Effect of Low Calcite Addition on Sulfate Resisting (SR) Portland Cements: Hydration Kinetics at Early Age and Durability Performance After 2 Years

  • Islem LabidiEmail author
  • Sonia Boughanmi
  • Abdelhafid Khelidj
  • Mohamed El Maaoui
  • Adel Megriche
Original Article


The limestone addition to sulfate resisting Portland (SR) cement is often used to adjust artificially the amount of aluminate (C3A) in order to keep it less than 3%. However, CaCO3 addition is well known to induce thaumasite formation (CaCO3·CaSO4·CaSiO3·15H2O) in Portland cement during external sulfate attack. This compound decreases the concrete durability because of its high volume expansion. In this work we studied the addition of limestone within low amounts (≤ 6.21%) into SR Portland cement. Thus, three commercial samples (A, B and C) containing limestone addition were analyzed by XRF technique and the results were used to estimate the mineralogical composition by Bogue method. XR diffraction coupled to Rietveld algorithm was then performed to determine the real amounts of mineralogical phases. The results were as following: C3S contents were 49.70, 45.30 and 50.60%; C3A: 1.40, 1.10 and 2.5% and calcium carbonate addition 6.21, 2.10 and 2.10%, respectively. The effect of calcium carbonate on hydration kinetics at early age has been studied and discussed. Durability of studied cement samples was followed during 2 years into Na2SO4 and MgSO4 solutions (10 g/L). Thaumasite formation was observed only in cement mortars corresponding to the sample 6.21% CaCO3, C3S (49.70%) and C3A (1.4%) at 22 °C, this result is different from earlier works in which thaumasite precipitates at low temperature (5 °C) and high CaCO3 contents (> 15%). This result proved the SR Portland cement must contain very low limestone addition, less than 6% and much lower if used at 5 °C or below.


SR Portland cement Limestone addition Calcite Hydration kinetics Thaumasite Durability 














\( \bar{S} \)







Loss in ignition

1 Introduction

Sulfate resisting (SR) Portland cement [1] is widely used in building works to improve concrete resistance against external sulfate attack. This chemical aggression causes cracks, expansion and chemical disintegration of cement hydration products. This is due to the formation of secondary ettringite (C3A.3CaSO4.32H2O) and secondary gypsum (CaSO4.2H2O) [2]. In fact these compounds have a large volume expansion and cause cracking of concrete surface. So, standard rules require very low content of aluminate (C3A ≤ 3%) [1]. This kind of cement (SR Portland cement) is very difficult to manufacture [3] since low alumina implies the use of high kiln temperature in order to keep a sufficient liquid phase necessary to maintain the stability of the cement manufacturing process [4, 5].

Many cement produces use the addition of limestone (carbonate calcium) during cement milling operation to low artificially the C3A content. In general, limestone addition in common Portland cement is used to lower the costs while maintaining the required compressive strength of cement product [6].

High calcite addition (10–20%) was enhanced the rate of hydration of cement phases and rises the compressive strength at early ages [7]. Also, Kakali et al. [8] proved that the replacement of 10–20% of clinker by limestone caused the transformation delay of the ettringite to monosulfoaluminate since the monocarboaluminate formed instead to the monosulfoaluminate.

However by ageing ordinary Portland cement containing limestone, in presence of sulfate ions, crystals of thaumasite (CaCO3·CaSO4·CaSiO3·15H2O) [9, 10] appear having formula similar to that of ettringite [11]. Precipitation occurs preferably at low temperature about 5 °C [12, 13]. Tsivilis et al. [14] studied the behaviour of Portland cement containing of 15–30% of limestone at 5 °C in 1.8% of MgSO4 solutions, they showed that the higher limestone addition is the more the deterioration of concrete is severe, thaumasite was appeared after 4–12 months.

Issar et al. [15] studied the sulfate resistance of low C3A Portland cements in function of C3S contents, natural pozzolana (20–40%) and limestone (10–20%) addition. They showed that low C3S Portland cements have good resistance to external sulfate attack.

In fact, high C3S content was increased the calcium hydroxide (CH) in early age which favors secondary ettringite formation [16].

It seems that studies dealing exclusively of the effect of low (< 10%) limestone addition on sulfate resisting Portland cements are not under taken.

We proposed to investigate the properties of three commercial sulfate resisting Portland cements, in which there are limestone as minor additive.

Our objective is not to follow the variation of compressive strength, but to evaluate the effect of a low calcite addition on sulfate resisting Portland cement performances.

This evaluation is divided on two steps: firstly, we investigate the effect of a small limestone additive on hydration kinetics of SR Portland cement at early age, pastes of cement samples were hydrated in various times intervals (1, 2, 7 and 14 days) and studied by means infrared spectroscopy (FTIR) and thermogravimetric analysis (TGA). The reactivity behavior of cement samples is followed by using isothermal calorimetry. In this part, hydration kinetics is highlighted from the determination of hydration degree α (t) by using two different methods: the calculation of the ultimate chemically bound water (Wt): calculation of bounded water by TG curve and measurement of the cumulative heat released during hydration.

Secondly, we study thaumasite formation into A, B and C cements after 2 years exposure to aggressive sulfate mediums (10 g/L Na2SO4 and 10 g/L MgSO4) at 22 °C.

2 Materials and Methods

2.1 Materials

Three commercial sulfate resisting Portland cements used in this study were taken from different cement industries located in Tunisia. The mineralogical compositions presented in Table 1 of these samples were determined by using Rietveld refinement method. This work is a subject of a research paper still under review [17].Agreement parameters: R expected (Rexp), R profile (Rp), Weighted R profile (Rwp), Goodness of fit (GOF) and R Bragg (Table 2) of all analysis samples confirm the good quality of the fitting procedure.
Table 1

A, B and C cements chemical and mineralogical analysis (in weight percentages) as well as Blaine surface (cm2/g) and bulk density (g/cm3)

Chemical composition (Wt%)a













Cement A












Cement B












Cement C












Mineralogical composition (Wt%) determined by Rietveld method










Free lime



Cement A












Cement B










Cement C










aThe chemical composition is determined from X-ray fluorescence analysis

bTHE amount of calcite (= CaCO3) is determined from thermogravimetric analysis of anhydrous cement

Table 2

Agreement parameters of Rietveld refinement of cement samples










Cement A







Cement B







Cement C







The chemical and mineral compositions (see Table 1) of three samples were similar.

3 Cement Pastes and Mortars Preparation

The pastes used to study the hydration process were prepared by mixing the dry cement powder with distilled water (using water to cement ratio (w/c) equal to 0.4), using automatic mixer during two minutes separately to obtain a paste. The paste samples were allowed to hydrate in plastic bottles at 22 °C, in ambient atmosphere. The hydration of each paste was stopped after different time intervals 1, 2, 7and 14 days by submerging small pieces in ethanol and filtration under vacuum for 10 min. Every sample was dried at 50° C. The dried specimens were grind using agate mortar and sieved to 80 µm for making of FTIR and TGA studies.

To perform durability test of cement samples, mortar prisms (40 mm × 40 mm × 160 mm) were prepared by mixing of 450 g cement, 1350 g normalized sand and 225 g distilled water with w/c equal to 0.5 [18]. Prisms were totally immersed at 22 °C in three different mediums: tap water, 10 g/L Na2SO4 and 10 g/L MgSO4. All different solutions were refitted every month [19]. Stopping of hydration of each mortar is performed as described in previous paragraph. The laboratory is equipped with a storage chamber of the cement mortars and pastes especially designed to maintain a constant temperature at 22° C and humidity equal to 98%.

The sulfate solutions were prepared at laboratory scale from Na2SO4 and MgSO4.7 H2O salts. In addition, the tap water is changed from season to season and the average of chemical composition of this water was donated in following Table 3. The tap water medium is not considered an aggressive environment as indicated in the classification of severity of sulfate environments (less than 150 mg/L) [20]. So the sulfate concentration in tap water is remained negligible towards the other aggressive sulfate environments (10 g/L Na2SO4 and 10 g/L MgSO4).
Table 3

The average of chemical composition of tap water used in this study











Concentration (mg/L)










4 X-Ray Diffraction Analysis

The mortars were analyzed by means of Philips PANalytical X’Pert PRO X-ray diffractometer with a CuKα radiation at a scan speed of 0.06° 2θ/min. Acceleration power applied was 40 kV, with a current of 40 mA.

The identification of crystalline phases of all samples was obtained by comparing the XRD diffraction patterns to those of individual phases given into the database (powder diffraction file) PDF provided by the ICDD (International center for diffraction data).

This identification is performed automatically by the X’Pert High Score Plus software.

5 Ftir

FTIR transmission spectra were measured by Perkin Elmer 783 dispersive spectrometer in the scan range of 4000–400 cm−1 for all the samples before and after hydration at different times.

6 Isothermal Calorimetry

To evaluate the effect of calcite on cement hydration kinetics, an isothermal calorimeter (TAM air brand) operating at 25 °C was used to measure the hydration heat flow and cumulative heat flow of the anhydrous cement samples (A, B and C) during 8 days. All isothermal calorimetry tests were performed at w/c ratio equal to 0.4.

7 Thermogravimetric Analysis

The thermal analysis of hydrated cement pastes were performed from 23 to 1030 °C, using a dry nitrogen atmosphere, with heating rate of 5 °C/min, in a Mettler tolerdo TGA/DSC 1 STARe System model instrument. DTG curves were obtained directly from the software of the equipment.

8 Methods to Determine Degree Hydration (α (t)) of Cement Blended

8.1 Water Non-Evaporable or Chemically Bound Water

Different methods were proposed to calculate the hydration degree of cement pastes based on the amount of chemically bound water (= water non-evaporable). The latter was determined from the thermogravimetric analysis of hydrated cement pastes at different times t.

In this study, the suitable way to assess the hydration degree was described by Mounanga [21]. Degree of hydration “α (t)” (Eq. (1)) is a ratio of chemically bound water at time t “Wt” (Eq. (2)) to the ultimate chemically bound water “W∞”.
$$\alpha \left( {\text{t}} \right) = \frac{\text{Wt}}{{{\text{W}}\infty \times {\text{mc}}}}$$
mc is the initial anhydrous cement sample mass (Eq. (3)).
$$\begin{array}{ll} {\text{Wt }} &= \left| {\Delta {\text{m 145}} \to 1 0 0 0 ^ {{\circ {\rm C(t)}}}} \right| - \left| {\Delta {\text{m 600}} \to 8 0 0 ^ {{\circ {\rm C(t)}}}} \right| \\ &\quad+ {\text{ m}}_{\text{d}} - {\text{ m}}_{\text{c}} \times {\text{ LOI}}_{\text{C}} , \end{array}$$
$${\text{m}}_{{\text{C}}} = \frac{{{\text{ms}}}}{{\left( {1 + \frac{{\text{W}}}{{\text{C}}}} \right) \times (1 + {\text{LOIc}})}},$$
where \(\Delta {{{\rm m} 145 - }} \to 1 0 0 0 ^ {{\circ\, {\rm C(t)}}}\) and \(\Delta {\text{m 600 - }} \to 8 0 0 ^ {{\circ\, {\rm C(t)}}}\) are the mass loss recorded in the range of 145–1000 °C (g), and between 600 and 800 °C, respectively.

md: the device’s drift (g), LOIC: Loss in ignition of anhydrous cement mass, ms: the initial mass of sample putted in the ceramic crucible to do ATG measurement.

The ultimate chemically bound water corresponds to the content of non-evaporable water after complete hydration. It is a theoretical value obtained from the chemical composition potential of the anhydrous cement and quantities of water necessary for the complete hydration of the different constituents. The values obtained from this calculation are W∞ = 0.22, 0.24 and 0.23 g for cement A, cement B and cement C, respectively. It is corresponded to the water quantity consumed after the hydration of 1 g of cement [22].

8.2 Calorimetric Results [23]

The degree of hydration measured by isothermal calorimetry (Eq. (4)) corresponds to the ratio between the amount of cumulative heat released at time t, Q(t) and that of heat released after complete hydration, Qtot(∞).
$$\alpha \left( {\text{t}} \right) = {\text{ Q}}\left( {\text{t}} \right)/{\text{ Q}}_{\text{tot}} \left( \infty \right),$$
Qtot(∞) is the potential heat of hydration liberated if all components reacted as described in the following formula (Eq. (5)):
$${\text{Q}}_{\text{tot}} \left( \infty \right) = \mathop \sum \limits_{{{\text{i}} = 1}}^{\text{n}} {\text{Qi }}({\text{Pi}}) \ldots \left( {{\text{J}}/{\text{g}}} \right),$$
where n: Principles cement phases (C3S, C2S, C3A and C4AF), Qi: Enthalpy of hydration of pure cement phases (J/g), Pi: Weight percentage of cement phases deduced by Rietveld analysis.

9 Results and Discussion

9.1 Early Age Hydration of Resisting Sulfate Portland Cements

9.1.1 Qualitative Analysis: FTIR

Diffractograms for dry cements only has been used to quantify the mineralogical compositions. XRD analysis for cement pastes is difficult to handle because the major hydrated phases namely C–S–H are amorphous. Even primary ettringite, monosulfoaluminate and monocarboaluminate which have low crystallinity are difficult to be detected [23]. Only the diffraction patterns performed after one or 2 years could be interpreted.

Figures 1, 2 and 3 represent the FTIR absorption spectra of the hydrated pastes after 1, 2 and 14 days, respectively.
Fig. 1

FTIR spectra of cement pastes after 1 day of hydration

Fig. 2

FTIR spectra of cement pastes after 2 days of hydration

Fig. 3

FTIR spectra of cement pastes after 14 days of hydration

We note that calcium monocarboaluminate (CaO·Al2O3·CaCO3·11 H2O) can be observed only into cement A paste after 14 days hydration, the corresponding absorption bands are: 1426 cm−1 and 716 cm−1 for ʋ3 and ʋ4 CO32−, respectively, 535 cm−1 corresponded to AlO6 associated to monocarboaluminate and a last typical band at1641 cm−1assigned to ʋ2 of H2O associated to CaO·Al2O3·CaCO3·11 H2O [24, 25]. This result is explained by the reaction between the aluminate compound C3A and calcium carbonate (CaCO3), it is the chemical effect of the limestone addition only shown in cement A pastes.

The FTIR spectra after 7 days was done but not presented because it does not have a particularity. It does not contain any monocarboaluminate.

Other bands of interest at 1120 and 966 cm−1, not involving CaCO3 are corresponding to asymmetric stretching of sulfate ion SO42− [26] in primary ettringite and anti-symmetric stretching vibration of Si–O [27] in hydrated calcium silicate (C–S–H) respectively.

9.2 Hydration Kinetics

10 Thermogravimetric Analysis

The thermogravimetric analysis was used in this study to evaluate the effect of limestone additive on cement hydration kinetics. Figures 4, 5 and 6 show the TGA and DTG curves for cement A, cement B and cement C hydrated pastes for 1, 2, 7 and 14 days. The analysis of thermogravimetric test results (Figs. 4, 5 and 6.) reveals three different zones of weight loss:
Fig. 4

TG-DTG curves of cement A after a 1 day, b 2 days, c 7 days and d 14 days of the hydration

Fig. 5

TG-DTG curves of cement B after a 1 day, b 2 days, c 7 days and d 14 days of the hydration

Fig. 6

TG-DTG curves of cement C after a 1 day, b 2 days, c 7 days and d 14 days of the hydration

  • From 25 to 400 °C: this weight loss is due to the escape of all the water present in the porosity of the mortar (free water and capillary water) and the departure of chemically bound water to the hydrates (C–S–H + Ettringite). In addition, the decomposition of gypsum (CaSO4. 2H2O) takes place at around 140–150 °C [28]. That’s why, in the determination of hydration degree procedure in the previous Sect.~II.8.1, we adopted the weight loss between 145 and 1000 °C.

  • From 400 to 600 °C: the endothermic peak is related to the deshydroxylation of portlandite (Ca(OH)2) according to the following reaction:

    $${\text{Ca }}\left( {\text{OH}} \right)_{ 2} \to {\text{CaO}} + {\text{ H}}_{ 2} {\text{O}} .$$
  • From 600 to 800 °C: the TGA and DTG of this region is the result of carbonated phases (CaCO3) decomposition [29] according to the reaction CaCO3 → CaO + CO2 [30].

We note that the peak at 700 °C vary with the amount of calcite just below 650 °C as indicates Figs. 4, 5 and 6, this latter wide band at 650 °C may be related to the release of H2O from the structural hydroxide groups of C–S–H precipitates [29].The evolution of Ca(OH)2 into cement pastes is represented by the peak in the range of 400 to 600 °C. When this peak is intense, the rate of cement hydration is high.

We note that Ca(OH)2 content increase continuously during hydration as indicated Fig. 7. The amount of Ca(OH)2 at early ages is always reported to the amount of alite (C3S) in dry cement paste [31]. In this case, cement A which contains 49.70% of C3S is that produces the less quantities of carbonate hydroxide. Instead of cement B which contain only 45.30% of C3S. It seems that the calcite content plays a role of retarder additive. The knowledge of Portlandite amount at early age hydration in cement pastes could be a first prediction of the durability of RS Portland cement against external sulfate attack (see the next paragraph).
Fig. 7

Content of Portlandite as a function of curing time for all cement pastes

As known from the previous research studies [32, 33, 34, 35], the principal factors that affect the rate and severity of external sulfate attack are:
  • Aluminate (C3A) content (C3A ≤ 3),

  • concentration of sulfate ions in the environment and the type of cation associated to sulfate ions,

  • portlandite content.

In an environment with high sulfate contents, the hydroxide ions of portlandite (Ca (OH)2=CH) are substituted by sulfate ions to form the secondary gypsum (how is different that the gypsum used as a setting regulator and added at the moment of milling clinker in Portland cement manufacturing) as described in following reaction in the case of Na2SO4 (N \(\bar{S}\)) solution:
$${\text{CH}} + {\text{N}}\bar{S} + 2 {\text{H}} \to {\text{C}}\bar{S}{\text{H}}_{ 2} + {\text{NH}} .$$
Then this formed gypsum is reacted with aluminate to produce the secondary ettringite, as described in reaction:
$${\text{C}}_{ 3} {\text{A}} + 3 {\text{C}}\bar{S}{\text{H}}_{ 2} + 2 6 {\text{H}} \to {\text{C}}_{ 6} {\text{A}}\bar{S}_{ 3} {\text{H}}_{ 3 2} .$$

The name “secondary” is used to designed the ettringite generated by external sulfate attack and to distinguish it from the primary ettringite that is formed during curing of Portland cement when the mortar paste is in plastic phase.

In sulfate magnesium (M \(\bar{S}\)) solution, the attack mechanism is described in these sequential reactions:
$${\text{CH}} + {\text{M}}\bar{S} + 2 {\text{H}} \to {\text{C}}\bar{S}{\text{H}}_{ 2} + {\text{MH,}}$$
$${\text{C}}_{ 3} {\text{A}} + 3 {\text{C}}\bar{S}{\text{H}}_{ 2} + 2 6 {\text{H}} \to {\text{C}}_{ 6} {\text{A}}\bar{S}_{ 3} {\text{H}}_{ 3 2} ,$$
$${\text{C}} - {\text{S}} - {\text{H}} + {\text{M}}\bar{S} + 2 {\text{H}} \to {\text{C}}\bar{S}{\text{H}}_{ 2} + {\text{M}} - {\text{S}} - {\text{H}} .$$

This mode of attack shows also the formation of secondary ettringite like the attack by sulfate sodium, but it leads to decalcification of C–S–H to generate M–S–H.

From this bibliographic background it is clear that the portlandite (CH) content is plays a significant role on the sulfate attack mechanism: so to mitigate as much as possible this type of chemical aggression, it is better to use a resisting sulfate portlandite cement which produces a minimum of portlandite amount in the system and containing a lower quantity of aluminate phase. Regarding the portlandite amount generated from all cement pastes at early age (1, 2, 7 and 14 days), cement A seems that it has the most resistive character against the external sulfate attack than the other studied specimens, since it produce a minimum amount of portlandite and furthermore its content of C3A (1.4%) is weak.

In order to determine the chemically bound water and the hydration degree of different cement pastes, the Mounanga [22] method was used in detail in this paper. The non evaporable water (Wt) was estimated from the mass loss between 145 °C and 1000 °C by Taylor [36] and Loukili [37] when studing cement pastes without carbonate. Table 4 presents the values (\(\Delta {\text{m 145}} \to 1 0 0 0 ^ {{\circ\, {\rm C}}}\)), (\(\Delta {\text{m 600}} \to 8 0 0 ^ {{\circ {\rm C}}}\)) and the percentage of the chemical bound water (%Wt) estimated from Eq. (2).
Table 4

\(\Delta {\mathbf{m}} 1 4 5\to 1 0 0 0 ^ {{\circ {\rm C}}}\), \(\Delta {\mathbf{m}} 6 0 0\to 8 0 0 ^ {{\circ {\rm C}}}\) and Wt calculated using Mounanga method


Age (days)

\(\Delta {\text{m 145}} \to 1 0 0 0 ^ {{\circ}}{\rm C}\) (mg)

\(\Delta {\text{m 600}} \to 8 0 0 ^ {{\circ}} {\rm C}\) (mg)


Cement A

















Cement B

















Cement C

















As shown in Table 4, cement A presents the lower content of the chemical bound water than the other cement pastes, this results is related to the CaCO3 content (6.21%) in this cement sample since it is higher than 5%. This finding is confirmed by the research work of El-Roudi el al [25].

The hydration degree of cement samples calculated using the Eq. 1 was presented in Fig. 8. It can be noticed that in beginning of hydration (1 day), all cement pastes presented the same degree of hydration (29.26, 29.88 and 28.77%). Then after 2 days, degree of hydration of B and C cement pastes were similar (40%) and higher than the one of cement A (28.41%). From 7 to 14 days, cement pastes containing the lower calcite content (1.82%) exhibited an enhancement from 50.37 to 54.81% of degree of hydration. The same manner for the cement B pastes, which increased from 42.87 to 50.32%. These results confirmed the previous works including the influence of the calcium carbonate on the hydration degree of Portland cement [38]. So it is clear that more limestone addition is the more the hydration kinetics will be delayed.
Fig. 8

Hydration degree assessment of A, B and C cement pastes calculated using Mounanga method

From the degree of hydration, we deduced the porosity by applying this following expression proposed by Powers and Brownyard [39]:
$$\varPhi \left( {\text{t}} \right) = \frac{{\rho \times \left( {\frac{W}{C}} \right) - (1.15 + 0.06 \times \rho ) \times \alpha (t)}}{{1 + \rho \times (\frac{W}{C})}}, \, \left( 6\right)\ {\text{While }}\rho \;{\text{is the density of cement }}\left( {{\text{g}}/{\text{cm}}^{ 3} } \right).$$
Figure 9 provides the porosity variation as a function of hydration time for cement A, B and C pastes calculated using Eq. 6. It can be noticed that the porosity of cement A pastes which containing about 6% of calcite is higher than that of the other cement samples. Although the pastes containing a lower amount of calcium carbonate (1.82%) present the lower porosity after 7 and 14 days of hydration. This observation does not agree with previous works [6] which say that the CaCO3 particles during hydration reduced the total pore volume, as a result the total porosity decreases.
Fig. 9

Porosity variation of cement pastes

11 Isothermal Calorimetry

Figure 10 presents the heat evolution rate curves for cements (A, B and C). There are some differences in intensity, but all the heat flow curves present the main stages of hydration [31] as shown in Fig. 10, the first and very intense heat flow(I) is due to the C3A and C3S, The second heat flow (II) appears slowly during the monosulfoaluminate formation and C–S–H precipitation.
Fig. 10

Heat evolution rate curves recorded during hydration of cements

The rate of hydration of cement A is clearly much lower than that cement B and cement C (see Fig. 10), this curve in well concordance with the low content of Ca(OH)2 corresponding to this paste (cement A). It can be seen (Fig. 10) that the rate of heat generation of cement containing 6.21% CaCO3 shifted to the right and it joined than the other curves [6]. This indicated that cement A paste was slightly delayed compared to the other cement paste’s hydration kinetics. The maximum of heat release by cement A is notably lower than that of cements B and C, but the width at middle height is larger, so the surface of the three peaks which represent the heat released are not very far one to another.

Figure 11 presents the total cumulative heat Q (J/g) which presented the enthalpy of each cement hydration process for cement pastes. The results show that cement C has the highest cumulative heat (187 J/g), this sample contains 1.82% of CaCO3. Then we found cement B (Q = 159 J/g) which contains 2.10% and finally cement A with cumulative heat equal to 152 J/g which contain of 6.21% of CaCO3. The three curves are constantly disposed in this order. We conclude that the increase of CaCO3 addition decreases always the cumulative heat and then the rate of hydration [38].
Fig. 11

Evolution of total cumulative heat Q (J/g) in function of the time

Figure 12 represents the evolution of hydration degree calculated as mentioned in Sect.~7.2 II-7-2 and using the Eq. 5. We found that the curves are placed in the same order than the Fig. 11 this confirms the same conclusions. It can be noticed that the hydration degree of the cement C was rapidly increasing. However, the evolution of hydration degree of cement A and cement C slowed down. After 200 h, the degree of hydration of cement A, cement B and cement C were found to be 48.55, 45.47 and 42.24%, respectively. All the results proved that the carbonate calcium presence in cement pastes delayed the hydration process. So, the degree of hydration decreases with an increase of calcite replacement amounts.
Fig. 12

Degree of hydration of cement from Isothermal calorimetry for three cements pastes (w/c = 0.4)

There are correspondence between the hydration degree of cement samples determined from ATG and those estimated by using isothermal calorimetric.

12 Durability of SR Portland Cements Under External Sulfate Attack (Na2SO4 and MgSO4)

12.1 Compressive Strength

The results of mechanical compressive strength of mortars of different cements immersed during 2 years into the following solutions, 10 g/L Na2SO4, 10 g/L MgSO4 and tap water are presented in Fig. 13.
Fig. 13

Mechanical compressive strength of mortars immersed in a 10 g/L Na2SO4, b 10 g/L MgSO4 and c tap water

Figure 13c show that the compressive resistance of samples immersed in tap water during 2 years increase continuously over all this period. We note that the compressive resistances have not yet reached stable values. The relative positions of curves are: C presents the better resistance then B then A. This order is the same of the calcite content, which is 1.8; 2.1 and 1% for C, B and A, respectively.

The results of compressive strength tests of all cement mortars (Fig. 13a and b) are presented the same tendency evolution: the mechanical compressive strength increase up to 90 days then it present a drop which contain up to 2 years of immersion in sulfate solutions.

The rise of compressive strength in the first three months as shown in Fig. 13.(a) and Fig. 13.(b) can be explained by the fact to the start of gypsum and ettringite (see reactions (7), (8), (9), (10) et (11)) growth in the system, that fill the hollow pores in the mortar structures and make it more denser, in this phase the mortar is still unaffected. The mechanical compressive strength decrease of all mortars exposed in sulfate solutions from 210 days is automatically due to the development of expansive products “ettringite and gypsum” which caused the destruction, expansion and cracking of mortars [40].

Moreover the measurement results of the compressive strength show the influence of the nature of the cation (Na+, Mg2+) associated to the sulfate ion: the mechanical strengths of all studied mortars stored magnesium sulfate solution were lower than those immersed in sodium sulfate medium. Hence the MgSO4 environment is more aggressive than the Na2SO4 medium. This finding is confirmed by the previous works of Neville [41] and Xiong [33] who mentioned that the external magnesium sulfate attack of concretes is the fastest and the most severe and it leads to a decline in compressive strength as indicated by our results. It is related directly to the formation of hydrates magnesium silicates (M–S–H) as described in Eq. 11, this product has not a binder aspect and leads to the loss of the mechanical properties of concrete, especially the compressive strength [41] as shown in our study.

The cement A based mortars developed the weakest mechanical compressive strength in the different studied environments (Tap water, 10 g/L Na2SO4 and 10 g/L MgSO4) compared to the other specimens, while cement C showed the highest strength as indicated in Fig. 13.

12.2 Visual Inspection

A visual evaluation of the mortars of each specimen was made after 2 years (720 days) of immersion by removing the mortars from the solutions and photographing them.

A special note was made of any change in the color, crumbling and precipitation of all materials.

After 2 years of prismatic mortars storage in 10 g/L of sodium sulfate solution (Fig. 14), we noticed only the appearance of microcracks at the faces of the cement A mortar. In addition there is degradation at the corners and edges in all cement mortars.
Fig. 14

Photographs of mortar prisms immersed for 720 days (2 years) in 10 g/L MgSO4 and g/L Na2SO4 at 22°C

The mortars exposed to 10 g/L MgSO4 solution showed a deposit of a white coloring layer on the mortars with degradation of the corners and edges in all the samples as shown in Fig. 14. The appearance of the white coating is related probably to the secondary gypsum formation.

This visual examination showed that the degree of deterioration of A cement mortar is very important compared to the other specimens.

13 Mineralogical Identification of Deterioration Products

13.1 XRD Analysis

The X-ray diffraction analysis of the mortars exposed to external sulfate attack (10 g/L Na2SO4 and 10 g/L MgSO4) allowed identifying and positioning the different crystalline phases encountered in the altered materials.

Figures 15 and 16 show the superposition of the XRD patterns of A, B and C mortars stored during 2 years in 10 g/L Na2SO4 and 10 g/L MgSO4 at 22 °C over the range from 4° to 27° 2θ.
Fig. 15

XRD patterns of a, b and c mortars immersed in 10 g/L Na2SO4 solution after 2 years at 22 °C

Fig. 16

XRD patterns of a, b and c mortars immersed in 10 g/L MgSO4 solution after 2 years at 22 °C

The XRD patterns of all mortars revealed the presence of expansive products such as secondary gypsum (2θ = 11.589°/PDF number: 00-033-0311), secondary ettringite (2θ = 9.091°, 15.784°/PDF number: 00-041-1451) and thaumasite (2θ = 9.214°, 16.014°/PDF number; 00-046-1360) in both sulfate solutions (magnesium sulfate and sodium sulfate).

This latter is probably detected only in cement A mortars in both aggressive environments (10 g/L Na2SO4 and 10 g/L MgSO4). The thaumasite product is the result of two possible mechanisms:
  • From the secondary ettringite (Ca6Al2[(OH)4SO4]3.26 H2O) by the substitution of the aluminate ions Al3+ with silicate ions Si4+ and the interstitial replacement of [(SO42−)3(H2O)2] by [(SO42−)3(CO32−)2], so thaumasite formation [42]: Ca6[Si(OH)6]2(CO3)2(SO4)2.24 H2O=CaCO3·CaSO4·CaSiO3·15H2O=C3\(C\bar{S}\) SH15.

  • The second way to obtain the thaumasite product, an interaction between sulfates ions, carbonates and the C–S–H gel as described in followed reaction [43]:

$${\text{C}} \bar{S}{\text{H}}_{ 2} \, + \,{\text{C}} \bar{C} + {\text{C}}{-}{\text{S}}{-}{\text{H}} + 1 2 {\text{H}} \to {\text{C}}_{ 3} CS{\text{SH}}_{ 1 5}$$

It is known that the thaumasite compound precipitation is favored in cold temperature below 15 °C [44], but there are another studies which found this product at 20 °C [45, 46].

XRD diffractograms show that the main peaks of ettringite and thaumasite are very weak and have very close reflections peaks (see Figs. 15 and 16). That’s why other technique is required to distinguish thaumasite compound.

The eventuality to detect the presence of thaumasite in cement A mortars is very large for the reason that the cement A containing 6.21% of calcium carbonate (see Table 1).

13.2 Infrared Spectroscopy

In this section the purpose of the use of infrared spectroscopy analysis is to prove the existence or the absence of thaumasite in the altered mortars.

The IR spectra of the cement A, cement B and cement C mortars prisms stored in 10 g/L Na2SO4 and 10 g/L MgSO4 for 720 days at 22 °C are presented in Fig. 17.
Fig. 17

IR spectra of A, B and C mortars immersed in 10 g/L Na2SO4 and 10 g/L MgSO4 for 2 years at 22 °C

As shown in Fig. 17, the thaumasite presence was confirmed only in cement A mortars stored in sodium and magnesium solutions by the detection of bands of SiO6 group vibration at 500, 673 and 750 cm−1. The octahedral silicon (SiO6) is very infrequent in mineral silicates which its band in IR spectrum is indicative the thaumasite compound must be present [43].

Moreover the IR results mortars revealed the peaks at around 600 and 1120 cm−1 that indicated the SO42−vibration groups in all studied mortars attributed exactly to secondary gypsum and harmful secondary ettringite [47]. All these observations, taken together with XRD results of the previous section, strongly suggest that the cement A based mortars contain thaumasite, ettringite and gypsum although the cement B and C mortars have only the secondary gypsum and ettringite. This finding is related directly to the presence of an important amount of calcite (6.21%) in cement A compared to the other cement samples (see Table 1) which concerned an essential condition to form thaumasite.

14 Conclusion

The main conclusions could be derived from this investigation are summarized as follows:
  • The rate of hydration of the resisting sulfate Portland cement is clearly influenced by the limestone addition amounts.

  • By means FTIR analysis, the monocarboaluminate calcium compound was detected at 14th day of hydration in cement paste containing more than 5% of calcite: it is the chemical effect of limestone.

  • The addition of high amount of calcite (> 5%) delayed the hydration kinetics of the RS Portland cement at short term hydration.

  • Thaumasite formation was observed into mortars stored in both sulfates solutions (10 g/L Na2SO4 and 10 g/L MgSO4) after 2 years at 22 °C. This precipitate occurs only into the cement relatively rich with calcium carbonate (6%) with 2.1 and 1.8% there is no thaumasite formation.

  • It is clear that calcium carbonate is harmful for sulfate resisting Portland cements.


  1. 1.
    En N (2001) 197–1-Ciment—Partie 1: composition, spécifications et critères de conformité des ciments courants. AFNOR, ParisGoogle Scholar
  2. 2.
    Gollop R, Taylor H (1995) Microstructural and microanalytical studies of sulfate attack III Sulfate-resisting portland cement: Reactions with sodium and magnesium sulfate solutions. Cem Concr Res 25(7):1581–1590CrossRefGoogle Scholar
  3. 3.
    IslemLabidi H (2017) Adel Megriche: feasibility of efficient RS Portland cement from An optimization combination of raw materials. J Tunis Chem Soc 19:294–300Google Scholar
  4. 4.
    Sorrentino F (2011) Chemistry and engineering of the production process: state of the art. Cem Concr Res 41(7):616–623CrossRefGoogle Scholar
  5. 5.
    Chatterjee AK (2011) Chemistry and engineering of the clinkerization process—Incremental advances and lack of breakthroughs. Cem Concr Res 41(7):624–641CrossRefGoogle Scholar
  6. 6.
    Hawkins P, Tennis P, Detwiler R (2003) The Use of Limestone in Portland Cement: a State-of-the-Art Review. Portland Cement Association, Skokie, Illinois, USAGoogle Scholar
  7. 7.
    Bonavetti V, Donza H, Rahhal V, Irassar E (2000) Influence of initial curing on the properties of concrete containing limestone blended cement. Cem Concr Res 30(5):703–708CrossRefGoogle Scholar
  8. 8.
    Kakali G, Tsivilis S, Aggeli E, Bati M (2000) Hydration products of C3A, C3S and Portland cement in the presence of CaCO3. Cem Concr Res 30(7):1073–1077CrossRefGoogle Scholar
  9. 9.
     Schmidt T (2007) Sulfate attack and the role of internal carbonate on the formation of thaumasite. Thesis EPFL, Lausanne, SwitzerlandGoogle Scholar
  10. 10.
    Schmidt T, Lothenbach B, Romer M, Neuenschwander J, Scrivener K (2009) Physical and microstructural aspects of sulfate attack on ordinary and limestone blended Portland cements. Cem Concr Res 39(12):1111–1121CrossRefGoogle Scholar
  11. 11.
    Collepardi M (1999) Thaumasite formation and deterioration in historic buildings. Cement Concr Compos 21(2):147–154CrossRefGoogle Scholar
  12. 12.
    Juel I, Herfort D, Gollop R, Konnerup-Madsen J, Jakobsen HJ, Skibsted J (2003) A thermodynamic model for predicting the stability of thaumasite. Cem Concr Compos 25(8):867–872CrossRefGoogle Scholar
  13. 13.
    Crammond N (2002) The occurrence of thaumasite in modern construction—a review. Cem Concr Compos 24(3–4):393–402CrossRefGoogle Scholar
  14. 14.
    Tsivilis S, Sotiriadis K, Skaropoulou A (2007) Thaumasite form of sulfate attack (TSA) in limestone cement pastes. J Eur Ceram Soc 27(2):1711–1714CrossRefGoogle Scholar
  15. 15.
    Irassar E, Gonzalez M, Rahhal V (2000) Sulphate resistance of type V cements with limestone filler and natural pozzolana. Cem Concr Compos 22(5):361–368CrossRefGoogle Scholar
  16. 16.
    Al-Amoudi OSB (2002) Attack on plain and blended cements exposed to aggressive sulfate environments. Cem Concr Compos 24(3):305–316CrossRefGoogle Scholar
  17. 17.
    Islem Labidi SB, Houcine T, Adel M (2018) Critical research study of quantification methods of mineralogical phases in cementitious materials. J Aust Ceram Soc 34:1541–1547Google Scholar
  18. 18.
    EN N (2006) 196-1, Méthodes d’essais des ciments-Partie 1: détermination des résistances mécaniques. French Standard 2006Google Scholar
  19. 19.
    Khelifa MR (2009) Effet de l’attaque sulfatique externe sur la durabilité des bétons autoplaçants Thèse de Doctorat, U. de Constantine et U. d’OrléansGoogle Scholar
  20. 20.
    Ismail S, Sashidhar C, Kumar DP (2015) Effect of sulphates in the presence of water on the strength characteristics of blended cement. Int Res J Eng Technol 2(5):1228–1234Google Scholar
  21. 21.
    Mounanga P (2003) Experimental study of the behavior of cement pastes at very young age: hydration, shrinkage, thermophysical properties. Adv Cem Res 16:95–103CrossRefGoogle Scholar
  22. 22.
    Mounanga P (2003) Etude expérimentale du comportement de pâtes de ciment au très jeune âge: hydratation, retraits, propriétés thermophysiques.Thèse de Doctorat, NantesGoogle Scholar
  23. 23.
    Kocaba V (2009) Development and evaluation of methods to follow microstructural development of cementitious systems including slags. Thesis, EPFL, LausanneGoogle Scholar
  24. 24.
    Trezza M, Lavat A (2001) Analysis of the system 3CaO· Al2O3–CaSO4 2H2O–CaCO3–H2O by FT–IR spectroscopy. Cem Concr Res 31(6):869–872CrossRefGoogle Scholar
  25. 25.
    El-Roudi A, Balboul BA, Abdelzaher M (2015) Influence of limestone on physico-chemical properties of white portland cement pastes. Int J 3(6):210–221Google Scholar
  26. 26.
    Fernández-Carrasco L,  Torréns-Martín D,  Morales LM, Martinez-Ramírez S (2012) Infrared spectroscopy in the analysis of building and construction materials. In: Theophanides T (ed) Infrared Spectroscopy — Materials Science, Engineering and Technology, INTECH, Rijeka, Croatia, pp. 369–382Google Scholar
  27. 27.
    Horgnies M, Chen J, Bouillon C (2013) Overview about the use of Fourier transform infrared spectroscopy to study cementitious materials. WIT Trans Eng Sci 77:251–262CrossRefGoogle Scholar
  28. 28.
    Collier NC (2016) Transition and decomposition temperatures of cement phases—a collection of thermal analysis data. Ceram Silik 60(4):338–343Google Scholar
  29. 29.
    Gabrovšek R, Vuk T, Kaučič V (2006) Evaluation of the hydration of Portland cement containing various carbonates by means of thermal analysis. Acta Chim Slov 53:159–165Google Scholar
  30. 30.
    Boualleg S, Bencheikh M, Belagraa L, Daoudi A, Chikouche MA (2017) The combined effect of the initial cure and the type of cement on the natural carbonation, the portlandite content, and nonevaporable water in blended cement. Adv Mater Sci Eng 2017:17CrossRefGoogle Scholar
  31. 31.
    Bullard JW, Jennings HM, Livingston RA, Nonat A, Scherer GW, Schweitzer JS, Scrivener KL, Thomas JJ (2011) Mechanisms of cement hydration. Cem Concr Res 41(12):1208–1223CrossRefGoogle Scholar
  32. 32.
    Santhanam M, Cohen MD, Olek J (2003) Mechanism of sulfate attack: a fresh look: part 2. Proposed mechanisms. Cem Concr Res 33(3):341–346CrossRefGoogle Scholar
  33. 33.
    Xiong C, Jiang L, Song Z, Liu R, You L, Chu H (2014) Influence of cation type on deterioration process of cement paste in sulfate environment. Constr Build Mater 71:158–166CrossRefGoogle Scholar
  34. 34.
    Shanahan N, Zayed A (2007) Cement composition and sulfate attack: part I. Cem Concr Res 37(4):618–623CrossRefGoogle Scholar
  35. 35.
    Cao H, Bucea L, Ray A, Yozghatlian S (1997) The effect of cement composition and pH of environment on sulfate resistance of Portland cements and blended cements. Cem Concr Compos 19(2):161–171CrossRefGoogle Scholar
  36. 36.
    Taylor H (1990) Proposed mechanisms. Academic Press, LondonGoogle Scholar
  37. 37.
    Loukili A (1996) Etude du retrait et du fluage de Bétons à Ultra-Hautes Performances. Thèse de Doctorat, Université de  NantesGoogle Scholar
  38. 38.
    Deboucha W, Leklou N, Khelidj A, Oudjit MN (2017) Hydration development of mineral additives blended cement using thermogravimetric analysis (TGA): methodology of calculating the degree of hydration. Constr Build Mater 146:687–701CrossRefGoogle Scholar
  39. 39.
    Powers TC, Brownyard TL (1943) Studies of the physical properties of hardened Portland cement paste. J Proc 1946:101–132Google Scholar
  40. 40.
    Yu Q, Sawayama K, Sugita S, Shoya M, Isojima Y (1999) The reaction between rice husk ash and Ca (OH) 2 solution and the nature of its product. Cem Concr Res 29(1):37–43CrossRefGoogle Scholar
  41. 41.
    Neville A (2004) The confused world of sulfate attack on concrete. Cem Concr Res 34(8):1275–1296CrossRefGoogle Scholar
  42. 42.
    Ramachandran VS, Paroli RM, Beaudoin JJ, Delgado AH (2002)  Handbook of Thermal Analysis of Construction Materials, Noyes Publications, New York, USAGoogle Scholar
  43. 43.
    Torres S, Sharp J, Swamy R, Lynsdale C, Huntley S (2003) Long term durability of Portland-limestone cement mortars exposed to magnesium sulfate attack. Cement Concr Compos 25(8):947–954CrossRefGoogle Scholar
  44. 44.
    Higgins D, Crammond N (2003) Resistance of concrete containing ggbs to the thaumasite form of sulfate attack. Cem Concr Compos 25(8):921–929CrossRefGoogle Scholar
  45. 45.
    Hartshorn S, Sharp J, Swamy R (2002) The thaumasite form of sulfate attack in Portland-limestone cement mortars stored in magnesium sulfate solution. Cem Concr Compos 24(3):351–359CrossRefGoogle Scholar
  46. 46.
    Santhanam M, Cohen M, Olek J (2006) Differentiating seawater and groundwater sulfate attack in Portland cement mortars. Cem Concr Res 36(12):2132–2137CrossRefGoogle Scholar
  47. 47.
    Abdalkader A (2014) Thaumasite sulfate attack in cement mortars exposed to sulfate and chloride and implications to rebar corrosion. Thesis University of Sheffield, SheffieldGoogle Scholar

Copyright information

© The Tunisian Chemical Society and Springer Nature Switzerland AG 2019

Authors and Affiliations

  • Islem Labidi
    • 1
    Email author
  • Sonia Boughanmi
    • 1
  • Abdelhafid Khelidj
    • 2
  • Mohamed El Maaoui
    • 1
  • Adel Megriche
    • 1
  1. 1.Laboratory of Applied Mineral Chemistry (UR11ES18), Department of chemistry, Faculty of Sciences of TunisTunis El Manar UniversityTunisTunisia
  2. 2.GeM (UMR CNRS 6183) Institut de Recherche en Génie Civil et Mécanique IUT de Saint-NazaireSaint-Nazaire CedexFrance

Personalised recommendations