EFFECT ON ALUMINIUM SOLUBILITY WHEN ACID IS USED TO

NEUTRALISE RESIDUE
Sturt A
Research and Development, Rio Tinto Alcan Yarwun, QLD, Australia
Abstract
The RTA Yarwun alumina refinery currently produces 1.3Mtpa of residue. Seawater is used to neutralise the residue
prior to mud stacking and effluent discharge. Equipment limitations and altered process conditions have resulted in
insufficient seawater being available for complete neutralisation. As a result, sulphuric acid is being used to supplement the
neutralisation process. This has significantly altered the neutralisation chemistry, particularly the solubility of aluminium.
It was also found that the pH required to precipitate dissolved aluminium from the supernatant liquor is significantly lower
when acid is used compared to seawater.

1. Introduction Initial Titrations: Incremental amounts of seawater were added

to the last washer/oxalate underflow mixture and the pH recorded.
Waste liquor generated by the Bayer process contains high
The same titration was then performed with 10% sulphuric acid.
concentrations of sodium aluminate. In order to meet
environmental requirements, the removal of aluminium and other Titrations were performed to enable calculation of the volume of
metal impurities from solution is required prior to discharge. Sea seawater or acid required, to neutralise the mixture to a specific
water is commonly used to effectively neutralise red mud. The pH.
magnesium in sea-water reacts with aluminium in solution to
form hydrotalcite, during which free hydroxide is also consumed, Neutralisation simulations: All neutralisation simulations were
reducing pH and removing the aluminium from solution (equation performed in a beaker with overhead stirrer, at room temperature
1). Excess sea water forces this equilibrium to the right: for 5 minutes. A series of last washer/oxalate underflow samples
were neutralised to various pH values with either seawater, acid
6Mg2+ + 2NaAl(OH)4 + 8NaOH + CO32- or a combination of the two. In the case of a combination, sea-
+ 4H2O ÅÆ Mg6Al2CO3(OH)16.4H2O + 10Na+ [1] water and acid were dosed simultaneously. Neutralised mud
(Hydrotalcite) samples were centrifuged and the liquor collected for analysis.
Soluble aluminum was measured using ICP.
Hydrotalcite belongs to a group of minerals known as double
layered hydroxides and has the ability to incorporate anionic 3. Results and discussion
species from solution (Smith and Parkinson, 2005). The general
Aluminium levels in neutralised supernatant
formula for hydrotalcite is [(Mg(1-x)Alx(OH)2]X.nH2O)n)n, where X
can be any anion present when precipitation occurs (Anderson, The concentration of dissolved aluminium in washed mud
2007). Carbonate (CO32-) is the most strongly bound anion due is reduced when neutralised with seawater. The dissolved
to its high charge density and strong affinity for Mg2+ and Al3+. aluminium concentration of neutralised LWU as a function of
Incorporation of other anions such as vanadates and molybdates, pH can be found in Figure 1. The neutralisation reaction was
also found in waste liquor, is dependant on the pH. The more performed with either seawater or 10% sulfuric acid over a range
strongly bound carbonated hydrotalcite is preferentially formed of pH values, ~10.00 – 8.45 (Tables 1a and 1b).
at high pH. It is not until pH 8.5 is achieved that other anions
can become included into the hydrotalcite matrix (Smith and
Parkinson, 2005). 250

At Yarwun, last washer underflow, LWU, and oxalate settler 200

underflow are sent to a neutralisation tank. Frequently, the seaw ater neutralised
Al (mg/L)

acid neutralised
quantity of seawater did not reduce the pH of the effluent 150
E xpon. (acid neutralised)

sufficiently for discharge to the environment. In order to overcome E xpon. (seaw ater neutralised)

this problem, Yarwun implemented a seawater-sulphuric acid 100

neutralisation process.
50

The purpose of this investigation was to assess the changes in

neutralisation chemistry, in particular aluminium solubility, as a 0
8 .4 8 .9 9.4 9 .9 10.4 10.9
result of the use of acid with seawater. Further, a model was pH
developed to enable a prediction of aluminium concentration
in the neutralised supernatant based on the refinery’s current Figure 1. Soluble Aluminium present in neutralised mud
neutralisation strategy. Species such as vanadium, molybdenum
and arsenic, which are also considered detrimental to the
environment and removed through the neutralisation process,
were not investigated.

2. Experimental method
Samples: Samples of last washer underflow and oxalate settler
underflow were collected and mixed in a volumetric ratio similar
to that observed in the refinery.

Proceedings of the 8th International Alumina Quality Workshop • 2008 93

AQW2008_Proceedings_FINAL-1-120.93 93 4/08/2008 11:30:34 AM

 Table 1. Dissolved aluminium levels in mud neutralised with • C1 and C2 are constants equal to 2.429 and 1.969
acid (a) or seawater (b) respectively.
a) b) As the amount of acid is maximised and SW minimised, β
pH Acid neutralised mud Seawater neutralised mud increases. This was evident when exponential curves were fitted
pH
Dissolved Al (mg/L) Dissolved Al (mg/L)
to experimental data in Figure 3.

9.85 258 10.02 3.07 When [SW] = 1 [A] = 0 β = 1.969

9.4 93 9.46 1.46
When [SW] = 0.35 [A] = 0.65 β = 2.276
8.87 20.4 9.06 0.66

8.53 9.4 8.94 0.38 When [SW] = 0 [A] = 1 β = 2.429

The soluble aluminium content of mud neutralised with acid The model is accurate for mud neutralised to pH values within
is significantly greater than mud neutralised by seawater the range 8.0 to 10.5. Within this range the model compared
for equivalent pH values. For example, at a pH of ~9.0, the well with plant data. Importantly, this is the range over which
dissolved aluminium concentration of acid neutralised mud is refinery neutralisation will occur. Lab data indicated that above
30 times greater than the aluminium concentration of sea water a pH of 10.7, the aluminium relationship with pH in the case of
neutralised mud. At a pH of ~10.0, a 90 fold increase in the a seawater-acid mixture, is no longer exponential but forms an
aluminium concentration is evident. S-curve relationship.

Figure 2 displays a curve representing the aluminium The model can be used to show that as the acid fraction
concentration versus pH of LWU after it was neutralised in the increases, the pH required to reduce aluminium concentration
laboratory with a seawater-sulfuric acid mixture. The volumetric to a specific level decreases significantly. This is best displayed
ratio of seawater to acid was kept constant for each reaction, and in Figure 3, which plots aluminium concentration as a function of
this uncovered an exponential relationship between dissolved pH for various acid fractions.
aluminium concentration and pH in the neutralised effluent 50

45
400
40
A = 0.8
350
SW =0.2
35
A = 0.6
Aluminium (mg/L)

300 SW = 0.4
30
A cid/seaw ater A = 0.4
A cid neutralised neutralised
250 SW = 0.6
2.4294x 25
y = 1E -08e
Al (mg/L)

2.276x A = 0.2
y = 1E -08e
SW = 0.8
200 20

15
150

10
100
seaw ater 5
50 neutralised
1.9686x 0
y = 1E -08e
7 .9 8 .1 8 .3 8 .5 8 .7 8 .9 9 .1 9 .3 9 .5 9 .7 9 .9 1 0 .1 1 0 .3 1 0 .5
0
pH
8.5 9 9.5 10 10.5 11
pH
Figure 3. Soluble Aluminium as function of pH for different sea water and
Figure 2. Soluble Aluminium in mud neutralised with seawater-sulphuric acid acid fractions
mixture
Chemistry during neutralisation of mud with seawater-sulphuric
Development of model to predict soluble aluminium in acid mixture
neutralised mud The increase in soluble aluminium in neutralised mud as sea-
Using data from Figure 2 a model was proposed to predict water is replaced with acid is best explained by the different
aluminium concentration in neutralised mud as a function of pH, products that precipitate and ultimately remove aluminium from
seawater and acid. The model (equation 2) states: solution. Hydrotalcite, formed when liquor reacts with Mg2+ in
seawater, has a pH of about 10.6. This was equivalent to the
Al (mg/L) = 1×10-8 exp (β × pH) [2]
endpoint of a titration curve when incremental amounts of
where β is dependant on the amounts of seawater and acid used seawater were added to LWU liquor, shown in Figure 4.
in the neutralisation. The formula for β (equation 3) states:
13
β = C1[A] + C2[SW] [3] 1 2 .5

• [SW] is the seawater fraction. [SW] is equal to the moles of 12

equivalent caustic-neutralising ions from seawater divided 1 1 .5
by the total moles of caustic-neutralising ions added to the 11
system;
pH

1 0 .5
[SW] = 2 × n(Mg2+) [4] 10
2 × n(Mg2+) + n(H+)
9 .5
• [A] is called the acid fraction and is equal to the moles of 9
H+ ions from the acid divided by the total moles of caustic-
8 .5
neutralising ions added to the system;
volume seawater →
[A] = 2 × n(H+) [5]
2 × n(Mg2+) + n(H+) Figure 4. Titration of LWU liquor with seawater

In order to reduce the pH below 10.6, excess magnesium is

required. Excess magnesium forces equation [1] to the right.

94 Proceedings of the 8th International Alumina Quality Workshop • 2008

AQW2008_Proceedings_FINAL-1-120.94 94 4/08/2008 2:55:35 AM

 Aluminium is precipitated as hydrotalcite, so there is very little aluminium is removed via precipitation of the neutral solid
sodium aluminate remaining in solution below pH 10.6. This gibbsite (Al(OH)3). As a result, a much lower pH value is required
was evident in lab simulations when, at a pH of 10.0, soluble to achieve acceptable aluminium levels if all the neutralisation
aluminium in sea-water neutralised mud was measured to be as demand is met with sulphuric acid.
low as 3.1mg/L.
When a seawater-sulphuric acid mixture is used to neutralise red
Gibbsite, Al(OH)3, thought to form when sulphuric acid reacts with mud, the pH needed to achieve the required aluminium level is
LWU liquor (Smith 2005), is a neutral compound. In aqueous lower as the portion of acid is increased.
solutions close to neutral pH (~7), the solubility of gibbsite is
very low (Wefers, 1987, p.3). The simplified reaction of sodium
aluminate with sulphuric acid is shown below:

H2SO4 + 2NaAl(OH)4 Æ Na2SO4 + 2Al(OH)3 + 2H2O [6]

Gibbsite is also amphoteric, so the endpoint of the LWU titration

using sulphuric acid is approximately 7 (Figure 5).

12

10

8
pH

0
volume acid →

Figure 5. Titration of LWU liquor with 1% sulphuric acid

Therefore, a titration of LWU with sulphuric acid requires a much

lower pH value (~8.0) to precipitate aluminium to below 3mg/L
when compared to seawater (~10.6).

Neutralising LWU with a seawater-sulphuric acid mixture results

in equations [1] and [6] occurring simultaneously. As the portion
of acid in the mixture increases, the pH required to reduce
aluminium concentration to an acceptable level is reduced. This
is because there is more of equation [6] occurring and less of
equation [1].

The formation of gibbsite in place of hydrotalcite not only causes

a shift in soluble aluminium levels in the neutralised mud but
is likely to have an affect on other elements such as vanadium.
Although not investigated, the formation of gibbsite in place of
hydrotalcite has the potential to reduce the incorporation of
species such as vanadate into the hydrotalcite matrix.

4. Conclusions
Neutralisation of last washer underflow reduces dissolved
aluminium levels regardless of whether neutralisation is
conducted with seawater, acid or a combination of the two.
Seawater neutralisation, however, is more effective at removing
aluminium from solution.

In the case of seawater neutralisation, soluble aluminium is

removed from solution via the precipitation of hydrotalcite.
At pH values below 10, there is very little aluminium in the
liquor, <3mg/L. However, when acid is used for neutralisation,

References
1. Smith H. D., Parkinson G. M., “Seawater Neutralisation: Factors affecting adsorption of anionic chemical species”, 7th International Alumina Quality
Workshop, 2005, pp 221-224.
2. Anderson J., Smith M., “Gove Seawater Neutralisation Testwork: Phase 1: Establishing Primary Neutralisation Parameters, Phase 2: Ecotoxicology
Testing”, QRDC Internal report, Q07/004 Sept 2007.
3. Wefers K., Misra C., Oxides and Hydroxides of Aluminium, Alcoa Laboratories, 1987.

Proceedings of the 8th International Alumina Quality Workshop • 2008 95

AQW2008_Proceedings_FINAL-1-120.95 95 4/08/2008 2:55:35 AM