Thorium in high-titania slag - J. NELL

NELL, J. Thorium in high-titania slag. The 6th International Heavy Minerals Conference ‘Back to Basics’, The Southern African Institute of Mining and
Metallurgy, 2007.

                                        Thorium in high-titania slag
                                                                    J. NELL
                                                             Mintek (now at Hatch)

                         Introduction                                         range of compositions were selected for bulk chemical
Heavy mineral deposits often contain relatively high levels                   analysis, scanning electron microscopy and micro-analysis.
of radioactive elements (thorium and uranium in particular).                  The samples were crushed to a maximum particle size of
It is difficult to obtain clean separation of ilmenite and                    1 mm and from each sample a 20-kg batch was riffled out
monazite (the main impurity mineral containing radioactive                    for analysis.
elements), and physical intergrowths of the two minerals
are, in fact, not uncommon. As a result, ilmenite                                            Bulk chemical composition
concentrates obtained from such deposits often contain high                   Chemical compositions of the samples selected for test
levels of thorium and uranium. Because of the large                           work are given in Table I. TiO2, Al2O3, SiO2, MgO, MnO,
negative free energies of formation of ThO2 and U3O8 both                     FeO and V2O5 were analysed by ICP-OES, Na2O, CaO and
thorium and uranium report to the slag during smelting and                    Cr2O3 by atomic absorption spectroscopy and U and Th by
if an ilmenite concentrate contaminated with monazite is                      XRF (the XRF detection limit for U and Th was around
used to make high titania slag, the concentrations of these                   4 ppm). The ‘B’ series has a higher silica content than the
elements in the slag end up about 40% higher than in the                      others as a result of small silica additions that were made
ilmenite feed.                                                                during smelting. All samples have elevated alumina levels
   Previous work, patented by RGC Mineral Sands Limited,                      as a result of furnace refractory contamination during
and subsequently confirmed by Mintek, showed that                             smelting.
radioactivity may be reduced by roasting an ilmenite                            As expected, there is an inverse relationship between the
concentrate with borax at 1000°C to 1100°C and leaching it                    concentrations of TiO2 and FeO in the samples (Figure 1).
with hydrochloric acid1. It is evident that the cost of an                    The TiO2 and FeO levels are similar to the levels found in
ilmenite roast/leach process will be high and there may be                    industrial slags2. Extreme conditions of reduction under
economic benefits in removing the radioactive elements                        which reduction of more ‘refractory’ oxides (e.g., chromia,
from the slag instead (the volume of slag to be treated is                    magnesia, silica, alumina and even urania and thoria) might
smaller than the volume of ilmenite feed).                                    take place were not explored in the study. Note that BB1
   To remove radioactive elements from the slag it is                         and BBB1 contain more than 7% combined silica and
necessary to know how these elements occur in the slag. In                    alumina (SiO2 as a flux addition during smelting, and Al2O3
other words, the distribution of uranium and thorium                          from refractory contamination) and they do not fall on the
between the various oxide, silicate and metallic phases in                    general trend defined by the other samples.
high titania slag must be determined. Once the deportment                       The U and Th concentrations of sample E2 is noteworthy.
of thorium and uranium between the various phases has                         In the preparation of slag E2, monazite concentrate was
been determined, an appropriate process for slag                              added to the ilmenite to increase the concentrations of
purification can be developed.                                                radioactive elements in the slag. This was done primarily to
                                                                              facilitate the detection of Th and U during the phase
                         Test samples                                         chemical characterization of the slag.
High-titania slag was prepared in a pilot-scale DC furnace
from ilmenite concentrates with different levels of uranium                                 Radioactivity measurements
and thorium. Subsequently, ten slag samples representing a                    Gamma ray emissions from the samples were measured

                                                                 Table I
                                 Bulk chemical compositions of the slag samples selected for test purposes

THORIUM IN HIGH-TITANIA SLAG                                                                                                                       7
solid-solution series (in high-titanium slags M may be Ti4+,
                                                                        Ti 3+ , Fe 2+ , Mg 2+ and Al 3+ , appropriately combined to
                                                                        maintain charge balance in the structure). All the slags
                                                                        contain TiO2 (rutile) in trace amounts.
                                                                          The shape of the background at small diffraction angles
                                                                        indicates the presence of small amounts of silicate glass in
                                                                        the slags. This feature was particularly prominent in the ‘B’
                                                                        series of samples and is consistent with the addition, during
                                                                        smelting, of SiO2 to these samples.

                                                                        The slags, after tapping, consist mostly of iron-titanium
                                                                        oxides (M3O5 solid solutions where M = Ti4+, Ti3+, Fe2+,
                                                                        Mg2+, Mn2+ and Al3+) and silicate phases. The aim of this
                                                                        part of the characterization is to determine how the
                                                                        radioactive elements are distributed between these phases.
                                                                        This information is critical to develop methods for the
        Figure 1. Relationship between the TiO2 and FeO                 removal of impurities from the slag.
               concentrations in the selected slags
                                                                        Proton induced X-ray emission spectroscopy (PIXE)
                                                                        Measurements on slag F 3, which was doped with
                              Table II                                  monazite, were performed by iThemba Laboratories. The
    Gamma ray emissions from samples selected for test work.
    Results are expressed as (counts.second-1.gram-1), standard
                                                                        advantage offered by PIXE is a relatively low detection
                deviations are given in parentheses                     limit for elements with high atomic mass, such as thorium
                                                                        and uranium. However, there are also draw backs to the
                                                                           • The inability to analyse elements with an atomic
                                                                             number below about 20 (Ca) when these are contained
                                                                             in phases with a high mean atomic number
                                                                           • The semi-quantitative nature of the analyses
                                                                             (measurements are standardless, although matrix
                                                                             corrections are performed).
                                                                          It is therefore not possible to perform comprehensive,
                                                                        quantitative chemical analyses of the phases present in the
                                                                        slag using this technique.
                                                                          However, the instrument can be used to generate
                                                                        concentration maps for areas of slag measuring almost
with a gamma ray detector. Measurements were performed                  0.5 mm2. This provides a powerful visual image of the
on 50-g samples, counting for 400 seconds. Repeated                     distribution of trace elements between the coexisting phases
measurements (between 3 and 6) were performed on each                   over a relatively large area of slag. Such maps for a grain of
sample to calculate mean values and standard deviations                 slag consisting of silicate and oxide phases are shown in
(Table II). There is good agreement between gamma ray                   Figure 3 and Figure 4. In Figure 3 the silicate phases are
emissions and the combined thorium and uranium                          characterized by high concentrations of iron and manganese
concentrations of the samples (Figure 2). A regression of               and a relatively low titanium content. The observations may
the data gives:                                                         be summarized as follows:
where NC is the normalized gamma-ray emission in units
of counts.gram -1.seconds -1. It is important to note that
Equation [1] is strictly valid for the slag samples that were
analysed and that it cannot be applied to other materials.
  Reconciliation of the radioactivity of the slags with the
concentrations of radioactive elements was not undertaken.
There can be no doubt that the radioactive decay series in
the slags are not in equilibrium owing to factors such as the
volatilization of some of the daughter elements during

    Phase-chemical characterization of the slag
X-ray diffractometry
Powder X-ray diffraction was used to identify the phases
present in the samples. In all the samples, the most                    Figure 2. Gamma ray emissions from the samples as a function of
abundant phase belongs to the M 3O 5 multi-component                           the combined uranium and thorium concentrations

8                                                                                                        HEAVY MINERALS 2007
900000                               300               Sr                1500
                                            Ti                                Tal
                                                             840000                               280                                 1400
                                                             780000                               260                                 1300
                                                             720000                               240                                 1200
                                                             650000                               220                                 1100
                                                             600000                               200                                 1000
                                                             540000                               180                                  900
                                                             480000                               160                                  800
                                                             420000                               140                                  700
                                                             360000                               120                                  600
                                                             300000                               100                                  500
                                                             240000                                                                    400
                                                             180000                                80
                                                                                                   60                                  300
                                                100 μm       120000                                                                    200
                                                              60000            100 μm              40                100 μm
                                                                                                   20                                  100
                                                                  0                                                                      0

       Fe                180000            Mn               900000                              1500                                  9000
                         168000                                               Y                                    Nb
                                                            840000                              1400                                  8400
                         156000                             780000
                         144000                                                                 1300                                  7800
                                                            720000                              1200                                  7200
                         132000                             650000                              1100                                  6600
                         120000                             600000
                         108000                                                                 1000                                  6000
                                                            540000                               900                                  5400
                          96000                             480000
                          84000                                                                  800                                  4900
                          72000                             360000                               700                                  4200
                          60000                             300000                               600                                  3600
                          48000                             240000                               500                                  3000
                          36000                             180000                               400                                  2400
        100 μm            24000
                                           100 μm           120000                               300                                  1800
                          12000                              60000             100 μm            200                100 μm            1200
                              0                                  0                               100                                   800
                                                                                                   0                                     0

Figure 3. Backscattered electron image (a) and distribution maps
    of titanium (b), iron (c) and manganese (d), to illustrate the     Figure 4. continued. Elemental maps of tantalum (a), strontium
  location of silicate and oxide phases in a fragment of slag from     (b), yttrium (c) and niobium (d) in silicate- and titanium oxide-
sample F3. In the backscattered electron image the silicate phases                  phases in the grains shown in Figure 3
 are black in colour and the titanium oxides are grey. Elemental
     maps were composed using the PIXE facility at iThemba
                             Laboratories                                                       6000                                  3000
                                                                               Ce                                   La
                                                                                                5600                                  2800
                                                                                                5200                                  2600
                                                                                                4800                                  2400
                                                                                                4400                                  2200
                                                                                                4000                                  2000
        Thi                600                                  60                                                                    1800
                                                 Ui                                             3200                                  1600
                           560                                  56                              2800
                           520                                                                                                        1400
                                                                52                              2400                                  1200
                           480                                  48                              2000                                  1000
                           440                                  44                              1600
                           400                                                                                                         800
                                                                40                              1200
                           360                                                                                                         600
                                                                36                100 μm         800                                   400
                           320                                  32                               400                    100 μm
                           280                                  28                                 0
                           240                                                                                                           0
                           200                                  20
                           160                                  16
                           120                                  12
         100 μm             80                    100 μm         8         Figure 4. (continued). Elemental maps of cerium (a) and
                            40                                   4
                             0                                   0          lanthanum (b) in silicate and titanium oxide phases in
                                                                                         the grains shown in Figure 3
                           6000                 Ba             1200
                           5600                                1120
                           5200                                1040
                                                                           while others such as U, La and Ba are concentrated in
                           4000                                 800        localized regions within silicate grains.
                           3600                                 720
                           3200                                 640      The partitioning of Th, U, Y, Zr, Nb, Ta, Ba and Ce into
                           2800                                 560
                           2400                                 480   silicate phases is noteworthy. These elements have large
                                                                      atomic masses and large ionic radii. This is also a
                           1200                                 240   characteristic of the radioactive daughter elements
         100 μm             800                  100 μm         160
                            400                                  90   generated by the decay series of thorium and uranium. By
                              0                                   0   implication these elements should also concentrate in the
                                                                      silicate phases, rather than in the oxide phases present in the
Figure 4. Elemental maps of thorium (a), uranium (b), zirconium          The PIXE data for chromium are more ambiguous.
  (c) and barium (d) in silicate and titanium oxide phases in the
                                                                      Several grains of slag containing silicate and oxide phases
 grains shown in Figure 3. Maps were composed using the PIXE
                facility at iThemba Laboratories
                                                                      were mapped; in one instance chromium showed a slight
                                                                      preference for silicate phases but in other cases the
                                                                      distribution was more uniform. Microprobe data (see
                                                                      below) are in agreement with the latter scenario, indicating
  • All the trace-elements detected by the instrument (Th,            a slight preference of chromium for M3O5 oxide phases.
    U, Y, Zr, Nb, Ta, Ba and Ce) partition preferentially                The thorium content of the silicate phases was found to
    into silicate phases (Figure 4)                                   be approximately 4 500 ppm. This means that the thorium
  • Some of the elements, such as Th, Zr, Nb and Ta, are              content of the titanium oxide phases must be below
    uniformly distributed throughout the silicate phases              220 ppm (the bulk thorium content of the sample). This

THORIUM IN HIGH-TITANIA SLAG                                                                                                               9
constrains the distribution coefficient for thorium between                               per analysis; this was considered to be counter
silicate and oxide phases (D Th                 = wt % Th in                              productive in terms of the objectives of the work
silicate/wt % Th in oxide) to be greater than 20. The                                   • The volatiliszation of sodium during analysis
approximate uranium content of the silicate phases is about                             • Complicated interelement correction procedures due to
100 ppm but the distribution coefficient could unfortunately                              the high thorium and zirconium concentrations in the
not be estimated because the uranium content of the bulk                                  phases.
sample was below the detection limit of the measurement                                In most slags more than one type of silicate could be
technique.                                                                           identified on the basis of the concentrations of SiO2, MnO
                                                                                     and FeO. The silicate phases are characterized by complex
Electron microprobe analysis                                                         chemical compositions and a lack of crystalline textures,
Electron microprobe analysis using wavelength dispersive                             i.e., they have the chemical and morphological
X-ray spectroscopy (WDS) was performed to quantify the                               characteristics of glass. The presence of silicate glass in the
thorium content of the silicate phases present in each of the                        samples is consistent with the presence of amorphous
samples (the uranium concentration in the silicate phases                            material detected by XRD. The glass represents silicate
was similar to the detection limit and it was not measured).                         melt that coexists with the titanium oxide melt at furnace
The detection limit for thorium is about 500 ppm (for a                              operating temperatures.
counting time of 300 seconds); well below the                                          Silica-poor glass is characterized by high concentrations
concentration level in the silicate phases, but above the                            of FeO and MnO. There is also a positive correlation
concentration in M3O5 titanium-oxide phases. The thorium                             between the concentrations of MnO and FeO in the silica-
content of the latter phases could therefore not be                                  poor glass. Texturally, silica-rich glass occurrs as rounded
measured.                                                                            droplets within the silica-poor glass (Figure 5), suggesting
  With rare exceptions, between 5 and 10 points of each                              immiscibility of two silicate melts at high temperatures.
phase were analysed per sample. Mean values and standard                             The ThO2 and ZrO2 concentrations in the different silicate
deviations for silicate and oxide phases are reported in                             phases in a particular slag are inversely proportional to the
Table III and Table IV respectively. Several factors may                             silica content, i.e., thorium and zirconium partition
contribute to the low analytical totals obtained for some of                         preferentially to the silica-poor glass. These relationships
the analyses:                                                                        are illustrated by the data for samples E2 and H4 in
   • The presence of sulphur (up to about 4% by weight in                            Table IIII (owing to the poor optical imaging on the
     extreme cases), potassium (maximum of about 1% by                               instrument used for the WDS analysis it was not possible to
     weight), and other trace elements such as niobium in                            distinguish between different silicate phases in the slag.
     the silicate phases. To include these elements would                            Consequently, no analyses of the silica-poor phase were
     have extended counting times well beyond 5 minutes                              obtained for B4, F3 and G3).

                                                                   Table III
              Electron microprobe analyses of silicate phases in the slag samples. (standard deviations are given in parentheses)

*Analyses of the silica-poor phase in E2 are incomplete due to instrumental difficulties at the time of analysis

10                                                                                                                    HEAVY MINERALS 2007
Table IV
      Electron microprobe analyses of iron-titanium oxide phases in the slag samples. (standard deviations are given in parentheses)

                                                                           The oxide phases contain MgO in quantities similar to
                                                                        those present in the coexisting silicate phase but virtually
                                                                        no CaO. To the extent that this signifies a tendency for
                                                                        alkali earth metals with large ionic radii to partition into
                                                                        silicate phases, this is highly significant (such a tendency is
                                                                        in agreement with the PIXE results that show preferential
                                                                        partitioning of barium to the silicate phases). Radium, an
                                          Fe-Ti-oxide                   alkali earth metal considered to be an important source of
                                                                        radioactivity in slag, has an ionic radius of 1.40 Å, much
          Silica-rich phase                                             larger than the radii of Ba2+ and Ca2+ (1.35 Å and 0.99 Å
                                                                        respectively). The observed partitioning behaviour of MgO,
                       Metal                                            CaO and BaO suggests that RaO also will partition to the
                                                                        silicate phases present in the slag, but even more so than the
                                        Silica-poor phase               lighter alkali earth metals.

                                                                          Distribution of elements between silicate and
                                                                                           oxide phases
                                                                        The distribution of elements between silicate and oxide-
                                                                        phases is a reflection of the thermodynamic behaviour of
   Figure 5. Backscattered electron image illustrating textural         the system. The distribution coefficient for element i, Di, is
    relationships between different silicate phases in slag H4.         defined as:
       Droplets of a silica-rich phase are surrounded by a                                                                             [2]
                          silica-poor phase
                                                                          The value of D i , depends on variables such as
                                                                        temperature, oxygen partial pressure (i.e., slag-grade) and
   The high thorium content of the silicate phases is one of            melt composition. If there is a sufficient dependence on
the most important observations of this study. It is                    melt composition it would, in principle, be possible to
important to note that the concentration levels of thorium in           manipulate the behaviour of impurities during smelting. In
silica-poor glasses are still well below the values at which            the context of this study it would be highly desirable to
silicate melts (glasses) become saturated with thorium.                 calculate D Th for each of the slags. Due to the low
High alkali-borosilicate glasses quenched from 1250°C                   concentration of thorium in the oxide phases it was
contain between 2% and 5% Th (2.3%–5.7% ThO 2 by                        unfortunately not possible to measure DTh directly, although
weight) in solution3; data cited by Farges 4 suggest similar            minimum values were estimated from mass-balance
solubility levels in alumino-silicate glasses quenched from             constraints.
the same temperature. He also noted an increase in the                    Compositional data for silicate- and oxide-phases were
solubility of thorium with increasing temperature. The                  used to calculate DMn, DMg, DAl, DFe, DTi and DCr for each
silicate glasses in the high-titanium slags are therefore not           of the slags (Table V). Di values for Ca and Si were not
expected to be saturated in thorium (the reason for the                 calculated since the concentrations of these elements in the
difference in thorium content of the different silicate                 M3O5 solid solutions were not accurately measured. Data
glasses will be discussed in a later section).                          for silicate phases with different compositions in the same
   The distribution of thorium between silicate and oxide-              slag were used to calculate average values for the D i.
phases will be reported later; as an initial observation it is          Where data for silica-poor glass were not available,
sufficient to note that thorium has a strong preference for             inequalities were used to indicate the bias (samples B4, E2,
silicate phases.                                                        F3 and G3).

THORIUM IN HIGH-TITANIA SLAG                                                                                                           11
Table V
                        Average distribution coefficients (silicate oxide) for Mn, Mg, Al, Fe, Ti and Cr for each of the slags
                                          (the range over which Di varies is reflected as an uncertainty)

Distribution coefficients for B4, E2, F3 and G3 are not well constrained due to a lack of analytical data. Symbols are used to indicate the bias resulting from
the lack of data.

                                                                   Table VI
                  Calculated mass-percentage of silicate and oxide phases in each slag and the calculated thorium distribution

‘Bulk Th’ is the bulk Th content of the slag (from Table I)
‘Th concentration in silicates’ is the average Th concentration in silicate phases determined by microprobe
‘Th in slag contained in silicates’ is the amount of thorium contained in the silicate phases in the slag
‘Th in slag contained in oxides’ is the amount of thorium contained in the oxide phases in the slag
‘Th conc. in oxides’ indicates the maximum Th concentration in oxide phases calculated from the Th distribution
DTh is the minimum value of the silicate-oxide distribution coefficient for Th
The Th distribution in E2 and H4 are subject to large uncertainties due to the large difference in the Th content of the silicate phases
Calculations are based on semi-quantitative PIXE data

   Elements may be classified according to their partition                                   The chemical behaviour of thorium in
coefficients:                                                                                         high-titania slag
    • Silicon, calcium (and thorium) have large partition
      coefficients which means that they are strongly                                 Distribution of thorium between silicate and oxide-
      concentrated in the silicate phases. Manganese also                             phases
      prefers silicate phases but its partition coefficient is                        Although the distribution coefficient could not be measured
      considerably smaller (about 10)                                                 directly, it is possible to constrain the distribution of
    • Iron, magnesium, chromium and aluminium have                                    thorium through calculation.
      distribution coefficients that are around unity, which                             First, the weight fractions of silicate and oxide phases in
      means that they are evenly distributed between silicate                         each slag are calculated. The average concentrations of
      and oxide phases                                                                MnO, SiO2, TiO2 and CaO in silicate and oxide phases are
    • Titanium has a very small distribution coefficient                              used for this purpose because of the magnitudes of the
      (~ 1/100), which means that it is strongly concentrated                         silicate-oxide distribution coefficients of these elements.
      in oxide phases.                                                                Next, the amount of thorium reporting to silicate phases (in
                                                                                      μg/g slag) is calculated from the weight percentage silicate
   The distribution coefficients indicate how the removal of
                                                                                      phases in the slag and the average thorium concentration in
silicate phases from the slag will affect the composition of                          the silicate phases. Standard error-propagation techniques
the oxide residue. Elements that preferentially partition to                          are used to calculate the uncertainties in these quantities.
silicate phases (Di > 1) will be depleted in the residue (i.e.,                       Within the uncertainty of the data, the amount of thorium
the composition of the residual oxide slag will be upgraded                           reporting to the silicate phases is enough to account for the
through the removal of these elements), while elements for                            mean total thorium content of the slags and, as a result, only
which Di < 1 will be enriched in the residue (providing                               a maximum value for the amount of thorium reporting to
insoluble oxides do not precipitate during leaching).                                 oxide phases can be calculated (this being the difference
Leaching of silicate phases should therefore result in a                              between the total thorium content of the slag and the
decrease in the concentrations of SiO2, CaO and Th (to a                              amount of thorium present in silicate phases—taking the
lesser extent also MnO), and an increase in the                                       uncertainties in the data into account). Following the
concentration of TiO 2 in the oxide residues. The                                     calculation of a maximum value for the amount of thorium
concentrations of FeO, MgO, Al2O3 and Cr2O3 in the oxide                              present in oxide phase, a minimum value for the thorium
residues will not be significantly affected by the removal of                         distribution coefficient can be calculated. Results of these
silicate phases.                                                                      calculations are reported in Table VI.

12                                                                                                                             HEAVY MINERALS 2007
The calculation is complicated by the large difference in           Using thermodynamic data of Barin6, the equilibrium
thorium content of the silica-rich and silica-poor glasses. In      oxygen partial pressure for reaction 1 at 1900 K is 10-9.3 bar
the case of slags B4 and G3, calculation of the distribution        (this is about 1.3 log10-units more reducing than the Fe-FeO
coefficient was not attempted because of a lack of data for         buffer). The oxidation of solid Ti4O7 to TiO2 (rutile) at
the thorium-rich, silica-poor, glass. Although WDS data are         1900 K takes place at a very similar oxygen partial pressure
not available for F3 either, semi-quantitative PIXE analyses        (10-9.2 bar), indicating that solid Ti3+-bearing oxide phases
for this sample are consistent with a value for the thorium         are stabilized at oxygen partial pressures approximately 7
distribution coefficient of at least 20. Consistently large         log 10 -units more oxidizing than the Ti-TiO 2 buffer.
distribution coefficients for thorium lead to an important          Assuming similar behaviour in the thorium-oxygen system,
conclusion:                                                         the formation of Th 3+-bearing oxide phases at 1900 K
   • The removal of silicate phases will also remove the            might be expected at oxygen partial pressures of about 10-17
      bulk of the thorium and cause a significant decrease in       bar. This is about 8 log10-units more reducing than the
      the thorium content of the oxide residue.                     minimum oxygen partial pressure likely to be attained
                                                                    during smelting (10-9.3 bar). Although these calculations are
  It is informative to consider the significance of a large
                                                                    not directly applicable to the oxidation state of multi valent
distribution coefficient for thorium between silicate and
                                                                    cations in melt phases, the data suggest that the ratio of
oxide phases. Consider a slag containing 5 wt% silicate
                                                                    Th3+ to Th4+ is likely to be very low, even under the most
phases and a bulk Th content of 200 ppm. Assuming
  Silicate–oxide                                                    reducing conditions expected during smelting.
DTh              = 100, it follows that the Th concentration in
                                                                      The structural environment around Th4+ in silicate glasses
the silicate and oxide phases will be 3360 ppm and 33.6
                                                                    containing between 1% and 3% (by weight) Th 4+ was
ppm respectively. Therefore, if the silicate phases could be
                                                                    investigated by Farges4 using extended X-ray absorption
fully leached from the slag, it should be possible to reduce
                                                                    fine structure (EXAFS) spectroscopy. Several conclusions
the thorium content by at least 84%.
                                                                    from the work by Farges4 are directly applicable to current
  A large value for D Th is consistent with the general
incompatible behaviour of thorium during the
crystallization of magma with different compositions5 (an              • Th4+ is mainly six co-ordinated by oxygen with a mean
element is said to be ‘incompatible’ if it is concentrated in            bond length, ⟨ VI Th-O.⟩ of about 2.3 Å. Eight-co-
the residual melt once different oxide and silicate minerals             ordinated Th4+ (⟨VIIITh-O⟩ ~ 2.4 Å) is present at high
start to crystallize during cooling). Unfortunately, there are           thorium concentration levels (3%) and/or when
no data for D Th between silicate melt and M 3O 5 oxide                  polymerization increases. Mean metal-oxygen bond
phases in the geological literature that could be used for               lengths for the two six-co-ordinated cation sites (M1
comparison with the results of this study.                               and M2) in orthorhombic Ti3O5-rich solid solutions
                                                                         ⟨VIM1-O⟩ and ⟨VIM2-O⟩ are approximately 2.04 Å and
                         Discussion                                      2.01 Å, respectively4. The much shorter metal-oxygen
                                                                         bond lengths in Ti3O5-rich solid solutions mean that
The silicate phases in the slags were previously identified              thorium is unlikely to partition into oxide phases
as quenched silicate melt (i.e., glass). It is therefore relevant        crystallizing from the slag during cooling
to consider the chemistry of thorium in silicate melts as it           • The amount of six-co-ordinated Th4+ (VITh4+) increases
may suggest options for optimizing the amount of thorium                 in the presence of major quantities of alkali cations
reporting to the silicate melt phases during smelting. The               such as Na+ and K+. Should the addition of such fluxes
most stable oxidation state for thorium is +4; with respect              be practical, it will facilitate the collection of thorium
to thorium metal, it is also one of the most stable oxides.              by silicate melt phases during smelting
The stability of ThO2 relative to Th plus oxygen is evident            • The bond strength around six-co-ordinated Th 4+ in
from the Gibbs free energy of formation of ThO 2 ;                       silicate melt or glass is high relative to that in Th-
approximately -870 KJ.mol -1 at 1900 K (1627°C). By                      bearing minerals such as ThO 2 and ThSiO 4 . This
comparison, the Gibbs free energy of formation of TiO2                   explains in part why these phases are not observed in
(relative to Ti and O2) at 1900 K is about -603 KJ.mol-1. In             the slags, even at high bulk thorium concentration
terms of oxygen partial pressure (PO2) it means that ThO2                levels (> 300 ppm)
will form when the PO2 exceeds 10-23.8 bar, while TiO2 will            • Increasing polymerization of the silicate glass or melt
(form from Ti and O2) when the PO2 exceeds 10–16.5 bar. A                (high concentrations of SiO2), will favour higher co-
practical implication of this result is that metallic thorium            ordinations around Th4+ due to a lack of non-bridging
will not form during smelting and the Th content of the pig              oxygens (NBOs) to which thorium preferentially
iron will be negligible.                                                 bonds. This explains why the silica-rich glass phase
  It is not known what the effective oxygen partial pressure             contains less thorium than the co existing, low silica,
during smelting is; it will also be different for the different          glass phase in the slags.
slags (hence the differences in Fe/Ti ratio of the slags).
Phase-chemically, the composition of the M3O5 phases that             To summarize: data on the chemical behaviour of
crystallize from iron-titanium oxide slag may effectively be        thorium in silicate melts and glasses support the mass-
described in terms of the FeTi 2O 5-Ti 3O 5 solid solution          balance calculations according to which the amount of
series. The activity of Ti3O 5 in the slag increases with           thorium in silicate glass phases is sufficient to account for
increasing grade, and when all the Fe 2+ is reduced to              the total thorium content of the slags.
metallic iron (Fe0) the activity of Ti3O5 is almost unity. An
indication of the minimum oxygen partial pressure during                                    Summary
smelting may therefore be obtained from the equilibrium             Elements with a large ionic radius preferentially partition to
between solid Ti3O5 and TiO2 (rutile):                              the immiscible silicate melt that forms during smelting.
                                                             [3]    These elements include thorium (which is present in such
                                                                    large quantities that its concentration in silicate phases

THORIUM IN HIGH-TITANIA SLAG                                                                                                    13
could be measured by electron microprobe), zirconium,                                   References
barium, lanthanum, niobium, cerium (the distribution of
these elements was determined by PIXE), and possibly also         1. RGC Mineral Sands Ltd., S.A. Patent no. 93/5474, 29
the radioactive daughter elements of the thorium- and                July 1993.
uranium-decay series.                                             2. BESSINGER, D., DU PLOOY, H., PISTORIUS,
   Mass balance calculations suggest that the thorium                P.C., and VISSER, C. Characteristics of some high
content of the silicate phases is high enough to account for         titania slags. Heavy Minerals. R.E. Robinson. (ed.)
the total thorium content of the slags. This is also reflected       Johannesburg, South African Institute of Mining and
by the silicate-oxide distribution coefficients for Th that are      Metallurgy, 1997. pp. 151–156.
overwhelmingly in favour of the silicate phases. Depending
on the total silica content, silicate phases constitute           3. ELLER, P.G., JARVINEN, G.D., PURSON, R.A.,
approximately 4 to 8 weight per cent of the slags that were          PENNEMAN, R.R., LYTLE, F.W., and GREEGOR,
selected for test work.                                              R.B. Actinide abundances in borosilicate glass.
   Spectroscopic data in the literature are consistent with the      Radiochimica Acta, vol. 39, 1985. pp. 17–22
partitioning of thorium into depolymerized (low-silica)           4. FARGES, F. Structural environment around Th4+ in
glasses and melts relative to crystalline phases and more            silicate glasses: Implications for the geochemistry of
polymerized glasses and melts. These results are in                  incompatible M 4+ elements. Geochimica et
agreement with the observations of this study and confirm            Cosmochimica Acta, vol. 55, 1991. pp. 3303–3319.
that silicate phases contain most of the thorium present in
                                                                  5. MAHOOD, G.A. and HILDRETH, W. Large partition
the slag.
   In principle, the radioactivity levels of high titania slag       coefficients for trace elements in high silica rhyolites.
can be lowered by the removal (physical or chemical) of              Geochimica et Cosmochimica Acta, vol. 47, 1983.
silicate phases from the slag. Disposal of the radioactive,          pp. 11–30.
silicate fraction will be onerous and the economics of such       6. BARIN, I. Thermochemical Data of Pure Substances,
a step are questionable.                                             Parts I and II. VCH Verlagsgesellschaft. 1989.

14                                                                                               HEAVY MINERALS 2007
You can also read
NEXT SLIDES ... Cancel