ARTICLE IN PRESS

WAT E R R E S E A R C H 40 (2006) 1941 – 1950

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/watres

Effect of a magnetic water treatment on homogeneous and

heterogeneous precipitation of calcium carbonate

Alimi Fathia,b, Tlili Mohameda, Gabrielli Claudeb, Georges Maurinb,, Ben Amor Mohameda
a
Laboratoire de Géochimie et Physicochimie de l’Eau, Institut Nationale de la Recherche Scientifique et Technique, BP 95,
2050 Hammam-Lif, Tunisia
b
Laboratoire Interfaces et Système Électrochimie (LISE), UPR15 du CNRS, Université P. et M. Curie, Tour 22, 5e étage,
4 Place Jussieu, 75252 Paris Cedex 05, France

art i cle info A B S T R A C T

Article history: In this paper are reported experimental results on the effect of a magnetic field on the
Received 5 October 2005 precipitation process of calcium carbonate scale from a hard water. Carbonically pure
Received in revised form water was circulated at a constant flow rate in a magnetic field. After this treatment,
6 March 2006 calcium carbonate precipitation was induced by degassing dissolved carbonic gas. The
Accepted 7 March 2006 nucleation time was identified from the variations of the pH and the Ca2+ concentration.
The ratio between homogeneous and heterogeneous nucleation was determined from the
Keywords: measurement of the mass of precipitated calcium carbonate.
Water treatment It is shown that the magnetic treatment increases the total amount of precipitate. This
Magnetic field effect depends on the solution pH, the flow rate and the duration of the treatment. In
Calcium carbonate precipitation addition, the magnetic treatment modifies the ratio between homogeneous/heterogeneous
Degasification nucleation. Homogeneous nucleation is promoted by an increasing the pH of water, the
Homogeneous flow rate as well as the residence time. The magnetic treatment enhances these effects
Heterogeneous nucleation with a maximum for a 15 min treatment time. It is shown that the presence of calcium
carbonate colloid particles is not necessary. It is advanced that the main magnetic effects
concern the associations of ionic species which are present in the solution and which are
involved in the nucleation process of calcium carbonate precipitation
& 2006 Elsevier Ltd. All rights reserved.

1. Introduction device was patented in Belgium in 1945 (Vemeiren, 1958).

Powerful electromagnets were used in hot water systems
The antiscale magnetic treatment (AMT) of hard waters is since the 1960s in the Soviet Union (Grutsch, 1977). The
currently used to prevent the build up of deposits on the walls application of AMT was reported in the United States since
of industrial systems, in particular in heat exchangers, as well 1975 (Grutsch, 1977; Grutsch and McClintock, 1984). In many
in domestic equipments (Proceedings of the International cases, the field is delivered by permanent magnets in various
meeting of Antiscale magnetic treatment, 1996; MAG 3, 1999). geometrical configurations. Several devices are based on AC
This kind of physical treatment presents the great advantage or pulsed fields. (Oshitani et al., 1999). According to the review
to avoid the use of chemicals such as strong acids or paper of Baker and Judd (1996), in spite of this long
polyphosphates which are expensive and can be harmful for experience, the efficiency of this treatment is still a con-
human life or deleterious for the environment. AMT has been troversial question and it is not possible to get a clear
employed for more than a half century. The first commercial explanation of the phenomenon.

Corresponding author. Fax: +33 1 44 27 40 74.

E-mail address: [email protected] (G. Maurin).
0043-1354/$ - see front matter & 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.watres.2006.03.013
 ARTICLE IN PRESS
1942 WAT E R R E S E A R C H 40 (2006) 1941– 1950

A great number of experimental researches were carried out In a previous work (Gabrielli et al., 2001), the scaling power of a
on the modification of the CaCO3 precipitation process by AMT. magnetically treated water was evaluated by mean of an
In most cases, a supersaturated solution was prepared by the electrochemical method. The crystal growth rate was directly
‘‘double decomposition method’’, by mixing two equimolar measured with an electrochemical quartz crystal microbalance.
solutions of CaCl2 and Na2CO3, respectively, but this method It was shown that the field strength, the composition and the
presents the drawback to introduce systematically an excess flow velocity of the water have significant effects on the scaling
of foreign Na2+ and Cl
ions which increase the conductivity. rate. The nature of the material of the tubing, in which the MF
Various electrochemical and non-electrochemical techniques was applied, appeared also to have an influence on AMT. It was
were used to estimate the scaling potentiality of water (Hui deduced that electrokinetic phenomena occurring in the vicinity
and Lédion, 2002). In the electrochemical methods, the strong of the walls of the tubing could be also implied (Gabrielli et al.,
increase of the interfacial pH on the cathode plate due to the 2001). However, it could be objected that the electrochemical
reduction of dissolved oxygen induced the heterogeneous precipitation test induces exclusively an heterogeneous nuclea-
nucleation of calcium carbonate. The scaling rate can be tion by a strong interfacial pH shift, this could be in contra-
evaluated from the decrease of the electrochemical current diction with what happens in usual conditions where the
intensity (Lédion et al., 1985), by measuring the deposited homogeneous and heterogeneous precipitation are mainly due
mass thanks to a quartz crystal microbalance (Gabrielli et al., to the escape of carbonic gas from the water, under the effect
1996b), or even by performing electrochemical impedance either of a change of pressure or an increase of the temperature.
analysis (Gabrielli et al., 1996a; Deslouis et al., 1997). Among For this reason, the experimental works which are presented
the non-electrochemical methods, some of them consist in in this paper were performed with the same magnetic treatment
forcing the calcium carbonate precipitation by increasing the system which proved its efficiency, but the scaling potentiality of
pH, some other are based on the heating or the evaporation of the treated water was evaluated by inducing the calcium
the solution (Euvrard et al., 1997). carbonate precipitation by a method based on the extraction of
According to the literature, the efficiency of the magnetic dissolved carbonic gas. From the measurement of the variations
treatment depends on numerous parameters. For example, of the solution pH, the Ca2+ concentration, and the mass of the
Chibowski et al. (2003) or Barrett and Parsons (1998) have homogeneous precipitate, it was possible to evaluate the effect
observed that a magnetic treatment applied during on a hard of the magnetic treatment on the nucleation time and on both
water decreased the quantity of scale deposited on the wall. the homogeneous and heterogeneous precipitation rates. The
The principle of the phenomenon is still not well understood morphology of the two kinds of precipitate was examined by
and various contradictory hypotheses were proposed. It was SEM. Their crystal structure was identified by X-ray diffraction.
often attributed to the Lorentz forces ~ F ¼ q ~v~ B exerted
either on moving ions or on charged solid particles (Higa-
shitani et al., 1993). The magnetic field (MF) would be able to 2. Experimental
disturb the double ionic layer surrounding the colloidal
particles and their zeta potential (Gamayunov, 1983; Higa- 2.1. Water preparation
shitani and Oshitani, 1998; Parsons et al., 1997). According to
several authors (Higashitani et al., 1993; Barrett and Parsons, In order to avoid any side effect by foreign ions, calcocarbo-
1998), the MF would tend to reduce the nucleation rate and to nically pure water (CCP water), (water containing only Ca2+,
accelerate the crystal growth. Ferreux assumed that the free CO2


3 and HCO3 ions) was used. It was prepared by dissolving
enthalpy of calcium carbonate formation would be modified
1
0.3 to 0.5 g dm reagent grade CaCO3 in deionized water, by
by the MF (Ferreux, 1992). The scale modification could also bubbling carbon dioxide during a full day according to
result from the preferential formation of the aragonite
CO2 þ H2 O þ CaCO3 ! Ca2þ þ 2HCO3
: (1)
polymorph (Coey and Cass, 2000; Knez and Pohar, 2005; Kobe
et al., 2001), which is usually formed above 60 1C, instead of This dissolved CaCO3 quantity is equivalent to a water
calcite. Aragonite, which may result from the transformation hardness of 30–50 French Degrees (1 1F corresponds to
of metastable vaterite nuclei (Gabrielli et al., 1999), exhibits a 10 mg dm
3 of dissolved CaCO3). This hardness can be
characteristic needle shape morphology with a rather weak considered as moderate by comparison with the values used
adhesion to the substrate. Therefore, they could be carried by many authors (Chibowski et al., 2003; Higashitani et al.,
away by the liquid flow (Kronenberg, 1985). On the contrary, 1993; Higashitani and Oshitani, 1998; Knez and Pohar, 2005).
calcite which is the more stable calcium carbonate poly- The water pH resulting from this preparation was about 5.7.
morph at room temperature (Plumber and Busenberg, 1982), Then, the calcocarbonic equilibrium was displaced towards
forms dense and tenacious layers, which are difficult to supersaturation by adjusting the pH between 6 and 7.5 thanks
remove mechanically. It was advanced that the magnetic to a strong stirring to exhaust a part of the dissolved CO2 in
effect concerns ferromagnetic impurities which are nuclea- the atmosphere. The supersaturation O depends on the (Ca2+)
tion seeds (Busch and Busch, 1997). Lundager Madsen has and (CO2
3 ) activity according to
concluded that the field accelerates the crystallization of
sparingly soluble diamagnetic salts of weak acids such as ðCa2þ ÞðCO2

3 Þ
O¼ , (2)
Ks
carbonates and phosphates. He suggested that MF is able to
change the orientation of the proton spin and to disturb Where Ks is the equilibrium solubility product for a given
dehydration phenomena by hindering the transfer of the calcium carbonate polymorph, more generally for calcite,
proton to a water molecule (Lungader Madsen, 1995, 2004). which is the most stable form.
 ARTICLE IN PRESS
WAT E R R E S E A R C H 4 0 (200 6) 194 1 – 195 0 1943

Table 1 – Supersaturation coefficient O versus water dilute NaOH solution (Feitler, 1972). As soon as the critical
hardness and pH supersaturation was reached, the first calcium carbonate
nuclei were formed and the pH began to decrease because of
Hardness (1F) pH the liberation of H+ ions according to:

5.7 6.0 7.0 7.5 Ca2þ þ HCO3
! CaCO3 þ Hþ : (4)

The induction time tind of CaCO3 nucleation which is

30 0.05 0.18 1.32 6.1
40 0.15 0.32 3.2 10.11 needed for the formation of the first nuclei, corresponds to
50 0.23 0.47 4.72 14.93 the change of the slope of the curve pH ¼ f(t). In the present
work, similar to the LCGE method (Laboratoire Chimie
et Genie de l’Environnement : Dedieu et al., 1994; Elfil and
Roques, 2001), the pH variations are induced by exhausting
CO2 from the water according to the reverse of Eq. (1). This
On Table 1, are reported the values of O relatively to calcite for method, which does not need any addition of chemicals,
various hardness of a CCP water at different pH. These data gives a good simulation of most natural precipitation
were calculated on the basis of the Legrand and Poirier model processes where scaling phenomena are due to CO2 depar-
of the calcocarbonic system (Legrand et al., 1981). For pH ¼ 5.7 ture.
and 6, o was less than 1, which means that the water was In the precipitation test set-up (Fig. 2), 0.5 dm3 of treated
undersaturated and aggressive with respect to solid calcium water was placed in a cell made of a dense polyamide
carbonate. For pH ¼ 7 and 7.5, the solution was moderately material and immersed in a thermostatic water bath to
supersaturated and, theoretically, CaCO3 nucleation could maintain the temperature at 30.0 1C. It was shown that this
spontaneously occur. But it is a matter of fact that it is polyamide material exhibits a good affinity for the CaCO3
necessary to reach a larger critical supersaturation Oc (generally heterogeneous precipitation (Ben Amor et al., 2004). CO2
OcE40 for homogeneous nucleation and OcE20 for hetero- degasification was provoked by a continuous flow of pure
geneous nucleation) to get calcium carbonate nucleation and nitrogen (8 dm3 min
1) through a diffuser located at the
growth. So, in the present conditions, these waters did not bottom of the cell. The solution pH and the conductivity
contain any calcium carbonate seeds, contrarily to the cases were continuously monitored to detect the nucleation time.
presented in the literature where the hardness was very high. In a first approximation, the conductivity is proportional to
the water hardness. Minute water samples (1 cm3) were
2.2. Water treatment drawn out every minute and their calcium ion concentration
was analysed by EDTA complexometry. (accuracy E0.10
3 M).
A volume V ¼ 0.50 dm3 of CCP water was circulated in close For example, Fig. 3 presents the typical variations of the pH
loop from a thermostated closed glass tank through a plastic and the calcium concentration versus time. The nucleation
tubing (TygonTM R3606, section area s ¼ 0.38 cm2, total length time tind evaluated from the two curves are in good agree-
150 cm) thanks to a volumetric gear pump (Fig. 1a). The flow ment. In the present example, the critical supersaturation
rate was moderate in order to have a laminar regime. A part of was reached at pHE8.5 and the induction time was equal to
the tubing (l ¼ 20 cm) was inserted between the polar pieces 10.570.5 min. The variations of the conductivity gave similar
of the magnetic device. This length can be doubled by passing results. At the end of the treatment, the full volume was
twice the pipe between the polar pieces. For the sake of filtered with a 0.45 mm pore size cellulose nitrate membrane to
comparison, experiments were also carried out without separate and to weight the calcium carbonate mass mb
passing in the MF. The magnetic device was described in formed in the bulk of the solution by homogeneous nuclea-
details in the paper of Gabrielli et al. (2001): a series of 5 pairs tion. The mass of calcium carbonate mw deposited on the
of permanent magnets with north and south faces facing walls of the cell by heterogeneous nucleation is deduced from
each other are associated alternately (Fig. 1b). the values of mb and [Ca2+]. The morphology of the calcium
The magnetic circuit of each pair was closed with a U carbonate grains formed in the solution or on the wall was
shaped iron yoke. The field strength was about 0.16 T in the air studied by SEM and their crystal structure was determined by
gap. The total time of treatment tT was between 5 and 30 min. X-ray diffraction.
In this configuration, the mean time of exposure of water to
the MF texp for n passes, is independent of the flow rate j:
3. Results
sl
texp ¼ n tT (3)
V
3.1. Effect of the magnetic treatment on the induction time

In preliminary experiments, it was verified that the circula-

2.3. Precipitation test tion of the CCP water in the plastic tubing has no significant
effect on the pH and on the Ca2+ concentration, whatever the
In order to evaluate the scaling potentiality of the treated solution pH and the flow rate. This result is valid in the
water, the precipitation of the calcium carbonate was induced absence as well in the presence of MF on a part of the plastic
by following a method adapted from the so called ‘‘critical pH tubing. The pH increase observed between the beginning and
method’’ where the pH was slowly increased by addition of a the end of the magnetic treatment did not exceed 0.1 units for
 ARTICLE IN PRESS
1944 WAT E R R E S E A R C H 40 (2006) 1941– 1950

Fig. 1 – Sketch of the set-up for the magnetic treatment of circulating water. (a) set up and (b) permanent magnets disposition.

conductimeter pH meter
Flow
control

EDTA

N2

Ca++ analysis
Thermostatic bath ( 30˚C)

Fig. 2 – Scheme of the set-up for the precipitation test by CO2 degassing.

a 15 min processing time and it was about 0.2 for 30 min. It the calcium carbonate nucleation rate. The effect of pH is
means that the treatment does not induce any significantCO2 easily understandable: when the water becomes more alka-
departure due to water agitation, and that no CaCO3 particles line, the calcocarbonic equilibrium is displaced towards a
can be formed during the treatment time. stronger supersaturation and the nucleation probability
In Fig. 4 are reported a set of experimental induction times becomes larger. However, the influence of the flow rate in
measured during the precipitation process, as evaluated by the absence of MF is somewhat surprising. It is worth noting
the variations of the pH and the calcium concentration. In all that the simple circulation of the water in the tubing has a
cases, the water hardness was 40 1F, the temperature 30 1C effect on the induction time of CaCO3 nucleation measured
and the total treatment time was 15 min, in the presence or thereafter during the precipitation test. For example, for an
absence of MF. The flow rate was varied from 0 to initial pH of 6, tind decreased from 22 to 12 min when the flow
0.94 dm3 min
1 and the solution pH from 6 to 7.5. This figure rate varied from 0 to 0.94 dm3 min
1. This effect was less
demonstrates clearly that these three parameters: flow evident when the solution pH was higher. In our knowledge
velocity, initial pH, and application of the MF tend to reduce this effect had never been mentioned until now. When the MF
the induction time. It is equivalent to say that they increase was applied along 20 cm of the plastic pipe, the induction
 ARTICLE IN PRESS
WAT E R R E S E A R C H 4 0 (200 6) 194 1 – 195 0 1945

9 0.006

pH
8.5 0.005

8 0.004

[Ca2+] / M
pH

7.5 0.003
Ca2+

7 0.002
tind

6.5 0.001
0 5 10 15 20 25 30 35 40 45 50
Time /min

Fig. 3 – Variations of the solution pH and the calcium ion concentration during a precipitation test (H ¼ 40 1F, initial pH ¼ 7).

25 Table 2 – Effect of the treatment-time on the induction

With magnetic field time at various solution pH and various flow rates
Without magnetic field
20 pH = 6
Flow rate Time of Induction time
(dm3 min
1) treatment (min)
pH ¼ 6 pH ¼ 7
15
tind / min

pH = 7 0.54 5 17 15
15 14 10
10 30 12 8

pH = 7.5 0.74 5 14 9
5 15 10 9
30 13 9

0.94 5 12 9
0 15 7 7
0 0.2 0.4 0.6 0.8 1
30 13 8
Flow rate / dm3. min-1

Fig. 4 – Variations of the induction time tind versus the flow

rate for various initial values of the initial pH, in the
presence or absence of magnetic field.
carbonate, expressed in percentage of the initial CaCO3
quantity dissolved in the water. This quantity was deduced
from the measurement of Ca2+ concentration after the
time was even more reduced and this effect increased with treatment. These measurements were carried out on 40 1F
the flow velocity. water hardness, in the presence or not of the MF, for various
In Table 2, are illustrated the influence of the processing flow rates, at pH 6–7.5. The treatment time was 15 min and
time on the induction time in the presence of MF, for 2 pH the CO2 degassing time was 90 min. In the absence of MF, the
values and for various flow rates in the case of a 40 1F water total amount of precipitated CaCO3 did not depend signifi-
hardness. One can observe that in all cases, there was an cantly from the flow rate, whatever the pH. It means that the
optimum treatment time of about 15 min, corresponding to total amount of precipitate was the same although the
the smaller induction time—or the higher nucleation rate. For nucleation induction time was decreased by the water
a longer treatment time, paradoxically the effect of the MF circulation in the tubing. In the presence of the MF, the total
was less pronounced. amount of precipitate was significantly increased by ca.
7–22% for j ¼ 0.54 dm3 min
1, and this effect was increased
3.2. Effect on homogeneous and heterogeneous nucleation even more with a higher flow rate (26% at pH ¼ 7.5 and
j ¼ 0.94 dm3 min
1).
In Fig. 5 are reported experimental measurements of the The amount of calcium carbonate precipitated in the bulk
percentage of the total quantity of precipitated calcium of the solution by homogeneous nucleation during the
 ARTICLE IN PRESS
1946 WAT E R R E S E A R C H 40 (2006) 1941– 1950

95 60
With Magnetic Field
pH = 7.5 Without Magnetic Field
90
Total precipitation ratio / %

With 50

Homogeneous precipitation ratio / %

magnetic treatment
85 pH = 7
40
pH = 6 pH=7.5
80
30
Without magnetic treatment
75

20
70 pH=6
0 0.2 0.4 0.6 0.8 1
Flow rate /dm3. min-1
10
Fig. 5 – Variations of the total precipitation ratio in % of the
amount of dissolved CaCO3, versus the flow rate, for treated
waters at various pH (full line : in the presence of MT, dotted 0
0 0.2 0.4 0.6 0.8 1
lines in the absence of MT, H ¼ 40 1F).
Flow rate / dm3.min-1

Fig. 6 – Variations of the homogeneous precipitation ratio in

precipitation test was determined after filtration of the % of the amount of dissolved CaCO3 versus the flow rate for
solution. Fig. 6 shows that the percentage of precipitated treated waters at two pH (full line: in the presence of MT,
calcium carbonate by homogeneous nucleation expressed in dotted lines in the absence of MT, H ¼ 40 1F).
% of the initial dissolved quantity increased with the pH and
with the flow rate. When MF was applied, the amount of 70
homogeneous precipitate was even larger. For example, in the 50˚FWMF
case of a 40 1F and pH ¼ 7.0 water, at j ¼ 0.74 dm3 mn
1, in the 60
40˚FWMF
30˚FWMF
absence of MF, the total precipitation ratio was 73% divided in 50˚F W/oMF
Homogeneous Precipitation Ratio / %

22.2% for the homogeneous precipitation ratio and 50.8% for 40˚F W/oMF
the heterogeneous ratio. In the presence of MF, the total 50 30˚F W/oMF

precipitation ratio was increased to 84%. The homogeneous

contribution was increased to 38.9% whereas the heteroge- 40
neous one was decreased to 45.1%. These experimental data
give evidence that the MF promotes preferentially the
30
homogeneous precipitation detrimentally to the scaling of
the walls. This scaling inhibition confirms previous results on
heterogeneous precipitation obtained by electrochemical 20
tests (Gabrielli et al., 2001).
As illustrated by Fig. 7, the homogeneous precipitation ratio 10
is increased by the water hardness and also by the flow rate,
especially at low hardness. The specific effect of the MF is
0
relatively more important for low hardness water.
0.5 0.6 0.7 0.8 0.9 1
The influence of the treatment time on the precipitation in
Flow rate / dm3.min-1
magnetic treated water was also investigated. For a 40 1F,
pH ¼ 7.0 water, Fig. 8a and b present the variations of the total Fig. 7 – Variations of the homogeneous precipitation ratio in
and the homogeneous precipitation ratio, versus the treat- % of the amount of dissolved CaCO3 versus the flow rate for
ment time, for various flow rates. They give evidence that treated water of various hardness (full line: in the presence
there is an optimal treatment time (about 15 min), which of MT, dotted lines in the absence of MT, pH ¼ 7).
corresponds to a larger total amount of precipitated calcium
carbonate and a larger contribution of homogeneous pre-
cipitate, whatever the flow rate.
In all previous experiments, the treatment time was homogeneous precipitation ratio from ca. 41% to ca. 52%,
maintained during 15 min with a single pass between the whereas, in the absence of MF, this ratio was only ca. 25%.
pole pieces. That corresponds to a mean exposure time This result shows that, as expected, a longer exposure time
texp ¼ 0.23 min, according to Eq. (1). In order to increase to MF increased the total amount of precipitate but also
the exposure time without changing the treatment time, a it modified the property of the treated water, in such a
double pass treatment was applied. In Fig. 9, it is shown way that the homogeneous precipitation was preferentially
that a double pass of a 40 1F pH ¼ 7.0 water increased the increased.
 ARTICLE IN PRESS
WAT E R R E S E A R C H 4 0 (200 6) 194 1 – 195 0 1947

85 70

Homogeneous Precipitation Ratio / %

O.94 dm3.min-1 60
Total Precipitation Ratio / %.

80 Double pass

50
Single pass
75
40
O.74 dm3.min-1
70 30 Without magnetic treatment

O.54 dm3.min-1 20
65

10
60
0 5 10 15 20 25 30 0
(a) Time of treatment / min 0.5 0.6 0.7 0.8 0.9 1
Flow rate / dm3.min-1
50
0.94 dm3.min- Fig. 9 – Variations of the homogeneous precipitation ratio in
Homogeneous Precipitation Ratio / %

45 1
% of the amount of dissolved CaCO3 versus the flow rate for
40 a treated water either in the absence of magnetic treatment
0.74 dm3.min- or after passing one time or two time in the magnetic device
35 1
(pH ¼ 7, H ¼ 40 1F).
30 0.54 dm3.min-
25 1

20 modified some species which were present in the calcareous

15 water during the treatment, and which were implied later in
the nucleation process during the precipitation test.
10
5
0 4. Discussion
0 5 10 15 20 25 30
(b) Time of treatment / min The main findings of this experimental investigation can be
summarized as follows:
Fig. 8 – Precipitation ratio in % of the amount of dissolved
CaCO3 versus the treatment time, for various flow rate
 In the absence of MF, the circulation of water in the tubing
(pH ¼ 7, H ¼ 40 1F). (a) total precipitation ratio and (b)
has a positive effect on the calcium carbonate nucleation
homogeneous precipitation ratio.
which is triggered later during the precipitation test.
 MF decreases the nucleation induction time and increases
the nucleation rate.
3.3. Calcium carbonate morphology  MF increases the precipitation potentiality of the treated
water. This effect is more pronounced for homogeneous
Calcium carbonate crystals formed in the solution bulk and nucleation than for heterogeneous nucleation.
separated by filtration, were systematically observed by SEM.  The efficiency of the treatment increases with the water
In the absence of any treatment, the crystals exhibit the flow rate in the tubing and with the time of exposure to
characteristic morphology of the vaterite form (Fig. 10a). The the MF.
identification of the crystal structure was confirmed by X-ray  The efficiency of the treatment increases with the hard-
diffraction and also by micro Raman spectroscopy, on the ness and with the pH of the water.
basis of previous studies (Gabrielli et al., 2000, 2003).  MF treatment is efficient even for a water having a
After MF treatment; as seen in Fig. 10a and b, the formation supersaturation coefficient lower than 1 (undersaturation)
of the calcite form is promoted. At pH ¼ 6, a part of the or having a supersaturation sufficiently small to be sure
precipitate is made of vaterite crystals partially transformed that no calcium carbonate colloidal particle be present.
into aragonite needles, an other part is made of calcite  MF treatment promotes the homogeneous precipitation of
crystals with their characteristic near cubic shape. The mean aragonite and calcite crystals instead of vaterite.
crystal size of precipitation test was about 10 mm after 90 min
of precipitation test. For a more alkaline treated water Numerous mechanisms have been proposed in the litera-
(Fig. 10c), only calcite crystals were observed, and their size ture to explain the MF effects on the precipitation of calcium
was reduced to about 5 mm. Knowing that the total amount of carbonate from hard waters. According to Knez and Pohar
homogeneous precipitate was increased and the nucleation (2005) they are belonging to two different approaches: either
time was decreased it means that the number of nuclei was magnetohydrodynamic (MHD) phenomena or hydration ef-
significantly increased. That demonstrates that the MF had fects.
 ARTICLE IN PRESS
1948 WAT E R R E S E A R C H 40 (2006) 1941– 1950

In the present case, similar effects in relation with hydro-

dynamic forces could be attributed to the circulation of the
water. It was observed that MF promotes aggregation of latex
particles only when the suspension is circulated Busch et al.
(1996). According to these authors, MHD phenomena induce
eddy currents which flattened the fluid velocity profile in the
tube. This effect would result in a larger velocity gradient,
along the walls. In addition, the streaming potential along the
walls which is velocity dependent should increase with the
MF. This phenomenon, by changing the surface charge, could
throw out of balance the calcocarbonic equilibrium in the
vicinity of the tubing walls (Knez and Pohar, 2005). MHD
phenomena could also concern the electrical double layer
near the charged surface of particles. This interpretation
seems receivable for many experimental results of the
literature where highly supersaturated waters were treated
by MF (Barrett and Parsons, 1998). The aggregation of the
CaCO3 colloidal particles under the influence of electrostatic
phenomena would contribute to accelerate the crystal growth
and the precipitation process.
However, in the present experimental conditions (CCP
water, moderate pH and low hardness), we can assume that
the water did not contain any calcium carbonate particles in
suspension. Nevertheless, since 1941, it is known that hard
waters contain various ionic associations such as ionic pairs
CaCO31 (Greenwald, 1941; Plumber and Busenberg, 1982).
According to Gal et al. (1996, 2002); calcium carbonate
crystallization is a very complex process involving several
steps: at first hydrated Ca2+ and CO2
3 ions are associated in
CaCO31 shape. The formation of loosely attached (Ca2+—CO2
3 )
ionic pairs weakens the water shells and hydrated micelle of
disordered pairs are formed, whose complexity increases
with the supersaturation. Their progressive dehydration leads
to CaCO3 solid forms. In agreement with several authors (Gal
et al., 1996; Elfil and Roques, 2001; Tlili et al., 2001, 2003) the
solid forms which are formed at first are made of hydrated
amorphous or hexagonal calcium carbonate crystal forms, so
these unstable hydrated forms can be considered as the
precursors of the three usual anhydrous polymorphs—na-
mely calcite vaterite or aragonite. We can assume that the
hydrodynamic forces as well as the electrokinetic phenomena
along the walls could favour the process of ionic pair
micellization. This hypothesis is supported by the fact that,
after a simple water recirculation in the absence of MF, the
Fig. 10 – SEM images of calcium carbonate crystals obtained induction time for calcium carbonate nucleation was sig-
by homogeneous precipitation and filtration: (a) non treated nificantly reduced during the precipitation test. So it can be
water , pH ¼ 6, H ¼ 40 1F, vaterite crystals. (b) magnetic supposed that the micellization of ionic pairs was more
treated water, pH ¼ 6, H ¼ 40 1F, flow rate ¼ 0.54 dm3 min
1: advanced, thus facilitating the effective nucleation process
calcite and vaterite transformed into aragonite and (c) during the following precipitation test. In the presence of MF,
magnetic treated water, pH ¼ 7.5, H ¼ 40 1F , flow these effects could be expected to be enhanced.
rate ¼ 0.54 dm3 min
1: pure calcite. MF effects were also often explained in terms of magneti-
cally induced changes in the hydration of ions or solid
surfaces. According to Lungader Madsen (1995, 2004) MF
MHD phenomena depend on the flow of the treated water. induces faster proton transfer from hydrogen carbonate to
The beneficial influence of the flow rate on the magnetic water (reaction 3), due to proton spin inversion in the field of
effect observed in the present results is in agreement with diamagnetic salts. The increased formation of CO2
3 ions
this approach. Busch et al. (1996) have assumed that the would explain the beneficial effect of MF on the amount of
Lorenz forces exerted on charged species induce local precipitate. Another explanation proposed by Higashitani
convection movements in the liquid which could contribute et al. (1993) is related to the specific influence of MF on the
to accelerate associations between ions or colloidal particles. hydration of CO2
3 ions which could directly modify the
 ARTICLE IN PRESS
WAT E R R E S E A R C H 4 0 (200 6) 194 1 – 195 0 1949

polymorph phase equilibrium during the precipitation. The Gabrielli, C., Jaouhari, R., Maurin, G., Keddam, M., 2001. Magnetic
same phenomenon could also affect the dehydration process water treatment for scale prevention. Water Res. 35 (13),
of the ionic pair associations and the hydrated calcium 3249–3259.
Gabrielli, C., Jaouhari, R., Joiret, S., Maurin, G., 2000. In situ Raman
carbonate forms which act as precursors for the formation
spectroscopy applied to electrochemical scaling. Determina-
of the anhydrous forms which constitute the calcium
tion of the structure of vatérite. J. Raman spectrosc. 31 (6),
carbonate precipitate. 497–501.
Gabrielli, C., Jaouhari, R., Joiret, S., Maurin, G., Rousseau, P., 2003.
Study of the electrochemical deposition of CaCO3 by in situ
5. Conclusion Raman Spectroscopy. I Influence of the substrate. J. Electro-
chem. Soc. 150 (7), C478–C484.
The present results show that the circulation of water in a Gabrielli, C., Keddam, M., Maurin, G., Perrot, H., Rosset, R.,
Zidoune, M., 1996a. Estimation of the deposition rate of
stationary MF reduces the nucleation induction time and
thermal calcareous scaling by the electrochemical impedance
increases the total amount of calcium carbonate obtained
technique. J. Electroanal. Chem. 412 (1–2), 189–193.
thereafter during the precipitation test by CO2 degassing. This Gabrielli, C., Keddam, M., Perrot, H., Khalil, A., Rosset, R., Zidoune,
effect is more pronounced for homogeneous nucleation. It M., 1996b. Characterization of the efficiency of antiscale
does not need the presence of CaCO3 particles in the treated treatments of water Part I: Chemical processes. J. Appl.
water. It can be assumed that the ionic associations which are Electrochem. 26 (11), 1125–1132.
present even in unsaturated solutions and which are involved Gabrielli, C., Maurin, G., Poindessous, G., Rosset, R., 1999.
in a complex process leading to the final formation of the Nucleation and growth of calcium carbonate by an electro-
chemical scaling process. J. Cryst. Growth 200 (1-2), 236–250.
anhydrous calcium carbonate crystals, are affected by MHD
Gal, J.Y., Bollinger, J.C., Tolosa, H., Gache, N., 1996. Calcium
and electrokinetic phenomena especially along the walls or carbonate solubility, a reappraisal of scale formation and
by magnetically induced changes of their hydration shell. inhibition. Talenta 43, 1497–1509.
Gal, J.Y., Fovet, Y., Gache, N., 2002. Mechanisms of scale formation
R E F E R E N C E S and carbon dioxide partial pressure influence. Part I: elabora-
tion of an experimental method and a scaling model. Water
Res. 36, 755.
Gamayunov, N.I., 1983. Coagulation of suspension after magnetic
Baker, J., Judd, S., 1996. Magnetic amelioration of scale formation.
treatment. J. Appl. Chem. USSR 56, 975–982.
Water Res. 30 (2), 247–260.
Greenwald, I., 1941. The dissociation of calcium and magnesium
Barrett, R.A., Parsons, S.A., 1998. The influence of magnetic fields
carbonates and bicarbonates. J. Biol. Chem. 141, 789–794.
on calcium carbonate precipitation. Water Res. 32 (3), 609–612.
Grutsch, J.F., McClintock, J.W., 1984. Corrosion and deposit control
Ben Amor, M., Zgolli, D., Tlili, M.M., Manzola, A.S., 2004. Influence
in alkaline cooling water using magnetic water treatment at
of water hardness, substrate nature and temperature on
Amoco’s largest refinery, ‘‘Corrosion 84’’ NACE, paper 330, New
heterogeneous calcium carbonate nucleation. Desalination
Orleans.
166, 79–84.
Grutsch, J.F., 1977. USA/USSR Symposium of Physical Mechanical
Busch, K.W., Gopalakrishnan, S., Busch, M.A., Tombacz, E., 1996.
Magnetohydrodynamic aggregation of cholesterol and poly- Treatment of Wastewaters. EPA, Cincinnati, 44.
styrene latex suspensions. J. Colloid Inter. Sci. 183, 528–538. Higashitani, K., Oshitani, J., 1998. Magnetic effects on thickness of
Busch, K.W., Busch, M.A., 1997. Laborator studies on magnetic adsorbed layer in aqueous solutions evaluated directly by
water treatment and their relationship to a possible mechan- atomic force microscope. J. Colloid Interf. Sci. 204, 363–368.
ism for scale reduction. Desalination 109 (2), 131–148. Higashitani, K., Kage, A., Katamura, S., Imai, K., Hatade, S., 1993.
Chibowski, E., Hotysz, L., Szczes, A., 2003. Adhesion of in situ Effects of magnetic field on formation of CaCO3 particles. J.
precipitated calcium carbonate in the presence and absence of Colloid Interf. Sci. 156, 90–95.
magnetic field in quiescent conditions on different solid Hui, F., Lédion, J., 2002. Evaluation methods for the scaling power
surfaces. Water Res. 37, 4685–4692. of water. Eur. J. Water Qual. 33 (1), 55–74.
Coey, J.M.D., Cass, S., 2000. Magnetic water treatment. J. Magn. Knez, S., Pohar, C., 2005. The magnetic field influence on the
Magn. Mat. 209, 71–74. polymorph composition of CaCO3 precipitation from carbo-
Dedieu, L., Hort, C., Martin-Dominguez, A., Rola, M., Roques, H., nized aqueous solutions. J. Colloid Interf. Sci. 281, 377–388.
1994. Contribution à l’étude des phénomènes d’entartrage. Kobe, S., Drazic, G., McGuiness, P.J., Strazisar, J., 2001. The
1ière partie : Généralités et méthodes d’étude LCGE, Tribune de influence of the magnetic field on the crystallisation form of
l’Eau 48 (N1 571), pp. 3–19. calcium carbonate and the testing of a magnetic water-
Deslouis, C., Gabrielli, C., Keddam, M., Khalil, A., Rosset, R., treatment device. J. Magn. Magn. Mater. 236 (1-2), 71–76.
Tribollet, B., Zidoune, M., 1997. Impedance techniques at Kronenberg, K.J., 1985. Experimental evidence for effects of
partially blocked electrodes by scale deposition. Electrochim. magnetic fields on moving water. IEEE Trans. Magn. 21 (5),
Acta 42 (8), 1219–1233. 2059–2061.
Elfil, H., Roques, H., 2001. Role of hydrated phases of calcium Lédion, J., Leroy, J.P., Labbé, J.P., 1985. Détermination du caractère
carbonate on the scaling phenomenon. Desalination 137, incrustant d’une eau par un essai d’entartrage accélère. TSM.
177–186. L’eau 80, 323–328.
Euvrard, M., Lédion, J., Leroy, P., 1997. Effect and consequences of Legrand, L., Poirier, G., Leroy, P., 1981. Les équilibres carboniques
electric treatment in preventing scaling in drinking water et l’équilibre calcocarbonique. Eyrolles, Paris.
systems. J. Water SRT-Aqua 46, 71–83. Lungader Madsen, H.E., 1995. Influence of magnetic field on the
Feitler, H., 1972. The scale meter: a new method for determining precipitation of some inorganic salts. J. Cryst. Growth 152,
the critical pH of scaling. Mater. Prot. Perform 11 (1), 31–35. 94–100.
Ferreux, M., 1992. Role d’un traitement magnétique sur la Lungader Madsen, H.E., 2004. Crystallization of calcium carbonate
cristallogenèse du carbonate de calcium dans les eaux in magnetic field ordinary and heavy water. J. Cryst. Growth
entartrantes. Ph. D. Thesis, University of Besanc-on, France. 267, 251–255.
 ARTICLE IN PRESS
1950 WAT E R R E S E A R C H 40 (2006) 1941– 1950

MAG 3, 1999. Proceedings of Anti-Scale Magnetic treatment and Proceedings of the International meeting of Antiscale magnetic
Physical Conditions. School of Water Sciences, Cranfield Uni- treatment, 1996. School of Water Sciences, Cranfield Univer-
versity, UK. sity, UK.
Oshitani, J., Uehara, R., Higashitani, K., 1999. Magnetic effects on Tlili, M.M., Ben Amor, M., Gabrielli, C., Joiret, S., Maurin, G.,
eletrolyte solutions in pulse and alternation field. J. Colloid Rousseau, P., 2001. Characterization of CaCO3 hydrates by
Interf. Sci. 209, 374–379. micro-Raman spectroscopy. J. Raman Spectrosc. 33, 10–16.
Parsons, S., Judd, S.J., Stephenson, T., Udol, S., Wang, B.L., 1997. Tlili, M.M., Ben Amor, M., Gabrielli, C., Joiret, S., Maurin, G., 2003.
Magnetically augmented water treatment. Trans. Inst. Chem. Calcium carbonate precursors during scaling process. Com-
Eng. 75(B), 98–104. munication First IWA Conference on Scaling and Corrosion in
Plumber, L.N., Busenberg, E., 1982. The solubilities of calcite, Water and Wastewater Systems, Cranfield University UK Full
aragonite and vaterite in CO2–H2O solutions between 0 and text in special volume.
90 1C, and an evaluation of the aqueous model for the system Vemeiren, T., 1958. Magnetic treatment of liquids for scale and
CaCO3–CO2–H2O. Geochim. et Cosmochim. Acta 46, 1011–1040. corrosion prevention. Corros. Technol. 5, 215–219.