Hydration of calcium aluminates and calcium

Cement and Concrete Research 47 (2013) 43–50

Contents lists available at SciVerse ScienceDirect

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

Hydration of calcium aluminates and calcium sulfoaluminate studied by Raman spectroscopy David Torréns-Martín a,⁎, Lucia Fernández-Carrasco a, Sagrario Martínez-Ramírez b a b

Department of Architectural Technology I and Center for Research in Nanoengineering, Universitat Politècnica de Catalunya, Barcelona Tech, Spain Instituto de Estructura de la Materia (IEM-CSIC), Spain

a r t i c l e

i n f o

Article history: Received 18 June 2012 Accepted 29 January 2013 Keywords: Cements (D) Hydration (A) Aluminates Spectroscopy (B)

a b s t r a c t Raman spectroscopy has been used to follow the hydration of the main calcium aluminate phases present in calcium aluminate cement (CAC) and calcium sulfoaluminate cement (CSA) clinkers, i.e. C3A, CA, C12A7, CA2, C4AF and C4A3S. We investigate the reaction products induced by hydration on these six compounds. Spectra of anhydrous pure samples and pastes hydrated for 48 h were recorded. In order to contrast the Raman analysis results, the samples were also characterized by XRD and FTIR techniques. Hydration of calcium aluminates led to the formation of C3AH6, C2AH8 and aluminum hydroxide, and hydration of ferrite phases led to hydrogarnet phases. Meanwhile the hydration of C4A3 S led to the formation of ettringite and AFm phases. The Raman spectra analysis developed gives the details of the vibration of the different functional groups present in the calcium aluminate hydrated samples and our results show the potential of Raman spectroscopy in the study of the aluminate hydration on the cement chemistry. The product identity was confirmed by XRD and infrared spectroscopy. © 2013 Elsevier Ltd. All rights reserved.

1. Introduction The composition of calcium aluminate cements within ordinary Portland cements (OPCs) is significantly different from the CaO–SiO2–Al2O3 ternary system even though it contains oxides of calcium, silicon, aluminum and iron [1]. Due to the requirements of refractory concretes, CACs are produced with a wide range of alumina (Al2O3) contents from around 40–80%. The calcium aluminates are obtained by having lime and alumina reacting at high temperature. For the lower aluminum grades (40–60% Al2O3), which are generally made by melting bauxite and limestone in a reverbatory furnace, the principal reactive phase is CA and it is responsible for properties of the material. Other mineralogical phases appear in minor amounts and are C2S (belite), C2AS (gehlenite), and ferrite solid solutions [2], meanwhile the refractory cements contain phases such as C12A7, CA and CA2 [3]. The calcium sulfoaluminate cement (CSA) is produced from gypsum, bauxite and limestone at 1300 °C. These starting materials lead to a final clinker based on the quinary system CaO–SiO2–Al2O3–Fe2O3–SO3 and are formed by three main minerals: C4A3S (yeelimite), C2S and CS (anhydrite), other minor phases such as C3A, C4AF (brownmillerite), C12A7 (mayenite) and C2AS can also be present [4]. The hydration of CAC cements gives principal stable phases such as C3AH6 and gibbsite [5]; however, depending on the temperature some unstable calcium aluminate phases can be developed such as CAH10 ⁎ Corresponding author. E-mail address: [email protected] (D. Torréns-Martín). 0008-8846/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.cemconres.2013.01.015

and C2AH8. On the other hand, the ettringite is the main crystalline phase formed during the hydration of CSA cement, characterized by a hexagonal structure with the structure built up of hexagonal columns of Ca6[Al2(OH)12·24H2O] 6+ oriented along the z-axis; channels rise in the inter-column space, filled with sulfate groups and crystallization water [6]. Moreover, in the hydration process of this cement, other minor hydrates such as the AFm phases can be involved. In terms of characterization, Raman spectroscopy has been reported in a limited number of papers for the investigation of the hydration of cements [7–9] and proved to be an appropriate technique for studying cement hydration; however due to fluorescence for the anhydrous phases, it can be the responsible of anomalous results. This fluorescence is due to the presence of trace impurities, most probably rare earth elements, and was firstly explained by Richardson et al. [10]. For solid samples the micro-Raman technique, allows obtaining spectra for surface sample just a few microns in depth, without fluorescence problems. Spectra can be obtained rapidly and no sample preparation is needed. The Raman spectroscopy was also used to study the structure of C–S–H gel. Kirkpatrick et al. [11] assigned the bands of a synthetic C–S–H gel and showed its similarity to the tobermorite structure. Garbev et al. [12] used a 633 nm laser to study the structural variations of C–S–H gel when changing the C/S ratio. They saw that the changes in the ratio lead systematic changes in their Raman spectra, from which structural information can be ascertained. In the hydration of Portland cement phases, Ibáñez et al. [13] studied the hydration and carbonation of C3S and C2S and confirmed the fast and slow hydration of C3S and C2S, respectively. Black et al. [14]

44

D. Torréns-Martín et al. / Cement and Concrete Research 47 (2013) 43–50

investigated the hydration of C3A and C4AF in the presence and absence of sulfate. They found a band corresponding to C3AH6 and carboaluminate. The sulfate phases were studied for Renaudin et al. [15]. They investigated a synthetic ettringite and monosulfoaluminate with microRaman. They focused their work in the vibrational group sulfate and OH and saw a shift in the ν1 sulfate signal from 990 cm−1 in the ettringite to 982 cm−1 in the monosulfoaluminate. In the OH group the shift in the vibrations were from 3638 cm−1 in the ettringite to 3688 cm−1 in the monosulfoaluminate. They also studied the structural changes in the ettringite when having a partial substitution of Al 3+ for Fe2+. In other more recent research, Mesbah et al. [16] detected the substitution of Cl− and CO32− in the AFm phases by means of Raman spectroscopy. When considering the analysis by Raman of other cements as calcium aluminate and calcium sulfate cements, there is no much research done especially concerning the hydrates. McMillan et al. [17] made a Raman study of calcium aluminate glasses and crystals. They explained that the CaAl2O4 glass spectrum may be interpreted in terms of a fully polymerized network of tetrahedral aluminate units, which is depolymerized on addition of CaO component analogous to binary silicate systems. Another important work of Gastaldi et al. [18] used Raman spectroscopy for a calcium sulfoaluminate cement clinker hydration study. On the hereby research, we present a very early stage hydration of C3A, CA, C12A7, CA2, C4AF and C4A3 S by using Raman spectroscopy analysis. To understand the Raman spectra of the hydrated phases, we pay special attention to the Raman spectra of anhydrous. The results we found show that Raman spectroscopy is a very valuable technique for the detection of these hydrated phases. To corroborate the interpretation of Raman spectra, all of the samples were also analyzed by X-ray diffraction and infrared spectroscopy.

scans were recorded to improve the signal-to-noise ratio. Correct calibration of the instrument was verified by measuring the Stokes and anti-Stokes bands and checking the position of the Si band at ± 520.6 cm − 1. Furthermore, checks either by collection on different sample portions or by averaging several measurements on different points always gave similar results. 3. Results and discussion 3.1. X-ray diffraction analysis

2. Materials and methods

X-ray characterization of the powders indicated that our powder was the required pure calcium aluminate or calcium aluminate sulfate phases. Table 1 shows the ICCD patterns (International Centre for Diffraction Data) used for the identification of detected phases in the samples. The corresponding patterns of hydrated samples are depicted in Fig. 1. Fig. 1 shows the diffraction patterns of CA, CA2, C12A7, C4AF, C4A3 S and C3A hydrated samples. In the patterns we can see a total hydration for the cubic-C3A powder to obtain the cubic C3AH6. Different hydration degrees were seen for CA, CA2 and C12A7 phases, and the diffraction lines due to the anhydrous phases were observed. Nonetheless, we found higher hydration of the C12A7 phase than CA. Both aluminates, CA and C12A7 presented crystalline C3AH6 and C2AH8. In the case of the CA2 hydrated sample we detected very weak diffraction lines due to the cubic C3AH6. As can be seen in Fig. 1, the main hydration product for C4AF was C3AH6 and still some C4AF un-hydrated was also present. Black et al. [14] found the same products in the C4AF hydration. In the case of C4A3 S hydration, anhydrous phase was identified together with ettringite and two forms of AFm phases. These results in the C4A3S hydration are the same with those found by Winnefeld and Lothenbach [20]. According with the literature, there is probably amorphous AH3 which is not detectable for XDR.

2.1. Materials

3.2. Infrared spectroscopy analysis

CA, C12A7, CA2, cubic-C3A, C4AF, and C4A3S were prepared by high temperature synthesis and subsequent grinding, by using the appropriate stoichiometric mixtures of CaCO3 and Al2O3. In the case of C4A3S also the adequate proportion of CaSO4 was used. The aluminate hydrates were obtained by mixing the appropriate anhydrous powder and water with a water/cement weight ratio equal to 0.5. After mixing with water, the pastes were stored at room temperature (21 °C) for a period of 2 days in a CO2-free atmosphere with a relative humidity of 100%. Afterwards, the pastes were treated with a solvent exchange method (acetone and ethanol) in order to stop the hydration progress [19].

Table 2 shows the main absorption band and the identification of the functional group signals for the studied pure anhydrous phases CA, C12A7, CA2, cubic-C3A, C4AF, and C4A3 S. As it was described by Tarte [21], for aluminate phases, two different regions can be identified: the first one in the region 900–750 cm−1 that was assigned to stretching vibrations of a lattice of interlinked AlO4 tetrahedra; and the second one, the absorption in the 500–400 cm−1 area, which can be due to bending mode of the AlO4 lattice [21,22]. The spectra are shown in Fig. 2. The FTIR spectra of the hydrated phases are shown in Fig. 3. The spectra of hydrated CA, C12A7 and CA2 showing similar absorption bands indicate the same hydration products for these phases. In the absorption region due to the O\H groups (4000–3000 cm−1), these spectra had two sharp bands sited at 3532 and 3467 cm−1 due to AH3 [32] and a shoulder at 3625 cm−1 probably due to C2AH8. Additionally, the FTIR spectra of CA and C12A7 hydrated samples, had a medium sharp band at 3670 cm−1 due to C3AH6 [23]. Moreover, a small band at 1030 cm−1 was assigned to Al\OH vibrations and two broad bands about 796 and 532 cm−1 due to bending Al\O bonds. The spectra of hydrated C3A and C4AF, both were similar with a sharped band about 3667 cm −1 due to C3AH6 and two broad bands at 796 and 535 cm −1 due to stretching vibration of Al\O bonds. The relative intensity was higher for the hydrated C3A. Additionally the spectrum of C4AF had bands at 3635 (weak), 3536 (sharp) and

2.2. Analytical methods X-ray diffraction analysis was performed using a Philips PW-1700 powder diffractometer (Bragg-Brentano geometry) equipped with a secondary graphite monochromator (CuKα12, flat sample) and standard operating conditions of 40 kV and 50 mA; the step side used was 0.020°, 1 s per step, in the 5–60 2θ range. The FTIR spectroscopy analysis was conducted in a NICOLET 6700 Thermo Scientific spectrometer with a detector DGTS CsI and 32 scans were recorded to register each sample. The samples were prepared by mixing 1 mg of sample with 300 mg of KBr. The scans were taken in the mid-infrared region at frequencies of 4000 cm−1 to 400 cm−1, with a spectral resolution of 4 cm−1. Micro-Raman measurements were performed at room temperature using a Renishaw inVia Raman microscope equipped with a Leica microscope, a CCD camera, and laser at 532 nm with 10 mW laser powder. Typical spectra from 90 to 4000 cm −1 were recorded with a resolution of 4 cm −1. The time acquisition was 10 s and 5

Table 1 Patterns of ICCD found in the cement phases. Phase Pattern

C3A 38–1429

C4AF 30–226

CA 70–134

C12A7 9–413

CA2 23–1037

C4A3S 33–256


45

Fig. 1. XRD patterns of hydrated phases at 48 h A1:CA, A2:C12A7, A3:CA2, c:C3AH6, H:C2AH8·ff:C4AF, Y:C4A3S, m1:3CA·CaSO4·12H, m2:3CA·CaSO4·14H, e:ettringite.

3417 cm −1 (weak and broad) due to O\H bonds of AH3. The spectrum of hydrated C4AF presented the ν1 [(Fe,Al)O45−], as significant absorption band, sited between 800 and 830 cm −1 with maxima close to 720 cm −1. A broad and less intense band with several maximums between 620 and 670 cm −1 was also present. Compared with the XRD patterns (Fig. 1), the results are similar excepting for the

Table 2 The main absorption band and its identification for the pure anhydrous phases. vw=very weak; s=strong; m=medium; b=broad; w=weak; sh=shoulder; vs= very strong. Cubic-C3A

CA

C12A7

CA2

C4AF

C4AS

941 sh 921

Stretching AlO4 898 863

3.3. Raman spectra analysis of anhydrous and hydrated powders

868

843 w 817

840 820 804

787

788 sh 780

762 sh

764 sh 723 686 573 542 vw

862 846

859 w 844 813

817

798

Bedding AlO4 521 508 467 vw 456

415

779 684 sh 688 640

574 540

461 448 425 sh 417

442 422 406

[SO]42− [(Fe,Al)O45−]

presence of AH3. In XRD it is not visible probably due to the amorphous state of this phase. In the zone of the high wavenumbers, when in the spectrum of hydrated C4A3 S, we can see several signals: towards 3632 cm−1 stretching vibrations of O\H free present in the ettringite [24]. There was another band about 3529 cm−1 due to AH3 and a broad band about 3471 cm−1 due to O\H [32]. This phase is not detected for XRD. Such a fact shows the amorphous nature of this species. It has a stretching vibrations of S\O centered in 1120 cm−1 due to ettringite. This band has a shoulder in 1170 cm−1 due to the presence of the monosulfoaluminate. There is a band to 1027 cm−1 due to Al\OH vibrations and several bands in the interval 800–400 cm−1 (789, 539 and 424 cm−1) due to Al\O bonds.

1191 1097 856 809 774 715 605

w vw sh s

3.3.1. Anhydrous phases Fig. 4(a) shows the Raman spectrum of the CA phase. There is a strong and sharp at 522 cm−1 with a weak band in 546 cm−1 and both are due to ν1 AlO45− group vibrations. The broad band at 793 cm−1 is due to ν3 AlO45− vibrations. At lower frequencies (below 300 cm−1) there exists a group of bands being the higher about 144 cm−1; these signals are due to Ca\O bonds. In the case of C3A (Fig. 4(b)) the spectrum shows two sharp bands at 508 and 757 cm−1. The C3A structure is based on isolated six-membered rings of AIO4 tetrahedra [25]. These rings are formed by corner-sharing of two oxygen per tetrahedron to give a structure with two non-bridged oxygen per AIO4 tetrahedron, defined here as AlO2 units. The strong Raman band at 757 cm−1 may be associated with a symmetric stretching vibration of these AlO2 groups. The weaker Raman bands and the strong infrared absorption in this region may be related to asymmetric aluminate stretching motions. The Raman peak at 508 cm−1 may be associated with symmetric motion of or about the bridged oxygen in the Al\O\Al linkages. These signals are characteristic of cubic C3A [26,8]. The sample also presents a weak band at 360 cm−1 probably due at bond Ca\O. The Raman spectrum of the anhydrous phases from C12A7 is presented in Fig. 4(c). In this spectrum we found two main signals, the first one at 772 cm−1 and the other at 521 cm−1 due to ν3 AlO45− and ν1 AlO45− respectively. The rest of the signals, below 300 cm−1, can be assigned to Ca\O bonds.

46


Fig. 2. FTIR spectra of anhydrous samples.

The Raman spectrum of the CA2 is presented in Fig. 4(d). This spectrum has many signals that can be divided in two groups. The first group is in the interval 640–950 cm−1 with two main bands (medium and sharp) at 661 and 689 cm−1 both are due to ν3 AlO45− vibrations. The second group of bands between 500 and 400 cm−1 has a strong and sharp band at 412 cm−1 and other medium and sharp band at 458 cm−1 both are due to ν1 AlO45− vibrations. It is interesting to show that there is also a small band in 3779 cm−1 due to O\H stretching vibrations. According with McMillan [17] the structure of CaAl2O4 is a fullypolymerized tetrahedral aluminate framework. When the aluminate content increases, there appears depolymerized aluminate tetrahedral. The CA2 presents this case: units of AlO2 that show many signals. The

high content of Al2O3 in C12A7 depolymerizes the aluminate framework in AlO and AlO2 unit. Fig. 4(e) shows the C4A3S Raman spectrum. We can see a very strong and sharp band at 993 cm−1 due to ν1 SO42− vibrations, other small signals at 1201 cm−1 and 614 cm−1 can be due to ν3 and ν4 from SO42− groups respectively [18]. The signals due to AlO45− move to lower frequencies than aluminate phases without sulfate, then the ν3 mode appears at 645 cm−1 and ν1 mode shifts to 482 cm−1. The C4AF spectrum (Fig. 4(f)) shows broad bands. This fact indicates a poor crystalline phase. It presents two broad bands, the first one around 280 cm −1 due to ν2 and ν4 [(Fe,Al)O45−] or [(Fe,Al)O69−] The second band is towards 770 cm−1 due to [(Fe,Al)O45−] or [(Fe,Al) O69−] [14,27].

Fig. 3. FTIR spectra of hydrated samples at 48 h.


47

Fig. 4. Raman spectrum of anhydrous phases.

3.3.2. Hydrated phases The Raman spectrum of CA is in Fig. 5(a). The spectrum is the same with that obtained from anhydrous phase. The relative intensity and the position of band do not change. Through XRD we saw that the hydration of this phase was poor, and the support of FTIR showed that

his hydration products were C2AH8, and low amounts of C3AH6 and AH3. The only difference between the spectra from the anhydrous and hydrated materials is in the high frequencies in the O\H stretching zone. There are two signals in this zone, one weak and broad about 3533 cm −1. According with Frost et al. [32] this band

48


may correspond to AH3 in gibbsite form. The other signal is a broad band about 3646 cm −1 due to ν1 OH − of C3AH6. The hydrated C3A Raman spectrum is depicted in Fig. 5(b). We can see a sharp band with medium relative intensity at 542 cm −1. According with Black et al. [14,8] this band is due to ν1 AlO45− vibrations from C3AH6. The spectrum presents others bands; a weak and broad band in 779 cm−1, and two weak bands in 334 and 168 cm−1.

Comparing with the anhydrous spectra the first band corresponding to ν3 AlO45− in C3A, and the lower frequencies are due to Ca\O bond. In the 3000–4000 cm−1 zone, it presents one single band in 3650 cm−1 in the zone of O\H stretching vibrations. From XDR results the only crystalline hydrated phase identified was C3AH6, therefore this signal can be attributed to ν1 OH− of C3AH6. Comparing with the synthetic hydrogarnet spectrum [28] we can see that they are the same.

Fig. 5. Raman spectrum of hydrated samples at 48 h.


The spectrum of C12A7 is in Fig. 5(c), presents higher fluorescence. In the XRD and FTIR analyses we could see strong signals of C3AH6. In the Raman spectrum we see a weak band in 538 cm −1 due to ν1 AlO45− of C3AH6 [8,15]. In the zone of O\H vibrations we have two signals. A weak band that is in 3530 cm −1 is due to AH3 [32]. The other signal is a sharp band around 3650 cm −1, this band matches with those found in the hydrated C3A and CA, therefore it is due to C3AH6. The hydration products obtained for CA2 were C2AH8 and AH3. The Raman spectrum for the hydrated CA2 is in Fig. 5(d). The bands of AlO45− disappear and the strong and sharp band in 3779 cm −1 is the only visible signal. The only difference is three weak bands between 3300 and 3680 cm −1. One signal is in 3530 cm −1 of AH3 and the others in 3427 and 3613 cm −1 that could be attributed to C2AH8. In the case of C4A3 S hydrated phase (Fig. 5(e)) we have several compounds in the system, as can be seen in FTIR and XRD, ettringite, AFm-12, AFm-14 (where the numeral indicates the number of molecules of water in the crystal structure) and unreacted C4A3S. Ettringite and monosulfoaluminate Raman spectrum have been studied by several authors in the range 100–2000 cm −1 [8,14,15,18,29–31] with little information in the O\H region [8,14,31], however for all the hydrated samples it is rather important to study the evolution of the water molecules. In previous works [28,31] there are the Raman spectra for synthetic ettringite and monosulfoaluminate (AFm). In the ettringite there are bands at 204, 346 (very weak), 451, 548, 989 (strong) and 1119 (very weak) cm −1. For the AFm the bands are 146, 451, 530, 615 (very weak), 979 (strong), and 1089 (very weak) cm −1. In the higher frequencies, ettringite has a broad band in 3460 cm −1 and other signal in 3641 cm −1. For the AFm the signals are in 3618 and 3692 cm −1. The results published in previous works are similar to our data presented in Fig. 5(d). The broad band towards 467 cm −1 is due to various vibrational modes. It presents the ν1 AlO45− of the C4A3 S, and the ν2 SO42− of the ettringite and AFm phases [8]. In 532 cm −1 there is a weak band due to Al\(OH) stretch of AFm phase. The sulfate phases have a weak band in 616 cm −1 due to ν4 SO42 − and sharp band in 991 due to ν1 SO42− these signals come from a mixture signals of ettringite and AFm. There exists a very weak band in 654 cm −1 due to ν3 AlO45− of unreacted C4A3 S. Between 1061 and 1226 cm −1 there is a broad band with several signals. These signals are ν3 SO42− of different sulfate phases (C4A3S, ettringite, AFm). In the zone of higher frequencies there is a broad signal due to O\H vibrations. We have a broad band towards 3446 cm −1 due to H2O stretching, and this signal has several bands. Also we can see a broad band within 3363–3463 due to gibbsite and bayerite and a weak signal towards 3524 cm −1 due to gibbsite [32]. Other band towards 3630 cm −1 could be due to the overlap between ettringite and AFm. The last band in 3680 cm −1 is due to the AFm. The Raman spectrum of C4AF is showed in Fig. 5(f). We can see several changes compared with anhydrous spectrum. So the ν2 and ν4 of [(Fe,Al)O45−] or [(Fe,Al)O69−] modes are 296 cm −1 and ν1 mode is 691 cm −1. Both are broad bands, this fact can be due to the damping effects of iron in solid solution [27]. The spectrum has two weak signals, one in 520 cm −1 probably due to ν1 AlO45− of C3AH6. The other band is in 1082 cm −1, and this may arise from an iron carbonate. Black et al. [14] showed that this band is due to calcite. There is also a medium band at 3639 cm −1 associated to OH. From XRD results the only phase hydrated is C3AH6, so this signal is probably due to this phase. In Table 3 there are the signals for all the phases (anhydrous, synthetic hydrated and experimental hydrated) where the changes in the signals depending of the phase can be seen. The vibrational modes of sulfate have not shifted in the position but present changes in the intensity. The principal information of this work is in the higher frequency region. We can assign the vibrational modes of OH for each phase. All compounds with C3AH6 have bands in 3650 cm−1. The

49

Table 3 The main Raman band and its identification for all phases studied. vw = very weak; s = strong; m = medium; b = broad; w = weak; sh = shoulder; vs = very strong. C3A

C4AF

Anhydrous phases Al\O\Al 508 Td AlO4 ν1 ν3

757

SO42− ν1 ν3 ν4 [(Fe,Al)O45−] [(Fe,Al)O69−]

C4A3S

CA

CA2

C12A7

482

522 s/546 w

412 s/ 458 661/689 s

521

C2AH8 CA2

C2AH8 C3AH6 C12A7

645

772

993 s 1201 614 280 b 770 b

Hydrated phases Present phases C3AH6 C3AH6 AFm-12, C4AF AFm-14, ettringite C4A3S Al\O\Al 532 w Td AlO4 ν1 521 520 w 467 779 b 654 ν3 2− SO4 ν1 991 1061–1226 ν3 616 ν4 [(Fe,Al)O45−] 296 b 691b [(Fe,Al)O69−] Stretching O\H 3446

C2AH8 C3AH6 CA

529

3427 vw 3533

3530 vw 3613 vw

3530

3639 3646 3650 s

3650 3680

vibration for the AH3 appears towards 3530 cm−1. For the C2AH8 assigned bands are in 3420 and 3610 cm−1. The sulfate phases comparing the synthetic and experimental hydrated phases can be assigned for the ettringite at 3445 and 3641 cm−1. For the monosulfoaluminate phase, the signals appear at 3557, 3620 and 3680 cm−1. 4. Conclusions From the analysis of cement phases by Raman spectroscopy performed in this work, we can draw several conclusions. The different anhydrous phases were characterized by XDR and FTIR, and the phases found in the hydration were confirmed by these techniques. The Raman spectrum obtained shows several differences for the same chemical species. In the case of AlO45− we have seen a change to minor frequencies when the amount of Al2O3 is high in the phase. The presence of sulfate ions overlaps the signals of aluminates, and when it comes to aluminosilicates the signals of silicates are higher but with a lower intensity. In the CA2 and C12A7 phase had bands in the zone of O\H stretching. For the C4AF phase we saw a poor crystalline phase, where the bands ν4 and ν2 overlap. For the hydrated phase, the principal result we can draw is in the zone of O\H stretching. We assigned the different signals ν1 OH − to C3AH6, C2AH8 and AH3. So for the C3AH6 towards 3650 cm −1, and for the C2AH8 near 3613 cm −1, we verify the signal in 3530 cm −1 for the AH3. In the case of sulfate phase we saw a different position for the ettringite and AFm.

50


Acknowledgments This study was funded by the MINECO (Ministerio de Educación y Competitividad) through BIA0767-2008 project. References [1] Karen L. Scrivener, Calcium aluminate cement, in: John Newman (Ed.), Advanced Concrete Technology, Constituent Materials, 2003. [2] B. Touzo, A. Glotter, K.L. Scrivener, Mineralogical composition of fondu revisited, in: R.J. Mangabhai, F.P. Glasser (Eds.), Calcium Aluminate Cements 2001, IOM Communications, London, 2001, pp. 129–138. [3] J. Bensted, Calcium aluminate cements, in: J. Bensted, P. Barnes (Eds.), Structure and Performance of Cements, 2nd ed., Spon Press, London, 2002. [4] W. Lan, F.P. Glasser, Hydration of calcium sulfoaluminate cements, Adv. Cem. Res. 8 (31) (July 1996) 127–134. [5] Karen L. Scrivener, A. Capmas, Calcium aluminate cements, in: Peter CHewlett (Ed.), Chemistry of Cement and Concrete, 2006. [6] H.F.W. Taylor, Crystal-structures of some double hydroxide minerals, Min. Mag. 39 (1973) 377–389. [7] M. Frias, S. Martínez-Ramírez, Use of micro-Raman spectroscopy to study reaction kinetics in blended white cement pastes containing metakaolin, J. Raman Spectrosc. 40 (12) (2009) 2063–2068. [8] L. Black, C. Breen, J. Yarwood, C.-S. Deng, J. Phipps, G. Maitland, Hydration of tricalcium aluminate (C3A) in the presence and absence of gypsum—studied by Raman spectroscopy and X-ray diffraction, J. Mater. Chem. 16 (13) (2006) 1263–1272. [9] D. Gastaldi, F. Canonico, E. Boccareli, Ettringite and calcium sulfoaluminate cement: investigation of water content by near-infrared spectroscopy, J. Mater. Sci. 44 (21) (2009) 5788–5794. [10] I.G. Richardson, J. Skibsted, L. Black, R.J. Kirkpatrick, Characterisation of cement hydrate phases by TEM, NMR and Raman spectroscopy, Adv. Cem. Res. 22 (4) (2010) 233–248. [11] R. James Kirkpatrick, J.L. Yarger, P.F. McMillan, P. Yu, X. Cong, Raman spectroscopy of C–S–H, tobermorite, and jennite, Adv. Cem. Bas. Mater. 5 (1997) 93–99. [12] K. Garbev, P. Stemmermann, L. Black, C. Breen, J. Yarwood, B. Gasharova, Structural features of C–S–H(I) and its carbonation in air—a Raman spectroscopic study. Part I: fresh phases, J. Am. Ceram. Soc. 90 (3) (2007) 900–907. [13] J. Ibáñez, Ll Artús, R. Cuscó, A. López, E. Menéndez, M.C. Andrade, Hydration and carbonation of monoclinic C2S and C3S studied by Raman spectroscopy, J. Raman Spectrosc. 38 (1) (2007) 61–67. [14] L. Black, C. Breen, J. Yarwood, J. Phipps, G. Maitland, In situ Raman analysis of hydrating C3A and C4AF pastes in presence and absence of sulphate, Adv. Appl. Ceram. 105 (4) (2006) 209–216. [15] G. Renaudin, R. Segni, D. Mentel, J.M. Nedelec, F. Leroux, C. Taviot-Gueho, A Raman study of the sulfated cement hydrates: ettringite and monosulfoaluminate, J. Adv. Concr. Technol. 5 (3) (2007) 299–312.

[16] A. Mesbah, C. Cau-dit-Coumes, F. Frizon, F. Leroux, J. Ravaux, G. Renaudin, A new investigation of the Cl−–CO32− substitution in AFm phases, J. Am. Ceram. Soc. 94 (6) (2011) 1901–1910. [17] P. McMillan, B. Piriou, Raman spectroscopy of calcium aluminate glasses and crystals, J. Non-Cryst. Solids 55 (2) (1983) 221–242. [18] D. Gastaldi, E. Boccaleri, F. Canonico, M. Bianchi, The use of Raman spectroscopy as a versatile characterization tool for calcium sulphoaluminate cements: a compositional and hydration study, J. Mater. Sci. 42 (20) (2007) 8426–8432. [19] J. Zhang, G.W. Scherer, Comparison of methods for arresting hydration of cement, Cem. Concr. Res. 41 (10) (2011) 1024–1036. [20] F. Winnefeld, B. Lothenbach, Hydration of calcium sulfoaluminate cements — experimental findings and thermodynamic modelling, Cem. Concr. Res. 40 (8) (2010) 1239–1247. [21] P. Tarte, Infrared spectra of inorganic aluminates and characteristic vibrational frequencies of AlO4 tetrahedra and AlO6 octahedra, Spectrochim. Acta A Mol. Biomol. Spectrosc. 23A (7) (1967) 2127–2143. [22] L. Fernández-Carrasco, D. Torréns-Martín, L.M. Morales, S. Martinez-Ramírez, Infrared spectroscopy in the analysis of building and construction materials. chapter in: Theophile Theophanides (Ed.), Infrared Spectroscopy — Materials Science, Engineering and Technology, INTECH, 2012, pp. 369–382. [23] L. Fernandez-Carrasco, T. Vazquez, Aplicación de la espectroscopia infrarroja al estudio de cemento aluminoso, Mater. Constr. 46 (241) (1996) 53–65. [24] J. Bensted, An infrared spectral examination of calcium aluminate hydrates and calcium aluminate sulphate hydrates encountered in Portland cement hydration, Alluminati di Calcio, Seminario Internazionale Torino, Italy, 1982. [25] H.G.W. Taylor, Cement Chemistry, Ed. Academic Press, London, 1990. [26] C.S. Deng, C. Breen, J. Yarwood, S. Habesch, J. Phipps, B. Crasterb, G. Maitland, Ageing of oilfield cement at high humidity: a combined FEG-ESEM and Raman microscopic investigation, J. Mater. Chem. 12 (10) (2002) 3105–3112. [27] M. Conjeaud, H. Boyer, Some possibilities of Raman microprobe in cement chemistry, Cem. Concr. Res. 10 (1) (1980) 61–70. [28] D. Torréns-Martín, L. Fernández-Carrasco, S. Martínez-Ramírez, J. Ibañez, L. Artús, T. Matschei, Raman spectroscopy of anhydrous and hydrated calcium aluminates. J. Am. Ceram. Soc. (submitted for publication). [29] K. Jallad, M. Santhaman, D. Menashi, D. Ben-Amotz, Chemical mapping of thaumasite formed in sulfate-attacked cement mortar using near-infrared Raman imaging microscopy, Cem. Concr. Res. 31 (6) (2001) 953–958. [30] S. Sahu, D. Exline, M. Nelson, Identification of thaumasite in concrete by Raman chemical imaging, Cem. Concr. Compos. 24 (3–4) (2002) 347–350. [31] S.K. Deb, M.H. Manghnani, K. Ross, R.A. Livingston, P.J.M. Monteiro, Raman scattering and X-ray diffraction study of the thermal decomposition of an ettringite-group crystal, Phys. Chem. Miner. 30 (1) (2003) 31–38. [32] L. Frost, J. Kloprogge, S. Russell, J. Szetu, Vibrational spectroscopy and dehydroxylation of aluminum (oxo)hydroxides: gibbsite, Appl. Spectrosc. 53 (4) (1999) 423–434.