- Email: [email protected]
High-resolution synchrotron powder diffraction analysis of ordinary Portland cements: Phase coexistence of alite
Nuclear Instruments and Methods in Physics Research B 238 (2005) 87–91 www.elsevier.com/locate/nimb
High-resolution synchrotron powder diffraction analysis of ordinary Portland cements: Phase coexistence of alite ´ ngeles G. de la Torre, Enrique R. Losilla, Aurelio Cabeza, A Miguel A.G. Aranda * Departamento de Quı´mica Inorga´nica, Cristalografı´a y Mineralogı´a, Universidad de Ma´laga, 29071 Ma´laga, Spain Available online 27 July 2005
Abstract The mineralogical composition of four commercial and NIST RM-8488 Portland clinkers have been analysed by Rietveld methodology using high-resolution synchrotron X-ray powder diffraction data. Alite phase coexistence has been observed in four patterns. White Portland clinkers show a single alite or a very small amount of a second alite with smaller volume due to higher magnesium content. Grey Portland clinkers show a much pronounced alite phase coexistence which has been related to higher magnesium contents. Details about these analyses are given. Furthermore, the full mineralogical composition (including the non-diffracting content) has been determined from the overestimation of the added standard, a-Al2O3, in the Rietveld analyses. White clinkers contain 15 wt.% of non-diffracting content while this fraction is much smaller in grey clinkers, 7 wt.%. Ó 2005 Elsevier B.V. All rights reserved. PACS: 61.10.Nz Keywords: Rietveld; Phase analysis; Clinker; Alite
1. Introduction There is a great interest in the Quantitative Phase Analysis (QPA) of Ordinary Portland Cements (OPC) as their final performances depend on the mineralogical composition (and texture). *
Corresponding author. Tel.: +34 9 5213 1874; fax: +34 9 5213 2000. E-mail address: [email protected] (M.A.G. Aranda).
Rietveld analysis [1] via Laboratory X-ray Powder Diffraction (LXRPD) is the most powerful as well as available technique for obtaining QPA. However, LXRPD data may contain systematic errors such as strong preferred orientation, optical aberrations which change with 2h, microabsorption and strong peak overlapping of the different phases. Furthermore, the small mean penetration depth (25 lm for Cu Ka) may lead to poor particle statistics if the sample is not rotated, especially
0168-583X/$ - see front matter Ó 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.nimb.2005.06.023
88
A´.G. de la Torre et al. / Nucl. Instr. and Meth. in Phys. Res. B 238 (2005) 87–91
for the large alite crystals. Hence, in order to validate the QPA obtained by LXRPD, it is desirable to compare the obtained results with an analytical technique that has not such drawbacks. High-Resolution Synchrotron X-ray Powder Diffraction (HRSXRPD) overcomes most of the drawbacks described above. Rotating capillary geometry (transmission) avoids preferred orientation. High energy X-rays minimize microabsorption and a large amount of sample is tested leading to good particle statistics. High-resolution data minimize the overlapping and parallel synchrotron X-rays geometry, with an analyser crystal, do not show optical aberrations. Hence, we have previously reported HRSXRPD data for Portland cement to achieve good QPA of commercial OPC [2] and to show that accurate Rietveld QPA for cements can be obtained [3]. Other authors have also used synchrotron radiation for the quantitative phase analysis of cements [4,5]. QPA using the Rietveld method is standardless but the crystal structures of all phases must be known as the process comprises the comparison of the measured and calculated patterns. This methodology has been applied to grey and white Portland clinkers and cements of different complexities [6–9] including the quantification of the non-diffracting contents [4,10]. It is also being adapted to carry out the on-line mineralogical analyses in cement factories [11,12]. The four main phases in OPC are alite (C3S or Ca3SiO5), belite (C2S or Ca2SiO4), aluminate (C3A or Ca3Al2O6) and ferrite (C4AF or Ca4Al2Fe2O10). In this work, we present results of QPA of commercial OPC (grey and white) by using HRSXRPD including the quantification of the non-diffracting contents. We have also analysed a NIST RM-8488 clinker.
HRSXRPD. The five samples were mixed with aAl2O3 standard.
2. Experimental section
3. Results and discussion
2.1. Samples
3.1. Phase coexistence of alite
Two commercial white Portland clinkers (W1, W2), two commercial grey Portland clinkers (G1, G2) [all from different factories] and a NIST RM-8488 clinker have been analysed by
Fig. 1 displays a selected region of W1, W2, G2 and RM-8488 HRSXRPD patterns. Alite is the high temperature monoclinic (MIII) polymorph of Ca3SiO5 [14] present in OPC since it is stabilised
2.2. X-ray data collection HRSXRPD patterns were collected on ID31 powder diffractometer of ESRF, European Synchrotron Radiation Facility, (Grenoble, France) using a short penetrating wavelength k = ˚ (30.97 keV) selected with a double-crys0.40027 A tal Si(1 1 1) monochromator and calibrated with Si ˚ ). The Debye–Scherrer conNIST (a = 5.43094 A figuration was used with the samples loaded in borosilicate glass capillaries, diameter of 2 mm, and rotating during data collection. The overall measuring time was 100 min to have very good statistics over the wide angular range 2.5–30° (in ˚ ). The data from the multi-ana2h) (9.15–0.77 A lyser Si(1 1 1) stage were normalised and summed into 0.003° step size. Powder data for one sample, G1, was collected with a slightly different wave˚. length, k = 0.42970 A 2.3. Synchrotron X-ray data analysis The powder patterns were refined by the Rietveld method with GSAS suite of programs [13] by using a Lorentzian peak shape function. The references of the crystal structures for the cement phases were reported in [3] including that of the corundum standard. The refined overall parameters were: background coefficients, cell parameters, zero-shift error, peak shape parameters and phase fractions. Brindley correction was not necessary. The peak shape parameter LY was refined freely for each phase and the anisotropic peak shape parameter (STEC in GSAS program) was refined when necessary.
I (a.u.)
A´.G. de la Torre et al. / Nucl. Instr. and Meth. in Phys. Res. B 238 (2005) 87–91
89
C3 S 2
C4 AF
C4 AF
C3 A
C3 S 1
C3 S 2 C3 S 1 C3 S 2
C3 S 1 C2 S
NKS
C2 S
NKS
G2 C3 S 1 C3 S 2
C3 S 1
C3 S 2
RM 8488
C3 S
C3 S
C3 S C3 A
C2 S
W1 C2 S
C3 S
C3 S
C3 S
W2
8
2θ /º
9
Fig. 1. Selected region of HRSXRPD raw patterns with peaks labelled in cement jargon for the W1 and G2 samples. The arrows highlight the alite coexistence in several clinkers. From bottom to top: two commercial white OPCs (W1 and W2), a commercial grey OPC (G2) and NIST RM-8848.
at room temperature by the presence of foreign ions. W1 does not show alite phase coexistence, see Fig. 1, whereas, W2, G1, G2 and RM 8488 display this coexistence. Fig. 2 shows a selected region of the Rietveld plot for G1 with peaks from the different phases labelled, as an example of the quality of the fits. Two alites computed with the same structural description [14] but with different
unit cells volumes were introduced to perform the refinements/quantifications. Rietveld results for all samples are given in Table 1. Table 2 gives the elemental compositions measured by X-ray Fluorescence for all clinkers. Alite MIII (monoclinic) is stabilised mainly by the substitution of Ca2+ by Mg2+. White Portland clinkers contain very small amounts of Mg2+ and
Fig. 2. Rietveld plot for G1 grey OPC with peaks labelled.
A´.G. de la Torre et al. / Nucl. Instr. and Meth. in Phys. Res. B 238 (2005) 87–91
90
Table 1 Rietveld QPA of HRSXRPD data for the analysed samples Phases/wt.%
W1
W2
G1
G2
RM-8488
C3S(1) C3S(2) C3S_total C2S C3A C4AF NaK3(SO4)2 K2SO4 MgO Al2O3 Al2O3 (weighed) Amorphous ˚3 V C3S(1)/A ˚3 V C3S(2)/A
51.7 (1) – – 19.7 (2) 2.4 (1) – – – – 26.2 (1) 21.99 20.6 (5) 120.91 (1) –
55.3 (1) 2.0 (1) 57.3 (1) 8.5 (2) 6.0 (1) – – 0.6 (1) – 27.6 (1) 24.34 15.6 (5) 120.87 (1) 119.90 (1)
32.1 (2) 16.7 (2) 48.8 (3) 10.9 (2) 6.3 (1) 10.6 (1) 1.5 (1) – 0.37 (4) 21.55 (7) 20.33 8.8 (4) 120.42 (1) 119.94 (1)
23.8 (2) 26.6 (2) 50.4 (3) 8.3 (1) 6.7 (1) 9.3 (1) 1.5 (1) – 0.43 (3) 23.40 (7) 21.79 7.1 (4) 120.43 (1) 119.93 (1)
35.6 (2) 11.5 (3) 46.3 (4) 10.7 (2) 2.6 (1)_2.5 (1)a 11.8 (1) – – – 25.3 (1) 23.75 8.0 (5) 120.68 (1) 120.13 (1)
The weighed amounts of a-Al2O3, the derived non-diffracting contents and the volumes (V/Z) for C3S are also given. a RM-8488 contains two polymorphs of C3A, cubic (2.6%) and orthorhombic (2.5%). Table 2 Elemental composition of each cement sample measured by X-ray fluorescence Sample
CaO/wt.%
SiO2/wt.%
Al2O3/wt.%
Fe2O3/wt.%
MgO/wt.%
SO3/wt.%
K2O/wt.%
Na2O/wt.%
W1 W2 G1 G2 RM-8488
68.74 69.16 65.42 63.88 66.50
22.84 21.92 20.99 21.20 22.68
4.53 3.70 5.26 5.89 4.90
0.57 0.29 3.17 3.67 4.07
0.60 0.60 1.24 1.62 0.98
1.03 0.4 1.15 1.43 0.31
– – 1.10 1.03 0.35
– – 0.41 – 0.11
consequently the volumes of C3S of these samples are larger, see Table 1. Furthermore, alite phase coexistence, if present, is very small (W2 clinker contains 2.0 (1)% of the C3S phase with the smallest volume). Grey clinkers (G1 and G2) have higher contents of Mg2+ and they display overall smaller volumes, see Table 1. The amount of the ˚3 alite phase with the smallest volume (119.9 A per formula unit) is higher for the analysed grey clinkers, see Table 1. Additionally, they have a small amount of MgO as free periclase. RM-8488 clinker also shows two alites with different volumes and it contains less Mg2+, see Table 2, and no periclase is present in the pattern. These results indicate that both alite phases contain magnesium cations. However, the Mg2+-content in the smallest volume phase is higher. The superstructures of these two coexisting alites should be different but we are using the same structural description in the analyses since there are no alternative structures available. More synthetic and structural
studies are needed to properly understand and describe the structural role of the impurities (Mg2+, Al3+, etc.) in the alite(s). 3.2. HRSXRPD and LXRPD studies The Rietveld analytical results given in Table 1 (and those shown in Fig. 1) indicates that the alite phase coexistence is very pronounced in grey clinkers which contain a high amount of magnesium. This is the first direct evidence of alite phase coexistence in commercial OPC which has been possible due to the extreme high resolution of the ID31 synchrotron powder diffraction data. The influence of the alite phase coexistence in the Rietveld analysis of LXRPD data remains to be fully studied. Preliminary results indicate that laboratory data with Cu Ka1,2 radiation do not show any signs of phase coexistence. The analyses are good when using a unique structural description for alite. However, when strictly monochromatic
A´.G. de la Torre et al. / Nucl. Instr. and Meth. in Phys. Res. B 238 (2005) 87–91
Cu Ka1 radiation is used, some signs of phase coexistence are evident in the patterns that need to be taken into account. A full comparison of accurate HRSXRPD analytical results and the mineralogical contents from LXRPD data will be reported elsewhere.
91
Acknowledgment We thank the ESRF for the provision of X-ray synchrotron beam time. References
3.3. Non-diffracting content analysis All clinkers were mixed with a-Al2O3 in order to indirectly obtain the amorphous contents [15]. This quantification is obtained from the small overestimation of the standard in the Rietveld analyses, see Table 1. Hence, the particle statistics must be very good and HRSXRPD is much better at providing that than LXRPD. Although more analyses are needed, the studied samples indicate that white Portland clinkers contain approximately 15–20 wt.% of non-diffracting/amorphous content. These values for grey OPC clinkers are much smaller and closer to 7–8 wt.%. These results are in agreement with the manufacturing processes and other reports [2,4]. The amount of liquid phase in grey-OPC kilns is larger and the particles are better crystallised (low non-diffracting contents), whereas the amount of liquid phase in white-clinker kilns is smaller and so particles grow containing more defects which leads to an increase of the non-diffracting content.
[1] R.A. Young, in: R.A. Young (Ed.), The Rietveld Method, IUCr Monographs on Crystallography-5, Oxford University Press, 1993. [2] A.G. De la Torre, A. Cabeza, A. Calvente, S. Bruque, M.A.G. Aranda, Anal. Chem. 73 (2001) 151. [3] A.G. De la Torre, M.A.G. Aranda, J. Appl. Crystallogr. 36 (2003) 1169. [4] P.M. Suherman, A.V. Riessen, B. Oconnor, D. Li, D. Bolton, H. Fairhurst, Powder Diffr. 17 (2002) 178. [5] V. Peterson, B. Hunter, A. Ray, L.P. Aldridge, Appl. Phys. A 74 (2002) S1409. [6] G. Walenta, T. Fullmann, Powder Diff. 19 (2004) 40. [7] T. Fullmann, G. Walenta, ZKG Int. 56 (2003) 45. [8] R. Schmidt, T. Feldman, World Cement (2003) 64. [9] N. Lundgaard, E.S. Jons, World Cement (2003) 59. [10] P.S. Whitfield, L.D. Mitchell, J. Mater. Sci. 38 (2003) 4415. [11] N.V.Y. Scarlett, I.C. Madsen, Powder Diffr. 16 (2001) 71. [12] M. Enders, ZKG Int. 56 (2003) 54. [13] A.C. Larson, R.B. Von Dreele, General Structural Analysis System, Los Alamos National Lab. Rep. No. LA-UR86-748, 1994. GSAS program. Available from:
High-resolution synchrotron powder diffraction analysis of ordinary Portland cements: Phase coexistence of alite ´ ngeles G. de la Torre, Enrique R. Losilla, Aurelio Cabeza, A Miguel A.G. Aranda * Departamento de Quı´mica Inorga´nica, Cristalografı´a y Mineralogı´a, Universidad de Ma´laga, 29071 Ma´laga, Spain Available online 27 July 2005
Abstract The mineralogical composition of four commercial and NIST RM-8488 Portland clinkers have been analysed by Rietveld methodology using high-resolution synchrotron X-ray powder diffraction data. Alite phase coexistence has been observed in four patterns. White Portland clinkers show a single alite or a very small amount of a second alite with smaller volume due to higher magnesium content. Grey Portland clinkers show a much pronounced alite phase coexistence which has been related to higher magnesium contents. Details about these analyses are given. Furthermore, the full mineralogical composition (including the non-diffracting content) has been determined from the overestimation of the added standard, a-Al2O3, in the Rietveld analyses. White clinkers contain 15 wt.% of non-diffracting content while this fraction is much smaller in grey clinkers, 7 wt.%. Ó 2005 Elsevier B.V. All rights reserved. PACS: 61.10.Nz Keywords: Rietveld; Phase analysis; Clinker; Alite
1. Introduction There is a great interest in the Quantitative Phase Analysis (QPA) of Ordinary Portland Cements (OPC) as their final performances depend on the mineralogical composition (and texture). *
Corresponding author. Tel.: +34 9 5213 1874; fax: +34 9 5213 2000. E-mail address: [email protected] (M.A.G. Aranda).
Rietveld analysis [1] via Laboratory X-ray Powder Diffraction (LXRPD) is the most powerful as well as available technique for obtaining QPA. However, LXRPD data may contain systematic errors such as strong preferred orientation, optical aberrations which change with 2h, microabsorption and strong peak overlapping of the different phases. Furthermore, the small mean penetration depth (25 lm for Cu Ka) may lead to poor particle statistics if the sample is not rotated, especially
0168-583X/$ - see front matter Ó 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.nimb.2005.06.023
88
A´.G. de la Torre et al. / Nucl. Instr. and Meth. in Phys. Res. B 238 (2005) 87–91
for the large alite crystals. Hence, in order to validate the QPA obtained by LXRPD, it is desirable to compare the obtained results with an analytical technique that has not such drawbacks. High-Resolution Synchrotron X-ray Powder Diffraction (HRSXRPD) overcomes most of the drawbacks described above. Rotating capillary geometry (transmission) avoids preferred orientation. High energy X-rays minimize microabsorption and a large amount of sample is tested leading to good particle statistics. High-resolution data minimize the overlapping and parallel synchrotron X-rays geometry, with an analyser crystal, do not show optical aberrations. Hence, we have previously reported HRSXRPD data for Portland cement to achieve good QPA of commercial OPC [2] and to show that accurate Rietveld QPA for cements can be obtained [3]. Other authors have also used synchrotron radiation for the quantitative phase analysis of cements [4,5]. QPA using the Rietveld method is standardless but the crystal structures of all phases must be known as the process comprises the comparison of the measured and calculated patterns. This methodology has been applied to grey and white Portland clinkers and cements of different complexities [6–9] including the quantification of the non-diffracting contents [4,10]. It is also being adapted to carry out the on-line mineralogical analyses in cement factories [11,12]. The four main phases in OPC are alite (C3S or Ca3SiO5), belite (C2S or Ca2SiO4), aluminate (C3A or Ca3Al2O6) and ferrite (C4AF or Ca4Al2Fe2O10). In this work, we present results of QPA of commercial OPC (grey and white) by using HRSXRPD including the quantification of the non-diffracting contents. We have also analysed a NIST RM-8488 clinker.
HRSXRPD. The five samples were mixed with aAl2O3 standard.
2. Experimental section
3. Results and discussion
2.1. Samples
3.1. Phase coexistence of alite
Two commercial white Portland clinkers (W1, W2), two commercial grey Portland clinkers (G1, G2) [all from different factories] and a NIST RM-8488 clinker have been analysed by
Fig. 1 displays a selected region of W1, W2, G2 and RM-8488 HRSXRPD patterns. Alite is the high temperature monoclinic (MIII) polymorph of Ca3SiO5 [14] present in OPC since it is stabilised
2.2. X-ray data collection HRSXRPD patterns were collected on ID31 powder diffractometer of ESRF, European Synchrotron Radiation Facility, (Grenoble, France) using a short penetrating wavelength k = ˚ (30.97 keV) selected with a double-crys0.40027 A tal Si(1 1 1) monochromator and calibrated with Si ˚ ). The Debye–Scherrer conNIST (a = 5.43094 A figuration was used with the samples loaded in borosilicate glass capillaries, diameter of 2 mm, and rotating during data collection. The overall measuring time was 100 min to have very good statistics over the wide angular range 2.5–30° (in ˚ ). The data from the multi-ana2h) (9.15–0.77 A lyser Si(1 1 1) stage were normalised and summed into 0.003° step size. Powder data for one sample, G1, was collected with a slightly different wave˚. length, k = 0.42970 A 2.3. Synchrotron X-ray data analysis The powder patterns were refined by the Rietveld method with GSAS suite of programs [13] by using a Lorentzian peak shape function. The references of the crystal structures for the cement phases were reported in [3] including that of the corundum standard. The refined overall parameters were: background coefficients, cell parameters, zero-shift error, peak shape parameters and phase fractions. Brindley correction was not necessary. The peak shape parameter LY was refined freely for each phase and the anisotropic peak shape parameter (STEC in GSAS program) was refined when necessary.
I (a.u.)
A´.G. de la Torre et al. / Nucl. Instr. and Meth. in Phys. Res. B 238 (2005) 87–91
89
C3 S 2
C4 AF
C4 AF
C3 A
C3 S 1
C3 S 2 C3 S 1 C3 S 2
C3 S 1 C2 S
NKS
C2 S
NKS
G2 C3 S 1 C3 S 2
C3 S 1
C3 S 2
RM 8488
C3 S
C3 S
C3 S C3 A
C2 S
W1 C2 S
C3 S
C3 S
C3 S
W2
8
2θ /º
9
Fig. 1. Selected region of HRSXRPD raw patterns with peaks labelled in cement jargon for the W1 and G2 samples. The arrows highlight the alite coexistence in several clinkers. From bottom to top: two commercial white OPCs (W1 and W2), a commercial grey OPC (G2) and NIST RM-8848.
at room temperature by the presence of foreign ions. W1 does not show alite phase coexistence, see Fig. 1, whereas, W2, G1, G2 and RM 8488 display this coexistence. Fig. 2 shows a selected region of the Rietveld plot for G1 with peaks from the different phases labelled, as an example of the quality of the fits. Two alites computed with the same structural description [14] but with different
unit cells volumes were introduced to perform the refinements/quantifications. Rietveld results for all samples are given in Table 1. Table 2 gives the elemental compositions measured by X-ray Fluorescence for all clinkers. Alite MIII (monoclinic) is stabilised mainly by the substitution of Ca2+ by Mg2+. White Portland clinkers contain very small amounts of Mg2+ and
Fig. 2. Rietveld plot for G1 grey OPC with peaks labelled.
A´.G. de la Torre et al. / Nucl. Instr. and Meth. in Phys. Res. B 238 (2005) 87–91
90
Table 1 Rietveld QPA of HRSXRPD data for the analysed samples Phases/wt.%
W1
W2
G1
G2
RM-8488
C3S(1) C3S(2) C3S_total C2S C3A C4AF NaK3(SO4)2 K2SO4 MgO Al2O3 Al2O3 (weighed) Amorphous ˚3 V C3S(1)/A ˚3 V C3S(2)/A
51.7 (1) – – 19.7 (2) 2.4 (1) – – – – 26.2 (1) 21.99 20.6 (5) 120.91 (1) –
55.3 (1) 2.0 (1) 57.3 (1) 8.5 (2) 6.0 (1) – – 0.6 (1) – 27.6 (1) 24.34 15.6 (5) 120.87 (1) 119.90 (1)
32.1 (2) 16.7 (2) 48.8 (3) 10.9 (2) 6.3 (1) 10.6 (1) 1.5 (1) – 0.37 (4) 21.55 (7) 20.33 8.8 (4) 120.42 (1) 119.94 (1)
23.8 (2) 26.6 (2) 50.4 (3) 8.3 (1) 6.7 (1) 9.3 (1) 1.5 (1) – 0.43 (3) 23.40 (7) 21.79 7.1 (4) 120.43 (1) 119.93 (1)
35.6 (2) 11.5 (3) 46.3 (4) 10.7 (2) 2.6 (1)_2.5 (1)a 11.8 (1) – – – 25.3 (1) 23.75 8.0 (5) 120.68 (1) 120.13 (1)
The weighed amounts of a-Al2O3, the derived non-diffracting contents and the volumes (V/Z) for C3S are also given. a RM-8488 contains two polymorphs of C3A, cubic (2.6%) and orthorhombic (2.5%). Table 2 Elemental composition of each cement sample measured by X-ray fluorescence Sample
CaO/wt.%
SiO2/wt.%
Al2O3/wt.%
Fe2O3/wt.%
MgO/wt.%
SO3/wt.%
K2O/wt.%
Na2O/wt.%
W1 W2 G1 G2 RM-8488
68.74 69.16 65.42 63.88 66.50
22.84 21.92 20.99 21.20 22.68
4.53 3.70 5.26 5.89 4.90
0.57 0.29 3.17 3.67 4.07
0.60 0.60 1.24 1.62 0.98
1.03 0.4 1.15 1.43 0.31
– – 1.10 1.03 0.35
– – 0.41 – 0.11
consequently the volumes of C3S of these samples are larger, see Table 1. Furthermore, alite phase coexistence, if present, is very small (W2 clinker contains 2.0 (1)% of the C3S phase with the smallest volume). Grey clinkers (G1 and G2) have higher contents of Mg2+ and they display overall smaller volumes, see Table 1. The amount of the ˚3 alite phase with the smallest volume (119.9 A per formula unit) is higher for the analysed grey clinkers, see Table 1. Additionally, they have a small amount of MgO as free periclase. RM-8488 clinker also shows two alites with different volumes and it contains less Mg2+, see Table 2, and no periclase is present in the pattern. These results indicate that both alite phases contain magnesium cations. However, the Mg2+-content in the smallest volume phase is higher. The superstructures of these two coexisting alites should be different but we are using the same structural description in the analyses since there are no alternative structures available. More synthetic and structural
studies are needed to properly understand and describe the structural role of the impurities (Mg2+, Al3+, etc.) in the alite(s). 3.2. HRSXRPD and LXRPD studies The Rietveld analytical results given in Table 1 (and those shown in Fig. 1) indicates that the alite phase coexistence is very pronounced in grey clinkers which contain a high amount of magnesium. This is the first direct evidence of alite phase coexistence in commercial OPC which has been possible due to the extreme high resolution of the ID31 synchrotron powder diffraction data. The influence of the alite phase coexistence in the Rietveld analysis of LXRPD data remains to be fully studied. Preliminary results indicate that laboratory data with Cu Ka1,2 radiation do not show any signs of phase coexistence. The analyses are good when using a unique structural description for alite. However, when strictly monochromatic
A´.G. de la Torre et al. / Nucl. Instr. and Meth. in Phys. Res. B 238 (2005) 87–91
Cu Ka1 radiation is used, some signs of phase coexistence are evident in the patterns that need to be taken into account. A full comparison of accurate HRSXRPD analytical results and the mineralogical contents from LXRPD data will be reported elsewhere.
91
Acknowledgment We thank the ESRF for the provision of X-ray synchrotron beam time. References
3.3. Non-diffracting content analysis All clinkers were mixed with a-Al2O3 in order to indirectly obtain the amorphous contents [15]. This quantification is obtained from the small overestimation of the standard in the Rietveld analyses, see Table 1. Hence, the particle statistics must be very good and HRSXRPD is much better at providing that than LXRPD. Although more analyses are needed, the studied samples indicate that white Portland clinkers contain approximately 15–20 wt.% of non-diffracting/amorphous content. These values for grey OPC clinkers are much smaller and closer to 7–8 wt.%. These results are in agreement with the manufacturing processes and other reports [2,4]. The amount of liquid phase in grey-OPC kilns is larger and the particles are better crystallised (low non-diffracting contents), whereas the amount of liquid phase in white-clinker kilns is smaller and so particles grow containing more defects which leads to an increase of the non-diffracting content.
[1] R.A. Young, in: R.A. Young (Ed.), The Rietveld Method, IUCr Monographs on Crystallography-5, Oxford University Press, 1993. [2] A.G. De la Torre, A. Cabeza, A. Calvente, S. Bruque, M.A.G. Aranda, Anal. Chem. 73 (2001) 151. [3] A.G. De la Torre, M.A.G. Aranda, J. Appl. Crystallogr. 36 (2003) 1169. [4] P.M. Suherman, A.V. Riessen, B. Oconnor, D. Li, D. Bolton, H. Fairhurst, Powder Diffr. 17 (2002) 178. [5] V. Peterson, B. Hunter, A. Ray, L.P. Aldridge, Appl. Phys. A 74 (2002) S1409. [6] G. Walenta, T. Fullmann, Powder Diff. 19 (2004) 40. [7] T. Fullmann, G. Walenta, ZKG Int. 56 (2003) 45. [8] R. Schmidt, T. Feldman, World Cement (2003) 64. [9] N. Lundgaard, E.S. Jons, World Cement (2003) 59. [10] P.S. Whitfield, L.D. Mitchell, J. Mater. Sci. 38 (2003) 4415. [11] N.V.Y. Scarlett, I.C. Madsen, Powder Diffr. 16 (2001) 71. [12] M. Enders, ZKG Int. 56 (2003) 54. [13] A.C. Larson, R.B. Von Dreele, General Structural Analysis System, Los Alamos National Lab. Rep. No. LA-UR86-748, 1994. GSAS program. Available from: