XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 1717
mother liquor filtered off and the recovered crystals were
washed with washing solution to remove any adsorbed
impurities. The wash solution was made from an acetone
and Millipore water mixture with the same final O/A ratio
as for the experiment (similar composition to the metal free
mother liquor). The crystals were air dried in the fume hood
to ensure minimal artefacts arose during drying, before
being used for further analysis. The powder samples were
analysed with SEM-EDS (Hitachi 3700N with Bruker
X-Sense operating at 5 keV) for morphology and elemental
composition, whilst PXRD (Panalytical XPert Pro) was car-
ried out for crystal phase identification.
RESULTS AND DISCUSSION
The effects of O/A ratio and acetone concentration on vana-
dium crystallization exhibited complex micro and macro
scale effects linked to the supersaturation levels. Some pre-
liminary insights into the expected behaviour of the system
were acquired through thermodynamic modelling using
OLI Studio 11.5 (OLI Systems, Inc. Cedar Knolls, USA).
Experimental results (Figure 2) from varying the O/A ratio
for the impure system at different acetone concentrations
showed that, at O/A ratio =0.5 and antisolvent diluted to
80% acetone (v/v), the V concentration remained fairly
constant (~28%) throughout the run. More concentrated
acetone (90–100%) resulted in increased V quantities in
the solid phase with increase in tb but the pure acetone
exhibited the fastest response (higher V quantities over the
same tb). At the higher O/A ratio (0.75) and antisolvent
diluted to 80% acetone (v/v), the V quantities in the solid
phase only showed notable increase at tb ≤ 2 h when com-
pared to the lower O/A ratio (0.5). It can be inferred that
the effects of dilution are more pronounced at the onset of
the batch but are soon superseded by the O/A ratio. In gen-
eral, pure acetone 100% (v/v) tends to support maximum
recovery of vanadium but tends to be least selective towards
impurities.
A closer look at the solid-liquid system behaviour, in
which vanadium salts are crystallizing, can be seen through
analysing the yield as shown in Figure 3.
By comparing the vanadium salt yields, at the same
O/A ratio and specific tb, it can be seen that the pure system
with 100% acetone (v/v) results in lower yields than the
impure system. As anticipated, an increase in O/A ratio at
the same acetone concentration had a positive effect on the
yield but the extent of differences, with increase in batch
time, were not as anticipated. The results affirm that the
effects of the O/A ratio and dilution are dominant at lower
ratios and acetone concentrations, as well as the onset of
the batch process. The combination of a lower O/A ratio
and higher dilution decreased the vanadium salt yield with-
out increasing selectivity. This suggests that supersaturation
control alone will not yield the desired effect of increased
selectivity.
The behaviour of other cations within the system
(see Table 1) and how they impact purity is presented in
Figure 4. The purity of the vanadium salts remained fairly
constant at a set O/A ratio and acetone concentration with
the exception of the later batch times (4 ≤ tb ≤ 6) at O/A
ratio of 0.75 and 90% acetone (v/v). It is hypothesized that
the increase in Al and Ca in the solid phase might have
slowed down the V conversion due to increased competi-
tion for the crystallizing resource.
The percentage purity generally decreases with batch
time for the lower O/A ratio of 0.5, but differences are
more significant at O/A ratio of 0.75. Taking both purity
and yield into account indicates that, at lower O/A ratio
of 0.5 and less concentrated acetone (80%), an optimum
batch processing time would be 4 ≤ tb ≤ 6, and that oper-
ating the crystallizer beyond this period is envisaged to
increase the yield.
An approximation of how changes in the aqueous
chemistry impacted the crystal habit were explored using
light microscopy (Figure 5).
At the O/A ratio of 0.75 with diluted antisolvent, the
crystal habit was impacted as presented in Figure 5. The
trend, for the pure and impure system at tb ≤ 2, is similar to
that at the lower O/A ratio (0.5) but light microscope vis-
ibility and crystal stability was quickly attained (circa 2 h).
However, the crystals are smaller, and the acicular crystals
appeared earlier as anticipated for systems in which higher
supersaturation levels prevail. Higher batch times (2 h ≤ tb
≤ 6 h) saw dominant crystal habit of the pure system transi-
tion from hexagonal laths to ellipsoidal and then to acicu-
lar particles whilst the most dilute (80% acetone) impure
system appeared to have a delayed onset of a similar trend.
The acicular nature of the crystals habit results from the
light microscopy were further investigated to establish the
stability of this habit by using SEM. The results (Figure 6)
for the filtered solids taken at the end of the batch corrobo-
rated this needle-like crystal habit and its dominance at the
higher O/A ratio of 0.75.
Powder-XRD identified the final crystal product as
predominantly hydrated metamunirite (NaVO3∙2H2O) for
the impure system. It is important to note that other vana-
dium salts in which the impurities appeared were picked
mother liquor filtered off and the recovered crystals were
washed with washing solution to remove any adsorbed
impurities. The wash solution was made from an acetone
and Millipore water mixture with the same final O/A ratio
as for the experiment (similar composition to the metal free
mother liquor). The crystals were air dried in the fume hood
to ensure minimal artefacts arose during drying, before
being used for further analysis. The powder samples were
analysed with SEM-EDS (Hitachi 3700N with Bruker
X-Sense operating at 5 keV) for morphology and elemental
composition, whilst PXRD (Panalytical XPert Pro) was car-
ried out for crystal phase identification.
RESULTS AND DISCUSSION
The effects of O/A ratio and acetone concentration on vana-
dium crystallization exhibited complex micro and macro
scale effects linked to the supersaturation levels. Some pre-
liminary insights into the expected behaviour of the system
were acquired through thermodynamic modelling using
OLI Studio 11.5 (OLI Systems, Inc. Cedar Knolls, USA).
Experimental results (Figure 2) from varying the O/A ratio
for the impure system at different acetone concentrations
showed that, at O/A ratio =0.5 and antisolvent diluted to
80% acetone (v/v), the V concentration remained fairly
constant (~28%) throughout the run. More concentrated
acetone (90–100%) resulted in increased V quantities in
the solid phase with increase in tb but the pure acetone
exhibited the fastest response (higher V quantities over the
same tb). At the higher O/A ratio (0.75) and antisolvent
diluted to 80% acetone (v/v), the V quantities in the solid
phase only showed notable increase at tb ≤ 2 h when com-
pared to the lower O/A ratio (0.5). It can be inferred that
the effects of dilution are more pronounced at the onset of
the batch but are soon superseded by the O/A ratio. In gen-
eral, pure acetone 100% (v/v) tends to support maximum
recovery of vanadium but tends to be least selective towards
impurities.
A closer look at the solid-liquid system behaviour, in
which vanadium salts are crystallizing, can be seen through
analysing the yield as shown in Figure 3.
By comparing the vanadium salt yields, at the same
O/A ratio and specific tb, it can be seen that the pure system
with 100% acetone (v/v) results in lower yields than the
impure system. As anticipated, an increase in O/A ratio at
the same acetone concentration had a positive effect on the
yield but the extent of differences, with increase in batch
time, were not as anticipated. The results affirm that the
effects of the O/A ratio and dilution are dominant at lower
ratios and acetone concentrations, as well as the onset of
the batch process. The combination of a lower O/A ratio
and higher dilution decreased the vanadium salt yield with-
out increasing selectivity. This suggests that supersaturation
control alone will not yield the desired effect of increased
selectivity.
The behaviour of other cations within the system
(see Table 1) and how they impact purity is presented in
Figure 4. The purity of the vanadium salts remained fairly
constant at a set O/A ratio and acetone concentration with
the exception of the later batch times (4 ≤ tb ≤ 6) at O/A
ratio of 0.75 and 90% acetone (v/v). It is hypothesized that
the increase in Al and Ca in the solid phase might have
slowed down the V conversion due to increased competi-
tion for the crystallizing resource.
The percentage purity generally decreases with batch
time for the lower O/A ratio of 0.5, but differences are
more significant at O/A ratio of 0.75. Taking both purity
and yield into account indicates that, at lower O/A ratio
of 0.5 and less concentrated acetone (80%), an optimum
batch processing time would be 4 ≤ tb ≤ 6, and that oper-
ating the crystallizer beyond this period is envisaged to
increase the yield.
An approximation of how changes in the aqueous
chemistry impacted the crystal habit were explored using
light microscopy (Figure 5).
At the O/A ratio of 0.75 with diluted antisolvent, the
crystal habit was impacted as presented in Figure 5. The
trend, for the pure and impure system at tb ≤ 2, is similar to
that at the lower O/A ratio (0.5) but light microscope vis-
ibility and crystal stability was quickly attained (circa 2 h).
However, the crystals are smaller, and the acicular crystals
appeared earlier as anticipated for systems in which higher
supersaturation levels prevail. Higher batch times (2 h ≤ tb
≤ 6 h) saw dominant crystal habit of the pure system transi-
tion from hexagonal laths to ellipsoidal and then to acicu-
lar particles whilst the most dilute (80% acetone) impure
system appeared to have a delayed onset of a similar trend.
The acicular nature of the crystals habit results from the
light microscopy were further investigated to establish the
stability of this habit by using SEM. The results (Figure 6)
for the filtered solids taken at the end of the batch corrobo-
rated this needle-like crystal habit and its dominance at the
higher O/A ratio of 0.75.
Powder-XRD identified the final crystal product as
predominantly hydrated metamunirite (NaVO3∙2H2O) for
the impure system. It is important to note that other vana-
dium salts in which the impurities appeared were picked