Active low-energy belite cement

Cement and Concrete Research 68 (2015) 203–210

Contents lists available at ScienceDirect

Cement and Concrete Research journal homepage: http://ees.elsevier.com/CEMCON/default.asp

Active low-energy belite cement Theodor Staněk a, Petr Sulovský b,⁎ a b

Research Institute for Building Materials, JSC, Brno, Czech Republic Dept. of Geology, Faculty of Science, Palacky University, Olomouc, Czech Republic

a r t i c l e

i n f o

Article history: Received 14 October 2013 Accepted 6 November 2014 Available online 21 December 2014 Keywords: Cement (D) Ca2SiO4 (D) Sulfate (D) Physical Properties (C) WDS

a b s t r a c t The term ‘low-energy cement’ is used for cements that can be in some applications used instead of OPC, and which are produced with less energy. A more extensive utilization of these cements would lead to the lowering of expenses on production of binders as well as a reduction of undesirable emissions. The belite-rich cement belongs to this group. Pure belite clinkers with interstitial matter consisting of C3A and C4AF have not been produced, as they have insufficient strength. This work describes the results of hydraulic activation of belite-rich clinkers with sulfate anions. The principle of activation is used for the preparation of belite-rich clinkers with an increased Ca:Si ratio in the structure of dicalcium silicate and partial substitution of SiO4− 4 by SO2− 4 . Cements, prepared from these belite-rich clinkers, containing up to 20% of alite, which are burned at 1350 °C, have the same technological properties, including early strengths, as OPC. © 2014 Published by Elsevier Ltd.

1. Introduction The research and the production of hydraulically active low-energy cements, especially those based on belite-rich clinkers, are very topical and for the future of cementitious binders highly prospective. It is expected to be one of the main directions in the development of the world cement industry. Mass production of low-energy cements would mean a considerable decrease in the total expense of cement production when compared to the current common Portland cement with high alite content. It would also mean an overall decrease in the environmental impact of cement production. This attenuated impact can be seen rather in the decrease of CO2 emissions owing to the composition of the raw meal for the production of low-energy clinkers, as they require less CaO and thus less consumption of CaCO3, than in reduction temperature of burning by 100 –300 °C. This approach supports sustainable development of high-quality raw materials – notably pure limestones – by facilitating the use of resources of a lesser grade with lower CaCO3/higher impurity content. It also encourages the (re-)utilization of industrial by-products and/or wastes. Belite in the common Portland clinker has considerably lower hydraulic activity than alite [1] and contributes significantly to strengths only after 28 days of hydration. This led to efforts to stabilize hydraulically active forms of belite, specifically related to high-temperature modifications. A chemical stabilization by suitable admixtures, usually complemented by fast quenching [2–5] is one of the possibilities. A newer method of belite hydraulic activation is a utilization of the

⁎ Corresponding author. E-mail address: [email protected] (P. Sulovský).

http://dx.doi.org/10.1016/j.cemconres.2014.11.004 0008-8846/© 2014 Published by Elsevier Ltd.

‘remelting reaction’ [6,7] or ‘sol-gel method’ [8,9]. Such methods can be realized only under conditions outside the possibilities of currently used industrial technologies. The research of the mechanism and kinetics of belite clinker formation has shown, that a quickly formed belite clinker has, in contrary to original expectations, lower hydraulic activity than a longer burned, recrystallized belite clinker [10]. Production of sulfoaluminate-belite (SAB) cement is performed to a limited extent, as these cements show appropriate properties [11,12]. Experiments with industrial production of the sulfo-aluminate-ferrite belite clinker (SAFB) and the high-Fe belite clinker (HFBC) gave cements with satisfactory strengths after 28 days of hydration, but low short-term strengths [13]. China has made great progress in testing such technologies with industrial production of SAB and also fluoraluminate belite cement and high belite Portland cement with 20–30 wt.% alite [14]. The main problem to be solved is the production of hydraulically active belite cement, with properties like those of standard “alite” Portland cement, using the existing production lines. The principle of activation is the formation of a belite clinker with an increased ratio of CaO:SiO2 in the structure of dicalcium silicate by substitution of SiO4 groups by SO4 groups. SO3 considerably decreases the viscosity and surface tension of the clinker melt [15,16]. It postpones the start of alite formation and decreases its nucleation and its overall content in the clinker. It can completely block alite formation [17–19]. Beyond this, it supports the formation of belite and enters in quantities into the belite structure. This fact can be used in the preparation of relatively high saturated belite clinkers. A relatively high saturation of a belite clinker (high content of CaO in the belite structure) ensures in the ensuing reaction with water the formation of increased portlandite (Ca(OH)2) and an overall alkalinity, accelerating the course of its hydration.

204

T. Staněk, P. Sulovský / Cement and Concrete Research 68 (2015) 203–210

In practice, waste desulfurization products and fluidized bed fly ash can be used in the burning of the ‘sulfobelite’ clinker. This clinker, burnt at lower temperatures, would not contain a greater extent of C4 A3 S calciumsulfoaluminate (Klein's complex, yeelimite) — in comparison with SAB cement.

2. Material and methods The raw materials used in the preparation of experimental raw meals were pure limestone with 97.5 wt.% CaCO3 (L1), limestone with increased SiO2 content — 30 wt.% (L2), clay shale with 41.5 wt.% SiO2 and 12.5 wt.% Al2O3 (CS), pure quartz with 99.5 wt.% SiO2 (Q), roasted pyrite as iron correction (Fe), and flue-gas desulfurization (FGD) gypsum. The composition and expected chemical parameters of the raw meals are given in Tables 1–3. The planned experiments involved the burning of belite clinkers with 4, 5, 6 and 8 wt.% of SO3 in the raw meal and a different lime saturation index, the referenced lowsaturation belite clinker, and referenced alite clinker without SO3 addition. Six kilograms of each raw meal mixture were ground until ~12 wt.% fraction passed b 0.09 mm. The raw meals were pressed into tablets 4 cm in diameter, weighing about 80 g. All belite clinkers were burned in a super-kanthal furnace in the following regime: rate of temperature rise 15 °C/min, final temperature 1350 °C or 1400 °C, and isothermal dwell time 30–60 min. The alite clinker was burned at 1450 °C for 120 min. Representative amounts of the clinkers thus produced were comminuted to b1 mm and sieved. Polished epoxy mounts of fractions 0.045– 1 mm were etched in fumes of acetic acid for reliable phase identification [20]. Modal content of constituent phases was determined using optical microscopy and point counting of 2000 points per section [21]. For the recalculation from volume to weight percentages, the following densities were used: C3S — 3.15, C2S — 3.28, C3A — 3.03, C4AF — 3.77, and free CaO — 3.35 g.cm−3. After re-polishing, the sections were analyzed with an electron microprobe (CAMECA SX100 with 5 crystal spectrometers) to determine the chemical composition of the clinker minerals. The quantitative spot analyses were performed under the following analytical conditions: accelerating voltage 15 kV, probe current 20 nA, and spot size 2 μm. The standards used for internal calibration were natural standards — wollastonite (Ca, Si), barite (S), orthoclase (K, Al), andradite (Fe), jadeite (Na), rhodonite (Mn), pyrope (Mg), hydroxylapatite (P), and synthetic TiO (Ti); all elements were analyzed on their Kα lines. The lower limits of detection varied between ~0.01 wt.% (e.g. for P, S Si, Al, Ca) and ~0.02 wt.% (e.g. Ca, K, Mn, Fe). The acquired counts, corrected for background, were recalculated to oxides using the PAP correction procedure included in the CAMECA PeakSight automation program. The EMPA results in wt.% oxides were recast into composition formulae in atoms per formula unit (apfu) according to the method described in [22].

Table 1 The composition of reference raw meals without the addition of SO3 in wt.% and expected basic chemical parameters. Raw meal

B75S0

B81S0

A95S0

A97S0

L1 L2 CS Q Fe EG LSF SR AR

55.27 19.59 23.72 – 1.42 – 75 2.6 1.4

69.70 – 22.56 6.74 1.00 – 81 3.2 1.5

64.41 14.10 20.62 – 0.87 – 95 2.6 1.6

71.83 – 23.40 3.73 1.04 – 97 2.6 1.5

Table 2 The composition of raw meals with the addition of 4 and 5 wt.% SO3 in wt.% and expected basic chemical parameters. Raw meal

B86S4

B80S5

B85S5

B89S5

B89.5S5

B90S5

L1 L2 CS Q Fe EG LSF SR AR

66.20 – 21.02 6.30 0.89 5.59 86 3.2 1.5

48.82 18.92 22.96 – 1.37 7.92 80 2.6 1.4

52.74 17.51 20.57 – 1.26 7.92 85 2.6 1.4

65.22 – 21.29 5.61 0.89 6.99 89 3.2 1.5

65.35 – 21.19 5.59 0.88 6.99 89 3.2 1.5

65.48 – 21.09 5.56 0.88 6.99 90 3.2 1.5

Qualitative determination of sulfate phases was done by powder X-ray diffraction of the clinkers with a Bruker D8 Advance diffractometer with variable slits and a position sensitive detector. The diffractometer was operated at 40 kV and 30 mA, using Ni-filtered CuKα of λ = 1,541,718 Ǻ, recording 6–80° in 0.02° 2Θ increments with 188 s counting time per step and total scan time of 1 h 5 m 59 s. β-C2S unit cell parameters were refined by the Rietveld method using the Topas 3 software. Finally, 2–4 wt.% of natural gypsum was added as a setting controller to all belite clinkers ground to produce cement with similar specific surface areas. These cements were then subjected to testing of their compressive and flexural strength and heat of hydration according to EN 196 standards.

3. Results and discussion 3.1. Composition of clinkers The result of the quantitative phase composition of all clinkers, the major oxides contents and the real basic chemical parameters (LSF, SR, AR) is given in Tables 4–6. These tables also contain further data about the clinker burning (temperature and burning time) and on the properties of cements prepared from them. The microstructure of the beliterich clinker doped with SO3 is exemplified by Fig. 1. The chemistry of belite grains was determined by electron microanalysis. The results are given in Tables 7–9. The charge sums for the cations and the framework elements are also included there. Compared with the composition of belite in common alite-rich clinkers, reported by various authors [23–25], belites from our experiments exhibited an elevated Ca/Si ratio (ranging from 2.4 to 2.6), that correspond to the partial substitution of Si by S. In common belites it is much closer to 2 (1.9 to 2.18). Herfort et al., who also experimented with burning raw meals enriched with sulfur [26], reported belites with atomic Ca/Si ratio between 2.37 and 2.42.

Table 3 The composition of raw meals with the addition of 6 and 8 wt.% SO3 in wt.% and expected basic chemical parameters. Raw meal

B89S6

B90S6

B91S6

B92S6

B96S6

B98S6

B92S8

L1 L2 CS Q Fe EG LSF SR AR

64.49 – 20.19 6.07 0.86 8.39 89 3.2 1.5

64.75 – 19.99 6.02 0.85 8.39 90 3.2 1.5

64.97 – 19.79 6.00 0.85 8.39 91 3.2 1.5

65.24 – 19.60 5.93 0.84 8.39 92 3.2 1.5

66.21 – 18.85 5.75 0.80 8.39 96 3.2 1.5

66.67 – 18.49 5.66 0.79 8.39 98 3.2 1.5

62.72 – 19.42 5.85 0.83 11.18 92 3.2 1.5

Where: LSF — lime saturation factor after Lea–Parker, SR — silica ratio, AR — alumina-iron ratio.

T. Staněk, P. Sulovský / Cement and Concrete Research 68 (2015) 203–210 Table 4 Composition of reference clinkers without the addition of SO3 and properties of cements prepared from these clinkers. Cement

B75S0

Phase C3 S C2 S C3 A C4AF C free Anhydrite Parameter SO3 SiO2 Al2O3 Fe2O3 CaO LSF SR AR Parameter Temperature Burning time Parameter Gypsum addition Density Specific surface Hydration time 2 days 7 days 28 days 90 days

Phase composition of used clinker [wt.%] 0.2 24.5 66.7 79.3 58.4 12.2 4.0 6.5 12.3 16.5 10.0 7.8 0.0 0.5 0.9 0.0 0.0 0.0 Chemical parameters of used clinker [wt.%, –] 0.06 0.07 0.03 27.03 25.11 22.63 5.49 4.70 4.80 3.90 3.68 2.77 59.65 62.50 66.23 70.4 79.9 93.5 2.88 3.00 2.99 1.41 1.28 1.73 Clinker burning parameters [°C, min] 1400 1350 1450 40 50 120 3 Cement parameters [wt.%, kg/m , m2/kg] 4.0 3.0 4.0 3230 3225 3179 435 397 437 Compressive strength [MPa] 2.0 4.5 16.6 2.3 12.8 44.7 14.6 63.5 66.3 38.3 81,1 69.5

B81S0

A95S0

A97S0 75.8 6.9 11.0 5.8 0.5 0.0 0.02 21.83 4.97 3.53 66.68 96.2 2.57 1.41 1450 120 4.0 3172 398 21.6 49.3 66.9 72,7

Calculations of atomic ratios of elements filling the Ca site (i.e. Ca, Mg, Na, K) versus elements filling the Si site (Si, S, P, Al, Fe3+, Ti) have shown that they are close to stoichiometric 2:1, assuming that all Mg resides in the Ca site and all Fe is trivalent and filling the Si site — see Tables 7–9. Charge calculations indicate only a negligibly small deficiency on the anions side. The average deficit from the suite of 9 clinkers

Table 5 Composition of clinkers with the addition of 4 and 5 wt.% SO3 and properties of cements prepared from these clinkers. Cement

B86S4

Phase C3 S C2 S C3 A C4AF C free

Phase composition of used clinker [wt.%] 17.2 0,0 8.3 13.8 16.6 66.9 81.3 71.5 71.3 70.3 1.6 2.6 4.1 1.0 2.1 12.0 14.6 13.6 11.1 8.8 2.3 0.0 0.0 1.7 1.4 0.0 1.5 2.5 1.1 0.8

CS Parameter SO3 SiO2 Al2O3 Fe2O3 CaO LSF SR AR Parameter Temperature Burning time Parameter Gypsum addition Density Specific surface Hydration time 2 days 7 days 28 days 90 days

B80S5

B85S5

B89S5

B89S5 II

B89.5S5

B90S5

20.6 63.5 1.5 12.7 1.1 0.6

23.8 59.1 3.3 11.6 1.5 0.7

Chemical parameters of used clinker [wt.%, –] 3.80 4.44 4.43 4.74 4.84 4.80 23.65 24.08 23.04 22.94 23.06 23.15 4.15 5.47 4.79 4.59 4.36 4.33 3.10 3.68 3.31 2.82 2.80 2.85 62.27 58.70 59.84 61.24 61.52 61.86 85.1 77.0 82.7 85,7 86.0 86.2 3.26 2.63 2.84 3.10 3.22 3.22 1.34 1.49 1.45 1.63 1.58 1.52 Clinker burning parameters [°C, min] 1350 1400 1400 1350 1350 1350 50 40 40 50 60 60 3 2 Cement parameters [wt.%, kg/m , m /kg] 3.0 4.0 4.0 2.0 2.0 2.0 3234 3261 3191 3239 3239 3249 397 436 438 407 399 401 Compressive strength [MPa] 14.4 1.8 11.0 20.0 24.3 20.9 34.6 12.2 24.2 40.9 47.1 45.3 57.1 47.4 51.1 58.4 64.4 63.1 64.0 60.5 65.8 66.9 70.4 71.0

205

Table 6 Composition of clinkers with the addition of 6 and 8 wt.% SO3 and properties of cements prepared from these clinkers. Cement

B89S6

Phase C3S C2S C3A C4AF C free

Phase composition of used clinker [wt.%] 17.0 20.0 16.1 33.4 28.6 65.2 63.8 67.4 50.5 55.7 3.5 5.6 0.8 4.0 1.2 10.1 8.0 10.8 8.7 9.6 1.6 0.8 1.9 0.8 2.6 2.6 1.8 3.0 2.6 2.3

CS Parameter SO3 SiO2 Al2O3 Fe2O3 CaO LSF SR AR Parameter Temperature Burning time Parameter Gypsum addition Density Specific surface Hydration time 2 days 7 days 28 days 90 days

B90S6

B91S6

B92S6

B96S6

Chemical parameters of used clinker [wt.%, –] 4.15 5.26 5.42 5.48 5.42 22.55 22.44 22.65 22.18 21.87 4.22 4.11 3.92 3.82 3.85 3.08 3.18 2.93 2.88 2.80 62.10 62.46 62.13 63.06 62.99 88.6 89.6 88.8 92.1 93.2 3.09 3.08 3.31 3.31 3.29 1.57 1.29 1.34 1.33 1.38 Clinker burning parameters [°C, min] 1350 1350 1350 1350 1350 50 60 50 60 60 Cement parameters [wt.%, kg/m3, m2/kg] 3.0 3.0 3.0 3.0 3.0 3201 3216 3209 3191 3206 406 477 353 410 350 Compressive strength [MPa] 20.1 26.2 15.8 31.7 18.0 39.4 50.7 27.5 51.6 26.7 59.1 65.7 49.1 64.7 44.0 67.6 70.1 67.0 76.2 68.7

B98S6

B92S8

39.0 43.6 2.3 8.6 2.7 3.8

15.4 66.8 3.1 12.3 0.4 2.0

5.50 21.28 3.52 2.68 63.57 97.1 3.43 1.31

6.97 22.45 4.26 2.95 60.73 87.0 3.11 1.44

1350 60

1350 50

3.0 3189 353

3.0 3210 402

19.8 28.9 43.7 60.4

16.9 26.1 36.4 64.9

presented in Tables 7–9 is only 1.2% of the bulk charge and is comparable to the analytical error due to both the analysis spot geometry and counting statistics. Accepting the assumption that iron is divalent presented in [26] would more than double this deficit. Correlation analysis of the average apfu values of the components forming the belite has shown several statistically significant (at the α = 0.05 level) correlations between certain elements. The significant positive correlation between Al and S (Pearson's r = 0.84) together with a strong negative correlation between Si on one side and Al and S on the other (Pearson's r being −0.83 and −0.91, respectively) indicates the existence of Si ↔ Al + S substitution (Table 10), reported in earlier studies [26–29]. Sodium shows a positive correlation to Si (r =

4.74 22.96 4.22 2.92 61.98 87.1 3.22 1.45 1350 60 2.0 3247 403 19.9 46.0 61.4 71.5

Fig. 1. Microstructure of belite-rich clinker S-B85S5. Rare alite forms euhedral crystals. Inclusion in alite crystal in the upper right corner is free lime. The remainder is made of oval belite grains and interstitial clinker matter.

206

T. Staněk, P. Sulovský / Cement and Concrete Research 68 (2015) 203–210

Table 7 Average chemical composition of belite grains in reference clinkers without the addition of SO3 determined by electron microprobe [wt.%]. Clinker

B75S0

SiO2 TiO2 Al2O3 Fe2O3 CaO MgO MnO SO3 K2O Na2O P2O5 Ca:Si (atomic) (Ca + K + Na + Mg): (Si + S + P + Al + Fe + Ti) (atomic)

32.01 0.21 0.96 1.14 63.23 0.38 0.05 0.06 0.79 0.26 0.09 2.116 2.037

B81S0 30.77 1.53 1.60 62,82 0.43 0.10 0.73 0.38 0.44 2.187 2.034

A95S0

A97S0

31.01 0.28 1.62 1.48 63.72 0.32 0.07 0.07 0.93 0.37 0.07 2.202 2.056

31.10 1.60 1.52 64.00 0.31 0.09 0.85 0.44 0.17 2.205 2.067

0.77), to K (r = 0.78), and Fe. There exists an inverse correlation between Fe and Al, indicating the probable concurrence of these elements in substitution for Si. As expected, both Na and K show a negative correlation with Ca, although significant only at α = 0.1, two-tail. Herfort et al. [27], presented results of similar experiments and reported a positive correlation between the fraction of sulfur occupying the tetrahedral sites and a silica/alkalis molar ratio. This relation was also observed in the samples prepared within this study (Fig. 2), as indicated by the correlation coefficient r = 0.756, significant on the confidence level of 0.05. Regression analysis of belite microanalyses has clearly disclosed the substitution of SiO4 groups by joint replacement with Al and SO4 (see Fig. 3), confirmed by the high value of the correlation coefficient

(0.843). Our results of belite analysis from most of the analyzed sulfobelite clinkers show the atomic ratio of Al/S to be close to 1.5:1 (1.414 on average). This corresponds more closely to the probable substitution 5[SiO4]4− → 3[AlO4]5− + 2[SO4]2− rather than the Al/S ratio of 2:1, reported in [28]. The overcharge on the left side of the above equation can be compensated by either vacancies or as suggested by Shimosaka [30], by involving calcium ions in the substitution: 6[SiO4]4− + 3Ca2+→ 4[SO4]2− + 2[AlO4]5− + 3□Ca. The substitution mechanism appears to be analogous to that revealed in the phosphorus entrance into the belite structure via the ‘berlinite substitution’ 2SiO4 ↔ AlPO4 [31]. The experiments with raw meals doped with various amounts of sulfur have shown that the fraction of sulfur atoms occupying the tetrahedral position is limited. The maximum value observed in individual belite grains was 0.11, and the average values did not exceed 0.085. This limit is probably set by the number of structural defects acceptable in the belite structure. Above that limit the structure would be destroyed due to the differences in the ionic radii of S, Al, and Si. Our results have shown that an increase in the lime saturation increases the content of SO3 and decreases SiO2 in belite. This results in a higher Ca:Si in belite, as has been previously reported [32]. The incorporation of SO3 into the belite structure, where it substitutes for SiO2, enables widening this substitution to other oxides, above all Al2O3. High-temperature microphotometry and high-temperature XRD studies have revealed that belites containing 2.4–3.2 wt.% SO3 in their

Table 8 Average chemical composition of belite grains in clinkers with the addition of 4 and 5 wt.% SO3 determined by electron microprobe [wt.%]. Clinker SiO2 TiO2 Al2O3 Fe2O3 CaO MgO MnO SO3 K2O Na2O P2O5 Total Ca:Si (atomic) (Ca + K + Na + Mg):(Si + S + P + Al + Fe + Ti) (atomic)

B86S4 28.22 2.48 1.68 64.09 0.70 2.75 0.22 0.27 0.19 100.60 2.43 2.04

B80S5

B85S5

B89S5

B89S5 II

B89.5S5

B90S5

30.56 0.22 1.81 1.17 63.74 0.25 0.06 1.61 0.10 0.11 0.10 99.73 2.24 1.97

27.70 0.20 2.65 1.69 62.65 0.59 0.07 2.98 0.14 0.15 0.09 98.91 2.42 1.98

28.17 0.05 2.34 1.43 64.40 0.46 0.01 2.50 0.17 0.22 0.19 99.94 2.45 2.06

27.89 n.d. 2.10 1.28 62.63 0.56 n.d. 2.56 0.21 0.29 0.19 97.71 2.41 2.06

28.49 0.28 2.33 1.28 65.08 0.56 0.05 2.57 0.16 0.21 0.17 101.18 2.45 2.06

25.23 n.d. 2.65 1.46 61.08 0.49 n.d. 7.85 0.33 0.23 0.17 99.49 2.59 1.89

Table 9 Average chemical composition of belite grains in clinkers with the addition of 6 and 8 wt.% SO3 determined by electron microprobe [wt.%]. Clinker

B89S6

B90S6

B91S6

B92S6

B96S6

B98S6

B92S8

SiO2 TiO2 Al2O3 Fe2O3 CaO MgO MnO SO3 K2O Na2O P2O5 CaO:SiO2 (molar) (Ca + K + Na + Mg):(Si + S + P + Al + Fe + Ti) (atomic)

27.31

27.04

26.77

26.79

26.83

27.02

27.53

2.60 1.60 64.76 0.36

2.79 1.72 64.75 0.62

2.69 1.68 64.20 0.49

2.76 1.69 63.92 0.46

2.81 1.53 63.26 0.41

2.62 1.69 64.36 0.45

2.70 1.57 64.45 0.51

3.33 0.21 0.18 0.19 2.541 2.057

3.17 0.13 0.12 0.16 2.566 2.070

3.48 0.19 0.19 0.08 2.569 2.064

3.87 0.32 0.27 0.11 2.556 2.034

4.24 0.36 0.44 0.10 2.526 2.006

3.63 0.23 0.23 0.10 2.552 2.052

3.06 0.13 0.15 0.11 2.508 2.048

T. Staněk, P. Sulovský / Cement and Concrete Research 68 (2015) 203–210

207

Table 10 The correlation matrix of the average apfu values of sets of belites from nine experimental samples with different input sulfur content and lime saturation values (listed in Tables 4–6). The critical value of Pearson's correlation coefficient is ±0.695 (ν = 7, α = 0.05, two-tailed); significant values are typed bold (significant positive correlation) or bold italics (significant negative correlation).

Na K Mg Ca Si Al S P FeIII

Na

K

1 0.777 0.598 −0.540 0.766 −0.853 −0.794 0.432 0.715

1 0.173 −0.424 0.401 −0.637 −0.363 0.499 0.619

Mg

Ca

Si

Al

S

P

FeIII

rcrit(ν = 9, α = 0.05) = 0.695 1 −0.681 0.455 −0.424 −0.523 0.210 0.592

1 −0.501 0.329 0.274 −0.566 −0.374

crystal lattice are practically 100% formed by the β modification. This corresponds with the results of Garbacik et al. [33]. Comparison of the unit cell parameters of belite without and with SO3 doping is demonstrated in Table 11. The unit cell volume of belite doped with SO3 is

1 −0.827 −0.911 0.385 0.343

1 0.843 −0.221 −0.726

1 −0.286 −0.430

1 0.070

1

slightly larger than that of belite without doping of SO3, which is in accordance with earlier findings of Morsli et al. [34].

3.2. Cement properties

Fig. 2. Fraction of sulfur atoms on the tetrahedral site versus the atomic ratio S/(Na + K) — average analyses of belite from this study and from Herfort et al. [27].

Tables 4–6 overviews the basic parameters of all cements prepared from burned clinkers. The parameters involve specific surface (Blaine), density, addition of setting a controller and compressive strength after 2, 7, 28 and 90 days hydration. For comparison, Table 4 gives parameters of clinkers prepared without SO3 addition (B — purely belite, B81S0 — belite with 25% alite content, A95S0 and A97S0 — high alite clinkers). The evolution of compressive strength with up to 28 days of hydration of chosen belite clinkers activated by SO3 added to the raw meal is shown in Fig. 4. The results show that a belite clinker doped with SO3 (sulfobelite clinker) is in contrast to a common belite clinker, distinctly more hydraulically active. Cement prepared from it compares with the technological properties of the Portland cement with prevailing alite. In contrast to a purely belite clinker doped with SO3, a small amount of alite distinctly supports the growth of early strength, because this alite is present in its more hydraulically active M1 modification [19], which is stabilized just by SO3. The alite formation also improves the parameters of burning and grindability. At an alite content of around 20 wt.% and a burning temperature of 1350 °C it is possible to achieve a common fineness of grinding compressive strength over 20 MPa after 2 days hydration. The relatively high lime saturation (yet still about 10% lower than in a common Portland clinker) in this belite clinker ensures a reaction with water and the formation of higher amounts of portlandite (Ca(OH)2). This increases the overall alkalinity, thus accelerating the course of hydration (Fig. 5). Also the presence of small amounts of anhydrite II, was found to have a positive impact on the development of cement strength [35]. The release of hydration heat for the same set of cements as in Fig. 4, is shown in Fig. 6. It is obvious, that although sulfur activated belite-rich cements have a similar evolution of compressive strength as the aliterich reference cements, they have lower values of heat of hydration. On the other hand, such a decrease in heat of hydration can be very beneficial in certain applications. A large part of these belite-rich cements

Table 11 Unit cell parameters of belite.

Fig. 3. Relationship between the average content of sulfur and aluminum in belite.

Clinker

B75S0

B85S5

a [Ǻ] b [Ǻ] c [Ǻ] beta [°] Volume [Ǻ3]

5.5034 6.7622 9.3306 94.085 346.36

5.5215 6.7747 9.3492 94.331 348.72

208

T. Staněk, P. Sulovský / Cement and Concrete Research 68 (2015) 203–210

Fig. 4. The 28-day evolution of the compressive strength of the chosen reference cements and cements prepared from belite-rich clinkers activated by sulfate ions (determined according to EN 196-1).

can therefore be classified as low heat of hydration cements with heat of hydration below 270 kJ/kg and some even in the category of cements with very low heat of hydration (below 220 kJ/kg) (see the values after 7 days of hydration in Fig. 6 and the specific hydration heat of cements during the 24 hours of hydration in Tab. 13). The time evolution of paste temperature during hydration is demonstrated in Fig. 7. These cements have soundness below 10 mm and have ordinary requirements for water and parameters of hardening as in the common high-alite (see examples in Table 12). 4. Conclusions The presented cement prepared in the laboratory represents a new type of hydraulically active low-energy belite cement. The clinker for its preparation, containing only around 20 wt.% of alite, was burned at a temperature of 1350 °C and is activated by the

addition of about 5 wt.% SO3 (related to the bulk clinker weight). Compared to SAB cements it does not contain C4 A3 S . It contains a small proportion of anhydrite. Activation was realized by the addition of sulfate ions, which in the structure of belite substitute SiO4, caused an increased entry of Al2O3 into the belite and increased the CaO:SiO2 in belite. The sulfur addition to the clinker also stabilized the hydraulically more active monoclinic alite M1 modification. Cements prepared from these clinkers have the same fineness of grinding technological parameters equal to those of alite-rich cement — including short-term strengths, which exceed 20 MPa. In addition, they have a favorably lower heat of hydration after 7 days of hardening. These results extend the knowledge in the field of heterogeneous reactions in clinker minerals formation. They indicate the possibility of separate industrial production of a special low-energy active belite clinker alongside with the production of the common alite clinker and the production of economically and ecologically expedient blended Portland cement, possessing suitable technological properties, or the goal-directed preparation of special cements with properties tailored to the intended use. Introduction of this cement into the cement making industry would lead to a decrease in energy consumption for

Table 12 Examples of results of normal consistence, hardening and volume stability of cements prepared from activated belite-rich clinkers, compared with reference cements (determined according to EN 196-3).

Fig. 5. The correlation between portlandite content (determined by DTA-TG) and compressive strength of high-belite cements.

Cement

Normal consistence [%]

Start of setting [min]

Time of setting [min]

Soundness [mm]

B75S0 B81S0 A97S0 B80S5 B92S8 B89S5 II B92S6

30.7 27.3 26.7 30.3 27.5 28.7 27.3

60 190 160 260 150 180 160

90 220 200 340 350 230 210

0.0 1.7 0.7 0.8 1.0 1.3 0.7

T. Staněk, P. Sulovský / Cement and Concrete Research 68 (2015) 203–210

209

Fig. 6. The development of heat of hydration in chosen reference cements and cements prepared from belite-rich clinkers activated by sulfate ions (determined by dissolution method according to EN 196-8).

production, a saving of high quality limestone, utilization of waste raw materials with high SO3 content and at the same time the lowering of CO2 emissions. This activated belite clinker could be used with regard to its good hydraulic properties for the production of self-contained sulfobelite cement or for blending with OPC for the production of special cements. In working practice, it would be advantageous to use the existing kiln lines for alternating production of the sulfobelite and the classical alite clinker. The lower burning temperature and very short clinkering zone during the belite clinker burning would enable the use of the sticking from the preceding burning of the alite clinker, because the sticking from belite burning is probably insufficiently stable.

Acknowledgment This research was done within project no. P104/12/1494 financed by the Czech Science Foundation.

References [1] A.M. Sharara, H. El-Didamony, E. Ebied, A. El-Aleem, Hydration characteristic of β-C2S in the presence of some pozzolanic materials, Cem. Concr. Res. 24 (5) (1994) 966–974. [2] A. Chatterjee, High belite cements—present status and future technological options: part I, Cem. Concr. Res. 26 (8) (1996) 1213–1225. [3] A. Emanuelson, A.R. Landa-Cánovas, S. Hansen, A comparative study of ordinary and mineralised Portland cement clinker from two different production units

Fig. 7. The time evolution of paste temperature during the first 24 hours of hydration.

210

[4]

[5]

[6] [7] [8] [9] [10]

[11]

[12] [13] [14]

[15] [16] [17] [18]

[19]

T. Staněk, P. Sulovský / Cement and Concrete Research 68 (2015) 203–210 part II: characteristics of the calcium silicates, Cem. Concr. Res. 33 (10) (2003) 1623–1630. A. Gies, D. Knöfel, Influence of alkalis on the composition of belite-rich cement clinkers and the technological properties of the resulting cements, Cem. Concr. Res. 16 (3) (1986) 411–422. K.H. Jost, R. Seydel, A. Müller, J. Stark, Relationship between phase-composition, cooling rate and Na2O content in belite clinkers, Zement Kalk Gips 41 (4) (1988) 169–170. K. Fukuda, I. Maki, S. Ito, Remelting reaction within belite crystals during cooling, J. Am. Ceram. Soc. 75 (10) (1992) 2896–2898. K. Fukuda, N. Wakamatsu, S. Ito, Improvement in hydration reactivity of alphaphase belite by remelting reaction, J. Am. Ceram. Soc. 84 (3) (2001) 639–641. R. Chrysafi, T. Perraki, G. Kakali, Sol-gel preparation of 2CaO∙SiO2, J. Eur. Ceram. Soc. 27 (2–3) (2007) 1707–1710. W. Kurdowski, S. Duszak, B. Trybalska, Belite produced by means of low-temperature synthesis, Cem. Concr. Res. 27 (1) (1997) 51–62. T. Staněk, The Relationship Between Parameters of Belite Clinker Preparation and its Hydraulic Properties(PhD Thesis) Brno University of Technology, Czech Republic, 2009. A.D.R. Brown, In Commercial, production, composition and properties of a calcium sulfoaluminate cement, Int. Conf. on Cements for the Future — Calcium Sulphoaluminates, London, 1998 (3 pp.). S.C. Mojumdar, I. Janotka, Thermophysical properties of blends from Portland and sulfoaluminate-belite cements, Acta Phys. Slovaca 52 (5) (2002) 435–446. C.D. Popescu, M. Muntean, J.H. Sharp, Industrial trial production of low energy belite cement, Cem. Concr. Compos. 25 (7) (2003) 689–693. T. Sui, Z. Wen, J. Wang, L. Fan, Development of belite based cements in China. 1st part: cement combinations for durable concrete, in: R.K. Dhir, T.A. Harrison, M.D. Newlands (Eds.), Proceedings of the International Conference Held at the University of Dundee, Dundee, Scotland, UK, 2005, pp. 323–328. A.P. Osokin, E.N. Potapova, Alitbildung in zusammengesetzten oxid-salz-schmelzen, Silikattechnik 37 (3) (1986) 79–80. A.P. Osokin, E.N. Potapova, Alitbildung in oxid-salz-schmelzen. Struktur und eigenschaften der flüssigen klinkerphase, Silikattechnik 40 (1) (1989) 4–6. I. Maki, K. Goto, Factors influencing the phase constitution of alite in Portland cement clinker, Cem. Concr. Res. 12 (3) (1982) 301–308. T. Staněk, Influence of minor oxides SO3 and MgO on the Portland clinker properties, in: J. Kostka (Ed.), Proceedings of the International Conference CEMENT 2002, High Tatras, Slovakia, 2002, pp. M1–M9 (2002). T. Staněk, P. Sulovský, The influence of the alite polymorphism on the strength of the Portland cement, Cem. Concr. Res. 32 (7) (2002) 1169–1175.

[20] S. Chromý, Dying of free lime and silicates in the polished section of Portland cement clinker, Zement Kalk Gips 27 (2) (1974) 79–84. [21] S. Chromý, Accuracy and precision of microscopic quantitative phase analysis of Portland clinkers, Silikáty 22 (3) (1978) 215–226. [22] W.A. Deer, R.A. Howie, J. Zussman, An Introduction to the Rock Forming Minerals. Appendix in 2nd Excerpted Student Edition, Longman, Burnt Mill, 1992. 696. [23] M. Kristmann, Portland cement clinker — mineralogical and chemical investigations, Cem. Concr. Res. 8 (1) (1978) 93102. [24] A. Ghose, P. Barnes, Electron microprobe analysis of Portland cement clinkers, Cem. Concr. Res. 9 (6) (1979) 747–755. [25] A.M. Harrisson, H.F.M. Taylor, Electron-optical analysis of the phases in a Portland cement clinker, with some observations on the calculation on quantitative phase composition, Cem. Concr. Res. 15 (5) (1985) 775–780. [26] C. Hall, K.L. Scrivener, Oilwell cement clinkers, X-ray microanalysis and phase composition, Adv. Cem. Bas. Mater. 7 (1) (1998) 28–38. [27] D. Herfort, J. Soerensen, E. Coulthard, Mineralogy of sulphate rich clinker and the potential for internal sulphate attack, World Cem. 28 (5) (1997) 77–85. [28] L. Bonafous, C. Bessada, D. Massiot, J.P. Coutures, 29Si MAS NMR study of dicalcium silicate: the structural influence of sulfate and alumina stabilizers, J. Am. Ceram. Soc. 78 (10) (1995) 2603–2608. [29] H.F.W. Taylor, Ettringite in cement paste and concrete, in: J.P. Bournazel, Y. Malier (Eds.), International RILEM Conference on Concrete: From Material to Structure, Arles, France, 1996, pp. 55–76. [30] K. Shimosaka, Study on the Additives in Clinker and the Properties of Cement(Diss. Thesis) Graduate School of Science and Engineering, Saitama University, 2005. 1–130. [31] T. Staněk, P. Sulovský, The berlinite substitution in clinkers; environmental aspects. 87, Tagung der Deutschen Mineralogischen Gesellschaft, Halle, Germany, in Hallesches, Jahrbuch für Geowissenschaften, 31, 2010, p. 249. [32] T. Staněk, P. Sulovský, Dicalcium silicate doped with sulfur, Adv. Cem. Res. 24 (4) (2012) 233–238. [33] A. Garbacik, T. Baran, M. Ostrowski, Energy saving low emissions belite cements, Proceedings of 13th ICCC, Madrid, Spain, 2011 (7 pp.). [34] K. Morsli, A.G. De la Torre, M. Zahir, M.A.G. Aranda, Mineralogical phase analysis of alkali and sulfate bearing belite rich laboratory clinkers, Cem. Concr. Res. 37 (5) (2007) 639–646. [35] T. Staněk, L. Tomancová, The influence of calcium sulfate forms on the properties of cement (in Czech), Silika 18 (1–2) (2008) 41–44.