Sie sind auf Seite 1von 6

Quantitative analysis of mineralized white Portland clinkers: The structure of Fluorellestadite

Isabel Pajares
n Eduardo Torroja, CSIC, C/Serrano Galvache s/n, Instituto de Ciencias de la Construccio 28033-Madrid, Spain

ngeles G. De la Torre A
mica Inorga nica, Cristalograf a y Mineralog a, Universidad de Ma laga, Departamento de Qu laga, Spain 29071-Ma

nez-Ram rez, Francisca Puertas, and Mar a-Teresa Blanco-Varela Sagrario Mart
n Eduardo Torroja, CSIC, C/Serrano Galvache s/n, Instituto de Ciencias de la Construccio 28033-Madrid, Spain

Miguel A. G. Arandaa)
mica Inorga nica, Cristalograf a y Mineralog a, Universidad de Ma laga, Departamento de Qu laga, Spain 29071-Ma

Received 1 April 2002; accepted 15 July 2002 Fluorellestadite, Ca10(SiO4 ) 3 (SO4 ) 3 F2 , has been synthesized as single phase. This compound crystallizes in the apatite type structure, s.g. P6 3 /m, with parameters a 9.4417(1) , c 6.9396(1) and V 535.76(1) 3 . The renement of its crystal structure converged to R W P 12.33% and R F 4.58%. The atomic parameters have been used to analyze the phase content of mineralized white Portland clinkers. These clinkers contain Ca3 SiO5 , Ca2 SiO4 , Ca12Al14O32F2 and Ca10(SiO4 ) 3 (SO4 ) 3 F2 . The agreement between the elemental composition inferred from the Rietveld phase analysis and that measured by XRF is noteworthy. This comparison does not take into account the presence of amorphous phases and unmodeled elemental substitutions in crystalline phases. Similar Rietveld studies on commercial white Portland clinkers are also shown to be feasible. 2002 International Centre for Diffraction Data. DOI: 10.1154/1.1505045
INTRODUCTION

For the production of Portland clinkers, mineralizers and/or uxes are often added to the raw mixes to accelerate reactions and enhance burnability. Traditional uxes such as Fe2 O3 and Al2 O3 have been partially substituted by the mineralizing pair CaF2 /CaSO4 to produce clinkers with low aluminate contents at temperatures between 1300 and 1400 C nez et al., 1991; Blanco-Varela et al., 1985, 1990; Gime Moir, 1982; Blanco-Varela et al., 1997; Wenxi et al., 1992. The use of this mineralizing pair to partially replace aluminates is particularly useful in manufacturing white clinkers, because of the potential for energy conservation and seawater resistance Pajares et al., 2001. The phases that may be present in this type of mineral nez and Blanco-Varela, 1995 are: alite, ized clinkers Gime C3 S, Ca3 SiO5 ]; belite, C2 S, Ca2 SiO4 ]; calcium aluminate C12A7 , Ca12Al14O33] or a related compound C11A7 CaF2 , and uorellestadite Ca12Al14O32F2 ]; 6CaO 3SiO2 3CaSO4 CaF2 or Ca10(SiO4 ) 3 (SO4 ) 3 F2 ]. The alite rate formation, in the presence of the mineralizing pair (CaF2 /CaSO4 ) is higher than that found for mixes mineralized only with CaF2 Christensen and Johansen, 1980. This faster rate for C3 S is mainly due to the thermodynamic effect of widening its stability range and the primary phase eld of crystallization. Additionally, ion mobility in the melt is enhanced by CaSO4 and CaF2 as a new liquid phase appears
a

Author to whom correspondence should be addressed: Departamento de mica Inorga nica, Cristalograf a y Mineralog a, Universidad de Ma laga, Qu laga, 29071-Ma Spain, Phone: Int 34 952131874, Fax: Int 34 952132000, electronic mail: g aranda@uma.es Powder Diffraction 17 (4), December 2002

nez and Blanco-Varela, 1995; Blancobelow 1200 C Gime Varela et al., 1986. It is of great importance to know the mineralogical compositions of these clinkers in order to understand and predict the mechanical strength of mortars or concrete elaborated with these cements. So far, calculations with the method of Bogue 1929 are used to deduce the mineralogical composition by using the elemental content usually measured by X-Ray Fluorescence XRF. This method has well known problems Taylor, 1997 mainly due to the lack of thermodynamical equilibrium in the kiln. Hence, alternative analytical methods are being studied to measure the mineralogical compositions of clinkers and cements. Laboratory X-Ray Powder Diffraction, LXRPD, is now widely used to carry out quantitative phase analyses of crystalline mixtures Hill and Howard, 1987; Bish and Howard, 1988; Madsen et al., 2001 by using the Rietveld method Rietveld, 1969; McCusker et al., 1999. The method compares the measured pattern and that calculated with the crystal structures. This methodology has recently been applied to Portland clinkers De la Torre et al., 2001, aluminous cements Guirado et al., 2000 and llmann, calcium sulpho-aluminate cements Schmidt and Po 2000. The interest on the application of XRD to Portland cements is recently increasing Goswami and Panda, 1999; Taylor et al., 2000. On the other hand, in order to apply the Rietveld method to mineralized clinkers it is necessary to know the structure of uorellestadite. Unfortunately, the crystal structure of this phase was not known although it can be prepared as a crystalline single phase. Its powder pattern is present in the PDF Database JCPDS-ICDD with a hexagonal cell, s.g. P6 3 /m, llman and Neubauer, of a 9.441 and c 6.939 Po
2002 JCPDS-ICDD 281

281

0885-7156/2002/17(4)/281/6/$18.00

TABLE I. Chemical analysis for the raw materials and mineralized clinker % w/w. Sample SiO2 Al2 O3 Fe2 O3 CaO MgO SO3 K2 O Na2 O CaF2 Lossd Ratioe
a b

Sand 89.43 5.11 0.22 1.46 0.42 0.00 0.93 0.11 2.30 12.45

Limestone 3.79 0.75 0.32 53.76 0.35 0.00 0.11 0.22 40.70 79.72

Kaolin 57.72 27.94 0.60 1.12 0.98 0.00 0.23 0.11 11.29 1.76

Fluorite 99.9 1.51

Gypsum 1.68 0.00 0.10 37.86 0.07 40.05 0.05 0.08 20.11 4.56

Clinker-ECa 23.04 2.61 0.45 67.73 0.53 2.75 0.32 0.29 2.28 -

Clinker-EAb 22.25 2.70 0.34 69.06 0.88 2.98 nd nd nd 0.48 -

Clinker-PAc 26.2 1.05 69.3 2.50 0.95 -

Expected composition for the mineralized clinker taking into account the raw materials ratio. Elemental analysis of the mineralized clinker by XRF and expressed as oxide content. c Phase analysis of the mineralized clinker by the Rietveld method and expressed as oxide content. d Weight loss by heating at 1000 C. e Raw materials ratio for the preparation of the mineralized clinker.

1993. The structure must be similar to that of apatite where the phosphate groups are randomly replaced by silicate and sulphate tetrahedra and the hydroxyl groups are replaced by the uoride anions. In fact, a structure for ellestadite approximately Ca10(SiO4 ) 3 (SO4 ) 3 (OH) 2 ] has been reported for a mineral from Cuba Organova et al., 1994. This crystal was monoclinic but with the apatite framework and it contained chloride anions and even a small amount of carbonate. The aim of this work is to determine the crystal structure of uorellestadite and use it for the quantitative analysis of the mineralized white Portland clinkers.

White Portland clinker

A typical commercial white Portland clinker was sampled from a factory and characterized by powder diffraction.
LXRPD studies

EXPERIMENT Fluorellestadite synthesis

The raw materials used for uorellestadite, FLELL, (6CaO 3SiO2 3CaSO4 CaF2 ) synthesis were CaCO3 , CaSO4 2H2 O, SiO2 and CaF2 with high purity grade in the stoichiometric ratio. The ground mixture was heated at 1000 C for 2 h. Then, the sample was ground again in an agate mortar and heated at 1000 C for 32 h. To follow the synthetic procedure, free lime concentration was determined by the ethylene-glycol method UNE Standard 80-243-86, 1986 giving 0.42% w/w of CaO. Routine LXRPD data indicated that the sample was a highly crystalline single phase.
Mineralized white Portland clinker preparation

The LXRPD pattern for FLELL was recorded on a Siemens D5000 /2 diffractometer at reection mode by using CuK 1,2 radiation 1.542 with a secondary curved graphite monochromator at 25 C. The samples were loaded in an aluminum holder by sample-front pressing. The experimental details are given in Table II. The LXRPD data for a commercial white Portland clinker were collected as for the mineralized clinker but spinning the sample at 15 rpm to improve the particles statistics.

RESULTS AND DISCUSSION

The nal goal of this work is the determination of the mineralogical composition of mineralized white Portland clinker by powder diffraction using the Rietveld methodology. However, to do so, the structural information of all crystalline phases is needed. As the crystal structure of FLELL was not known, this structure was determined prior to the mineralogical quantication.
Crystal structure of Fluorellestadite

The chemical analysis of the raw materials used for the preparation of the mineralized clinker is presented in Table I. In order to get the required mineralizers content, the dosage given at the bottom of Table I was selected. The resulting silica modulus was 7.50 % w/w. The raw materials were ground to particle size smaller than 45 m except limestone which was ground to particle size smaller than 125 m. The mixes were hand homogenized, using ethanol as dispersive medium, taking great care to avoid any modications of particle size and then dried. Cylindrical pellets of 2.5 cm diameter and 1.52.0 cm length were prepared. The clinkering temperature was 1350 C for 1 h in a laboratory furnace. The chemical composition of the clinker was determined according to UNE Standard 80-225-93 1993.
282 Powder Diffr., Vol. 17, No. 4, December 2002

FLELL crystallizes in an hexagonal cell, s.g. P6 3 /m , PDF no. 45-0009, with edges very similar to those of the apatite type structure. Hence, the uorapatite structure has been used as starting model Mackie and Young, 1973. The powder pattern was analyzed by the Rietveld method with the GSAS suite of programs Larson and Von Dreele, 1994 by using the pseudo-Voigt peak shape function Thompson et al., 1987 corrected for axial divergency Finger et al., 1994. The background was tted by linear interpolation function. The P position in uorapatite was substituted by Si and S 1 occupation factor for each type of atoms. Cell paramwith 2 eters and the spatial atomic positions were rened. Finally,
Pajares et al. 282

TABLE II. Data collection details and rened parameters for FLELL. Soller slits Divergence slit/mm Antidivergence slit/mm Receiving slit/mm X-ray tube, V/kV and I/mA Step-size/2 Angular range/2 Sample spinner Counting time/s/step Chemical formula a / c / V / 3 M /g/mol Z No. reections ( K 1 ) GW/0.01 2 LX/0.01 LY/0.01 S/L H/L R W P /% R P /% R F /% 2 sets 2 2 0.2 4030 0.03 10120 none 14 Ca10(SiO4 ) 3 (SO4 ) 3 F2 9.441 749 6.939 648 535.761 1003.25 1 292 5.82 0.51 8.53 0.031 0.016 12.33 8.85 4.58

the isotropic temperature factors were optimized. Preferred orientation was corrected by the March-Dollase algorithm Dollase, 1986 along 0 0 1 direction and the rened parameter was 0.9782. The pseudo-Voigt parameters were GW Gaussian part, LX ( d * -independent Lorentzian part and LY ( d * -dependent Lorentzian part. Some details of the FLELL renement are given in Table II. The last renement converged to R W P 12.3%, see Table II. The rened atomic positions and isotropic vibration temperature factors are given in Table III. The Rietveld plot is displayed in Fig. 1. The inset shows an enlarged view for the t in the high angle part of the pattern. Inter-atomic bond distances range between 1.54 and 1.56 for Si/SO bonds, 2.39 and 2.85 for CaO bonds, and the CaF bond distance is 2.30 . The angles for the tetrahedra (SiO4 or SO4 ) ranges between 107.4 and 110.8 The structure of FLELL belongs to the well known apatite type structure, so no discussion will be reported.
Phase analysis of the mineralized white Portland clinker

The sample was rotated during LXRPD data collection as the particles statistic increases notably. The intensities are

better and so, the phase analysis gives more accurate results. First, the phases present in the synthetic clinker were identied: C3 S De la Torre et al., 2002, C2 S Mumme et al., 1995, C11A7 CaF2 Williams, 1973 and uorellestadite. The LXRPD multi-phase pattern was analyzed with the GSAS suite of programs as above. Scale factors and unit cell parameters were rened for each phase and the positional and thermal atomic parameters were NOT rened for any phase. It is worthy to point out that the peak shape parameters for phases in low concentrations are unstable and usually hard to rene. In this case, a number of variables as low as possible should be used. Hence, only two parameters to describe the pseudo-Voigt were used and only one parameter LY was rened for the phases in low concentrations. Alite presents preferred orientation due to its large particle size. In this case the effect was corrected using the spherical-harmonic correction Von Dreele, 1997. This method gives much better ts for alite than the MarchDollase algorithm Dollase, 1986. It used the cylindrical symmetry and the order of the spherical harmonics was 10. The nal texture index was 4.79 1 texture index represents an ideal random powder and stands for a single crystal. The Rietveld plot for the mineralized white Portland clinker is given in Fig. 2. An enlarged view of the most informative part of the pattern is given in Fig. 3 top with the peaks due to a given phase labeled. Details of the Rietveld renement of this pattern, including the phase analysis, are given in Table IV. The phase content obtained from the Rietveld analysis has been transformed to overall elemental content expressed as oxides in Table I. However, one should exercise care when comparing these values with those expected from the dosage of the raw materials and those measured by XRF. In order to compare these numbers, the atomic substitution in the crystalline phases and the presence of amorphous phases was neglected. This approximation is coarse. However, under these constraints, the agreement in Table I is remarkably good. Some considerations can be drawn: 1 Overall CaO is always slightly overestimated in the Rietveld analysis as it includes the Ca/Mg substitution which cannot be modeled. Consequently, MgO is not estimated. 2 Similarly, SiO2 is overestimated in the Rietveld analysis as the Si/Al substitution cannot be taken into account. Consequently, Al2 O3 is underestimated. 3 The agreement for SO3 content is good and CaF2 is underestimated in the Rietveld analysis probably due to the presence of amorphous uorides which are the last crystallizing fraction.

TABLE III. Rened atom parameters and thermal factors for Ca10(SiO4 ) 3 (SO4 ) 3 F2 . Atom O1 O2 O3 Sia Sa Ca1 Ca2 F
a

Wyck. pos. 6h 6h 12i 6h 6h 4f 6h 2a

x 0.32347 0.58657 0.33855 0.39613 0.3961 1/3 0.23993 0.0

y 0.48466 0.46838 0.25685 0.36843 0.3684 2/3 0.99203 0.0

z 1/4 1/4 0.07095 1/4 1/4 0.00165 1/4 1/4

U iso / 2 0.0142 0.0182 0.0221 0.0121 0.012 0.0311 0.0221 0.0263

The occupation factor is 0.5. Quantitative analysis of mineralized white Portland clinkers 283

283

Powder Diffr., Vol. 17, No. 4, December 2002

Figure 1. Rietveld plot (10-120/2 ) for uorellestadite. The inset shows an enlarged view of the high angle part of the pattern.

Figure 2. Rietveld plot (20-70/2 ) for the mineralized white Portland clinker.

284

Powder Diffr., Vol. 17, No. 4, December 2002

Pajares et al.

284

Figure 3. Selected region (28-35/2 ) of the Rietveld plot for the mineralized white Portland clinker top. Same selected region for a commercial white Portland clinker bottom.

TABLE IV. Selected data for the LXRPD Rietveld renement of the mineralized clinker. Phase C3 S C2 S C11A7 CaF2 FLELL
a b

s.g. Cm P2 1 /n I 4 3d P6 3 /m

a / 33.0584 5.5092 11.9593 9.4461

b / 7.05944 6.7562 11.959 9.446

c / 18.5361 9.3132 11.959 6.9341

/
94.2076 94.412 90.0 90.0

V / 3 4314.17 345.61 17101 535.82

GWa 12.89 5.0 5.0 5.0

LYb 16.76 394 258 504

w/w % 72.33 15.17 2.13 10.54

S / L 0.015 and H / L 0.015 were used to describe the asymmetry due to axial divergence for all the phases. GW is expressed in (0.01) 2 units. c LY is expressed in 0.01 units. 285 Powder Diffr., Vol. 17, No. 4, December 2002 Quantitative analysis of mineralized white Portland clinkers 285

Phase analysis of commercial white Portland clinker

The Rietveld analysis of a diffraction pattern for a commercial white Portland clinker manufactured without CaF2 was carried out for the sake of comparison. This sample was simpler and it contained C3 S, C2 S and C3 A. The phase analysis was carried out as described above and the most informative part of the Rietveld plot is given in Fig. 3 bottom. The results for white Portland clinkers, including the quantication of the non-diffracting part, will be reported elsewhere.

CONCLUSIONS

This paper reports the crystal structure of uorellestadite which belongs to the apatite type structure. This structure is used to quantify the mineralogical content of mineralized white Portland clinkers. These clinkers contain C3 S, C2 S, C11A7 CaF2 and uorellestadite. Some guidelines for these type of Rietveld studies are presented. The agreement between the elemental composition inferred from the phase analysis and that measured by XRF is noteworthy taken into account the possible presence of amorphous phases. Similar Rietveld studies on commercial white Portland clinkers are also shown to be feasible.

ACKNOWLEDGMENTS

This work has been supported by the FEDER 1FD97nez-Ram rez 0894 and PGC97-1144 research grants. S. Mart thanks DGES MEC for the award of Contrato de Incor n de Doctores. I. Pajares and A. G. De la Torre poracio a, thank two studentships by MCYT and Junta de Andaluc respectively.
Bish, D. L. and Howard, S. A. 1988. Quantitative phase analysis using the Rietveld method, J. Appl. Crystallogr. 21, 86 91. zquez, T., and Palomo, A. 1985. Spanish Patent Blanco-Varela, M. T., Va N 54269. zquez, T., and Palomo, A. 1986. A study of a Blanco-Varela, M. T., Va new liquid-phase to obtain low-energy cements, Cem. Concr. Res. 16, 97104. zquez, T., Palomo, A., Puertas, F., and Gime nez, S. Blanco-Varela, M. T., Va 1990. Spanish Patent N 9001906. zquez, T. 1997. Blanco-Varela, M. T., Palomo, A., Puertas, F., and Va CaF2 and CaSO4 in white cement clinker production, Adv. Cem. Res. 9, 105113. Bogue, R. H. 1929. Industrial Engineering Chemistry Analytical Edition 1, 192. Christensen, N. H. and Johansen, V. 1980. Mineralizers and uxes in the clinkering process: Kinetics effects on alite formation, 7th Inter. Cong. Chem. Cement Paris II, I-1-I-16. De la Torre, A. G., Cabeza, A., Calvente, A., Bruque, S., and Aranda, M. A. G. 2001. Full phase analysis of Portland clinker by penetrating synchrotron powder diffraction, Anal. Chem. 73, 151156. De la Torre, A. G., Bruque, S., Campo, J., and Aranda, M. A. G. 2002. The superstructure of C3 S from synchrotron and neutron powder diffraction and its role in quantitative phase analyses, Cem. Concr. Res. 32, 13471356. Dollase, W. A. 1986. Correction of intensities for preferred orientation in powder diffractometry: application of the March model, J. Appl. Crystallogr. 19, 267272. Finger, L. W., Cox, D. E., and Jephcoat, A. P. 1994. A correction for powder diffraction peak asymmetry due to diaxial divergence, J. Appl. Crystallogr. 27, 892900. 286 Powder Diffr., Vol. 17, No. 4, December 2002

nez, S., Blanco-Varela, M. T., Palomo, A., and Puertas, F. 1991. Gime Production of low energy requirements cement. An industrial test, Zement-Kalk-Gips 44, 1215. nez, S. and Blanco-Varela, M. T. 1995. Solid state phases relationGime ship in the CaO-SiO2 -Al2 O3 -CaF2 -CaSO4 system, Cem. Concr. Res. 25, 778 782. Goswami, G. and Panda, J. D. 1999. Application of XRD in a rapid quality control system of cement, Powder Diffr. 14, 114 117. , S., and Chincho n, S. 2000. Quantitative Rietveld analyGuirado, F., Gal sis of aluminous cement clinker phases, Cem. Concr. Res. 30, 1023 1029. Hill, R. J. and Howard, C. J. 1987. Quantitative phase analysis from neutron powder diffraction data using the Rietveld method, J. Appl. Crystallogr. 20, 467 474. Larson, A. R. and Von Dreele, R. B. 1994. General Structural Analysis System, Los Alamos National Lab. Rep. No. LA-UR-86-748. GSAS program @ http://public.lanl.gov:80/gsas/. Mackie, P. E. and Young, R. A. 1973. Location of Nd dopant in uorapatite, Ca5 (PO4 ) 3 F: Nd, J. Appl. Crystallogr. 6, 26 31. Madsen, I. C., Scarlett, N. V. Y., Cranswick, L. M. D., and Lwin, T. 2001. Outcomes of the International Union of Crystallography Commission on powder diffraction round robin on quantitative phase analysis: samples 1a to 1h, J. Appl. Crystallogr. 34, 409 426. r, D., and Scardi, P. McCusker, L. B., Von Dreele, R. B., Cox, D. E., Loue 1999. Rietveld renements guidelines, J. Appl. Crystallogr. 32, 36 50. Moir, G. K. 1982. Mineraliser high alite cements, World Cem. 374 382. Mumme, W. G., Hill, R. J., Bushnell-Wye, G., and Segnit, E. R. 1995. Rietveld structure renement, crystal chemistry and calculated powder diffraction data for the polymorphs of dicalcium silicate and related phases, Neues Jahrb. Mineral., Abh. 169, 35 68. Organova, N. I., Rastsvetaeva, R. K., Kuzmina, O. V., Arapova, G. A., Litsarev, M. A., and Finko, V. I. 1994. Crystal structure of lowsymmetry ellestadite in comparison with other apatitelike structures, Kristallograya 39, 278 282. zquez, T., and Mart nezPajares, I., Puertas, F., Blanco-Varela, M. T., Va rez, S. 2001. Inuencia del contenido de aluminatos en la aptiRam n, la hidratacio n y el comportamiento meca nico de cetud a la coccio mentos blancos mineralizados con CaF2 y CaSO4 , Cemento y n 825, 626 635. Hormigo llmann, H. and Neubauer, J. 1993. ICDD Grant-in-Aid. Po Rietveld, H. M. 1969. A prole renement method for nuclear and magnetics structures, J. Appl. Crystallogr. 2, 6571. llmann, H. 2000. Quantication of calcium sulphoSchmidt, R. and Po aluminate cement by Rietveld analysis, Mater. Sci. Forum 321 324, 10221027. Taylor, H. F. W. 1997. Cement Chemistry, 2nd ed. Thomas Telford, London. Taylor, J. C., Hinczak, I., and Matulis, C. E. 2000. Rietveld full-prole quantication of Portland cement clinker: the importance of including a full crystallography of the major phase polymorphs, Powder Diffr. 15, 718. Thompson, P., Cox, D. E., and Hasting, J. B. 1987. Rietveld renement of Debye-Scherrer synchrotron X-ray data from Al2 O3 , J. Appl. Crystallogr. 20, 79 83. todos de ensayo de cementos. UNE STANDARD 80-243-86. 1986. Me n del o xido de calcio libre; me todo del etilenglicol. Determinacio todos de ensayo de cementos. UNE STANDARD 80-225-93. 1993. Me lisis qu mico. Determinacio n del dio xido de silicio (SiO2 ) reactivo Ana en los cementos, en las puzolanas y en las cenizas volantes. Von Dreele, R. B. 1997. Quantitative texture analysis by Rietveld renement, J. Appl. Crystallogr. 30, 517525. Wenxi, H., Guanglliant, X., and Zhongyuan, L. 1992. The effect of Calcium Fluorsulpho-Silicate on Formation of Calcium Silicate, 9th Inter. Congr. Chem. Cem. New Delhi, 2, 379385. Williams, P. P. 1973. Renement of the structure of (CaO) 11(Al2 O3 ) 7 CaF2 , Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem. 29, 15501551. Pajares et al. 286

Das könnte Ihnen auch gefallen