transfer resistance.

3.3.3 Kinetics of the adsorption process

The kinetics of the adsorption process was studied with five identically prepared
mixtures by measuring the uptake amount of lithium at different times. The initial
lithium concentration was 10 ppm, pH was 10.2 and 0.02 g of adsorbent was used
for each parallel experiment.

25,00

20,00

15,00
Lithium Uptake
(mg Li/g)
10,00

5,00

0,00
0 5 10 15 20 25 30
Time (hours)

Figure 3.28 The kinetics of adsorption process

It is seen in Figure 3.28 that, the adsorption process is completed in 20 hours with
the maximum lithium uptake value of 22 mg Li/g adsorbent. When the results in
Table 3.28 and Figure 3.28 were evaluated together, it was seen that the same uptake
value of 22 mg Li/g was acquired for both the experiments having initial lithium
concentration of 4.73 and 10.0 ppm Li in 24 h. If the mass transfer limitations
89
govern the kinetics of the process, one must find a lower value for initial lithium
concentration of 4.73 ppm Li. The equal values of uptake ratio for different initial
lithium concentrations, at equal adsorption times validate the negligible mass
transfer limitation in the process.

3.3.4 Equilibrium Isotherms

Equilibrium isotherm calculations were done with solutions containing different

initial lithium concentrations between 1 and 5 ppm.

Equilibrium concentration of lithium in solution with respect to uptake value is

shown in Figure 3.29.

25,0

20,0
Lithium uptake (mg/g)

15,0

10,0

5,0

0,0
0,00 0,05 0,10 0,15 0,20 0,25 0,30 0,35
Lithium concentration at equilibrium (ppm)

Figure 3.29 The equilibrium concentration of lithium in solution vs uptake of lithium

ions

- Langmuir Isotherm

The general Langmuir isotherm equation is shown below;

90
 𝐶𝑒 1 𝐶𝑒
= + Eqn. 7
𝑄𝑒 𝐾𝑎𝑑𝑠 .𝑄𝑚 𝑄𝑚

Here,

Ce: Lithium concentration in solution at equilibrium (mg/L)

Qe: Uptake amount of lithium at equilibrium (mg/g)

Kads: Empirical equilibrium constant (L/mg)

Qm: Theoretical maximum adsorption capacity (mg/g)

According to the data labeled in Figure 3.29, the Langmuir isotherm curve is drawn
and shown in Figure 3.30 with a best line having R2 value of 0.997.

0,016

0,014

0,012

0,010
Ce/Qe

0,008

0,006

0,004

0,002

0,000
0,000 0,010 0,020 0,030 0,040 0,050
Ce

Figure 3.30 Langmuir isotherm curve for LiMnO at 25 oC.

The equation of the best line is,

Ce
= 0.2944Ce +0.0012 Eqn. 8
Qe

91
Qm value was found as 3.4 mg/g and Kads was found as 184.8 L/mg. Kads value was
higher than values cited in the literature, but there is a contradiction in the Kads
values at different studies [83], [85], [100]. This difference in Kads values may result
in insufficiency of Langmuir isotherm to explain what is happening in separation
process. The ion exchange mechanism and surface adsorption mechanism are two
possible mechanisms taking place simultaneously in the process. The maximum
uptake value was found as 3.37 mg Li/g, which should be around 20 mg Li/g
according to the previous experiments.

- Freundlich isotherm

Freundlich isotherm equation is shown below,

1/𝑛
𝑄𝑒 = 𝐾𝑓 . 𝐶𝑒 Eqn. 9

Where, Kf indicates relative capacity and n indicates adsorption intensity.

By taking the logarithm of both sides, the equation shown below is acquired,

1
𝑙𝑜𝑔𝑄𝑒 = 𝑙𝑜𝑔𝐾𝑓 + 𝑛 . 𝑙𝑜𝑔𝐶𝑒 Eqn. 10

The log Qe vs log Ce plot is shown in Figure 3.31 with a best line having R2 value of
0.9504.

1,60
1,40
1,20
1,00
logQe

0,80
0,60
0,40
0,20
0,00
-3 -2,5 -2 -1,5 -1 -0,5 0
logCe

Figure 3.31 Freundlich isotherm curve for LiMnO at 25 oC.

92
The slope gives 1/n, n value was found as 2.14, which means high tendency of
lithium ions towards manganese oxide particles. The intercept value was equal to log
(Kf) which gives Kf as 45.7 L/mg. The Freundlich isotherm actually works for dilute
systems, thus fits to the concerning system with R2 value of 0.95.

3.3.5 Stability experiments

In order to understand the stability of lithium manganese oxides, a cycle of

adsorption and desorption with lithium chloride feed solution and hydrochloric acid
leaching solution was repeated. The experiment was started with the adsorption of 5
ppm lithium chloride following by the desorption with 0.5 M HCl and continues
with the adsorption of lithium chloride. This cycle was repeated for six times and the
loss of capacity in the adsorbents was tabulated in Table 3.29.

Table 3.29 The stability of lithium manganese oxide particles in lithium recovery

Capacity change with respect to

Cycle no Capacity (mg Li/g ads.)
first cycle (%)
1 21.1 100
2 20.7 98.1
3 20.1 95.3
4 19.8 93.8
5 19.6 92.9
6 19.4 92

As it can be deduced from Table 3.29, the capacity of the adsorbents was decreased
by around 1 % in each additional cycle. If these values were extrapolated linearly,
the half-life of the adsorbents was found at 50th. adsorption-desorption cycle. The
mechanism of adsorption in lithium manganese oxides is explained by ion exchange
or redox reactions occurring in crystal structure. The exchange of ions in the crystal
is happening in tapotactic way which means there is no distortion of the crystal
lattice due to interchange of the ions. Treatment of LiMnO with acid causes the
removal of nearly all the lithium from the tetrahedral sites while maintaining a
spinel-type structure stable. However it is also known that, Mn (IV) atoms are placed
in octahedral sites of the adsorbent crystal and transition metal atoms at octahedral
93
position may result in Jahn-Teller distortion [72], [87]. This distortion leads the
crystal to lower its energy which is favorable. In lithium manganese oxide, this type
of distortion may cause tiny distractions in crystal lattice and result in a loss of
capacity of the adsorbent.

3.3.6 Selectivity experiments

The selectivity experiments were conducted by a feed solution which consists of

equal concentration of lithium, sodium, potassium and magnesium ions at pH:10.
The results and conditions of the experiments were tabulated in Table 3.30.

Table 3.30 Selectivity of lithium manganese oxides with respect to Na, K and Mg.

Feed Initial Final Adsorbent Capacity

Metal Selectivity with
solution Conc. Conc. amount (mg ion/g
ion respect to lithium
(ml) (ppm) (ppm) (g) adsorbent)
Li 100 9.60 0.77 0.051 17.31 1
Na 100 9.68 9.35 0.051 0.64 0.037
K 100 10.37 9.57 0.051 1.38 0.08
Mg 100 9.40 8.75 0.051 1.30 0.075

Table 3.30 shows that, lithium manganese oxide adsorbent has strong selectivity
towards lithium; 27 times higher than sodium, 12.5 times higher than potassium and
13.3 times higher than magnesium ions. The relatively high selectivity for Li+ can be
explained by the ion-sieve effect of the spinel lattice with a three-dimensional (1×3)
tunnel, suitable in size for fixing Li+ in cubic phase MnO2 ion sieves obtained from
LiMnO precursors, yet other metal ions can only adsorb on the adsorbent surface
sites because of their large ionic radii, similar to the cases of the MnO2 ion-sieves in
other studies [77], [87]. Moreover the tapotactic insertion-desertion mechanism of
Li+ ions in the crystal lattice does not let other ions to take place of the lithium ions.
This ability of the adsorbent makes it unique for selective lithium separation
operations compared to other candidate chemicals.

94
3.4 Separation of lithium by PSMA-LiMnO beads in an adsorption column

Column experiments with PSMA-manganese oxide adsorbents were first done with
artificial brine solutions in order to quantify the column performance. The main
parameters that affect the performance of the column are pH, feed concentration,
flow rate and particle size of the beads. The column dimensions and adsorbent
amount was fixed for all experiments.

All the experiments were performed with the same bead which was labeled as T16.
The MnO2 content of T16 is 51 %. The column diameter is 20 mm, the column
height is 385 mm and total weight of the beads is 38.5 g. Bead size is 3mm in size
unless otherwise stated.

3.4.1 The breakthrough plots for solutions containing lithium chloride only

3.4.1.1 Effect of pH

The most critical parameter that affects the behavior of adsorption process is pH. As
explained in Section 3.3.1, increase in pH strictly improves the capacity of LiMnO
particles. On the other hand, the swelling ratio of PSMA content in the beads
increases at high pH. Swelling of the beads increases the transfer rate of lithium ions
in and out, however reduces the flow rate in the column. At pH values higher than 9,
flow in the column was even cut down by the expansion of the beads. Therefore the
maximum pH value was determined as pH:9.0 by the restriction of swelling behavior
of the beads. The minimum pH value was chosen as 7.5, where the activity of
LiMnO particles becomes significant.

Three experiments were performed at pH: 7.5, 8.2 and 9.0. The column conditions
were tabulated in Table 3.31.

95
Table 3.31 The column operating conditions for the pH dependency of adsorption
process

Total
Feed Li Feed buffer Average Outlet
Run Feed processed
conc. content* flow rate initial and
# pH solution
(ppm) (ml/min) final pH
(ml)
1 20 0.23/1.65 7.5 15 14440 2.25 - 2.98
2 20 0.33/1.65 8.2 15 14440 2.25 – 3.21
3 20 0.48/1.65 9.0 15 16350 2.25 – 3.45
*(mlNH3(25%)/gNH4Cl)

pH 7.5 pH 8.2 pH 9.0

20
18
16
14
Li conc (ppm)

12
10
8
6
4
2
0
0 2000 4000 6000 8000 10000 12000 14000 16000 18000
Processed water (ml)

Figure 3.32 The breakthrough curves at different pH values

Figure 3.32 shows the behavior of the breakthrough curves at different pH values. As
pH increases, curves shift to the right and become steeper. The higher capacity of
LiMnO at higher pH values leads to this behavior. The capacity values were
calculated as 128.8, 148.8 and 221.2 mg Li for pH 7.5, 8.2 and 9.0 respectively while
the theoretical capacity values were 137.5, 153.2 and 235.6. The capacity values
increase according to theoretical capacity values which implies that the process has
low/negligible mass transfer limitation.

96
3.4.1.2 Effect of initial lithium concentration

The lithium concentration in brines as natural lithium sources varies from 0.17 ppm
(sea water) to 1500 ppm (Salar de Atacama, Chile). The lithium concentration in
feed solutions does not change the performance of the column but affects the shape
of the breakthrough curves. In order to observe this effect, three runs were carried
out by changing the concentration in feed solution. The conditions are shown in
Table 3.32.

Table 3.32 The column operating conditions for different initial lithium
concentration in adsorption process

Total
Feed Li Feed buffer Average Outlet
Run Feed processed
conc. content* flow rate initial and
# pH solution
(ppm) (ml/min) final pH
(ml)
4 5 0.48/1.65 9 15 14440 2.25 – 3.45
3 20 0.48/1.65 9 15 16350 2.25 – 3.45
5 120 0.48/1.65 9 15 21110 2.25 – 3.45
*(mlNH3(25%)/gNH4Cl)

97
Figure 3.33 The breakthrough curves at different initial lithium concentrations.

Figure 3.33 shows the change in shapes of the breakthrough curves with respect to
change in initial lithium concentration. At 5 ppm, the curve has many small
fluctuations and cannot reach the saturation in the end. The capacity of the column
was found as 60.1 mg Li which is 26 % of the theoretical capacity. At 20 ppm initial
concentration, a smooth breakthrough curve is observed and the capacity is found as
212.2 mg Li (94 % of the theoretical capacity). At 120 ppm, the curve has a sharp
increment at the beginning of the process and reaches to its final concentration. The
breakthrough behavior cannot be observed. The capacity value was found as 230.0
mg Li which is 98 % of the theoretical capacity. The column capacity reaches near to
its saturation point in fast and effective way. As the initial lithium concentration
increases from 5 to 120 ppm, the curves get steeper as expected. In the scope of

98
lithium recovery, in order to separate more lithium from brines, feeding the column
until its saturation becomes important.

3.4.1.3 Effect of flow rate

In order to show the effect of flow rate on breakthrough curves, a series of

experiments were conducted. The operating conditions for the experiments are listed
in Table 3.33.

Table 3.33 The column operating conditions for different flow rates in adsorption
process

Total
Feed Li Feed buffer Average Outlet
Run Feed processed
conc. content flow rate initial and
# pH solution
(ppm) (ml/min) final pH
(ml)
6 20 0.48/1.65 9 5 4800 2.25 – 3.15
3 20 0.48/1.65 9 15 16350 2.25 – 3.45
7 20 0.48/1.65 9 30 28800 2.25 – 4.10
*(mlNH3(25%)/gNH4Cl)

99
Figure 3.34 The breakthrough curves at different flow rates.

Figure 3.34 shows the effect of increasing flow rates on breakthrough behavior.
Under 5 ml/min flow rate, the breakthrough cannot be reached similar to the case in
5 ppm Li feed. Lithium concentration in the effluent of the column is below 0.16
ppm during the process and the column capacity was calculated as 95.35 mg Li. The
low lithium concentration in the effluent is due to the increasing space time of the
lithium solution in the column. As flow rate increases to 30 ml/min, the space time
decreases hence, curves become steeper. The oscillations observed in initial stages of
the process are attributed to the channeling or dead volume effect. The column
capacity is calculated as 160.2 mg Li, which is 68 % percent of the total theoretical
capacity. The percent saturation of the column is smaller than expected, probably
due to the low space time in the process. The breakthrough curve at 15 ml/min
seemed to be the ideal one to observe the breakthrough point and the general
characteristics of the adsorption column.

100
3.4.1.4 Effect of bead size

The size of the beads in the column may change the performance of the system due
to mass transfer effects. In order to investigate the effect of bead size on efficiency of
the column, two experiments were carried out with two columns filled with different
size of beads. The operating parameters of the columns are tabulated in Table 3.34.

Table 3.34 The column operating conditions for different bead sizes in adsorption
process

Total
Feed Li Feed buffer Bead Average Outlet
Run Feed processed
conc. content* size flow rate initial and
# pH solution
(ppm) (mm) (ml/min) final pH
(ml)
3 20 0.48/1.65 3 9 15 16350 2.25 – 3.45
8 20 0.48/1.65 8 9 15 10560 2.25 – 4.10
*(mlNH3(25%)/gNH4Cl)

Figure 3.35 The breakthrough curves for different bead sizes

Figure 3.35 shows the effect of bead size and mass transfer clearly. When bead size
is increased from 3 mm to 8 mm, the breakthrough curve becomes broader. It is
important to note that as the bead size increases, the PSMA content in the beads
increases and LiMnO content decreases. 8 mm sized beads can only hold 43 %
LiMnO content, while 3 mm sized beads can hold 51 % LiMnO. This brings
additional barrier to the diffusion of lithium ions through the beads which have
higher PSMA content. Also decrease in the total mass transfer area in the system
101
leads the broadening of the breakthrough curve. In case of 3 mm sized beads, larger
amount of beads can be used in filling the same volume. In this case, 38.5 g of 3 mm
sized beads can be filled with 35.4 g of 8 mm sized beads. The capacity usage in
each case also validates the findings. In 3 mm sized beads, the capacity of the
column is calculated as 221.2 mg Li which is 94 % of the theoretical capacity.
Meanwhile the capacity of column filled with 8 mm sized beads is found as 128.4
mg Li which is 70 % of the theoretical capacity. Another significant fact is
employing 3 mm sized beads in 20 mm diameter column. The ratio of column
diameter to bead diameter is around 7 and coincides with the ratio of 8 mentioned as
rule of thumb for adsorption column experiments.

3.4.2 The breakthrough plots for solutions containing foreign ions

Lithium chloride sources are generally poor in lithium content but reach in NaCl,
KCl and MgCl. The behavior of adsorbents with solutions containing foreign ions
(Na, K, Mg) is questioned in this section. Sodium chloride, potassium chloride and
magnesium chloride are very rich in sea water with respect to lithium chloride. For
instance, sea water contains 0.17 ppm Li besides 10800 ppm Na, 1290 ppm Mg and
390 ppm K at 3.5 % salinity. Before feeding the column with the mixture of ions, the
adsorption capacity of the beads for each foreign ion was analyzed individually. The
conditions of the experiments were arranged similar to experiments done with
lithium chloride solutions.

3.4.2.1 The breakthrough curve for sodium chloride feed solution

The conditions of the experiments were given in Table 3.35.

Table 3.35 The column operating conditions for feed solutions containing NaCl

Total
Feed Na Feed Average Outlet
Run Feed processed
conc. buffer content* flow rate initial and
# pH solution
(ppm) (ml/min) final pH
(ml)
9 10.1 0.48/1.65 9 15 14440 1.84-3.75
*(mlNH3(25%)/gNH4Cl)

102
 10

8
Na conc (ppm)

0
0 5000 10000 15000
Processed water (ml)

Figure 3.36 The breakthrough curve for sodium chloride feed solution

Figure 3.36 shows that the outlet concentration of the column suddenly reaches to
the concentration of the inlet which depicts that the affinity of the adsorbent towards
sodium is very low. Generally, a mass transfer zone can be observed in the very first
part of the curve which results in a gradual increase in the effluent ion composition.
The sudden increase in the composition was attributed to the very low affinity of the
beads towards sodium. The capacity of the column was calculated as 2.51 mg Na
which was 88 times less than the capacity value for lithium (221.2 mg Li). The
adsorption capacity of single LiMnO towards Na was 27 times less than for Li. The
selectivity of the beads in column operation towards lithium ions is three times larger
with respect to batch operation with single LiMnO.

3.4.2.2 The breakthrough curve for potassium chloride feed solution

The conditions of the experiment were given in Table 3.36.

103
Table 3.36 The column operating conditions for feed solutions containing KCl

Total
Feed Feed buffer Average Outlet
Run Feed processed
K conc. content* flow rate initial and
# pH solution
(ppm) (ml/min) final pH
(ml)
10 10.3 0.48/1.65 9 15 14440 1.94-3.37
*(mlNH3(25%)/gNH4Cl)

10

8
K conc (ppm)

0
0 5000 10000 15000
Processed water (ml)

Figure 3.37 The breakthrough curve for potassium chloride feed solution

It is seen from Figure 3.37 that the affinity of the adsorbent towards potassium ions
is very low similar to sodium ion. The capacity of the column was found as 5.78 mg
K which was 38 times less than the capacity value for lithium. The adsorption
capacity of single LiMnO towards K was 12.5 times less than for Li. The selectivity
of the beads in column operation towards lithium ions was three times larger with
respect to batch operation with single LiMnO similar to the case with sodium.

3.4.2.3 The breakthrough curve for magnesium chloride feed solution

The conditions of the experiment were given in Table 3.37.

104
Table 3.37 The column operating conditions for feed solutions containing MgCl

Total
Feed Feed buffer Average Outlet
Run Feed processed
Mg conc. content* flow rate initial and
# pH solution
(ppm) (ml/min) final pH
(ml)
11 11 0.48/1.65 9 15 14880 1.05-3.43
*(mlNH3(25%)/gNH4Cl)

10

8
Mg conc (ppm)

0
0 5000 10000 15000
Processed water (ml)

Figure 3.38 The breakthrough curve for magnesium chloride feed solution

In Figure 3.38, it is seen that the outlet of the column does not reach the inlet
magnesium concentration. Magnesium ions were adsorped by the beads partially.
The capacity of the column was calculated as 52.9 mg Mg which was 4.2 times less
than the capacity value for lithium. The adsorption capacity of single LiMnO
towards Mg was 13 times less than for Li, but the selectivity of the beads in column
operation towards lithium ions get decreased. The reason of this decrease was
attributed to the PSMA content in the beads. The polar OH groups in PSMA may
attract the Mg ions and lead to this unexpected result.

105
3.4.3 The breakthrough plots for the artificial and real brine solutions

The concentrated sea water in Çamaltı Salina (İzmir) and the extract from boron clay
in Bigadiç (Balıkesir) are two potential lithium sources that were studied in this
thesis work. The concentration of lithium ion in concentrated sea water and boron
clay extract is much lower than foreign ion concentration in those solutions.

In order to describe the behavior of LiMnO-PSMA beads when opposed to high salt
concentrations, an artificial brine solution of concentrated sea water was prepared
first and the following experiments were conducted.

3.4.3.1 The breakthrough plots for the artificial brine solutions

Artificially prepared brine concentration and the experimental conditions were given
in Table 3.38.

Table 3.38 The column operating conditions for artificial brine solutions

Total
Average Outlet
Run Feed solution Feed buffer Feed processed
flow rate initial and
# conc. (ppm) content* pH solution
(ml/min) final pH
(ml)
0.43 ppm Li
53430 ppm Na
12 0.2/1.65 8.1 15 37164 0.5-2.52
4464 ppm K
9210 ppm Mg
*(mlNH3(25%)/gNH4Cl)

106
 0,5
0,45
0,4
Li conc. ppm 0,35
0,3
0,25
0,2
0,15
0,1
0,05
0
0 5000 10000 15000 20000 25000 30000 35000 40000
Processed water (ml)

Figure 3.39 The effluent lithium concentration in a column fed with artificially
prepared brine solution

Figure 3.39 shows that the outlet concentration of the column oscillates around 0.1
ppm through the experiment. The high ionic strength of the feed solution due to high
concentrations of the foreign ions probably plays the major role in this behaviour.
The capacity of the column was found as 11.2 mg Li. Breakthrough curves for Na, K
and Mg cannot be observed. Due to high concentration of those ions, the feed
concentration was suddenly observed in the outlet of the column.

3.4.3.2 The breakthrough plots for the real brine solutions

3.4.3.2.1 The breakthrough plot for concentrated sea water taken from
Çamaltı Salina

The sea water in Çamaltı Salina is concentrated around 9 to 10 fold during NaCl
production in playa and salt concentration becomes 35 Bome. The concentration of
the sea water and the operational conditions of the column were given in Table 3.39.

107
Table 3.39 The column operating conditions for concentrated brine solution taken
from Çamaltı Salina

Total
Average Outlet
Run Feed Feed processed
flow rate initial and
# conc. (ppm) pH solution
(ml/min) final pH
(ml)
1.25 ppm Li
76750 ppm Na
13 8.1 15 11473 0.5-2.84
11023 ppm K
21000 ppm Mg
*(mlNH3(25%)/gNH4Cl)

1
0,9
0,8
0,7
Li conc. (ppm)

0,6
0,5
0,4
0,3
0,2
0,1
0
0 5000 10000 15000
Processed water (ml)

Figure 3.40 The effluent lithium concentration in a column fed with concentrated
brine solution from Çamaltı Salina

Figure 3.40 shows that the lithium concentration at the outlet remains constant at
0.78 ppm through the experiment. The capacity of the column was calculated as 5.39
mg Li. The behavior of the column is very similar to the case in artificial brine. The
high ionic strength of the concentrated sea water probably suppressed the capacity of
the adsorbent by forming an ionic layer around the beads. The breakthrough
behavior for the foreign ions was observed similarly to the case in artificial brine
solution and the concentration in the outlet of the column suddenly increased to the
concentration of the inlet stream.

108
3.4.3.2.2 The breakthrough plot for boron clay extract taken from Bigadiç

The concentration of the clay extract and the operating conditions are given in Table
3.40.

Table 3.40 The column operating conditions for boron clay extract taken form
Bigadiç

Total
Average
Run Feed Feed processed
flow rate
# conc. (ppm) pH solution
(ml/min)
(ml)
388 ppm Li
4880 ppm Na
14 8.73 15 780
530 ppm K
15110 ppm Mg

The total volume of the boron extract was low to feed to the column continuously;
therefore the feed solution was fed to the column in batchwise, where the outlet of
the column was circulated back to the feed tank. After 8 h of operation, the following
concentrations listed in Table 3.41 were measured for the initial and final
concentration in the feed tank.

Table 3.41 The initial and final concentrations in the feed tank before and after the
operation

Run # Li Na K Mg
Run 14 initial conc. (ppm) 388 4880 530 15110
Run 14 final conc. (ppm) 105 4250 460 12400

Table 3.41 shows that nearly 75 percent of the lithium content in the clay extract is
adsorbed by the beads. However, besides lithium, sodium and magnesium ions were
also adsorbed by the beads, which affect the selectivity of the total process
negatively. The selectivity analysis of the process will be covered in following
sections.

109
3.4.4 Recovery of lithium and regeneration of the beads

The recovery of lithium and regeneration of the beads were done with 0.5 M HCl.
This process is generally called “elution process” and functions to interchange
hydrogen ions with lithium ions and regenerate the beads while recovering lithium.
Elution of the columns fed with different feed solutions was analyzed separately.

3.4.4.1 Elution of the column fed with solution having lithium chloride only

The elution was first done with the column fed with solution containing lithium
chloride in Run 3.

600

500

400
Li conc (ppm)

300

200

100

0
0 1000 2000 3000 4000 5000 6000
Processed water (ml)

Figure 3.41 The elution curve after Run 3

Figure 3.41 shows the elution curve of the column after Run 3. The total amount of
lithium eluted from the column was calculated as 220 mg Li which was nearly equal
to capacity value (221.2 mg Li) acquired after Run 3.

In order to have an elution solution with high lithium concentration, Run 3 was
repeated and in this case the column was fed with 350 ml 0.5 M HCl solution two
times consecutively. The operation was batchwise i.e., the outlet of the column was

110
returned back to the feed tank. The concentration of the solutions of two consecutive
leaching processes is shown in Table 3.42.

Table 3.42 Elution process data of the column after Run 3.

Run # Total volume (ml) Li conc. (ppm)

15 350 445
16 350 115

In Run 3, the column capacity was calculated as 221.2 mg Li. After first leaching in
Run 15 157.5 mg of Li was leached from the column which was 70 % of the total
lithium amount adsorbed in the column. In Run 16, additional 40.3 mg of Li was
leached from the column which adds up to total 197.8 mg of Li recovered from the
column. At the end, 90 % of total lithium adsorbed in the column was eluted. The
leaching process seems to be very effective and nearly all lithium in the column is
taken to acid solution. After elution processes, beads were washed with plenty of
acid solution and regenerated to be used in following separation process.

3.4.4.2 Elution of the column fed with artificial brine solution

The high concentration of the foreign ions in artificial brine solution leads to
adherence of those ions on bead surfaces and partially on LiMnO particles.
Therefore a sweeping operation was applied before elution process. In sweeping
process the column was fed with tap water or slightly acidic water (pH:6.1) and
foreign ions were washed through the column. In that manner, the column fed with
artificial brine solution in Run 12 was washed with tap water in Run 17 and washed
with acidic water in Run 25 before elution. The results after each sweeping process
were analyzed in the following sections.

3.4.4.2.1 Sweeping of the column with tap water

After Run 12, the column was washed with plenty of tap water to get rid of foreign
ions. The change of ion concentrations at the outlet of the column is shown in Figure
3.42.
111
 Sodium Potassium Magnesium

16000

14000

12000
Metal ion conc. ppm

10000

8000

6000

4000

2000

0
0 500 1000 1500 2000
Processed water (ml)

Figure 3.42 The change of ion concentrations in sweeping process applied after Run
12

It is seen from Figure 3.42 that the sweeping rate of sodium ions is very high and
3.33 g Na was swept by 2 L of tap water. The swept amount of potassium ions and
magnesium ions were calculated as 234 mg K and 509 mg Mg respectively. During
the sweeping process, only 0.326 mg of lithium ions were lost which was 10214
times lower than swept Na, 718 times lower than swept K, 1560 times lower than
swept Mg. The sweeping process efficiency decreases after 500 ml of tap water was
fed to the system. The concentration of K and Mg becomes 13 ppm and 6 ppm
respectively which are nearly equivalent values in ASKİ tap water analysis report
[101]. Another graphic representation of the high sweeping rate of foreign ions is
given in Figure 3.43, which shows the molar ratio of K and Mg with respect to Li in
the column effluent.

112
 K/Li Mg/Li

600

500

400
X/Li

300

200

100

0
0 500 1000 1500 2000
Processed water (ml)

Figure 3.43 The molar ratio of potassium and magnesium with respect to lithium
during sweeping process

After the sweeping process was completed, the column was fed with 350 ml 0.5 M
HCl batchwise for elution step. The concentration of the elution solution and the
summary of the steps up to elution are summarized in Table 3.43

113
Table 3.43 The elution of the column saturated with artificial brine solution and
swept by tap water

Total vol.
Run
Operation processed Li (mg) Na (mg) K (mg) Mg (mg)
#
(ml)
Adsorbed amount
12 of ions from 37164 11.2 Saturated Saturated Saturated
artificial brine
Amount of ions
17 2000 0.326 3330 234 509
swept by tap water
Eluted amount of
18 350 10.2 53.2 27.7 171.5
ions by 0.5 M HCl

Table 3.43 shows the amount of ions adsorbed from artificial brine in Run 12, the
amount of ions swept through the column in Run 17 and eluted amount of ions in
Run 18 by 0.5 M HCl. The total mass balance holds for lithium and shows small
amount of lithium remaining in the column. In order to evaluate the effectiveness of
the overall process (feeding-sweeping-eluting), the concentration of the artificial
brine and the elution solution are compared in Table 3.44.

Table 3.44 The comparison of the artificial brine and elution solution composition
after sweeping by tap water

The composition Li (ppm) Na (ppm) K (ppm) Mg (ppm)

Articial brine
0.43 53430 4464 9210
conc.
Elution solution
29 152 79 490
conc.
Change of Increased by 67 Decreased by 352 Decreased by 57 Decreased by 19
concentration folds folds folds folds

Table 3.44 shows that the overall process is very highly selective towards lithium
compared to sodium and potassium ions and highly selective compared to
magnesium ions. The results coincide with the findings in Section 3.4.2 where the
column capacity for magnesium was found greater than sodium and potassium.

114
In order to enrich the lithium content in the eluted solution and decrease the foreign
ion content, the overall process was repeated once more with a feed solution of 5 L
prepared at a composition of elution solution given in Table 3.44. The new feed
solution was prepared in 0.5 M HCl, and the pH of the solution was adjusted to 9 by
ammonia addition. After 5 L was fed to the column in batchwise, the column was
swept by 4 L of tap water continuously and lastly eluted by 350 ml 0.5 M HCl
solution. The compositions of the solutions in each step are shown in Table 3.45.

Table 3.45 The second adsorption-sweeping-elution process composition with the

feed solution having concentration of elution process in Run 18

The composition Li (ppm) Na (ppm) K (ppm) Mg (ppm)

The new feed
29 152 79 490
solution
Elution solution
310 151 38 1076
conc.
Change of Increased by 10.7 Decreased by 2 Increased by 2.2
No change
concentration folds folds folds

Table 3.45 shows that an additional adsorption-sweeping-elution cycle increases the

composition of lithium by 10.7 folds, while decreasing the potassium content by 2
folds and increase the magnesium content by 2.2 folds. The composition of sodium
does not change. Increase in magnesium concentration and no change in sodium
concentration was attributed to the tap water used in sweeping process. The sodium
and magnesium content in tap water probably hinder the desorption of those ions.
The low potassium content in tap water (3.1 ppm) explains the decrease in the
potassium composition with respect to feed solution.

115
3.4.4.2.2 Sweeping of the column with acidic water at pH 6.1

As described in Section 3.4.4.2.1, the column was washed with acidic water at pH
6.1 after Run 12 instead of tap water. The logic behind washing the column with a
slightly acidic solution was to increase the swept amount of foreign ions, especially
Mg ions, while maintaining the lithium content in the column. After sweeping
process, the column was eluted with 0.5 M HCl solution, to have the lithium rich
solution finally. The concentration of the elution solution and the summary of the
steps up to elution are summarized in Table 3.46.

Table 3.46 The elution of the column saturated with artificial brine solution and
swept by acidic water at pH 6.1.

Total
Run volume
Operation Li (mg) Na (mg) K (mg) Mg (mg)
# processed
(ml)
Adsorbed amount
12 of ions from 37164 11.2 Saturated Saturated Saturated
artificial brine
Amount of ions
26 swept by acidic 2000 0.984 3350 241 612
water
Eluted amount of
27 350 9.6 28.4 21.4 82.1
ions by 0.5 M HCl

Table 3.46 shows that sweeping the column with an acidic water increases the swept
amount of all ions through the column including lithium. However, the swept
amount of magnesium ions increases to 612 mg from 509, while swept of lithium
ions only increases to 0.984 mg from 0.326 mg. By losing 0.658 mg of lithium, 103
mg of magnesium could be swept through the column.

In order to evaluate the effectiveness of the overall process (feeding-sweeping (at

pH:6.1)-eluting), the concentration of the artificial brine and the elution solution are
compared in Table 3.47.

116
Table 3.47 The comparison of the artificial brine and elution solution composition
after sweeping by acidic solution (pH 6.1)

The composition Li (ppm) Na (ppm) K (ppm) Mg (ppm)

Articial brine
0.43 53430 4464 9210
conc.
Elution solution
27.4 81.1 61.1 234.6
conc.
Change of Increased by 64 Decreased by 659 Decreased by 73 Decreased by 39
concentration folds folds folds folds

Table 3.47 obviously shows that sweeping of the column by an acidic water
increases the lithium content with respect to foreign ion content after elution.

In order to enrich the lithium content in the solution, the adsorption-sweeping-

elution cycle was repeated once more with an acidic sweeping solution. The results
were tabulated in Table 3.48.

Table 3.48 The second adsorption-sweeping(at pH:6.1)-elution process composition

with the feed solution having concentration of elution process in Run 26

The composition Li (ppm) Na (ppm) K (ppm) Mg (ppm)

The new feed
27.4 81.1 61.1 234.6
solution
Elution solution
261 32 20.6 102.5
conc.
Change of Increased by 9.5 Decreased by 2.5 Decreased by 3 Decreased by 2.3
concentration folds folds folds folds

Table 3.48 represents that, an additional adsorption-sweeping-elution cycle increases

the lithium content to 261 ppm while reducing the foreign ion concentrations, unlike
sweeping with tap water in the previous section. The decreasing ratio of foreign ion
concentration is reduced to a number between 2 and 3 which actually implies that the
system approaches near its equilibrium point. The repetition of the adsorption-
sweeping-elution cycle once more did not change the concentration of the ions
significantly and gave nearly equal values.

117
3.4.4.3 Elution of the column fed with concentrated sea water taken from
Çamaltı Salina

Similar to the case in artificial brine solution, sweeping process was applied before
elution due to high concentration of foreign ions in concentrated sea water, but the
sweeping process was applied batchwise by deionized water or acidic water at pH
6.1 in this case.

3.4.4.3.1 Sweeping of the column with deionized water

After feeding the fresh column with concentrated sea water in Run 13, 1 L of fresh
deionized water was fed to the column, 3 times for 2 hours. After that a leaching
solution of 0.5 M HCl was fed to the column and the contents of the column were
eluted. The overall process starting from feeding the column with concentrated sea
water followed by three consecutive sweeping processes and leaching process were
given in Table 3.49.

Table 3.49 The elution of the column saturated with concentrated brine solution and
swept by demineralized water

Total
Run volume
Operation Li (mg) Na (mg) K (mg) Mg (mg)
# processed
(ml)
Adsorbed amount
11473
13 of ions from real 5.39 Saturated Saturated Saturated
(continuous)
concentrated brine
Amount of ions 1000
19 0.72 4443 289.5 617
swept by water (batch)
Amount of ions 1000
20 0.19 479 19.3 8.2
swept by water (batch)
Amount of ions 1000
21 0.1 49 8.4 5.1
swept by water (batch)
Eluted amount of
22 350 4.20 24.4 28.0 72.7
ions by 0.5 M HCl

118
Table 3.49 shows the amount of ions adsorbed from real concentrated brine in Run
13, the amount of ions swept from the column in Run 19 through Run 21 by
demineralized water and eluted amount of ions in Run 21 by 0.5 M HCl. The trend
of the overall process is very similar to the case with artificial brine solution. The
mass balance holds for lithium and shows that only 3.3 % of lithium ions were left in
the column and 78 % of lithium ions adsorbed by the column was eluted by 0.5 M
hydrochloric acid. In order to evaluate the effectiveness of the overall process
(feeding-sweeping-eluting), the concentration of the real concentrated brine and the
elution solution are compared in Table 3.50, which shows high selectivity towards
lithium similar to the process with artificial brine feed.

Table 3.50 The comparison of the real concentrated brine and elution solution
composition after sweeping by deionized water

The composition Li (ppm) Na (ppm) K (ppm) Mg (ppm)

Real concentrated
1.25 76750 11023 21000
brine conc.
Elution solution
12 69.7 80 208
conc.
Change of Increased by Decreased by Decreased by Decreased by
concentration 10 folds 1101 folds 138 folds 101 folds

3.4.4.3.2 Sweeping of the column with acidic water at pH 6.1

After Run 13, the column was washed with acidic water at pH 6.1 in this case. The
sweeping process was applied in three consecutive runs with 1 L acidic solution in
each. The amount of ions adsorbed, swept and eluted is given in Table 3.51.

119
Table 3.51 The elution of the column saturated with concentrated brine solution and
swept by acidic water at pH 6.1.

Total vol.
Run
Operation processed Li (mg) Na (mg) K (mg) Mg (mg)
#
(ml)
Adsorbed amount
11473
13 of ions from real 5.39 Saturated Saturated Saturated
(continuous)
concentrated brine
Amount of ions
1000
28 swept by acidic 0.84 4474 292 819
(batch)
water
Amount of ions
1000
29 swept by acidic 0.27 512 28.2 53
(batch)
water
Amount of ions
1000
30 swept by acidic 0.13 48 9.3 18.4
(batch)
water
Eluted amount of
31 350 4.06 16.3 28.0 61.3
ions by 0.5 M HCl

Table 3.51 shows that the swept amount of lithium and foreign ions do not change
much with respect to the case with deionized water. The reason of this behavior may
be attributed to the strong buffer content in concentrated sea water which consumes
the hydrogen ions in acidic water. The concentrations for the overall process are
shown in Table 3.52.

Table 3.52 The comparison of the real concentrated brine and elution solution
composition after sweeping by acidic water at pH 6.1

The composition Li (ppm) Na (ppm) K (ppm) Mg (ppm)

Real concentrated
1.25 76750 11023 21000
brine conc.
Elution solution
11.6 46.6 62 175.1
conc.
Change of Increased by Decreased by Decreased by Decreased by
concentration 9.3 folds 1647 folds 178 folds 120 folds

120
It is seen that, even though the amount of foreign ion swept through the column did
not change much, the foreign ion concentration of the elution solution is reduced
around 15-20 % with respect to the case where sweeping process was done with
deionized water.

3.4.4.4 Elution of the column fed with boron clay extract taken from Bigadiç

The sweeping of the column was done with tap water and acidic water separately
after Run 14.

3.4.4.4.1 Sweeping of the column with tap water

After Run 14 was completed, the column was swept by 4000 ml tap water in order to
get rid of foreign ions. The outlet concentration through sweeping process is shown
in Figure 3.44.

Na K Mg

9000
8000
7000
6000
5000
ppm

4000
3000
2000
1000
0
0 500 1000 1500 2000 2500 3000 3500 4000 4500
Processed water (ml)

Figure 3.44 The change of ion concentrations in sweeping process applied after Run
14

121
Figure 3.44 shows that, after 500 ml of tap water was processed in the column, the
effect of sweeping process reduces and the ion concentrations in the outlet of the
column get equal to tap water ion concentrations. Up to that point, 0.44 g Na, 0.042
g K and 2.0 g Mg were swept through the column while 5.6 mg Li was lost in the
operation. After the sweeping process was completed, the column was fed with 350
ml 0.5 M HCl batchwise, in two consecutive steps for elution. The concentration of
the elution solution and the summary of the steps before elution were summarized in
Table 3.53.

Table 3.53 The elution of the column saturated with clay extract and swept by tap
water

Total
Run volume
Operation Li (mg) Na (mg) K (mg) Mg (mg)
# processed
(ml)
Adsorbed amount
780
14 of ions from clay 220.7 491.4 54.6 2113.8
(batch)
extract
Amount of ions
23 500 5.60 440 42.2 1195.6
swept by tap water
Eluted amount of
24 ions by 0.5 M HCl 350 144.0 21.7 6.6 673.4
(First elution)
Eluted amount of
25 ions by 0.5 M HCl 350 41.3 9.8 1.6 188.0
(Second elution)

Table 3.53 shows the overall process sequence starting with feeding the column by
clay extract and followed by two consequent sweeping processes and finally elution
step. The overall process selectivity towards lithium compared to potassium and
magnesium were seen obviously but similar to the previous processes, magnesium
ion seems to behave like lithium ion and decrease the selectivity.

The comparison of the clay extract feed composition with elution solution
composition is given in Table 3.54.

122
Table 3.54 The comparison of the boron clay extract and elution solution
composition after sweeping with tap water

The composition Li (ppm) Na (ppm) K (ppm) Mg (ppm)

Clay extract conc. 388.0 4880.0 530.0 15110.0
Elution solution
411.5 62.1 18.9 1924
conc.
Change of Increased by Decreased by Decreased by Decreased by
concentration 1.06 folds 78.6 folds 28 folds 7.9 folds

It is deduced from Table 3.54 that, sodium and potassium ion concentrations were
reduced by 28 and 78.6 folds respectively which shows the high selectivity of beads
towards lithium ions. However magnesium ion concentration decreases by 7.9 folds
while the lithium concentration stays nearly constant.

3.4.4.4.2 Sweeping of the column with acidic water at pH 6.1

Following Run 14, a sweeping process was applied with acidic water at pH 6.1. The
amount of processed ions in each step are given in Table 3.55.

Table 3.55 The elution of the column saturated with clay extract and swept by acidic
water at pH 6.1

Total
Run volume
Operation Li (mg) Na (mg) K (mg) Mg (mg)
# processed
(ml)
Adsorbed amount
780
14 of ions from clay 224.7 512.7 61.1 2184.6
(batch)
extract
Amount of ions
32 swept by acidic 500 17.4 491 52.3 1491.4
water at pH 6.1
Eluted amount of
33 350 182.5 17.4 4.3 229.8
ions by 0.5 M HCl

123
In sweeping process with acidic solution, the magnesium could be swept more than
the case with deionized water similar to the previous results. The final concentrations
of the overall process are given in Table 3.56.

Table 3.56 The comparison of the boron clay extract and elution solution
composition after sweeping with acidic water

The composition Li (ppm) Na (ppm) K (ppm) Mg (ppm)

Clay extract conc. 412.0 4964.0 584.0 15972.0
Elution solution
521.4 49.7 12.3 656.6
conc.
Change of Increased by Decreased by Decreased by Decreased by
concentration 1.06 folds 100 folds 48 folds 24 folds

When Table 3.54 and Table 3.56 are compared, it is obviously seen that using an
acidic sweeping solution helps to remove more foreign ions from the column and
increase the lithium/foreign ion ratio, thus increase the overall separation efficiency.
However, higher magnesium ion concentration than lithium ion concentration in clay
extract seems to be a problem for recovery of pure lithium carbonate in a real
process. A repetition of the overall process would probably decrease the magnesium
concentration and increase the lithium concentration however an additional
separation unit must be utilized in order to get rid of magnesium content in the
resulting lithium chloride rich solution.

After the analysis of column experiments done in Section 3.4, the following
conclusions can be written:

- Lithium was obviously separated by adsorbents selectively compared to

foreign ions Na, K and Mg.

- Foreign ions were also adsorbed by PSMA content in the beads partially.

- High foreign ion content in feed solution suppressed the capacity of LiMnO
beads.

- In sweeping operation, huge amount of Na, K and Mg ions could be swept

through the column.
124
 - Employing a slightly acidic solution (pH:6.1) in sweeping process obviously
increased the swept amount of foreign ions especially magnesium.

- Most of the lithium adsorbed by the beads stayed within LiMnO content in
the column during sweeping by tap water, deionized water and acidic water
at pH 6.1.

- It is possible to further concentrate lithium content in a solution by

consecutive adsorption-sweeping-elution cycles up to a point where
thermodynamic limitation of the process emerges.

3.4.5 Separation of lithium by a post-treatment with LiMnO particles

following column adsorption

In this section, the concentrated brine taken from adsorption column was further
treated with LiMnO particles in order to increase the lithium and reduce the foreign
ion content. The experiments were done batchwise by direct contact of adsorption
column effluent with LiMnO particles at pH:9.8 in a large volume beaker. The initial
and final concentrations of adsorption process are shown in Table 3.57.

Table 3.57 The initial and final concentrations of the adsorption process done with
column effluent

Brine effluent of Capacity

Volume Adsorbent Initial conc. Final conc.
adsorption (mg ion/
(ml) (g) (ppm) (ppm)
column g ads.)
261 ppm Li 0.3 ppm Li 17.38 mg Li
32 ppm Na 21.1 ppm Na 0.73 mg Na
Artificial brine 1000 15
20.6 ppm K 0.35 ppm K 1.35 mg K
102.5 ppm Mg 83.5 ppm Mg 1.27 mg Mg
521.4 ppm Li 0.36 ppm Li 17.37 mg Li
49.7 ppm Na 28.4 ppm Na 0.71 mg Na
Clay extract 1000 30
12.3 ppm K 0.3 ppm K 0.4 mg K
656.6 Mg 617.9 ppm Mg 1.29 mg Mg

After the adsorption process, LiMnO particles were let to settle down and liquid part
was decanted. Then, LiMnO particles were mixed with 100 ml demineralized water
125
to sweep the foreign ions which was followed by elution of the remaining ion
content in the particles. The elution of the particles was done in three consecutive
steps by using 100 ml of 0.5 M HCl in each step. The eluted amount of ions and
concentrations in each step were tabulated and given in Table 3.58.

Table 3.58 The amount of eluted ions and concentrations in elution process of post-
treatment

Vol. of Reco- Eluted amount of ions Elution solution concentration

Brine Elut. # elut very (mg) (ppm)
soln (ml) (%) Li Na K Mg Li Na K Mg

1. elut. 100 64.2 167.6 6.4 12.9 12.4 1675.6 64.4 128.9 124.1
Artificial

2. elut. 100 21.3 55.6 2.2 4.2 4.1 555.9 22.2 42.1 41.2
brine

3. elut. 100 8 21.1 0.8 1.5 1.5 211.4 8.3 15.4 15.3
TOTAL 300 93.6 244.3 9.4 18.8 18.1
1. elut. 100 61.7 321.7 12.6 8.4 23.3 3217.0 126.2 83.8 233.1
2. elut. 100 19.8 103.2 4.1 2.5 7.5 1032.4 40.6 25.1 75.4
extract
Clay

3. elut. 100 7.4 38.6 1.5 0.9 2.7 385.8 14.9 8.8 27.5
TOTAL 300 88.9 463.5 18.2 11.8 33.6

Table 3.58 shows that the effluent lithium concentration from adsorption column can
be further concentrated significantly by direct act of LiMnO particles while reducing
the foreign ion content of the brine. The efficiency of the overall process starting
from feeding the artificial brine/clay extract to the adsorption column in consecutive
steps which is followed by the post-treatment with LiMnO particles were analyzed in
terms of lithium concentrations, foreign ion concentrations and the amount of
lithium lost in each step.

3.4.5.1 Efficiency of the overall process starting from artificial brine

In order to have 1675 ppm Li, 64.4 ppm Na, 128.9 ppm K and 124.1 ppm Mg in 100
ml of water, 1517 L of concentrated sea water was processed with a concentration of
0.43 ppm Li, 53430 ppm Na, 4464 ppm K and 9210 ppm Mg. In the first adsorption-
sweeping-elution cycle, 458 mg Li was adsorbed by the LiMnO-PSMA beads, while
65 mg of Li was lost during sweeping process. In the second adsorption-sweeping-

126
elution cycle, 59 mg of Li was lost during sweeping processes. And lastly, in the
post-treatment with LiMnO particles, 17 mg of Li was lost during sweeping
processes. In the overall process, 141 mg of Li was lost, which means that 69.2 % of
the initially adsorbed lithium could be recovered in the final solution. The lithium
concentrations, lithium to foreign ion ratio and lithium loss in each step were
summarized in Table 3.59.

Table 3.59 Li concentration, Li/foreign ion ratio, Li loss in each separation step
starting from artificial brine

Li
Li percantage
Li Loss Li Loss
Brine Na/Li K/Li Mg/Li conc. among total
(mg) (%)
(ppm) cations
(%)
Initial brine 124,256 10,381 21,419 0.43 6.408x10-4 - -
After first ads.-
3.0 2.2 8.6 27.4 6.8 65 14.2
sweep.-elut. cycle
After second ads.-
0.12 0.08 0.39 261.0 62.7 59 12.9
sweep.-elut. cycle
After post
0.04 0.08 0.07 1675.6 84.0 17 3.7
treatment
TOTAL 141 30.8

A feasibility study of recovering lithium from the concentrated brine in Çamaltı

Salina is given in APPENDIX D.

3.4.5.2 Efficiency of the overall process starting from clay extract

In order to have 3217 ppm Li, 126.2 ppm Na, 83.8 ppm K and 233.1 ppm Mg in 100
ml of water, 2.23 L of clay extract was processed with a concentration of 412 ppm
Li, 4964 ppm Na, 584 ppm K and 15972 ppm Mg. In the adsorption-sweeping-
elution cycle, 642 mg Li was adsorbed, while 42 mg of Li was lost during sweeping
process. In the post-treatment with LiMnO particles, 58.4 mg of Li was lost during
sweeping processes. In the overall process, 100.4 mg of Li was lost, which means

127
that 84.4 % of the initially adsorbed lithium could be recovered. The lithium
concentrations, lithium to foreign ion ratio and lithium loss in each step were
summarized in Table 3.60.

Table 3.60 Li concentration, Li/foreign ion ratio, Li loss in each separation step
starting from clay extract

Li
Li percantage
Li Loss Li Loss
Brine Na/Li K/Li Mg/Li conc. among total
(mg) (%)
(ppm) cations
(%)
Initial clay extract 12.1 1.4 38.8 412.0 1.9 - -
After ads.-sweep.-
0.09 0.02 1.26 521.4 42.1 42.0 6.5
elut. cycle
After post
0.04 0.02 0.07 3217 88.5 58.4 9.1
treatment
TOTAL 100.4 15.6

An overall process was proposed for the mass production of lithium carbonate from
boron clay and a feasibility study was done which was shown in APPENDIX D.

After the series of operations, lithium concentration in the artificial brine and clay
extract was increased to 1675.6 and 3217 ppm respectively while reducing the Li/Na,
Li/K, Li/Mg ratio significantly. The concentrated brine would then be sent to pools
to let water content evaporate and bischofite salt (MgCl26H20) and lithium carnalite
salt (LiClMgCl27H2O) precipitate in conventional processes. The remaining
magnesium content in the brines may be removed by utilizing lime in the processes.

128
 CHAPTER 4

CONCLUSIONS

4 CONCLUSIONS

The aim of this study is to increase the efficiency of lithium separation from brines
by liquid-liquid extraction and adsorption methods with various materials
synthesized throughout the study. The conclusions driven from the study can be
divided into three parts in terms of the techniques employed.

Liquid-liquid extraction

- Hexyl, octyl, dibutyl and dihexyl formamides were synthesized by direct

reaction of formic acid and related alkyl amines in high yields.
- Among those formamides, hexyformamide showed the highest distribution
coefficient 0.14 (at 25 oC and pH=4) for LiCl extraction. This is the highest
value cited in the literature for a single solvent so far.
- The LiCl extraction process was shown to be highly selective for Li,
compared to Na, K ions.
- Li selectivity is somewhat low with respect to calcium ions. However
calcium ions are generally absent in concentrated sea water due to its
precipitation as sulfates and carbonates during evaporation in ponds.
- It was found that extracted lithium could be eluted with 1.0 M HCl solution,
so that 93 % lithium was stripped in the first contact. Full recovery of lithium

129
 could be achieved by repeating the acid treatments. In these conditions no
hydrolysis of formamide was detected.
- These results implied that hexyl formamide seemed to be a promising
candidate for practical use in extraction of lithium chloride from concentrated
brines.

Batch adsorption

- Lithium manganese oxides (LiMnO) were synthesized by solid-solid and

hydrothermal reaction procedures. The highest capacity was found as 22.8
mg Li/g adsorbent at pH:10.1 by the method used in the study of Ooi et al.
- As pH increases from 6.3 to 10.2, the capacity values increases from 5.2 to
22.8 mg Li/g adsorbent.
- The initial lithium concentration did not affect the capacity value of the
adsorbents.
- The kinetics study showed that the adsorption process was totally completed
in 20 hours.
- The equilibrium isotherm calculations showed that, the adsorption process
was mostly fitted into Langmuir isotherm with an R2 value of 0.997.
- The stability experiments of the adsorbents indicated that the capacity of the
adsorbents decreased by around 1 % in each consecutive cycle.
- LiMnO adsorbents showed strong selectivity towards lithium; 27 times
higher than sodium, 12.5 times higher than potassium and 13.3 times higher
than magnesium ions.

Column adsorption

- The LiMnO particles were integrated into poly(styrene-maleic anhydride-

glycidyl methacrylate) (PSMA) polymer and employed in column adsorption.
- PSMA has been utilized as organic metal oxide support for the first time and
gave superior results compared to similar polymers cited in the literature.
- The properties which make PSMA superior than the other polymers are its
self-crosslinking and swelling ability.

130
- Self-crosslinking makes PSMA-LiMnO beads easy to process, tough and
stabile.
- PSMA-LiMnO beads were stable up to pH:9. At basic pH 9 it swells 4 %,
which increased the diffusion rate of lithium and hydrogen ions in and out of
the beads.
- The capacity of the column filled with LiMnO beads was calculated as 221.2
mg Li at pH 9 with a 20 ppm Li feed solution which was the 94 % of the
theoretical capacity.
- The capacity of the column was found as 2.51 mg Na, 5.78 mg K and 52.9
mg Mg when each foreign ion was fed to the column in equal operating
conditions.
- It was found out that, the capacity value of the column for lithium was
suppressed in the presence of high foreign ion concentrations.
- Sweeping process before eluting the column decreased the foreign ion
concentration in elution solution. Sweeping with acidic water at pH:6.1 gave
the most satisfactory results.
- For artificial brine, after two adsorption-sweeping-elution cycles and post-
treatment by LiMnO, the lithium concentration was increased to 1675.6 ppm
from 0.43 ppm. The lithium concentration among the cations in the brine was
increased to 84 % from 6.408x10-4 % while 30.8 % of the lithium in the brine
was lost during the process.
- For clay extract, after one adsorption-sweeping-elution cycle and post-
treatment by LiMnO, the lithium concentration was increased to 3217 ppm
from 412 ppm. The lithium concentration among the cations in the brine was
increased to 88.5 % from 1.9 % while 15.6 % of the lithium in the brine was
lost during the process.
- It is finally concluded that, adsorption via LiMnO-PSMA beads can be used
to concentrate lithium content in brines selectively which makes adsorption
process a promising method to be employed in conventional lithium
separation processes.

131
132
 CHAPTER 5

RECOMMENDATIONS

RECOMMENDATIONS

Adsorption via LiMnO particles is a promising method to be conjugated with

conventional lithium carbonate production process from brines especially where the
foreign ion content is remarkably high. Boron clay disposed in Bigadiç is a valuable
source of lithium with its 3000 ppm Li content. The experiment held on this study
shows a way to recover lithium from the clay in a feasible way. As a future
prospect, a pilot plant project can be proposed to produce an industrial grade lithium
carbonate which will also incorporate the knowledge and infrastructure in ETİ
Mining Company.

For further investigation,

- The optimum pH of the sweeping process can be determined where the

minimum amount of lithium is lost and maximum amount of foreign ions are
swept
- The recycle pathways must be determined in the overall process to loose
minimum amount of lithium in consecutive elution operations
- The precipitation regimes of the magnesium salts like bischoffite and
carnalite in pools must be determined to get rid of the magnesium content in
the processed brine

133
 - A more realistic feasibility analysis can be done which takes into account the
total cost and time needs of the process while considering to produce whole
possible goods (i.e KCl, MgCl2, boron derivatives) from the clay as by-
products of lithium carbonate.

134
 REFERENCES

REFERENCES

[1] D. Linden and T. B. Reddy, Handbook of Batteries. 2002, p. 328.

[2] J. H. Fishwick, Applications of lithium in ceramics. Boston Cahners, 1974, p.

156.

[3] USGS Report, “Mineral Commodity Summaries 2010”

[4] Kirk-Othmer, “Lithium and lithium compounds,” Kirk-Othmer Encyclopedia

of Chemical Technology, Volume 15. 2005.

[5] “Is Lithium-ion the Ideal Battery?,” 2010. [Online]. Available:

http://batteryuniversity.com/learn/article/is_lithium_ion_the_ideal_battery/3.
[Accessed: 17-Aug-2014].

[6] E. Cairns, “Batteries and Fuel Cells,” Newsletter, Coll. Chem., U. Calif.,
Berkeley, vol. 10, no. 3, p. 4, 2002.

[7] Anon, “New Lithium Battery Design,” Sci. News 158, p. 399, 2000.

[8] M. A. Delucchi and T. E. Lipman, “An analysis of the retail and lifecycle cost
of battery-powered electric vehicles,” Transp. Res. Part D, vol. 6, pp. 371–
404, 2001.

[9] E. E. Richard L. Nailen, “Grease: What it is; How it Works,” Electr. Appar.,
vol. April, 2002.

[10] K. Murakami, H. Sato, and K. Watanabe, “A new bubble-point-pressure

correlation for the binary LiBr/H2O solution,” Int. J. Thermophys., vol. 16, no.
3, pp. 811–820, May 1995.

[11] E. A. Starke, T. H. Sanders, and I. G. Palmer, “New Approaches to Alloy

Development in the Al-Li System,” JOM, vol. 33, no. 8, pp. 24–33, Dec.
2013.

[12] G. W. Edward W. Abel, Francis Gordon Albert Stone, Comprehensive

Organometallic Chemistry 2. Elsevier, p. 667, 1995.

135
[13] I. Natori, “Synthesis of Polymers with an Alicyclic Structure in the Main
Chain. Living Anionic Polymerization of 1,3-Cyclohexadiene with the n -
Butyllithium/N,N,N‘,N‘-Tetramethyl-ethylenediamine System ”
Macromolecules, vol. 30, no. 12, pp. 3696–3697, Jun. 1997.

[14] “FMC Lithium-More than lithium-World Pharmaceutical Frontiers.” [Online].

Available: http://www.worldpharmaceuticals.net/contractors/chemicals-and-
catalysts/fmc-lithium/. [Accessed: 17-Aug-2014].

[15] H. Zerriffi, Tritium: The environmental, health, budgetary, and strategic

effects of the Department of Energy’s decision to produce tritium. Institute for
Energy and Environmental Research, 1996.

[16] N. Tsoulfanidis, Measurement and Detection of Radiation. Washington, D.C.:

Taylor & Francis, pp. 467 – 501, 1995.

[17] National Research Council (U.S.). Committee on Separations Technology and

Transmutation Systems and N.R.C. (U.S.) C. on S.T. and T. Systems, Nuclear
wastes: technologies for separations and transmutation. National Academies
Press, 1996.

[18] M. Makara-studziñska, A. Koœlak, J. Morylowska-topolska, and A.

Urbañska, “Lithium Therapy – The effectiveness of the medicine, side
symptoms, complications and their influence on the quality of the life in
affective diseases,” J. Elem., vol. 15, no. 2, pp. 393–403, 2010.

[19] T. Ichikawa, S. Isobe, N. Hanada, and H. Fujii, “Lithium nitride for reversible
hydrogen storage,” J. Alloys Compd., vol. 365, no. 1–2, pp. 271–276, Feb.
2004.

[20] Z. Xiong, C. K. Yong, G. Wu, P. Chen, W. Shaw, A. Karkamkar, T. Autrey,

M. O. Jones, S. R. Johnson, P. P. Edwards, and W. I. F. David, “High-
capacity hydrogen storage in lithium and sodium amidoboranes.,” Nat.
Mater., vol. 7, no. 2, pp. 138–41, Feb. 2008.

[21] Donald E. Garrett, Handbook of Lithium and Natural Calcium Chloride.

Elsevier, 2004.

[22] USGS Report, “Mineral Commodity Summaries 2012.”

[23] USGS Report, “Mineral Commodity Summaries 2013.”

[24] USGS Report, “Mineral Commodity Summaries 2014.”

[25] M. . D. M. Saller, “Lithium takes charge, supply and demand reviewed,” Ind.
Miner., vol. 37, no. March, 2000.

136
[26] Schaller W.T., “Lithium Minerals Industrial Minerals and Rocks,” Am. Inst.
Min. Met. Eng., p. 427, 1937.

[27] E. Siame and R. D. Pascoe, “Extraction of lithium from micaceous waste from
china clay production,” Miner. Eng., vol. 24, no. 14, pp. 1595–1602, Nov.
2011.

[28] S. D. M. Norton J. J., “Lithium Resources of North America,” Geol. Surv.

Bull., vol. 1027, no. G, 1955.

[29] M. Tamlin, A. Sheth, and D. J. McCracken, “Lithium,” Min. Eng., 2002.

[30] M. J. J., “Lithium and lithium compounds,” Encyclopedia of Chemical

Processing and Design. p. Vol 28, 1988.

[31] W. A. Averill and D. L. Olson, “A review of extractive processes for lithium

from ores and brines,” Energy, vol. 3, no. 3, pp. 305–313, Jun. 1978.

[32] G. O. G. Löf and W. K. Lewis, “Lithium Chloride from Lepidolite,” Ind. Eng.
Chem., vol. 34, no. 2, pp. 209–216, Feb. 1942.

[33] “Salar Brines.” [Online]. Available: http://www.sqm.com/en-

us/acercadesqm/recursosnaturales/salmuera.aspx. [Accessed: 19-Aug-2014].

[34] “Resources & Recycling ǀ Rockwood Lithium.” [Online]. Available:

http://www.rockwoodlithium.com/resources-recycling/. [Accessed: 19-Aug-
2014].

[35] “Company Overview.”

[Online].Available: http://www.orocobre.com/About_Us.htm. [Accessed: 19-Aug-

2014].

[36] “Who We Are: the sustainable materials technology company| Simbol, Inc.”
[Online]. Available: http://www.simbolmaterials.com/who_we_are.htm.
[Accessed: 20-Aug-2014].

[37] R. and S. J. Hollender, “Bolivia and its Lithium,” 2010.

[38] W. Huang, X. Wang, .Z. Sun, Z. Nie, Z. Sha, “Zabuye Salt Lake solar pond in
Tibet , China : Construction and operational experience,” Nat. Resour.
Environ. Issues, vol. 15, 2009.

[39] “South Korea aims to produce lithium from seawater | Business Spectator.”
[Online].Available:
http://www.businessspectator.com.au/news/2011/1/21/cleantech/south-korea-
aims-produce-lithium-seawater. [Accessed: 20-Aug-2014].

137
[40] “Posco cuts lithium recovery time from 12 months to 8 hours - BNamericas.”
[Online]. Available: http://www.bnamericas.com/news/mining/posco-cuts-
lithium-recovery-time-from-12-months-to-8-hours. [Accessed: 20-Aug-2014].

[41] Anon, “Nevada Brine Supports a Big New Lithium Plant,” Chem. Eng., pp.
86–88, 1966.

[42] D. E. Garrett, Potash; Geology, Processing, Uses and Phase Data. London:
Chapman & Hall, 1996.

[43] D. E. Garrett, Borates : Handbook of Deposits , Processing , Properties , and

Use , 1998 , 483 pages ,. San Diego, CA: Academic Press, 2001.

[44] Coad M., “Lithium production in Salar de Atacama,” Industirial Miner., vol.
Nov., p. 27, 1984.

[45] I. F. Wilkomirsky, “Extractive Metallurgy of Lithium,” Miner. Rev., vol.

53(221), pp. 31–36, 1998.

[46] J. Susan, “Production of lithium metal grade lithium chloride from lithium-
containing brine,” 1990.

[47] W. C. Baumann, “Composition for the recovery of lithium values from brine
and process of making/using said composition,” 2001.

[48] A. N. Mordoğan H., Helvacı C., Malayoğlu U., “Lityumun Tersiyer bor
yatakları ve güncel göllerdeki varlığı, dağılımı ve kazanılma olanakları,”
1994.

[49] B. Holdorf, H., Ziegenbalg, G., and Schmidt, “Recovery of Lithium

Compounds from Natural Salt Brines.,” in Seventh Symposium on Salt, pp.
571–595, 1993

[50] G. G. Gabra and A. E. Torma, “Lithium chloride extraction by n-butanol,”

Hydrometallurgy, vol. 3, no. 1, pp. 23–33, Jan. 1978.

[51] J. A. Hermann, “Preparation and purification of lithium chloride,” 1966.

[52] S. Kaneko and W. Takahashi, “Adsorption of lithium in sea water on

alumina—magnesia mixed-oxide gels,” Colloids and Surfaces, vol. 47, pp.
69–79, Jan. 1990.

[53] K. Scbwochau, “Extraction of Metals from Sea Water,” Top. Curr. Chem.,
vol. 124, pp. 91–133, 1984.

[54] R. D. Goodenough, “Recovery of Lithium,” 1960.

[55] M. P. Neipert and C. K. Bon, “Method of lithium recovery,” 1967.

138
[56] K. Rothbaum, H. P., and Middendorf, “Lithium extraction from Wairakei
geothermal waters,” NZ J. Technol., vol. 2, no. 4, pp. 231–235, 1986.

[57] H. Pauwels and C. Fouillac, “Lithium recovery from geothermal waters of

Cesano ( Italy ) and Cronembourg (Alsace , France ),” in 12th New Zealand
Geothermal Workshop, pp. 117–123, 1990.

[58] I. Pelly, “Pelly,” Biotechnol., J. Appl. Chem., vol. 28, pp. 469–475, 1978.

[59] Y. Epstein, J. A., Feist, E. M., Zmora, J., and Marcus, “The Recovery of
Lithium from the Dead Sea,” Hydrometallurgy, vol. 6, pp. 269–275, 1981.

[60] A. D. Ryabtsev, L. T. Menzheres, and A. V. Ten, “Sorption of Lithium from

Brine onto Granular LiCl·2Al(OH)3·mH2O Sorbent under Dynamic
Conditions,” Russ. J. Appl. Chem., vol. 75, no. 7, pp. 1069–1074, Jul. 2002.

[61] A. H. Hamzaoui, B. Jamoussi, and A. M’nif, “Lithium recovery from highly

concentrated solutions: Response surface methodology (RSM) process
parameters optimization,” Hydrometallurgy, vol. 90, no. 1, pp. 1–7, Jan. 2008.

[62] M. Abe, Y. Kanzaki, and R. Chitrakar, “Synthetic inorganic ion-exchange

materials. 44. NMR study of the ion-exchange reaction of lithium on titanium
antimonate and tin antimonate,” J. Phys. Chem., vol. 91, no. 11, pp. 2997–
3001, May 1987.

[63] L. Ma, B. Chen, X. Shi, and K. Zhang, “Li+ extraction/adsorption properties

of Li-Sb-Mn composite oxides in aqueous medium,” Trans. Nonferrous Met.
Soc. China, vol. 21, no. 7, pp. 1660–1664, Jul. 2011.

[64] Y. S. Kim, K. S. No, K. S. Chung, J. C. Lee, and K. Ooi, “Li+ extraction

reactions with spinel-type LiM0.5Mn1.5O4 (M=Ti, Fe) and their electronic
structures,” Mater. Lett., vol. 57, no. 26–27, pp. 4140–4146, Sep. 2003.

[65] C.-M. Park, Y.-U. Kim, and H.-J. Sohn, “Topotactic Li Insertion/Extraction in
Hexagonal Vanadium Monophosphide,” Chem. Mater., vol. 21, no. 23, pp.
5566–5568, Dec. 2009.

[66] L. Fransson and K. Edstro, “Structural investigation of the Li 1 ion insertion /

extraction mechanism in Sn-based composite oxide glasses,” vol. 62, pp.
1213–1218, 2001.

[67] C. V. Ramana, a. Mauger, F. Gendron, C. M. Julien, and K. Zaghib, “Study of

the Li-insertion/extraction process in LiFePO4/FePO4,” J. Power Sources, vol.
187, no. 2, pp. 555–564, Feb. 2009.

[68] E. I. Kachibaya, R. a. Imnadze, T. V. Paikidze, and R. a. Akhvlediani,

“Cathodic materials for lithium-ion batteries based on spinels LixMn2−yMeyO4:
Synthesis, phase composition, and structure of LixMn2−y CryO4 at x = 1.0−1.2
139
 and y = 0−0.5,” Russ. J. Electrochem., vol. 42, no. 11, pp. 1224–1234, Nov.
2006.

[69] Z. Gong and Y. Yang, “Recent advances in the research of polyanion-type

cathode materials for Li-ion batteries,” Energy Environ. Sci., vol. 4, no. 9, p.
3223, 2011.

[70] R. Chitrakar, H. Kanoh, Y. Miyai, and K. Ooi, “Recovery of Lithium from

Seawater Using Manganese Oxide Adsorbent ( H1.6Mn1.6O4 ) Derived from
Li1.6Mn1.6O4,” Ind. Eng. Chem. Res., vol. 40, pp. 2054–2058, 2001.

[71] X. M. Shen, “Phase transitions and ion exchange behavior of electrolytically

prepared manganese dioxide,” J. Solid State Chem., vol. 64, p. 270, 1986.

[72] K. Ooi and M. Abe, “Topotactic Li+ Insertion to X-Mn02 in the Aqueous
Phase,” Langmuir, no. 5, pp. 150–157, 1989.

[73] K. Ooi, Y. Miyai, and J. Sakakihara, “Mechanism of lithium(1+) insertion in

spinel-type manganese oxide. Redox and ion-exchange reactions,” Langmuir,
vol. 7, no. 6, pp. 1167–1171, Jun. 1991.

[74] Q. Feng, Y. Miyai, H. Kanoh, and K. Ooi, “Li+ Extraction/Insertion with

Spinel-Type Lithium Manganese Oxides. Characterization of Redox-Type and
Ion-Exchange-Type Sites,” Langmuir, vol. 8, pp. 1861–1867, 1992.

[75] H. Welfare, “Manganese oxide minerals : Crystal structures and economic

and,” vol. 96, no. March, pp. 3447–3454, 1999.

[76] Q. Feng, “Hydrothermal Synthesis of Lithium and Sodium Manganese Oxides

and Their Metal Ion Extraction-Insertion Reactions,” no. 23, pp. 1226–1232,
1995.

[77] R. J. Ammundsen Brett, Jones Deborah J., “Mechanism of Proton Insertion

and Characterization of the Proton Sites in Lithium Manganate Spinels,”
Chem. Mater., vol. 7, no. 11, pp. 2151–2160, 1995.

[78] M. H. Kazuharu YOSHIZUKA, Ayuko KITAJOU, “Lithium recovery from

geothermal waters of Cesano (Italy) and Cronembourg (Alsace , France ),”
Ars Separatoria Acta, vol. 4, pp. 78–85, 2006.

[79] S. N. and K. Y. A Kitajou, T Suzuki, “Selective recovery of lithium from

seawater using a novel MnO2 type adsorbent II – Enhancement of lithium ion
selectivity of the adsorbent.,” Ars Separatoria Acta, vol. 2, pp. 97–106, 2003.

[80] K. Yoshizuka, A. Kitajou, and M. Holba, “Selective recovery of lithium from

seawater using a novel MnO2 type adsorbent III - benchmark evaluation,” Ars
Separatoria Acta, vol. No. 4, pp. 78–85, 2006.

140
[81] S. Nishihama, K. Onishi, and K. Yoshizuka, “Selective Recovery Process of
Lithium from Seawater Using Integrated Ion Exchange Methods,” Solvent
Extr. Ion Exch., vol. 29, no. 3, pp. 421–431, May 2011.

[82] L. Wang, W. Ma, R. Liu, H. Y. Li, and C. G. Meng, “Correlation between Li +

adsorption capacity and the preparation conditions of spinel lithium
manganese precursor,” Solid State Ionics, vol. 177, no. 17–18, pp. 1421–1428,
Jul. 2006.

[83] L. Wang, C. G. Meng, M. Han, and W. Ma, “Lithium uptake in fixed-pH

solution by ion sieves,” J. Colloid Interface Sci., vol. 325, no. 1, pp. 31–40,
Sep. 2008.

[84] L. Wang, C. G. Meng, and W. Ma, “Study on Li + uptake by lithium ion-sieve

via the pH technique,” Colloids Surfaces A Physicochem. Eng. Asp., vol. 334,
no. 1–3, pp. 34–39, Feb. 2009.

[85] L. Tian, W. Ma, and M. Han, “Adsorption behavior of Li+ onto nano-lithium
ion sieve from hybrid magnesium/lithium manganese oxide,” Chem. Eng. J.,
vol. 156, no. 1, pp. 134–140, Jan. 2010.

[86] C. Özgür, “Preparation and characterization of LiMn2O4 ion-sieve with high

Li+ adsorption rate by ultrasonic spray pyrolysis,” Solid State Ionics, vol. 181,
no. 31–32, pp. 1425–1428, Oct. 2010.

[87] K. Ooi and J. Sakakihara, “Mechanism of Li+ Insertion in Spinel-Type

Manganese Oxide. Redox and Ion-Exchange Reactions,” Langmuir, vol. 7,
pp. 1167–1171, 1991.

[88] Q.-H. Zhang, S.-P. Li, S.-Y. Sun, X.-S. Yin, and J.-G. Yu, “LiMn2O4 spinel
direct synthesis and lithium ion selective adsorption,” Chem. Eng. Sci., vol.
65, no. 1, pp. 169–173, Jan. 2010.

[89] R. R. Rashmi, “Preparation and characterization of manganous manganic

oxide (Mn3O4),” J. Mater. Sci. Mater. Electron., vol. 3, pp. 257–262, 2002.

[90] Z. M. O. Rzaev, “Complex-Radical Terpolymerization of Glycidyl ( Methyl )

Methacrylates , Styrene , and Maleic Anhydride,” no. May 1998, pp. 1095–
1102, 1999.

[91] L. Alders, Liquid-liquid extraction : theory and laboratory practice / Alders L.

Elsevier, 1959.

[92] Y. Uludag, H. Ö. Özbelge, and L. Yilmaz, “Removal of mercury from

aqueous solutions via polymer-enhanced ultrafiltration,” J. Memb. Sci., vol.
129, no. 1, pp. 93–99, Jun. 1997.

141
[93] W. G. H. George E. P. Box , J. Stuart Hunter, Statistics for Experimenters,
2nd ed., vol. 21, no. 3. J. Wiley, 2005.

[94] J. M. Paulsen and J. R. Dahn, “Phase Diagram of Li-Mn-O Spinel in Air,”

Chem. Mater., vol. 4, pp. 3065–3079, 1999.

[95] et al Gerhard Ertl, Helmut Knözinger, Ferdi Schüth, Handbook of

Heterogeneous Catalysis. WILEY, 2008, p. 4270.

[96] C. Valley, “Product Bulletin,” 2014.

[97] G. Xiao, K. Tong, L. Zhou, J. Xiao, S. Sun, P. Li, and J. Yu, “Adsorption and
Desorption Behavior of Lithium Ion in Spherical PVC − MnO2 Ion Sieve,”
Ind. Eng. Chem. Res., vol. 51, pp. 10921–10929, 2012.

[98] L.-W. Ma, B.-Z. Chen, Y. Chen, and X.-C. Shi, “Preparation, characterization
and adsorptive properties of foam-type lithium adsorbent,” Microporous
Mesoporous Mater., vol. 142, no. 1, pp. 147–153, Jun. 2011.

[99] T. Ryu, J. Shin, J. Ryu, I. Park, H. Hong, B.-G. Kim, and K.-S. Chung,
“Preparation and Characterization of a Cylinder-Type Adsorbent for the
Recovery of Lithium from Seawater,” Mater. Trans., vol. 54, no. 6, pp. 1029–
1033, 2013.

[100] Q. H. Zhang, S. Sun, S. Li, H. Jiang, and J. G. Yu, “Adsorption of lithium

ions on novel nanocrystal MnO2,” Chem. Eng. Sci., vol. 62, no. 18–20, pp.
4869–4874, Sep. 2007.

[101] “Kaliteli ve Güvenli Şebeke Suyu, ASKİ Genel Müdürlüğü.” [Online].

Available: http://www.aski.gov.tr/tr/kaliteli-sebeke-suyu/index.html.
[Accessed: 24-Aug-2014].

[102] A. Seidell and W. F. Linke, Solubilities of Inorganic and Organic

Compounds. Van Nostrand, 1952.

[103] Rockwood Lithium, Quotation # 20019109, no. 047. 2014, pp. 1–3, 2014

142
 APPENDIX A

A.

ALTERNATIVE MATERIALS EMPLOYED IN LITHIUM SEPARATION

In this thesis work, various alternative materials were synthesized and employed in
lithium separation besides the materials used in liquid-liquid extraction and
adsorption methods. Successfully synthesized alternative materials are comb like
polyethylene glycol (C-PEG), poly glycidyl methacrylate, poly vinyl formamide and
epoxy-lithium complex. In addition to that, synthesis trials for water soluble crown
ether polymer and water insoluble crown ether polymer were done, unfortunately the
synthesis of those polymers cannot be achieved throughout the study.

The synthesis, characterization and the lithium separation performance of the

successfully synthesized materials were described below.

Comb like poly ethylene glycol (C-PEG)

C-PEG has tines in the form of long poly ethylene glycol chains which are supposed
to form coordination complexes with lithium ions. The aim was to dissolve C-PEG
in lithium brine and form coordination bonds between lithium ions and oxygen
atoms in tines. The coordination bond would selectively separate lithium from brines
with multi-ion content. After the coordination bonds formed the mixture was
ultrafiltrated in a ultrafiltration set-up, so that the lithium would be separated from
the rest of the solution in an efficient and selective way.

143
The synthesis of comb like poly ethylene glycols (C-PEG) was first started from the
monomer synthesis. Polyethylene glycol with a certain molecular weight was reacted
with methacryloyl chloride in toluene. After that the resulting monomer was
radically polymerized and comb-like poly ethylene glycol was synthesized. The
reaction sequence is shown in Figure A.1.

Figure A.1 Reaction sequence of comb like poly ethylene glycol

The characterization of the monomer was done by 1H-NMR spectrum as given in

Figure A.2.

144
Figure A.2 1H-NMR spectrum of C-PEG monomer

The signals observed at different ppm values were paired with relevant chemical
groups. The signal at 3.3 ppm belongs to methyl end group attached to oxygen. The
strong signal at 3.7 ppm shows the methylene bridge protons in polymer backbone.
Two signals at 5.6 and 6.2 belong to olefinic hydrogens and signal at 2 ppm shows
the methyl group protons attached to double bond.

The characterization of the polymer (C-PEG) was done by 1H-NMR spectrum as

given in Figure A.3.

145
Figure A.3 1H-NMR spectrum of C-PEG

In the structure of polymer, two types of protons were observed in 3.3 ppm and 3.6
ppm which belong to methyl end group and methylene bridge protons respectively.
The detailed synthesis procedure was given in two steps.

- Synthesis of PEG functionalized with methacrylate;

75 g (100 mmol) of polyethylene glycol (PEG, Mn:750 Da) was weighed, dissolved
in toluene and put into a 250 ml volume flask. Than 11.2 g (110 mmol) of distilled
triethyl amine was added to the mixture. When this mixture was stirred at 0 oC,
13.25 g (110 mmol) methacryloyl chloride was added dropwise. When the first drops
fed into the mixture, a precipitate of triethyl ammonium was appeared in the
solution. After the precipitate was filtered from the solution, the excess of toluene
was evaporated and PEG with functionalized with methacrylate was precipitated in
hexane. The resulting product (MPEG) was dried in vacuum dryer at 40 oC for 12
hours.

- Synthesis of comb like PEG (C-PEG)

8.35 g (100 mmol) of MPEG was weighed and dissolved in 100 ml water in a 100 ml
volume erlenmeyer. After nitrogen gas passed through the mixture, the erlenmeyer

146
 o
was placed into an oil bath at 60 C and 54 mg (2mmol) 2,2'-
Azobis(isobutyramidine) dihydrochloride (AABH), ([AABH]/[MPEG]: 1/50) was
added to the mixture. The reaction continued until the magnet in the Erlenmeyer
cannot mix the solution due to the increase in the viscosity. Then, the resulting
solution was filtrated with a proper size membrane in order to use in polymer
enhanced ultrafiltration (PEUF) experiments. The final concentration was measured
by taking a sample from it and dried in a drying oven.

The separation efficiency of C-PEG was analyzed by utilizing a polymer enhanced

ultrafiltration set-up. To observe the time where the permeate flux and concentration
reaches steady state, results of a representative experimental run was given in Table
A.1. It was seen that steady state was attained after two hours.

Table A.1 A representative experimental run for a PEUF experiment

Polymer : C-PEG

Polymer MW : ~50 kDa

Loading : 0.01

pH : 7.3

∆P : 200kPa

MWCO : 20 kDa

Time (hour) 0 1 2 3 4

Fp (L/m2h) 4.8 4.6 4.6 4.6

pH(Feed) 7.3 7.4 7.3 7.3 7.3

T (oC) 25.0 25.2 25.4 25.3 25.3

Li Cf (ppm) 8.0 7.9 7.9 7.9 7.9

Li Cp (ppm) 6.8 6.7 6.7 6.7

R 0.14 0.15 0.15 0.15

Fp,avg=0.15 Cf,avg=7.9 Cp,avg=6.7 Ravg=0.15

147
In order to understand whether C-PEG works in PEUF or not, some preliminary
experiments at different pH values with constant loading (0.01) and temperature (25
o
C) were performed and retention values were calculated. The results are shown in
Figure A.4.

0,40

0,35

0,30

0,25

Retention 0,20
0,15

0,10

0,05

0,00
4 5 6 7 8 9 10 11
pH

Figure A.4 Retentions for different pH values at 25 oC and constant loading (0.01)
for C-PEG

As it can be deduced from Figure A.4, a maximum retention value of 0.35 could be
acquired with C-PEG (Mn:950) around pH 9.

Effect of loading value on retention was also investigated at pH=9 and the results
listed in Table A.2 were acquired.

Table A.2 Retention values for different loadings at 25 oC and pH:9

Retention Loading
0.21 0.1
0.35 0.01

0.28 0.002

148
It was deduced that, when loading value increased above 0.01, retention value started
to decrease. This unusual behavior was attributed to the decrease in polymer
solubility resulting in coiling up of polymer chains.

Next, the selectivity of C-PEG polymers on PEUF was investigated with

concentrated sea water taken from Çamaltı Salina under varying loading values and
the results listed in Table A.3 were obtained.

Table A.3 Retention values for concentrated brine taken from Çamaltı Salina in
PEUF utilizing C-PEG under pH:7.4 and 25 oC

Retention values (R)

Metal ion
Loading:0.1 Loading:0.01 Loading:0.001
Na 0.060 0.090 0.041
K 0.014 0.049 0.001
Mg 0.093 0.134 0.136
Li 0.086 0.178 0.136

It was seen that C-PEG is selective to lithium near sodium and potassium in 2 to 4
folds. For magnesium, the retention values came out to be similar for different
loading values. Only when loading value is 0.01, the retention value of lithium
seems to be higher than magnesium. At this point, the selectivity values of C-PEG
were found as SLi/Na=2, SLi/K=3.6, SLi/Mg=1.3.

Even though the selectivity values are below the expectations, when the high
concentration of potassium (around 10000 ppm), magnesium (around 20000 ppm)
and sodium (around 75000 ppm) near lithium (around 1.25 ppm) is considered, the
results were found meaningful for utilizing PEUF method in separation of lithium.

After these results were examined, the synthesis of C-PEG was done by different
molecular weight PEG methacrylate to highlight the definite separation efficiency of
PEUF with C-PEGs. Additional C-PEGs were synthesized having different Mn
values (300, 500, 950 and 2000). The most satisfactory results were obtained by C-
PEG with a monomer MW around 950 g/mol. The higher MW polymer which was

149
synthesized by PEG methacrylate with Mn: 2000 had a solubility problem and PEUF
set-up could not be utilized.

The synthesis experiments with monomers having Mn values of 300 and 500 did not
give expected retention values. The retention values for C-PEG (Mn: 300) and C-
PEG (Mn: 500) were given in Table A.4 for different pH values at loading: 0.01.

Table A.4 Retention values for different Mn C-PEG under various pH at loading:
0.01

C-PEG C-PEG
pH (Mn: 300) (Mn: 500)
Retention Retention
7 0.04 0.11
8 0.05 0.21
9 0.12 0.21
10 0.33 0.21
11 0.20 0.15

After that, PEUF experiments were conducted without any polymer in order to
understand the effect of membrane on retention values, and the results were
tabulated in Table A.5Table .

Table A.5 Retention values for membrane (PES MWCO: 5kDa) on separation of
lithium via PEUF under varying pH.

pH Retention
6.7 0.15
7.5 0.16
9 0.24
10 0.32
11 0.19

Table A.4 and Table A.5 show that C-PEG does not actually interact with lithium
ions and cause their separation, the main reason of separation is the membrane
barrier. The membrane employed in PEUF experiments was polyether sulfone with a

150
molecular weight cut-off 5 kDa. The molecular structure of PES membrane is seen in
Figure A.5.

Figure A.5 The molecular structure of PES membranes

The interaction of lithium ions with PES membrane can be occurred via unshared
electrons of sulfone and ether group. Since oxygen atoms tend to be more
electrophilic then sulfur atoms, the electron cloud between sulfur and oxygen atoms
is oriented towards oxygen, which means slightly negative charge on oxygen atoms.
The remaining positive charge on sulfur atom can also be stabilized by the resonance
effect of benzene rings in the structure. If we consider that lithium ions present in
solution are in Li+ form, an interaction of lithium ions with sulfone group oxygens is
predictable.

After these findings, the selectivity of membrane towards lithium near foreign ions
was questioned and a mixture of lithium, sodium, potassium, magnesium each
having a concentration around 10 ppm in solution was fed to the PEUF set-up with
the same membrane (PES MWCO: 5 kDa). The results are tabulated in Table A.6.

151
Table A.6 Retention values for Na+, K+, Mg2+ and Li+ with PES membrane at
varying pH values.

Retention Retention Retention Retention Retention

pH + + + 2+
(Li ) (Na ) (K ) (Mg ) (Total)

7.1 0 0.03 0.04 0.05 0.12

7.5 0.02 0.03 0.04 0.01 0.10

8.25 0.01 0.04 0.04 0.03 0.12

9.5 0.03 0.05 0.05 0.10 0.23

10.1 0.03 0.05 0.07 0.07 0.22

As it is obviously deduced from Table A.6, retention values for lithium near foreign
ions are smaller which means that the membrane does not have selectivity towards
lithium. Mg2+ ions have the highest retention value around pH=9.5. It can be
explained by the +2 charge, which attracts magnesium towards slightly negative
sulfur more than other cations. When we look to the total retention values and
compare them with the values in Table A.6, it is seen that the capacity of membrane
to hold cations is somewhat filled by separate individual cations. The small
difference in total retention values can be attributed to the steric effects especially
caused by the magnesium ion which is the largest cation among them with the
highest charge.

152
Poly glycidyl methacrylate

The synthesis of poly glycidyl methacrylate was started with the reaction of glycidyl
methacrylate with dilute sulfuric acid for overnight in order to enhance its water
solubility with introduced hydroxyl groups. The reaction sequence is as shown in
Figure A.6;

O O

O H2SO4 O

OH
O
OH
glycidyl methacrylate hydrolyzed glycidyl methacrylate

Figure A.6 The reaction sequence for hdrolyzed glycidyl methacrylate synthesis

Afterwards, the polymerization of the hydrolyzed monomer was done with 2,2'-
Azobis (isobutyramidine) dihydrochloride in 1/100 molar ratio with 15 % monomer
concentration in water at 65 oC for 4 hours. The resulting polymer has the molecular
structure shown in Figure A.7.

Figure A.7 Molecular structure of hydrolyzed poly glycidyl methacrylate

The resulting poly glycidyl methacrylate polymer becomes insoluble in water due to
slight crosslinking occurs during the polymerization reaction. The gel like polymer
153
was than chopped for 2-3 minutes at 300 rpm in a blender and washed with water
several times. The hydroxyl groups in gel-like polymer are supposed to make
coordination bonds with lithium ions and then separate the lithium ions from
aqueous phase. The logic behind this technique is very similar to the technique used
with comb like poly ethylene glycols. Expect that, this material does not dissolve in
water and does not require a polymer enhanced ultrafiltration technique.

In the separation experiments simply the lithium solution was mixed with gel like
polymer and pH of the solution was adjusted to 10 by ammonium chloride and
ammonia buffer. The solution was mixed for 24 hours and then wait for another 24
hours to gel-like polymers settle down. The supernatant solution was analyzed in
atomic absorbance spectrophotometer. 47.2 ppm feed solution was mixed with 0.33
g gel-like polymer and the capacity was found as 1.2 mg Li/g polymer. This value
appears to be so small compared to the values acquired with manganese oxide
adsorbents. Therefore further experiments with hydrolyzed poly glycidyl
methacrylate gel-like polymers were not held on.

154
Poly vinyl formamide

Formamides are very polar molecules and have strong solvation powers. In Chapter
3.2, alkyl formamides were employed as extraction solvents for lithium chlorides
and fairly high distribution coefficient values were acquired in the experiments.
Therefore, it was thought that, to synthesize a polymer of vinyl formamide can lead
us to high retention values in polymer enhanced ultrafiltration method. The synthesis
of poly vinyl formamide was accomplished by addition polymerization technique
starting from vinyl formamide monomer. 2,2'-Azobis (isobutyramidine)
dihydrochloride was used as an initiator. The final product was dissolved in water
and the polymer solution was first filtrated with a membrane having molecular
weight cut-off: 3-20 kDa. Adjustment of the pH was done by ammonia-ammonium
chloride buffer. The final mixture was fed to the polymer enhanced ultrafiltration
(PEUF) set-up and the experiment was repeated for two different pH values at
loading:0.01 and 25 oC. The resulting retention values are tabulated in Table A.7 and
shown below.

Table A.7 PEUF experiment results by poly vinyl formamide

pH: 8 pH:9
Feed conc. (ppm) 45.68 45.165
Permeate conc.
44.64 43.21
(ppm)
Retention 0.023 0.043

As it can be seen from Table A.7, poly vinyl formamide has no functionality in
separation of lithium. Therefore further experiments were not held on.

155
Epoxy-lithium complex

Epoxy lithium complex was synthesized by mixing 1,2 epoxy octane by lithium
chloride solution in NMP and react the epoxide groups by triethyl amine addition at
high temperature. The final product was supposed to have a structure shown in
Figure A.8.

Figure A.8 4 octyl 12-crown-4 ether

This hypothetical molecule does not mix with water due to alkyl groups and can
make complexes with lithium ions by oxygen atoms. Therefore it could be used as
selective liquid-liquid extraction agent to separate lithium from brines.

The H-NMR spectrum of the resulting product is given in Figure A.9.

156
Figure A.9 H-NMR spectrum of 4 octyl 12-crown-4 ether

Figure A.9 shows that the peak which is divided into three at 0.9 ppm represents
methyl groups at the end of four octyl groups. The signals around 1.3 and 1.5 show
the methylene bridge protons near to oxygen. In signal at 3.5 ppm, -CH2- groups on
the ring are detected. They are shifted to right which may be caused by the lithium
ions coordinated in the middle of the ether ring. The synthesis of 4 octyl 12-crown-4
ether started with dissolving lithium chloride in minimum amount of N-methyl
pyrrolidone. Then, 1,2 epoxy octane was added into the mixture. The molar ratio of
lithium chloride to 1,2 epoxy octane was 1:4. The mixture was then heated to 140 oC
and 2 ml of triethyl amine was added to the mixture. Under a reflux, the reaction
continued for 12 hours. The resulting black mixture was mixed with water and boiled
to recover the desired product with water. After 10 hours of distillation, the water
and slightly yellow colored organic material were condensed and mixed with diethyl
ether to extract the organic material from water. The ether phase was separated by a
separation funnel and remaining solvent was evaporated in rotary evaporator which
left the desired product. The resulting product was very low in quantity, so the

157
extraction experiments with that solvent could not be held on. The synthesis route
was presented for further research about topic.

158
Water soluble 12-crown-4 etheric polymer

12-Crown-4 ethers are known to bind with lithium ions by coordination bonds. In
order to employ ethers in PEUF system, water soluble polymers with 12-crown-4
ether side groups must be synthesized. The hydrophobic nature of crown ethers
makes it very difficult to integrate them into a water soluble polymer structure. The
problem was handled by a new synthesis pathway which was proposed to integrate
crown ethers in a water soluble polymer as shown in Figure A.10.

Figure A.10 The pathway of synthesis of water soluble crown ether polymers

The synthesis was started with dripping 0.2 mol sulfonyl chloride by a funnel into an
250 ml erlenmeyer which had 0.1 mol triethylene glycol. The reaction took place in
159
the presence of triethyl amine at 0 oC by the aid of an ice bath. In the second step, 0.1
mol LiOH was dissolved in N-methyl pyrrolidone and mixed with the product. After
the addition of LiOH, no reaction was observed and the sequence was terminated.

The synthesis experiments couldn’t be successful throughout the study, because

during the synthesis, candidate crown ether monomer structure with SO4 leaving
groups must dissolve the lithium hydroxide in step 2. However the candidates could
not dissolve the lithium. Therefore a successful synthesis route could not be
established, but the proposed reaction sequence was given for further investigation.

160
Water insoluble 14-crown-4 etheric polymer

Lastly, another synthesis route was studied to synthesize a 14-crown-4 ether

monomer in order to polymerize it and get a water insoluble adsorbent. The proposed
synthesis has eight steps.

In the first step, equimolar 1,3 propanediol and metallic sodium was mixed (See
Figure A.11). Although sodium metal is very reactive against OH groups the
synthesis could only be done by adding additional methanol (petrol ether is also tried
but does not give reaction). The reaction proceeded under vacuum (100 mbar) at 90
o
C for 24 hours. In order to get rid of methanol, the mixture was heated up to 130 oC
under vacuum in final step.

methanol
Na (step 1)
+Na-O OH
HO OH

Figure A.11 14-crown-4 monomer synthesis step 1

In the second step, sodium propanediol salt was mixed with dibromoethane (See
Figure A.12). In order to sustain a well-mixed system, propanediol salt was first
heated to 60 oC, then dibromoethane was dripped onto it. The reaction took place
very fast, and due to the heat release during reaction, the temperature of the mixture
increased rapidly. The reaction continued for 24 hours at 70 oC. The resulting
product was tried to be separated by dissolving it in diethyl ether, but it didn’t work
out. Than methanol was used to dissolve the mixture and the precipitated salt was
filtered. During the next step which was the evaporation of methanol from the
mixture, additional salt was precipitated, so it was thought that, methanol also
dissolved the salt. In order to get rid of salt, the liquid part was decanted and used in
the next step.

161
 24 h O O
Br 70-80 oC
+
(step 2)
Na-O OH Br
OH HO

Figure A.12 14-crown-4 monomer synthesis step 2

In third step, the product was mixed with butyl lithium in a molar ratio of 1:2 (See
Figure A.13). Butyl lithium could be stored in hexane solution, therefore the reaction
continued in hexane medium at room temperature for 8 hours.

O O O O

2 (step 3)
Li
OH HO OLi LiO

Figure A.13 14-crown-4 monomer synthesis step 3

In the fourth step, glycidol was dissolved with equivolume dichloromethane (See
Figure A.14). Then, benzoyl chloride was dissolved with equivolume
dichloromethane and pyrridine was dissolved with equivolume dichloromethane.
Equimolar benzoyl chloride and pyrridine solutions were dripped onto equimolar
glycidol solution simultaneously. The reaction was held on in an ice bath under
vigorous stirring for 15 hours. After the reaction was completed, the product mixture
was taken into separation funnel. Ice-water and sodium chloride was added to funnel
and the funnel was shaken vigorously. Then the phase separation was observed in the
funnel. The dichloromethane phase was taken and mixed with sodium sulfate in
order to separate remaining water residues. In the rotary evaporator, the
dichloromethane was evaporated and an orange colored viscous liquid was taken as a
product.

162
 O O

HO Cl (step 4)
O
O
O

Figure A.14 14-crown-4 monomer synthesis step 4

In the fifth step, the product was mixed with mesyl chloride (See Figure A.15).
Mesyl chloride was put into a reaction balloon which was placed in an ice bath and
the product of step 4 was dripped onto it. After dripping, the mixture was taken into
room temperature and mixed for 60 hours.

O O
O O

O O S
S Cl (step 5)
O
O Cl O
O

Figure A.15 14-crown-4 monomer synthesis step 5

In the sixth step, the product after step 5 was mixed with the product which was
acquired after step 3 and a condensation reaction was expected to synthesize a
crown-4-ether structure (See Figure A.16). While dripping, a phase separation
occurred in the mixture. The reaction was continued for 15 hours at 80 oC. The
hexane phase was separated into another erlenmeyer and put aside. The remaining
black residue was held in the balloon.

O O
Li+
O O
O
O O O
O S (step 6)
O
O OLi LiO
Cl O
O

Figure A.16 14-crown-4 monomer synthesis step 6

Afterwards a TLC experiment was done in order to understand how many products
were in the hexane phase. TLC paper showed, 7 separate dots, which designates 7

163
different products. At that point, different products could not be separated and the
next two steps could not be done.

The final two steps were designed as shown in Figure A.17 and Figure A.18.

In step 7, the product at step 6 was supposed to mix with acryloyl chloride and a
monomer of 14-crown-4 ether form would form.

O O O O
+
Li Li+
O O O
O O
(step 7)
Cl
O O
O O

Figure A.17 14-crown-4 monomer synthesis step 7

In the final step, 14-crown-4 ether monomer would be polymerized by AIBN at 65

o
C and finally a candidate water insoluble polymer to be used as an adsorbent in
lithium separation from brines would be synthesized.

*
* n
O
O
O AIBN O
(step 8)
65 oC
O O
O O
+
Li
Li+
O O
O O

Figure A.18 14-crown-4 monomer synthesis step 8

164
 The final product could not be syntesistudy but reported for the sake of
folowing investigations.

APPENDIX B

XRD MEASUREMENTS OF SYNTHESIZED

LITHIUM MANGANESE OXIDES

[1], d=4.744(5), 2-theta=18.69(2)

400

350

300
[2], d=2.455(4), 2-theta=36.57(6)

[3], d=2.031(2), 2-theta=44.58(6)

[5], d=1.569(3), 2-theta=58.81(13)

250
[4], d=1.866(5), 2-theta=48.75(13)
Intensity (cps)

200

150

100

50

0
10 20 30 40 50 60
Intensity (cps)

29

-21

-71
10 20 30 40 50 60

2-theta (deg)

Figure B.1 XRD pattern of spinel lithium manganese oxide LMO (350 oC, Li/Mn=1)

165
 Intensity (cps) Intensity (cps) Intensity (cps) Intensity (cps)

166
Li/Mn=1)
Li/Mn=1)
10
0

-55
-15
25
20
40
60
80
100
120
140
160
180
200
220
240
260
280
300
320

10

10
4
0

-96
-46
54
50
100
150
200
250
300
350
400
450
500
550

10
[1], d=4.666(5), 2-theta=19.00(2)

20
20
[1], d=4.729(3), 2-theta=18.748(13)

20
20
[2], d=4.248(13), 2-theta=20.90(6)

30
30

30
30
2-theta (deg)

2-theta (deg)
[2], d=2.430(3), 2-theta=36.96(5)
[3], d=2.465(2), 2-theta=36.43(3)
40
40

[4], d=2.360(4), 2-theta=38.11(7)

40
40
[3], d=2.0226(12), 2-theta=44.77(3)
[5], d=2.0361(10), 2-theta=44.46(2)
50
50

[6], d=1.8660(13), 2-theta=48.76(4)

50
50
60
60

[7], d=1.5671(15), 2-theta=58.88(6)

60
60

Figure B.3 XRD pattern of the spinel lithium manganese oxide LMO (550 oC,
Figure B.2 XRD pattern of the spinel lithium manganese oxide HMO (350 oC,
 Intensity (cps) Intensity (cps)

Li/Mn=1)
10
0

-68
-18
32
50
100
150
200
250
300
350
400

10
[1], d=4.662(4), 2-theta=19.021(17)

20
20

30
30

2-theta (deg)
[2], d=2.426(2), 2-theta=37.02(3)
[3], d=2.320(2), 2-theta=38.78(4)
40
40

[4], d=2.0221(12), 2-theta=44.78(3)

[5], d=1.858(3), 2-theta=48.99(8)

50
50

60
60

Figure B.4 XRD pattern of the spinel lithium manganese oxide HMO (550 oC,

167
 Intensity (cps) Intensity (cps) Intensity (cps) Intensity (cps)

168
10
10

0
0

-75

-275
125
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
1600
-73
27

-173
127
100
200
300
400
500
600
700
800
900
1000

10
10
[1], d=4.7142(8), 2-theta=18.808(3) [1], d=4.7334(10), 2-theta=18.731(4)
[2], d=4.6635(18), 2-theta=19.014(7)

20
20

20
20
[2], d=4.246(3), 2-theta=20.902(13) [3], d=4.282(7), 2-theta=20.73(3)

30
30

30
30

[3], d=2.4651(3), 2-theta=36.416(5) [4], d=2.4303(7), 2-theta=36.958(11)

[4], d=2.4198(9), 2-theta=37.124(14)
[5], d=2.3645(12), 2-theta=38.02(2)
[5], d=2.3309(17), 2-theta=38.59(3)

40
40

2-theta (deg)
2-theta (deg)

40
40

[6], d=2.0477(4), 2-theta=44.194(9) [6], d=2.0246(5), 2-theta=44.724(12)

[7], d=2.0190(4), 2-theta=44.855(9)

[8], d=1.8806(5), 2-theta=48.359(13)

[9], d=1.8577(8), 2-theta=48.99(2) [7], d=1.852(2), 2-theta=49.14(7)

50
50

50
50

Figure B.6 XRD pattern of LMO for the synthesis at 750 oC

[10], d=1.5771(6), 2-theta=58.47(2)
Figure B.5 XRD pattern of HMO for the synthesis at 750 oC [8], d=1.5625(12), 2-theta=59.07(5)

60
60

60
60

[11],
[12], d=1.4492(5),
d=1.4395(5), 2-theta=64.22(3)
2-theta=64.70(3) [9], d=1.4421(10), 2-theta=64.57(5)
[13], d=1.4196(3), 2-theta=65.720(14) [10], d=1.4258(7), 2-theta=65.40(4)
[14], d=1.3852(7), 2-theta=67.57(4)
[15], d=1.3595(4), 2-theta=69.02(2) [11], d=1.3665(11), 2-theta=68.62(6)

70
70
70
70
 Intensity (cps) Intensity (cps) Intensity (cps) Intensity (cps)

-47
53

-147
100
200
300
400
500
600
700
800

10
10

10
0

-45
55

-245
-145
155
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
1400

10
[1], d=4.677(2), 2-theta=18.959(9) [1], d=4.7190(17), 2-theta=18.789(7)

20
20

20
20
[2], d=4.295(8), 2-theta=20.66(4)
[2], d=4.227(18), 2-theta=21.00(9)

30
30

30
30

[3], d=2.4385(12), 2-theta=36.828(18) [3], d=2.466(2), 2-theta=36.40(3)

[4], d=2.4231(14), 2-theta=37.07(2)
[4], d=2.336(2), 2-theta=38.51(3)
40

40

40
2-theta (deg)

2-theta (deg)
40

[5], d=2.0270(9), 2-theta=44.67(2) [5], d=2.0423(7),

[6], d=2.0205(8), 2-theta=44.317(15)
2-theta=44.820(18)

[6], d=1.8566(13), 2-theta=49.02(4) [7], d=1.8726(7), 2-theta=48.58(2)

50

50

50
50

Figure B.7 XRD pattern of LMO for the synthesis at 650 oC

Figure B.8 XRD pattern of HMO for the synthesis at 650 oC

[8], d=1.5743(7), 2-theta=58.59(3)
[7], d=1.5593(14), 2-theta=59.21(6)
60

60
60
60

[9], d=1.4455(4), 2-theta=64.40(2)

[8], d=1.4313(9), 2-theta=65.12(5) [10], d=1.4209(7), 2-theta=65.65(4)

[9], d=1.375(4), 2-theta=68.1(2) [11], d=1.3831(7), 2-theta=67.69(4)

[12], d=1.3607(7), 2-theta=68.96(4)
70
70

70
70

169
 Intensity (cps) Intensity (cps) Intensity (cps) Intensity (cps)

170
10

10
0

0
-74
-24
26
50
100
150
200
250
300
350
400

-64
-14
36
50
100
150
200
250
300
350
10

10

Li/Mn=1.25)
Li/Mn=1.25)
[1], d=4.681(3), 2-theta=18.944(10) [1], d=4.697(2), 2-theta=18.878(10)

20

20
20

20
[2], d=4.168(17), 2-theta=21.30(9)

30

30
30

30
2-theta (deg)

2-theta (deg)
[2], d=2.435(3),
[3], d=2.43(8), 2-theta=36.88(5)
2-theta=37.0(13) [3], d=2.447(3), 2-theta=36.70(4)

40

40
40

40
[4], d=2.0197(18), 2-theta=44.84(4) [4], d=2.0289(13), 2-theta=44.63(3)

[5], d=1.864(4), 2-theta=48.83(10)

50
50

50
50
[6], d=1.573(4), 2-theta=58.65(16)
60
60

60
60

Figure B.10 XRD pattern of the spinel lithium manganese oxide HMO (450 oC,
Figure B.9 XRD pattern of the spinel lithium manganese oxide LMO (450 oC,
 Intensity (cps) Intensity (cps) Intensity (cps) Intensity (cps)

10
10

0
0

-64
-14
36
50
100
150
200
250
300
350
-67
-17
33
50
100
150
200
250
300
350

10
10

Li/Mn=0.8)
Li/Mn=0.8)
[1], d=4.673(5), 2-theta=18.976(19) [1], d=4.703(4), 2-theta=18.854(16)

20
20

20
20

30
30

30
30

2-theta (deg)
2-theta (deg)

[2], d=2.432(3), 2-theta=36.93(5) [2], d=2.452(3), 2-theta=36.61(5)

40
40

40
40

B.
[3], d=2.0214(11), 2-theta=44.80(2) [3], d=2.0298(17), 2-theta=44.60(4)

[4], d=1.877(2), 2-theta=48.46(6)

[4], d=1.857(5), 2-theta=49.01(14)

50
50

50
50

60
60
60
60

Figure B.11 XRD pattern of the spinel lithium manganese oxide LMO (450 oC,

Figure B.12 XRD pattern of the spinel lithium manganese oxide HMO (450 oC,

171
172
 APPENDIX C

SEM PHOTOGRAPHS OF SYNTHESIZED PSMA-LiMnO BEADS

Figure C.1 SEM photographs for Bead T14

173
Figure C.2 SEM photographs for Bead T15

174
 APPENDIX D

THE FEASIBILITY STUDY TO RECOVER LITHIUM CARBONATE

FROM ÇAMALTI SALINA AND BORON CLAY EXTRACT

The potential sources to produce lithium conventionally in Turkey are Çamaltı

Salina, Salt Lake and boron clay extract as mentioned in this study. The low
concentration of lithium especially in seawater is the major problem with the high
magnesium ion content. However a selective separation process including lithium
manganese oxides can resolve the problems and lead a way to an industrial
production of lithium carbonate.

In this section, the feasibility studies to recover lithium from the potential resources;
Çamaltı Salina and boron clay extract in Bigadiç.

Li2CO3 production from Çamaltı Salina

Lithium carbonate production may be achieved by a system which is drawn in Figure

D. 1.

175
 II II I
I II

C1 C2 C3
HCl III HCl III HCl III
LiMnO

Conc.
Sea 261 ppm
Water II Li
I Mixing &
1.5 ppm settling
II Tap
Li I tank
water II

1675 ppm
IV IV Li
IV pH:9
Alkali-
zation
Pools

NH3 3-6 % Li
Bischofite, lithium carnalite
precipitates
T=95 C
Soda ash Ammonium chloride,
sodium chloride
solution

Lithium
carbonate

Figure D. 1 Li2CO3 production process from concentrated brine in Çamaltı Salina

The aim of the proposed process is to enrich the lithium content in concentrated
brine solution from 1.2-1.5 ppm to 3000 ppm. The adsorption of lithium chloride is
done by feeding the concentrated brine through (Line I) series of columns filled with
PSMA-LiMnO beads (C1, C2 and C3). The effluent of C3 which has trace amount
of lithium is sent back to sea. After that, tap water is fed to the columns for a while
until the outlet concentration of foreign ions remains constant Line 2. The outlet of
tap water may be recycled back to C1, when it has a lithium content higher than 1.5
ppm. After sweeping process is completed, dilute hydrochloric acid solution is fed to
each column batchwise and the lithium content in the column is eluted (Line 3).
After elution process is completed, 261 ppm Li solution is acquired. After that, a
post-treatment process with LiMnO particles is done and the concentration of lithium
is increased to 1675 ppm. The neutralization of hydrochloric acid content in solution
can be done by ammonia. The concentrated brine would then be sent to pools to let
water content evaporate and bischofite salt (MgCl26H20) and lithium carnalite salt
(LiClMgCl27H2O) precipitate in conventional processes. The remaining magnesium
176
content in the brines may be removed by utilizing lime in the processes. The
solutions enriched in lithium content are finally precipitated by adding soda ash
(Na2CO3) in the form of Li2CO3. Before this operation, the lithium content must be
at least saturated to 3000 ppm and the solution should be heated to 95 oC. The high
temperature lowers the carbonate solubility, therefore increases precipitated lithium
carbonate amount. The dependency of lithium carbonate solubility in water towards
temperature is given in Table D. 1.

Table D. 1 Temperature dependency of lithium carbonate solubility [102]

Temperature (oC) Solubility (g/100 ml)

25 1.29
100 0.69

In this study, an experiment was performed in order to produce lithium carbonate

from 3000 ppm LiCl solution. The simple experiment set-up is as shown below.

Figure D. 2 Li2CO3 production from LiCl

By this experiment, some amount of lithium carbonate was produced while using
artificially prepared 3000 ppm LiCl solution. Sodium and potassium carbonates
dissolves in water rapidly but magnesium carbonate solubility in water is 0.0063
g/100 ml [102] which means that, all the magnesium remaining in the solution
precipitates with lithium in the last step. Therefore enriching the brine in lithium
content is as important as decreasing the magnesium content in the brine.

177
In Çamaltı Salina, 500,000 tons of NaCl is produced in a year. That means
14,285,714 tons of sea water is processed in NaCl production. The sea water is
concentrated up to 10 folds which means the total amount of sea water reduces
approximately to 1.5 M tons. After NaCl is produced, the residue is fed back to sea
directly. In order to process that much of water in a year one has to use pumps with a
capacity of 260 kg/min. Totally 13 tons of Li2CO3 can be produced from that amount
of sea water. The price of lithium carbonate is 8540 USD/ton commercially [103]
which means 111,020 USD annual income.

The cost of the system is roughly estimated by utilizing a standard 22” diameter
column with a 4 m height. The volume of each column is calculated as 1 m3. The
density of the beads is known as 1 g/cm3. 1 ton of beads can adsorb 7 kg Li, which
means 30 kg Li2CO3. In order to have 13 tons of Li2CO3 as a product, columns must
be saturated 430 times in a year. In the basis of lab-scale experiments done in this
thesis work, approximately 0.5 L of water is enough to sweep 38.5 g of beads which
scales up to 12.5 m3 of water to sweep 1 ton of beads. 14.35 ml HCl is enough to
elute 38.5 g of beads, which scales up to 372 kg HCl to elute 1 ton of beads. Half of
the acid is required as ammonia to neutralize it.

According to these rough estimations, the expanses were listed in Table D. 2.

Table D. 2 Rough estimation of the cost of 13 tons lithium carbonate production in a

year from Çamaltı Salina

Expenditure Amount Price ($)

Fixed costs
Pumps 3 12,000
Columns, tanks and flow
3 100,000
elements
Total 112,000

Operational costs
Tap water 5,000 tons 25,000
160 ton HCl
HCl, NH3, Na2CO3 50 ton NH3 30,000
18 ton Na2C03

178
Table D.2 continued.

PSMA-LiMnO beads
3 tons 100,000
(d: 1000 kg/m3)
Electrical cost of pumps in one
0.7 kWh(10 TL/hour) 4,500
year
Total 159,500
GRAND TOTAL 271,500

It is obviously seen that, under these circumstances there is no chance to extract

lithium carbonate from Çamaltı Salina in a feasible way due to the very low
concentration of lithium in sea water. However, in terms of energy considerations
and the strategic role that the lithium element will play in the future, the economical
drawback of the process might be reconsidered.

To reduce the magnesium content in the concentrated brine, a pretreatment process

can be integrated to the design which utilizes soda ash before feeding the
concentrated brine into the adsorption column. In this study, preliminary experiments
were done about using soda ash to reduce the magnesium content in the concentrated
brine. The brine and equimolar amount of sodium carbonate with respect to
magnesium was heated to 95 oC and mixed in a beaker. Suddenly a white precipitate
was observed and after 15 min of mixing time, the precipitate was filtered. When
the magnesium content of the initial and final solutions was analyzed in AAS, it was
seen that the magnesium content was reduced to 8614 ppm from 21000 ppm while
maintaining the lithium concentration constant at 1.25 ppm. This additional process
would destress the sweeping duty of the system but would not be sufficient to
overcome the high magnesium concentration problem.

179
Li2CO3 production from boron clay in Bigadiç-Balıkesir

Besides acquiring lithium from sea water, boron clay in Bigadiç seems to be a very
promising candidate to gather lithium with its 3000 ppm Li content. The system will
not require that much of tap water used for sea water and also it won’t need high
performance pumps and operational costs. On the other hand it will need additional
sulfuric acid and ammonia expenses.

The proposed flow diagram to extract lithium carbonate from boron clay is shown in
Figure D. 3.

Ammonia
Tap water

Sulfuric acid

Filter 1 Filter 2

Reactor 1 Reactor 2
400 ppm Li 400 ppm Li
acidic basic

Disposed Clay
from field
Residue Residue

Tap water desalination unit

(cation exchanger)

LiMnO

3200 ppm HCl Lithium

Mixing &
Li chloride
settling 500 ppm Li
stock tank
tank

Pools

Tap water source

3-6 % Li
Bischofite, lithium carnalite
precipitates
Fresh water

Soda Lithium
Reactor 3
ash carbonate

o
95 C

Figure D. 3 Flow diagram of extraction lithium carbonate from boron clay.

180
The aim of the proposed process is to recover lithium from boron clay and make a
production of lithium carbonate. The process starts by the treatment of clay with
sulfuric acid so that, leaching the lithium content and taking into brine. The mixture
is then sent to filter where clay residues are separated. In the second reactor,
ammonia is added to the lithium solution to increase the pH near 9 which is the
operational pH of the beads. Filter 2 after Reactor 2 helps to get rid of white
precipitate formed after ammonia addition. The basic lithium chloride rich solution is
then sent to the adsorption column where adsorption, sweeping and elution processes
are done. Following the post-treatment of brine with LiMnO in settling tanks, the
brine is sent to pools to further concentrate. After that by addition of ammonia and
soda ash the final lithium carbonate product is acquired. When the expenses in the
system are roughly estimated, the following table is acquired.

Table D. 3 Expenses of the lithium carbonate plant from Bigadic boron clay.

Expenditure Amount Properties Price (TL)

Fixed costs

316 Stainless steel

Reactors 4 100,000
4 m3

Columns 2 316 Stainless steel 50,000

316 Stainless steel
Filters 2 250 kg/h filtering 25,000
capacity
316 Stainless steel
Flow elements 1 10,000
316 Stainless steel
Storage tanks 4 20,000
4 m3
Desalination unit 1 Cation exchanger 20,000
Pumps 5 10,000
Total 235,000 TL
Operational costs
1000 kg/ton Needs a production
PSMA-LiMnO beads 30,000
clay/year unit
Commercial
Sulfuric acid 600L/ton clay 120
Ammonia 300L/ton clay Commercial 120
Soda ash 25 kg/ton clay Commercial 20
HCl 300 L/ton clay Commercial 120
Industrial water
Tap water 5000 L/ton clay 10
(or well water)
Heating and electricity 7 kWh/ton clay 10

400 TL/ton clay +

Total
30,000 TL/year

181
 Table D.3 continued.

235,000 TL Fixed
GRAND TOTAL 10,000TL/year Oper.
+ 400 TL/ton clay

28.7 kg lithium carbonate can be produced from 1 ton of clay (3000 ppm lithium
content). Lithium carbonate price is 8540 USD/ton commercially. If 1000 tons of
clay is processed in a year, 430,000 TL is spent to the operational cost and 525,000
TL will be income. Gross profit in a year is around 100,000 TL. 2.5 years will be
enough to pay the expenses for the fixed costs.

Two and a half year time to return investment money will make the process feasible.
Also, the strategic importance of lithium in energy market and expected raises in
lithium carbonate price make the process promising.

182
 CURRICULUM VITAE

PERSONAL INFORMATION

Surname, Name: Erdoğan, Barış

Nationality: Turkish (TC)
Date and Place of Birth: 25 February 1983, Zonguldak
Marital Status: Single
Phone: +90 312 297 61 57 (128)
Fax: +90 312 297 61 57 (128)
email: barise@logoskimya.com

EDUCATION

Degree Institution Year of Graduation

MS METU Chemical Engineering 2008
BS METU Chemical Engineering 2006
High School KDZ Ereğli Anadolu High School, Zonguldak 2001

WORK EXPERIENCE
Year Place Enrollment
2012- Present LOGOS Chemical Technologies Co-Founder
2008-2012 METU Department of Chemical Engineering Research Assistant
2005 July İstanbul Elyaf Intern Engineering Student
2004 August ERDEMİR Intern Engineering Student

FOREIGN LANGUAGES
Advanced English

PUBLICATIONS
Sinan Korpe, Baris Erdogan, Goknur Bayram, Serkan Özgen, Yusuf Uludag, Niyazi
Bıçak, Crosslinked DADMAC polymers as cationic superabsorbants, Reactive
Functional Polymers, 69, 9, September 2009, sayfa 660-665

183
PUBLICATIONS IN REVIEW
Baris Erdogan, H. Onder Ozbelge, Niyazi Bicak, Direct synthesis of N-alkyl
formamides and their utility in lithium extraction from artificial brines and
concentrated seawater, Journal of Environmental Chemical Engineering

ABSTRACTS IN INTERNATIONAL CONFERENCES

Baris Erdogan, H. Onder Ozbelge, Niyazi Bıçak, Extraction of Lithium from
Seawater by N-Alkyl Formamides, 12. Mediterranean Congress of Chemical
Engineering, November 2011; Barcelona, Spain

Baris Erdogan, H. Onder Ozbelge, Niyazi Bıçak, Separation of Lithium by Polymer

Enhanced Ultrafiltration, 12. Mediterranean Congress of Chemical Engineering,
November 2011; Barcelona, Spain

HOBBIES
Mountaineering, Microcontrollers

184