Advances in Chemical Engineering and Science, 2013, 3, 20-26
doi:10.4236/aces.2013.34B004 Published Online October 2013 (http://www.scirp.org/journal/aces)
Precipitation and Crystallization of Struvite from
Synthetic Wastewater under Stoichiometric Conditions
Anna Kozik1, Nina Hutnik1, Krzysztof Piotrowski2*, Agata Mazienczuk1, Andrzej Matynia1
1Wroclaw University of Technology, Faculty of Chemistry, Wybrzeze Wyspianskiego, Wroclaw, Poland
2Silesian University of Technology, Department of Chemical & Process Engineering, M. Strzody Gliwice, Poland
Email: *krzysztof.piotrowski@polsl.pl
Received June, 2013
ABSTRACT
Phosphate (V) ions were continuously removed from synthetic wastewater containing inorganic impurities using mag-
nesium and ammonium ions. The product was magnesium ammonium phosphate (V) hexahydrate, struvite, MgNH4PO4
× 6H2O. Research ran in stoichiometric conditions in DT MSMPR type crystallizer with internal circulation of suspen-
sion. Increase in process environment pH from 9 to 11 resulted in 3-time decrease of mean struvite crystals size (from
40.1 to 12.6 m). Elongation of mean residence time of suspension in a crystallizer up to 3600 s resulted in improve-
ment of the product quality. Mean size of struvite crystals enlarged up to 50.2 m. Based on kinetic calculations results
(SIG MSMPR model) it was concluded, that linear struvite crystal growth rate varied within 5.04 × 10–9 – 1.69 × 10–8
m/s range, whereas nucleation rate within 1.4 × 107 – 1.7 × 1010 1/(s m3) limits. In solid product, besides struvite, also
all impurities present in wastewater were identified analytically as hydroxides, phosphates and other salts.
Keywords: Struvite; Precipitation; Continuous Reaction Crystallization; Phosphate(V) Ions; Impurity Ions; Continuous
DT MSMPR Crystallizer; Crystal Size Distribution; Kinetics
1. Introduction
Inexpensive and easily available, secondary phosphorus
sources can be industrial, municipal liquid wastes, liquid
manure, urine and other waste solutions containing
phosphate (V) ions [1]. Technological idea of recovery
from them some useful phosphorus compounds is based
on reaction crystallization of sparingly soluble phosphate
salts, mainly struvite MgNH4PO4 × 6H2O (MAP) [2, 3].
Controlled reaction crystallization of struvite is not sim-
ple process. Its course and final results are strongly af-
fected by temperature, concentrations of main reacting
substances (phosphate(V), magnesium and ammonium
ions), process environment’s pH, reaction crystallization
of not only expected product, but also co-precipitating
sparingly soluble salts or/and hydroxides of some metals
present in wastewaters [1, 4-6]. Process success depends
also on the continuous crystallizer construction and its
work mode/parameters, including: mean residence time
of suspension in working volume of the crystallizer, in-
tensity of mixing and circulation of suspension inside the
apparatus, inlet places and contact method of the reacting
substances, etc. [1, 7-9].
The experimental test results concerning recovery of
phosphate(V) ions from solution of similar chemical
composition to agricultural, animal breeding or mineral
fertilizer industry wastewater are presented. Precipitation
of phosphate (V) ions with magnesium and ammonium
ions in alkaline aqueous environment, followed by stru-
vite mass crystallization was carried out in a laboratory
continuous DT MSMPR (Draft Tube, Mixed Suspension
Mixed Product Removal) type crystallizer with propeller
stirrer. Crystallizer was provided with feed solution of
determined chemical composition, prepared earlier in a
mixer from chemically pure substances and deionized
water. The solution contained, besides phosphate (V)
ions, also: magnesium, ammonium, aluminium, calcium,
copper, iron, potassium, zinc and nitrate (V) ions. In this
environment also sodium and chloride ions were present
from dissolved in water crystalline salts from which syn-
thetic wastewater was made. Concentration of phosphate
(V) ions was assumed to be 1.0 mass %. This concentration
is ca. 2 - 5-time higher than reported in real wastewaters.
However, one can thus compare the presented results
with the data concerning struvite production process
from solutions containing phosphate (V) ions only, of
concentration 1.0 mass % [10]. The research was carried
out at molar ratio of the substrates PO4
3– : Mg2+ : NH4
+ as
1 : 1 : 1 in temperature 298 K. Influence of pH (from 9 to
11) and mean residence time of suspension in a crystal-
lizer (from 900 to 3600 s) on product crystal size distri-
*Corresponding author.
Copyright © 2013 SciRes. ACES
A. KOZIK ET AL. 21
butions, mean size and homogeneity within product
population were investigated. Linear growth rate of stru-
vite crystals and its nucleation rate were estimated. Cal-
culations were based on the simplified model of mass
crystallization kinetics in MSMPR crystallizer – SIG
(Size Independent Growth) model.
2. Material and Methods
Photo of experimental plant is shown in Figure 1. It is
fully automated. Continuous Bioengineering RALF plus
Solo plan. Steering, control and acquisition of measure-
ment data streams were carried out with the use of PC
computer (driven by BioScadaLab software). Process ran
in DT MSMPR type crystallizer of working volume Vw
0.6 dm3 (total volume Vt 1.3 dm3). Crystallizer, made of
glass, was equipped with heating/cooling coat providing
stable process temperature, as well as with the system
delivering compressed air, required for stripping of ab-
sorbed CO2 and oxidation of possible organic substances
present in struvite continuous reaction crystallization
environment. Crystallizer diameter was d 100 mm, its
working part height hw 90 mm, total height ht 200 mm.
Inside the crystallizer circulation profile (DT, Draft Tube ,
ddt 52 mm, hdt 50 mm) was installed, inside which four-
paddle propeller stirrer of diameter dm 48 mm operated.
Mixer speed, process temperature, inlet stream of air,
inflows of feed and alkalising solution, as well as out-
flow of product crystal suspension from the crystallizer
were strictly controlled and adjusted by computer.
Synthetic wastewater feeding the crystallizer was aqueous
solution of ammonium di hydrogen phosphate (V)
NH4H2PO4, magnesium chloride MgCl2×6H2O, chlorides
of impurity cations (AlCl3×6H2O, CaCl2×2H2O, CuCl2 ×
2H2O, FeCl3×6H2O, KCl and ZnCl2), as well as sodium
salt of impurity anion (NaNO3). The mixture was pre-
pared in external mixer using crystalline substances (p.a.,
POCh Gliwice, Poland) and deionized water (Barnstead–
NANOpure DIamond). Concentrations of main substrates:
phosphate (V) ions (1.0 mass %) [10], magnesium
(0.256 mass %) and ammonium (0.190 mass %) resulted
from their assumed molar ratio 1 : 1 : 1. Detailed chemi-
cal composition of the feed was presented in Table 1.
This solution was continuously introduced into circula-
tion profile (mixer speed: 4.0 1/s; suspension movement
– downward). Between crystallizer body and circulation
profile (suspension movement – upward) aqueous solu-
tion of sodium hydroxide, of concentration 3 mass %
NaOH was dosed in amount providing the assumed, con-
trolled pH of continuous struvite reaction crystallization
environment. Tests ran in temperature 298 ±0.2 K as-
suming pH 9, 10 or 11 (±0.1) and mean residence time of
suspension in a crystallizer 900, 1800 or 3600 (±20) s.
Compressed air flow was established on the 100 Ndm3/h
level (pressure ca. 2.5 bar). After stabilisation in a crys-
tallizer the assumed parameter values, process in a steady
state ran through another 5. After this time whole crys-
tallizer content was transferred on vacuum filter. Product
crystals were not washed. Using appropriate analytical
methods and procedures there were determined: solid
phase concentration in product crystals suspension (MT),
chemical composition of mother solution and solid phase
(using, among others, plasma emission spectrometer
ICP–AES CPU 7000, spectrometer IR PU9712, atomic
absorption spectrometer iCE 3000, spectrophotometer
UV–Vis Evolution 300), product crystal size distribu-
tions (solid particle laser analyser Beckman Coulter LS
13 320) and crystal habit (scanning electron microscope
JEOL JSM 5800LV). Accuracy of process data determi-
nation in the continuous plant used was estimated to be
ca. 10%.
Kinetic parameters of the investigated continuous struvite
reaction crystallization process were determined based
on population density distributions n(L) of product crys-
tals [11]. The most simplified kinetic model for continu-
ous MSMPR crystallizer – SIG model [12], was used for
the calculations. Crystal population density distribution
equation resulting from the assumed SIG kinetic model
constraints can be presented in the form of Eq. (1):
a
b
Figure 1. Photo of experimental plant for continuous reaction crystallization of struvite: (a) general view, (b) continuous DT
MSMPR type crystallizer unit with internal circulation of suspension.
Copyright © 2013 SciRes. ACES
A. KOZIK ET AL.
22
Table 1. Chemical composition of synthetic wastewater from
fertilizer industry.
Component Concentration mass %
PO43– 1.0
Mg2+ 0.256
NH4+ 0.190
Al3+ 0.002
Ca2+ 0.05
Cu2+ 2·10–5
Fe3+ 2·10–4
K+ 0.025
Zn2+ 2·10–5
NO30.02
pH 3.7

0exp L
nL nG


(1)
from which for L = 0 one can determine the nuclei popu-
lation density n0 value, as well as crystal linear growth
rate G for the known mean residence time of suspen-
sion in a crystallizer. Nucleation rate B can be calculated
from Eq. (2):
0
BnG (2)
3. Results and Discussion
Statistical parameter values of product crystal size dis-
tributions are presented in Table 2.
From the table it results, that diversified struvite crys-
tals of mean size Lm from 12.6 to 50.2 m (L50 from 10.7
to 38.7 m) were produced from the synthetic wastewa-
ter depending on pH of reaction crystallization process
environment and mean residence time of suspension in a
crystallizer. These are large differences, speaking about
significant influence of process parameters on product
crystal sizes. With the pH increase homogeneity within
product crystals population increased (as CV decreased
by more than 10%, from 78.4 to 69.6%), simultaneously
their mean size significantly decreased. Rise of pH from
9 to 11 caused, that Lm values decreased from 40.1 to
12.6 m, thus more than 3 times. Also second statistical
parameter of crystal size distribution L50 decreased
analogously by ca. 67%. With the pH increase struvite
solubility decreases (minimal value corresponds to pH 10,
3 [13] or 10,7 [14]), and its precipitation potential in-
creases [2,3]. All these make, that nuclei population den-
sity enlarges (Table 3), shifting of mean or median crys-
tal size towards smaller values.
Table 2. Experimental test results concerning continuous
struvite reaction crystallization process in DT MSMPR type
crystallizer. Process temperature: 298 K.
Process parameters Crystal product characteristics
No.
pH
s
Lm
m
L50
m
Ld
m
CV
% La/Lb
1 9 900 40.1 32.9 38.0 78.45.4
2 10 900 15.6 13.3 13.6 71.14.6
3 11 900 12.6 10.7 11.6 69.64.4
4 9 1800 43.9 36.4 41.7 79.75.6
5 9 3600 50.2 38.7 43.7 82.56.1
*Without product crystal washing; Lm = ΣxiLi, where: xi – mass fraction of
crystals of mean fraction size Li; L50 – median crystal size for 50 mass %
undersize fraction; Ld – crystal mode size; CV = 100(L84L16)/(2L50), where:
L84, L16, L50 – crystal sizes corresponding to 84, 16 and 50 mass % undersize
fractions).
Table 3. Nucleation rate B and crystal linear growth rate G estimated for struvite reaction crystallization process in a con-
tinuous DT MSMPR type crystallizer with SIG MSMPR model. Process conditions – see Table 2.
Kinetic parameters of the process (SIG MSMPR model)
No.
n(L)*) R2
(for linear segment*) G ×10–9 m/s B ×109 1/(s m3)
1 n = 2.344×1016 exp(–6.570×104L) 0.986 16.9 0.39
2 n = 6.218×1017 exp(–1.489×105L) 0.994 7.46 4.6
3 n = 3.406×1018 exp(–2.206×105L) 0.982 5.04 17
4 n = 1.348×1016 exp(–5.793×104L) 0.992 9.59 0.12
5 n = 2.045×1015 exp(–3.797×104L) 0.971 6.98 0.014
*for L > 50 m (pH 9), L > 20 m (pH 10), L > 10 m (pH 11).
Copyright © 2013 SciRes. ACES
A. KOZIK ET AL. 23
Elongation of mean residence time of suspension in a
crystallizer was responsible for growth of product crys-
tals sizes, even by more than 25%. Struvite crystals
reached mean size Lm 50.2 m for mean residence time
3600 s and pH 9. With the elongation of mean residence
time of suspension mean supersaturation in solution de-
creased, resulting thus in decrease of both kinetic com-
ponents of the process: nucleation rate of solid phase and
its linear growth rate (Tabl e 3). Longer residence time of
crystals in supersaturated solution produced, however,
that their sizes enlarged significantly. In solution of
lower mean supersaturation crystals grew slower, how-
ever longer and more stable. Homogeneity within crystal
population, however, decreased (CV increased from 78.4
to 82.5%), resulting mainly from increase in intensity of
co-running crystal attrition and breakage processes cor-
related with the elongation of their residence time in a
mixed and circulated suspension. From aqueous solutions
containing only phosphate (V) ions of concentration 1.0
mass % at the same crystallizer struvite of generally lar-
ger crystal sizes (by ca. 15% on average) was produced
[10]. These crystals were also more homogeneous. The
largest differences were observed for mean residence
time of suspension in a crystallizer elongated up to 3600
s. For example, for pH 9 and 3600 s: Lm 63.0 m, CV
63.2% (system without impurities, [10]) and Lm 50.2 m,
CV 82.5% (synthetic wastewater, Table 2). Main reason
of these differences is impurities presence. Individual
influence of each impurity is different. Some of these
affect shape and sizes of struvite crystals advantageously
[5], while others, for example, raise nucleation rate or
limit linear growth rate [15]. Their interaction – effect of
not only presence, but also concentration of particular
impurities in a process system, is usually disadvanta-
geous [6,9]. In the discussed case study (see Table 2) net
effect of impurities also turned out to be unfavourable.
Exemplary volumetric (mass) crystal size distributions
of the products are presented in Figure 2. From these it
results, that raise of pH from 9 (Figure 2(a)) up to 11
(Figure 2(b)) resulted in shift of crystal dominant size
(Ld, maximum of differential distribution), towards
smaller values: from 38.0 to 11.6 m. Size and amount
of the largest struvite crystals in a product decreased. The
largest crystal size at pH 9 reached 210 m and then de-
creased to 50 m – at pH 11, thus decreased more than
four times. Simultaneously the smallest crystals fraction
in the product enlarged from 7.4% (pH 9) up to 22.7%
(pH 11) – for the particles of size smaller than 5 m. In
result mean size of struvite crystals shrank more than 3
time (Lm 40.1 12.6 m). Elongation of mean residence
time of suspension up to 3600 s (at pH 9) produced,
however, increase in Ld value of struvite crystals: from
38.0 m (Figure 2(a)) to 43.7 m (Figure 2(c)). Maxi-
mal crystal size reached 250 m, larger by 40 m com-
pared to results for 900 s. Particle fraction of sizes
smaller than 5 m practically not modified in comparison
to 900 s (Figure 2(a)) and was 7.3 – 7.4% (Figure
2(c)). In net result, mean crystal size increased by ca. 10
m (Table 2).
Population density distributions of product crystals
which size distributions are presented in Figure 2 are
presented in Figure 3. From these distribution courses,
presented in lnn L coordinates it results, that for stru-
vite particles of size L > 50 μm (pH 9) and L > 10 m
(pH 11) these courses can be with satisfactory accuracy
approximated with linear function. Appling Equation (1)
one can thus calculate linear crystals growth rate G,
while from Equation (2) their nucleation rate B. Parame-
ters of population density distribution functions for stru-
vite crystals of size L < 50 μm or L < 10 μm (Equation
(1)) and calculated on this basis G and B values are pre-
sented in Table 3. Nonlinearity in population density
distribution courses for the crystals of size L < 50 m or
L < 10 μm (in lnnL coordinates, Figure 3) speaks
about more complex process kinetics than it results from
the preliminary assumed SIG MSMPR model [11,12].
Determined this method kinetic parameter values should
be regarded as the estimated ones only. It especially
concerns nucleation rate values calculated with Equation
(2), with the use of significantly devaluated nuclei popu-
lation density n0 (n(L) for L = 0). As it results from Fig-
ure 3, the differences between n0 values predicted by
a
b
c
Figure 2. Exemplary differential (left scale) and cumulative
(right scale) volumetric (mass) size distributions of crystals
products: a) pH 9, 900 s, b) pH 11, 900 s, c) pH 9, 3600
s (corresponding to No. 1, 3 and 5 in Table 2).
Copyright © 2013 SciRes. ACES
A. KOZIK ET AL.
24
020406080100 120 140 160 180 200 220 240 260
108
109
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
pH 9, 900 s
n = 2.344×1016exp(–6.570×104L)
pH 11, 900 s
n = 3.406×1018exp(–2.206×105L)
pH 9, 3600 s
n = 2.045×1015exp(–3.797×104L)
Population density n, 1/(m m3)
Crystal size L, m
Figure 3. Influence of pH and mean residence time of sus-
pension in DT MSPR type crystallizer on population den-
sity distribution of crystal product: points – experimental
data, dashed lines – n(L) values calculated with Equation (1)
(Table 3) for crystal fractions L > 50 m (pH 9) and L > 10
m (pH 11).
extrapolation using linear SIG kinetic model and real
values reach even 103 – 106. Calculated nucleation rates
of struvite crystals B are thus useful only for relative,
conventional comparison of process parameters influence
on its course and results.
Analysing the kinetic data presented in Table 3, one
can notice decreasing trend of linear struvite crystal
growth rate G with the increase in environment’s pH and
with elongation of mean residence time of suspension in
the crystallizer. Generally larger crystal growth rates are
observed for the shortest mean residence times in appa-
ratus, what is in conformity with the observations con-
cerning the classical continuous mass crystallization
processes [11]. Increase in environment pH from 9 to 11
results in decrease of linear struvite crystals growth rate
from 16.9×10–9 to 5.04×10–9 m/s. It is significant de-
crease of this rate (by more 3 times). It is accompanied
by increase in nuclei population density n0, thus increase
in nucleation rate B (from 0.39×109 to 17×109 1/(s m3)).
As a consequence final mean crystal size Lm decreased
(40.1 12.6 m). Elongation of mean residence time of
suspension in a crystallizer effectively confined struvite
nucleation rate (Table 3 ). Crystal linear growth rate also
decreased, however significantly lower nucleation rate
and longer contact time of crystal phase with supersatu-
rated mother solution resulted in visible growth of mean
struvite crystal size (Table 2). With the elongation of
mean residence time more convenient conditions of mass
transfer between the liquid and solid phases established,
additionally providing more stable growth of crystal
phase. In process conditions characterized by relatively
long mean residence time of suspension in a crystallizer
higher quality product is manufactured. However unit
productivity is lower, thus economical effectiveness of
the production plant is smaller.
Solid products, without water washing of crystals on
the filter (ca. 20-25 mass % of mother solution in a filter
cake) and after drying, contained mainly struvite, but
also hydroxides and salts of impurities from mother solu-
tion (indirectly – from synthetic wastewater). In Figure 4
there are presented scanning electron microscope images
of the exemplary crystalline products. Diversified sizes
of struvite particles are clearly visible. Other solid parti-
cles, co-precipitated from wastewater in the process con-
ditions are also visible. The most often these form ag-
glomerates on struvite crystal surfaces. The best shaped
struvite crystals, of the most advantageous size distribu-
tion, were produced at pH 9 and mean residence time of
suspension elongated up to 3600 s (test No. 5 in Tabl e 2 ,
Figure 2(c) and Figure 4(b)). Based on microscope im-
ages analysis one can conclude, that struvite crystal sur-
face was taken up by co-precipitated solid particles of
hydroxides and impurity salts, what in turn generated
large tensions and stresses within struvite crystal struc-
tures. Thus many crystal cracks, irregular surfaces, de-
formed edges, tubular crystals presence, etc. are observed
(Figure 4).
100
m
b
100
m
a
Figure 4. Scanning electron microscope images of struvite
crystals produced from synthetic wastewater in continuous
DT MSMPR type crystallizer. Process parameters: a) pH 11,
900 s, b) pH 9, 3600 s (corresponding to Figure 2b and c,
respectively).
Copyright © 2013 SciRes. ACES
A. KOZIK ET AL. 25
Table 4. Chemical composition of solid phase and mother
solution after filtration of crystal suspension removed from
continuous DT MSMPR type crystallizer (see Table 2).
Concentration in:
Component
Mother solution mg/kg Solid phase1 mass %
PO43– 15 – 66 37.2 – 40.2
Mg2+ 31 – 66 8.6 – 9.6
NH4+ 70 – 110 6.5 – 7.0
Al 0.05 – 0.5 0.05 – 0.16
Ca 20 – 60 1.8 – 3.4
Cu 0.002 – 0.004 (6 – 11)×10–4
Fe 0.005 – 0.008 0.006 – 0.018
K 63 – 160 0.20 – 0.66
Zn < 0.1 0.09 – 0.18
NO3100 – 150 0.20 – 0.29
1after drying, without water washing of crystals on a filter.
Based on planimetric measurement results involving
50 crystals of each product, randomly selected from three
scanning electron microscope images it was concluded,
that average ratio of their length La to their width Lb var-
ied from 4.4 to 6.1, depending on the process parameter
values (Table 2). Struvite produced in the same crystal-
lizer from aqueous solutions of phosphate (V) ions with-
out impurities characterized with La/Lb ratio ca. 6 (for
[PO4
3–]RM 1.0 mass %, pH 9 and 900 s) [10]. Increase
in pH from 9 to 11 caused, that struvite crystals became
clearly shorter and thinner (La/Lb 5.4 4.4, see Figure
4(a)). Contrary, elongation of mean residence time of
suspension up to 3600 s favoured production of longer
and wider crystals, and length La increment was clearly
larger than width Lb increment. The La/Lb ratio increased
thus up to 6.1 (see Figure 4(b)). One can assume that
struvite crystal sizes and their shape are the net result of
wastewater impurities action and parameters of reaction
crystallization process. From the microscope images it
also results, that agglomeration within struvite crystals
was not significant. It generally speaks advantageously
about process conditions established in the crystallizer in
respect to nucleation and growth of struvite crystals.
Taking, however, under consideration all components of
struvite continuous reaction crystallization process in DT
MSMPR type crystallizer one can notice, that main factor
influencing the process course is supersaturation in
mother solution, very strongly dependent (at constant
composition of feed solution, constant temperature and
constant mixing/circulation intensity) on environment pH
and on mean residence time of suspension in the crystal-
lizer working volume.
In Table 4 the concentration ranges of components in
the post processed mother solution and in solid phase
(without water washing of crystals on the filter and after
their drying) removed from the crystallizer were pre-
sented. Crystal product, as it results from Table 4, be-
sides main component MgNH4PO4 × 6H2O, contained
also all impurities present in synthetic wastewater (hy-
droxides, phosphates (V), chlorides, nitrates (V)). From
the data analysis it results, that aluminium, calcium,
copper, iron and zinc ions practically totally precipitated
(compare concentration of these ions in wastewater (Ta-
ble 1) and in post processed mother solution (Table 4)).
One can also notice, that concentration of phosphate
(V) ions in a post processed mother solution varied from
66 mg/kg (pH 9, 900 s) to 15 mg/kg (pH 9, 3600 s).
This concentration values decreased regularly with the
pH raise and with elongation of mean residence time of
struvite crystals suspension in the crystallizer. From the
comparison it results, that concentration of phosphate (V)
ions can be decreased even 4 times. It is connected with
decrease of struvite solubility with the rising of reacting
mixture pH while longer contact time of crystals with
mother solution in a crystallizer is responsible for more
thorough discharge of the generated supersaturation. The
values of phosphate (V) ions concentration in mother
solution can be regarded small, thus effectiveness of their
removal from the feed solution as fully satisfactory.
4. Conclusions
The sparingly soluble salt, MgNH4PO4 × 6H2O, struvite
was produced from synthetic wastewater, simulating
wastewater from mineral fertilizer industry, agricultural
industry or liquid manure. Process ran in continuous DT
MSMPR type crystallizer. Struvite crystals of mean size
Lm from ca. 13 to ca. 50 m were removed from the
crystallizer. It was proved, that increase in pH (from 9 to
11) of struvite reaction crystallization process environ-
ment produced decrease of crystal mean size by more
than 3 times (Lm 40.1 12.6 m, 900 s). Contrary,
elongation of mean residence time of suspension in a
crystallizer from 900 to 3600 s produced significant
enlargement of this characteristic size by ca. 25% (Lm
50.2 m at pH 9, 3600 s). Products of moderate crystal
homogeneity (CV ca. 76%) were removed from the
crystallizer. It is complex, net effect of feed chemical
composition, pH, mean residence time of suspension, as
well as crystals attrition and breakage on the supersatura-
tion level establishing in mother solution.
From the population density distribution of product
crystals nucleation rate and linear growth rate of struvite
were estimated. It was concluded, that linear crystal
growth rate varied within 5.04×10–9 – 16.9×10–9 m/s
range, while nucleation rate changed within the
0.014×109 – 17×109 1/(s m3) limits depending on process
parameter values. With the increase in environment pH
Copyright © 2013 SciRes. ACES
A. KOZIK ET AL.
Copyright © 2013 SciRes. ACES
26
nucleation rate increased while linear crystal growth rate
decreased, what influenced final size of struvite crystals
disadvantageously. Elongation of mean residence time of
suspension in a crystallizer caused, that both kinetic pa-
rameter (B and G) values decreased. On the other hand
longer contact time of crystals from the supersaturated
mother solution compensated with excess lower crystal
growth rate. In result struvite crystals of relatively large
sizes were produced.
Concentration of phosphate (V) ions decreased from
1.0 mass % in a feed to 15 – 66 mg/kg in a postprocessed
mother solution. It can be regarded as a very good result
of their removal process from the feed. In solid product,
besides main component – struvite, all impurities from
wastewater appeared in form of hydroxides, phosphates
and other salts.
5. Acknowledgements
This work was supported by the National Science Centre
of Poland under grant No. NN 209 0959 40 (2011–2014).
REFERENCES
[1] K. S. Le Corre, E. Valsami–Jones, P. Hobbs and S. A.
Parsons, “Phosphorus Recovery from Wastewater by
Struvite Crystallization: A Review,” Critical Reviews En-
vironmental Science and Technology, Vol. 39, 2009, pp.
433-477. doi:10.1080/10643380701640573
[2] S. Parsons, “Recent Scientific and Technical Develop-
ments: Struvite Precipitation,” CEEP Scope Newslett.,
Vol. 41, 2001, pp. 15-22.
[3] J. Doyle and S. A. Parsons, “Struvite Formation, Control
and Recovery,” Water Research, Vol. 36, 2002, pp.
3925-3940. doi:10.1016/S0043-1354(02)00126-4
[4] A. Matynia, B. Wierzbowska, N. Hutnik, K. Piotrowski,
R. Liszka, T. Ciesielski and A. Mazienczuk, “Method for
Struvite Recovery from Wastewater of Mineral Fertilizer
Industry,” Przem. Chem., Vol. 89, 2010, pp. 478-485.
[5] N. Hutnik, K. Piotrowski, J. Gluzinska and A. Matynia,
“Effect of Selected Inorganic Impurities Present in Real
Phosphate(V) Solutions on the Quality of Struvite Crys-
tals Produced in Continuous Reaction Crystallization
Process,” Progr. Environ. Sci. Technol., Vol. 3, 2011, pp.
559-566.
[6] N. Hutnik, A. Kozik, A. Mazienczuk, K. Piotrowski, B.
Wierzbowska and A. Matynia, “Phosphates(V) Recovery
from Phosphorus Mineral Fertilizers Industry Wastewater
by Continuous Struvite Reaction Crystallization Process,”
Water Research, Vol. 47, 2013.
doi:10.1016/j.watres.2013.04.026
[7] J. Koralewska, K. Piotrowski, B. Wierzbowska and A.
Matynia, “Kinetics of Reaction–crystallization of Struvite
in the Continuous Draft Tube Magma Type Crystallizers
– influence of Different Internal Hydrodynamics,” Chi-
nese Journal of Chemical Engineering, Vol. 17, 2009, pp.
330-339. doi:10.1016/S1004-9541(08)60212-8
[8] A. Mazienczuk, N. Hutnik, K. Piotrowski, B. Wierz-
bowska and A. Matynia, “Continuous Crystallizers with
Jet Pump Driven by Recirculated Mother Solution in Pro-
duction of Struvite,” Przem. Chem., Vol. 91, 2012, pp.
890-895.
[9] A. Kozik, K. Piotrowski, B. Wierzbowska and A.
Matynia, “Recovery of Phosphate (V) Ions from Waste-
waters of Different Chemical Compositions,” Progr. En-
viron. Sci. Technol., Vol. 3, 2011, pp. 550-558.
[10] A. Kozik, A. Matynia and K. Piotrowski, “Continuous
Reaction Crystallization of Struvite from Diluted Aque-
ous Solutions of Phosphate (V) Ions in DT MSMPR
Crystallizer,” Proceedia Engineering, Vol. 42, 2012, pp.
313-322. doi:10.1016/j.proeng.2012.07.422
[11] J. W. Mullin, “Crystallization”, Butterworth–Heinemann,
Oxford, 1993.
[12] A. D. Randolph and M. A. Larson, “Theory of Particulate
Processes: Analysis and Techniques of Continuous Crys-
tallization”, Academic Press, New York, 1988.
[13] K. N. Ohlinger, T. M. Young and E. D. Schroeder, “Pre-
dicting Struvite Formation in Digestion,” Water Research,
Vol. 32, 1998, pp. 3607-3614.
doi:10.1016/S0043-1354(98)00123-7
[14] V. L. Snoeyink and D. Jenkins, “Water Chemistry,”
Wiley, New York, 1980.
[15] N. Hutnik, K. Piotrowski, A. Kozik and A. Matynia,
“Continuous Reaction Crystallization of Struvite from
phosphate(V) Solutions Containing Inorganic Impurities
–SIG kinetic Approach,” Proceedings of the 38th Inter-
national Conference of Slovak Society of Chemical Engi-
neering SSCHE, Tatranské Matliare, 23-27 May 2011, pp.
325-0333.