THE APPLICATION OF CALCIUM PHOSPHATE PRECIPITATION CHEMISTRY
TO PHOSPHORUS RECOVERY: THE INFLUENCE OF ORGANIC LIGANDS
Jacqueline A.M. van der Houwen* (jacv@nhm.ac.uk) and Eugenia Valsami-Jones (evj@nhm.ac.uk)
Department of Mineralogy, The Natural History Museum, Cromwell Road, SW7 5BD, London, UK
Abstract
Current knowledge of calcium phosphate precipitation chemistry is brought together and compared with the
needs for understanding application in processes for recovering phosphates from wastewaters. The effects of the
high concentrations of organic species present in such waters appears as a key factor requiring further research
and laboratory experiments in this area are described. The supersaturation required for precipitation of calcium
phosphate at 25ºC, pH 7, 0.1 M ionic strength and near-stoichiometric (for hydroxylapatite) calcium to
phosphate molar ratio was determined under homogeneous precipitation conditions. The experiments were
carried out in air. The phase precipitated at the critical concentration was allowed to grow using a constant
composition method. The influence of organic ligands on the precipitation was investigated using two small
molecular weight organic ligands, acetate and citrate, present at a concentration of 10 -3 M. The precipitate was
studied using X-ray diffraction. Good reproducibility of the experiments, which were carried out in triplicate,
was observed.
The study assessed the supersaturation degree necessary for precipitation of hydroxylapatite to be 10.28,
assuming a solubility constant of log K=-57.13. The required supersaturation was not affected by the presence of
acetate. However, citrate was found to increase the degree of supersaturation to 11.07. It is proposed that this is
due to binding of citrate on the active sites of newly formed nuclei, thereby inhibiting precipitation. All
experiments showed formation of a poorly crystalline hydroxylapatite.
Keywords: citrate, homogeneous precipitation, hydroxylapatite, supersaturation
INTRODUCTION
The precipitation of calcium phosphates has been studied extensively in the past 30 years with the main focus on
biological processes as, for example, bone and tooth mineralisation. In the last two decades further interest in
calcium phosphate precipitation has arisen from its potential as a mechanism of phosphorus removal from
wastewater (1,2), although few studies have considered the underpinning chemistry. There is currently a
particular need for research on calcium phosphate precipitation in organic rich systems that simulate wastewater
environments. A small number of studies exist, which have assessed the effect of organic ligands on the
precipitation of dicalcium phosphate dihydrate (DCPD) (3,4), octacalcium phosphate (OCP) (5) and the
inhibition of hydroxylapatite (HAP) precipitation (6). All these studies were carried out as seeded growth
experiments (heterogeneous nucleation) and all reported inhibition of precipitation kinetics by the presence of
the organic ligands. Adsorption of the ligands on active growth sites of the seeding material was given as the
cause for this inhibition. The influence of citrate on calcium phosphate precipitation onto OCP seeding material
has been studied (7). This study concluded that phospho-citrate complexes formed on active growth sites
inhibited precipitation kinetics.
The study of calcium phosphates is further challenged by uncertainty over the first phase to precipitate. Previous
research suggested the formation of an amorphous calcium phosphate preceding the formation (or
transformation) of a more crystalline calcium phosphate (8). Which specific crystalline calcium phosphate forms
will depend mostly on pH and kinetics (9). The phase predicted to be stable is DCPD (CaHPO 4*2H2O) at acidic
pH around 5, OCP (Ca4H(PO4)3*2.5H2O) at pH around 6 and HAP (Ca5(PO4)3OH) at pH of 7 and above (10).
Previous studies into homogeneous nucleation proceeded by mixing of calcium and phosphate solutions at a
desired supersaturation degree and leaving the nuclei enough time to grow into crystals whilst keeping the pH
constant (11). A disadvantage of this method is that the solution composition will change continuously (i.e.
calcium and phosphate concentrations fall), as soon as precipitation starts. This, in turn, may result in variation
in the nature (mineralogy and/or crystallinity) of the precipitate. Homogeneous nucleation has been criticised for
giving non-reproducible results due to chance nucleation onto foreign particles in the solution (12). The aim of
the present study was to elucidate the chemical principles of calcium phosphate precipitation at neutral pH,
particularly in the presence of organic ligands. The critical supersaturation degree for homogeneous precipitation
of calcium phosphate at pH of 7 was determined. The small molecular weight organic ligands studied were
acetate and citrate, which possess one and three functional groups (carboxylates) respectively. Investigation into
homogeneous nucleation is important in order to compare and contrast with previous work on heterogeneous
nucleation, but also to help understand problems such as the formation of fines nucleating spontaneously in
industrial processes.
MATERIALS AND METHODS
Homogeneous precipitation calculations
For homogeneous precipitation to occur, the solution needs to be supersaturated with respect to a
mineral phase. The degree of supersaturation is given by the following equation (i).
SCa/P
=
log(IAP/Ksp)
(i)
Where IAP is the ionic activity product and Ksp is the solubility constant for the calcium phosphate mineral (13).
To determine the supersaturation degree required for precipitation it is necessary to state the ionic products for
each of the expected calcium phosphate minerals (equations ii-v).
Log IAPDCPD
= log (Ca2+) + log (HPO42-)
Log Ksp
= -6.68 (14)
Log IAPTCP
= 3 log (Ca2+) + 2 log (PO43-)
Log Ksp
= -28.9 (15)
Log IAPOCP
= 4 log (Ca2+) + 3 log (PO43-) + log (H+)
Log Ksp
= -46.9 (16)
Log IAPHAP
= 5 log (Ca2+) + 3 log (PO43-) + log (OH-)
Log Ksp
= -57.4 (17)
(ii)
(iii)
(iv)
(v)
For the calculation of the free ion concentrations of the lattice parameters and subsequently the supersaturation
degree the computer program PHREEQC was used with the Minteq database (18). The activity was calculated
using the Davies equation (vi).
Log f = -Az2I/(1+I-0.3I)
f=
activity
z=
valency
I=
Ionic strength
(vi)
A is 0.5 for water at 25C (19).
Solution preparation
The solutions were made from stock solutions of  0.1 M organic acid. All chemicals used were of
Analar® purity or better. Acetic acid was made up from glacial acetic acid solution and citric acid was made up
using citric acid salt. The stock solutions were standardised with 0.1 M NaOH and were found not to deviate
more than 7%. The NaOH solution was standardised with potassium hydrogen phthalate (COOHC 6H4COOK)
solutions. These titrations were carried out with a Mettler DL55 automatic titrator. Stock solutions of 0.1 M
calcium and 0.06 M phosphate were made from calcium chloride and sodium phosphate salts; their
concentrations were confirmed by ICP-AES analyses. Working solutions of 10-3 M organic acid were made in
NaCl with a total ionic strength of 0.1 M. The initial calcium to phosphate concentrations were determined by
calculation of concentrations of calcium and phosphate at a low supersaturation degree of 2. For these
calculations the computer program PHREEQC was used (18). The solubility constant used in the calculations
was determined in pH-drift experiments at 25C, ionic strength 0.1 M and final pH 7 ( 0.1) and was found to be
log K= -57.4 (mol l-1)9 (17). The calcium and phosphate solutions added subsequently in order to slowly increase
supersaturation, and the titrants were made at the molar ratio stoichiometric to hydroxylapatite (calcium to
phosphate: 1.67  0.1) from the stock solutions, and adjusted to the required pH before making up the volume,
in order to maintain exact concentrations. All solutions were made up in 0.1 M NaCl in order to maintain
constant ionic strength during the experiment. The pH of the solutions was adjusted by addition of 0.1 M NaOH
and equilibrated for 24 hours in air. All solutions used in these experiments were filtered through a 0.2 m
Millipore® filter to minimise the possibility of introducing particles (e.g. dust) which could interfere with
nucleation.
Precipitation experiments
In these experiments an overhead propeller stirrer was used at a speed of 180 rpm. The experiments
were carried out in a water bath at 25C. Experiments were carried out in triplicate to ensure reproducibility.
To determine the saturation degree necessary for precipitation, calcium and phosphate were introduced stepwise.
After each addition of calcium and phosphate solutions, 30 minutes was given as an induction period before the
next addition. Two samples were taken at each increment: sample 1 was taken 1 minute after addition and
sample 2 was taken immediately before the next addition to check stability of the calcium and phosphate
concentrations during this induction period. A continuous drop in pH indicated precipitation. In order to
maintain constant experimental conditions during precipitation, and to produce enough precipitate for
identification, a constant composition method was used (20). The principle of the method is that at the critical
supersaturation, where nucleation begins, the drop in pH triggers the addition of calcium, phosphate and a pH
adjusting solution (base). This maintains calcium, phosphate and pH at a constant level and allows the
precipitate to grow. The experimental set-up is given schematically in figure 1.
Pump
Titrants
Ca
P
OH
pH-controller
Overhead stirrer
pH-electrode
Figure 1.
Schematic experimental set-up for precipitation under constant composition conditions.
The titrant concentrations were provided at the stoichiometric ratio of hydroxylapatite (calcium to phosphate
1.67  0.1) for calcium and phosphate. Preliminary experiments showed that the pH was only controllable when
using a higher (than stoichiometric) concentration of base. The molar ratio of calcium: phosphate: hydroxyl-ions
used for the titrants was 5:3:4. At the end-point of the experiment the solution was rapidly filtered through a 0.2
m syringe filter and then dried at 40C for identification.
Analyses
During the experiments 1.5 ml samples were taken to determine calcium and phosphate concentration
of the solution. 0.8 to 1 ml of sample was diluted 10 times and acidified with 2% HNO 3. Calcium and
phosphorus were measured using Inductively Coupled Plasma Atomic Emission Spectrometry (ICP-AES) with
an ARL 3410. The precipitate was identified using X-ray diffraction (Enraf Nonius).
RESULTS AND DISCUSSION
Table 1 lists the initial concentrations, which show that the experiments started at the stoichiometric
ratio for HAP of 1.67  0.08. The experiments were controlled at pH 7.03 with an accuracy of  0.05 and
experiments were found to be reproducible.
Table 1.
Initial calcium and phosphate concentrations of the experiments.
Calcium (10-4 mol l-1)
Experiment
Phosphate (10-4 mol l-1)
Molar ratio (Ca:PO4)
Control (no organic ligand)
3.37
1.96
1.714
Acetate
3.48
1.99
1.746
Citrate
3.29
2.00
1.644
Experimental results are shown in figures 2a, b and c. The figures display one of each of the triplicate result
graphs for the three different experimental conditions studied here (control, acetate, citrate). These graphs show
the stepwise increase of calcium and phosphate over time until precipitation takes place.
Precipitation in the control at pH=7
0.006
concentration (mol l -1)
0.005
0.004
Precipitation in the presence of 10-3M acetate
Phosphate
Calcium
0.006
0.003
concentration (mol l -1)
0.005
0.002
0.004
0.001
Critical point of
precipitation
Phosphate
Calcium
0.003 0
0
50
100
150
200
250
time
(min)
Critical
point of
precipitation
0.002
0.001
0
0
50
100
150
time (min)
200
250
Precipitation in the presence of 10-3M citrate
0.008
0.007
concentration (mol l-1)
0.006
0.005
Phosphate
Calcium
0.004
0.003
Critical point of
precipitation
0.002
0.001
0
0
50
100
150
200
250
time (min)
Figures 2a, b and c.
Determination of the critical point of precipitation in the different experiments.
In these figures the calcium and phosphate concentrations between additions is stable which indicates no
precipitation occurs until the slow drop in pH (critical point of precipitation shown as the dashed line). The
calcium to phosphate molar ratio slightly increases during the stepwise addition process but was found not to
increase over 1.8 (on average). At the critical point of precipitation the automatic addition of the titrants
commences.
Figures 2a, b and c also show that that the calcium and phosphate concentrations remain constant during
precipitation. The addition of the titrants had to continue for up to 10 minutes, in order to recover enough
precipitate. In table 2 the critical concentrations are given upon which precipitation took place. These
concentrations are the average of three experiments and the range in concentration is also shown.
Table 2.
Critical calcium and phosphate concentrations in the experiments.
Experiment
Calcium (10-3 mol l-1)
Phosphate (10-3 mol l-1)
Control (i.e. no organic ligand)
4.868  0.306
2.644  0.078
Acetate
4.877  0.451
2.683  0.341
Citrate
6.883  0.184
3.874  0.065
These concentrations are used in the calculation of the supersaturation degree using the computer program
PHREEQC (18). The supersaturation degree calculated for the control and the system with acetate present was
the same, within error, at 10.55 and 10.56 respectively. This is to be expected, from the concentrations listed in
table 2. In the presence of citrate however, much higher concentrations of calcium and phosphate were necessary
for precipitation to occur. The higher concentration is partly due to complex formation with citrate in the
solution, which decreases free ion concentration of calcium and phosphate. However the supersaturation degree
is calculated based on free ion concentration, and therefore excludes the effect of complexation. In other words,
variations in the degree of supersaturation imply an effect at the mineral surface. The supersaturation degree in
the presence of citrate was found to be 11.33, which is significantly higher than the control or the acetate
experiment. As this is not an aqueous effect, it has to be the result of a surface inhibition. It is therefore likely
that it is due to citrate binding onto active sites of the newly formed nuclei and inhibiting growth, until there is
enough critical mass of phosphate and calcium in solution for precipitation to occur regardless.
The precipitate was studied by X-ray diffraction. In figure 3 the pattern of one of the precipitates is given in
comparison with the pattern of natural hydroxylapatite; all precipitates showed a similar pattern.
Comparison X-ray diffraction patterns
10000
120 degrees position-sensitive detector
(PSD), NHM-London.
Flat (spinning) powder sample on silicate
deep well.
Fixed beam-sample detector geometry.
Beam to sample angle = 7.5 degrees
Pattern acquisition time = 600 minutes
9000
intensity (arbitrary units)
8000
7000
6000
5000
4000
3000
2000
1000
Natural HAP
0
12
22
32
42
52
62
72
82
92
2 theta
Figure 3.
XRD-pattern of the precipitate formed in comparison with natural hydroxylapatite.
As can be seen in this figure, the pattern shows a poorly crystalline hydroxylapatite but no other phase is present.
These results imply that the first and only crystalline phase to precipitate in all experiments presented here is
hydroxylapatite.
The observation of HAP precipitation in this study is in agreement with the results of Boskey and Posner (11),
who investigated HAP homogeneous precipitation at low supersaturation (saturation degree between 5 and 9) at
pH=7.4 and ionic strength 0.15 M. Their experiments were carried out by rapidly mixing a calcium and
phosphate solution to the desired final supersaturation degree. After nuclei formation the crystals were allowed
to grow keeping the pH constant during this time. The precipitate was identified using electron microscopy and
was found to be identical, even though poorly crystalline, to hydroxylapatite. They concluded the precipitation of
HAP without the formation of a precursor phase.
Nancollas and Tomazic (21) investigated heterogeneous precipitation of calcium phosphate onto HAP seeds at
different levels of supersaturation. Precipitation in seeded experiments can take place at lower supersaturations
than homogeneous nucleation. The range of supersaturation degree for calcium phosphate to precipitate in the
presence of seeding material in their study was between 7 and 11. The saturated solutions were prepared by
mixing calcium and phosphate solutions in which precipitation was initiated by the addition of a seed. The pH
during the experiments was kept constant. The precipitate was studied using X-ray diffraction and infrared
spectroscopy. They found that at high supersaturation (SI=11) an amorphous calcium phosphate (ACP) was
formed as a precursor phase. At a low supersaturation degree (SI=7) the study found formation of
hydroxylapatite as the first phase to precipitate. These previous studies (11, 21) worked with a solubility
constant for HAP of 1.8*10-58 (log K = -57.74) which is similar to the solubility constant used here. Also
Koutsoukos et al (22) reported the direct precipitation of a highly crystalline hydroxylapatite onto seeding
material using a constant composition method and low supersaturation (SI=7). The precipitate was studied using
X-ray diffraction.
A comparison of these studies with the results presented here indicates that the formation of ACP may depend on
the supersaturation degree. In seeded experiments at high supersaturation, similar to the supersaturation degree
necessary for homogeneous nucleation to take place (SI = 11), it is possible for ACP to form as precursor.
Another possibility might be that both ACP and HAP will form simultaneously at a degree of supersaturation
high enough for homogeneous precipitation to occur.
It should be noted that even though no precursor phase to HAP was observed here, and although every care was
taken to minimise the chance of any transformations (e.g. the filtration of the precipitate was carried out as soon
as the precipitate was visible), it is still possible that the first phase formed was another calcium phosphate,
which rapidly transformed into HAP. Appropriate experiments are currently being designed to address this
possibility (precipitation of calcium phosphate from supersaturated solution in a specially designed chamber
within an X-ray diffractometer with a position sensitive detector). The possible formation of a mixture of HAP
with ACP cannot be disregarded as a non-crystalline phase cannot be identified using X-ray diffraction.
Chemical analyses of the precipitate can only provide information on the calcium to phosphate molar ratio of the
precipitate, but will not necessarily prove which calcium phosphate phase has formed, as it is very likely a nonstoichiometric hydroxylapatite may precipitate (23, 24). Finally, it should be noted that the supersaturated
solutions used had the stoichiometric HAP molar ratio, although the precipitated apatites had a molar ratio
ranging from 1.49 to 1.65.
CONCLUSIONS
The results of this study show that reproducible homogeneous precipitation experiments can be
performed under thoroughly controlled conditions. It was found that precipitation in a control experiment at pH
7 took place at a supersaturation degree of 10.55. The influence of acetate was found to be negligible as
supersaturation was similar, at 10.56. Citrate, however, had a pronounced effect, as it required a higher degree of
supersaturation (11.33) for precipitation to take place. This is interpreted as the result of citrate binding to the
active sites of the nuclei inhibiting precipitation. All experiments at pH 7 showed the precipitation of poorly
crystalline hydroxylapatite with no apparent precursor phase.
ACKNOWLEDGMENTS
The authors wish to thank CEEP (Centre Européen d’ Etudes des Polyphosphates, a sector group of
CEFIC, the European Chemical Industry Council) for funding of this project. Gordon Cressey assisted with XRD
identification. Vic Din and Gary Jones are thanked for support and advice with chemical analyses and
experimental set-up. Alan Pethybridge’s academic supervision of Jacqueline van der Houwen is acknowledged.
REFERENCES
1. Joko I. Phosphorus removal from wastewater by the crystallization method. Water Science and Technology,
vol. 17, pp.121-132 (1984).
2. Seckler, M.M., Bruinsma, O.S.L. and van Rosmalen, G.M. Calcium phosphate precipitation in a fluidized bed in
relation to process conditions: a black box approach. Water research, 30: 1677-1685 (1996).
3. Frèche M., Rouquet N., Koutsoukos P. and Lacout J.L. Effect of humic compounds on the crystal growth of
dicalcium phosphate dihydrate. Agrochimica, vol. 36, no.6, pp. 500-510 (1992).
4. Grossl P.R. and Inskeep W.P. Precipitation of dicalcium phosphate dihydrate in the presence of organic acids.
Journal of the American Soil science society, vol. 55, pp. 670-675 (1991).
5. Grossl P.R. and Inskeep W.P. Kinetics of octacalcium phosphate crystal growth in the presence of organic
acids. Geochimica et Cosmochimica acta, vol. 56, pp. 1955-1961 (1992).
6. Inskeep W.P. and Silvertooth J.C. Inhibition of hydroxyapatite precipitation in the presence of fulvic, humic
and tannic acids. Journal of the American Soil science society, vol. 52, pp. 941-946 (1988).
7. Sharma V.K., Johnsson M., Sallis J.D. and Nancollas G.H. Influence of citrate and phosphocitrate on the
crystallisation of octacalcium phosphate. Langmuir, vol. 8, pp. 676-679 (1992).
8. Kibalczyc W. Study of calcium phosphate precipitation at 37 C. Crystal research and technology, vol. 24(8),
pp. 773-778 (1989).
9. De Rooij J.F., Heughebaert J.C. and Nancollas G.H. A pH study of calcium phosphate seeded precipitation.
Journal of colloid and interface science, vol. 100, no. 2, pp. 350-358 (1984).
10. Madsen L.H.E. and Christensson F. Precipitation of calcium phosphate at 40 C from neutral solution.
Journal of crystal growth, vol. 114, pp. 613-618 (1991).
11. Boskey A.L. and Posner A.S. Formation of hydroxyapatite at low supersaturation. Journal of physical
chemistry, vol. 80, pp. 40-45 (1976).
12. Nancollas G.H., Amjad Z. and Koutsoukos P. Calcium phosphates-speciation, solubility and kinetic
considerations, in «Chemical modelling in aqueous systems», pp. 475-497 (1979).
13. Appelo C.A.J. and Postma D. Geochemistry, groundwater and pollution. Balkema, Rotterdam (1993).
14. Marshall R.W. and Nancollas G.H. The kinetics of crystal growth of dicalcium phosphate dihydrate. Journal
of physical chemistry, vol. 73, pp. 3838-3844 (1969).
15. Gregory T.M., Moreno E.C., Patel J.M. and Brown W.E. Solubility of ß-Ca3(PO4)2 in the system Ca(OH)2H3PO4-H2O at 5, 15, 25 and 37ºC. Journal of research of the National Bureau of Standards (A)- Physics and
Chemistry, vol. 78A, no. 6, pp. 667-674 (1974).
16. Moreno E.C., Brown W.E. and Osborn G. Stability of dicalcium phosphate dihydrate in aqueous solutions
and solubility of octacalcium phosphate. Soil science society proceedings, vol. 24, pp. 99-102 (1960).
17. Van der Houwen J.A.M. and Valsami-Jones E. The effect of organic ligands on the dissolution of synthetic
hydroxylapatite. (2001) (in prep.)
18. Parkhurst D.L. and Appelo C.A.J. Users guide to PHREEQC (version 2)- a computer program for speciation,
batch-reaction, one-dimensional transport, and inverse geochemical calculations. U.S Geological Survey, Earth
Science Information center (1999).
19. Stumm W. and Morgan J.J. Aquatic chemistry. John Wiley and Sons, Inc. (1996).
20. Amjad Z., Koutsoukos P.G., Tomson M.B. and Nancollas G.H. The growth of hydroxyapatite from solution.
A new constant composition method. Journal of dental research, vol. 57, pp. 909 (1978).
21. Nancollas G.H. and Tomazic B. Growth of calcium phosphate on hydroxyapatite crystals. Effect of
supersaturation and ionic medium. The journal of physical chemistry, vol. 78, no. 22, pp. 2218-2225 (1974).
22. Koutsoukos P., Amjad Z., Tomson M.B. and Nancollas G.H. Crystallisation of calcium phosphates. A
constant composition study. Journal of the American chemical society, vol. 102, pp. 1553-1557 (1979).
23. Nancollas G.H. and Zawacki S.J. Calcium phosphate mineralisation. Connective tissue research, vol. 21, pp.
239-246 (1989).
24. Zawacki S.J., Heughebaert J.C and Nancollas G.H. The growth of non-stoichiometric apatite from aqueous
solution at 37C: Effect of pH upon the precipitated phase. Journal of colloid and interface science, vol. 135,
no.1, pp. 33-44 (1990).