JCG3382012189 194
JCG3382012189 194
JCG3382012189 194
net/publication/256746144
CITATIONS READS
22 1,682
2 authors:
Some of the authors of this publication are also working on these related projects:
“Theoretical and experimental study of the controlled precipitation of inorganic Salts in granular and consolidated Porous Media. SPM” - Research Project: ARISTEIA II
National Strategic Reference Framework (NSRF) 2007-2013, Coordinator: Prof. Petros Koutsoukos, Chemical Eng. Department, University of Patras View project
All content following this page was uploaded by Aikaterini I Vavouraki on 09 February 2018.
a r t i c l e i n f o abstract
Article history: The crystallization of sodium sulfate decahydrate (Na2SO4 10H2O, mirabilite) from supersaturated
Received 23 February 2011 solutions was investigated using stable supersaturated solutions seeded with mirabilite seed crystals.
Received in revised form The experiments were done in batch, stirred reactors in which the supersaturated solutions were
7 September 2011
prepared either by dissolution of sodium sulfate anhydrous at 32 1C followed by cooling to 18 or 20 1C
Accepted 3 November 2011
or by mixing equal volumes of equimolar ammonium sulfate and sodium hydroxide solutions at 20 1C.
Communicated by B.A. Korgel
Available online 12 November 2011 Inoculation of the solutions supersaturated only with respect to mirabilite with seed crystals was
accompanied with temperature increase of the thermostated solution. Despite the fact that crystal growth
Keywords: was initiated with seed crystals, the process started past the lapse of induction times inversely proportional
A1. Supersaturated solutions
to the solution supersaturation. The rates of crystal growth were measured both from the temperature rise
A2. Crystal growth rates
and from the concentration–time profiles, which were linearly correlated. The measured crystal growth
A2. Growth from solutions
A3. Seed crystals rates showed a parabolic dependence on supersaturation at low supersaturations. For higher values
B1. Mirabilite this dependence changed to linear, a behavior consistent with the BCF spiral crystal growth model.
B1. Sodium sulfate decahydrate The morphology of the crystals growing at 20 1C showed typical prismatic habit, while at 18 1C when
crystallized from cooled sodium sulfate solutions changes in the crystal habit to a leaf like morphology
were observed.
& 2011 Elsevier B.V. All rights reserved.
0022-0248/$ - see front matter & 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jcrysgro.2011.11.007
190 A.I. Vavouraki, P.G. Koutsoukos / Journal of Crystal Growth 338 (2012) 189–194
long time periods (kinetic stabilization) has not yet clearly been mixing equal volumes (100 mL each) of equimolar solutions
described. This supersaturation domain allows for the perfor- ammonium sulfate and sodium hydroxide, prepared from the
mance of careful kinetics measurements and for the precise respective stock solutions. Solutions were stirred magnetically
determination of the effect that additives may have on the crystal throughout the crystal growth experiments. The stock solutions
growth process. Clarification of the mechanism at specific opera- were prepared from the respective crystalline solids (Merck, Pro
tional conditions is of particular importance for the design of the Analisi) using triply distilled water followed by filtration through
appropriate strategies to control the crystal growth of mirabilite membrane filters (0.22 mm Millipore). The sodium hydroxide
in aqueous media. solutions were standardized by titrations of potassium hydrogen
It is the aim of this study to obtain mechanistic information on phthalate. The solutions were left in the reactor under magnetic
the crystal growth of mirabilite from supersaturated solutions stirring (ca. 200 rpm) for 30 min at 32 1C. Next, the solution was
using the seeded growth technique as the most appropriate for cooled at constant rate (0.5 1/min) to 18 and/or 20 1 depending on
relatively low supersaturations. The results of measurements of the experimental conditions in each case. The temperature
the crystallization kinetics of mirabilite in aqueous supersatu- reduction was sufficient for the development of supersaturation
rated solutions, stable for very long time periods, were obtained with respect to mirabilite. The solutions were allowed to equili-
at temperatures of 18 and 20 1C. The two temperature values brate at the new temperature for at least 2 h under stirring. The
selected were sufficient not only for the development of differ- crystallization of mirabilite was initiated by inoculation of the
ences in the solution supersaturation but most important could supersaturated solutions with mirabilite (ca. 250 mg, 71 mg)
provide indications concerning changes in the morphology of the seed crystals (mirabilite, 99þ %, Merck, A.C.S. reagent). The
mirabilite crystals, which grow out of the supersaturated solu- stability of the seed crystals, kept at 4 1C at all times, was
tions. The investigation of the crystal growth kinetics at different regularly checked both with respect to water content and the
temperatures was limited to small differences due to the large X-ray diffraction pattern to ensure that no dehydration took
variation of mirabilite solubility with temperature [22]. Two place. Following the addition of the seed crystals, crystal growth
different methods of preparation of the supersaturated solutions of mirabilite started past the lapse of very small yet reproducibly
were tested: dissolution of sodium sulfate at higher temperature (725 s) measured induction time. The crystal growth of mirabil-
and mixing ammonium sulfate and sodium hydroxide solutions. ite, is exothermal at ambient pressure [23,24] and the process
The stability domain of the system was defined through a series of was therefore followed through the temperature variation as a
preliminary experiments. Subsequent crystal growth experiments function of time using a thermocouple connected to a computer
were done in the stable domain of the supersaturated solutions with the appropriate interface and recording software. In all
using mirabilite seed crystals. experiments of the present work, the crystal growth of mirabilite
on the seed crystals was rather fast (duration 2 min). During the
course of the crystallization, samples were withdrawn, filtered
2. Experimental
through membranes filters (0.22 mm Millipore) in order to sepa-
rate the liquid from the solid. Solids on the filter were character-
All experiments were done in a double walled jacketed reactor,
ized and the filtrates were analyzed for sodium sulfate. Solids
thermostated with circulating water from a programmable water-
were examined by scanning electron microscopy (SEM) and
bath (Cole-Parmer, Polyscience) equipped with heater and refrig-
characterized by powder X-ray diffraction (XRD, Philips 1830/40).
erating unit ( 70.1 1C). The experimental setup is shown in Fig. 1.
Both SEM and XRD analyses were done as fast as possible to avoid
The temperature of the water bath was initially set at 32 1C
dehydration of mirabilite. In SEM, due to the high vacuum, the
and the reactor was thermostatted pumping the thermostatted
onset of dehydration could be clearly seen and at this point
water through the walls of the glass jacket. The working solutions
morphology analysis was terminated. The sodium sulfate con-
were prepared in the reactor at this temperature (at which they
centration analysis, during the course of crystal growth experi-
were undersaturated with respect to mirabilite) either by the
ments was done by evaporation of the liquid (water) in weighing
dissolution of anhydrous sodium sulfate (thenardite, Merck,
crystal vials at 80 1C. The respective mass was measured until
Puriss) in a carefully measured mass of water ( 70.01 g) or by
constant weight. Duplicate analyses were done by conductivity
titrations of the filtrates with standard barium nitrate solutions.
5
Since mirabilite crystals used to seed the solutions were
obtained commercially, their solubility was determined, since it
was possible that the presence of impurities, which could not be
4 detected by our analytical methods could yield differences in the
solubility thus leading to erroneous calculations of the solution
6 supersaturation. The solubility of mirabilite crystals used to seed
the supersaturated solutions in the present work was measured at
1 the experimental temperatures 18 and 20 1C. The measurements
3
were done approaching equilibrium from both under- and super-
saturated solutions. In the former case seed crystals were intro-
duced in undersaturated solutions and were allowed to reach
equilibrium, while in the latter mirabilite was allowed to reach
equilibrium from supersaturated solutions past seeded growth.
More specifically, accurately weighted amounts of anhydrous
2 sodium sulfate were dissolved in 200 mL distilled water by
successive solid additions, until no more salt could be dissolved.
In the case of supersaturated solutions mirabilite (250 mg) was
Fig. 1. Experimental set-up for the investigation of the crystallization of mirabilite introduced in stable supersaturated solutions close to the esti-
from aqueous supersaturated solutions. 1: heated–cooled circulating thermostat;
2: magnetic stirrer; 3: double walled water jacketed glass reactor; 4: thermo-
mated (literature values) equilibrium. The mirabilite crystals,
couple temperature sensor; 5: interface for data acquisition and transfer and with a high surface to volume ratio, required only a short time
6: computer for data collection. to reach equilibrium. Samples were withdrawn over a time period
A.I. Vavouraki, P.G. Koutsoukos / Journal of Crystal Growth 338 (2012) 189–194 191
of 24 h in order to confirm the constancy of the solute concentra- prepared mixing equimolar solutions of ammonium sulfate and
tion. In both cases, the precise concentration of sodium sulfate in sodium hydroxide as described in the experimental section.
the bulk solutions was measured by gravimetric analysis and the Following the introduction of seed crystals in the supersaturated
solubility of sodium sulfate was found to be equal to 1.03 at 18 1C solutions and past a short induction time the solution tempera-
( 70.03) and 1.23 ( 70.05) mol kg 1 H2O at 20 1C. Solubility ture increased and after reaching a maximum decreased to return
values were in good agreement with solubility values reported to the initial temperature of the thermostated reactor. Sampling
earlier [22,25]. and analysis for the concentration of the remaining in solution
The temperature– time and concentration–time profiles were sodium sulfate during the time corresponding to the temperature
used for the calculation of the crystal growth rates of mirabilite at changes, showed that this time period corresponded to the de-
different supersaturations (RT and RC, respectively). The reported supersaturation of the solution, which returned to a point close
values are the average of experiments done in triplicate. The but not exactly to equilibrium. The temperature and concentra-
supersaturations investigated in the experiments of the present tion changes of the sodium sulfate concentration in solution for a
work with respect to mirabilite ranged between 0.24–0.66 at typical experiment are shown in Fig. 3.
18 1C and 0.18–0.54 at 20 1C. The temperature increase, Dy, in the solution following the
addition of seed crystals in the supersaturated solutions was
attributed to the heat of crystallization. In all cases the tempera-
3. Results and discussion ture rise was proportional to the solution supersaturation, as may
be seen in Fig. 4.
The thermodynamic driving force for the crystallization of The experimental conditions (i.e. temperature and concentra-
mirabilite is the solution supersaturation, S. Assuming that at tions) were such that heat transfer effects in the process of crystal
relatively small deviations from equilibrium (as in the present growth may be neglected [26]. The induction times preceding the
work) activity coefficients in the supersaturated solutions and at
equilibrium are the same, S may be defined as
ms
S¼ ð1Þ 3.0
m1
where ms and mN are the solute molal concentrations in the
2.5
supersaturated solution and at equilibrium, respectively. The
relative supersaturation, s, is
cs c1 2.0
s¼ ¼ S1 ð2Þ
c1
Δθ/°C
4.0
1.8
Thenardite
m Na2SO4/mol Kg-1
1.6
3.0
Solubility / mol Kg-1
Heptahydrate
1.4
2.0 Mirabilite
1.2
1.0 1.0
0.0 0.5 1.0 1.5 2.0
t / min
10 15 20 25 30 35 Fig. 3. Temperature– and concentration–time profiles for the seeded growth of
Temperature /°C sodium sulfate decahydrate from supersaturated solutions; 20 1C, s ¼ 0.38.
(a) Temperature–time profile following the introduction of mirabilite seed
Fig. 2. Solubility isotherms for sodium sulfate. Solubility data for mirabilite and crystals. Dashed line indicates the slope of the rising part, which was used for
thenardite from Ref. [23] and for heptahydrate from Ref. [13]; (J) solubility the calculation of the rates of crystal growth; (b) concentration–time profile of the
measurements of the present work; (): crystal growth experiments of mirabilite sodium sulfate decahydrate concentration remaining in the solution for the
at 18 1C (Table 1); () crystal growth experiments of mirabilite at 20 1C (Table 1). experiment corresponding to (a).
192 A.I. Vavouraki, P.G. Koutsoukos / Journal of Crystal Growth 338 (2012) 189–194
4.0 15
3.5
1
3.0 2
3 10
2.5
RT / °C min-1
4
Δθ/°C
2.0 5
6
1.5 7
8 5
1.0
9
10
0.5
0.0
0
0 5 10 15 20 25 30 0 30 60 90
t / min G / ms-1
Fig. 4. Crystal growth of sodium sulfate decahydrate following seeding in super- Fig. 6. Seeded growth of sodium sulfate decahydrate from supersaturated solu-
saturated solutions with mirabilite seed crystals at 20 1C, at different values of tions: correlation between the rate measured from the respective rate of the
relative supersaturation, s; 1:0.54; 2:0.50; 3:0.46; 4:0.42; 5:0.38; 6:0.34; 7:0.30; temperature rise, following the addition of mirabilite crystals, RT, and the linear
8:0.26; 9:0.22 and 10:0.18. growth rate, G, calculated from the decrease of the solute concentration as a
function of time; supersaturated solutions prepared by the dissolution of sodium
sulfate at 32 1C: () 18 1C; (~) 20 1C and (J) 20 1C, solutions prepared mixing
3.0
ammonium sulfate and sodium hydroxide.
2.5
of sodium sulfate decahydrate. For the same temperature, the
linear relationship is valid regardless of the method of prepara-
2.0
tion of the supersaturated solutions.
The initial experimental conditions and the crystal growth
1.5
Δθ/°C
0.0 where 3a/b is a shape factor equal to 1 (cubes and spheres) and
rmirabilite is the density of the mirabilite crystals.
The plots of the linear growth rates as a function of the relative
0.0 0.5 1.0 1.5 2.0 supersaturation, shown in Fig. 7, suggested that there is a
t / min parabolic dependence at the low supersaturation range followed
by linear at higher supersaturations in agreement with the
Fig. 5. Crystal growth of mirabilite from supersaturated solutions prepared from
sodium sulfate solutions, on mirabilite seed crystals at 20 1C, s ¼ 0.38. Plot of the prediction of the Burton–Cabrera–Frank model of spiral growth
change of solution temperature as a function of time; S: addition of seed crystals [29]. This complex dependence of the growth rates on the relative
and t: induction time preceding crystallization. supersaturation for both systems investigated, suggested that the
mechanism is surface diffusion controlled.
onset of crystal growth and the rates of crystal growth were It is interesting to note that large differences in the measured
measured from the temperature–time profile variation for each rates were observed at the different temperatures measured,
experiment as shown in Fig. 5. apart only by 2 1C. These differences implied large values of the
Induction times preceding crystal growth in seeded growth of corresponding apparent activation energy. The small temperature
sparingly soluble salts has been reported in the literature and has difference investigated was justified by the fact that the solubility
been used as a tool for the investigation of the mechanism of of mirabilite is strongly affected by temperature. The rate deter-
crystal growth, in combination with results from experiments in mining step in the crystal growth of mirabilite from supersatu-
which the precipitation was spontaneous [27]. The crystal growth rated solutions is surface diffusion, suggesting the possibility to
rates were doubled upon doubling the amount of the inoculating control kinetics by changes of the surface of mirabilite, which
seed crystals, suggesting that it was the seed crystals, which grew may be done e.g. by the adsorption of compounds capable of
and that secondary nucleation was absent. electrostatic or other type of interactions at the solid/water
Measurements of the rates of crystal growth done measuring interface [30]. The strong temperature dependence (high appar-
both the rate of temperature rise, RT, and the rate of concentration ent activation energy) corroborated further the mechanistic con-
decrease of sodium sulfate in the supersaturated solutions, RC (or clusions drawn from the kinetics measurements. Finally, it is also
the corresponding linear growth rate defined in Eq. (3)) showed a interesting to note that the source of the crystallizing mirabilite
linear relationship, as may be seen in Fig. 6. This linear relation- lattice ions did not seem to play an important role at low
ship, entirely empirical, suggested that it is possible, at least for supersaturations, while at higher values the rates measured were
the experimental conditions of the present work (temperature lower exceeding the experimental error710%.
and supersaturation range) to calculate mass crystal growth rates The magnitude of the crystal growth rates of mirabilite
from the rate of temperature rise as a result of the crystallization measured in our experiments were comparable with the rates
A.I. Vavouraki, P.G. Koutsoukos / Journal of Crystal Growth 338 (2012) 189–194 193
Table 1
Seeded crystal growth experiments of mirabilite (250 mg mirabilite seed crystals),
8.0x10-5
from supersaturated solutions prepared by dissolution of thenardite at 32 1C at
different relative supersaturation values, s, induction times, t, maximum change
in temperature due to crystallization, Dy, rate of crystal growth measured from
1600
Table 2
Seeded crystal growth experiments of mirabilite (250 mg mirabilite seed crystals),
Intensity / a.u.
of crystal growth of sodium sulfate decahydrate measured in where FWMirabilite and rMirabilite are the formula weight and the
cooling crystallization from supersaturated solutions [2,31] and in density of the solid (322.20 and 1.49 g cm 3, respectively), NA
MSMPR crystallizer [1]. The dependence of the induction time on Avogadro’s number and n the number of ions in the formula of
the solution supersaturation showed a typical profile as expected mirabilite [33]. The values calculated from the linear fit of the
from the classical nucleation theory [28,32] according to which experimental data according to Eq. (4) yielded values for the
the relationship between the logarithm of the induction time and surface energy of 3.9 and 2.1 mJ m 2 for 18 and 20 1C, respec-
the inverse of the square of the logarithm of the relative super- tively. The low values were anticipated since the nucleation is
saturation is linear: mainly heterogeneous on the inoculating seed crystals. The trend
however of decreasing surface energy with increasing tempera-
16pu2m g3s 1 ture is both expected and rather large for a temperature variation
log t ¼ A þ ð4Þ
3ðkTÞ3 ðlog SÞ2 of only 2 1C.
The crystalline solid was identified as mirabilite by powder X
In Eq. (4), A is a constant, 16p/3 is a shape factor corresponding ray diffraction (XRD) as may be seen in the spectrum shown in
to spherical shape of the new nuclei, um the molecular volume of Fig. 8, where the reflections corresponding to d ¼5.49, 3.83, 3.48
mirabilite, gs is the surface tension of the new nuclei developing and 3.11 Å typical of synthetic mirabilite (ICDD 11-647). The XRD
on the surface, k Boltzmann’s constant and T the absolute patterns obtained were in agreement with the results of the
194 A.I. Vavouraki, P.G. Koutsoukos / Journal of Crystal Growth 338 (2012) 189–194
Acknowledgments
References