An Investigation into Aqueous Titanium
Speciation Utilising Electrochemical Methods for
the Purpose of Implementation into the Sulfate
Process for Titanium Dioxide Manufacture
Samala Shepherd, BSc. (Hons)
Masters of Philosophy in Chemistry
University of Newcastle
March, 2013
STATEMENT OF ORIGINALITY
This thesis contains no material which has been accepted for the award of any other degree or diploma in any university or other tertiary institution and, to the best of my knowledge and belief, contains no material previously published or written by another person, except where due reference has been made in the text. I give consent to this copy of my thesis, when deposited in the University Library**, being made available for loan and photocopying subject to the provisions of the Copyright Act 1968.
**Unless an Embargo has been approved for a determined period.
Samala L. Shepherd
i Acknowledgements
There are many people who have helped me and contributed to my work in a number of ways and I’d like to thank them. I’d like to thank BHP Billiton Newcastle Technology Centre for making this project possible. The ARC for support and funding. Dr. Scott Donne for the vast knowledge he provided me with and the friendship and support. Thank you to Carolyn Freeburn, Vicki Thompson, and Stephen Hopkins for the ‘store/equipment room’ when I was in need. Thank you to Dianna Brennan for keeping me supplied. Michael Fitzgerald for his continued support. Last but not least a big thanks to my family, they are stuck with me but bare the burden with smiles and support and I thank them greatly.
ii Declaration
To the World of Chemistry. May you never stop amazing people!
iii Table of Contents
STATEMENT OF ORIGINALITY ...... i
Acknowledgements ...... ii
Declaration ...... iii
Table of Figures ...... 5
Abstract ...... 9
1 Introduction to Titanium and its Electrochemistry ...... 10
1.1. The Sulfate Process ...... 10
1.2. Aqueous Titanium Chemistry ...... 15
1.2.1. Introduction ...... 15
1.2.2. Aqueous Titanium(II) Chemistry ...... 17
1.2.3. Aqueous Titanium(III) Chemistry ...... 18
1.2.4. Aqueous Titanium(IV) Chemistry ...... 19
1.2.4.1. Monomeric Species of Titanium(IV) ...... 20
1.2.4.2. Oligomeric species of Titanium(IV) ...... 23
1.2.5. Mixed Oxidation State, Ti(III) and Ti(IV) Species ...... 25
1.3. Aqueous Titanium(IV) Electrochemistry ...... 27
1.3.1. Thermodynamics ...... 27
1.3.2. Redox Mechanisms ...... 28
1.3.3. The CE or ECE Mechanism ...... 29
1 1.3.3.1. Low Acid Concentration Mechanisms ...... 29
1.3.3.2. High Acid Concentration Mechanisms ...... 31
1.4. Electroactivity of Various Ti(IV) Species ...... 32
1.5. Polarography of Titanium(IV) ...... 35
1.6. Constant Current Electrolysis of Titanium containing solutions .. 37
1.7. Hydrogen Evolution Reaction (HER) ...... 38
1.7.1. Proton Diffusion ...... 38
1.7.2. Kinetics ...... 39
1.7.3. Cathodic Substrate Effects ...... 40
1.8. Leachate Impurities of Ilmenite ...... 42
1.9. Design ...... 42
1.10. References ...... 43
2 Experimental ...... 46
2.1. Solution Preparation ...... 46
2.2. Cyclic Voltammetry Measurements ...... 47
2.3. Impedance Spectroscopy (EIS) Measurements ...... 49
2.4. Rotating Disc Electrode Voltammetry ...... 51
2.5. Electrochemical Protocol for Polarography ...... 52
2.6. References ...... 54
3 Electrochemistry Results ...... 55
2 3.1. Cyclic Voltammetry ...... 55
3.1.1. Cathodic Substrate ...... 55
3.1.2. Titanium Concentration Effects ...... 55
3.1.3. Sulfuric Acid Concentration Effects ...... 61
3.1.4. Kinetic Information ...... 65
3.1.5. Solution Age Effects ...... 69
3.2. Electrochemical Impedance Spectroscopy (EIS) ...... 72
3.3. Rotating Disc Electrode (RDE) Voltammetry ...... 82
3.4. Conclusions ...... 88
3.5. References ...... 89
4 Polarography ...... 90
4.1. Polarographic Behaviour...... 90
4.2. Effect of H2SO4 Concentration ...... 97
4.3. Effect of Electrolyte Age ...... 99
4.4. Effect of Different Acid Media ...... 103
4.4.1. Backgrounds ...... 103
4.4.2. Different Supporting Electrolytes ...... 103
4.4.3. Titanium Metal ...... 105
4.4.4. Titanyl oxysulfate Addition to Various Acid Media ...... 107
4.5. Conclusions ...... 110
3 4.6. References ...... 110
5 Sulfate Process Implementation ...... 112
5.1. Overview of Electrochemical Techniques ...... 112
5.2. Implementation ...... 112
5.3. Industrial Design ...... 115
5.4. References ...... 115
6 Appendix A ...... 116
7 Appendix B ...... 118
8 Appendix C ...... 131
4 Table of Figures
Figure 1-5. Curves of Ti(IV) in SWV, with forward scan ∆I < 0 and reverse scan ∆I > 0. The numbers on the curves are the concentrations of H2SO4 (M) [28]...... 31
Figure 1-6. Curves of Ti(IV) in SWV, with forward scan ∆I < 0 and reverse scan ∆I > 0. The
numbers on the curves are the concentrations of H2SO4 (M) [28]...... 34
-6 2 -1 Figure 1-7. Diffusion coefficients, D, of Ti(IV) in H2SO4 at 25°C. D0 = 4.2 x 10 cm s [28].... 36
Figure 1-8. The exchange current densities for hydrogen evolution on several cathodic substrates
[41]. Data points are the negative logarithm of the exchange current densities in Am-2...... 41
Figure 2-1. An example of three full cycles from cyclic voltammetry of a solution containing 0.0
2+ M TiO in 0.1 M H2SO4, at 21 ± 1°C...... 48
Figure 2-5. An example of a Lorentzian lineshape graph...... 54
Figure 3-1. Cathodic sweep only from the cyclic voltammogram, of solutions containing different
2+ TiO concentrations in 0.1 M H2SO4 at a scan rate of 5 mV/s using a Pb cathode, at 21 ± 1°C.
The numbers in bold brackets, (1) – (4) are explained in the text...... 56
Figure 3-2. The maximum peak currents corresponding to peaks (1) – (4) from Figure 3-1 of the
2+ solutions containing varying TiO concentrations in 0.1 M H2SO4...... 59
Figure 3-12. As the H2SO4 concentration was increased, the maximum RCT values decreased which is an indication that the reaction is becoming more facile. The maximum RCT values for 1
M and 5 M H2SO4 are roughly the same at 520 Ω and 530 Ω respectively. This indicates that the kinetics of the reaction are the same for the different solutions. However, these maximum RCT values occur at different potential which suggests different species...... 77
Figure 3-12. Charge transfer resistance (RCT) versus voltage for solutions containing varying
2+ [H2SO4] in 0.06 M TiO from EIS measurements at 21 ± 1°C...... 78
5 2+ Figure 3-13. Series resistance (RS) versus voltage for solutions with varying [TiO ] in 1 M
H2SO4 from EIS measurements at 21 ± 1°C...... 79
2+ Figure 3-14. Double layer capacitance (Cdl) versus voltage for solutions with varying [TiO ] in 1
M H2SO4 calculated from EIS measurements, at 21 ± 1°C...... 81
Figure 3-15. Cathodic sweep a of rotating disc electrode (RDE) scan at 5 mV/s with increasing
2+ rotation rate from 200-3000rpm on a solution containing 0.13 M TiO and 5 M H2SO4, at 21 ±
1°C...... 83
Figure 7-1. Cathodic sweep only from cyclic voltammetry of a solution containing varying
2+ [TiO ] and 0.01 M H2SO4, with a scan rate of 5 mV/s at 21 ± 1°C...... 119
Figure 7-2 Cathodic sweep only from cyclic voltammetry of a solution containing varying [TiO2+] and 1 M H2SO4, with a scan rate of 5 mV/s at 21 ± 1°C...... 120
Figure 7-3 Cathodic sweep only from cyclic voltammetry of a solution containing varying [TiO2+] and 5 M H2SO4, with a scan rate of 5 mV/s at 21 ± 1°C.
6 1800 0.0 M TiO2+ 1600
1400
1200 0.03 M TiO2+
) 1000 Ω / ( CT
R 800 0.06 M TiO2+
600
0.09 M TiO2+ 0.13 M TiO2+ 400
200
0 -1.3 -1.2 -1.1 -1 -0.9 -0.8 -0.7 -0.6 -0.5 Potential / (V vs SCE)
...... 121
Figure 7-4 Charge transfer resistance (RCT) versus voltage for solutions containing varying
2+ [TiO ] in 0.1 M H2SO4, from EIS measurements at 21 ± 1°C...... 122
Figure 7-5 Charge transfer resistance (RCT) versus voltage for solutions containing varying
2+ [TiO ] in 0.01 M H2SO4, from EIS measurements at 21 ± 1°C...... 123
Figure 7-6 Charge transfer resistance (RCT) versus voltage for solutions containing varying
2+ [TiO ] in 5 M H2SO4, from EIS measurements at 21 ± 1°C...... 124
2+ Figure 7-7 Series resistance (RS) versus voltage for solutions with varying [TiO ] in 0.1 M
H2SO4, from EIS measurements at 21 ± 1°C...... 125
7 2+ Figure 7-8 Series resistance (RS) versus voltage for solutions with varying [TiO ] in 0.01 M
H2SO4, from EIS measurements at 21 ± 1°C...... 126
2+ Figure 7-9 Series resistance (RS) versus voltage for solutions with varying [TiO ] in 5 M H2SO4,
from EIS measurements at 21 ± 1°C...... 127
2+ Figure 7-10 Double layer capacitance (Cdl) versus voltage for solutions with varying [TiO ] in
0.1 M H2SO4 calculated from EIS measurements, at 21 ± 1°C...... 128
2+ Figure 7-11 Double layer capacitance (Cdl) versus voltage for solutions with varying [TiO ] in
0.01 M H2SO4 calculated from EIS measurements, at 21 ± 1°C...... 129
2+ Figure 7-12 Double layer capacitance (Cdl) versus voltage for solutions with varying [TiO ] in 5
M H2SO4, calculated from EIS measurements, at 21 ± 1°C. A log scale has been used to see the relationship clearer...... 130
8 Abstract
Titanium dioxide, TiO2, plays an important role in many areas of modern society due to its
very high refractive index. In its powder form, it is most commonly found in paints, plastics,
toothpaste, foods and medicines. Industrially, titanium dioxide has been mixed with glass or
cement to produce superior materials.
TiO2 pigment is produced by two common methods, the Sulfate and Chloride Processes. This thesis will focus on the Sulfate process. During the sulfate process, the titanium ends up being in
aqueous form in concentration acid, and little is known about the speciation of aqueous titanium.
Therefore it is the focus of this thesis to investigate the electrochemistry of Ti(IV) reduction to
Ti(III) in sulfuric acid media, using known electrochemical techniques, such as; cyclic
voltammetry, impedance, RDE, polarography, to determine kinetic information, i.e. diffusion co-
efficients, to better understand how titanium reacts in a sulfuric acid medium.
The experimental results revealed that in sulfuric acid media, compared to other acid
media, the sulfate ions form three different bridging species either as oligomers or monomers.
These species depend on various solution conditions such as titanium and acid concentration as
well as solution age. Indications of the best possible conditions have been provided.
9 1 Introduction to Titanium and its
Electrochemistry
Titanium is relatively common, being the ninth most abundant element in the Earth’s crust [1]. However, due to restrictions and difficulties in the refinement process [1], titanium only
begun being manufactured industrially in the last century. The most important industrial use of
titanium is in the form of titanium dioxide, TiO2. This naturally occurring dioxide is found to
have a crystal structure of rutile, anatase or brookite. These structures have been classified as tetragonal for rutile and anatase and orthorhombic for brookite. Each titanium atom is coordinated to six equidistant oxygen atoms, and each oxygen atom to three titanium atoms [2]. Titanium dioxide powder is used in paints, plastics, toothpaste and paper [1]. TiO2 pigment is produced by
two common methods, the Sulfate and Chloride Processes. Both these processes use ilmenite as
their starting material [2]. Ilmenite is a black ore containing various amounts of TiO2 in the form
of FeO.TiO2.
1.1. The Sulfate Process
Chemically, the Sulfate Process is straightforward but the industrial process is more
complex, as shown schematically in Figure 1-1. It begins with the digestion of ground ilmenite
ore (FeO.TiO2) in concentrated sulfuric acid (85-90%) [2] i.e.,
FeTiO3 + 2H2SO4 → TiOSO4 + FeSO4 + 2H2O (1.1)
10 At 100°C, the mixture is converted to a porous cake, containing Fe(II), Fe(III) and Ti(IV) sulfates.
11 Scrap Fe
Ilmenite Digestion Porous Purification FeSO4⋅7H2O (FeTiO3) ° Cake and Filtration (Copperas) (~100 C)
H2SO4 H2SO4 Black (85-92%) Regeneration Liquor
SO Hydrolysis - 2 OH (~100°C/hours)
Post Calcination Production TiO2 TiO ⋅nH O (~1000°C) 2 2 (e.g., coating)
Scrubbers SO3 H2O Figure 1-1. Schematic of The Sulfate Process.
12 This cake is then redissolved in dilute H2SO4 to form a black-liquor (impure), followed by the crystallisation of FeSO4.7H2O (copperas) upon cooling, which also has the effect of removing any un-reacted ore. Hydrolysis of this cooled liquor with the aid of a TiO2 seed of either rutile or anatase structure leads to the precipitation of hydrous TiO2 [3, 4], i.e.,
OH - TiOSO4 + (n+1)H2O → TiO2.nH2O + H2SO4 (1.2)
Also at this stage the iron (in the form of copperas) needs to be removed. It is easier to remove iron from the solid in the ferrous state, so scrap iron is added (Eqn. 1.4), followed by a washing stage. The hydrous pigment is then calcined at 1000°C to bring about thermal decomposition
(Eqn. 1.3); i.e.,
Heat TiO2.nH2O → TiO2 + nH2O (1.3)
To finish off, the TiO2 pigment is often coated with an inorganic coating, such as a silicate or
ZrO2, which may both inhibit the gradual yellowing observed in some paints and improve the dispersibility of the pigment in certain media [2].
The main disadvantages of the Sulfate Process on an industrial scale are that it does have a high risk factor with the use of strong, concentrated acid, and certain steps carry a high operating cost; i.e., the precipitation of copperas as a waste product due to the addition of extra scrap iron. These issues are however minimal compared to the Chloride Process which requires large quantities of chlorine to be used, which has to be removed from the pigment at the end of production. BHP-Billiton have developed a new metallurgical route to producing TiO2 pigment that is more economical and environmentally friendly than the traditional Sulfate route. In the
BHP-Billiton process, the ilmenite is leached with the aid of a reducing agent (scrap iron or
13 Ti(III)) in much more dilute sulfuric acid (50%), and the acid is recycled later in the process. In the traditional Sulfate Process, the acid is waste, and must be treated and disposed, which has its own economical and environmental burdens.
An important point to note is that the production of titanium dioxide should lead to a pure white material. However, the presence of Fe(III) in the pigment tints the product brown. Because iron is present in ilmenite, rectification and purification steps in both the Chloride and Sulfate
Processes, need to be taken to remove it. During the hydrolysis stage, Fe(III) is also hydrolysed but not Fe(II), which remains soluble. Therefore, if all the iron present is in the Fe(II) oxidation state, it will crystallise and then can easily be washed out of the titanium dioxide. To achieve this, scrap iron or zinc is added so that it reduces Ti(IV) to Ti(III), and that in turn reduces Fe(III) to
Fe(II); i.e.,
2+ + 2+ 3+ 2TiO + 4H + Fe → Fe + 2Ti + 2H2O (1.4)
or
2+ + 2+ 3+ 2TiO + 4H + Zn → Zn + 2Ti + 2H2O (1.5)
then
3+ 3+ 2+ 2+ + Ti + Fe + H2O → TiO + Fe + 2H (1.6)
These extra additions of Zn or Fe are inefficient and uneconomical. Iron usage leads to more Fe(II) being produced which must be washed out of the hydrous pigment. An alternative is the electrogeneration of Ti(III) as it would decrease the costs involved with the addition of scrap iron as a reductant. This in turn would decrease the production of FeSO4.7H2O and the need to dispose of this waste. So the electrogeneration of Ti(III) offers an attractive alternative both economically and technically.
14 1.2. Aqueous Titanium Chemistry
1.2.1. Introduction
Titanium, in aqueous form, exists in the oxidation states of Ti(III) and Ti(IV). The evidence for titanium existing as Ti(II) is contradictory, but it is believed that it is not stable, due to complexes of Ti(II) being extremely reducing and requiring careful handling [1]. Figure 1-2 contains an Eh-pH diagram for titanium species at 298 K [5]. This figure was constructed using
22 relative stability equations. The dashed lines indicate the domains of the dominant aqueous titanium species with the following:
2+ − 1` TiO / HTiO3 pH = 5.64 (1.7)
2+ 3+ 2` Ti / Ti E0 = -0.368 (1.8)
2+ 2+ 3` Ti / TiO E0 = -0.135 – 0.0591 pH (1.9)
2+ − 4` Ti / HTiO3 E0 = 0.362 – 0.1475 pH (1.10)
3+ 2+ 5` Ti / TiO E0 = 0.100 – 0.1182 pH (1.11)
2+ 2+ 6` TiO / TiO2 E0 = 1.800 – 0.0591 pH (1.12)
- 2+ 7` HTiO2 / TiO2 E0 = 1.303 + 0.0295 pH (1.13)
The bold lines indicate the boundaries between two solid species (8, 9 and 11 [Appendix A]) or one solid species and one dissolved species (13, 17, 18, 19 and 20 [Appendix A]). The two letters
‘a’ and ‘b’, represent hydrogen evolution and oxygen evolution, respectively. The use of question marks in the figure represents the regions where the equilibrium between titanium and water is not exact as the data collected for the free energies is questionable.
15
Figure 1-2. Potential-pH equilibrium diagram for the titanium-water system, at
25°C [5].
16
The areas of interest for this research are the dashed lines between Ti2+ and Ti3+ and the
2+ - 2+ dashed lines between the species TiO , HTiO3 and TiO2 , as indicated by Eqns. 1.7 – 1.13.
1.2.2. Aqueous Titanium(II) Chemistry
There is reason to believe that Ti2+ exists as it is included in graphs and tables throughout the electrochemical literature, as in Figure 1-2 [5]. However, it’s reduction potential is given (Eqn
1.14) as the Ti(II)/Ti(III) couple [6]; i.e.,
Ti 3+ + e − ↔ Ti 2+ E0 = −0.23V (1.14)
Given that a reduction potential has been reported it is unusual to find no reference to the Ti(II) oxidation state at all. One reason for this is that the preparation of a stable Ti(II) solution was impossible, and thus it was concluded that titanium(II) rapidly disproportionates to Ti(III) and
Ti(0) [1]. Therefore, Ti(II) in aqueous solutions has barely been investigated [1].
This uncertainty lead a research group, Kölle et al., in 2003 to investigate further the occurrence of Ti(II) [7]. They discovered a green solution in HF media which they identified as
Ti(II), which when exposed to air turned brownish red, then colourless. This finding supports the colour of the solutions already observed in previous experiments. They concluded that Ti(II) can
2+ 2+ exist, most likely as the Ti hexaaqua ion ([Ti(H2O)6] ), which is typical for transition metals.
Compounds containing Ti(II) are known to exist, but as soon as they are dissolved in water, they are no longer stable due to the oxidation by water [8]. However, in chemistry text books [8], it is written that solutions of TiO in dilute HCl at 0°C have been known to contain
Ti(II) ions which persist for some time [8].
17 From literature results there is some evidence for the existence of Ti(II) species.
However, since these are low potential species it is not likely that we will encounter them in this work, since the focus is on the reduction of Ti(IV) to Ti(III), and this potential range occurs at a higher potential.
1.2.3. Aqueous Titanium(III) Chemistry
The hexaaqua Ti(III) ion is a bright violet species and is used to good effect as a reducing agent, and hence needs to be handled and stored under nitrogen and/or hydrogen [8]. The potential at which Ti(III) occurs has been calculated using Gibbs free energy values [6] (Eqn
(1.15)) and estimated (Eqn (1.16)) [8], i.e.,
4+ − 3+ Ti + e ↔ Ti E0 = 0.2 V (1.15)
2+ + - 3+ Ti(OH)2 + 2H + e ↔ Ti + 2H2O E0 = 0.008 V (1.16)
Eqn (1.16) is just one example of a redox reaction involving Ti3+, and depends upon hydrogen ions (H+) being available in solution. Other examples include the aerial oxidation of the Ti2+ hexaaqua ion mentioned above.
Ti(III) has also been used as a selective reductant in organic syntheses by Foller et al. [3].
The reducing quality of the Ti(III) ion is especially important in the Sulfate Process as mention above, as it reduces Fe(III) to Fe(II) via Eqn (1.5).
Ti(III) is generally present in dilute acid solutions (< 2 M) as the hexa-aqua ion,
3+ [Ti(H2O)6] . Related works indicate that one of the coordinated water molecules may undergo hydrolysis [1]; i.e.,
3+ 2+ + -4 [Ti(H2O)6] ↔ [Ti(H2O)5(OH)] + H K=1.3x10 (1.16)
18
2- However, in the presence of sulfuric acid it has been reported that the SO4 acts as a ligand and a series of Ti(III)-sulfate complexes may be formed [9]; i.e.,
3+ 2− + Ti + SO 4 ↔ TiSO 4 β1 = 24.3 ± 3.8 (1.17)
3+ 2− − Ti + 2SO 4 ↔ Ti(SO 4 )2 β2 = 180 ± 11 (1.18)
3+ 2− 3− Ti + 3SO 4 ↔ Ti(SO 4 )3 β3 = 376 ± 83 (1.19)
where β is the cumulative stability constant.
The halides of Ti(III) are well defined and exist as complexes with different ligands of
3+ - the type [ML6] X3 where M is a metal, L is an unidentate ligand, and X is a halide. These, however, are all highly unstable with respect to oxidation, and in the presence of moisture nearly all hydrolyse instantly [1].
1.2.4. Aqueous Titanium(IV) Chemistry
Ti(IV) is the most stable of the oxidation states, but it too has been hard to isolate in different media as a lone Ti(IV) ion. It is possible that Ti(IV) may exist as a hydrolysed species
2+ 2+ 2+ for which some examples have been proposed, including Ti(OH)2 , Ti(OH)Cl or Ti2(OH)6 , depending on the media used [8]. There is no definite or conclusive proof of these species, however, within the literature. In strong acid media there exists an equilibrium between various hydrolysed Ti(IV) species [8], e.g.,
2+ + 2+ 4+ 2TiO + 2H ↔ Ti(OH)2 + Ti (aq) (1.20)
19 2+ 2+ where TiO is the titanyl ion, and Ti(OH)2 is a partially hydrolysed Ti(IV) species. This equilibrium of course relies on the availability of a titanyl unit, which was once thought not to exist. For its neighbouring element, vanadium, there are well known “true” yl-type bonds that exist, where the vanadium and oxygen are double bonded; i.e. V=O2+. It is therefore reasonable to propose that these yl-type bonds also exist for titanium, even tough it was generally accepted that true titanyl cores were very scarce [10, 11]. In older literature by Comba et al., [12] it is described as either “non-existent”, or at least “still unproven”, but more recent investigations have shown that yl-type bonds do exist with various monomers, sulfate complexes, and oligomers in H2SO4 solutions having also been investigated [8, 12].
1.2.4.1. Monomeric Species of Titanium(IV)
The stoichiometry of the species in sulfuric acid has been examined and reported in the literature by various authors [1, 12, 13], where it was concluded that all monomeric species present had a +2 charge. This was confirmed by Raman spectroscopy measurements on solutions of TiO2+. As there exists a true vanadyl (V=O) moiety, Gratzel et al., assumed there should also be a true titanyl (Ti=O) moiety [13]. These authors compared the vibrational frequencies for
v~ (V=O) which were observed in the region 940 - 1000 cm-1, to that of a titanyl solution which produced a weak shoulder at v~ = 975 cm-1. The possibility of this shoulder occurring from a (M-
OH) vibration was ruled out as these vibrations are expected to occur around 500 cm-1.
Following this revelation, further investigations into the equilibrium between dihydroxo
2+ Ti(IV) species ([Ti(OH)2] (aq)), protonated monomers and oligomeric species in dilute acidic solutions was undertaken by Comba, et. al., [12]. One such equilibrium that was proposed is illustrated in Figure 1.3.
20 2+ [TiO(H2O)5 ]
+ M + M + M 4+ −H (Ka1 ) 3+ −H (Ka2 ) −H (Ka3 ) + [Ti(H2O)6 ] ←→[Ti(H2O)5 OH] ←→ ←→[Ti(H2O)3 (OH)3 ]
2+ [Ti(H2O)4 (OH)2 ]
Figure 1-3. Possible hydrolysis scheme for the equilibrium of Ti(IV) species in dilute
M acid solutions [12]. Estimates for the acidity constants were given: Ka3 = 2.0;
M M M M Ka2 Ka1 > 25 (Ka2 < Ka1 ).
M M In the scheme proposed above, the authors have stated that Ka2 < Ka1 , implying that the second protonation of titanyl is easier than the first. The authors also observed that there was a proton dependence of the oxygen-exchange rate of the titanyl; i.e., the exchange rate increases with increasing [H+]. The authors were also surprised by the lability of the -yl oxygen in Ti=O when compared to the -yl oxygen in V=O. They found that the titanyl exchanges its -yl oxygen about nine orders of magnitude faster, and that the exchange mechanism must be different for titanium when compared to vanadium. Since the second protonation is easier than the first, it implies that the -yl bond stabilises the titanyl ion and thus the equilibria is expected to lie to the
2+ side of the titanyl ion (Figure 1.3). Therefore the concentration of Ti(OH)2 is unlikely to be large, and is one reason why direct experimental evidence on these species remains to be found
[12].
As mentioned above, the vanadyl ion exchange mechanism must be different to the titanyl ion mechanism. The vanadyl ion is affected by the presence of OH- ions in the coordination sphere [14], where a hydroxyl group in an equatorial position converts to a -yl oxygen by proton transfer from the original -yl oxygen. This is an example of an internal electronic rearrangement (IER) mechanism [14]. These authors, Johnson et al., also suggest that this IER mechanism could be a general feature of other -yl oxygen-solvent exchanges. As
21 mentioned the titanyl ion behaves differently. Analysing the Ballhausen and Gray molecular orbital (MO) scheme for vanadyl [15], and adjusting to suit titanyl, they showed that the higher energy of the 3d orbitals leads to σ and π molecular orbitals that have more oxygen character than those of their vanadyl counterparts [12]. Thus the -yl oxygen in titanyl has a greater electron density than in vanadyl, and is therefore more easily protonatable. In fact, vanadyl is practically not protonatable in aqueous solutions [16]. The apparent fact that the titanyl ion is readily protonatable is also responsible for the rapid exchange rate as shown by the [H+] proton dependence on the oxygen exchange rate [12].
The existence of the titanyl ion in acidic solutions was also studied by measuring 17O
NMR chemical shifts [12]. Compared to other oxo species arising from other metal ions such as
Mo(V), Mo(IV) and V(V), the 17O NMR chemical shifts are expected to occur in the regions
1000-1100, 700-800 and 0-100 ppm. However, no 17O NMR signals were able to be measured at a temperature of 263 K as the -yl type oxygen exchanges rapidly and causes exchange broadening. The temperature was dropped to 195.85 K and a peak was observed at 1028 ppm.
This is a chemical shift region where -yl type oxygens are expected to occur. Furthermore, under these experimental conditions, the authors observed that TiO2+ undergoes protonation equilibria, as the signal disappears with increasing pH.
Overall, it seems that the literature is in agreement that a possible intramolecular
2+ protonation species of Ti(OH)2 is the responsible reactive intermediate in the rapid -yl oxygen
3+ exchange [12], but it only occurs at low concentrations. Other species, such as [Ti(OH2)5(OH)]
4+ or Ti(OH2)6 may occur at different concentrations [12].
Older literature by Babko et al., stated that, as with the Ti(III) case in sulfuric acids solutions, free sulfate is capable of complexing the various Ti(IV) species. These reactions for the complexation of the titanyl ion, along with the associated equilibrium constants are given in Eqns
(1.21 - 23) [17]; i.e.,
22 2+ 2− TiO + SO 4 ↔ TiOSO 4 K1 = 169.5 (1.21)
2− 2− TiOSO 4 + SO 4 ↔ TiO(SO 4 )2 K2 = 76.92 (1.22)
2− 2− + 2− TiO(SO 4 )2 + SO 4 + 2H ↔ Ti(SO 4 )3 + H 2O K3 = 1.35 (1.23)
2+ Note that in Eqn (1.23), the -yl oxygen in the TiO core has been eliminated as a water molecule.
1.2.4.2. Oligomeric species of Titanium(IV)
The formation of oligomers in acidic solutions containing Ti(IV) has not been widely studied. However, it is known that Zr(IV) and Hf(IV) both have a strong tendency to form oligomeric species in acidic solutions [18, 19]. Also, oligomerisation occurs for many other
5+ transition metal ions upon hydrolysis; e.g., Cr(III) as Cr4O(OH)5 [20], Mo(II–VI) [21] and
Fe(III) [22], so it is expected and not unreasonable for oligomerisation to occur for acidic Ti(IV) solutions.
As mentioned in the monomeric species section above, the titanyl ion has a more labile oxo-oxygen, and is readily protonatable. The protonation equilibria that is responsible for the rapid -yl oxygen exchange (Figure 1-3) is believed to be also responsible for the formation of a series of oligomeric species of Ti(IV) in acidic media [12]. It was also observed by these authors that there was no oligomeric species containing any trace of a titanyl fragment, which implies that the monomeric titanyl is unlikely to be the building block for the oligomeric framework [12]. The
17O NMR investigations into the speciation of Ti(IV), which provided evidence of the titanyl ion and monomeric species, also discovered certain oligomeric species [12]. These oligomers exist in appreciable concentrations where the Ti(IV) concentration is between 0.1 – 0.5 M, the H+ concentration between 1.0 - 2.5 M, and the temperature between 236 - 323 K. The best fit to their
23 4+ α β 5+ data is a model comprising of two trimers and one tetramer; i.e., Ti3O4 , Ti3O O 2H3 , and
α β 8+ Ti4O 4O 2H4 , respectively, the formation of which are given below; i.e.,
2+ 4+ + 3 3TiO + H2O ↔ Ti3O4 + 2H K3 = 0.38 ± 0.06 (1.24)
2+ α β 5+ + 2 -1 3TiO + 2H2O ↔ Ti3O3 O2 H3 + H K3 = 1.64 ± 0.06 m (1.25)
2+ α β 8+ 1 -3 4TiO + 2H2O ↔ Ti4O4 O2 H4 K4 = 2.31 ± 0.03 m (1.26)
where α and β represent the charge associated with each oxygen (at this stage they have not been assigned values).
Note that these equilibrium constants were determined by Comba et al., in noncomplexing perchlorate media, and some differences in sulfuric acid media are expected due
- 2- to the ability of HSO4 and SO4 to complex the monomeric and oligomeric species. An activity- pH diagram for these oligomeric species has been calculated, as shown in Figure 1-4 [23]. This diagram was produced by using the standard Gibbs energies of the species, where sometimes it
α β 8+ was necessary for the values to be estimated. From this plot we can see that the Ti4O 4O 2H4 ions predominate over the TiO2+ ions for pH < -0.53. This needs to be noted, as such low pHs have been used in the study of the electrode kinetics of the Ti(IV)/Ti(III) redox couple [24]. Up to five oligomers were discovered in the 17O NMR study; however, due to limitations in the nuclearity of the oligomers, the others were determined to be negligible.
24
Figure 1-4. Activity-pH for aqueous oligomeric species of Ti(IV) in equilibrium with
TiO2 [23].
1.2.5. Mixed Oxidation State, Ti(III) and Ti(IV) Species
When Ti(III), a bright violet species, and Ti(IV), a colourless species, are mixed together in a sulfuric acid media, the resulting solution is brown-violet in colour [25]. That is to say, the addition of the two species does not exhibit a straightforward additive effect. Therefore, an
25 association between Ti(IV) and Ti(III) was assumed to be responsible for the colour change.
However, in chloride media, this additive effect was not observed using UV-visible absorption spectroscopy [9]. Recent studies by Cservenyák et al. suggest the formation of the Ti(III)-Ti(IV)- sulfate species could occur by reactions such as that shown in Eqn (1.27) [9]; i.e.,
(2q−3−p)− O + 3+ + − + + − 2− ↔ IV III + (1.27) TiOSO4 Ti (p 2)H (q 1)SO4 Ti Ti (SO4 )q (H )p O
From these spectrophotometric studies we see that there is an increase in absorptivity with increasing acid concentration, indicating that this process is an equilibrium process. Previous spectrophotometric studies into Ti(III) in sulfate solutions may now be in doubt due to the presence of a mixed oxidation state species. Due to the high sensitivity to oxidation of Ti(III) in solution, any commercial solutions have most likely been partially oxidised by dissolved oxygen and will hence shift the wavelength maximum. It was observed that an increase in the absorbance maximum occurred by an order of five times [9] initially, and thus interfered with the Ti(III)-
Ti(IV)-sulfate complex. This increase in the absorbance maximum is likely due to an intramolecular charge transfer process between the Ti(III) and Ti(IV) centres [9].
The non-straightforward additive effect of Ti(III) to Ti(IV) was investigated by utilising
UV-visible spectroscopy [9]. The absorbance maximum for a Ti(III) solution was measured to be
520 nm, and this shifted to 470 nm when an equimolar ratio of 1:1 Ti(III) - Ti(IV) solution was measured [9]. This shift in absorbance maxima lead to further spectrophotometric studies, as well as possible concentration effects; i.e., having an excess of Ti(IV) in solution. With an equimolar ratio the species exhibited a deviation from Beer-Lambert behaviour; i.e., the absorbance was not a linear function of concentration [9], which again reiterates a non additive effect. However, with
26 excess Ti(IV), the absorbance increased significantly which indicates the presence of an equilibrium process (Eqn 1.27) [9]. This therefore makes it more difficult to determine reliable values for the molar absorptivity. It was also determined that the mixed Ti(III) – Ti(IV) species is
- dependant on the acid concentration, as HSO4 ions are more predominant in concentrated sulfuric acid solutions [9]. Consequently, this has hindered the research as absorptivity values need to be taken into consideration. When the absorbance was measured after addition of Na2SO4 it was noted that only a slight decrease in absorbance was observed, whereas with the addition of
H2SO4, the absorbance increased significantly [9]. Therefore, for this mixed oxidation species, the strength of the acid is going to have more of an effect on the outcome than the complexing effect from the sulfate ions.
1.3. Aqueous Titanium(IV) Electrochemistry
Polarographic methods have been utilised to study Ti(IV), especially as mercury is a good electrode to use since its affinity for hydrogen is low. The study of the electrochemistry of the Ti(III)/Ti(IV) redox couple in aqueous sulfuric acid has been conducted in previous literature by Habashy [27], to try and develop a technique for analytically determining Ti(IV). An irreversible electrode reduction was observed which became reversible with an increase in acid concentration up to ~10 M. This was rationalised as being due to the formation of hydrated sulfate complexes of Ti(IV) at high sulfuric acid concentrations. More recently published research
[28] reproduced this reversibility result using square wave voltammetry (SWV) on a mercury drop electrode, as well as DC polarography of Ti(IV) solutions in sulfuric acid.
1.3.1. Thermodynamics
The titanyl ion is reduced in aqueous solutions following Eqn. (1.28) [6]; i.e.,
27
2+ + − 3+ 0 TiO + 2H + e ↔ Ti + H2O E = +0.099 V (1.28)
Various other non-stoichiometric titanium oxides with the general formula TinO2n-1 are known to exist and can also be formed by reduction in aqueous solution [23]. These oxides only occur as intermediate species up until TiO2 is the dominant, stable species. Some examples of these reductions are given in Eqn. (1.29 – 30); i.e.,
2+ − + 3TiO + 2H2O + 2e → Ti3O5 + 4H
E(V) = 0.5568 + 0.1183 pH + 0.0887 log(TiO2+) (1.29)
2+ − + 4TiO + 3H2O + 2e → Ti4O7 + 6H
E(V) = 0.9957 + 0.1775 pH + 0.1183 log(TiO2+) (1.30)
Actual thermodynamic data could not be located for the reduction of other Ti(IV) species, such as
2+ 4+ Ti2(OH)6 or Ti3O4 for example, in aqueous solution.
1.3.2. Redox Mechanisms
The oxidation of Ti(III) to Ti(IV) occurs via a simple one electron transfer, followed by hydrolysis of one coordinated water to form an oxo ligand. However, the opposite reduction of
Ti(IV) to Ti(III) has been proposed to be a much more complex kinetic mechanism [28], as is discussed below.
28 1.3.3. The CE or ECE Mechanism
A CE, EC, ECC or ECE mechanism is a reaction mechanism involving a combination of electron transfer at the electrode surface; i.e., an electrochemical (E) step(s), coupled with a homogeneous chemical (C) step(s) [29]. So a CE mechanism involves a homogeneous chemical step followed by an electrochemical step. The other combinations mention before are just different combinations of these steps. As an example, consider the following generic reaction:
− A ↔ B e →D (1.31)
where species A is not electrochemically active, but species B is electrochemically active. There exists a chemical equilibrium between the species A and B. Also B can be reduced, reversibly or irreversibly, in an electrochemical step to species D. Therefore, Eqn (1.31) is an example of a CE mechanism.
A similar process is believed to occur for the electrochemical reduction of Ti(IV) in sulfuric acid media [30, 31]. In other words, there exists various equilibria between species of
Ti(IV), some of which are electrochemically active, while others are not. The presence of these electrochemically active species was found dependent on the acid concentration, and divided the species into two cases for low (1 - 4 M) and high (>4.5 M) acid concentrations [28].
1.3.3.1. Low Acid Concentration Mechanisms
In relatively low acid concentration solutions; i.e., 0.4 M H2SO4, and assuming only monomeric titanium species exist, the reaction sequence can be represented by a first order CE mechanism
(Eqn. 1.32). This sequence involves a slow electrode reaction and a rate determining
29 homogeneous chemical reaction [28], where S1, S2, and S3 are monomeric titanium species, which are as yet, to be identified; i.e.,
← →k1 SS12k 2 (1.32) e− SS23 →
Through the use of square wave voltammetry, it was observed that at low potentials, (-
1180 mV vs SCE), an anodic wave of S3 appears close to the potentials where the irreversible reduction of S2 occurs (Figure 1-5), and then the oxidation of S3 lead to an S1 species [28]. Thus, from Eqn. (1.32), a reaction scheme was proposed [28]; i.e.,
(1.33)
The arrow between the S1 and S3 species indicates the direction of the slow electrochemical reaction. This reaction scheme is a three-stage ECE mechanism, where two oxidised species are bound by a slow pseudo-first order chemical reaction [28].
30 Species H Species F
Species G
Figure 1-5. Curves of Ti(IV) in SWV, with forward scan ∆I < 0 and reverse scan ∆I
> 0. The numbers on the curves are the concentrations of H2SO4 (M) [28].
1.3.3.2. High Acid Concentration Mechanisms
When the acid concentration is increased to greater than 4.5 M, only two overlapping waves were observed when SWV was performed [28]. The first wave corresponds to a relatively
31 fast charge transfer, while the second wave was irreversible. Under these conditions the scheme proposed above (Eqn. (1.33)) becomes Eqn. (1.34) [28]; i.e.,
(1.34)
This ECE scheme correlates closely with that stated by other earlier work [31]. It is believed that the formation of an intermediate species, most likely a mixed oxidation state species (as in Eqn
(1.27)) occurred in the reduction of Ti(IV) to Ti(III). Neglecting any speciation of Ti(IV) and
Ti(III), the ECE process may be as follows [31].
E Step: TiIV + e− → TiIII (1.35)
C Step: TiIV + TiIII ↔ [TiIV − TiIII ]
or TiIV + TiIII kc →[TiIV − TiIII ] (1.36)
E Step: [TiIV − TiIII ]+ e− → 2TiIII (1.37)
where the intervening chemical step may be either an equilibrium or a second order irreversible reaction [31].
1.4. Electroactivity of Various Ti(IV) Species
The use of square wave voltammetry by Fatouros et al, with 2.1×10-3 M solutions of
Ti(IV) in sulfuric acid showed the presence of three different electroactive species [28]. These
32 species were given generic labels F, G and H and they are likely to correspond to those listed in
Eqns. (1.21 – 23). Looking at the curves obtained from the forward scans from SWV, species F is ascribed to the second peak observed in a 0.4 M H2SO4 solution, or the third peak in a 2.6 M
H2SO4 solution (Figure 1-5). Species G is ascribed to the first peak in a 0.4 M H2SO4 solution, or the second peak in a 2.6 M H2SO4 (Figure 1-5) and 4.5 M H2SO4 solution (Figure 1-6). Species H is ascribed to the first peak in a 2.6 M H2SO4 solution and a 4.5 M H2SO4 solution (Figures 1-5 and 1-6). However, a preceding protonation chemical reaction is believed to form electroactive species H.
33 Species H
Figure 1-6. Curves of Ti(IV) in SWV, with forward scan ∆I < 0 and reverse scan ∆I
> 0. The numbers on the curves are the concentrations of H2SO4 (M) [28].
Table 1.1 summarises the acid concentrations at which these species were found (from
Figures 1-5 and 1-6), along with their respective peak potentials and any knowledge of the chemical forms of these species, as described above [28].
34 Table 1.1. Summary of the electroactive species, F, G and H [28].
Approximate Ep /mV Species [H2SO4] / M Chemical Form (vs SCE)
-850 0.4 2_ - F SO4 , HSO4 -750 2.6
-200 0.4 2_ - G SO4 , HSO4 -300 2.6
-100 2.6
H -50 4.5 H+ step *
+100 8.8
* Protonation chemical step in a CE mechanism forms the electroactive species.
No values for the potentials, or the electroactivity of the oligomers, as in Eqns. (1.17 - 19), has been reported.
1.5. Polarography of Titanium(IV)
The polarographic behaviour of aqueous titanium solutions has been investigated previously by various authors [28, 32]. The earliest report by Lingane et al., [32], published in the
1950’s, compared the half-wave potentials of the titanic-titanous (Ti(IV)/Ti(III)) couple, and concluded that in dilute acid, the reduction of Ti(IV) required a larger overpotential than the oxidation of Ti(III). The authors go on to propose that in fairly strong sulfuric acid, the couple behaves reversibly, and the complexation of Ti(IV) in sulfuric acid must involve the hydrogen
- 2- sulfate (HSO4 ) ion, not just the sulfate (SO4 ) ion. However, in weak hydrochloric acid, the couple is highly irreversible, but as the concentration of hydrochloric acid increases, it becomes more reversible. In strong perchloric acid Ti(III) has been oxidized by the perchlorate ion and so
35 does not produce a reversible wave. This indicates that both H+ and Cl- ions are required for any reversible behaviour to occur. It is interesting to add that they also investigated phosphoric acid solutions where the couple behaves perfectly reversibly, even in dilute solutions.
In more recent work [28], it was observed that whatever the concentration of sulfuric acid that was used, the limiting poloarographic currents, immediately preceding proton reduction, are diffusion controlled. Furthermore, the diffusion coefficients (D) were calculated (Figure 1-7)
[28], and were shown to be practically inversely proportional to the sulfuric acid viscosity. D0 is
4.2 x 10-6 cm2s-1 and is included as it takes into account the sulfuric acid viscosity as a function of water viscosity.
-6 2 - Figure 1-7. Diffusion coefficients, D, of Ti(IV) in H2SO4 at 25°C. D0 = 4.2 x 10 cm s
1 [28].
36 1.6. Constant Current Electrolysis of Titanium containing solutions
Constant current electrolysis is based on Faraday’s law, which relates the total charge passed in an experiment (Q) to the number of molecules electrolysed (N) and the number of electrons involved in the electron transfer reaction (n) [33]; i.e.,
Q = n F N (1.38)
where F is Faraday’s constant (96486.7 C/mol).
This technique has been applied to solutions of Ti(IV) in acid solutions by Foller et al.
[3], involving a method of reducing Ti(IV) to Ti(II) via an electrochemical cell. The literature describes the difference in the rate and current efficiency of the reduction of Ti(IV) when certain variables are changed; i.e., Ti(IV) and sulfuric acid concentrations, temperature, current density, cathodic substrate and electrolyte flow rate. The authors state that if the electrolysis is carried out at a temperature above 35°C, it is most beneficial for the electrochemical kinetics of Ti(IV) reduction. Plus, if the temperature is high, generally the current efficiencies are also high.
However, the authors state that the temperature must not exceed 70°C, as at higher temperatures hydrolysis of Ti(IV) to hydrous TiO2 occurs. They believe that the optimal temperature was in the range 45 – 55°C, Ti(IV) concentration was in the range 5 – 10 wt%, sulfuric acid concentration was in the range 20 – 25 wt%, and the optimal cathodic substrate was lead due to its large hydrogen evolution overpotential. As the current density was increased (≥ 500 Am-2), hydrogen evolution at the electrode surface also increased and thus current efficiency decreased greatly.
37 1.7. Hydrogen Evolution Reaction (HER)
To prevent the hydrolysis of Ti(IV), a highly acidic media must be used to ensure the
Ti(IV) stays soluble; e.g., 3 – 7 M H2SO4 [4], and in our experience leads to a solution that is stable indefinitely. Under such circumstances there will be a competition between the possible electrochemical reductions at the cathode. One of the major competing reduction reactions is that involving protons; i.e., the hydrogen evolution reaction (HER):
+ − 0 2H + 2e ↔ H 2 E = 0.00 V (1.38)
As we can see by comparison with the reduction of the titanyl ion (Eqn. 1.28), the titanyl ion is actually thermodynamically favoured by 0.190 V. But due to the fact that in highly acidic solutions, H+ is present in larger concentrations than Ti(IV), and is also the most mobile ion, this reduction may not be the most kinetically favoured.
1.7.1. Proton Diffusion
+ -9 The diffusion coefficient (D) of H in aqueous solutions is abnormally high; i.e., 7.3×10
2 -1 m s [34], which is 10 – 100 times greater than other ionic species. To rationalise this observation, various mechanisms have been proposed that have been the subject of much debate, though the most notable is the ‘Grotthuss Mechanism’ [35]. There are a few similarities between these mechanisms, one of them being the fact they include some form of proton ‘hopping’. These
+ ‘hops’ are thought to be between a hydronium (H3O ) ion to a freely rotating water molecule [36].
Others have said that the water molecule is not freely rotating, but rather is due to a field induced
+ by an attraction between lone pairs and neighbouring H3O ions [37]. Others describe an ordered
H-bonded network of water molecules [38]. From these various mechanisms, it can be suggested
38 that this ‘hopping’ occurs in a predetermined manner along an ordered network of water molecules, and that addition of H2O units to the network may be rate determining [38], due to either structural rearrangement or the ‘hopping’ effect.
1.7.2. Kinetics
The adsorption of a hydrogen ion to the surface, coupled with a one electron reduction to form an adsorbed hydrogen atom is the first step to occur in the HER mechanism [39, 40]. This is the Volmer reaction; i.e.,
+ − H + e ↔ Hads (1.39)
The cathode substrate and its affinity for protons at its surface determines the surface excess of adsorbed H (θH) at any given potential. The next part of the mechanism is still debatable and the reaction can either proceed via the Heyrovsky reaction (Eqn (1.40)) or the Tafel reaction (Eqn
(1.41)) [26, 27]; i.e.,
+ − Hads + H + e ↔ H2 (1.40)
2Hads ↔ H2 (1.41)
However, in reality it is highly probable that the reaction proceeds via either Eqns (1.40) and
(1.41), and could also occur in parallel depending on the conditions under which the reaction occurred. This would alter the rate determining step and therefore which pathway the reaction would take.
39 1.7.3. Cathodic Substrate Effects
The cathodic substrate which needs to be used to study Ti(IV) reduction needs to have a low affinity for hydrogen. This is vital because, as mentioned above, the HER involves surface adsorption of a hydrogen ion, which leads to an excess of hydrogen adsorbed to the cathode surface. If the surface of the substrate has a high affinity for hydrogen, then the cathodic current that contributes to the HER is large and this is unwanted.
Cathodic substrates with low overpotentials for the HER produce a large exchange current density. The exchange current density (i0) is the maximum current density achievable with no deviation from the equilibrium potential of the redox couple. The exchange current densities for hydrogen evolution have been determined for a range of cathodic substrates as seen in Figure
(1.8) [41].
40 13
Zn Hg 12 Pb
11
10 ) -2 9 ) / (Am
0 Al 8 Ti
-log(i Sn
7
Au 6
Ni 5 Pd Pt 4
Figure 1-8. The exchange current densities for hydrogen evolution on several cathodic substrates [41]. Data points are the negative logarithm of the exchange current densities in Am-2.
As mentioned above, the HER is in direct competition with the reduction of Ti(IV) in sulfuric acid solutions, and thus it would be favourable to choose a cathodic substrate that has a low exchange current density for hydrogen evolution. However, a cathodic substrate for which
41 Ti(IV) reduction has good kinetics should be chosen and so a substrate that is the best fit for both qualities should be chosen.
1.8. Leachate Impurities of Ilmenite
Considering ilmenite is the starting material for the Sulfate Process [2], there is always going to be Fe(II) and Fe(III) present in relatively high concentrations in the process liquor. There is evidence of possible associations between Fe(II), Fe(III) and Ti(IV) species [23, 26], and the effect these connections might have on the electrochemical properties of Ti(IV) are still being investigated.
Other impurities can enter the system via the addition of scrap iron and effect the current efficiency of Ti(IV) reduction. This scrap iron is needed for the generation of Ti(III) so it can then in turn reduce Fe(III) to Fe(II) [2]. These impurities may include other metals such as Ni, Co and
Sb. It is known that in the electrowinning of Zn, the presence of these metals is detrimental as they catalyse the HER and promote Zn dissolution [42]. This increase in the production of hydrogen thus decreases the current efficiency and one can assume that the same will occur in our solutions if any of these impurities are present.
1.9. Design
From the literature, if Ti(IV) is going to be successfully reduced to Ti(III), a few considerations need to be taken for optimal reduction. It has been shown that when exposed to air, absorbed oxygen will oxidise Ti(III) to Ti(IV), where the titanium is most stable. Therefore, if oxygen evolution is used as the anode reaction, an ion permeable barrier will be required to isolate this compartment so that oxygen does not diffuse into the cathode compartment, and that
Ti(III) does not diffuse into the anode compartment where it can be electrochemically oxidised.
42 As well as any electrochemically evolved oxygen, dissolved oxygen needs to be taken into consideration. This will need to be purged with an inert gas such as nitrogen for example, so as to not oxidise the electrochemically generated Ti(III).
A flow through cell design would be the best as mention above, it allows for fresh solution to reach the cathode and hence produces the maximum current efficiency for Ti(IV) reduction.
1.10. References
[1] Clark, R.J.H. “The Chemistry of Titanium and Vanadium”, Elsevier: Amsterdam
1968.
[2] Tioxide Group, “Manufacture and General Properties of titanium dioxide pigments’, 1999.
[3] Foller, P.C., Vora, R., Allen, R.G., US Patent Number: 5,250,162 (1993).
[4] Rahm, J.A., Cole, D.G., US Patent Number 4,288, 415, (1981).
[5] Pourbaix, M., “Atlas of Electrochemical Equilibria in Aqueous Solutions”, National
Association of Corrosion Engineers, Texas, (1974) 217.
[6] Aylward, G., Findlay, T., “SI Chemical Data”, 6th Ed, John Wiley & Sons Australia, Ltd.,
2008.
[7] Kölle, U., Kölle, P., Angew. Chem. Int. Ed., 42 (2003) 4540 – 4542.
[8] Cotton, F.A., Wilkinson, G. “Advanced Inorganic Chemistry”, 5th Ed., Wiley-
Interscience: New York 1988.
[9] Cservenyák, I., Kelsall, G.H., Wang, W., Electrochim. Acta 41 (1996) 563.
[10] Clark, R.J.H., “The Chemistry of Titanium”, Pergamon Press: New York 1975.
[11] Wieghardt, K., Quilitzsch, U., Weiss, J., Nuber, B., Inorg. Chem., 19 (1980) 2514.
[12] Comba, P., Merbach, A., Inorg. Chem. 26 (1987) 1315.
[13] Gratzel, M, Rotzinger, F.P., Inorg. Chem. 24 (1985) 2320.
43 [14] Johnson, M.D., Murmann, R.K., Inorg. Chem., 22 (1983) 1068.
[15] Ballhausen, C.J., Gray, H.B., Inorg. Chem., Vol. 1, 1 (1962) 111.
[16] Nagypal, I., Fabian, I., Connick, R.E., Acta Chim. Acad. Sci. Hung., 110 (1982) 447.
[17] Babko, A.K., Mazurenko, E.A., Nabivanets, B.I., Russian J. Inorg. Chem., 14 (8) (1969)
1091.
[18] Clearfield, A., Vaughan, P.A., Acta Christallogr., 9 (1956) 555.
[19] Angstadt, R.L., Tyree, S.Y., J. Inorg. Nucl. Chem., 24 (1962) 913.
[20] Stünzi, H., Rotzinger, F.P., Marty, W., Inorg. Chem., 23 (1984) 2160.
[21] Richens, D.T., Sykes, A.G., Comments Inorg. Chem., 1 (1981) 141.
[22] Schneider, N., Comments Inorg. Chem., 3 (1984) 205.
[23] Kelsall, G.H., Robbins, D.J., J. Electroanal. Chem., 283 (1990) 135.
[24] Bard, A.J., Encyclopedia of Electrochemistry of the Elements, Vol. 5, Marcel Dekker, New
York, 1976, pp. 305.
[25] Shlenk, M., Helvetica Chimica Acta, 19 (1936) 625.
[26] Reynolds, M.L., J. Chem. Soc., (1965) 2993.
[27] Habashy, G.M., Z. Anorg. Chem., 306 (1960) 312.
[28] Fatouros, N., Krulic, D., Labari, N., J. Electroanal. Chem., 568 (2004) 55.
[29] Bard, A.J., Faulkner, L.R., “Electrochemical Methods, Fundamentals and Applications”,
John Wiley and Sons, Inc., 1980.
[30] Fatouros, N., Krulic, D., Labari, N., J. Electroanal., Chem., 549 (2003) 81.
[31] Cservenyák, I., Kelsall, G.H., Wang, W., Electrochim. Acta, 41 (1996) 573.
[32] Lingane, J.J., Kennedy, J.H., Analytica Chimica Acta, 15 (1956) 294.
[33] Bott, A.W., Current Separations, 18:4 (2000) 125.
[34] Adams, R.N., “Electrochemistry at Solid Electrodes”, Marcel-Dekker Inc., New
York, 1969.
[35] Agmon, N., Chemical Physics Letters, 244 (1995) 456.
44 [36] Conway, B.E., “Modern Aspects of Electrochemistry”, Vol. 3, Butterworths,
London, 1964.
[37] Bernal, J.D., Fowler, R.H., J. Chem. Phys., 1 (1933) 515.
[38] Eigen, M., Ang. Chem. Intern. Ed., 3 (1964) 1.
[39] Gennero, de Chialvo, M.R., Chialvo, A.C., Electrochem. Comm., 1 (1999) 379.
[40] Saraby-Reintjes, A., Electrochim. Acta., 31(2) (1986) 251.
[41] Harinipirya, S., Sangaranaryanan M.V., Langmuir, 18 (2002) 5572.
[42] Ivanov, I., Hydrometallurgy, 72 (2004) 73.
45 2 Experimental
2.1. Solution Preparation
Solutions for analysis were prepared by dissolving an appropriate amount of hydrated titanyl oxysulfate (TiOSO4.nH2O; >98%; Sigma Aldrich) in the minimum amount of Milli-Q ultrapure water (>18 MΩ/cm) in a clean 100 mL volumetric flask. Under these circumstances the titanyl sulfate took 1 – 2 hours to dissolve completely. Then an appropriate amount of sulfuric acid (H2SO4; 98%; Sigma Aldrich) was added after complete dissolution to produce the desired acid concentration in the final volume. The solution volume was then made up to the mark with additional Milli-Q water. A matrix of solutions were prepared covering the compositional range from 0.1 – 0.5 M in terms of Ti(IV) concentration, and 0.01 – 5 M in terms of H2SO4 concentration. Each solution was examined by various electrochemical methods after varying periods of time to determine the effects of solution age, since eventually each solution will hydrolyse to precipitate titanium dioxide, although solutions with acid concentrations of 1 M and above were stable for at least 4 weeks. Note that selected solutions were also prepared in the reverse fashion for comparison; i.e., the titanyl sulfate was dissolved in an appropriate amount of sulfuric acid and then made up to the mark on the volumetric flask with Milli-Q water.
The exact composition of the as-supplied TiOSO4.nH2O was unknown due to the variable water content. Its composition was therefore determined by thermogravimetric analysis (ambient to 600°C at 10°/min in static air). The analysis determined that there was 24% water in the solid.
This solid was used throughout the experiments, of course taking into account this water content.
During polarographic analysis Cu2+ was used as an internal standard. Therefore, a
2+ solution of 1.0 M Cu was prepared from anhydrous copper sulfate (CuSO4; AR Grade; Sigma
Aldrich) and Milli-Q water.
46 Solutions of Ti(IV) in other acid media; i.e. perchloric acid (HClO4; 70%; Ajax Finechem
Pty Ltd.), nitric acid (HNO3; 69%; Aristar), acetic acid (CH3COOH; Sigma-Aldrich), and hydrochloric acid (HCl; 32%; Merck) were prepared by dissolving titanium metal under heat for a week in each acid and then filling up to mark in a volumetric flask. The titanium was weighed before and after, as the titanium did not dissolve completely in some of the acids. The acid concentrations were all made to be 3 M to aid the titanium digestion. Hydrogen peroxide (H2O2;
30%; Analytical Reagent; Chem Supply) was added to aid oxidation in some cases.
The appropriate copper salt, relative to the acid used, was prepared and used as the internal standard for each acid; i.e. copper(II) perchlorate hexahydrate ((CuClO4)2.6H2O; 98%;
Aldrich), copper(II) nitrate pentahemihydrate (Cu(NO3)2.2½H2O; 98%; Riedel-de Haën), copper(II) acetate monohydrate (Cu(CH3COO)2.H2O; 98%; ACS Reagent; Sigma-Aldrich) and copper(II) chloride (CuCl2; AR Grade; Sigma-Aldrich).
All solutions were freshly made and the electrochemical methods were conducted at 21 ±
1°C.
2.2. Cyclic Voltammetry Measurements
Cyclic voltammetry was conducted using a Pine Instruments Bipotentiostat Model
AFCBP1, in the potential range of +0.5 to -1.5 V vs SCE, with a scan rate of 5 mV/s. Typically, three full cycles were carried out to ensure consistency within the measurements as shown in
Figure 2-1.
47 0.02
0.015
0.01
0.005 )
-2 0 -1.5 -1 -0.5 0 0.5
-0.005
-0.01 Current / (Acm
-0.015
-0.02
-0.025
-0.03 Potential / (V vs SCE)
Figure 2-1. An example of three full cycles from cyclic voltammetry of a solution
2+ containing 0.0 M TiO in 0.1 M H2SO4, at 21 ± 1°C.
Measurements were taken on the prepared solutions using a cell designed to keep the surface area of the lead electrode (described below) constant in each measurement (see Figure 2-
2). The scan rate was kept low and not varied, as at high scan rates, the capacitive current needs to be taken into account [1]. Another reason for keeping the scan rate low is to enable the different species to be able to reach the electrode surface.
The cell consisted of a cylindrical Teflon tube, which was placed on top of the lead foil electrode. The Teflon tube was secured in place using a Perspex cover and base plate by screwing
48 down the top piece via four screws attached at each corner. This ensured that no liquid was lost when added. A platinum foil counter electrode and a saturated calomel reference electrode were used for each experiment. The lead was polished using 800 grit emery paper before each experiment was conducted, and the same amount of liquid electrolyte was used each time. Other metals such as tin, zinc, nickel and iron, were also investigated as the electrode substrate and all the appropriate steps were taken as per the lead electrode procedure.
Pt Foil Saturated
El t d Calomel
Screw Cylindrical
T fl t b Lead
El t d Plastic Base Figure 2-2. A schematic of the cell designed for cyclic voltammetry experiments.
2.3. Impedance Spectroscopy (EIS) Measurements
The EIS tests were conducted on the prepared solutions detailed above. They were carried out using a Solartron SI 1287 Electrochemical Interface coupled with a Solartron
Analytical 1252A Frequency Response Analyser, using the ZPlot software. The experiment was designed to collect an impedance spectrum (at 20 kHz to 0.1 Hz using a 5 mV AC excitation signal), at a high potential, step down the potential, allow the system to equilibrate for 10 minutes, then collect another impedance spectrum again, and so forth across a potential range of -
1.5 to -0.5 V vs SCE (see Figure 2-3).
49 (V) Run Impedance
St
10 mins
(Time) Figure 2-3. A construal schematic of the program setup for EIS measurements.
The data was then analysed using complex non-linear least squares regression with curve fitting (Figure 2-4).
50 10
9
8
7
6 Ω 5 -Z" /
4
3
2
1
0 0 5 10 15 20 25 30 35 40 Z' / Ω
Figure 2-4. An example of curve fitting applied to EIS data.
2.4. Rotating Disc Electrode Voltammetry
Rotating disc electrode measurements were conducted on the prepared solutions using a
Pine Instrument Company MSRV Speed Control coupled with a Model AFCBP1 Bipotentiostat and using various rotation rates ranging from 100 – 3000 rpm. The disc was a lead disc, and it was polished before each experiment with an alumina (0.1 mm) micropolish suspension (Pine
51 Research). A saturated calomel reference electrode and a Pt counter electrode were used.
The conditions used were a potential range from -1.5 to +0.5 V vs SCE and a scan rate of
5 mV/s.
2.5. Electrochemical Protocol for Polarography
50.0 mL of the TiOSO4/H2SO4 solution being analyzed was pipetted into the polarography cell. 200 µL of the 1 M Cu2+ internal standard was added and the solution was then purged with water-saturated N2 for at least 10 minutes to remove dissolved oxygen. After this time a differential pulse polarogram of the solution was recorded using a Metrohm 663 VA Stand
Polarograph with 626 Polarecord attached, with the settings shown in Table 2.1. Note that the
Cu2+ internal standard was modified for each different acid media, as the titanium concentration was much lower, so therefore the copper concentration was lowered.
Table 2.1. Polarographic settings.
Differential pulse polarography 10 Mode mV pulse
SMDE
Voltage range +0.1 V to -1.3 V vs Ag/AgCl
Scan rate 5 mV/s
Voltage pulse 10 mV
Drop time 2 s
Drop Size 1
52 So that the results from individual polarograms could be compared, a Lorentzian lineshape
(Figure 2-5) was firstly fit to each of the peaks present in the polarogram; i.e.,
W 2 i (2.1) = max i 2 2 4 (W / 2) + (V −Vp )
where i and imax are the measured and peak currents (µA), respectively, W is the peak width at half height (V), V is the potential (V vs Ag/AgCl) and Vp is the potential (V) at which the current maximum occurs. In some instances the peaks were distorted, in which case the peak was fit with multiple lineshapes with the same Vp and imax values; i.e., only W varies on either side of Vp. The area under each fitted polarographic peak was then determined using numerical integration.
53 6
5
4 A) µ
3 Current / (
2
1
0 -1.4 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 Potential / (V vs Ag/AgCl)
Figure 2-5. An example of a Lorentzian lineshape graph.
2.6. References
[1] Bard, A.J., Faulkner, L.R., “Electrochemical Methods, Fundamentals and Applications”, John
Wiley and Sons, Inc., 1980.
54 3 Electrochemistry Results
3.1. Cyclic Voltammetry
3.1.1. Cathodic Substrate
Cyclic voltammetry was conducted on electrodes made of the metals Zn, Sn, Ni and Pb.
It was observed that only Pb showed signs of Ti(IV) reduction before hydrogen evolution. This agrees with the literature found on the subject [1], and thus lead was chosen as the cathodic substrate throughout the remainder of these electrochemical measurements.
3.1.2. Titanium Concentration Effects
Figure 3.1 shows the data obtained from the cyclic voltammetry sweeps on
2+ solutions of 0.1 M H2SO4 with varying TiO concentrations using a Pb cathode. Peak (1) at -
0.440 V is present in all solutions that have titanyl oxysulfate present. As this is absent in the solution not containing TiO2+, we can presume that this peak is due to TiO2+ reduction. Also, since the size of the peak doesn’t vary as we increase the TiO2+ concentration, we can say that this reduction is not concentration dependent. That is, normally the reaction rate increases as the concentration increases according to the rate law, but this is not occurring in this system.
55 0.0 M TiO2+
(2)
(1) 0.03 M TiO2+
(4) 2+ (2) (1) 0.06 M TiO ) -2 (4) (1) 2+ (2) 0.09 M TiO Current / (Acm
0.005 Acm-2
(4) 2+ (2) (1) 0.13 M TiO
(3)
(4)
-1 -0.8 -0.6 -0.4 Potential / (V vs SCE)
Figure 3-1. Cathodic sweep only from the cyclic voltammogram, of solutions
2+ containing different TiO concentrations in 0.1 M H2SO4 at a scan rate of 5 mV/s using a Pb cathode, at 21 ± 1°C. The numbers in bold brackets, (1) – (4) are explained in the text.
56 Peak (2) at -0.526 V was unexpectedly seen in the top curve, which only consists of 0.1
M H2SO4. This peak does not correspond to lead sulfate reduction as this reduces at a lower potential (Eqn. (3.1)), and there is no peak on the graph corresponding to this species. Therefore, we can presume that this reduction does not occur at all in these solutions under these conditions.
- 2- PbSO4 + 2e → Pb + SO4 E° = -0.604 V (vs SCE) (3.1)
Another lead species that could be reduced is PbO, which reduces at -0.824 V (Eqn.(3.2)). But under acidic conditions, this lead species reduces at +0.956 V (Eqn. (3.3)), so we do not need to consider this reduction occuring under our experimental conditions.
- - PbO + H2O + 2e → Pb + 2OH E° = -0.824 V (vs SCE) (3.2)
+ - PbO + 2H + 2e → Pb + H2O E° = +0.956 V (vs SCE) (3.3)
Still, looking at Peak (2), intrestingly when TiO2+ is added the peak is suppressed suggesting interaction between TiO2+ in the solution and PbO on the electrode surface. Also, as the concentration of TiO2+ increases the potential at which Peak (2) occurs, does not change.
Peak (3) at -0.597 V is seen in the most concentrated TiO2+ solution only. However, the other solutions do show long sholders on their curves leading up to Peak 4. Normally peaks are very definate, but when they lose a nice shape, its an indication of the reduction of species with similar reduction potentials. Thus, the species reducing at Peak (3) is present in all solutions containing TiO2+, but is not concentrated enough in solution to produce a reduction potential until the TiO2+ is high enough.
Peak (4) at -0.683 V is seen in all solutions that contain TiO2+. In the most concentrated solution (0.13 M) the peak has the highest current. This peak is well defined in this higher
57 concentration solution, whereas in the other solutions with 0 – 0.09 M TiO2+ concentraions, this peak has a much smaller current at the same potential (within ± 0.01 V).
Therefore, in this solution (0.1 M H2SO4), the peaks occurring at -0.44 ± 0.01 V can be assumed to be a species non dependant on titanium concentration, as it is not observed in the solution with no TiO2+, and the current generated for the increasing titanium concentration also increases; i.e., more TiO2+ present means more of the species is present and more current is generated. Also, the peaks observed at -0.597 and -0.72 ± 0.04 V will most likely be due to the presence of TiO2+ as well, as no peak is present in the 0 M TiO2+ solution but is present in all the solutions containing TiO2+.
The peak observed at position (2) is present in all solutions, so we can attribute this to be from the H2SO4 presence, which we will look at closer further on.
Overall, we can see that as the titanium concentration increases, the peak currents for the
Peaks (2) and (4) (from Figure 3-1) also increase, except for the 0.13 M TiO2+ solution, which has a smaller peak current for the peak (4) (see Figure 3-2). Normally, peak current is proportional to the concentration of the solution via the Randles-Sevcik equation; i.e.,
5 2/3 1/2 1/2 * Ip = (2.69x10 ) n A D0 v C0 (3.4)
2 where n is the number of electrons used in the reaction, A is the area of the electrode (cm ), D0 is
2 * the diffusion coefficient (cm /s), v is the scan rate (V/s), and C0 is the bulk concentration
(mol/cm3).
58 0 Peak (1)
Peak (2) -0.002
-0.004 ) -2 Peak (3)
-0.006 Current / (Acm -0.008
-0.01
Peak (4)
-0.012 -0.02 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 TiO2+ Concentration / (M)
Figure 3-2. The maximum peak currents corresponding to peaks (1) – (4) from
2+ Figure 3-1 of the solutions containing varying TiO concentrations in 0.1 M H2SO4.
So from Figure 3-2, as the concentration of TiO2+ increases, so should the peak current.
However, Figure 3-2 clearly shows that this is not the case for Peaks (2) and (4) for this solution.
When we reach a TiO2+ concentration of 0.09 M for Peaks (2) and (4), the peak current increases greatly, but then for 0.13 M TiO2+, the peak current decreases. Since these peaks all occur at similar potentials, we can say that they may be due to different sized oligomers. If this system
59 was an ideal system we would see that when the concentration of titanium is increased, the concentration of the species formed would increase and thus an increase in current would occur.
Since we see a deviation from this reasoning; i.e., the peak current is not proportional to concentration, we can presume that some speciation is occurring and inhibiting the process. With the occurrence of Peak (4), we can say a new species is formed and that it detracts from the other species formed, and that is why we see a decrease in the current as there is a decrease in that species formed.
2+ The solution containing 0.01 M H2SO4 and varying TiO concentrations (Appendix B,
Figure 7-1) shows the solutions with 0.09 M and 0.13 M TiO2+ have two peaks. These peaks correspond to Peaks (3) and (4), and the solutions with 0.03 M and 0.06 M TiO2+ only have Peak
(4). Peaks (1) and (2) are not present at all.
2+ The solution containing 1 M H2SO4 and varying TiO concentrations (Appendix B,
Figure 7-2) shows the solution containing no TiO2+ with a peak at -0.66 V, which is not seen in
2+ the other H2SO4 concentrated solutions. The 0.09 M and 0.13 M TiO solutions also have this peak at -0.66 V. These two solutions also show another peak at -0.61 V (0.09 M TiO2+) and -
0.71 V (0.13 M TiO2+) which correspond to Peak (3) and Peak (4) from Figure 3-1 respectively.
The solutions containing 0.03 M TiO2+ and 0.06 M TiO2+ have a peak each at -0.61 V and -0.58 V respectively.
2+ The solution containing 5 M H2SO4 and varying TiO concentrations (Appendix B,
Figure 7-3) shows the solution containing no TiO2+ with a peak at -0.66 V, like the one seen from
2+ the 1 M H2SO4 concentrated solutions. The rest of the TiO solutions all have this peak at -0.66
V with solutions 0.06 M TiO2+ and 0.09 M TiO2+ having another peak at -0.61 V. This corresponds to Peak (3) from Figure 3-1. There is a lot of “noise” seen in these curves due to the high concentration of H2SO4 encouraging hydrogen evolution to occur quicker.
2+ Overall, Peaks (3) and (4) are observed in all solutions no matter the H2SO4 or TiO concentration. The peak current changes depending on the H2SO4 concentration so we can say
60 that the oligomers formed are the same species, as they occur at the same potential, but they differ in size from solution to solution.
3.1.3. Sulfuric Acid Concentration Effects
Figure 3-3 shows the data obtained from cyclic voltammetry conducted on a 0.09 M
2+ TiO solution with varying H2SO4 concentrations.
-0.05 1 M H2SO4
0.01 M H2SO4
0.1 M H2SO4 -0.25
5 M H2SO4 ) -2 -0.45
Current / (Acm / Current -0.65
-0.85
-1.05 -1.5 -1.2 -0.9 -0.6 Potential / (V vs. SCE)
61 Figure 3-3. Cathodic and Anodic sweep from cyclic voltammetry of solutions
2+ containing different [H2SO4] in 0.09 M TiO at a scan rate of 5 mV/s using a Pb cathode, at 21 ± 1°C.
The arrows indicate the direction that the sweep has been made. It is interesting to note that for the anodic sweep the cathodic current increases in magnitude, which is unusual as normally the current generated is lower. This indicates that the surface has become activated
2+ towards TiO reduction and possibly H2 evolution. The nature of this activation is unknown at this stage, but may involve surface cleaning. This possible effect from the hydrogen evolution reaction is also worsened by the increase in acid concentration.
The voltammograms obtained in Figure 3-3 have been re-scaled to highlight the area where the reduction peaks occur for the titanium species; i.e., where the titanium species are active (see Figure 3-4), and not PbSO4 (Eqn. 3.1).
62 -0.002
0.01 M H2SO4
-0.006 ) -2
-0.01 Current / (Acm
0.1 M H2SO4
-0.014
1 M H2SO4
5 M H2SO4
-0.018 -0.85 -0.8 -0.75 -0.7 -0.65 -0.6 -0.55 Potential / (V vs. SCE)
Figure 3-4. Cathodic sweep only from cyclic voltammetry of solutions containing
2+ different [H2SO4] in 0.09 M TiO at a scan rate of 5 mV/s using a Pb cathode, at 21
± 1°C, highlighting the electroactive potential region.
As the acid concentration is increased the current for the peak generated also became greater in magnitude. It is interesting to note that for both the 1 M and 5 M acid solutions, the current obtained for the peaks is the same (±0.2 mA) which is an indication that they may contain oligomeric species that are of similar size. However, these peaks occur at different potentials; i.e.,
63 -0.63 V for 5 M acid and -0.60 V for 1 M acid. This indicates that the two peaks may represent different species, as each are possibly electrochemically active at different potentials. Therefore, as well as proposing that they may contain similar sized oligomers, and thus have similar mass transport rates, they are in fact different species, the formation of which may be acid dependent.
This could be due to either a mixed oxidation state titanium species forming, or maybe the
- + bridging species is the HSO4 ion, instead of just the sulfate ion, as there are more H ions available in the higher acid concentration solution. So these species could be similar size, but have different ions forming the species.
The shapes of the waves are also different for the two lower acid concentrations compared to the higher acid concentrations. The two lower acid concentrations have a smaller change in current for their waves, whereas the two higher acid concentrations have relatively sharp waves. The shallow, long waves suggest slow kinetics for the solution, whereas a sharp peak suggests relatively facile kinetics.
The other solutions described in the Experimental (Chapter 2), containing various TiO2+ concentrations, i.e., 0.03, 0.06 and 0.13 M, were all examined via cyclic voltammetry. For all these solutions, those having the highest acid concentration always generated the largest peak current relative to that series; i.e., if the TiO2+ concentration was 0.03 M, the peak current generated in the 5 M H2SO4 solution was larger than for all the other acidic concentrations. Also,
2+ 2+ the solution with the highest TiO and H2SO4 concentrations (0.13 M TiO and 5 M H2SO4), generated the largest peak current for all the solutions, at a value of 0.05 Acm-2. Furthermore, at this high concentration of TiO2+, we all but lose any form of peak occurring for the low acid concentration solutions.
64 3.1.4. Kinetic Information
The exchange current density can be estimated from the cyclic voltammetry data collected from each sample, as for example, shown in Figure 3-7. This is done by first calculating the reversible cell potential, ECell, for each solution using the Nernst equation; i.e.,
2+ + - 3+ TiO + 2H + e → Ti + H2O (3.5)
2 2++ 0 RT TiO H EE=23++ + ln (3.6) TiO /Ti nF Ti3+
where E0 is calculated from the literature [5], n = 1 for this case, [TiO2+] is 0.03 – 0.13 M dependant on the solution, [H+] is 0.02 – 10 M dependant on the solution, and [Ti3+] will be very small, approximately at 1×10-6 M as an impurity level, since it has not been deliberately added.
For the higher acid concentrations, the activity coefficients were taken into account as found in
Reference [6]. This value is then taken and used on the respective graphs log(i(Acm-2)) versus potential for each solution and the current line extrapolated as shown by the example in Figure 3-
5.
65 0 Reversible redox potential
-1
-2
i o -3 ) -2
-4 /Acm i Tafel slope log( -5
-6
-7
-8 -1.5 -1.3 -1.1 -0.9 -0.7 -0.5 -0.3 -0.1 0.1 0.3 0.5 Potential / (V verses SCE)
Figure 3-5. log(i/Acm-2) versus potential for exchange current density calculation of
2+ a 0.1 M H2SO4 and 0.03 M TiO solution at 21 ± 1°C, for peak (4) from Figure 3-1.
The exchange current densities are now shown in Figure 3-6.
66 -2
-2.5
-3 0.06 M TiO2+ 0.13 M TiO2+ ) -2 -3.5 ) / (Acm i -4 0.09 M TiO2+ log(
-4.5
-5
0.03 M TiO2+ -5.5 0.01 0.1 1 10
log ([H2SO4]) / (M)
Figure 3-6. Exchange current densities as a function of H2SO4 concentration, of the solutions containing varying TiO2+ concentration, at 21 ± 1°C.
This figure shows the exchange current density measured for all the solutions as a function of acid concentration. As the H2SO4 concentration increased, the exchange current densities decreased. The values are all quite low which overall suggests slow kinetics. As we go to 1 M H2SO4, the exchange current density for all solutions increase, indicating an increase in reaction kinetics. This makes sense as there is more sulfuric acid in solution and therefore there is
67 a higher concentration of a titanium-sulfuric acid species formed. It could also be due to the presence of more H+ in solution.
Examining the 5 M H2SO4 solutions, the exchange current densities have all increased except for the most concentrated TiO2+ solution. This solution has decreased quite substantially in terms of exchange current density. As mentioned before, a larger exchange current density
2+ indicates a faster reaction. So for the 5 M H2SO4 and 0.13 M TiO solution, the reaction that is occurring is happening at a relatively slow rate. This could be again from the increased concentration of H+ that are in solution, as well as a higher TiO2+ concentration. The more ions in solution, the larger the oligomers formed and thus, the slower the reaction.
Since the exchange current density is dependent on concentration, we will also look closer at the effect TiO2+ concentration has on this value (see Figure 3-7). It is interesting to note that as the TiO2+ concentration increased, so does the exchange current density. Here we can see that at low TiO2+ concentrations, the exchange current densities are quite different with the
2+ solution in 1 M H2SO4 having the smallest value. As we increase the TiO concentration, the exchange current densities in most solutions increased slightly and then decreased and finally increased again at higher TiO2+ concentration. However, again we see an irregularity for the 0.13
2+ M TiO and 5 M H2SO4 solution as was seen in Figure 3-6. This irregularity could be due to the formation of oligomers, and/or a different mechanism could be taking place; e.g., an ECE or a
CEC.
68 -2
0.01 M H2SO4 -2.5
-3
0.1 M H2SO4 )
-2 -3.5 /Acm i
log( -4 5 M H2SO4
-4.5
1 M H SO -5 2 4
-5.5 0.02 0.04 0.06 0.08 0.1 0.12 0.14 [TiO2+]/ (M)
Figure 3-7. Exchange current density as a function of TiO2+ concentration, for solutions containing varying H2SO4 concentrations, at 21 ± 1°C.
3.1.5. Solution Age Effects
2+ Figure 3-8 is an example of the age effects of a 0.01 M H2SO4 and 0.09 M TiO solution over 24 hours.
69 0 Hours
1 Hour
6 Hours ) -2 Current / (Acm 0.01 Acm-2 15 Hours
18 Hours
24 Hours
-1.5 -1.3 -1.1 -0.9 -0.7 -0.5 Potential / (V vs SCE)
Figure 3-8. Cathodic sweep only from cyclic voltammetry, of a solution containing
2+ 0.09 M TiO and 0.01 M H2SO4 over 24 hours, at a scan rate of 5 mV/s using a Pb cathode, at 21 ± 1°C. Dotted lines indicate baselines.
70 Physically, at this low H2SO4 concentration, the solution became milky over the 24 hour study
2+ period due to hydrolysis of the TiO leading to TiO2 precipitation. The changing nature of this precipitate in various solutions was interesting as it is an indication of various particle sizes of
TiO2 forming. This will undoubtedly have a great effect on the electroactivity of the solution as titanium is precipitating out of the solution as a solid, and is hence not electroactive. This solid could also contain crystals of such a small size that they cannot be seen with the naked eye and if they were not electroactive, this would also impact in the current produced by the solutions.
The current generated changes constantly over the 24 hours. From 0 to 6 hours, the current decreased by a factor of two. Then there is a massive spike of current at 15 hours to ~ -
1.69 x 10-2 Acm-2, three times higher than the current generated at 1 hour. The current then decreased again to about half its magnitude. This indicates that the electroactivity of the species in the solution changes with age. As mention before, this could be due to titanium precipitating out as a solid, but it also could be due to oligomer formation.
The potentials at which the peak currents occur, also constantly change over the 24 hours, as seen in Table 3.1. After the massive spike in current at 15 hours, we can see that in the wave obtained at 18 hours, a second bump appears in the curve, indicating that another species with substantially different activity is now present in the solution. The fact that just before this new species was formed there was a huge increase in the amount of current generated suggests that the species was extremely electrochemically active and that due to hydrolysis, the current is greatly decreased. Hydrolysis followed by agglomeration of the hydrolysed species is likely to occur in this solution, as it is only 0.01 M H2SO4, so not much sulfate is present for any sulfate bridging to form. However, it will still depend on how many sulfate ions are used to form the bridges and how many bridges are actually formed. From Table 3.1 we can see that this extra peak present in the 18 and 24 hour measurements occurred at -0.599 V which is similar to Peak (3) present at -
0.61 V in Figure 3-1. We proposed that this species was due to the presence of sulfuric acid, but from these measurements it seems that it may be due to hydrolysis.
71
Table 3.1 Potential and Current Values for the Peaks from Figure 3-8.
Peak 1 Peak 2 Time / (hrs) Potential / (V vs SCE) Current / (Acm-2) Potential / (V vs SCE) Current / (Acm-2)
0 -0.704 -0.01183 NA NA
1 -0.625 -0.00568 NA NA
6 -0.586 -0.00504 NA NA
15 -0.752 -0.01693 NA NA
18 -0.599 -0.00410 -0.721 -0.00645
24 -0.599 -0.00391 -0.775 -0.00713
It is also interesting to note that the potential for the peak current measured at 6 hours is -
0.586 V, which is again a similar potential to Peak (3) obtained from Figure 3-1. This could
indicate that the species formed at this potential is the dominant electroactive species at this time
and then as hydrolysis occurs, other species become the dominant forms.
3.2. Electrochemical Impedance Spectroscopy (EIS)
The EIS measurements were conducted using a SPECS/EIS combination, as described in
the Experimental section (Chapter 2). Figure 3-9 shows the circuit that was used to analyse the
EIS data. This is a modified Randles circuit [7] which includes a constant phase element (CPE) to
account for the microscopic roughness of the electrode surface (Z”CPE), and another to account for
irregular diffusion characteristics (Z’CPE).
72 RCT
Z’CPE
RS
Z’’CPE
Figure 3-9. Electrochemical schematic of the impedance model used, a Randles circuit [7].
A CPE is a constant phase element and can be used instead of a capacitor to take into account non-ideal behaviour. As mentioned above, the electrode surface is not completely smooth and has microscopic roughness, and thus this causes it to not behave ideally as it results in intersecting double layers at the solid-electrolyte interface. The equation for a CPE is given by:
−m mmππ ZCPE = σω cos− jsin (3.7) 22
where σ is the pre-exponential constant (dependant on the solution), ω is the angular frequency, and m is a coefficient which depends on the role of the CPE; i.e., 0 ≤ m ≤ 1, where if m = 1, then it is an ideal capacitor, but if m = 0, it is an ideal resistor. Additionally, if m = 0.5 then it can be considered a Warburg diffusion element.
The data that is collected from these measurements were plotted as Nyquist plots, an example of which is shown in Figure 3-10, for a range of potentials in a solution containing 0.03
2+ M TiO and 1 M H2SO4. We can see that it produces regular semi-circular shaped curves, which
73 are indicative of a kinetic controlled reaction. We can also see that the radius of the semi-circles decreases with decreasing potential, meaning the electrochemical reaction is becoming much more activated when applying more cathodic potentials.
300
250
200 ) Ω
150 -Z" / (
-0.80 V
100
-0.85 V 50 -0.90 V
-0.95 V 0 0 50 100 150 200 250 300 Z' / (Ω)
Figure 3-10. A complex plane plot of Z’ versus -Z”, i.e., a Nyquist plot of a solution
2+ containing 0.03 M TiO and 1 M H2SO4 at a range of potentials. Note: Z’ denoted the real values and Z” denoted the imaginary values.
74
Figure 3-11 shows the charge transfer resistance, RCT, values for the various solutions studied. Typically, resistance is calculated by Ohm’s law, but in this circuit an overpotential is needed to force a current through and we deviate from the Ohm’s Law. This non-ohmic behaviour gives us the RCT value being proportional to 1/i0 (where i0 is the exchange current density) and so can be used as an indication of how facile the reaction is.
0.0 M TiO2+ 700
600 0.03 M TiO2+ 0.06 M TiO2+ 500 )
Ω 400 / ( 2+ CT 0.13 M TiO R 300
200
0.09 M TiO2+ 100
0 -1.05 -0.95 -0.85 -0.75 -0.65 -0.55 Potential / (V vs SCE)
75 Figure 3-11 Charge transfer resistance (RCT) versus voltage for solutions containing
2+ varying [TiO ] in 1 M H2SO4 from EIS measurements at 21 ± 1°C.
When there is no TiO2+ present in the solution, the reactions are occurring at the slowest rate, hence the largest RCT values. In this solution a reaction involving the lead electrode forming a lead oxide or a lead hydroxide species, within the H2SO4 media, is occurring to give us RCT values. The more TiO2+ that is added, the faster the reactions occur, except for 0.13 M TiO2+, which again is irregular. Instead of decreasing in value, it is larger than the values obtained for the
0.09 M TiO2+ solution, and thus has slower kinetics, which could be an indication that larger oligomers are being formed and thus taking longer to react. At TiO2+ concentration of 0.09 M, the sulfate bridges must be binding more tightly to the titanium ions to form larger species, possibly with the aid of some back bonding from titanium. Back bonding occurs when the metal is donating electron density back to the ligands [10]. This was observed in the literature by Comba et al. [11], where they observed σ and π molecular orbitals with high electron density arising from the high energy 3d orbitals of the -yl oxygen in the titanyl ion.
The shapes of the curves also become skewed, and the potentials where the peak maxima occur are shifted to higher potentials. This again could indicate that the oligomers are indeed growing in size as the potentials are similar but slightly higher for increasing TiO2+ concentration; i.e. the oligomers are still the same species, just more chains due to more TiO2+ present.
The curve for the 0.09 M TiO2+ solution has a clear shoulder before reaching the apex of the curve, with multiple underlying processes apparently present. This indicates that different species are being reduced at the electrode surface, as the multiple peaks correspond to multiple charge transfer resistance maxima.
76 2+ The phenomenon occurring for the 0.09 M TiO in 1 M H2SO4 solution is a rarity when compared to the rest of the RCT results (Appendix B, Figures 7-4 to 7-6). The 0.01 M and 0.1 M
2+ H2SO4 solutions show the RCT value decreased with increased TiO concentration. The 0.1 M
H2SO4 solution again has varying potentials at which the peak maxima occur. The peak maximum has the same potential for the 0.03 M and 0.06 M TiO2+ solutions (-0.85 V), and the peak maximum has the same potential for the 0.09 M and 0.13 M TiO2+ solutions (-0.75 V). The solutions having the same potential indicates that the same species are present in both solutions.
There is also a second peak for the 0.06 M, 0.09 M and 0.13 M TiO2+ solutions at -0.60 V. This second peak indicates another species is being reduced at the electrode surface. The 0.01 M
2+ H2SO4 solution shows multiple peaks for all TiO solutions. This solution also follows the trend
2+ of decreased RCT value with an increased TiO concentration. We again have some of the peaks occurring at the same potentials (-0.85 and -0.60 V), so the same species are again present. This low H2SO4 concentration forms the most variety of species in solution, as seen by the multiple
2+ peaks. The 5 M H2SO4 solution, the trend of decreased RCT with increased TiO concentration is not seen. The 0.06 M TiO2+ solution is the slowest reaction, followed by the 0.03 M TiO2+ solution, then the 0.0 M TiO2+ solution. The 0.06 M TiO2+ solution shows two peaks at different potentials compared to the other solutions, but a peak at -0.60 V is still observed, so this species is
2+ present in solutions containing any matrix of TiO and H2SO4 concentrations.
The RCT values obtained from the solutions as a function of H2SO4 concentration are shown in Figure 3-12. As the H2SO4 concentration was increased, the maximum RCT values decreased which is an indication that the reaction is becoming more facile. The maximum RCT values for 1 M and 5 M H2SO4 are roughly the same at 520 Ω and 530 Ω respectively. This indicates that the kinetics of the reaction are the same for the different solutions. However, these maximum RCT values occur at different potential which suggests different species.
77 1000
900 0.01 M H2SO4
800 0.1 M H2SO4
700
600
) 5 M H SO 1 M H2SO4 2 4 Ω
/ ( 500 CT R 400
300
200
100
0 -1.3 -1.2 -1.1 -1 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 Potential / (V vs SCE)
Figure 3-12. Charge transfer resistance (RCT) versus voltage for solutions containing
2+ varying [H2SO4] in 0.06 M TiO from EIS measurements at 21 ± 1°C.
Therefore, we can believe that the reaction is not dependent on sulfuric acid concentration. If it were dependant on the sulfuric acid concentration, you would expect the maximum RCT value for the 5 M H2SO4 solution to be smaller than the maximum RCT value for the 1 M H2SO4 solution.
78 Also of note is the presence of multiple peaks in the curve for all the sulfuric acid concentrations, except for the 1 M H2SO4 solution. As mentioned previously, these multiple peaks are an indication of different species in the solutions.
2+ Figure 3-13 shows the series resistance, RS, values for various TiO concentrations in a 1
M H2SO4 solution.
1.8
0.09 M TiO2+
1.6 0.03 M TiO2+
1.4
0.0 M TiO2+
1.2 ) 0.06 M TiO2+ Ω / ( S R 1 0.13 M TiO2+
0.8
0.6
0.4 -1.1 -1 -0.9 -0.8 -0.7 -0.6 -0.5 Potential / (V vs SCE)
2+ Figure 3-13. Series resistance (RS) versus voltage for solutions with varying [TiO ] in 1 M H2SO4 from EIS measurements at 21 ± 1°C.
79
The RS value is physically based on the position of the electrodes which may vary slightly between experiments. Also, when the RS value is low, it is an indication that the electrodes have a clean surface free of an oxide layer. The sharp increase in RS, at -0.5 V, corresponds to the potential where TiO2+ reduction occurs which may be indicative of a surface layer being formed. At lower potentials the RS value was effectively constant. Another reason for a low RS value, especially in the more concentrated H2SO4 solutions, is due to an increase in conductivity of the solution.
2+ The RS values for varying TiO concentrations in 0.1M H2SO4 (Appendix B, Figure 7-7) are similar to the 1 M H2SO4 solution. A sharp increase occurs at -0.5 V, after which the values remain constant as the potential is decreased. The remaining two H2SO4 based solutions (0.01 M
(Appendix B, Figure 7-8) and 5 M (Appendix B, Figure 7-9)) are very similar, with the RS graphs exhibiting no peaks. So at these lowest (0.01 M) and highest (5M) H2SO4 concentrations, no surface layer is being formed on the electrode surface. The species are staying in solution more and not apparently reacting with the electrode surface.
Figure 3-14 shows the double layer capacitance for the solutions of 1 M H2SO4 with varying TiO2+ concentration. The double layer capacitance demonstrates the charge on the electrode surface not directly associated with redox reactions. These values were calculated by utilising the equation derived by Jovic [8] relating a CPE to the Cdl (an example of which is shown in Appendix C). We observe a large spike for all concentrations at a potential of -0.5 V, after which the data decreases progressively. Again the Cdl value is used as an indicator of when a
2+ reaction initially occurs. The Cdl value for the 0.13 M TiO solution is the largest as the double layer has access to more ions that are present in that solution. The other interesting feature to note
2+ from Figure 3-14, is that as the Cdl values increased they were not increasing with the TiO concentration. This is possibly due to different speciation occurring in the solutions, as larger oligomers have a lower capacitance value and smaller oligomers have a higher capacitance value.
80 These high and low values for the capacitance arise from the arrangement of ions in the double layer. Fewer large oligomers have ions near the electrode surface than small oligomers.
0.00025 0.13 M TiO2+
0.03 M TiO2+
0.0002
0.00015 ) -2 / (Fcm dl C 0.0001
0 M TiO2+
0.00005 0.06 M TiO2+
0.09 M TiO2+
0 -1.1 -1 -0.9 -0.8 -0.7 -0.6 -0.5 Potential / (V verses SCE)
Figure 3-14. Double layer capacitance (Cdl) versus voltage for solutions with varying
2+ [TiO ] in 1 M H2SO4 calculated from EIS measurements, at 21 ± 1°C.
2+ The Cdl results for the solution containing 0.1 M H2SO4 with varying TiO concentrations (Appendix B, Figure 7-10) are very similar to the 1 M H2SO4 solution. A large
81 current peak occurred at -0.60 ± 0.05 V, followed by progressively decreasing values. Again, the
2+ Cdl values don’t follow any perceivable trend with respect to TiO concentration. The solution
2+ containing 0.01 M H2SO4 with varying TiO concentration (Appendix B, Figure 7-11) has this initial peak at -0.60 V, which then decreased in value except for the 0.03 M TiO2+ solution. This solution contains many more peaks, indicating many more small oligomers present. The solution
2+ containing 5 M H2SO4 with varying TiO concentration (Appendix B, Figure 7-12) has one peak at -0.55 V for all solutions, except again for the 0.03 M TiO2+ solution. At this potential, this solution has a very large Cdl value compared to the other solutions in H2SO4. This reinforces the presence of small oligomers forming in the 0.03 M TiO2+ solution, i.e., low TiO2+ concentration means limited titanium ions to which the sulfates can form bridges with.
3.3. Rotating Disc Electrode (RDE) Voltammetry
Figure 3-15 shows an example of the data resulting from a series of RDE measurements at different rotation rates. As the rotation rate was increased, the current produced also increased as expected, as the diffusion layer is thinner at faster speeds giving a higher diffusion current
(Eqn. 3.8) [10], that is the reaction is diffusion controlled.
δ = 1.61 . D1/3.v1/6.ω-1/2 (3.8)
where δ is the diffusion layer, D is the diffusion coefficient (m2/s), v is the kinematic viscosity
(m2/s) and ω is the rotation rate (radians/s).
However, the potentials of the peaks stay the same, as indicated by the dotted lines on
Figure 3.15, which is an indication of kinetic control.
82 0
-1 200 rpm
-2
-3 Current / (mA)
-4 3000 rpm
-5
-6 -1.3 -1.2 -1.1 -1 -0.9 -0.8 -0.7 -0.6 Potential / (V vs SCE)
Figure 3-15. Cathodic sweep a of rotating disc electrode (RDE) scan at 5 mV/s with increasing rotation rate from 200-3000rpm on a solution containing 0.13 M TiO2+ and 5 M H2SO4, at 21 ± 1°C.
Also from Figure 3-15, there is an initial peak at -0.61 V followed by another peak at -
0.67 V. This indicates that there are at least two electroactive species present in the solution. We can also see that when all the curves overlap each other at -0.61 V, this is the stage of reduction
83 that is not mass transport limited; i.e., kinetically limited. Then as we move to lower potentials, the reduction becomes mass transport limited as the current generated for each curve differs, i.e., the lines have separated.
The data series were analysed using the Koutecky-Levich Equation, i.e.,
1 1 1 = + i i i k L (3.9) 1 1 = + 2/3 1/ 2 1/ 6 * ik 0.620nFAD0 ω v C0
where ik is the kinetically controlled current, iL is the mass transfer limited current, n is the number of electrons transferred, F is Faraday’s constant, A is the electrode area, D0 is the diffusion coefficient, ω is the angular rotational velocity, v is the kinematic viscosity and C0 is the concentration of the bulk solution.
A Koutecky-Levich plot (Figure 7-16) is calculated using the relationship between the current and angular frequency of rotation from Equation 3.9. From this plot, we take the y- intercept to give us a value for 1/ik. We then plot log(ik) against the potential to obtain a Tafel plot. The slope of the Tafel plot is proportional to the diffusion coefficient, D, and thus we can calculate the diffusion coefficient as we know all the other values.
84 0.8
-1.0 V -0.9 V 0.7 -1.1 V
0.6 -1.2 V -0.8 V ) -2 0.5
-0.7 V
/ (mA.cm -0.6 V i 0.4 1/ -1.3 V
0.3
0.2
0.1 0.007 0.012 0.017 0.022 0.027 ω-1/2 / (s-1/2)
Figure 3-16. A Koutecky-Levich plot for a solution containing 0.13 M TiO2+ and 5 M
H2SO4 over a range of potentials, at 21 ± 1°C.
85 -0.55
-0.65
-0.75
-0.85
-0.95
-1.05 Potential / (mV verses SCE)
-1.15
-1.25
-1.35 0.7 0.75 0.8 0.85 0.9 0.95 1 1.05 1.1 1.15 -2 Log (ik) / (mA.cm )
2+ Figure 3-17. The Tafel plot for a solution containing 0.13 M TiO and 5 M H2SO4 at
21 ± 1°C.
86
6 2 -1 2+ Table 3.2. Diffusion coefficient (10 cm s ) for solutions with varying [TiO ] and [H2SO4] at 21
± 1°C.
Acid Concentration / (M)
Ti(IV) Concentration / (M) 0.01 0.1 1 5
0.03 6.62 0.04 0.08 0.86
0.06 0.03 0.03 0.09 0.40
0.09 1.12 0.01 0.18 0.71
0.13 0.01 0.01 0.04 0.17
2+ Table 3.2 shows at 0.01 M H2SO4 concentration, there is no trend as the TiO concentration increased. Interestingly, the diffusion coefficient for the solution containing 0.01 M
2+ H2SO4 and 0.03 M TiO , has the highest diffusion coefficient of all the solutions indicating large oligomers have been formed in this solution. The solution containing 0.01 M H2SO4 and 0.09 M
TiO2+ is high as well compared to the other solutions. The rest of the diffusion coefficients are low in value with the 5 M H2SO4 solutions having a slightly higher value. In this high H2SO4 concentrated solution, there are more sulfate ions which leads to more sulfate bridges, which leads to larger oligomers.
Comparing the values obtained from Table 3.2 to other diffusion coefficients, it is interesting to note that for other solutions, an example of which is given below, with varying concentrations, the diffusion coefficients hardly change. For example, with the ferrocyanide ion as the electroactive species, and KCl as the supporting electrolyte, Table 3.3 shows the diffusion coefficients obtained [9]. This shows us that this system is not really affected by the presence of the supporting electrolyte, and thus the diffusion coefficient is due mainly to the presence of the electroactive species. Because our diffusion coefficients are so varied, we can say that the
87 supporting electrolyte plays a major role in the diffusion coefficient for this system via bridging, oligomer formation, etc.
Table 3.3 Diffusion coefficients for Ferrocyanide in KCl [9].
Concentration of KCl / (M) D x 105 / (cm2.sec-1)
0.05 0.662
0.1 0.650
0.5 0.639
1.0 0.632
2.0 0.629
3.0 0.621
3.4. Conclusions
Cyclic voltammetry was used successfully to determine that there are multiple Ti(IV) species present in sulfuric acidic solutions. It was discovered that the electrochemical activity of the titanium species so formed is dependent on the initial titanium concentration and the solution age. From the exchange current density calculation (Figure 3-7) it was shown that the reaction is more facile with more TiO2+ in the solution, so the more TiO2+ there is the more electrochemically active the species are. The impedance experiments also support this finding
2+ (Figure 3-11), as it was shown that the RCT values decreased for increased TiO concentration which means that the reaction is more facile.
The RDE measurements also support these findings, as Figure 3-15 shows the presence of at least two electroactive species. Also, from the diffusion coefficients calculated (Table 3.2) for the solutions, we can see that the variation of these diffusion coefficients is an indication of different species in solutions. Typically diffusion coefficient values do not alter with
88 concentration as this value is based on a pair of species, so the species in our solutions are changing with different concentrations.
The most reproducible result from the above electrochemical methods, is the deviation seen for the 0.09 M TiO2+ solution. In every different method used, the 0.09 M TiO2+ solution produced a result not fitting to the rest of the data obtained. This could be due to either stronger and/or longer sulfate bridges forming, as the titanium ions are not in excess and thus bind to more sulfate ions. Or the electroactive species is in its most favourable formation.
3.5. References
[1] Harinipirya, S., Sangaranaryanan, M.V., Langmuir, 18 (2002) 5572.
[2] Mineral Sands Report, Issue 115, May 2005, pp 15,16.
[3] Foller, P.C., Vora, R., Allen, R.G., US Patent Number: 5,250,162 (1993).
[4] Rahm, J.A., Cole, D.G., US Patent Number 4,288, 415, (1981).
[5] Pourbaix, M., “Atlas of Electrochemical Equlibria in Aqueous Solution”, National
Association of Corrosion Engineers, Houston, Texas, USA, 1974.
[6] Stapled, B.R., J. Phys. Chem., Vol. 10, 3 (1981) 779.
[7] Orazen, M.E., Tribollet, B., “Electrochemical Impedance Spectroscopy”, John Wiley & Sons
Inc., New Jersey, 2008.
[8] Jovic, V.D., Gamry Instruments, Inc., Center for Multidisciplinary Studies, Belgrade, (2003).
[9] Adams, R.N., “Electrochemistry At Solid Electrodes”, Marcel Dekker, Inc., 1969.
[10] Cotton, F.A., Wilkinson, G., Murillo, C.A., Bochmann, M., “Advanced Inorganic
Chemistry”, 6th Ed., Wiley-Interscience: New York 1999.
[11] Comba, P., Merbach, A., Inorg. Chem. 26 (1987) 1315.
89 4 Polarography
4.1. Polarographic Behaviour
Figure 4-1 shows a typical differential pulse polarogram resulting from the analysis of an
2+ aqueous acidic titanium-containing solution, in this case 0.1 M TiO in 0.1 M H2SO4.
5
4
3 A) µ
Cu2+
Current / ( 2
1
0 -1.3 -0.9 -0.5 -0.1 0.3 Voltage / (V verses SCE)
90 2+ Figure 4-1. Sample polarogram for the electrolyte 0.1 M TiO in 0.1 M H2SO4, at 21
± 1°C. Note the reduction process for Cu2+, which was added as an internal standard
(normalised).
Firstly, note that the peak at +0.01 V corresponds to the reduction of Cu2+ which was added as an internal standard for the measurements. The values obtained for the potentials for the Cu2+ peak maximum were used to normalise the curves as the mercury drop size was prone to differ in size for the measurements. The literature value for the polarographic reduction of copper(II) is
+0.05 V vs SCE [1] which is similar to the values obtained from these measurements.
At lower potentials there appear multiple peaks corresponding to the reduction of various aqueous titanium species, despite the fact that a single phase starting material was used. Since multiple peaks are present we can conclude initially that there are multiple redox processes occurring, either as a result of the continuing reduction of TiO2+ to Ti(III) and then perhaps to
Ti(II), or alternatively as a result of the reduction of different Ti(IV) species in the original solution.
To determine whether TiO2+ to Ti(III) reduction was occurring, the initial solution was placed into a standard H-cell in which the anode and cathode electrolytes were separated by a porous glass frit to avoid mixing, as shown in Figure 4-2.
91 Ag/AgCl Reference
Electrode Pt Counter Pb Working Electrode Electrode
Porous
Glass Frit
Magnetic Stirrer Bars
2+ Figure 4-2. H-cell used to electrolyse the 0.1 M TiO in 0.1 M H2SO4 electrolyte to increase its Ti3+ content, at 21 ± 1°C.
A cylindrical Pb electrode was used as the cathode/working electrode, while a Pt foil electrode was the anode/counter electrode. A Ag/AgCl reference electrode was also included in the cathode/working electrode compartment. Steady, even stirring was also used in both cell compartments to ensure adequate mixing. The catholyte was then electrolysed at a constant potential of -0.6 V versus the Ag/AgCl reference electrode for ~3 hours to effectively reduce a portion of the TiO2+ to Ti(III); i.e., to produce more of the potentially intermediate Ti(III) in the electrolyte. After this electrolysis period it was noted that the catholyte had changed colour from a clear solution to purple, which has been reported previously to be characteristic of a Ti(III) containing solution [2]. A sample of the catholyte was then taken and subjected to the same polarographic experiment as before, with the resultant polarogram shown in Figure 4-3. For other polarograms measured with an acid concentration greater than 0.3 M H2SO4, we observe three
92 peaks, whereas this polarogram resembles the polarogram seen in Figure 4-1. However, the first peak after the copper(II) reduction is smaller in magnitude, whereas the second peak after the copper(II) reduction is much larger in magnitude.
5 (2)
4
3 A) µ
Cu2+ (1)
Current / ( 2
1
0 -1.3 -0.9 -0.5 -0.1 0.3 Potential / (V verses SCE)
4+ 3+ Figure 4-3. Polarogram for the electrolyte 0.1 M Ti reduced to Ti in 1 M H2SO4.
Note the reduction process for Cu2+, which was added as an internal standard
(normalised).
93 This indicates that the multiple processes observed in the original polarogram are indeed due to the reduction of different Ti(IV) species, rather than its sequential reduction to lower valent Ti species. The question then becomes what is the origin of the multiple reduction processes? As has been discussed previously, there is evidence in the literature to suggest that the titanyl ion (TiO2+) forms oligomers in sulfuric acid media as a result of sulfate ion bridging between the individual Ti(IV) units. Therefore, it is possible that the multiple peaks observed in the polarogram are due to the reduction of different size oligomers.
A comment needs to be made at this point concerning the appearance of the reduction peaks in the polarogram, in particular the width of the peaks. When the peaks corresponding to the reduction of Ti(IV) are compared with that for the reduction of the internal standard Cu2+, it is clear that the Ti(IV) reduction peaks are substantially broader. Given that differential pulse polarography (DPP) produces a result that is essentially the derivative of a sampled DC polarogram, we can draw some conclusions as to the possible cause of the DPP peak broadening.
Firstly, it may be due to slow reaction kinetics for the reduction of Ti(IV). It is quite convenient for polarographic analysis, in particular identifying active species in a sample via their E1/2 values, that many reduction reactions on the Hg surface behave reversibly, essentially exhibiting very fast reduction kinetics or a large standard rate constant. However, if the rate constant is not so large then the system is expected to be quasi-reversible, which qualitatively would have the effect of spreading the potential range over which a sampled DC polarogram would be observed, or broadening a DPP peak, as shown in Figure 4-4.
94 Sampled DC Current Current (abitrary units) (abitrary Current
DPP Current
-0.8 -0.6 -0.4 -0.2 0 PotentialVoltage
Figure 4-4. An interpretation of how reduction kinetics can affect peak broadening in a differential pulse polarography experiment.
The cause of slower reduction kinetics may be the fact that we have oligomeric Ti(IV) species in the electrolyte, for which the reduction of some or all Ti(IV) monomer units is kinetically slower because of the need for charge distribution throughout the oligomeric unit.
95 An alternative explanation is that the DPP current peak we observe is actually composed of multiple overlapping processes, that when combined, give rise to a broader peak, as shown schematically in Figure 4.5.
Current (arbitrary units) (arbitrary Current
-0.8 -0.6 -0.4 -0.2 0 VoltagePotential
Figure 4-5. An interpretation of how multiple peaks can affect the overall peak in a differential pulse polarography experiment.
96 This is of particular relevance to this work since there is evidence in the literature, and also in Figure 4-1, to suggest that there are multiple oligomers of Ti(IV) containing species present. If this were the explanation, then rather than a constant integral number of monomers combining to produce a fixed size oligomer, there is a distribution around a mean oligomer size, and so the two
Ti(IV)-related peaks in Figure 4-1 reflect two rather localized distributions of oligomer sizes.
Being able to differentiate between these two options is not possible at this stage.
4.2. Effect of H2SO4 Concentration
Figure 4-6 contains polarograms conducted on a range of solutions containing 0.1 M
2+ TiO with different H2SO4 concentrations ranging from 0.01 to 0.5 M. As the concentration of
H2SO4 increased, particularly for 0.3 M H2SO4 and higher, the third peak at lower potentials becomes definitely more pronounced relative to lower H2SO4 concentration solutions.
97 Cu2+
0.01 M H2SO4 A)
µ 0.1 M H2SO4 Current / (
0.3 M H2SO4
1 µA
0.5 M H2SO4
-1 -0.8 -0.6 -0.4 -0.2 0 Potential / (V verses SCE)
Figure 4-6. The effect of H2SO4 concentration on the polarographic behaviour of 0.1
M TiO2+ at 21 ± 1°C.
98 An explanation for the emergence of this peak may revolve around the extent to which sulfate ions are available to form bridging species (oligomers) with the Ti(IV); i.e.,
2+ 2- 2-2n m(TiO + nSO4 ↔ TiO(SO4)n ) (4.1)
where n represents the number of sulfate ligands coordinated to each TiO2+ ion, and m is the number of monomer units linked together to form the oligomer. In lower H2SO4 concentration electrolytes it appears that there is only sufficient sulfate present to form two main oligomeric species, what we might expect to be relatively small oligomers since there is only limited sulfate present to coordinate with TiO2+; i.e., m in Eqn (4.1) is small. However, with more sulfate added
2-2n there is a greater driving force to shift the equilibrium in Eqn (4.1) to produce more TiO(SO4)n units, and hence allow for the formation of larger oligomer units. Furthermore, there is the potential to actually increase the number of coordinated sulfate ligands (n increases) to produce a fundamentally different oligomer unit.
4.3. Effect of Electrolyte Age
Figure 4-7 shows the polarographic response obtained from a solution of 0.1 M TiO2+ in
0.3 M H2SO4 recorded periodically over an extended period of time (up to 10 days). Initially the behaviour of the solution remains quite constant, with very little change apparent up to ~20 hours, except for a slight shift towards more negative potentials. However, after 10 days of solution aging there is a dramatic difference in the resultant polarogram, with the two higher potential peaks dropping dramatically in current, while the lowest potential peak increasing considerably in current. Assuming that after this time these peaks still correspond to the same Ti(IV) species, then what is occurring is a drop in the concentration of the higher potential oligomeric units, and an increase in the concentration of the lowest potential oligomer.
99
4.5
4 0 Hours A) µ 3.5 Current / (
3 10 Days
2.5
2 -1 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 Potential / (V verses SCE)
Figure 4-7. The effect of electrolyte age on the polarographic behaviour of a solution
2+ containing 0.1 M TiO in 0.3 M H2SO4, at 21 ± 1°C.
To explain this behaviour we have noted that aqueous acidic solutions of TiO2+ are only metastable, hydrolysing with time to form a suspension of TiO2. Of course the rate of hydrolysis does depend on the composition of the electrolyte, with low H2SO4 concentration solutions hydrolysing relatively quickly (~24 hours) compared to the more concentrated electrolytes, which
100 are stable almost indefinitely. This latter observation on the effects of H2SO4 concentration on hydrolysis, when combined with our previous data, does suggest that the bridging sulfate ligands contribute to solution stabilization. On a molecular level, for hydrolysis to occur, individual TiO2+ units must first undergo some form of hydroxylation to form a species that can then condense with other similar units via oxolation and/or olation to form a progressively increasing 3D network, that eventually can be considered a particle when it reaches a suitable size. A possible reaction scheme for this could be:
2+ + + Hydroxylation: TiO(OH2)5 ↔ TiO(OH2)4(OH) + H (4.2)
+ 2+ Oxolation: 2TiO(OH2)4(OH) ↔ [(H2O)4OTi-O-TiO(OH2)4] + H2O (4.3)
Olation:
+ + (4.4) OH HO2 O (H O) OTi + TiO( OH ) ↔ (H O) OTi TiO( OH ) ++H O 2H 2 3 2 3 22332 OH2 HO O
The role that sulfate plays is perhaps not so clear. It apparently does function as a bridging ligand between TiO2+ units, and by analogy with the hydroxylation, oxolation and olation mechanisms shown in Eqns (4.2)-(4.4), we might suspect that with time the size of the oligomer should increase, or alternatively, the concentration of a larger oligomer should increase.
In other words, after a longer time the equilibrium in Eqn (4.1) shifts further to the product side.
Quantitatively, the change in DPP peak areas can be shown to vary with time as shown in
Figure 4-8. As was mentioned in the Experimental section, a Lorentzian lineshape was fit to the experimental DPP data, and then by numerical integration, the area of each peak was determined.
This was then plotted in Figure 4-8 as an area ratio relative to the area of the Cu2+ internal standard peak, as a function of time.
101 10
Peak 3
Peak 1
1 Log Area Ratio Log Area
Peak 2
0.1 0 50 100 150 200 250 Time / (Hrs)
Figure 4-8. Variation in peak area ratio (Ti peak relative to the Cu2+ internal standard) as a function of time.
The real interest is after 10 days (240 hours). The first two peaks are seen to drop dramatically whereas the third peak increases. Over time, the solution is hydrolysing as well and the observation of electro-active species diminishing, whilst the solution stays the same. These electroactive species first formed, must be less stable over time and could be small oligomers. As time increases, these species attach themselves onto longer chains to make larger oligomers.
102
4.4. Effect of Different Acid Media
4.4.1. Backgrounds
Figure 4-9 shows the polarograms of the solutions of the various acids used at a concentration of 1 M. Copper was still added to each solution as the internal standard and this was used to normalise the data.
4.4.2. Different Supporting Electrolytes
As can be seen in Figure 4-9, the polarogram of sulfuric acid does not produce any electrochemically active species until we reach hydrogen evolution at potentials lower than -1.3
V.
The polarogram for perchloric acid (see Figure 4-9) shows no peaks corresponding to any electrochemically active species until we reach hydrogen evolution at lower potentials. However, the copper peak is slightly broader at the bottom for perchloric acid than it is for sulfuric acid, and the hydrogen evolution peak is a bit steeper.
The polarogram for acetic acid (see Figure 4-9) has another reduction peak immediately after the reduction of copper. This species also reduces as well and then no other electrochemically active species is present. Not even hydrogen evolution is present yet on this polarogram.
103 Sulfuric Acid
Perchloric Acid A) µ
Acetic Acid Current / (
Hydrochloric Acid
1 µA Nitric Acid
-1.3 -1.1 -0.9 -0.7 -0.5 -0.3 -0.1 0.1 Potential / (V verses SCE)
Figure 4-9. Polarograms of the base lines of the different acids used, 1 M
(normalised to copper peak).
104 The polarogram for hydrochloric acid (see Figure 4-9) has a copper(II) reduction peak which is broader than the same peak observed in sulfuric acid. There also appears to be a reduction occurring straight after the copper peak. Again hydrogen evolution is seen to occur at the lower potentials.
The polarogram for nitric acid (see Figure 4-9) has a well defined reduction peak for copper(II) and then it is reduced very quickly again at -0.9 V. This could be due to the nitrogen reduction as it has been reported that nitric acid as a media for polarographic measurements is detrimental as it can dissolve the mercury to form nitrate salts at low concentrations.
4.4.3. Titanium Metal
The concentration of titanium in each acid was calculated to be; 26.11 mM (all dissolved) in H2SO4 ; 1.34 mM in HClO4; 0.13 mM in CH3COOH; 25.6 mM (all dissolved) in HCl; and 1.57 mM in HNO3. Due to the low titanium concentration in acetic acid the resulting polarogram cannot easily be considered to be due to titanium reduction so will not be discussed in detail.
Figure 4-10 shows the polarograms resulting from dissolving titanium metal into various acidic media. The copper peak has been left off as it is not needed, but the polarograms have still been normalised to the copper peak.
105 Sulfuric Acid
Perchloric A id
Acetic Acid A) µ Current / (
0.2 µA
Hydrochloric Acid
Nitric Acid
-1.3 -1.1 -0.9 -0.7 -0.5 -0.3 -0.1 PotentialVoltage / (V verses SCE)
Figure 4-10. Polarograms of titanium metal dissolved in different acidic media.
106 Titanium metal dissolved completely in sulfuric acid and produced three peaks for its polarogram. This fits with our previous observations that in a sulfuric acid concentration greater than 0.3 M, we observe three reduction peaks. This is also further proof that the presence of sulfate ions produces different species, other than just the reduction of Ti(IV) to Ti(III).
The polarogram of titanium metal in perchloric acid produces one peak at -0.82 V. This could indicate that this is the reduction of Ti(IV) to Ti(III) as the polarogram of just perchloric acid had no peak at this potential. There is a reported half-reaction (Eqn. 4-5) [3], which is what could be happening.
2+ + - 0 TiO + 2H + 4e → Ti + H2O E = -0.882 V (4.5)
The polarogram of titanium metal dissolved in hydrochloric acid (it all dissolved, and produced a white precipitate when heated for too long) shows two peaks. We could propose that, the first peak at -0.24 V is due to Ti(IV) reduction and that the peak at -0.9 V is due to the reduction of Ti(III).
The polarogram for titanium metal dissolved in nitric acid shows a massive current generated starting at -0.5 V and quickly exceeding the graphs scale. The same large peak at -0.9
V is seen and as mentioned above, this could be due to the reduction of nitrate. The appearance of the peak at -0.5 V could be that of Ti(IV) reduction.
4.4.4. Titanyl oxysulfate Addition to Various Acid Media
Figure 4-11 shows the polarograms of titanyl-oxysulfate dissolved in sulfuric, perchloric, acetic, hydrochloric and nitric acids. Again copper is used as an internal standard, and all the polarograms are normalised to the copper(II) reduction peak.
107 As mentioned previously, we can see that in sulfuric acid, titanyl oxysulfate produces three peaks for a 1 M solution.
The polarogram for titanyl oxysulfate dissolved in perchloric acid shows a peak at -0.7 V with a shoulder at -0.95 V. This is a different peak potential from the peak potential obtained from just titanium metal. This different potential and the appearance of the shoulder leads us to believe that even at this low concentration of titanyl oxysulfate (0.1 M TiO2+), the sulfate ions begin to form bridging species.
The polarogram for titanyl oxysulfate dissolved in acetic acid produces a peak at -0.8 V which looks to be similar in shape to the peak observed from the perchloric acid media, but not has big. Again, is this due to the sulfate ions forming sulfate bridged species.
The polarogram for titanyl oxysulfate dissolved in hydrochloric acid shows two peaks at -
0.35 V and -0.55 V. These two peaks occur at completely different potentials to the peak observed when titanium metal was dissolved. This indicates that either, another species has been formed or the dissociation of oligomers has occurred.
108 Cu2+ Sulfuric Acid
Perchloric Acid A) µ
Acetic Acid Current / (
Hydrochloric Acid
1 µA
Nitric Acid
-1.3 -0.9 -0.5 -0.1 PotentialVoltage / (V vs SCE)
Figure 4-11. Polarograms of TiOSO4 dissolved in various acidic media.
109 The polarogram for titanyl oxysulfate dissolved in nitric acid shows two peaks at -0.49 V and -0.67 V and then the very large generated amount of current as seen in all the other nitric acid solutions. These two peaks, not seen when just titanium metal was dissolved, is another indication that the sulfate present in the titanyl oxysulfate is enough to form some kind of bridged species.
4.5. Conclusions
When titanium is dissolved in acid, specifically sulfuric acid, the presence of the sulfate ion has been found to have two major impacts on the species so formed in solution. Firstly, it was demonstrated that during polarographic measurements, the reduction of different Ti(IV) species was occurring rather than lower valent Ti(III) species reductions. Secondly, it was shown that as time elapses the sulfate ions do in fact act as ligands and form oligomers with the titanium.
However, full analysis of the polarograms is not possible yet to determine whether these curves are due to multiple overlapping or kinetics.
Use of different acid mediums with titanium metal and TiOSO4 favourably indicates that the sulfate ions do form oligomers by bridging species.
4.6. References
[1] Reynolds, G.F., Shalgosky, H.I., Webber, T. J., Analytica Chimica Acta, 9 (1953) 91.
[2] Cotton, F.A., Wilkinson, G. “Advanced Inorganic Chemistry”, 5th Ed., Wiley-
Interscience: New York 1988.
[3] Bard, A.J., “Encyclopedia of Electrochemistry of the Elements”, Vol. 5, Marcel Dekker,
Inc., 1976.
110
111 5 Sulfate Process Implementation
5.1. Overview of Electrochemical Techniques
Cyclic voltammetry, EIS, RDE and Polarographic measurements were successfully employed in the analysis of the electroactivity of the reduction of Ti(IV) to Ti(III). There are, in fact, multiple Ti(IV) species present when in sulfuric acidic solutions. The electrochemical activity of the titanium species so formed is dependent on the initial titanium concentration and the solution age. The reaction is more facile with more TiO2+ in the solution, so the more TiO2+ there is the more electrochemically active the species are.
The most reproducible result from these electrochemical methods, is the deviation seen for the 0.09 M TiO2+ solution. In every technique used, the 0.09 M TiO2+ solution produced a result not fitting to the rest of the data obtained.
5.2. Implementation
The implementation of electrochemically reducing Ti(IV) to Ti(III) in the sulfate process would have to be implemented into the process in the purification stage, Figure 5.1. However, as witnessed by the different electrochemical methods, TiO2+ in sulfuric acid media behaves differently than in other media. Therefore limitations would need to be set on the process to ensure optimal Ti(IV) reduction.
Limitation 1: A Ti(IV) concentration between 0.06 – 0.09 M would ensure that the third species observed in a more concentrated Ti(IV) solution, would not form.
Limitation 2: A sulfuric acid concentration that is not too high (<1 M) as there was not a big difference between the 1 M and 5 M H2SO4 solutions and kinetic properties. However from
112 diffusion coefficients, we learnt that having sulfuric acid as the supporting electrolyte does make a difference, so again a low acid concentration would be beneficial.
Limitation 3: Solutions which are no older than 15 hours would provide the best results because again, after this time, three different species are observed to form, slowing kinetics down.
Limitation 4: Applying a potential of -0.75 V would ensure that any Ti(IV) or Ti(III) intermediate species formed would not reduce at this potential.
113 Ti(IV)
ElectrochemicalScrap Fe
Ilmenite Digestion Porous Purification FeSO4⋅7H2O
(FeTiO3) (~100°C) Cake and Filtration (Copperas)
H2SO4 H2SO4 Black (85-92%) Regeneration Liquor
Hydrolysis SO2 OH- (~100°C/hours)
Post Calcination
Production TiO2 TiO ⋅nH O (~1000°C) 2 2 (e.g., coating)
Figure 5-1. Schematic of the SulfateScrubbers Process with Ti(IV) SOreduction.3 H2O
114 5.3. Industrial Design
There is a patent that describes a liquid sample flow through analysis cell [1]. However this is designed for low volume liquid samples. Considering, that this kind of apparatus exists, modifications could be made to this cell to try and reduce Ti(IV) to Ti(III) electrochemically. The biggest problem faced by this process is the perceptibly of Ti(III) to be oxidised by dissolved oxygen in solution. So not only would a flow through cell be needed to reduce the titanium, it will also need to be conducted under nitrogen gas.
From the limitations proposed, some properties that the flow through cell would need to contain would be:
• a high cathode area, for example, using a mesh lead cathode
• using a membrane to separate the anode
• a short residence time
• if the system is not used continually, flushing the lines with water.
5.4. References
[1] Davidson, R.A., Walker, E.B., US Patent No. 5,574,232 (1996).
115 6 Appendix A
The reactions explaining the bold lines from Figure 1-2 indicating the boundaries between two solid species (8, 9 and 11), or one solid species and one dissolved species (13, 17, 18, 19 and 20).
+ - 8. Ti + H2O = TiO + 2H + 2e E0 = -1.306 – 0.0591 pH
+ - 9. 2TiO + H2O = Ti2O3 + 2H + 2e E0 = -1.123 – 0.0591 pH
= -0.894 – 0.0591 pH
+ - 11. Ti2O3 + H2O = 2TiO2 + 2H + 2e E0 = -0.556 – 0.0591 pH
= -0.139 – 0.0591 pH
= -0.786 – 0.0591 pH
= -0.091 – 0.0591 pH
2+ + 2+ 13. Ti + H2O = TiO + 2H log(Ti ) = 10.91 – 2 pH
2+ - 2+ 17. Ti = Ti + 2e E0 = -1.630 + 0.0295 log(Ti )
2+ + - 2+ 18. 2Ti + 3H2O = Ti2O3 + 6H + 2e E0 = -0.478 - 0.1773 pH – 0.0591 log(Ti )
= -0.248 - 0.1773 pH – 0.0591 log(Ti2+)
2+ + - 2+ 19. Ti + 2H2O = TiO2 + 4H + 2e E0 = -0.502 - 0.1182 pH – 0.0295 log(Ti )
= -0.169 - 0.1182 pH – 0.0295 log(Ti2+)
116
3+ + - 3+ 20. Ti + 2H2O = TiO2 + 4H + e E0 = -0.666 - 0.2364 pH – 0.0591 log(Ti )
= -0.029 - 0.2364 pH – 0.0591 log(Ti3+)
117 7 Appendix B
0.03 M TiO2+
0.06 M TiO2+ ) -2
0.09 M TiO2+ Current / (Acm
0.13 M TiO2+
0.005 A.cm-2
-1 -0.8 -0.6 -0.4 Potential / (V vs SCE)
118 Figure 7-1. Cathodic sweep only from cyclic voltammetry of a solution containing
2+ varying [TiO ] and 0.01 M H2SO4, with a scan rate of 5 mV/s at 21 ± 1°C.
0.0 M TiO2+
0.03 M TiO2+ ) -2
0.06 M TiO2+ Current / (A.cm 0.09 M TiO2+
0.13 M TiO2+
0.005 A.cm-2
-1 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 Potential / (V vs SCE)
119
Figure 7-2 Cathodic sweep only from cyclic voltammetry of a solution containing
2+ varying [TiO ] and 1 M H2SO4, with a scan rate of 5 mV/s at 21 ± 1°C.
120 0.0 M TiO2+
0.03 M TiO2+
0.06 M TiO2+
2+
) 0.09 M TiO -2
0.13 M TiO2+ Current / (Acm
0.005 Acm-2
-1 -0.8 -0.6 Potential / (V vs SCE)
Figure 7-3 Cathodic sweep only from cyclic voltammetry of a solution containing
2+ varying [TiO ] and 5 M H2SO4, with a scan rate of 5 mV/s at 21 ± 1°C.
121 1800 0.0 M TiO2+ 1600
1400
1200 0.03 M TiO2+
) 1000 Ω / ( CT
R 800 0.06 M TiO2+
600
0.09 M TiO2+ 0.13 M TiO2+ 400
200
0 -1.3 -1.2 -1.1 -1 -0.9 -0.8 -0.7 -0.6 -0.5 Potential / (V vs SCE)
Figure 7-4 Charge transfer resistance (RCT) versus voltage for solutions containing
2+ varying [TiO ] in 0.1 M H2SO4, from EIS measurements at 21 ± 1°C.
122 2500
2000
0.0 M TiO2+
0.03 M TiO2+ 1500 ) Ω
/ ( 0.06 M TiO2+ CT R 1000
0.09 M TiO2+
500 0.13 M TiO2+
0 -1.5 -1.3 -1.1 -0.9 -0.7 -0.5 Potential / (V vs SCE)
Figure 7-5 Charge transfer resistance (RCT) versus voltage for solutions containing
2+ varying [TiO ] in 0.01 M H2SO4, from EIS measurements at 21 ± 1°C.
123 600
0.06 M TiO2+ 500
400 0.0 M TiO2+
0.03 M TiO2+ ) Ω
/ ( 300 CT R
200
0.09 M TiO2+ 100
0.13 M TiO2+
0 -1 -0.9 -0.8 -0.7 -0.6 -0.5 Potential / (V vs SCE)
Figure 7-6 Charge transfer resistance (RCT) versus voltage for solutions containing
2+ varying [TiO ] in 5 M H2SO4, from EIS measurements at 21 ± 1°C.
124 7
0.13 M TiO2+ 6 0.03 M TiO2+
5 ) Ω 4 2+ / ( 0.0 M TiO S R
3 2+ 0.06 M TiO 2+ 0.09 M TiO
2
1 -1.25 -1.15 -1.05 -0.95 -0.85 -0.75 -0.65 -0.55 Potential / (V vs SVE)
2+ Figure 7-7 Series resistance (RS) versus voltage for solutions with varying [TiO ] in
0.1 M H2SO4, from EIS measurements at 21 ± 1°C.
125 60
50 0.0 M TiO2+
40
) 0.03 M TiO2+ Ω 30 / ( S R 0.09 M TiO2+
20
0.06 M TiO2+
10 0.13 M TiO2+
0 -1.3 -1.2 -1.1 -1 -0.9 -0.8 -0.7 -0.6 -0.5 Potential / (V vs SCE)
2+ Figure 7-8 Series resistance (RS) versus voltage for solutions with varying [TiO ] in
0.01 M H2SO4, from EIS measurements at 21 ± 1°C.
126 5
0.13 M TiO2+ 4.5
4
3.5
3 ) Ω 2.5 / ( S R
2
1.5 0.0 M TiO2+ 0.09 M TiO2+ 1 0.06 M TiO2+ 0.5
0.03 M TiO2+ 0 -1.1 -1 -0.9 -0.8 -0.7 -0.6 -0.5 Potential / (V vs SCE)
2+ Figure 7-9 Series resistance (RS) versus voltage for solutions with varying [TiO ] in
5 M H2SO4, from EIS measurements at 21 ± 1°C.
127 0.00025 0.0 M TiO2+
2+ 0.0002 0.09 M TiO
0.13 M TiO2+ 0.00015 ) -2 / (Fcm dl C 0.0001
0.00005
0.06 M TiO2+ 0.03 M TiO2+
0 -1.25 -1.15 -1.05 -0.95 -0.85 -0.75 -0.65 -0.55 Potential / (V vs SCE)
Figure 7-10 Double layer capacitance (Cdl) versus voltage for solutions with varying
2+ [TiO ] in 0.1 M H2SO4 calculated from EIS measurements, at 21 ± 1°C.
128 0.00018
0.00016
0.03 M TiO2+ 0.00014
0.00012 ) -2 0.0001 0.0 M TiO2+
0.13 M TiO2+ / (Fcm
dl 0.00008 C
0.00006
0.00004 0.06 M TiO2+
0.00002 0.09 M TiO2+ 0 -1.5 -1.3 -1.1 -0.9 -0.7 -0.5 Potential / (V vs SCE)
Figure 7-11 Double layer capacitance (Cdl) versus voltage for solutions with varying
2+ [TiO ] in 0.01 M H2SO4 calculated from EIS measurements, at 21 ± 1°C.
129 0.01
0.03 M TiO2+
0.06 M TiO2+
0.09 M TiO2+ 0.001 0.0 M TiO2+ ) -2 / (Fcm dl C
0.0001
0.13 M TiO2+
0.00001 -1.1 -1 -0.9 -0.8 -0.7 -0.6 -0.5 Potential / (V vs SCE)
Figure 7-12 Double layer capacitance (Cdl) versus voltage for solutions with varying
2+ [TiO ] in 5 M H2SO4, calculated from EIS measurements, at 21 ± 1°C. A log scale has been used to see the relationship clearer.
130 8 Appendix C
Cdl calculations were made using the equation below;
1 − σ m = Cdl (m−1) R S
131