A new, energy-efficient chemical pathway for extracting ti metal from ti minerals

Article abstract of DOI:10.1021/ja408118x

Titanium is the ninth most abundant element, fourth among common metals, in the Earth's crust. Apart from some high-value applications in, e.g., the aerospace, biomedicine, and defense industries, the use of titanium in industrial or civilian applications has been extremely limited because of its high embodied energy and high cost. However, employing titanium would significantly reduce energy consumption of mechanical systems such as civilian transportation vehicles, which would have a profound impact on the sustainability of a global economy and the society of the future. The root cause of the high cost of titanium is its very strong affinity for oxygen. Conventional methods for Ti extraction involve several energy-intensive processes, including upgrading ilmenite ore to Ti-slag and then to synthetic rutile, high-temperature carbo-chlorination to produce TiCl₄, and batch reduction of TiCl₄ using Mg or Na (Kroll or Hunter process). This Communication describes a novel chemical pathway for extracting titanium metal from the upgraded titanium minerals (Ti-slag) with 60% less energy consumption than conventional methods. The new method involves direct reduction of Ti-slag using magnesium hydride, forming titanium hydride, which is subsequently purified by a series of chemical leaching steps. By directly reducing Ti-slag in the first step, Ti is chemically separated from impurities without using high-temperature processes.

Full text of DOI:10.1021/ja408118x

Communication
pubs.acs.org/JACS
A New, Energy-Efficient Chemical Pathway for Extracting Ti Metal
from Ti Minerals
Zhigang Zak Fang,* Scott Middlemas,† Jun Guo, and Peng Fan
Metallurgical Engineering, University of Utah, Salt Lake City, Utah 84112, United States
S
* Supporting Information
Communication focuses on the fundamental principles of the
ABSTRACT: Titanium is the ninth most abundant
element, fourth among common metals, in the Earth’s
crust. Apart from some high-value applications in, e.g., the
aerospace, biomedicine, and defense industries, the use of
titanium in industrial or civilian applications has been
extremely limited because of its high embodied energy and
high cost. However, employing titanium would signifi-
cantly reduce energy consumption of mechanical systems
such as civilian transportation vehicles, which would have a
profound impact on the sustainability of a global economy
and the society of the future. The root cause of the high
cost of titanium is its very strong affinity for oxygen.
Conventional methods for Ti extraction involve several
energy-intensive processes, including upgrading ilmenite
ore to Ti-slag and then to synthetic rutile, high-
temperature carbo-chlorination to produce TiCl4, and
batch reduction of TiCl4 using Mg or Na (Kroll or Hunter
process). This Communication describes a novel chemical
pathway for extracting titanium metal from the upgraded
titanium minerals (Ti-slag) with 60% less energy
consumption than conventional methods. The new
method involves direct reduction of Ti-slag using
magnesium hydride, forming titanium hydride, which is
subsequently purified by a series of chemical leaching
steps. By directly reducing Ti-slag in the first step, Ti is
chemically separated from impurities without using high-
temperature processes.
new method and the experimental data that demonstrate its
feasibility. The methods and data of the energy analysis are
documented in the Supporting Information.
The root cause of the high cost of titanium is its very strong
affinity for oxygen. A quick reference to the Ellingham diagram
of oxides1 will show that, among common metals, only Al, Mg,
Ca, and Li form more stable oxides than Ti. The challenges for
producing Ti are not only the extractive processes to separate
its oxides from other compounds in Ti feed stocks, but more so
the process that reduces titanium oxide to metal. Figure 1
Figure 1. Comparison of conventional and developing processes for
production of Ti sponge or Ti powder.
illustrates the major steps of existing methods of producing Ti
sponge from ilmenite ore. Over 90% of global titanium reserves
is in the form of ilmenite (∼45−65% TiO2 equivalent),2 with
only 10% occurring as natural rutile (∼95% TiO2). Ilmenite can
be smelted using a carbothermal process to yield pig iron and
Ti-slag.3 The main component of Ti-slag is ∼80% Ti-based
oxides which are bonded with other metals in the form of metal
titanates,4 including Fe, Mg, Ca, Al, Si, and other transition
metals. Pig iron is used for making iron and steel, whereas the
Ti-slag can be used to produce TiO2 pigment or Ti metal.
The existing technologies, including both commercial and
developmental processes, can be categorized into two groups:
those by reduction of TiCl4 (e.g., Kroll,5 Armstrong/ITP6) and
those by reduction of TiO2 (e.g., FFC process.7) The emphasis
in the former approach is to optimize the TiCl4 reduction
processes, while the primary motivation of the latter approach
is to avoid the high-temperature chlorination process. The
Armstrong process is a significant advancement compared to
itanium is the ninth most abundant element, fourth
Tamong common metals, in the Earth’s crust. It has been
used as a high-performance light metal in aerospace, defense,
biomedical, petrochemical, and other high-value applications
owing to its high specific strength, extreme corrosion resistance,
and biocompatibility. However, the use of titanium in industrial
or civilian applications has been extremely limited because of its
high embodied energy and high cost, despite the fact that
employing titanium would significantly reduce energy con-
sumption of mechanical systems such as civilian transportation
vehicles, which would have a profound impact on the
sustainability of a global economy and the society of the
future. Developing a technology to significantly reduce the
energy consumption, and thus the cost, of extracting titanium
has challenged scientists for nearly a century. In the work
described herein, a novel method for extracting titanium metal
from upgraded titanium minerals with 60% less energy
consumption than conventional methods is discovered. This
Received: August 8, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja408118x | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
the conventional Hunter8 process, as it is a continuous process
with molten sodium being pumped continuously into the
reactor to reduce TiCl4. The FFC process, published in 2000,
was also a breakthrough that demonstrated the potential of the
electrochemical approach. In either of the two approaches,
titanium ore must first be refined through a series of chemical
extractive processes to obtain highly refined rutile (TiO2)
before reduction to Ti metal.
The relative cost for each of the processing steps in the Kroll
process was reported by Hartman et al.9 After upgrading TiO2
(10% of total), chlorination and Mg reduction are the most
costly steps of making Ti sponge (24 and 66%, respectively). As
it will be shown, the new approach described in this
Communication changes the chemical pathway by first directly
reducing TiO2 to titanium hydride (TiH2) in the slag to
chemically isolate Ti from other compounds and then removing
impurities by a series of targeted leaching stages. The new
approach eliminates the need for carbo-chlorination of TiO2
and the reduction of TiCl4 by Kroll or Hunter types of
processes.
that the insolubility of TiH2 in water is attributed to its kinetic
passivation in water. These properties set up a condition by
which the product of the direct reduction of Ti-slag can be
sequentially leached to remove impurities and purify TiH2.
The feasibility of the first step, direct reduction of Ti-slag
with MgH2 (DRTS), lies in several underlying principles.
Thermodynamic modeling (HSC Chemistry 5.11, commercial
software from Outotec, Finland) showed that when MgH2 is
put in a system with Ti-slag in a hydrogen atmosphere, TiH2
and MgO are the favored products of the reduction of slag with
MgH2. Specifics of the thermodynamic analysis are shown in
the Supporting Information. Small fractions of other
intermetallic compounds may also form as the results of
DRTS, which will be removed through leaching processes.
From a kinetics perspective, one of the key concerns
regarding the new approach is that the reaction between MgH2
and TiH2 is in the “solid” state, and solid-state reactions are
usually too slow for such production methods. However, a
unique feature of MgH2 and Mg is that MgH2 dehydrogenates
between 300 and 400 °C, when the hydrogen pressure is at or
below 1 atm. Solid Mg has an unusually high vapor pressure at
temperatures below its melting point (649 °C). Thus, when the
reaction is carried out at approximately 500 °C and 1.0 atm
pressure of hydrogen (see results that follows), there is
substantial presence of the vapor phase of Mg, which improves
the reaction kinetics dramatically.
Recognizing the energy intensiveness involved in existing
technologies, the present authors discovered that the titanium
oxide contained in the minerals can be reduced directly with
magnesium hydride (MgH2) to form TiH2, which could be
purified subsequently by a series of chemical leaching steps.
The concept of the new chemical pathway is shown in three
primary steps in Figure 2.
The reaction kinetics and thermodynamic driving force for
forming TiH2 are further aided by the differences in equilibrium
pressure of hydrogen between TiH2 and MgH2. At 500 °C, the
equilibrium pressure of TiH2 is less than 1.0 atm (approx-
imately 0.5 atm), while the equilibrium pressure of MgH2 is
higher than 1.0 atm. Therefore, when the process is carried out
at 500 °C and 1.0 atm H2, Ti will hydrogenate, while MgH2 is
dehydrogenated, leaving Mg in a solid−vapor phase equilibrium
with substantial presence of Mg vapor.
The feasibility of reaction (1) can also be inferred in the
literature reports by Borok and Teplenko11 and Froes et al.,12
who both reported the reduction of purified TiO2 to Ti and/or
TiH2 using CaH2. Our preliminary work showed that, although
thermodynamically both CaH2 and MgH2 should work at
different temperatures, in the experiments CaH2 did not reduce
slag at similarly low temperature conditions. This was attributed
to the fact that CaH2 does not dehydrogenate until much
higher temperatures, and solid Ca has a much higher melting
point and not as high a vapor pressure as does Mg.
Additionally, the recovery of Mg through electrolysis has
been reported to consume 13.0 kWh of electrical energy per kg
of Mg, much less than the energy required to recover Ca
through electrolysis, which is 33−55 kWh/kg.13
Figure 2. Schematic illustration of the three-step process for extracting
Ti from upgraded Ti mineral (Ti-slag): (1) direct reduction of Ti-slag
using MgH2, (2) leaching to purify TiH2, and (3) dehydrogenation of
TiH2 to form Ti.
The feasibility of the second step is founded in the principles
of fractional chemistry and extractive metallurgy. The
impurities can be sequentially leached using common chemical
reagents. Many of the specific techniques are already proven in
the metallurgical industry. First, to remove MgO from the
reduced powder mixture, hot ammonium chloride (NH4Cl)
solutions can be used,14 and the reaction proceeds according to
The TiO2 contained within the slag reacts with the MgH2
reductant as follows:
TiO2 + 2MgH2 → TiH2 + 2MgO + H2(g)
(1)
The strategy of the new method is that, when MgH2 is used
to react with Ti-slag and form TiH2, Ti is chemically isolated
from the rest of the compounds in slag. Forming TiH2, rather
than Ti metal, is deliberate and necessary because Ti metal is
more prone to forming alloys with other elements such as Fe,
which would be extremely difficult to separate. Moreover, TiH2
has very unique chemical properties. It is insoluble in water,
resistant to dilute acid solutions,10 and has minimal or no
solubility for other impurities in the slag. It should be noted
MgO + 2NH4Cl(aq) → MgCl2(aq) + 2NH3(aq) + H2O
(s)
(2)
It has also been found that adding a chelating agent, such as
sodium dihydrogen citrate (NaC6H7O7), to the NH4Cl solution
can also significantly improve the solubility of MgO.15 Similarly,
leaching with hot NaOH solution can remove any remaining
B
dx.doi.org/10.1021/ja408118x | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Figure 3. X-ray diffraction patterns of (a) as-received Ti-slag, (b) the product of the reduction of slag with MgH2 at 500 °C, (c) the product of
leaching the reduced powder with NH4Cl solution, and (d) the product of dehydrogenation at 400 °C in an inert Ar atmosphere.
silicates from the reduced powder that would otherwise be
insoluble in mineral acids other than HF, and will also remove
Al2O3 with minimal dissolution of TiH2. It should be
emphasized that the majority of the oxides, including FeO,
Al2O3, and SiO2, are all reduced along with TiO2 during the
direct reduction process using MgH2. Intermetallic compounds
such as Fe2Ti, TiAl, and Ti5Si3 may form on the basis of the
thermodynamic analysis. Only residual oxides are still of
concern.
also found that milling the slag also greatly enhanced the degree
of conversion. Ti-slag was ball-milled with 20 wt% of the
eutectic salt mixture and then mixed with MgH2. The slag/
reductant mixture was heated in a tube furnace with flowing
hydrogen to 500 °C. Figure 3 compares the X-ray diffraction
(XRD) spectra of the reaction product and the as-received Ti-
slag, as well as those of the products after selected steps of
leaching and purification. It is evident that the primary phase in
the reaction products is TiH2, which was not present in as-
received Ti-slag. The other main compound in the reduction
product is MgO.
Any remaining metal or metal oxide compounds such as Fe
can be dissolved using a relatively dilute mineral acid,14 such as
HCl:
The total composition of the reduced powder was
determined using microwave digestion in concentrated
HNO3/HF acid solution and ICP-MS analysis. Another sample
of reduced powder was leached in an HCl solution at an
optimized concentration and temperature that allowed for the
dissolution of converted TiH2 but minimal Ti from any
unreacted slag. The Ti measured in the leach solution is divided
by the total Ti content in the powder to obtain the degree of
conversion. Conversions as high as 95% have been obtained.
The product of the direct reduction experiment was then
subjected to leaching as described above. MgO is removed from
the reduced powder mixture by NH4Cl solutions according to
eq 2. In Figure 3, a sharp decrease in the intensity of the MgO
peaks is observed after NH4Cl leaching. Remaining metal and
metal oxide compounds such as Fe were dissolved using dilute
HCl, as shown in eq 3. Figure 4 shows the morphology and
particle size of the TiH2 powder after leaching and purification.
The chemical compositions of the powder after all leaching
steps were analyzed using the energy-dispersive spectroscopy
(EDS) technique. Figure 4 shows that the TiH2 particles are
99% pure, free of impurity particles. The exact compositions of
impurity elements, however, were not determined because their
composition levels were below the accuracy capability of the
EDS technique. Ongoing efforts are focused on developing
methods to measure exact impurity compositions of the
powder.
FeO + 2HCl(aq) → FeCl2(aq) + H2O
(s)
(3)
TiH2 has been found to have a lower dissolution rate in HCl
acid than in H2SO4,10 and FeO has been found to have a higher
reactivity in HCl than in H2SO4;16 therefore, HCl was chosen
as the preferred mineral acid. Further, residual MgO in the
mixture can also be dissolved by the dilute HCl solution. In the
event that other residual contaminants are present at levels that
exceed acceptable limits, coordination compounds such as
carotenes, porphyrins, supramolecules, polycyclic aromatics,
ethylene diamine tetraacetic acid, carboxylic acids, N-hetero-
cyclic carboxylic acids, pyrimidine carboxylic acids, and
tetrazole carboxylic acid may be used to improve leaching
and impurity separation.
An experimental study was conducted on the reduction of
Ti-slag with MgH2. The slag used in this investigation is a solid
solution of various titanate compounds with a pseudo-brookite
structure.4 The composition of the slag includes 79.8% total
TiO2 (15.8% Ti2O3 reported as TiO2), 9.1% FeO, 5.6% MgO,
2.7% SiO2, 2.2% Al2O3, 0.6% total other metal oxides, and 0.1%
metallic Fe, as assayed by Rio Tinto Iron and Titanium. No
lanthanide or actinide compounds were detected.
It was found that the extent of reduction of Ti-slag with
MgH2 was facilitated by the inclusion of a eutectic mixture of
sodium and magnesium chloride salts that would be molten at
the reaction temperature. Molten salts have very high
conductivity and appear to facilitate electron transfer during
the reduction reaction. The exact molten salt composition was
determined after several trials using either single or mixed
chlorides salts. More details on the compositions of the
chloride salts are given in the Supporting Information. It was
The resulting hydride powder was dehydrogenated to
produce Ti metal by heating it above 400 °C in an argon
atmosphere. The XRD spectrum of reduced powder after
NH4Cl leaching and dehydrogenation at 400 °C is shown in
Figure 3. The hydride peaks have clearly been replaced by
C
dx.doi.org/10.1021/ja408118x | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
(9) Hartman, A. D.; Gerdemann, S. J.; Hansen, J. S. JOM 1998, 50,
16.
(10) Hydrogen Materials Science and Chemistry of Metal Hydrides;
Veziroglu, T. N., Zaginaichenko, S. Yu., Schur, D. V., Trefilov, V. I.,
Eds.; Kluwer Academic Publishers: Amsterdam, 2002.
(11) Borok, B. A.; Teplenko, V. G. Trans. Central Res. Inst. Ferrous
Metall., Moscow, Russia 1965, 43, 69−80.
(12) Froes, F. H. J. Metall. 1998, 50, 41.
(13) Best Available Techniques in the Non Ferrous Metals Industries;
European Commission: Seville, Spain, 2001.
(14) Lide, D. R. CRC Handbook of Chemistry and Physics, 67th ed.;
CRC Press: Boca Raton, FL, 1987.
(15) Fink, J. Petroleum Engineer’s Guide to Oil Field Chemicals and
Fluids, 1st ed.; Gulf Professional Publishing: Waltham, MA, 2007.
(16) Cotell, C. M.; Sprague, J. A.; Smidt, F. A. ASM Handbook; ASM
International: Materials Park, OH, 1994.
Figure 4. SEM/EDS analysis of final titanium hydride product using
the DRTS process as described.
alpha-Ti peaks, demonstrating the flexibility of the process to
make either Ti metal or TiH2 powders.
Finally, the change in the chemical pathway for Ti extraction
facilitates the potential for substantial energy reductions. The
DRTS approach is based on low-temperature and ambient-
pressure processes that eliminate the need for high-temperature
carbo-chlorination, molten Mg reduction, and vacuum
distillation. The energy-saving potential of the DRTS process
was estimated to be approximately 62% less than that of the
Kroll process. Details of the analysis of energy consumption are
included in the Supporting Information
ASSOCIATED CONTENT
* Supporting Information
Experimental procedures and energy analysis. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
zak.fang@utah.edu
■
Present Address
†S.M.: Army Research Laboratory, Aberdeen, MD
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work received financial support from the U.S. Department
of Energy on titanium-related technologies. We thank Rio
Tinto Iron and Titanium for supplying the Ti-slag for the
experiments. S.M. would like to acknowledge support from the
US Army Research Laboratory (ARL) administered by the Oak
Ridge Institute for Science and Education.
REFERENCES
■
(1) Ellingham, H. J. T. J. Soc. Chem. Ind., London 1944, 63, 125.
(2) U.S. Geological Survey. Mineral Commodity Summaries; U.S.
Government Printing Office: Washington, DC, 2010; p 172.
(3) Sahu, K. K. Waste Manage. Res. 2006, 24, 74.
(4) Gueguin, M.; Cardarelli, F. Miner. Process. Extr. Metall. Rev. 2007,
28, 1.
(5) Kroll, W. Trans. Metall. Soc. AIME 1959, 215, 546.
(6) Armstrong, D. R.; Borys, S. S.; Anderson, R. P. U.S. Patent
5779761, 1998.
(7) Chen, G.; Fray, D.; Farthing, T. Nature 2000, 407, 361.
(8) Barksdale, J. Titanium: its occurrence, chemistry, and technology,
2nd ed.; Ronald Press Co.: New York, 1996.
D
dx.doi.org/10.1021/ja408118x | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX