Hydrometallurgy 163 (2016) 198–203

Contents lists available at ScienceDirect

Hydrometallurgy

journal homepage: www.elsevier.com/locate/hydromet

The digestion of New Zealand ilmenite by hydrochloric acid

Richard G. Haverkamp ⁎, Desiderius Kruger 1, Ranjeeth Rajashekar 2
School of Engineering and Advanced Technology, Massey University, Private Bag, 11222, Palmerston North 4442, New Zealand

a r t i c l e i n f o a b s t r a c t

Article history: Ilmenite (FeTiO3) is an important source of titanium, for which the main application is TiO2 pigment. One route to
Received 15 September 2015 the production of TiO2 from ilmenite may be the digestion of both the Fe and Ti components in hydrochloric acid
Received in revised form 18 April 2016 and the subsequent direct hydrolysis of the Ti. The process requires that the Ti be solubilized. A kinetic study is
Accepted 24 April 2016
performed of the leaching of New Zealand (Barrytown) placer ilmenite (ground to a 10 μm median particle diam-
Available online 10 May 2016
eter) by hydrochloric acid in a stirred reactor. Around 90% extraction of both Fe and Ti is achieved with a 32% w/w
Keywords:
HCl concentration, with a 2:1 w/w ratio of HCl to FeTiO3 over the temperature range 60 to 90 °C. The shrinking
Ilmenite sphere model is a good representation of the kinetics if particle size distribution is accounted for. The reaction
Titanium dioxide rate is estimated to be 4th order in HCl concentration. The temperature rise due to an exothermic reaction is
Hydrochloric acid calculated to be c. 41 °C. Undesirable, in situ hydrolysis of Ti during digestion can be limited by control of temper-
Leach ature, reaction time and acid concentration. The data and models presented may facilitate the preliminary design
Digestion, kinetics of an industrial process to digest ilmenite to dissolve and retain Ti in solution.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction with world production capacity at 6.5 million tonnes in 2012

(U.S.GeologicalSurvey, 2013). Titanium metal sponge production
Extensive deposits of ilmenite occur as placer deposits (beach sands) has also grown steadily to 214 thousand tonnes in 2012
on the West Coast of the South Island of New Zealand. The size of (U.S.GeologicalSurvey, 2013) from 51 thousand tonnes in 1997, and is
the resource has been extensively surveyed, with the Barrytown expected to continue to grow with the increasing use of the metal in
old beach sands deposit near Westport containing approximately the aerospace and other industries.
6.9 million tonnes of ilmenite, found at an average concentration of For the production of TiO2 via the chloride route (with TiCl4 as an
13.8%. Nearby, at Carters Beach, surveys found 6.5–9.6 million tonnes intermediate), the dominant feedstock is a high-grade rutile or synthet-
at 7.6–15% (Christie and Brathwaite, 1998). The source of the ilmenite ic rutile feedstock. The upgrading of ilmenite to synthetic rutile
is the metamorphic rocks of the Southern Alps (Grapes and Watanabe, containing N85% TiO2 is commonly performed with hydrochloric acid
1992). Associated with the ilmenite are the potentially valuable acces- to leach out the Fe, usually after an oxidation and reduction step, leaving
sory minerals gold, monazite and zircon. There have been several the Ti (Gireesh et al., 2015; Girgin, 1990; Sinha, 1984; Jackson and
attempts in the past to develop the resource, including extensive work Wadsworth, 1976; Jain and Jena, 1977). The reactivity of ilmenite varies
in the 1980's on processing technologies (Mann and James, 1989). The with the mineral's source. Most ilmenites are relatively insoluble in HCl
Barrytown ilmenite deposit is not currently being exploited but there and must first undergo an oxidation and a reduction step (Ogasawara
has been recent commercial interest in developing the resource (Mills, and de Araujo, 2000; Wahyuningsih et al., 2013); a few are relatively
2013). soluble directly in HCl (Olanipekun, 1999).
Ilmenite supplies about 90% of the world's demand for titanium (Ti) The digestion of ilmenite by HCl may be investigated with the view
minerals. World Ti resources are estimated to be N 2 billion tonnes, with to producing a pigment-grade TiO2 directly from the Ti- and Fe-
world production of ilmenite at 6.2 million tonnes and rutile ore at containing digestion solution. Since the Ti needs to be in a soluble
7.0 million tonnes in 2012 (U.S.GeologicalSurvey, 2013). There is a form, and its subsequent hydrolysis must occur in a controlled fashion
growing demand for TiO2 for use as a pigment and for Ti metal for use (i.e. no significant hydrolysis can have occurred already), the conditions
in alloys. To meet the demand, TiO2 pigment production is increasing, required for digestion are different from those normally.
Various factors affecting ilmenite digestion in HCl have been studied
(Duncan and Metson, 1982a; Sasikumar et al., 2004; Sarker et al., 2006;
⁎ Corresponding author. Sasikumar et al., 2007; Janssen and Putnis, 2011). Where acid concen-
E-mail address: [email protected] (R.G. Haverkamp).
1
Present addresses: Corrosion Control Engineering, PO Box 330, Bulimba QLD 4171,
trations are too low, or temperatures too high, the already-hydrolyzed
Australia. Ti can accumulate and block the pores that should be formed during di-
2
Present addresses: No.1 Muniga Layout Maruthi Sevangar, Bangalore 5600033, India. gestion, thereby preventing further leaching and inhibiting digestion.

http://dx.doi.org/10.1016/j.hydromet.2016.04.015
0304-386X/© 2016 Elsevier B.V. All rights reserved.
 R.G. Haverkamp et al. / Hydrometallurgy 163 (2016) 198–203 199

More specifically, leaching occurs selectively, leaving a lattice structure Table 1

of undigested material (Duncan and Metson, 1982b). Ilmenite composition.

Here, we describe the reactivity of New Zealand-sourced ilmenite in Oxide Weight (%)
concentrated HCl under conditions that might be suitable for industrial TiO2 47.1
production of TiO2, and the data are used to model the digestion FeO 37.5
process. Fe2O3 3.0
SiO2 5.2
Al2O3 2.3
2. Experimental section MnO 1.7
MgO 0.2
CaO 1.1
2.1. Ilmenite
Na2O 0.3
K2O 0.3
Ilmenite concentrate from Barrytown, Westland, New Zealand was P2O5 0.2
supplied by Rio Tinto Pty Ltd. The composition of the ilmenite was V2O5 0.1
determined using borate fusion and X-ray fluorescence spectrometry
by Spectrachem Analytical Services, Lower Hutt, New Zealand.
3.3. Digestion profiles
The ilmenite was then ground in small batches (35 g for 2 min) in a
150-mm disc mill (Tema, UK) for the digestion experiments. A Malvern
Slightly more Fe was digested than Ti (Fig. 2), reflecting the fact that
Mastersizer 3000 (Malvern Instruments, UK) was used to measure
the ilmenite does not have a stoichiometric ratio of Fe and Ti (i.e. the
particle size of the ground ilmenite.
elemental analysis does not correspond with FeTiO3 but rather to
Fe0·97Ti1.03O3), and that the slight excess of Ti may be present as the
2.2. Digestion procedure more insoluble rutile structure.
The ultimate decreases in proportion of Ti digested at 85 °C and 90 °C
The ground ilmenite was digested in 32% w/w HCl (8.8 mol kg−1 HCl were due to the hydrolysis and precipitation of titanium hydroxide,
or 10.3 M HCl at 20 °C) at temperatures of 60, 80, 85 and 90 °C. The ratio TiO(OH)2. It is essential that this phenomenon be avoided if the diges-
of acid to ilmenite was 2 to 1, calculated as kg 100% HCl to kg ilmenite in tion liquor is to be used for subsequent hydrolysis to TiO2 for several
ore by analysis, the latter calculated as equivalent ilmenite based on reasons. The TiO(OH)2.precipitate may block the pores or partially
Fe2O3 being treated as FeO and taking (mol Ti + mol Fe)/2. This coat the dissolving ilmenite, thereby inhibiting further digestion. More
amounts to 8.50 mol HCl to 1 mol “FeTiO3”, bearing in mind that the importantly, it is likely to become associated with gangue, such as fine
Ti-to-Fe ratio in the ilmenite is not 1:1. silicates, introducing impurities in the TiO2. Also this uncontrolled
Digestion of ground ilmenite in hot HCl took place in a 3 L round- nucleation of TiO2 reduces the degree of control possible over the final
bottom glass flask, fitted with a reflux condenser, and with a Teflon form of any TiO2 pigment produced.
paddle stirrer on a glass shaft. The HCl was preheated in the flask on a Digestion of over 80% of Ti, without hydrolysis, was achieved
heating element before ilmenite addition. The element was also used (concentration of about 81 g kg−1, expressed as weight of TiO2, as is
in the 90 °C experiment to provide rapid initial heating after ilmenite the normal convention) and Fe digestion reached 84% (a concentration
addition since the boiling point of 32% w/w HCl is 84 °C. of 80 g kg−1 as FeO). Had ilmenite particles of a narrower size distribu-
Samples of liquor were collected at timed intervals and rapidly fil- tion been used, such that none of the larger, slower-to-digest particles
tered through glass fiber paper while still hot. The Ti and Fe contents had been present, the percentages of digestion would have been higher.
were analyzed by reduction with Al metal followed by quantitative ox- One trial (data not shown), using more finely ground ilmenite and less
idation by titration with cerium sulfate solution using the indicators acid (therefore with and acid-to-ilmenite ratio of 1.44:1 w/w, yielded
methyl blue (for Ti) and N- phenylanthranilic acid (for Fe). The results over 90% digestion of both Ti and Fe, which therefore lead to higher
are presented as percentage dissolution, based on the total Ti and Fe concentrations of Ti and Fe in solution (up to 87 g kg−1 TiO2).
contents of the added ilmenite and the Ti and Fe contents of the solution Complete digestion of the Ti component will not be possible, in part
adjusted for solution mass increases due to dissolution and removed because some Ti is present as rutile inclusions (Duncan and Metson,
sample volumes. 1982b), and rutile is not soluble in HCl.

3. Results and discussion

3.4. Acid concentration
3.1. Ilmenite composition
The digestion proceeds according the stoichiometric reaction:
The elemental composition of the Barrytown ilmenite, including
FeTiO3 + 4HCl = Fe2+ + TiO2+ + 2H2O + 4Cl−
inclusions, is shown in Table 1. It has been shown previously that the
ilmenite phase contains most of a sample's Ti, Fe and Mn (Duncan and
with 4 mol HCl for each mol of ilmenite. The reaction does not proceed
Metson, 1982b). The inclusions consist mainly of silicate minerals such
to Ti4+ ions in solution because Ti, under these conditions, is the TiO2+
as quartz and feldspars, and these are the repository of most of the
complex, most likely as TiOCl4 − (Cservenyak et al., 1996; Kelsall and
oxides listed in Table 1 and are resistant to dissolution in HCl. In
Robbins, 1990; Sole, 1999). This reaction has been carried out here in
addition, apatite is present as inclusions and is soluble in HCl (Duncan
a way that might have direct relevance to an industrial process, without
and Metson, 1982b).
a huge excess of HCl. Therefore the concentration of HCl declines signif-
icantly with reaction.
3.2. Particle size distribution The 2:1 w/w ratio of HCl:TiO2 used here represents 8.3 mol of HCl for
each mole of FeTiO3, which is an excess of 4.3 mol of HCl per mol of
Being from a placer deposit, the mined ilmenite had a narrow size FeTiO3. The calculated approximate HCl concentrations are shown in
distribution, centered of around 150–250 μm (Fig. 1a). Grinding result- Fig. 3, with HCl concentration changing from 32% w/w initially to around
ed in the particle size distribution shown in Fig. 1b, with D50 = 10.0 μm, 15% w/w when digestion peaked. If the digested Ti is hydrolyzed during
D10 = 1.3 μm and D90 = 76 μm. the reaction (an undesirable situation if subsequent direct pigment
200 R.G. Haverkamp et al. / Hydrometallurgy 163 (2016) 198–203

Fig. 1. Particle size distribution of a) ilmenite as mined; b) ground ilmenite used for digestion trials.

production is the aim), then the concentration of HCl is increased where t is time, k is the rate constant and a is the extent of reaction
through the reaction TiO2+ + H2O = TiO(OH)2 + 2H+, and this can (i.e. 0 = no reaction, 1 = complete digestion).
be seen for the 90 °C curve. As mentioned in Section 2.2, the boiling An alternative model, the shrinking sphere model, where the rate
point of 32% w/w HCl is 84 °C, so that for the 90 °C digestion in a non- also depends on the exposed surface area of the particles (which
pressurised vessel some heating must occur in the first few minutes decreases with time) and but may be due to one, or a combination, of
(at 30% w/w HCl the boiling point is 90 °C) but this is aided by the the diffusion to (and from) the surface and chemical reactions at the
exothermic reaction. surface (Eq. (2)), was also considered in the Nigerian ore analysis but
rejected as a poor fit to the data:

3.5. Kinetic analysis kt ¼ 1−ð1−aÞ1=3 ð2Þ

For a model to adequately represent the digestion of ilmenite in HCl, The shrinking core model fits the digestion data presented here well
many factors must be taken into account, not all of which are well for the 60 °C and 80 °C curves, and for the 85 °C and 90 °C, curves the
known or readily included. Here, the aim is not to provide an exhaustive onset of Ti hydrolysis is very pronounced (Fig. 4). Such plots could be
analysis of all the factors but rather to consider enough of them to allow valuable tools to indicate the onset of hydrolysis.
development of an industrially useful model for a limited range of By comparsion, a plot of the shrinking particle model fits the data
conditions. poorly (Fig. 5).
The hard rock ilmenite from the Jos Plateau in Nigerian is amenable However, there are some significant flaws with this model compar-
to direct HCl digestion and the kinetics of this reaction have been stud- ison. The observed distribution of particle size (Fig. 2) would be expect-
ied (Olanipekun, 1999). The standard shrinking core model (Levenspiel, ed to steepen the initial slope as small particles react more rapidly
1972), where diffusion through a reacted layer of increasing thickness is (owing to their higher surface-to-volume ratio) than the larger
the main determinant of reaction rate (represented by Eq. (1)), was particles. This has been modeled in detail for typical particle size distri-
found to be a good fit to the Nigerian experimental data: butions, and where the shrinking core model might have the best fit,
once the particle size distribution is taken into account, the more

kt ¼ 1 þ 2ð1−aÞ−3ð1−aÞ2=3 ð1Þ

Fig. 3. Approximate HCl concentration during digestion (not corrected for all aspects of
density changes, but accounting for the minor Fe3+ component), corresponding to the
Fig. 2. Digestion of Ti and Fe in ground ilmenite. Solid lines Ti, dashed lines Fe. Green digestion curves in Fig. 2. Green (circle) 60 °C, red (triangle) 80 °C, blue (square) 85 °C,
(circle) 60 °C, red (triangle) 80 °C, blue (square) 85 °C, black (diamond) 90 °C. black (diamond) 90 °C.
 R.G. Haverkamp et al. / Hydrometallurgy 163 (2016) 198–203 201

Fig. 6. Digestion curve for ilmenite at 85 °C for Ti (points) with the fitted curves
Fig. 4. Ilmenite digestion curves for Ti plotted to test the shrinking core model. Green
representing the particle size distribution applied in the shrinking sphere model
(circle) 60 °C, red (triangle) 80 °C, blue (square) 85 °C, black (diamond) 90 °C. Solid
(dashed line) and with a single particle size (solid line). Inset: particle size distribution
lines are linear fits (not using final two points for 85 °C, not using final four points for
measured (solid) and used (points connected by dashed line) for the analysis.
90 °C); dashed lines are a guide to the eye.

model throughout but should transition from one mechanism to the

other.

physically likely shrinking sphere model can be seen to apply (Gbor and
3.6. Temperature dependence of digestion
Jia, 2004).
For the data recorded here, the shrinking sphere model has been ap-
The temperature dependence of the reaction rate can be investigat-
plied with a five-point particle size range (approximating measured
ed with Arrhenius plots (Fig. 7). The slope of the Arrhenius plot is often
particle size; Fig. 6 inset). This is calculated from a best fit of the sum
taken to represent the activation energy for a reaction. However, while
of the calculated Ti concentrations from the dissolution of each particle
this applies to solution-phase reactions, for a heterogeneous reaction
size fraction with a particle size-dependent reaction rate constant,
between a solid and a solution, unless the reaction is purely chemically
where smaller particles completely dissolve earlier and result in the
controlled, the slope is actually a combination of activation energy and
plot being initially steeper.
diffusion since diffusion also varies with temperature. Furthermore,
Now the plot of kt=1 − (1− a)1/3 becomes non-linear and matches
the temperature dependence of diffusion of ions in a liquid also follows
the data well compared with the straight line based on a single-particle
the Arrhenius rate equation, and this would apply to either of the
size, given by kt= 1− (1− a)1/3 (Fig. 6). Clearly, accounting for particle
models discussed above. Therefore, despite the widespread use in the
size distribution does improves the fit of the shrinking sphere model
literature of “activation energies” when referring to Arrhenius plots
(even when our rather primitive five-point distribution is used).
for heterogeneous reactions, the slopes represent a confounding of
The Barrytown ilmenite is known to dissolve in HCl by the hollowing
diffusion and chemical reactions, and in ilmenite dissolution, the rate
out of the particles, initially along the basal planes, followed by disinte-
is probably mostly due to diffusion. Regardless, the plots are useful in
gration of the partially dissolved particles (Duncan and Metson, 1982b).
predicting (interpolating) reaction rate as a function of temperature.
This process is, therefore, similar to a shrinking core followed by a
shrinking particle, so does not precisely match either theoretical
3.7. Acid concentration and digestion rate

In the experimental data presented here, where the HCl concentra-

tion varied during the reaction, it was not easy to determine the
relationship between the ilmenite digestion reaction and acid concen-
tration. However, data for the digestion of Nigerian ilmenite
(Olanipekun, 1999), where the acid concentrations were held essential-
ly constant, can be plotted in either the form kt= 1 + 2(1− a) − 3(1−
a)2/3 or kt= 1 −(1− a)1/3, assuming a monodisperse particle size distri-
bution (Fig. 8).
Based on the stoichiometry of the reaction, if the digestion is
diffusion-controlled (as either a shrinking core or shrinking sphere),
then the process could be expected to depend on the fourth power of
the H+ concentration. If chemical reaction control is the driving force
behind digestion rate, then the rate would depend on the order of the
reaction mechanism which, except in rather simple cases, is not normal-
ly a direct reflection of the overall reaction stoichiometry, and is unlikely
to be higher than second or third order in a reactant. The data plotted
using both models show a fairly good fit to fourth-order rate depen-
dence. We were not able to include particle size distribution in the
Fig. 5. Ilmenite digestion curves for Ti plotted to test the shrinking sphere model. Green
kt = 1 − (1 − a)1/3 plot because this information was not provided by
(circle) 60 °C, red (triangle) 80 °C, blue (square) 85 °C, black (diamond) 90 °C. Solid Olanipekun, but it is likely that its incorporation would have resulted
lines are linear fits; dashed lines are a guide to the eye. in an even a closer fit of a fourth-order curve to the data.
202 R.G. Haverkamp et al. / Hydrometallurgy 163 (2016) 198–203

Fig. 7. Arrhenius plots for ilmenite digestion with a) k taken from kt=1−(1−a)1/3; b) k taken from kt=1+2(1−a)−3(1−a)2/3. Circles are Ti, triangles are Fe. Note that particle size or
particle size distribution is included in the values of k. The slope for a) is 15.0 and for b) is 11.3.

3.8. Saturation concentrations and digestion rate reaction from available enthalpies of formation of the components
(Table 2) and heat capacities of the solution.
An important experimental factor to consider is the saturation con- The reaction is considered to be:
centration of Fe2+ or TiO2+: as these are approached, the components
begin to limit reaction rate (Haverkamp and Welch, 1998). It seems
likely that the solubility of TiOCl2 in acid solution is high FeTiO3(s) + 4 HCl(aq, 32% w/w) = Fe2+(aq) + TiO2+(aq) + 4Cl−(aq)
(Madekufamba et al., 2006), so its effect on the diffusion rate of TiO2+ + 2H2O.
from the site of dissolution is not significant. However, FeCl2 is known
to have limited solubility in HCl solutions (i.e. c. 30% w/w in 10% HCl The heat capacity of the system is approximated as the sum of
at 60–90 °C; (Werner and Bender, 2010). In the dissolution results the heat capacity of water (4180 J kg− 1 K−1) and ilmenite
presented here, the concentration in the ilmenite digestion liquors (652 J kg− 1 K−1). The error due to neglecting HCl and the change
reached c. 12.7% w/w FeCl2. A preliminary test was carried out by adding with digestion is estimated to be small, c. 5%.
10.0% w/w FeCl2 to the acid before ilmenite digestion for the 90 °C For a ratio of 2:1 HCl-to-ilmenite, with an initial HCl concentration of
conditions and little effect was observed on the dissolution rate of Fe 32% w/w as used here, the adiabatic temperature rise of the digestion
or Ti (detailed results not presented here). mix for complete digestion would be 41 °C using a ΔHf for ilmenite of
1235 (Vitvitskii, 2009), but in the range 30–70 °C using the ΔHf for
ilmenite of 1125–1278 kJ mol−1 cited in Kelsall and Robbins (1990). If
3.9. Heat of reaction a more concentrated ilmenite mix is used, of 1.44:1 HCl-to-ilmenite,
the temperature rise would be 57 °C (or a range of 41–96 °C using the
The reaction of ilmenite with HCl is exothermic. In industrial diges- range of 1125–1278 kJ mol−1 for the ilmenite ΔHf). On the small scale
tion processes, the ratio of HCl-to-ilmenite is likely to be low (as it of the laboratory experiments, the system was far from adiabatic and
was in these experiments), so the temperature rise would be significant. the heat of reaction was easily accommodated by heat loss from the
The reactions were not adiabatic here; however, it is possible to esti- reaction vessel; however, for larger scale systems, it will be critical to
mate, based on some approximations, the temperature rise due to the consider the heat of reaction. Further investigation is needed, and it

Fig. 8. Dependence of rate constant on HCl concentration for data taken from Nigerian ilmenite digestion (Olanipekun, 1999). a) kt=1+2(1− a)−3(1−a)2/3; b) kt=1− (1− a)1/3.
Curves are best fit to a fourth power.
 R.G. Haverkamp et al. / Hydrometallurgy 163 (2016) 198–203 203

Table 2 References
Literature values of enthalpies of formation.
Christie, A.B., Brathwaite, R.L., 1998. Mineral commodity report 16 — Titanium. N. Z.
Component ΔHf (kJ mol−1) Source Mining 23, 15–25.
Cservenyak, I., Kelsall, G.H., Wang, W., 1996. Reduction of Ti-IV species in aqueous sulfuric
FeTiO3 Range 1125–1278 Cited in Kelsall and Robbins (1990)
and hydrochloric acids.1. Titanium speciation. Electrochim. Acta 41 (4), 563–572.
1235 Vitvitskii (2009)
Duncan, J.F., Metson, J.B., 1982a. Acid attack on New-Zealand ilmenite.1. The mechanism
Fe2+(aq) 90 Parker and Khodakovskii (1995)
of dissolution. N. Z. J. Sci. 25 (2), 103–109.
TiO2+(aq) 690 Madekufamba et al. (2006) Duncan, J.F., Metson, J.B., 1982b. Acid attack on New-Zealand ilmenite. 2. The structure
Cl−(aq) 167 and composition of the solid. N. Z. J. Sci. 25 (2), 111–116.
H2O 286 Gbor, P.K., Jia, C.Q., 2004. Critical evaluation of coupling particle size distribution with the
HCl(aq, 32% w/w) 158 Wagman et al. (1982) shrinking core model. Chem. Eng. Sci. 59 (10), 1979–1987.
HCl(aq, 15% w/w) 162 Wagman et al. (1982) Gireesh, V.S., Vinod, V.P., Nair, S.K., Ninan, G., 2015. Catalytic leaching of ilmenite using
hydrochloric acid: a kinetic approach. Int. J. Miner. Process. 134, 36–40.
Girgin, I., 1990. Leaching of ilmenite in HCl-H2O, HCl-CH3OH-H2O and HCl-CH3OH
solutions. Hydrometallurgy 24 (1), 127–134.
seems feasible to actually measure the heat of this digestion reaction Grapes, R., Watanabe, T., 1992. Paragenesis of titanite in metagraywackes of the Franz
directly in a suitably designed system. Josef-Fox Glacier area, Southern Alps. N. Z. Eur. J. Mineral. 4 (3), 547–555.
Haverkamp, R.G., Welch, B.J., 1998. Modeling the dissolution of alumina powder in
cryolite. Chem. Eng. Process. 37, 177–187 (Copyright (C) 2014 American Chemical
Society (ACS). All Rights ReservEd.).
3.10. Avoiding Ti hydrolysis Jackson, J.S., Wadsworth, M.E., 1976. A kinetic study of the dissolution of Allard Lake
ilmenite in hydrochloric acid. Light Met. 1976 (1), 481–540.
If the goal of ilmenite dissolution in HCl is for the direct production of Jain, S.K., Jena, P.K., 1977. Hydrochloric-acid leaching of Gopalpur ilmenite. Indian
J. Technol. 15 (9), 398–402.
TiO2 (rather than upgrading of the ore for a high Ti feedstock), then it is
Janssen, A., Putnis, A., 2011. Processes of oxidation and HCl-leaching of Tellnes ilmenite.
very important to avoid the initiation of Ti hydrolysis during digestion. Hydrometallurgy 109 (3–4), 194–201.
All the Ti must remain in solution, but more crucially, Ti hydrates – Kelsall, G.H., Robbins, D.J., 1990. Thermodynamics of Ti-H2O-F(−Fe) systems at 298 K.
J. Electroanal. Chem. 283 (1–2), 135–157.
that may nucleate hydrolysis and thereby render it more difficult to
Levenspiel, O., 1972. Chemical Reaction Engineering. Wiley, New York.
control of TiO2 hydrolysis (to ensure top-quality pigment production) Madekufamba, M., Trevani, L.N., Tremaine, P.R., 2006. Standard enthalpy of formation of
– must not be generated. A combination of high temperature, long reac- aqueous titanyl chloride, TiOCl2(aq), at T = 298.15 K. J. Chem. Thermodyn. 38
tion times, and low acid concentrations lead to hydrolysis, and therefore (12), 1563–1567.
Mann, G., James, D., 1989. The evaluation and development of the Barrytown ilmenite
these all constraint the digestion of ilmenite. As seen in Fig. 4, when the deposit as a source of titanium dioxide pigment. Monogr. Ser. - Australas. Inst. Mining
thresholds for temperature, time, and (lower) acid concentration are Metall. 13, 169–172.
exceeded, hydrolysis can be rapid; establishing the conditions for Mills, L., 2013. Coast Ilmenite Chase back on. Greymouth Star, Greymouth, New Zealand.
Ogasawara, T., de Araujo, R.V.V., 2000. Hydrochloric acid leaching of a pre-reduced
complete digestion while avoiding hydrolysis is not straightforward. Brazilian ilmenite concentrate in an autoclave. Hydrometallurgy 56 (2), 203–216.
Olanipekun, E., 1999. A kinetic study of the leaching of a Nigerian ilmenite ore by hydro-
chloric acid. Hydrometallurgy 53, 1–10.
4. Conclusions Parker, V.B., Khodakovskii, I.L., 1995. Thermodynamic properties of the aqueous ions (2+
and 3+) of iron and the key compounds of iron. J. Phys. Chem. Ref. Data 24 (5),
1699–1745.
It has been shown that ilmenite from Barrytown, New Zealand, may Sarker, M.K., Rashid, A.K.M.B., Kurny, A.S.W., 2006. Kinetics of leaching of oxidized and re-
be readily digested in concentrated HCl, without pretreatment, and that duced ilmenite in dilute hydrochloric acid solutions. Int. J. Miner. Process. 80 (2–4),
hydrolysis of the Ti fraction can be avoided. A rate equation for the reac- 223–228.
Sasikumar, C., et al., 2004. Effect of mechanical activation on the kinetics of sulfuric acid
tion has been fitted that enables modeling of the digestion reaction leaching of beach sand ilmenite from Orissa, India. Hydrometallurgy 75 (1–4),
rates. The temperature dependence of the reaction has been established 189–204.
and the dependence of the reaction rate on acid concentration has been Sasikumar, C., Rao, D.S., Srikanth, S., Mukhopadhyay, N., Mehrotra, S.P., 2007. Dissolution
studies of mechanically activated Manavalakurichi ilmenite with HCl and H2SO4.
approximated. Calculations of the heat of reaction have lead to further Hydrometallurgy 88 (1–4), 154–169.
modeling, in this instance of the temperature rise during the digestion Sinha, H.N., 1984. Hydrochloric acid leaching of ilmenite. Symposia Series — Austral-
process. As the models are applicable at an industrial scale, this work asian Institute of Mining and Metallurgy, Extractive Metallurgy Symposium,
pp. 163–168.
forms the basis for the development of an industrial process to digest Sole, K.C., 1999. Recovery of titanium from the leach liquors of titaniferous magnetites by
Barrytown ilmenite for the direct production of TiO2. Research on the solvent extraction — part 1. Review of the literature and aqueous thermodynamics.
leaching of beach sand ore that has not been ground will be presented Hydrometallurgy 51 (2), 239–253.
U.S. GeologicalSurvey, 2013. Mineral Commodity Summaries 2013. U.S.GeologicalSurvey,
in the future, as will work on the hydrolysis of these solutions to pro-
(Reston, Virginia).
duce pigment-grade TiO2. Vitvitskii, A.I., 2009. New equation of boiling curve. Russ. J. Appl. Chem. 82 (3), 505–508.
Wagman, D.D., et al., 1982. The NBS tables of chemical thermodynamic properties -
selected values for inorganic and C-1 and C-2 organic-substances in SI units. J. Phys.
Acknowledgements Chem. Ref. Data 11, 1 (-&).
Wahyuningsih, S., et al., 2013. Decomposition of ilmenite in hydrochloric acid to obtain
high grade titanium dioxide. Asian J. Chem. 25 (12), 6791–6794.
Westland Ilmenite Ltd. supplied ilmenite concentrate from the Werner, A., Bender, R., 2010. Ferrous Chlorides (FeCl2, FeCl3), Corrosion Handbook.
Barrytown deposit. Sue Hallas of Nelson assisted with editing the DECHEMA e.V., Frankfurt.
manuscript.