Rheological Behavior of Gypsum, Plaster, and

Hydroxyapatite Gel Blends

Mohamed Raii, Francisco Javier Escudero Sanz, Ange Nzihou

To cite this version:

Mohamed Raii, Francisco Javier Escudero Sanz, Ange Nzihou. Rheological Behavior of Gypsum,
Plaster, and Hydroxyapatite Gel Blends. Industrial and engineering chemistry research, 2012, 51
(34), pp.11163-11169. �10.1021/ie301154d�. �hal-01632401�

HAL Id: hal-01632401

https://hal.science/hal-01632401
Submitted on 17 Aug 2018

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
Rheological Behavior of Gypsum, Plaster, and Hydroxyapatite Gel
Blends
Mohamed Raii,* Francisco Javier Escudero Sanz,† and Ange Nzihou‡
Centre RAPSODEE, Mines Albi, CNRS, Université de Toulouse, Campus Jarlard, F-81013 Albi Cedex 09, France

ABSTRACT: The rheological behavior of water−calcium sulfates and hydroxyapatite−calcium sulfates binary blends was
considered in this study. These blends are to be used as permeable reactive barriers for groundwater remediation. The influence
of gypsum and plaster on the rheological behavior of Ca-HAGel has been studied. The results show that all blends and
formulations exhibit a shear-thinning effect and thixotropic behavior. The viscosity also depends on the level of concentration of
the solid and the reactivity of the blends. The hydration of plaster causes cohesion and agglomeration between particles due to
electrostatic forces, while the hydration of gypsum leads to dispersion due to repulsive forces between particles. The ζ potential
was used in this study to understand the interaction between particles and its effect on the global behavior of the blends. The
dissolution in water of sulfate ions containing calcium sulfate leads to a change in the surface charge of the blend. This allowed
the dispersion or agglomeration of the mixtures to be predicted.

1. INTRODUCTION There are three anhydrous or anhydrite forms produced by

Calcium phosphate is a well-known biomaterial frequently used sintering of hemihydrates or gypsum: the compact form α-
CaSO4 or CaSO4-I, obtained at 1200 °C and characterized by a
in the medical domain and as an agricultural fertilizer. It can
crystalline structure; the stable form β-CaSO4 or CaSO4-II,
come from natural sources such as fish or cow bones or as a
obtained at 400 °C, which is a natural insoluble anhydrite; the
result of a synthesis process using calcium and phosphate
metastable form γ-CaSO4 or CaSO4-III, obtained at low
compounds. Calcium phosphate has been investigated as a
temperature between 130 and 200 °C, which is a soluble
reactive material, especially in its thermodynamically stable
anhydrite.9 The stability of gypsum is also influenced by the
structure with the stoichiometric ratio Ca/P = 1.67
relative humidity (RH). At 10−35% RH, the migration of water
(hydroxyapatite, Ca-HA), to determine its usefulness in treating
molecules takes place toward the crystals, while at 100% RH,
contaminated water or as a stabilizing agent for hazardous
the migration takes place in the opposite direction.10,11 The
waste.1−3 Many studies have also shown the capacity of Ca-HA crystalline structure of the gypsum, characterized by a perfect
to remediate soil and groundwater.4,5 cleavage orientation (010), results in the circulation of water,
Gypsum can have two origins: natural gypsum and synthetic leading to dissolution on the surface (010).11,12 The blending
gypsum. Natural gypsum, characterized by a high purity, is an of hemihydrates with water gives a rigid and compact solid
anionic mineral with a C62h space group. This gypsum is a rock body, due to the hydration process.
resulting from the evaporation of water from saturated calcium The main purpose of this work is to investigate the
sulfate and geological sedimentation. Synthetic gypsum is a rheological behavior of water/gypsum, water/plaster, Ca-HA/
byproduct of industrial activities. It is obtained from a reaction gypsum, and Ca-HA/plaster blends. This is a preliminary step
between sulfuric acid and phosphate rock to produce in a project devoted to the formulation and characterization of
phosphoric acid (H3PO4). The precipitate, usually named a Ca-HA-based permeable reactive barrier. The evaluation of
phosphogypsum (PG), is dihydrated calcium sulfate. This the rheological behavior of a mixture of gypsum and plaster on
gypsum contains a significant amount of impurities and is Ca-HAGel aims to better understand the development of the
considered to be a waste product. The environmental impact matrix microstructure and, consequently, to assess its work-
associated with PG storage and disposal, as reviewed by Tayibi ability.
et al., is negative on the surrounding land, water, and air.6
Valorization seems to be by far the best way to reuse this
2. MATERIALS AND METHODS
hazardous material. Industrial PG can be valorized as a material
for soil stabilization. It can be added to Portland cement or to 2.1. Calcium Sulfates. The gypsum (G, CaSO4·2H2O) and
reactive materials such as fly ash to enhance the mechanical the plaster calcium sulfate (Pα, CaSO4·0.5H2O, prepared by the
properties of the soil and to reduce the plasticity index, wet method) used in this work are industrial byproducts of a
moisture, and water content.7 It can also be used as a hydraulic phosphoric acid (H3PO4) preparation from phosphate rock
binder in the design of calcium sulfoaluminate cement.8 The (PRAYON, Belgium). The plaster was used as received, and
calcium sulfates used in this work are industrial gypsum
(CaSO4·2H2O) and plaster (CaSO4·0.5H2O). They are by-
products of phosphoric acid production and are washed to
eliminate the residual fraction (phosphate, fluor, sodium,
radionuclides, etc.).
gypsum was sieved to under 1 mm and the larger fraction applied by a double-helical ribbon impeller. Samples were
discarded. homogenized before the shear stress τ (Pa) and viscosity versus
2.2. Hydroxyapatite Gel (Ca-HAGel). Ca-HAGel was the applied shear rate γ̇ (s−1) were measured. The dynamic
obtained by the reaction between calcite (CaCO3; 98% pure, trails applied on Ca-HAGel were performed with the same
Fisher Scientific) and ammonium phosphate (NH4H2PO4; apparatus using an oscillation mode with a preselected constant
99.4% pure, Fisher Scientific) at 25 °C with a molar ratio of shear stress of 50 Pa and a frequency ranging from 1 to 50 Hz.
Ca/P = 1.67, and the amount of demineralized water is 3 times The loss and storage moduli measurements were recorded for
higher than that of calcite. The equation below describes the 300 s at room temperature.
overall reaction pathway for the synthesis of Ca-HA. During the 2.4. Formulations. Rheological tests were carried out on
formation of Ca-HAGel, bubbling corresponding to the release pure Ca-HAGel, on binary water/gypsum (W/G), and on
of CO2 gas was observed. water/plaster (W/P) blends. Binary blends were prepared by
mixing the selected material with deionized water according to
10CaCO3 + 6NH4H 2PO4 + 2H 2O different mass ratios ranging from 0.4 to 1.4. The solid
concentrations of binary blends (ΓB1) and the W/G and W/P
→ Ca10(PO4 )6 (OH)2 + 6NH4OH + 2H 2O + 10CO2
mass ratios were calculated as follows:
At the beginning of the reaction, the pH value increased Vs W 1 − ΓB W 1 − ΓB
gradually from 6 to 8 and then stabilized after 12 h of synthesis. ΓB1 = ; = ; =
Vs + Vw G ρG ΓB P ρP ΓB
The stabilization of the pH revealed the formation of an
ammonium hydroxide buffer and the consumption of calcium where Vs is the volume of the solid phase, Vw is the volume of
carbonate. Ca-HAPowder was prepared by drying Ca-HAGel at water, and ρG and ρP are the densities of gypsum and plaster,
105 °C for 24 h to obtain a powder with monomodal particle- respectively. The Ca-HAGel content considered was from 0 to
size distribution. The water content calculated after drying at 50 wt % Ca-HAGel. The rheological behaviors of water-free
105 °C was 70%. Ca-HAGel, destined to be used for chemical formulations (Ca-HAGel/G and Ca-HAGel/P) with mass ratios
characterization, was filtered and rinsed with deionized water to of 50/50, 40/60, and 30/70 were not considered because the
dissolve any unreacted phosphate. pastes obtained were very thick and almost solids). The solid
2.3. Physical and Chemical Characterization. Powder concentrations of binary blends (ΓB2) containing either gypsum
densities were measured using a helium pycnometer (AccuPyc or plaster (calcium sulfate powder) with Ca-HAGel were
1330 V2.04N). The specific surface areas of the materials (G, P, calculated as follows:
and Ca-HA) were determined on a Micromeritics Gemini 2360,
by surface analysis using the Brunauer−Emmett−Teller VsT
ΓB2 = ; VsT = VCa ‐ HAPowder + VCSPowder ;
method, and the measurements were performed by nitrogen VsT + VwT
absorption at liquid-nitrogen temperature followed by
desorption (adsorption gas N2; heating temperature 150 °C). VwT = VCa ‐ HAWater
The particle-size distribution was determined by LASER
granulometry using a Malvern Laser Master Sizer Hydro 2000 where VsT is the total volume of the solid material equal to the
instrument (with ethanol as a dispersant shaken by ultrasound). sum of the Ca-HAPowder volume (VCa‑HAPowder) and the calcium
The pH measurements were made after dispersion of 10 g of sulfate powder volume (VCSPowder) and VwT is the total water
powder materials in 25 mL of deionized water and stirring for 1
h at room temperature, with the measurement being carried out equal to the Ca-HA water volume (VCa‑HAWater).
after 2 h. The particle charge was evaluated relative to the net
charge on the particles’ surfaces by evaluation of the pH point 3. RESULTS AND DISCUSSION
zero charge (pHPZC). This determination is based on the 3.1. Characteristics of the Materials. The results of the
electrochemistry and involves the introduction of 50 cm3 of a bulk density, specific surface area, and pH of the materials are
NaCl (0.01 M) solution in closed vials and adjustment of the given in Table 1. It can be seen that the bulk densities were
initial pH to values between 2 and 12 by the addition of NaOH used to calculate the solid concentrations in the blends. The pH
or HCl (0.1M). A mass of 0.15 g was added to the vials, and the of plaster is acidic, which provides an indication of the presence
final pH was measured and recorded after stirring for 48 h at of phosphoric acid. The specific surface area of plaster is 3
room temperature. The pHPZC value corresponds to the
intersection point of the titration curves obtained at one ionic Table 1. Physical Characteristics of the Materials
strength for the suspension (pHinitial) and blank solution
(pHfinal). Ca- Ca-
X-ray diffraction (XRD) analysis of the powders was GPowder PPowder HAPowder HAGel
performed on a Siemens D5000 XRD spectrometer using Cu bulk density ρ (g cm−3) 2.25 2.81 2.58 1.23
Kα (λ = 1.540 Å) anticathode radiation (power 40 mA, 45 kV). surface specific area SSA(m2 g−1) 26.27 72.81 138
The diffraction angle was set up from 10 to 80° 2θ with a step pH 8.44 2.88 7.85 8.41
count of 0.02° 2θ and a counting time of 3 s/step. ζ-Potential pHPZC 7.50 3.30 8.11
measurements were performed on a Zetasizer 3000 HSA residual moisture at 105 °C (%) 20.00 5.00 67
(Malvern Instruments Ltd.). loss upon ignition (at 1000 °C) 22.00 8.00 59.00
The rheological measurements were performed using a (%)
Rheostress (HAAKE RS 150 rheometer). The rheological granulometric fraction (%)
behavior and viscosity η (Pa·s) were evaluated for shear rates sand 79.6 48.9 43.4 9.0
ranging from 0 to 500 s−1 at constant temperature (20 ± 5 °C). silt 18.7 49.0 51.7 88.0
The sample was placed in a cylindrical vessel, and the strain was clay 1.7 1.7 4.9 3.0
Figure 1. Powder XRD pattern of Ca-HA heated for 10 h at 1000 °C [hydroxylapatite (+), hydroxyapatite (×), plaster (∧), gypsum (*), and
anhydrite (○)].

times higher than that of gypsum. The specific surface area of 4. RHEOLOGICAL BEHAVIOR
Ca-HAPowder indicates the surface reactivity of the particles. The 4.1. Steady Rheological Properties of Ca-HAGel. Apatite
net charge of the materials can be evaluated by referring to the gel is characterized by a specific texture caused by absorption of
pH of the aqueous solutions with the metal surface in a neutral water and interaction between fine particles. The identification
electric state and pHPZC. The plaster and Ca-HA particles are of the behavior and the structure Ca-HAGel requires an
positively charged (pH < pHPZC), and the gypsum particles are understanding of the influence of parameters such as material
negatively charged (pH > pHPZC). deformation, strength, and stability.
The classification and identification of materials following the As shown in Figure 2, the shear stress increases with the
geotechnical designation reveals that plaster and Ca-HAPowder shear rate when the stress exceeds a threshold value τ0 (1.425
seem to have a silty-sand texture, while gypsum, having a Pa), indicating that Ca-HAGel exhibits a shear thinning or,
diameter of less than 1 mm, is sandy-silt and Ca-HAGel is specifically, a viscoplastic behavior. Studies published by
identified as silty (Table 1). Knowles et al. carried out on a commercial Ca-HA suspension
XRD analysis of synthesized Ca-HA after heat treatment agree with this obtained result.14
The viscosity reaches an apparent value of η∞ = 0.144 Pa·s
(calcination at 1000 °C for 10 h to crystallize the structure and
above γ̇ = 200 s−1. The critical shear rate corresponding to the
eliminate calcite residue) showed the existence of a crystalline
Ca-HA dispersion is characterized by an extrapolated maximum
phase illustrated by hydroxyapatite in large quantities that had yield stress τ0e (2.289 Pa) and a minimum yield stress τ0 (1.425
synthesized stoichiometrically (Ca10(PO4)6OH2). The XRD Pa), related to the response of the gel to the deformation. The
patterns of gypsum and plaster reveal that there are almost no viscosity of Ca-HAGel is 0.144 Pa·s at a higher shear stress, and
common peaks. The Pα and G are characterized by five major above this value, the suspension flows. The increase in the shear
peaks at 14.7°, 25.7°, 29.6°, 31.8°, and 49.3° 2θ and 11.6, 20.7, stress leads to the disruption of the gel structure, and when the
29.0°, 31.1°, and 33.3° 2θ, respectively. The presence of an shear stress stops, the suspension tends to recover its original
anhydrite peak at 25.7° 2θ appears for the first time after heat structure (i.e., the flow is reversible).
treatment (Figure 1). The difference in the structures of both The viscosity does not only depend on the shear rate, but it
plaster and gypsum is related to the heat treatment, which is also correlated to the duration of the applied stress. The
modifies the crystalline lattice. The study carried out by Mandal internal structure of Ca-HAGel undergoes an initial homoge-
et al. demonstrated that heating the gypsum (CaSO4·2H2O; nization at low shear rate values corresponding to low levels of
plasticity. The constraint force is too small to disorganize the
98% pure) at 90 and 350 °C for 10 h revealed the presence of
structure; it requires an increase in the magnitude of the
two kinds of water: loosely and strongly held water molecules.13
applied force for the flow to occur. Beyond the shear stress
Consequently, the loss of water molecules modifies the threshold, with an increasing shear rate, the flow is enhanced by
monoclinic structural morphology of gypsum, where the the gradual dispersion of particles, promoting the separation
elimination of 1.5 water molecules allows the apparition of effect, which causes a decrease in the viscosity. The viscosity
hexagonal hemihydrates. Above 200 °C, the symmetry changes (apparent viscosity) value stabilizes at high shear rates. Once
to the orthogonal configuration due to the elimination of the apparent viscosity is reached, there is no further
strongly held water molecules. modification observed for the structure of Ca-HAGel.
 During synthesis, the storage and loss moduli (G′ and G″) rose
with an increase in the frequency and their values decreased
with the aging of the gel.
At low frequencies, the 12 h age dispersion (Figure 3a)
seems to have a heterogeneous structure at the beginning of the
periodic oscillation application. It is observed that G′ is greater
than G″, which corresponds to an elastic behavior. Therefore,
when the frequency is increased, the modulus values increase
strongly because of the dispersed state of the particles.
For a 24 h age dispersion (Figure 3b), the values of the
storage and loss moduli are almost zero below 15 Hz. The
strain response at constant stress after 24 h of synthesis is
delayed compared to the 12 h age dispersion measurement.
This change in behavior, identified at low frequencies, is due to
the evolution of the reaction. This response is related to
precipitation of the Ca-HA suspension occurring when the pH
value buffers around 8, leading to the formation of a colloidal
suspension. The behavior remains elastic at high-frequency
values.
After 48 h of synthesis (Figure 3c), we observe a complete
change in behavior at low frequencies. For frequencies below
25 Hz, G′ and G″ increase exponentially at low stress. Below 25
Hz, G″ is greater than G′, indicating viscous behavior. This
simply means that the suspension follows a constant structural
deformation where the stress is proportional to the strain rate.
Figure 2. Rheological behavior and viscosity of Ca-HAGel after 48 h of When the frequency reaches values higher than 25 Hz, G″
synthesis: (a) τ = f(γ̇); (b) log(τ) = f(γ̇) [−×−, shear stress (Pa); − becomes smaller than G′, indicating an elastic behavior: the
+ −, viscosity (Pa·s)]. stress is proportional to the deformation. However, the
viscosity increases progressively, indicating the agglomeration
4.2. Dynamic Rheological Properties of Ca-HA. To of colloids and the formation of gel-paste calcium phosphate.
explain the viscoelastic behavior observed for Ca-HAGel, This result is important because it shows the transition between
dynamic oscillatory tests were carried out to investigate the a suspension and pasty behavior. As demonstrated by Knowles
internal structure and response of the gel during Ca-HA et al.14 and Bao et al.,15 the mutual approach of particles in
synthesis. The dynamic rheological (oscillatory) test aims to solution is correlated to the positive charge of the Ca-HA
determine the storage (G′) and loss (G″) moduli. Figure 3 surface. The behavioral changes show the formation of gel
represents the results of the frequency−sweep response of a particles or gel-paste structures after 2 days of reaction,
Ca-HA suspension after 12, 24, 48, and 72 h of synthesis. exhibiting a viscoelastic behavior. The dynamic rheological
property study carried out by Liu et al.16 on a calcium
phosphate cement slurry (a composite of equimolar tetracal-
cium and dicalcium phosphate anhydrous) under the
conditions of a 2.0 powder/liquid ratio and a 1% strain showed
that, at the range of low strain and low frequency, the slurry
changed its behavior with frequency increases. The slurry began
to flow and presented a shear-thinning behavior.
For the 72 h age dispersion (Figure 3d), the loss modulus
values are higher than the storage modulus values, thus
revealing a viscous behavior. Nevertheless, when the frequency
increases and the strain exceeds the yield strain, the viscosity
decreases. The decrease in the G′ and G″ values is due to gel
aging after 48 h of synthesis, which causes deformation of the
gel structure. The disruption of extractible water molecules
caused by bonds between particles leads to flow and
demonstrates that Ca-HAGel exhibits a shear-thinning behavior.
4.3. Rheological Behavior of Binary Blends. The
viscosity of the binary matrix was measured and plotted versus
the concentration of added powder, as illustrated in Figures 4
and 5. W/G and W/P blends exhibit shear-thinning behavior.
Figure 4 shows that the viscosity of W/G blends decreases with
Figure 3. Frequency−sweep curves of Ca-HA as a function of the decreasing gypsum concentration until an almost constant
synthesis time (−Δ−, storage modulus G′, −□−, loss modulus G″). apparent viscosity for W/G ratios above 0.6 was achieved. The
The description of the behavior of Ca-HAGel is based on two viscosity plateau reveals no influence of water on the hydrated
conditions: if G′ > G″, the behavior is elastic; if G′ < G″, the behavior gypsum particles. Hydrated gypsum particles present similar
is viscous. double-electric layers of a common charge (negatively
Figure 4. Viscosities of W/G and W/P blends at different mass ratios.

Figure 5. Viscosities of Ca-HA/G and Ca-HA/P formulations.

charged), which generates electrostatic repulsion between particle constructing a growth slice.17 The predominance of reactive
surfaces (due to the overlap of the electrical double-layer sites on the grains attracts cations, which form a rigid network
theory). Gypsum particles are mainly positively charged and are that surrounds the layer of adsorbed water molecules. It is also
covered by a layer of positive average charge. This positive- known that the attraction of plaster particles requires cohesion
charge layer stabilizes the dispersion through particle repulsion, between atoms to form the contiguous flakes. The development
avoiding particle aggregation. At rest, particles tend to of this structure (agglomerated structure or plaster paste)
precipitate. If a stress is applied, the suspension is dispersed, depends on the amount of hydration water. According to the
leading to a decrease in the viscosity. The gypsum stochiometry, the α-hemihydrates require a W/P ratio higher
concentration controls the dispersed state; the repulsive forces than 0.186 to be transformed into gypsum. To improve the
tend to stabilize the suspension because the particles are consistency of the blends, the hydration rate was chosen from a
negatively charged. The apparent viscosity is obtained at a high range of 0.4−1.4. In the dynamical study in the linear
shear rate, where gravity forces are negligible. It can be seen viscoelastic domain published by Finot et al., the time−sweep
that, above a 0.6 W/G ratio, the interparticle distance is large test at 0.015% (ω = 10 rad·s−1), carried out on W/P = 0.7,
enough to make the apparent viscosity independent of the indicates that the rise in the storage modulus reflects the
gypsum content. hydration and setting processes. This reveals an increase in the
With regards to the plaster blends (W/P), the viscosity level of crystallized gypsum.20 Obviously, the viscosity
decreases with decreasing plaster concentration. The hydration decreases steadily with an increasing amount of water, and
of hemihydrate particles, requiring 1.5 water molecules, leads to the application of shear stress destroys the links between
crystallization. In general, the contact of plaster particles with particles, facilitating decoagulation of the structure and leading
water triggers dissolution, which induces the saturation or to flow.
supersaturation of the solution in Ca2+ and SO42−, leading to Plaster hydrates during viscosity measurement under stress.
the formation of gypsum nuclei.17 Consequently, the hydration The plaster viscosity value depends on the volume fraction of
control scenario is correlated with the nucleation and crystal- the solids (solid concentration ΓB1), particle size, and water
growth mechanism.18,19 excess (free of water after full plaster hydration). At a high
The high specific surface area of plaster particles plays a shear rate, it is possible to measure the apparent viscosity of the
major role in the multiplication of intergranular links. dispersion of agglomerated hydrated particles. The apparent
Interaction between plaster particles occurs by electrostatic viscosities of W/G blends are lower than those of W/P blends.
forces (i.e., the spread of ions in an extended diffuse layer) and Because plaster hydration held the measurement long, the
attractive forces (van der Waals) responsible for adhesion. difference observed in the apparent viscosity between W/G and
Furthermore, adhesion is described as the formation of two W/P cannot be solely attributed to hydration. The difference
Ca2+/SO42− layers bound together by ionic interaction, observed in a specific surface area may provide an explanation.
 Figure 5 shows that the viscosity increases when the solid
concentration (ΓB2) increases, with the latter being calculated
relative to the water (by wt %) present in Ca-HAGel. It should
be noted that the formulations having Ca-HAGel/G or Ca-
HAGel/P ratios higher than 1.5 are solids and their rheological
properties were not characterized. The water content in Ca-HA
is more than sufficient to hydrate calcium sulfate hemihydrate
particles. In this case, an almost solid paste is formed. Replacing
20% of Ca-HAGel by gypsum or plaster produced little influence
on the viscosity, indicating that for this blend there is little
influence of calcium sulfate on the rheological properties of Ca-
HAGel. With regards to the already hydrated gypsum particles,
increasing the gypsum concentration of the blend up to 40%
produced a slight increase in the viscosity. Blends containing Figure 7. ζ potential of Ca-HA/G and Ca-HA/P formulations without
more than 40% of gypsum are almost solids. The addition of added water [− −, ζ potential of gypsum (≈20 mV); − · −, ζ potential
plaster to Ca-HA requires a high stress to cause the flow. of plaster (≈5 mV); − ·· −, ζ potential of Ca-HA (≈−20 mV); −□−,
Plaster particles in contact with Ca-HA water cause plaster ζ potential of Ca-HA/G formulations; −Δ−, ζ potential of Ca-HA/P
hydration, creating strong bonds that encompass the Ca-HA formulations].
particles and make the flow of the blend difficult. Hydrated
plaster strongly increases the viscosity of Ca-HAGel by its own generating the change from negative to positive charge. The
weight. Figure 6 shows the thixotropic behavior of binary adsorbed amount of anionic sulfate on the Ca-HA particles is
equal to the Ca-HA electrical charge sites, and the coating of
gypsum sulfate Ca-HA particles is the main reason for
modification of the ζ potential. The surfaces of the plaster
particles are characterized by a positive charge, and the
formation of a plaster solid body is explained by cohesion
between particles and illustrated by a low ζ-potential value.
Figure 7 represents the ζ potential of formulations containing
Ca-HA and different proportions of the W/P ratio. With an
increase in the amount of plaster, the ζ potential remains
constant (with values similar to those of the plaster at the
beginning of the hydration process) and the surface charge of
Ca-HA changes. The nucleation process of hemihydrated
particles starts by the formation of Ca2+/SO42−-bound double
layers, and the concentration of sulfate decreases, indicating
that it is consumed during hydration. However, fixation of the
Figure 6. Example of a thixotropic hysteresis loop of composites 80% remaining sulfate in solution on the Ca-HA surface leads to a
Ca-HA/20% G and 80% Ca-HA/20% P [-*-, Ca-HAGel (η∞ = 0.144 change in the charge.
Pa·s); -×-, 80% Ca-HAGel/20% P (Γ = 0.369; η∞ = 0.348 Pa·s); -+-,
80% Ca-HAGel/20% G (Γ = 0.384; η∞ = 0.318 Pa·s)]. 5. CONCLUSION
The rheological test carried out reveals shear thinning and the
blends containing a mass ratio blend of 80% of Ca-HA and 20% thixotropic behavior of Ca-HAGel for formulations considered.
of plaster and a mass ratio of 80% of Ca-HA and 20% of The internal structure of Ca-HAGel exhibits viscoelastic
gypsum, indicating that the blends recover their initial structure. behavior after 48 h of synthesis.
In this case, viscosity values increase slightly compared to the The positively charged gypsum particles cause repulsion of
viscosity of Ca-HAGel while keeping their flow characteristics. the particles (repulsive forces), and the viscosity of the W/G
4.4. Interaction between Particles. The interpretation of (0.2−1.4) blends remains constant. Plaster is characterized by
interactions between these compounds seems to be complex. attractive forces, and the particles are positively charged. The
The ζ potential was considered as a means to understanding hydration of plaster leads to cohesion between particles, which
the interactions, based on the net charge of the particles, and explains the increase in the viscosity of the W/P blends
hence describing the suspension structure. The ζ potential compared to the W/G blends. The ζ potential of Ca-HA is
indicates the repulsion forces between particles promoting the negative, and dissolution of the calcium sulfate particles
dispersion stability. Figure 7 shows that the Ca-HA particles are saturates the solution in sulfate ions (SO42−); the charge of
characterized by a negative ζ potential, indicating that the the particles becomes positive because of fixation of SO42− on
particles are negatively charged (≈−20 mV)14 and the gypsum the Ca-HA surface. The adsorption of sulfate on the Ca-HA
particles are positively charged (≈20 mV).21 The Ca-HA ζ- particle surfaces influences the structures of the formulations
potential value indicates the instability of the gelled particles. and explains why the viscosity is independent of the Ca-HAGel
Adding gypsum to Ca-HA leads to charge modification. The content. Plaster hydration plays a major role in the Ca-HAGel
presence of repulsive forces is correlated to the high value of stabilization structure, and gypsum particles undoubtedly
the ζ potential predicting the stability. These results support enhance the hydraulic performance. Modification in the ζ
the assumption of sulfate attraction on the Ca-HA surface, potential can affect the retention capacity, reactivity, and
indicating dispersion of the particles. However, sulfate mechanical resistance. These aspects will be the subject of a
adsorption changed the charge distribution of Ca-HA, later study.
 This study has led to the understanding of the rheological (17) Singh, N. B.; Middendorf, B. Calcium sulphate hemihydrate
and physicochemical behavior on water/calcium sulfates and hydration leading to gypsum crystallization. Prog. Cryst. Growth
hydroxyapatite/calcium sulfate blends. These results will be Charact. Mater. 2007, 53, 57−77.
useful in the formulation of a permeable reactive barrier. (18) Bosbach, D.; Junta-Rosso, J. L.; Becker, U.; Hochella, M. F., Jr.
Gypsum growth in the presence of background electrolytes studied by

■ AUTHOR INFORMATION
Corresponding Author
Scanning Force Microscopy. Geochim. Cosmochim. Acta 1996, 60 (No.
17), 3295−3304.
(19) Sievert, T.; Wolter, A.; Singh, N. B. Hydration of anhydrite of
*Tel: +(00)33563493241. E-mail: Mohamed.Raii@mines-albi. gypsum (CaSO4.II) in a ball mill. Cem. Concr. Res. 2005, 35, 623−630.
(20) Finot, E.; Lesniewska, E.; Mutin, J.-C.; Goudonnet, J.-P.; Mutin,
fr.
J.-C.; Domenech, M.; Ait Kadi, A. Correlation surface forces with
Notes surface reactivity of gypsum crystals by atomic force microscopy.
The authors declare no competing financial interest. Comparison with rheological properties of plaster. Solid State Ionics
2001, 141−142, 39−46.
■ ACKNOWLEDGMENTS
The authors are thankful to PRAYON Belgium for providing
(21) Peng, J.; Qu, J.; Zhang, J.; Chen, M.; Wan, T. Adsorption
characteristics of water-reducton agents on gypsum surface and its
effects on the rheology of gypsum plaster. Cem. Concr. Res. 2005, 35,
gypsum and plaster and their interest in the present study. 527−531.

■ REFERENCES
(1) Chen, X.; Wright, J. V.; Conca, J. L.; Peurrung, L. M. Evaluation
of heavy metal remediation using mineral apatite. Water, Air, and Soil
Pollution 98; Kluwer Academic Publishers: Dordrecht, The Nether-
lands, 1997; pp 57−78.
(2) Nzihou, A.; Sharrock, P. Calcium phosphate stabilization of fly
ash with chloride extraction. Waste Manage. 2002, 22 (2), 235−239.
(3) Baillez, S.; Nzihou, A.; Bernache-Assolant, D.; Champion, E.;
Sharrock, P. Removal of aqueous lead ions by hydroxyapatites:
Equilibria and kinetic processes. J. Hazard. Mater. 2007, A139, 443−
446.
(4) Moore, R. C. Stabilization and Solidification of Hazardours. In
Radioactive and Mixed Wastes; Roger, D. S., Shi, C., Eds.; 2005; CRC
Press, pp 98−112; available on http://www.books-google.fr.
(5) Conca, J.; Wright, J. An Apatite II permeable reactive barrier to
remediate groundwater containing Zn, Pb and Cd. Appl. Geochem.
2006, 21 (12), 2188−2200.
(6) Tayibi, H.; Choura, M.; Lopez, F. A.; Alguacil, F. J.; Lopez-
Delgado, A. Environmental impact and management of phosphogyp-
sum. J. Environ. Manage. 2009, 90, 2377−2386.
(7) Degirmenci, N.; Okucu, A.; Turabi, A. Application of
phosphogypsum in soil stabilization. Building Environ. 2007, 42,
3393−3398.
(8) Kuryatnyk, T.; Angulski da luz, C.; Ambroise, J.; Pera, J.
Valorization of phosphogypsum as hydraulic binder. J. Hazard. Mater.
2008, 160, 681−687.
(9) Iribarne, A. P.; Iribarne, J. V.; Anthony, E. J. Reactivity of calcium
sulfate from FBC systems. Fuel 1997, 76 (4), 321−327.
(10) Auvray, C.; Homand, F.; Hoxha, D. The influence of relative
humidity on the rate of convergence in an underground gypsum mine.
Int. J. Rock Mech. Mining Sci. 2008, 45 (8), 1454−1468.
(11) Finot, E.; Lesniewska, E.; Mutin, J.-.C; Goudonnet, J.-P.
Reactivity of gypsum faces according to the relative humidity by
scanning force microscopy. Surf. Sci. 1997, 384, 212−217.
(12) Fan, C.; Teng, H. Surface behavior of gypsum during
dissolution. Chem. Geol. 2007, 245 (3−4), 242−253.
(13) Mandal, P. K.; Mandal, T. K. Anion water in gypsum
(CaSO4·2H2O) and hemihydrates (CaSO4·1/2H2O). Cem. Concr. Res.
2002, 32, 313−316.
(14) Knowles, J. C.; Callcut, S.; Georgiou, G. Characterisation of the
rheological properties and zeta potential of a range of hydroxyapatite
powders. Biomaterials 2000, 21 (13), 1387−1392 . Retrieved from
http://www.ncbi.nlm.nih.gov/pubmed/10850933.
(15) Bao, Y.; Senos, A. M. R.; Almeida, M.; Gauckler, L. J.
Rheological behavior of aqueous suspensions of hydroxyapatite
(HAP). J. Mater. Sci.: Mater. Med. 2002, 13 (7), 639−43 . Retrieved
from http://www.ncbi.nlm.nih.gov/pubmed/15348572.
(16) Liu, C.; Shao, H.; Chen, F.; Zhen, H. Rheological properties of
concentrated aqueous injectable calcium phosphate cement slurry.
Biomaterials 2006, 27, 5003−5013.