You are on page 1of 6

Light Metals 2005 Edited by Halvor Kvande TMS (The Minerals, Metals & Materials Society), 2005

BOEHMITE vs. GIBBSITE PRECIPITATION


Joanne Loh, Chris Vernon, Melissa Loan, Greta Brodie
AJ Parker CRC for Hydrometallurgy (CSIRO Minerals); P.O. Box 90; Bentley, Western Australia, 6982, Australia

Keywords: Gibbsite, Boehmite, Growth Mechanism, Organic Impurities

Abstract precipitation in comparison with gibbsite precipitation. The effect


of common refinery variables (such as caustic concentration,
Boehmite (monohydrate) precipitation, in place of gibbsite initial supersaturation, inorganic impurities), solution cations and
(trihydrate), has been discussed in the literature as a possible model organic compounds were also investigated.
energy-saving step in alumina refining. Indeed, the calcination of
monohydrate rather than trihydrate would save approximately Experimental
12% of the energy consumed in an alumina refinery. These
previous investigations have not focused on the economic and Seed Material and Experimental Conditions
technical feasibility of the process. The current investigation takes
a pragmatic approach, measuring precipitation rates, determining Table I lists the median particle size and surface areas of the seed
product phase and particle size distribution, and assessing the material used in the present work.
impact of impurities and probe species in an attempt to determine
the viability and possible improvements to such a process. The Table I. Median Particle Size (D[v, 0.5]) and Surface Area of
present work indicates that boehmite precipitation rates are Boehmite and Gibbsite Seed Material
approximately 2 orders of magnitude slower than for gibbsite Seed Material D[v, 0.5] (Pm) Surface Area (m2 g-1)a
under identical conditions. Agglomeration seems not to occur. Boehmite 6.02 3.9
The particle size distribution is too fine and the product is Gibbsite 86 0.11
a
unsuitable for calcination. However, the work has further measured by 5 point N2 BET
improved the fundamental understanding of gibbsite precipitation
and a hypothesis for the mechanism will be presented. Boehmite seed material was prepared by a hydrothermal reaction
of gibbsite with water at 200 °C (4 h) in an autoclave. The
Introduction product was pressure filtered through 0.45 Pm Supor membrane,
washed with hot deionised water and dried overnight (60 °C).
Gibbsite (Al(OH)3) is the usual precipitation product from Bayer Boehmite was the only phase as determined by powder X-ray
alumina refineries. Gibbsite undergoes calcination to produce the diffraction. Scanning electron microscopy (SEM) revealed large
final product alumina (Al2O3) (1). It has been estimated that the hollow particles, artefacts of the hydrothermal reaction consisting
energy required to calcine gibbsite to alumina is ~ 2 GJ t-1 Al2O3 of diamond shaped particles (Figure 1a) with smaller aggregates
[1]. This represents ~ 18 % of the total energy consumption of an of diamond shaped particles (Figure 1b).
alumina refinery (based on an average total energy consumption
of ~ 11 GJ t-1 Al2O3 [2]). In order to make valid comparisons between phases, gibbsite-
seeded experiments were performed under corresponding
2Al(OH)3 o Al2O3(s)  3H 2O(g) (1) conditions for boehmite precipitation. Gibbsite (C31 hydrate,
Alcoa World Alumina) was used as gibbsite seed (Table I). Due
Substantial energy savings achieved through the precipitation and to the very low surface area, matching the total boehmite seed
calcination of boehmite (AlOOH), in place of gibbsite, was surface area at a seed loading of 230 g L-1 required a
proposed by Filippou and Paspaliaris [3]. Calculated from corresponding gibbsite seed loading of 8150 g L-1. The resulting
standard thermodynamic data, the energy required to convert gibbsite precipitation rate would be too fast for a sensible
boehmite to alumina is ~ 34 % of the energy required for gibbsite comparison thus, gibbsite was also used at a seed loading of
conversion. This translates to an energy saving of 12 % of the 230 g L-1 and appropriate allowances made when comparing rates.
total energy consumption equating to a saving of AUS$60 M per Gibbsite seed consists of large aggregates of hexagonal crystals
annum in Australia alone (based on average cost and energy (Figure 1c).
consumption of Australian refineries).
Synthetic sodium aluminate liquors were prepared by dissolving
On the basis of this preliminary analysis, the prospect of replacing gibbsite (C31 hydrate) into a heated NaOH solution (~ C600) at
gibbsite precipitation with boehmite precipitation in the Bayer atmospheric pressure. The solution was pressure filtered through
process is clearly worth exploring. While much research into 0.45 Pm Supor membrane and the organic additives (as required)
boehmite precipitation has been carried out in the last decade, the were added to the cooled solution prior to the volume being made
focus was not on the economic viability of boehmite precipitation, up (to C200 g L-1 as Na2CO3). The solution was pressure filtered
comparative precipitation rates, nor on the product quality again prior to use.
required by smelters as end users of alumina.
Bayer plant liquor was also used in the present work. The liquor
The present work was undertaken to replicate previous work and was redigested at atmospheric pressure to increase the A/C ratio to
to provide a more commercially focused assessment (kinetics, match the synthetic liquor initial concentrations. Redigestion
product characteristics and product phase) of boehmite occurred at temperature just above boiling (~ 110 °C, 1 h) to

203
reduce the further degradation of the organic material present in Analytical Techniques
the liquor. The plant liquor was pressure filtered through 0.45 Pm
Supor membrane and diluted to the required volume and C. The liquor filtrate was analysed for alumina, caustic and total
caustic concentrations using the Worsley titration method [6].
20 Pm 2 Pm The solids were characterized by X-ray diffraction (XRD),
scanning electron microscopy (SEM) and particle size distribution
(PSD, laser diffraction).

Results and Discussion

Previous Work
a b
The majority of the previous investigations into boehmite
20 Pm precipitation have been carried out by a research group led by
Paspaliaris and Panias. Very few descriptions of the seed source
or general characteristics of the seed used in those experiments
were given. The few SEM images published recently [5, 7] show
very large ‘seed crystals’ with nuclei on the surface. The size of
the ‘seed crystals’ ranged from ~50 to 100 Pm (based on the scale
bar provided) which is much larger than the reported size of the
c seed. The hexagonal and blocky shape suggests that the seed
Figure 1. SEM images of (a) large and (b) small particles of crystals could be gibbsite. The presence of gibbsite in the product
Boehmite seed and (c) Gibbsite seed. was not reported by those authors.

Boehmite precipitation leads to a very high apparent solubility. While high efficiencies (72 g Al2O3 L-1) have been reported for
The alumina concentration in a boehmite-seeded solution, left to boehmite precipitation [8], they were calculated over 48 hours of
precipitate for 216 hours, was approximately 2.3 times higher than precipitation at a seed loading of 1200 g L-1. The median
the measured boehmite solubility [4]. In general, gibbsite exhibits diameter of the products was 10 Pm. The transfer of these
only a slight increase in apparent solubility, approximately 2 % precipitation conditions to industry would significantly decrease
higher than measured gibbsite solubility after 264 hours of both productivity and product quality.
precipitation. This metastability has been noted previously by
Panias [5]. In the present work, repetition of standard boehmite precipitation
experiments initially produced only boehmite, however after 24
Due to the length of the experiments falling around 216 hours, the hours, the product contained significant amounts of gibbsite.
apparent boehmite solubility (A*Boehmite apparent = 79 g Al2O3 L-1, Boehmite was reported as the only product in the literature. A
estimated from experimental results), was used to calculate the doubling of the median particle size through seed recycling (5
initial supersaturation (E=A/A*) rather than the equilibrium cycles of 24 hours, 1200 g L-1 seed loading) was reported [5].
boehmite solubility that is normally used. The experimental However, this was not observed in the present work, where the
conditions which promote boehmite precipitation were as final size of the product observed was only 1.3 times that of the
follows:Eapparent = 1.67, A/C 0.71, C205 at 90 °C. The sodium original seed. Using a coarser boehmite seed (D[v, 0.5]=32.7 Pm)
aluminate solutions used for gibbsite precipitation were prepared [9], particles > 80 Pm in size were produced. Agglomeration was
to this supersaturation for comparison. The solubility of boehmite reported for small particles as well as the larger particles, although
in potassium hydroxide solutions was measured and there was no no conclusive data was reported and no SEM or XRD evidence
significant difference to the reported boehmite solubility in was presented.
sodium hydroxide [4]. Model organic compounds were also
added to the liquor to determine their effect on boehmite Precipitation Kinetics
precipitation in comparison to their known effects on gibbsite
precipitation. It was assumed that the low concentration of the While the literature contains no quantitative comparisons between
organics (mmol L-1) would have no significant effect on the the rates of precipitation of gibbsite and boehmite, kinetic rate
solubility of gibbsite or boehmite. These compounds were used equations for boehmite precipitation have been reported [7, 10].
without further purification. Boehmite precipitation has been reported to be up to one to
several hundred times slower than gibbsite for the same
Experimental Procedure supersaturation [10], although no information about the
experimental conditions was given.
50 mL aliquots of the liquor were pipetted into a number of
250 mL Nalgene polypropylene bottles and equilibrated to the Boehmite and gibbsite seeded precipitation experiments were
desired temperature for 15 minutes in an end-over-end tumbling carried out under identical conditions (except for total seed
bottle roller. Seed was then added to the liquor and the bottle surface areas and seed size, see Seed Material and Experimental
resealed. The bottles were then returned into the bottle roller for Conditions). The gibbsite precipitation rate was significantly
the required time. Sampling consisted of the removal of one faster than the boehmite precipitation rate (Figure 2). Indeed, the
bottle and the whole contents constituted the sample for that time majority of the precipitation of gibbsite had already occurred by
interval. The sample was vacuumed filtered (0.45 mm Supor 24 hours. In addition, the surface area of the gibbsite seed is
membrane) and the solids washed and dried. approximately 3 % of the total surface area of the boehmite seed.

204
Using standard expressions for the precipitation rate for gibbsite, a As expected, a high initial supersaturation (E=4.2 based on
calculated rate constant for gibbsite precipitation is approximately equilibrium boehmite solubility or Eapparent=1.8, A/C 0.7)
200 times higher than for boehmite. increased the initial boehmite precipitation rate. However, the
1.7 increased rate can be attributed to gibbsite formation (gibbsite
Gibbsite accounted for 60 % of the sample at 100 g L-1 seed charge, while
*
1.6 Eapparent=A/A apparent 16 % of the sample was identified as gibbsite at 230 g L-1 seed
Boehmite
Supersaturation E

charge after 24 hours). The high initial supersaturation promoted


1.5


gibbsite nucleation.
1.4
SEM images revealed boehmite crystals adhering onto the surface
1.3 of gibbsite crystals (Figure 3a). Further adhesion of boehmite
1.2 crystals onto those already attached did not occur. Further growth
of the gibbsite occurred around the boehmite particles such that
1.1 *
while boehmite is further incorporated into the gibbsite crystal, it
E=A/A is never completely grown over (Figure 3b). These observations
1.0 indicated that with a high initial supersaturation, boehmite, having
0 25 50 75 100 125 150 175 200 225 250 a different crystal structure to gibbsite, could adhere onto gibbsite.
Time (h) However, as precipitation proceeds and supersaturation decreases,
gibbsite only precipitates onto gibbsite crystals, resulting in an
Figure 2. The comparison of desupersaturation rates in boehmite increase in size, and further boehmite adhesion ceases.
and gibbsite seeded experiments under identical experimental
conditions. 2 Pm 2 Pm

Gibbsite was the dominant phase from the gibbsite-seeded


experiment with very small amounts of boehmite (< 1 %).
Gibbsite nuclei agglomerated to form large aggregates (> 20 Pm).
The product from boehmite precipitation experiments was
predominantly boehmite with trace amounts of gibbsite (< 1 %).

Effect of Caustic Concentration a b


Figure 3. SEM images of (a) gibbsite nuclei covered with
Increasing the caustic concentration (at constant supersaturation boehmite particles and (b) gibbsite growth around the boehmite
and temperature) from 100 to 300 g Na2CO3 L-1 decreased the rate particles.
of boehmite precipitation, similar to the effect on gibbsite
precipitation [11, 12]. Co-precipitation of gibbsite significantly changed the PSD of the
product. There is a large increase in the percentage (by volume)
Significant changes in the crystal morphology of the gibbsite of particles > 10 Pm in size and this has been attributed to the
products precipitated from these solutions were observed [12]. At agglomeration of gibbsite crystals and the adhesion of boehmite
low caustic concentration (~100 g Na2CO3 L-1) acicular particles on the gibbsite agglomerates.
(elongated) crystals are produced. At higher caustic
concentrations (~300 g Na2CO3 L-1) gibbsite crystals tend to be Decreasing the initial supersaturation (E=3.4 based on equilibrium
more blocky. No significant morphological changes occurred in boehmite solubility or Eapparent=1.5, A/C 0.57) correspondingly
the boehmite products with changes in caustic concentrations. decreased the precipitation rate. Boehmite was the major phase in
the products with gibbsite accounting for < 1 %.
At all the investigated caustic concentrations, only gibbsite was
observed from gibbsite-seeded experiments. Increasing the Due to the interference of gibbsite in the boehmite-seeded
caustic concentration from 100 to 300 g Na2CO3 L-1 increased the experiments, no definite conclusions could be made with regards
amount of gibbsite present in the product from boehmite-seeded to whether the growth processes observed in gibbsite precipitation
experiments from 0 to 8 % (as estimated from XRD). This occur in boehmite precipitation at different initial
implies that higher caustic (or sodium ion) concentration favours supersaturations. However, it has been observed that boehmite
gibbsite precipitation in a boehmite-seeded experiment. agglomeration does not occur to any significant extent.
Effect of Initial Supersaturation Effect of Solution Cation – Sodium vs. Potassium
It is well established that by changing the initial supersaturation of The substitution of potassium for sodium in a gibbsite-seeded
the liquor, the different growth processes (nucleation, aluminate liquor increased the gibbsite precipitation rate [16]. For
agglomeration, slow growth) observed in gibbsite precipitation the first five hours of the precipitation experiment, the boehmite
may be separated and investigated individually to some extent precipitation rate in potassium aluminate liquor was significantly
[13-15]. The effect of supersaturation (at constant caustic faster relative to the precipitation rate from a sodium aluminate
concentration and temperature) on boehmite precipitation was liquor. By 168 hours, the yield had increased by 20 g L-1. Using
investigated to firstly determine if similar growth processes a common equation applicable to gibbsite growth (2), where SA is
occurred in boehmite precipitation and secondly, whether they the surface area and E is the supersaturation with respect to
could be separated and investigated individually. boehmite, it was possible to arrive at a temperature-dependent

205
growth rate constant (kT) for each experiment. occurred at less energetically favourable sites, indicating that
these sites can still grow (at high supersaturations) while the more
dAl2O3 active sites have been poisoned. The new sites create additional
- kT SAE 2 (2) adsorption and growth sites until these outweigh the poisoning
dt
capability of the gluconate and growth then continues as per
Where potassium was substituted for sodium, the rate constant normal. Gibbsite was the only phase detected.
was approximately 5 times greater for boehmite precipitation.
Applying the same equation to gibbsite precipitation, the rate The effect of sodium gluconate on boehmite precipitation is more
constant was approximately three times greater from the pronounced than for gibbsite precipitation. The induction period
potassium liquor. Thus the potassium ion has had a greater effect for boehmite was approximately 2 to 3 times that of the induction
in increasing the precipitation rate for boehmite than for gibbsite. period of gibbsite at the same sodium gluconate dosage
concentration. Blocky crystals were observed in the product
Boehmite was the only phase present in the product as determined (Figure 5b). Boehmite was identified as the only product.
by powder X-ray diffraction. No gibbsite formation occurred (up
2 Pm 2 Pm
to 168 hours) while gibbsite had been observed by this time in the
other boehmite precipitation experiments in sodium aluminate
liquors. This could indicate an ‘inhibitory effect’ of potassium on
gibbsite nucleation in boehmite seeded experiments under
boehmite precipitation conditions, although there is no such effect
on gibbsite precipitation where the liquor is seeded with gibbsite.

The substitution of potassium for sodium in synthetic aluminate


a b
liquors changed the mechanism of gibbsite growth, resulting in Figure 5. The effect of 4 mmol L-1 sodium gluconate on (a)
acicular crystals (Figure 4a). No acicular crystals were observed nucleation on the gibbsite seed surface and (b) the crystal
from the boehmite-seeded potassium aluminate liquor. However, morphology of boehmite.
changes in the crystal morphology were observed and consisted of
straight edged diamond crystals becoming more rounded (Figure The total surface area of boehmite seed used was approximately
4b). 38 times that of the gibbsite seed. This, coupled with the
inhibitory effect of sodium gluconate being a surface process,
5 Pm 2 Pm indicates that the boehmite seed surface has significantly fewer
active sites than gibbsite and the less active (inactive) sites do not
readily form new growth sites as is observed for gibbsite.

Sodium Tartrate (4 mmol L-1 concentration). Sodium tartrate had


a reduced inhibitory effect on both gibbsite and boehmite
precipitation than sodium gluconate at the same concentration.
a b This is in agreement with the observation that an organic
compound with a higher number of carbon atoms and hydroxyl
Figure 4. Products from potassium aluminate liquor; (a) gibbsite groups would have a greater extent of inhibition [18]. No
seeded and (b) boehmite seeded. induction periods were observed for gibbsite precipitation.
Scanning electron micrographs revealed mass (elongated) nuclei
Effect of Organic Impurities formation on the seed crystals, predominantly in areas of rough
surfaces and on prismatic faces (Figure 6a). The observed
The mechanism by which organic compounds inhibit gibbsite morphology is in agreement with previously reported observations
precipitation is commonly attributed to a surface process [17]. [19, 20]. Gibbsite was the only phase detected.
This occurs through the reduction of the number of active sites on
the seed surface, thereby reducing the number of available sites The boehmite precipitation rate was only slightly reduced in the
where precipitation can occur. Comparing the effects of model presence of sodium tartrate and the induction period was
organic compounds on gibbsite precipitation with the effects on increased by one hour. This indicated that tartrate adsorption is
boehmite precipitation will increase knowledge and understanding less specific on boehmite and that growth sites may not be the
of the surface properties of boehmite. Two well-known gibbsite preferred adsorption sites. SEM images reveal rounded diamond
inhibitors (sodium gluconate and sodium tartrate) and an organic shaped crystals (similar to Figure 4b) with some plate-like nuclei
impurity present in Bayer liquor (sodium acetate) were and needle formation also present (Figure 6b). Boehmite was the
investigated. only phase in the product, indicating that gibbsite nucleation had
been suppressed.
Sodium Gluconate (4 mmol L-1 concentration). Sodium gluconate
is one of the strongest known inhibitors of gibbsite precipitation in Sodium Acetate (0.5 mol L-1 concentration). There was no effect
sodium aluminate solutions [13]. From the present results, of sodium acetate on gibbsite or boehmite precipitation. No
sodium gluconate inhibits growth (presence of an induction significant changes to the crystal morphology of the seed and
period, decreased precipitation rates) and suppressed secondary nuclei were observed for either seed types. The products from
nucleation (increase in the percentage by volume of particles with these experiments were the same as the seed type – i.e. gibbsite
sizes > 80 Pm) of gibbsite by adsorbing at the active growth sites. from gibbsite-seeded experiment and boehmite from boehmite-
While surface nucleation was still observed (Figure 5a), these seeded experiment. While sodium acetate appears to have no

206
effect on gibbsite precipitation, it seems to suppress gibbsite nucleation and precipitation rates, crystal morphology and the
nucleation in boehmite-seeded experiments. incorporation of impurities into the crystal. The majority of
results from investigations into solution speciation in caustic
2 Pm 2 Pm aluminate liquors confirm the general consensus that aluminium is
predominantly present as a four-coordinated monomer, the
aluminate ion (Al(OH)4-). However, there is evidence that
polymeric or oligomeric aluminate species (clusters) are also
present in the solution [21-23]. Although the exact nature of their
composition is unknown, it is likely that the clusters are gibbsite-
like in structure as they are precursors to gibbsite nuclei. Cluster
a b formation occurs in the bulk solution and may exist also on the
crystal surface. While no distinct interfacial layer at the growing
Figure 6. The effect of 4 mmol L-1 sodium tartrate on nuclei
gibbsite crystal surface has been observed spectroscopically [22],
formation on (a) gibbsite and (b) boehmite seed surfaces.
liquor-like clusters may still be present in the interfacial layer.
Clusters having the same composition, both in the bulk solution
Boehmite Precipitation from Plant Liquor
and in the interfacial layer, would be spectroscopically
indistinguishable from each other and a ‘distinct’ layer would not
Boehmite precipitation tests were also carried out in plant liquor
be detectable.
containing native organic impurities. As expected, the
precipitation rate was lower in the plant liquor relative to a pure
Precipitation occurs through the addition of growth units onto the
synthetic Bayer liquor under the same conditions. Boehmite was
surface. This may occur through the movement of the growth unit
the only phase identified in the product.
onto the surface or through the rearrangement of a cluster near the
surface. Boehmite precipitation occurs with the addition of
Conclusions
boehmite growth unit onto a boehmite surface. Clusters, close to
Despite optimism in the literature, it is unlikely that boehmite the boehmite growing surface, could impede the movement of
precipitation will ever compete with or replace gibbsite boehmite growth units to the surface. Structural rearrangement of
precipitation in the Bayer process. The most compelling reasons a gibbsite-like cluster to form a boehmite growth unit is less
for this are: energetically favourable due to its high stability. Gibbsite is more
likely to form due to the existence of the pseudo-gibbsite structure
x The precipitation rate for boehmite is approximately
of the clusters and explains why gibbsite was a major product in
200 times slower than that of gibbsite.
our boehmite experiments.
x Boehmite forms with a small crystallite size and does
not seem to grow any larger than 30 Pm. Disruption or the prevention of gibbsitic cluster formation would
x Boehmite does not tend to agglomerate to any great then remove the propensity for gibbsite to form. This has been
extent and apparent particle aggregation processes seem demonstrated through the substitution of potassium ions for
to be a consequence of nucleation on existing slow- sodium ions in solution. Only boehmite was precipitated from a
growing particles. potassium aluminate solution while gibbsite was the major phase
x Boehmite ‘agglomerates’ are fragile and tend to have from a sodium aluminate solution under conditions promoting
morphologies that would not survive handling or boehmite precipitation. Molecular dynamics modeling has
calcination, resulting in fine particles. shown that less stable clusters are formed with potassium ions
than with sodium ions [14, 24]. Sodium ions appear to have a
The solutions proposed in the literature to overcome the problems stabilizing effect on the clusters of aluminate ions in solution.
of slow kinetics and low yield of boehmite precipitation are not This stabilizing effect was also observed by dynamic light
economically feasible. For example, kinetic inhibition has been scattering measurements – sodium aluminate solutions had a
described as being due to poisoning of the active surface area by longer nucleation induction period than for potassium aluminate
hydroxyl ions. To ‘overcome’ this inhibition, it was proposed that solutions [14]. The substitution of potassium for sodium results
neutralization of caustic would increase the precipitation rate. in the formation of ‘looser’ clusters (preventing gibbsitic cluster
This is highly unattractive to an alumina refinery, as caustic soda formation), decreasing the predisposition for gibbsite nucleation
will have to be replaced or regenerated. in seeded solutions.
Despite the negative findings on the economic feasibility of Reduction of the initial supersaturation and caustic concentration
boehmite precipitation as a replacement for gibbsite precipitation (at constant supersaturation) also produced boehmite as the only
in the Bayer process, the commonalities and differences between product from boehmite-seeded sodium aluminate liquor. The
boehmite and gibbsite precipitation still provided valuable formation of (gibbsitic) clusters may also be reduced in less
information about gibbsite precipitation mechanisms. A new concentrated liquors, leading to the precipitation of boehmite as
growth mechanism is presented, incorporating present and past the dominant growth mechanism. In addition to changes in the
observations. product phase, the precipitation rate for both boehmite and
gibbsite increased (7 to 8-fold) with a decrease in caustic
The ‘active’ aluminium species or the composition of the ‘growth concentration. This may be attributed to the increased mobility
units’ participating in gibbsite precipitation remains yet to be through the solution by the growth units to the growing surface.
identified. While precipitation may be dependent upon the The same magnitude of the increase in the precipitation rates in
addition of ‘growth units’ from the solution onto the growing boehmite and gibbsite most likely indicates that the cluster
crystal surface, interactions of the aluminium species in solution
could play an important role in influencing outcomes such as

207
formation is similar in both solutions and that cluster formation is 10. J.M. Lamerant and Y. Perret, “Boehmitic Reversion in a
reacts similarly to changes in liquor concentration (speciation). Double Digestion Process on a Bauxite Containing Trihydrate and
Monohydrate”, Light Metals, 2002, 181-184.
The addition of simple organic compounds and the presence of
native organic compounds in plant Bayer liquor also prevented 11. C. Vernon, D. Lau and G. Parkinson, “Towards a
gibbsite precipitation in boehmite-seeded experiments. The effect Fundamental Rate Equation for Gibbsite Growth in Bayer
of the organic compounds appears to be through either the Liquors”, Proceedings of the 5th International Alumina Quality
disruption of the gibbsitic clusters (preventing the cluster Workshop (Bunbury, Western Australia), 1 (1999), 129-139.
formation) or the stabilization of the clusters (such that the energy
advantage for gibbsite to precipitation is decreased) rather than 12. C.F. Vernon, M.J. Brown, D. Lau and M.P. Zeiba,
through the promotion of boehmite precipitation. This is evident “Mechanistic Investigations of Gibbsite Growth”, Proceedings of
in the lack of morphological changes to boehmite in the presence the 6th International Alumina Quality Workshop (Brisbane,
of these additives while there are significant changes to gibbsite. Australia), 2002, 33-39.
This indicates fewer interactions between boehmite and the 13. H. Watling, J. Loh and H. Gatter, “Gibbsite Crystallization
organic additives. These observations also indicate that boehmite inhibition 1. Effects of Sodium Gluconate on Nucleation,
growth may occur through a different intermediate species to Agglomeration and Growth”, Hydrometallurgy, 55 (3) (2000),
gibbsite growth. 275-288.
Acknowledgements 14. J.S.C. Loh, “The Effect of Cations on Gibbsite
Crystallization”, (Ph.D. thesis, Curtin University of Technology,
The authors gratefully acknowledge Phan Khanh for particle size Perth, Western Australia, 2001) 334p.
distribution measurements and Ian Davies for powder X-ray
diffraction spectra collection. The authors also gratefully 15. D. Ilievski, “Development and Application of a Constant
acknowledge the Light Metals Flagship Program (CSIRO) for its Supersaturation Semi-Batch Crystallizer for Investigating
generous sponsorship of this work. Gibbsite Agglomeration”, J. Crystal Growth, 233 (2001), 846-
862.
References
16. J.S.C. Loh, H.R. Watling and G.M. Parkinson, “The Effect of
1. D.J. Brodie and H.W. Schmidt, “Custom Designed Fluid Bed Sodium, Potassium and Caesium on Gibbsite Agglomeration in
Calciner for Nabalco Pty. Ltd.” Proceedings of the 5th Alumina Bayer Solutions”, Proceedings of the 5th Alumina Quality
Quality Workshop (Bunbury, Western Australia), 1 (1999), 30-40. Workshop (Bunbury, Western Australia), 1 (1999), 118-128.

2. Alumina Refineries of the World, Aluminium-Verlag, ISBN 3- 17. D. Rossiter, D. Ilievski, P. Smith and G.M. Parkinson, “The
87017-276-2. Mechanism of Sodium Gluconate Poisoning of Gibbsite
Precipitation”, Chem. Eng. Res. Des., 74 (A7) (1996), 828-834.
3. D. Filippou and I. Paspaliaris, “From Bayer Process Liquors to
Boehmite and, Then, to Alumina: An Alternative Route for 18. H.R. Watling, “Gibbsite Crystallization Inhibition 1.
Alumina Production”, Light Metals, 1993, 119-123. Comparative Effects of Selected Alditols and Hydroxycarboxylic
Acids”, Hydrometallurgy, 55 (3) (2000), 289-309.
4. D. Panias and I. Paspaliaris, “Solubility of Boehmite in
Concentrated Sodium Hydroxide Solutions: Model Development 19. A.M. Paulaime, I. Seyssiecq and S. Veesler, “The Influence
and Assessment”, Hydrometallurgy, 59 (2001), 15-29. of Organic Additives on the Crystallization and Agglomeration of
Gibbsite”, Powder Technol., 130 (2003), 345-351.
5. D. Panias and I. Paspaliaris, “Boehmite Process – A New
Approach in Alumina Production”, Erzmetall., 56 (2) (2003), 75- 20. I. Seyssiecq, S. Veesler, G. Pepe and R. Boistelle, “The
81. Influence of Additives on the Crystal Habit of Gibbsite”, J.
Crystal Growth, 196 (1999), 174-180.
6. W.L. Connop, “A New Procedure for the Determination of
Alumina, Caustic and Carbonate in Bayer Liquor”, Proceedings of 21. H.R. Watling, “Spectroscopy of Concentrated Sodium
the 4th International Alumina Quality Workshop (Darwin, Aluminate Solutions”, Applied Spectrosc., 52 (2) (1998), 250-258.
Australia), 2 (1996), 321-330.
22. H.R. Watling, S.D. Fleming, W. van Bronswijk and A.L.
7. C. Skoufadis, D. Panias and I. Paspaliaris, “Kinetics of Rohl, “Ionic Structure in Caustic Aluminate Solutions and the
Boehmite Precipitation From Supersaturated Sodium Aluminate Precipitation of Gibbsite”, J. Chem. Soc. Dalton Trans., 1998,
Solutions, Hydrometallurgy, 68 (2003), 57-68. 3911-3917.

8. D. Panias and I. Paspaliaris, “Boehmite Process – A New 23. P.D. Fawell and H.R. Watling, “Multi-Angle Laser Light
Approach in Alumina Production”, Travaux, 29 (33) (2002), 94- Scattering to Measure Nucleation Induction Periods”, Appl.
104. Spectrosc., 52 (8) (1998), 1115-1117.

9. I. Paspaliaris, D. Panias, A. Amanatidis, D. Mordini, G. 24. S.D. Fleming, “Computer Modeling of Gibbsite
Werner, G. Panou and D. Ballas, “Precipitation and Calcination of Crystallization”, (Ph.D. thesis, Curtin University of Technology,
Monohydrate Alumina from the Bayer Process Liquors”, Perth, Western Australia, 1999), 127-147.
Proceedings of the 3rd Annual Workshop – Eurothen (Lisbon,
Portugal), EUR19370 (2000), 540-547.

208

You might also like