ARTICLE

pubs.acs.org/JAFC

Improved Sugar Cane Juice Clarification by Understanding Calcium

Oxide-Phosphate-Sucrose Systems
William O. S. Doherty*
Centre for Tropical Crops and Biocommodities, Queensland University of Technology, Brisbane, QLD 4001, Australia
bS Supporting Information
ABSTRACT: It is accepted that the efficiency of sugar cane clarification is closely linked with sugar juice composition (including
suspended or insoluble impurities), the inorganic phosphate content, the liming condition and type, and the interactions between
the juice components. These interactions are not well understood, particularly those between calcium, phosphate, and sucrose in
sugar cane juice. Studies have been conducted on calcium oxide (CaO)/phosphate/sucrose systems in both synthetic and factory
juices to provide further information on the defecation process (i.e., simple liming to effect impurity removal) and to identify an
effective clarification process that would result in reduced scaling of sugar factory evaporators, pans, and centrifugals. Results have
shown that a two-stage process involving the addition of lime saccharate to a set juice pH followed by the addition of sodium
hydroxide to a final juice pH or a similar two-stage process where the order of addition of the alkalis is reversed prior to clarification
reduces the impurity loading of the clarified juice compared to that of the clarified juice obtained by the conventional defecation
process. The treatment process showed reductions in CaO (27% to 50%) and MgO (up to 20%) in clarified juices with no apparent
loss in juice clarity or increase in residence time of the mud particles compared to those in the conventional process. There was also a
reduction in the SiO2 content. However, the disadvantage of this process is the significant increase in the Na2O content.
KEYWORDS: Calcium phosphate, sugar cane, clarification, clarified juice, defecation

’ INTRODUCTION particles and the clarified juice quality.1 The design improve-
Raw cane juice and its subsequent form (mixed juice, MJ) are ments have led to significant increases in the capacity to size ratio
stable suspensions that contain large numbers of suspended par- of Australia’s Sugar Research Institute (SRI) clarifiers as com-
ticles. The soluble components of juice include sucrose, glucose, pared to multitray clarifiers. The decrease in juice residence times
fructose, proteins, oligosaccharides, polysaccharides, organic has been remarkable; up to 3.5 h residence time for multitray
acids, amino acids, and salts. The suspended material largely clarifiers has been reduced to as low as 20 min for the SRI
clarifier.1 However, the effectiveness of the conventional clarifi-
consists of bagacillo (a lignocellulosic byproduct), soil, sand,
cation process depends, to a certain extent, on the nature and
clay, starch, waxes, fats, and gums. The relative amounts of the
type of the calcium phosphate precipitated.2 The formation of
components in both phases depend on the variety, maturity, and
calcium phosphate from supersaturated solutions has been
condition of the cane plant, soil type, and the harvesting method widely studied and is readily accessible in the literature.3-12
and conditions. In general, the composition of the dispersed The nature of the calcium phosphate phases formed in sugar juice
phase is not critically important in terms of clarification behavior. depends on (a) calcium and phosphate ion concentrations, (b)
Rather, it is the nature or state of the surface of the suspended magnesium ions, (c) the pH of the solution, and (d) the presence
particles in the juice that determines the overall behavior of the of impurities such as organic acids, and carbonates.4,5,13-17 Thus,
juice. The aims of the clarification process are to remove the efficiency of sugar juice clarification is closely linked with the
suspended and colloidal particles, remove nonsucrose impurities juice composition (including suspended or insoluble impurities),
(e.g., proteins, polysaccharides and inorganic materials), and the inorganic phosphate content in the mixed juice, the liming
raise the juice pH to minimize inversion of sucrose during conditions, and interactions between juice components at high
subsequent processing. Clarification affects juice filterability, heat pH and temperatures. The addition of lime to the juice, with
transfer coefficients of evaporators, sucrose crystallization, and heating, precipitates both inorganic (mainly calcium phosphate)
the quality and quantity of raw sugar produced. It affects the and organic calcium salts.
color, crystal morphology, crystal content, ash, and polysaccha- The mode of clarification in many sugar cane factories around
ride content of raw sugar. In the raw sugar refinery, poor filtration the world is simple defecation. Lime addition may occur as milk
and reduced refined sugar yields can be traced back to the poor of lime or lime saccharate (lime dissolved in juice or syrup). In
clarification of mixed juice in sugar mills. Clarifier performance Australia, the process is based on the addition of lime as lime
and clarified juice quality affect factory throughput, molasses
purity, and raw sugar quality. Received: November 7, 2010
From the 1960s to early 1980s research in Australia on the Accepted: January 19, 2011
utilization of flocculants and optimization of clarifier design Revised: January 18, 2011
led to considerable improvements in settling rates of the juice Published: February 15, 2011

r 2011 American Chemical Society 1829 dx.doi.org/10.1021/jf1043212 | J. Agric. Food Chem. 2011, 59, 1829–1836
Journal of Agricultural and Food Chemistry ARTICLE

saccharate, after primary heating of the juice to 72-76 °C, to Table 1. Composition of Synthetic Juices
raise the pH of the juice to approximately 7.8-8.0. The juice is
juice A juice B juice C juice D juice E
then heated further to about 102 to 105 °C and is flashed to the
atmosphere before entering the clarifier. Polyacrylamide floccu- constituents pH 5.7 pH 4.4 pH 4.4 pH 3.0 pH 5.8
lant is added to the juice in the feed launder of the clarifier, and
CaO (mg/L) 393 257 257 60 399
the combined action of the flocculant and precipitating calcium
P2O5 (mg/L) 318 319 193 319 319
phosphates form large flocs that settle out in the clarifier. Because
the phosphate level in the juice is particularly important, phos- aconitic acid (mg/L) 1000 1000 1000 1000 1000
phate can be added to the mixed juice, prior to heating, in the sucrose (mass%) 12 12 12 12 0
form of superphosphate fertilizer or phosphoric acid. The major
disadvantage of using lime or lime saccharate to raise juice pH As aconitic acid is by far the most predominant organic acid in juice and
for calcium phosphate precipitation is the very high levels of CaO constitutes about two-thirds of the organic acid content, it was used in
in clarified juices arising not only from the added lime but also the study to model the total organic acids present in juice. Juice A is
from the endogenous calcium in raw juice. Excessive CaO levels regarded as an ideal juice since it contains not only sufficient amounts of
in clarified juices affects clarified juice quality which in turn affects CaO and P2O5 to form Ca3(PO4)2, but contains sufficient amount of
scaling of the evaporators, pans and fugals, and molasses exhaus- inorganic phosphate necessary for good clarification.11,18
tion. In order to address this issue and get a better and improved The experimental form was a 5
3
3
3
2 design with
understanding of calcium phosphate chemistry in sugar solu- 5 synthetic juices, 3 bases (lime saccharate, sodium hydroxide, and
tions, laboratory studies have been carried out on the formation ammonium hydroxide), 3 pH values (7.0, 7.5, and 8.0), and 3 tempera-
tures (50, 76, and 90 °C). The experiments were performed in duplicate.
of calcium phosphate in synthetic and factory juices in the
The treated synthetic juices were filtered through a 0.45 μm filter and
presence of different bases.
the filtrate acidified with HCl and stored frozen prior to analysis by ICP-
OES for Ca and P. The experimental error involved in the precipitation
’ MATERIALS AND METHODS studies was less than 5%. The error was determined by estimating the
confidence interval of the standard deviation based on pooled duplicate
Materials. Concentrated solutions of sodium hydroxide, ammo- measurements.19 The precipitates obtained with juice A at pH 8 were
nium hydroxide, phosphoric acid, aconitic acid, and hydrochloric acid washed with water and dried in a vacuum oven to a constant weight and
were prepared with analytical reagent chemicals (Merck, Victoria, the weight noted. The proportions of Ca and P in the precipitates were
Australia) and Millipore deionized water. Lime saccharate was prepared also determined by ICP-OES after dissolution in dilute HNO3. This was
by mixing 100 g of 200 g/kg solution of analytical grade calcium oxide to establish the composition of the primary precipitate in juice A.
(APS Specialty Chemicals, Brisbane, Australia) and 100 g of 65 mass% Investigations with Filtered Primary Juice. Similar calcium phos-
sucrose. Sucrose was an analytical grade reagent (Sigma-Aldrich, Sydney, phate precipitation experiments as those outlined for the synthetic juices
Australia). were carried out with a filtered factory primary juice. The factory juice
Factory primary juice (i.e., MJ after the primary heaters) that had had been previously stored frozen. To remove the effect of suspended
been previously stored frozen was defrosted and filtered through and coagulated particles (which will vary from one aliquot to the other)
0.45 μm filter prior to use. Samples of primary juices were obtained on the amount of precipitate formed, the juice was filtered through a
from five Australian sugar factories located in different sugar cane Whatman No. 52 filter paper. The alkalis examined were lime saccharate,
growing districts. The pH of both the synthetic and factory juices was sodium hydroxide solution, and ammonium hydroxide solution, a two-
measured on a Hanna Model H1 9025 m with the high temperature stage process involving adding lime saccharate to a heated juice to a set
probe Uteck EC-FG-7350401B. pH and then adding sodium hydroxide solution/ammonium hydroxide
The inorganic ions were analyzed using a Varian Vista-MPX CCD solution to the final pH. The precipitates were washed with distilled
simultaneous inductively coupled argon plasma optical emission spec- water and dried in a vacuum oven to a constant weight. The experiments
trometer (ICP-OES). To reduce the interference of the sugar organic were performed in duplicate and the results recorded as average values.
matrix, it was necessary to dilute the samples to a sucrose concentration The amounts of free calcium ions in the filtrate after the removal
of ∼1 mass%. of the precipitated residue were determined with a calcium electrode
Methods. The experimental procedure has been divided into (a Radiometer ion-selective membrane electrode type ISE25Ca) and
three stages in order to build a better understanding of calcium oxide/ 1 M KCl calomel reference electrode.
phosphate/sucrose systems that would impact on scale-forming ions in Investigations with Factory Primary Juices. For these series of
clarified sugar cane juice. The first stage of the laboratory investigation experiments, samples of factory primary juices were clarified with the
was to determine in some detail the amounts of residual CaO and P2O5 two-stage addition process of lime saccharate followed by sodium
remaining after treating different juices, at selected temperatures, with hydroxide and sodium hydroxide followed by lime saccharate at 3 pH
different alkalis. This stage of the work was also to provide insights on values (7.5, 7.8, and 8.0) and two liming temperatures (76 and 90 °C),
the solution chemistry of calcium phosphate under various processing boiled and settled in standard settling tubes [each tube of 1 L capacity
conditions. The second stage of the investigation on filtered primary with dimension of 460 mm (h)
55 mm (i.d.)] in a SRI-designed
juice built on the information obtained in the first stage and was aimed at heated illuminated clarification box. A dose equivalent to 3 mg/kg of
establishing whether a two-stage alkali treatment would result in reduced superfloc (a polyacrylamide-co-sodium acrylate flocculant; Cytec,
amounts CaO and P2O5 in the treated juices. The third stage of the Australia) was applied before settling. As the mud was unstirred, the
investigation was a practical demonstration of the two-stage treatment mud heights were not indicative of values obtained in commercial
process in factory juice clarification. clarifiers. Control experiments were also conducted with lime saccharate
Investigations with Synthetic Juices. To examine the effect of at pH 7.8 and 76 °C, typical conditions used by Australian sugar factories.
different alkalis on calcium phosphate precipitation, experiments were The clarification experiments were conducted in triplicate. Samples of
conducted on synthetic juices using lime saccharate, sodium hydroxide clarified juices were siphoned off for storage and subsequent analysis.
and ammonium hydroxide at different temperatures and pH values. The The settling rate of the flocculated mud particles were determined
compositions of the juices labeled A, B, C, D, and E are shown in Table 1. by measuring the level of the mud interface at 0.5, 1.0, 2.0, and 3.0 min.

1830 dx.doi.org/10.1021/jf1043212 |J. Agric. Food Chem. 2011, 59, 1829–1836

Journal of Agricultural and Food Chemistry ARTICLE

The initial settling rate (mL/min) was obtained from the graphical
analysis of the initial linear slope.
Turbidity of the clarified juices was determined by measuring the
absorbance (A) at 900 nm in 1 cm cells against distilled water.20 The
turbidity of the clarified juice was defined as 100A. The percentage error
for turbidity determination was (0.3%.
Color was determined at pH 7 by measuring the absorbance at
420 nm in 1 cm cells.21 For syrup color determinations in the range of
34000 to 42000 ICU, the reported 95% repeatability is ∼5% of the mean
color value.21 This repeatability value has been assumed to be the case
in the present study. Dry substance was measured using an air oven at
98 °C.22
Sucrose contents of the factory juices were determined by high
performance liquid chromatography analysis. Samples were diluted to
give ∼1% sucrose concentration and filtered through a 0.45 μm filter Figure 1. Effect of added caustic soda, lime saccharate, and ammonium
for analysis. The column used was a 250
7.8 mm i.d., 10-15 μm, 10% hydroxide on the concentration of CaO in synthetic juice A at 50 °C, 76
Naþ with a guard column 28
7.8 mm i.d. of the same material (Shodex °C, and 90 °C.
KS801). Sucrose was separated using this column equilibrated at 65 °C
and was detected by a Waters 410 Refractive Index Detector. The eluent
was water at a flow rate of 0.5 mL/min. The purity of the clarified juice
was calculated as the percentage ratio of sucrose content to dry
substance.
The inorganic ion concentrations of the clarified juices were deter-
mined by ICP-OES. The experimental error for the clarification experi-
ments for the inorganic ions was ∼5%.19

’ RESULTS AND DISCUSSION

Solution Chemistry. The main reactions that occur in limed
juice to produce calcium phosphate phases10-12 are as follows:
Ca2þ þ 2H2 PO4 - f CaðH2 PO4 Þ2 ðmonocalcium phosphateÞ
ð1Þ
Figure 2. Effect of added caustic soda, lime saccharate, and ammonium
hydroxide on the concentration of P2O5 in synthetic juice A at 50 °C, 76
Ca2þ þ HPO4 2- f CaHPO4 ðdicalcium phosphateÞ ð2Þ °C, and 90 °C.

The results of the calcium phosphate precipitation studies

3Ca2þ þ 2PO4 3- f Ca3 ðPO4 Þ2 ðtricalcium phosphateÞ in juice A are presented in Figures 1 and 2, with the changes in
ð3Þ residual CaO and P2O5 concentrations at different pH values and
temperatures. The results show that the amount of untreated
2CaHPO4 þ 2Ca3 ðPO4 Þ2 f Ca8 H2 ðPO4 Þ6 CaO and P2O5 are highly dependent on the juice's final pH and
working temperature. The amount of residual CaO and P2O5
ðoctacalcium phosphateÞ ð4Þ were also found to depend on juice type. For a majority of the
juices, irrespective of the alkali used, juices limed at 50 °C
Ca3 ðPO4 Þ2 þ 2Ca2þ þ HPO4 2- þ H2 O f Ca5 ðPO4 Þ3 OH contained the largest proportions of unreacted CaO and P2O5
þ 2Hþ ðhydroxyapatiteÞ ð5Þ compared to the juices limed at 76 and 90 °C, respectively. This is
due to increased solubility of the calcium phosphate phases at the
In addition, tetracalcium phosphate [Ca4(PO4)2OH2] is also lower temperature. However, some treated juices do not follow
formed. this trend. These are (a) juice A treated with sodium hydroxide at
Juices with insufficient phosphate generally clarify poorly 90 °C and at pH 7, (b) juice B treated with ammonium hydroxide
and produce clarified juices of high turbidities.11,19 Juices with at 76 °C and pH 7, (c) juice C treated with sodium hydroxide
excessive phosphate increases the level of soluble calcium and at 90 °C and pH 8, and (d) juice E treated at 90 °C and at pH
phosphate in clarified juices, thereby increasing scaling intensity values of 7.0 and 7.5. For these juices, with the unavailability of
and the amount of scale formed in the evaporators and pans. The sufficient calcium ions, it is probable that the type of phosphate
increase in the phosphate level in the clarified juices also reduces phases formed under these conditions is controlled by stoichi-
sugar filterbility.3 Excessive calcium ions in clarified juices ometry rather than inverse solubility effect. The dominant
increase scaling in the evaporators, pans, and centrifugals, phosphate species at pH ∼7 is H2PO4- followed by HPO42-,
increases ash in sugar, and reduces molasses exhaustion. As such, whereas at higher pH the reverse is true.11 Hence, at a pH of ∼7
the concentrations of soluble calcium and phosphate both in the stoichiometry of calcium phosphate phases is different from
juice before and after clarification are critical in determining the other systems that follow the inverse solubility rule.
the efficiency of not only the clarification process but also the As the highest levels of CaO and P2O5 contents were obtained
productivity of the sugar manufacturing process and the quality with juices treated at pH 7 and at 50 °C, only the data obtained
of the product sugar. for synthetic juices heated at 76 and 90 °C, with pH values of
1831 dx.doi.org/10.1021/jf1043212 |J. Agric. Food Chem. 2011, 59, 1829–1836
Journal of Agricultural and Food Chemistry ARTICLE

7.5 and 8.0, respectively, have been critically examined. As shown

in Figures 1 and 2, where lime saccharate is the base, all the juices
have precipitated the majority of their P2O5 though significant
proportions of CaO remain in solution. Where the initial levels of
P2O5 contents in the juices are the same (i.e., juices A, B, D and
E), the average level of CaO remaining after precipitation (at pH
7.5 and 8.0) is about 235 mg/L for juices limed at 76 °C and
220 mg/L for juices limed at 90 °C. In the case of juice C (where
there is insufficient phosphate ions available for interaction with
calcium ions), once calcium phosphate precipitation is complete,
any extra added lime saccharate adds proportionately to the CaO
content remaining in the juice. These results highlight the critical
role played by the initial concentration of P2O5 in juice in
determining the CaO content of treated juices. The results also Figure 3. Effect of added caustic soda, lime saccharate, and ammonium
raise the issue of high levels of unreacted CaO in juice when lime hydroxide on the concentration of CaO in synthetic juice D at 50 °C,
saccharate is used. Aconitic acid in its dissociated forms contains 76 °C, and 90 °C.
the species H2C6H3O6-, HC6H3O62-, and C6H3O63-. The first
dissociation constant (K1) of aconitic acid (4.8
10-3 at 25 °C)
is weaker than that of phosphoric acid (7.52
10-3 at 25 °C).23
However, the second dissociation constant (K2 = 4.6
10-5 at
25 °C) is well over 500 times that of phosphoric acid (6.23

10-8 at 25 °C). As the K2 values occur predominantly at the pH
range between 4 and 7, it is expected that HC6H3O62- would
readily form complexes with calcium ions in solution, making the
calcium ions inaccessible for interaction with phosphate ions in
situ. Therefore, in situations where free calcium ions are available,
such as with lime saccharate addition, the calcium phosphate
precipitation process preferentially occurs with these ions instead
of the calcium present in the juice which is predominantly
present as a complex.
As shown in Figures 1 and 2 where sodium hydroxide or
ammonium hydroxide are the alkalis, all the juices have surren- Figure 4. Effect of added caustic soda, lime saccharate, and ammonium
dered large proportions of the initial CaO and P2O5 contents to hydroxide on the concentration of P2O5 in synthetic juice D at 50 °C,
calcium phosphate formation. Similar results were obtained with 76 °C, and 90 °C.
juices B and C. With juice D, practically all the initial CaO
content was used up in calcium phosphate precipitation, while Composition of Solid Phases. The results of the composition
the majority of the initial P2O5 content of the juice remained of the precipitates formed with juice A at a pH of 8.0 are given in
unreacted (Figures 3 and 4). A plausible explanation for these Table 2. No analysis was carried out on the solids obtained with
effects is that during precipitation a fixed amount of soluble CaO other juices. The results show that increasing amounts of calcium
and P2O5 is removed from solution and that once one of the (as CaO) is precipitated as the juice temperature is increased
lattice forming ions has been depleted (in this case calcium), from 50 to 90 °C and that the precipitates obtained with lime
precipitation stops, and the other ion is left in excess. The ratio of saccharate contain the highest proportions of CaO to P2O5 at a
the concentration (in mg/L) of CaO to P2O5 that was withdrawn given temperature. The actual calcium phosphate phases are not
from solution was found to be about 1.3 for all treated juices with known. The molar ratio of CaO to P2O5 in juice A is 1.04,
the exception of juice D (treated with sodium hydroxide), which implying that there are enough lattice ions to form Ca3(PO4)2,
had a ratio of 2.2. which has a molar ratio of 1. None of the precipitates formed with
When the results of juice A were compared to those of juice E the various bases attained the ratio of 1 (except ammonium
(where the reaction was carried out in the absence of sucrose), hydroxide at 50 °C). The precipitates formed with lime saccha-
there were no noticeable differences between the precipitation rate and ammonium hydroxide have slightly higher ratios, while
path of calcium phosphate carried out in the presence and those of sodium hydroxide have slightly lower values. This
absence of sucrose. This is not to say that the physical nature probably implies that the precipitates formed with lime sacchar-
of the calcium phosphate is not different between the two ate and ammonium hydroxide contain the highest proportions
systems. of Ca3(PO4)2 relative to the other precipitates prepared with
From the foregoing, the major disadvantage of using of lime sodium hydroxide. It is worth mentioning that Ca3(PO4)2 is the
saccharate in calcium phosphate precipitation is the high levels of precursor for hydroxyapatite formation, which is an effective
residual CaO in the juice at high pH. However, the use of lime adsorbent for protein removal;1 therefore, it is beneficial to
saccharate results in significant low levels of P2O5, relative to the include lime saccaharate in raising juice pH for clarification.
other alkalis. As no one alkali can be used to achieve both Calcium Phosphate Precipitation in Filtered Primary
maximum calcium and phosphate removal, a combination be- Juice. In order to provide additional information on the solution
tween lime saccharate and sodium hydroxide or between lime chemistry of calcium phosphate in the presence of juice con-
saccharate and ammonium hydroxide should reduce the con- stituents, some selected experiments were conducted with a
centrations of these ions in clarified juices. filtered factory primary juice using a combination of alkalis.
1832 dx.doi.org/10.1021/jf1043212 |J. Agric. Food Chem. 2011, 59, 1829–1836
Journal of Agricultural and Food Chemistry ARTICLE

Table 2. Composition of Calcium Phosphate Precipitate for Trials Using Juice A at a pH of 8

type of base juice temp. (°C) CaO (mmol/L) P2O5 (mmol/L) CaO/3P2O5

NaOH 50 22.95 ( 0.51 7.92 ( 0.05 0.97

76 23.79 ( 1.61 8.27 ( 0.55 0.96
90 25.23 ( 0.41 8.49 ( 0.51 0.99
lime saccharate 50 26.28 ( 2.20 8.08 ( 0.07 1.08
76 27.40 ( 1.20 8.46 ( 0.22 1.08
90 29.09 ( 0.64 8.54 ( 0.31 1.14
NH4OH 50 21.16 ( 0.41 7.04 ( 0.11 1.00
76 21.96 ( 0.71 7.16 ( 0.51 1.02
90 21.98 ( 0.51 7.40 ( 0.11 1.04

Table 3. Effects of Added Lime Saccharate, Sodium Hydroxide, and Ammonium Hydroxide on the Concentrations of CaO, MgO,
and P2O5 in Filtered Primary Juice

composition of treated juice

type of base pH of juice temp of juice (°C) CaO (mg/L) P2O5 (mg/L) MgO (mg/L) free Ca as CaO (mg/L) wt of precipitate (mg)

none 5.1 40 301 338 318 nd nd

lime saccharate 7.0 76 339 210 321 55 120
90 359 191 336 58 90
8.0 76 299 91 294 55 230
90 358 82 303 44 260
NaOH 7.0 76 202 231 322 66 50
90 182 228 313 40 70
8.0 76 126 142 279 51 90
90 131 138 279 44 140
NH4OH 7.0 76 210 273 354 53 20
90 173 207 320 50 70
8.0 76 173 197 323 52 80
90 112 122 276 41 140
lime sacch/NaOH 6.0/8.0a 76 149 122 285 39 130
90 149 122 274 46 190
6.5/8.0 76 162 90 270 44 240
90 175 90 269 40 260
7.0/8.0 76 195 91 311 44 190
90 221 91 279 43 210
lime sacch/NH4OH 6.0/8.0 76 144 98 278 37 180
90 178 119 273 38 180
6.5/8.0 76 171 82 291 39 240
90 177 67 293 36 260
7.0/8.0 76 207 86 296 37 190
90 182 72 269 35 200
a
The first number is the initial pH of liming, and the second number is the final juice pH.
Table 3 shows the results of added lime saccharate, sodium The P2O5 values of juices treated with a two-stage alkali addition
hydroxide, and ammonium hydroxide on the concentrations of process were similar to the lime saccharate system. Table 3 also
CaO, MgO, and P2O5 in filtered primary juice, at different liming shows the levels of unreacted MgO and the weight of precipitates
temperatures and at different juice pH values. It also shows the formed after the addition of alkali. The two-stage alkali process
results of the two-stage processes involving adding lime sacchar- reduced the MgO content of the treated juice relative to the MgO
ate to a heated juice to a set pH and then adding sodium content in juice obtained with lime saccharate treatment. This
hydroxide solution/ammonium hydroxide solution to the final may be because some of the Mg2þ ions have occupied vacant
pH, The magnesium ion (as MgO) concentrations in the juices Ca2þ sites in the crystal lattice of calcium phosphate based on the
were monitored because of their influence on calcium phosphate concept of structural substitution in the crystal lattices of ions of
precipitation.5 As expected, samples that were treated first with similar sizes.11
lime saccharate and then with either sodium hydroxide or The weight of the precipitate gives an indication of the total
ammonium hydroxide gave significant lower CaO values com- amount of impurities removed from solution. An adequate
pared with the samples treated with only lime saccharate. quantity of precipitate must be formed in the juice to enhance
1833 dx.doi.org/10.1021/jf1043212 |J. Agric. Food Chem. 2011, 59, 1829–1836
Journal of Agricultural and Food Chemistry ARTICLE

sweep flocculation (i.e., the ability of the flocs to adsorb (27% to 50%) were obtained with samples clarified with the lime
impurities as the flocs settle) and increase the settling rates of saccharate and sodium hydroxide solution mixtures compared to
the juice particles. In order to be able to ascertain the best the clarified juice obtained by the normal process (i.e., control).
treatment option, the levels of unreacted CaO, P2O5, and MgO The correlation equation between the amount of CaO and Na2O
together with the corresponding weights of precipitate formed content is given by:
were examined as an entity. On the basis of this, the procedure
which involved first adding lime saccharate to juice to a pH of Y ¼ 2
10-4 x þ 89:7810-2 ; R 2 ¼ 0:60 ð6Þ
6.5 and then adding sodium hydroxide to a pH 8 at a juice
temperature of either 76 or 90 °C was identified as the best where Y is the ratio of the CaO content for the two-stage process
option for further investigation. However, it should be noted to the conventional process, and x is the amount of sodium
that the largest quantity of precipitate was formed with the juice hydroxide added as Na2O in mg/L.
limed at 90 °C than the juice limed at 76 °C, although the level As such, the amount of CaO in clarified juice can be regulated
of unreacted CaO was higher at the higher temperature. Similar on the basis of the amount of sodium hydroxide added during
results were obtained for the two-stage processes involving lime the clarification process. The clarification results of Table 4 also
saccharate and ammonium hydroxide addition. One advantage show a reduction in the MgO contents of up to 20% for clarified
of using ammonium hydroxide over sodium hydroxide is that juices obtained with mixed alkalis, though this is variable. For
ammonium ions are readily volatilized by heat and so do not have these juices, some reductions in SiO2 values (juice 4, Table 4)
the melassigenic action associated with sodium ions on sucrose. were also obtained.
However, the use of ammonium hydroxide is compromised by Although these studies focused on the use of lime saccharate in
the release of corrosive ammonia to the atmosphere. the two-stage process, it is envisaged that similar results would be
The levels of free calcium ions (reported as CaO) in the obtained with milk of lime. It should be noted that clarified juices
treated samples are shown in Table 4. These values are 6 times obtained with lime saccharate alone have higher CaO levels but
lower than the values obtained by ICP-OES. This means that lower P2O5 contents than clarified juices obtained with milk of
most of the Ca2þ ions in juice are bound and are not readily lime.2 Thus, if milk of lime was used to replace lime saccharate
available for chemical interaction because of the formation of there would be lower CaO levels in the clarified juices for these
complexes with the organic acids in the juice as has been deduced systems.
from the synthetic juice study. As such, during the clarification of Typically, during sugar cane juice processing, the heating
juice with lime saccharate, the inorganic phosphate ions prefer- surface area of evaporators becomes increasingly scaled as a
entially react first with the available free endogeneous calcium result of fouling. With increasing scale deposition, the rate of heat
ions in juice before reacting with the calcium ions from the added transfer across the evaporator heating surface area is diminished,
lime saccharate, leaving the calcium complex in solution. This and the net effect is a reduction in the capacity that juice can be
results in high residual calcium in clarified juice. It is presumed processed to concentrated juice. To maintain a nominated sugar
that the degree of association between calcium and saccharate is cane processing rate, the first effect calandria pressure will
far less than that between calcium and organic acids. In situations increase so as to provide a larger temperature difference across
where sodium hydroxide or ammonium hydroxide is used, after the evaporator set. However, once the low pressure steam valve
the initial reaction between phosphate ions and the free endo- to the first effect is fully opened, the evaporation rate for the set
geneous calcium ions, further reactions will inevitably proceed will decline with increasing scaling. There are three main options
with the calcium complex in solution. that can be employed to maintain the crushing rate at the target
Clarification Experiments with Factory Primary Juice. value. These are as follows:
Calcium phosphate precipitation studies on filtered primary juice (1) Reduce the addition of water to the mills and filters.
indicated significant reduction in the levels of CaO and to a lesser However, a reduction in the amount of added water to the
extent MgO in samples treated with the two-stage process, with milling train and the filter station generally reduces the
initial liming to pH 6.5, compared to samples treated with only amount of sucrose recovered from bagasse (i.e., sugar
lime saccharate (i.e., control). In light of this, clarification experi- cane fiber) and the residue from clarification (i.e., mud).
ments using this two-stage process were carried out on factory Consequently, during the period when the amount of
juices on-site at three sugar mills. In one set of experiments for added water is reduced, the effect of scale deposition is to
the two-stage process, lime saccharate was added first to raise the reduce sugar recovery and hence reduce the revenue to
juice pH to 6.5, followed by the addition of sodium hydroxide the mill.
to raise the juice pH to the desired value. In another set of (2) Allow the syrup concentration to reduce and increase the
experiments, the order of addition of lime saccharate and sodium evaporation load on the pan stage. The impact of reduced
hydroxide was reversed. Juices 1 and 2 were from different syrup concentration on the factory depends on many
factories, while juices 3, 4, and 5 were from the same factory factors including the capacity of the pan stage (where
(Table 4). As shown in Table 4, the turbidity, color, and sucrose crystallization is carried out), the sucrose content
phosphate levels of the control samples were marginally lower of the cane supply, and the pressure of the vapor/steam
than the samples which had been treated with the two-stage supplied to the pans.
process. The differences in the settling rate and mud volume (3) A combination of options 1 and 2.
between the control and the two-stage process were not sig- In practice, sugar factories generally adopt option 3 in order to
nificant. The apparent purity of the clarified samples was similar maintain the crushing rate at a target value. Despite this practice,
(except the control of juice 3). sugar factories record sugar and molasses losses. Therefore, by
The two-stage processes give lower levels of CaO and P2O5 in adopting a strategy of using a clarification procedure involving a
clarified juices, compared to the conventional clarification pro- combination of calcium and sodium alkalis will reduce scale
cess (Table 4). As shown in Table 4, significant CaO reductions formation and deposition and thus limit the extent option 3 is
1834 dx.doi.org/10.1021/jf1043212 |J. Agric. Food Chem. 2011, 59, 1829–1836
Journal of Agricultural and Food Chemistry ARTICLE

Table 4. Results of Clarification Experiments on Factory Primary Juices

settling rate turbidity apparent color at CaO mg/kg MgO mg/kg P2O5 mg/kg SiO2 mg/kg
liming typea (cm/min) mud vol. (%) Per 100 Bx purity (%) pH 7 on dry solids on dry solids on dry solids on dry solids

juice 1 10300 883 1749 745 453

L-76-7.8 37 27 52.7 87.1 9200 2322 1701 15.8 396
LC-76-7.5 60 17 62.5 86.8 10700 1119 1786 52.1 390
LC-76-7.8 48 22 67.3 87.0 11300 1425 1688 46.5 330
LC-76- 8.0 37 29 67.3 87.1 11000 1440 1683 62.1 409
juice 2 13200 999 1709 749 483
L-76-7.8 38 27 52.9 88.5 11400 2359 1650 21.2 418
LC-90-7.5 43 30 68.0 88.7 11900 1355 1675 36.6 377
LC-90-7.8 51 23 67.8 88.5 12000 1335 1580 41.7 375
LC-90-8.0 54 21 78.8 88.4 12000 1145 1429 31.5 383
juice 3 1422 1730 908 432
L-76-7.8 48 22 72.9 86.8 12300 2354 1563 20.8 401
CL-76-7.5 48 24 93.2 88.4 14400 1471 1367 46.6 377
CL 76-7.8 42 28 83.8 88.3 13700 1518 1277 31.4 375
CL-76-8.0 55 20 83.7 88.2 12300 1565 1240 20.9 383
juice 4 1214 1556 1041 526
L-76-7.8 48 23 47.1 89.1 10700 2261 1533 15.7 460
CL-90-7.5 47 24 52.2 89.0 11200 1648 1403 31.3 136
CL-90-7.8 27 56.9 89.1 11000 1619 1107 20.7 264
CL-90-8.0 32 33 57.3 88.9 11000 1515 942 20.8 255
juice 5b 1323 1894 653 495
L-76-7.8 38 28 7 87.7 8400 2499 1794 15.7 424
LC-76-7.8 46 24 8 87.9 9800 1463 1668 26.3 337
LC-76-7.8 46 26 7 87.8 9000 1479 1688 31.3 381
a
L is lime saccharate clarification; LC is the two-stage process with lime saccharate added to juice to raise the to juice pH to 6.5, followed by the addition
of sodium hydroxide solution to raise the juice pH to the desired value prior to clarification; CL is the two-stage process with sodium hydroxide solution
added to juice to raise the pH to 6.5, followed by the addition of lime saccharate to raise the juice pH to the desired value, prior to clarification. The first
number is the temperature of liming, and the second number is the final pH, e.g., L-76-7.8 represents liming juice to pH 7.8 at 76 °C. b Clarification
carried out with the P2O5 level in the juice increased to 300 mg/L. The P2O5 level in the juice prior to clarification was 120 mg/L.

used. The down side of using this clarification procedure is the has the potential benefits of reducing fouling of sugar mill
possible melassigenic property (i.e., a substance that binds to evaporators, pans, and centrifugals, hence extending the times
water and sucrose) associated with sodium ions, reducing between evaporator cleaning. The use of such a clarification pro-
sucrose crystallization. Previous laboratory and limited factory cess may minimize options employed to maintain target crushing
experiments by Saska and Cuddihy,24 indicated that soda ash rate as a consequence of evaporator fouling.
(i.e., sodium carbonate) may be used to replace lime in juice
clarification with a different type of flocculant. The sugar
produced in the laboratory using the process was of quality ’ ASSOCIATED CONTENT
similar to that of commercial raw sugar, in terms of color and ash
contents.24 As the experiments were limited, the impact of the bS Supporting Information. Information on the concentra-
process of scale formation was not accessed, including on the tions of residual CaO and P2O5 of juice A after calcium
effect of sucrose crystallization. phosphate precipitation at 76 °C; information on the composi-
In summary, the results from the study of CaO/phosphate/ tions of treated synthetic juices B, C, and E, heated at various
sucrose systems have provided further insights on the chemistry temperatures and pH values; graphical plots showing the effects
of calcium phosphate formation in sugar solutions and have of added NaOH on CaO and MgO levels in clarified juices. This
clearly shown that no single alkali can be used alone to achieve material is available free of charge via the Internet at http://pubs.
maximum calcium phosphate precipitation with reduced acs.org.
amounts of CaO and P2O5 in treated juices. Excessive CaO in
clarified juice can be minimized by using a combination of alkalis, ’ AUTHOR INFORMATION
at a particular initial pH set-point, during juice clarification as has Corresponding Author
been demonstrated with factory primary juices. The present *Tel: þ61 7 3138 1245. Fax: þ61 7 3138 4132. E-mail:
study has also shown that the two-stage clarification process can w.doherty@qut.edu.au.
operate under similar operating conditions (e.g., same final lime
pH and liming temperature) and use the same type of flocculant Funding Sources
as the conventional process. This, therefore, makes it easier to be I thank Sugar Research and Development Corporation and
adopted by sugar mill process engineers. The two-stage process Australian Sugar Mills for financial assistance in the project.
1835 dx.doi.org/10.1021/jf1043212 |J. Agric. Food Chem. 2011, 59, 1829–1836
Journal of Agricultural and Food Chemistry ARTICLE

’ REFERENCES
(1) Doherty, W. O. S.; Edye, L. A. An overview on the chemistry of
clarification of cane sugar. Proc. Aust. Soc. Sugar Cane Technol. 1999, 21,
381–388.
(2) Doherty, W. O. S.; Greenwood, J.; Pilaski, D.; Wright, P. G. The
effect of liming conditions in juice clarification. Proc. Aust. Soc. Sugar
Cane Technol. 2002, 24, 443–451.
(3) Bennett, M. C.; Regnauth, J. M. The effects of calcium and
phosphate in cane juice clarification. Int. Sugar J. 1960, 60, 13–16.
(4) Nancollas, G. H.; Amjad, Z.; Koutsoukos, P. Calcium phosphate
speciation solubility kinetics considerations. ACS Symp. Ser. 1979, 93,
475–497.
(5) Feenstra, K. P.; Hop, S.; De-Bruyn, A. The influence of small
amounts of magnesium ions on the formation of calcium phosphate in
moderately supersaturated solutions. J. Colloid Interface Sci. 1981, 83,
583–588.
(6) Kibabezyc, C. W.; Zielenkeiwicz, A.; Zielenkeiwicz, W. Colori-
metric investigation of calcium phosphate precipitation in relation to
solution composition and temperature. Thermochim. Acta 1988, 131,
47–55.
(7) Elliott, J. C. Studies in Inorganic Chemistry, 18. Structure and
Chemistry of the Apatites and Other Calcium Orthophosphates; Elseiver:
London, 1994.
(8) Jourani, A.; Bounahmidi, T. The kinetics of calcium phosphate
precipitation in clarification: II Sucrose solutions with Mg2þ and SO42-
impurities and industrial cane juice. Int. Sugar J. 1995, 97, 586–592.
(9) Simpson, R. The chemistry of clarification. Proc. S. Afr. Sugar
Technol. Assoc. 1996, 70, 267–271.
(10) Honig, P. Principles of Sugar Technology, 2nd ed.; Elsevier:
Amsterdam, The Netherlands, 1953.
(11) Deben, J. G. The chemistry of calcium phosphate precipitation
in cane juice clarification, Parts I, II and III. Int. Sugar J. 1976, 58 (35-
39), 73–75.
(12) Jourani, A.; Bounahmidi, T.; Laguerie, C. The kinetics of
calcium phosphate precipitation in clarification: II Sucrose solutions
with Mg2þ and SO42- impurities and industrial cane juice. Int. Sugar J.
1995, 97, 586–592.
(13) Boskey, A. L.; Posner, A. S. Magnesium stabilization of amor-
phous calcium phosphate, kinetic study. Mater. Res. Bull. 1974, 9, 907–
916.
(14) Tomazic, B.; Tomson, M.; Nancollas, G. H. Growth of calcium
phosphates on hydroxyapatite crystals: the effect of magnesium. Arch
Oral Biol. 1975, 20, 803–808.
(15) Nancollas, G. H.; Tomazic, B.; Tomson, M. A. The precipita-
tion of calcium phosphates in the presence of magnesium. Croat. Chem.
Acta 1976, 48, 431–438.
(16) Hidi, P. The effect of magnesium on clarification and the
phosphate content of raw sugars. Aust. J. Appl. Sci. 1964, 15, 35–40.
(17) Sutherland, D. N.; Tan, C. T. Sedimentation of a porous sphere.
Chem. Eng. Sci. 1970, 25, 1948–1950.
(18) Steindl, R.J Dirt - its implications for the clarifier and filter
stations. Proc. Soc. Sugar Cane Technol. 1998, 20, 484–490.
(19) Taylor, J. K. Quality Assurance of Chemical Measurements; Lewis
Publishers: Chelsea, MI, 1987.
(20) BSES. Laboratory Manual for Australian Sugar Mills; Method
10; Bureau of Sugar Experiment Stations: Indooropily, QLD, Australia,
2001; Vol. 2.
(21) ICUMSA method GS1/3-7. 2002.
(22) BSES. Laboratory Manual for Australian Sugar Mills; Method
19; Bureau of Sugar Experiment Stations: Indooropily, QLD, Australia,
2001; Vol. 2.
(23) Perrin, D. D.; Dempsey, B.; Serjeant, E. P. Prediction for Organic
Acids and Bases; Chapman and Hall: London, 1981.
(24) Saska, M.; Cuddihy, J. Replacement of lime with soda ash in
cane juice clarification. Sugar J. 2005, 68, 36–41.

1836 dx.doi.org/10.1021/jf1043212 |J. Agric. Food Chem. 2011, 59, 1829–1836