Incorporation of aluminium in calcium-silicate-hydrates

Cement and Concrete Research 75 (2015) 91–103

Contents lists available at ScienceDirect

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

Incorporation of aluminium in calcium-silicate-hydrates E. L’Hôpital a, B. Lothenbach a,⁎, G. Le Saout a, D. Kulik b, K. Scrivener c a b c

Empa, Laboratory for Concrete & Construction Chemistry, CH-8600 Dübendorf, Switzerland Paul Scherrer Institute, Laboratory for Waste Management, 5232 Villigen PSI, Switzerland EPFL, Laboratory of Construction materials, CH-1015 Lausanne, Switzerland

a r t i c l e

i n f o

Article history: Received 10 September 2014 Accepted 21 April 2015 Available online xxxx Keywords: C-S-H (B) Aluminium KOH (D) Characterization (B) Kinetics (A)

a b s t r a c t Calcium silicate hydrate (C-S-H) was synthetized at 20 °C to investigate the effect of aluminium uptake (Al/Si = 0–0.33) in the presence and absence of alkalis on the composition and the solubility of a C-S-H with a Ca/Si equal to 1.0. C-S-H incorporates aluminium readily resulting in the formation of C-A-S-H at Al/Si ≤0.1. At higher Al/Si ratios, in addition to C-A-S-H, katoite and/or stratlingite are present. Aluminium is mainly taken up in the bridging position of the silica dreierketten structure, which increases the chain length. The aluminium uptake in C-S-H increases with the aqueous aluminium concentrations. The presence of potassium hydroxide leads to higher pH values, to the destabilisation of stratlingite and to higher dissolved aluminium concentrations, which favours the aluminium uptake in C-S-H. Potassium replaces partially the calcium ions on the surface and interlayer, thus leading to more negative surface charge and to shortening of chain length. © 2015 Elsevier Ltd. All rights reserved.

Contents 1. 2.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Material and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Solid phase analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Solution analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Thermodynamic modelling . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Kinetics of C-A-S-H formation . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Influence of aluminium on C-S-H . . . . . . . . . . . . . . . . . . . . . . . 3.2.1. Solids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.2. Aqueous concentrations . . . . . . . . . . . . . . . . . . . . . . . 3.2.3. Correlation between dissolved aluminium and aluminium uptake in C-S-H 3.3. Influence of KOH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1. Solids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.2. Aqueous concentrations . . . . . . . . . . . . . . . . . . . . . . . 3.3.3. Influence of high pH values on the aluminium uptake . . . . . . . . . 3.3.4. Comparison of synthetic C-S-H with C-S-H in cement . . . . . . . . . . 4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix A. Samples studied . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix B. Dissolved concentrations . . . . . . . . . . . . . . . . . . . . . . . . . Appendix C. Saturation indices . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

⁎ Corresponding author. E-mail address: [email protected] (B. Lothenbach).

http://dx.doi.org/10.1016/j.cemconres.2015.04.007 0008-8846/© 2015 Elsevier Ltd. All rights reserved.

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . .

92 92 92 92 93 93 94 94 94 94 96 97 98 98 100 100 101 101 101 101 101 101 102

92

E. L’Hôpital et al. / Cement and Concrete Research 75 (2015) 91–103

1. Introduction Partial substitution of Portland cement by supplementary cementitious materials (SCM), such as fly ash, slag or calcined clays, offers the potential to reduce CO2 emissions. However such systems inevitably have different chemistry, form different hydrates than Portland cements and will therefore perform and behave differently in concrete. In the presence of silica-rich SCM, the Ca/Si ratio of the C-S-H drops from ≈1.5–1.9 in Portland cements [1] to 1.5 or lower in cements blended with slag, fly ash or metakaolin [2–4] while the Al/Si ratio may increase up to about 0.25 [3]. These changes in chemistry modify the structure and can affect the performance, mechanical properties and durability of the system. The composition and properties of low Ca/Si C-S-H phases are significantly different from those of C-S-H formed in Portland cement systems [5,6] with different capacity for the incorporation of alkalis or other ions. Calcium silicate hydrate (C-S-H) is the main hydration product in cement and it contributes significantly to the physical, chemical and mechanical properties. The C-S-H present in Portland cements is poorly crystalline with a variable composition and can contain other ions such as aluminium or alkalis within its structure [5,6]. C-S-H can be described as calcium (hydr)oxide layers with silica chains attached on both sides organized in a dreierketten structure, a repeating chain of three silica tetrahedra (see Fig. 1). Two of these silica tetrahedra are linked to the calcium oxide layer (pairing tetrahedra), while the third one, the bridging tetrahedron, is linked to two pairing tetrahedra. In a C-S-H particule, a few sheets are connected by an interlayer containing water, calcium, alkalis and other ions. The solubility of C-S-H as a function of the Ca/Si ratio can be described by a family of solubility curves depending on the synthesis method, the equilibration time and the temperature history as shown in Chen et al. [7], which indicates, at least after shorter equilibration times, the existence of several metastable C-S-H phases. Several NMR studies have investigated the incorporation of aluminium in C-S-H. The substitution of aluminium occurs primarily as tetrahedrally coordinated Al(IV) in the bridging position (Q2b) [9–12], as indicated in Fig. 1. In tobermorite, Al(IV) is reported to be present also in Q3 sites which link two dreierketten chains together [9,11]. At higher Ca/Si ratios, the relative amount of Al(IV) in C-S-H decreases, instead more octahedrally coordinated Al(VI) is observed [10, 11,13,14]. The octrahedrally coordinated Al(VI) has been suggested to represent either an amorphous aluminium hydroxide or a calcium aluminate hydrate at the C-S-H surface: third aluminium hydrate (TAH)

Q2u

Q2p(1Al)

-

H H

Q1

Q2p

Interlayer

Dreierketten chain Calcium oxide layer Dreierketten chain Q2b

Q3(1Al)

Interlayer

Fig. 1. Schematic structure of C-A-S-H. Grey circle: calcium ion; empty circle: species in the interlayer (water or alkali); light grey tetrahedra: SiO−, 4 dark grey tetrahedra: AlO4; −: negative charge (compensated by proton, calcium or other cations). Qn(mAl): n indicates the numbers of Si neighbours and m the number of aluminium neighbours, b: bridging position, p: pairing position. Adapted from [8].

[15] or in the interlayer within the C-S-H [13]. In addition to Al(IV) and Al(VI), approximately 10% of the aluminium associated with C-SH is present as pentacoordinated Al(V) regardless the Ca/Si ratio of the C-S-H [10,13,14]. Sun et al. [11] suggested that Al(V) and Al(VI) could compensate the negative charge introduced by the replacement of Si(+ IV) by Al(+ III) in the silica chain. However, the amount of Al(V) and Al(VI) does not correlate with the amount of Al(IV) in the silica chains [10,13]. NMR studies indicate that the presence of alkali hydroxides increases the uptake of aluminium in synthetic C-S-H [14] and white Portland cement [16]. While there is a number of studies focused on the effects of aluminium uptake on the C-S-H structure [10,11,13–15], very little data is available on the relations between the dissolved aluminium, calcium, silicon and hydroxide concentrations in the aqueous (pore) solution and the aluminium uptake in C-S-H, with the exception of a recent paper of Pardal et al. [17], where relatively high aluminium concentrations (1 mmol/L) and short reaction times (1 day) have been used. The lack of systematic experimental data at equilibrium conditions hinders the development of adequate thermodynamic models which would allow prediction of the aluminium uptake in C-S-H. In this paper, C-S-H with a constant Si/Ca = 1.0 is investigated. The effect of aluminium at Al/Si ratios from 0 to 0.33 is studied in the first part and the effect of alkalis (potassium hydroxide) on C-S-H and C-AS-H are considered in the second part. The solubility, the structure and the chemical composition of the resulting C(-A)-S-H gel are investigated after equilibration times of 6 months or longer. 2. Material and methods 2.1. Synthesis Calcium oxide (CaO), silica fume (SiO2) and monocalcium aluminate (CA: CaO.Al2O3) were used to synthesize C(−A)-S-H. CaO was obtained by burning calcium carbonate (CaCO3, Merck, pro analysis) at 1000 °C for twelve hours. SiO2, provided by Aerosil 200, Evonik, was chosen for its high specific surface area of 200 m2/g. CA (CaO.Al2O3) was used due to high reactivity [18]. CA was synthesized from CaCO3 and Al2O3 (Sigma Aldrich).The homogenized powder mixture was heated for 1 hour at 800 °C, 4 hours at 1000 °C and 8 hours at 1400 °C and cooled with a rate of 600 °C/h. The resulting solid was ground to a Blaine surface area of 3790 cm2/g. The CA contained 99.1% CA and 0.9 % C12A7 as determined by X-ray diffraction and Rietveld refinement analysis using X'Pert HighScore Plus. C(−A)-S-H samples were prepared by adding a total of 2 g of CaO, SiO2 and CA to 90 ml of Milli-Q water (water/solid = 45). The proportions of CaO, SiO2 and CA were varied to obtain C-A-S-H with a Ca/Si ratio of 1.0 and Al/Si ratios from 0 to 0.33 as indicated in Appendix A. The C-S-H was equilibrated in 0.5 M KOH solutions to mimic alkali concentrations relevant for Portland cements [19,20] and to study the effect of alkali on aluminium uptake. Synthesis and all sample handling were carried out in a N2 filled glovebox to minimize CO2 contamination. The samples were stored in 100 mL PE-HD containers placed on a horizontal shaker moving at 100 rpm and equilibrated at 20 °C. For each equilibration time a separate sample was prepared. After different equilibration times, the solid and the liquid phase were separated by filtration using 0.45 μm nylon filter, and subsequently analysed. 2.2. Solid phase analyses After filtration, the solid was washed with approximately 30 mL of a 50–50% water-ethanol solution, afterwards with approximately 30 mL of pure ethanol, dried for seven days by freeze drying and then stored until analysis in N2 filled desiccators in the presence of saturated CaCl2 solutions (≈ 30% RH) and NaOH pellets as CO2 trap. The solid

E. L’Hôpital et al. / Cement and Concrete Research 75 (2015) 91–103

phases were analyzed by thermogravimetric analysis (TGA), X-ray diffractometry (XRD) and Rietveld analysis, as well as by 29Si and 27Al nuclear magnetic resonance (NMR). TGA data were recorded with a TGA/SDTA851e Mettler Toledo. The weight loss of the samples was recorded from 30 °C up to 980 °C with a heating rate of 20 °C/min under nitrogen atmosphere. The total water bound in C-A-S-H was quantified from the total water loss between 30 and 550 °C. The amount of portlandite is quantified from the water loss around of the peak around 450 °C and the AH3 from the water loss around of the peak around 215 °C. Other solids such as stratlingite and katoite were not quantified by TGA as the weight loss of C-A-S-H and the different aluminium containing phases overlapped. X-ray diffraction (XRD) data were collected using a PANalytical X’Pert Pro MPD diffractometer in a θ–2θ configuration using incident beam monochromator and CuKα radiation (λ = 1.54 Å) with a fixed divergence slit size 0.5° and a rotating sample stage. The phase composition was identified with X’Pert HighScore Plus. The method of external standard (G-factor approach) using CaF2 was used to quantify the crystalline phases in the samples [21,22]. The mass attenuation coefficients of the samples were determined by mass balance (initial composition (Appendix A)–composition of the solution (Appendix B)) taking into account the amount of bound water determined by TGA. The structure models used for the Rietveld refinement are ICSD 94630 for the katoite [23] and ICSD 69413 for the stratlingite [24]. The 29Si NMR measurements were performed at room temperature using a Bruker Avance 400 MHz NMR spectrometer with a 7 mm CP/ MAS probe. The 29Si MAS NMR single pulse experiments were recorded at 79.49 MHz using the following parameters: 4500 Hz spinning speed, 9200 scans, π6 pulses of 2.5 ms, 20 s relaxation delays and RF field strengths of 33.3 kHz during TPPM decoupling sequence. The chemical shifts of the 29Si MAS NMR spectra were referenced to an external sample of tetramethylsilane (TMS). The observed 29Si resonances were analysed using the Qn(mAl) classification, where a Si tetrahedron is connected to n Si tetrahedral with n varying from 0 to 4 and m is the number of neighbouring AlO4 tetrahedra. Decompositions were carried out using the dmfit software [25]. The peak shape were constraint with a Lorentzian/Gaussian ratio = 0.5, full width at half height ≤ 3 ppm, constant chemical shifts for each peak, and constrained peak amplitudes, as described in [26]. The structure of tobermorite (Fig. 1) was respected by keeping 2 pairing tetrahedra (Q2p) for one bridging tetrahedron (Q2b: silica tetrahedron in bridging and Q2u: a bridging tetrahedra connected to another bridging SiO2 tetrahedra by hydrogen bonding [27]): Q2p/(Q2b + Q2u) = 2. The main chain length (MCL) and the Al/Si ratio were calculated as shown below.

MCL ¼


 3 2 Q 1 þ Q 2p þ Q 2b þ Q 2u þ Q 2 ð1AlÞ 2

Q1 1 2 Q ð1AlÞ Al 2 ¼ 1 2 Si Q þ Q p þ Q 2b þ Q 2u þ Q 2 ð1AlÞ

With Q1: end of chain and Q2(1Al): pairing silica tetrahedra neighbouring an aluminium in bridging position. The 27Al NMR measurements were performed at room temperature with a 2.5 mm CP/MAS probe. The 27Al MAS NMR single pulse experiments were recorded at 104.26 MHz using the following parameters: π pulses of 1 ms without 1H 20 kHz spinning speed, 4000 scans, 12 decoupling and 1 s relaxation delays. The chemical shifts of the 27Al MAS NMR spectra were referenced to 1.0 mol/L AlCl3-6H2Osolution at 0.0 ppm.

93

2.3. Solution analyses The elemental concentrations of calcium, silicon and aluminium in the filtrates were determined with a Dionex DP ICS-3000 ion chromatograph. Each sample was diluted by a factor 1, 5, 10, 100 and/or 1000 depending on the ion concentration and measured in duplicate. Standards from 0.1– 50 mg/L were used. The relative error of the measurements was ≈ 10%. To quantify the hydroxide concentration, pH measurements were made at room temperature with a Knick pH meter (pH-Meter 766) equipped with a Knick SE100 electrode. The pH electrode was calibrated against potassium hydroxide solutions of known concentrations (from 0.0001–0.5 mol/L) in order to minimize any error due to the type of alkali ion. The measurements were corrected to correspond to 20 °C (analytical error ±0.1 pH unit). The zeta ζ-potential was measured with an acoustophoresis electroacoustic method using a Zeta Probe from Colloid dynamics. The calibration was made with potassium tungstosilicates, KSiW. The measurements were carried out on unfiltered C-A-S-H samples equilibrated for 42 days prior to measurement under nitrogen atmosphere. During the measurement, the sample was continuously agitated in order to avoid any sedimentation at the bottom of the measurement cell. A second measurement was carried out on the solution (after filtration) to determine any interferences due to ions present in solution, which is considered as background and thus deducted from the initial measurement.

2.4. Thermodynamic modelling Thermodynamic modelling was carried out using the Gibbs free energy minimization program GEM-Selektor [28,29]. GEM-Selektor is a geochemical modelling code which computes equilibrium speciation in aqueous electrolyte solution, as well as the kind and amount of solids precipitated. Chemical interactions involving solids, solid solutions, aqueous electrolyte and gas phase are considered simultaneously. The thermodynamic data for aqueous species as well as for many solids were taken from the GEMS version of the PSI-Nagra thermodynamic database [30,31], while the solubility products for cement minerals were taken from the cemdata07 database [32,33] updated with the recent solubility products of calcium aluminate hydrates and hydrogarnets [34]. To model C-S-H, the downscaled CSHQ model recently suggested by Kulik [35,36] is used. The CSHQ model uses four different endmember C-S-H compositions (C1.5S0.67H2.5, C1.33SH2.17, C0.83S0.67H1.83 and C0.67SH1.5) which allow covering the complete range of Ca/Si ratios observed in C-S-H. The CSHQ model is able to reproduce the changes in silicate chain length and in the aqueous concentrations with changing Ca/Si ratio in the solid C-S-H because it is based on a sublattice concept that describes the different structural units in C-S-H and their possible variations. It considers i) the coupled replacement of the bridging tetrahedra (SiO2) and two 2H+ in the adjacent interlayer space by H2O and Ca2+ and ii) the replacement of vacant interstitial sites by Ca(OH)2. The CSHQ model as given in [35] calculates, based on an ideal solid solution between the end-members, compositions of the solid C-S-H from a Ca/Si = 0.67 to approximately Ca/Si = 1.6 in the presence of portlandite. Calculation of saturation indices from the concentrations determined in the pore solutions offers the possibility of assessing independently which solid phases can form or will dissolve from a thermodynamic point of view. The saturation index (SI) with respect to a solid is given by log(IAP/KS0), where the ion activity product (IAP) is calculated from activities derived from the measured concentrations in the solution and KS0 corresponds to the solubility product of the respective solid. A saturation index near 0 indicates that the phase is at or near equilibrium with the solution; while a negative saturation index indicates undersaturation and that the phase should not form or dissolve. Conversely, a positive saturation index indicates oversaturation and that the phase should precipitate.

94

E. L’Hôpital et al. / Cement and Concrete Research 75 (2015) 91–103

The activities of the different species were calculated from the measured total concentrations at the respective temperatures using GEMSelektor. The activity coefficients of aqueous species γi were computed with the built-in expanded extended Debye-Hückel equation in Truesdell-Jones form with common parameter ai of 3.67 for potassium hydroxide solutions:

logγi ¼

pffiffi −Ay z2i I pffiffi þ by I 1 þ By ai I

ð1Þ

where zi denotes the charge of species i, I the effective molal ionic strength, by is a semi-empirical parameter (by ≈ 0.123 for KOH at 25 °C), and Ay and By are P,T-dependent coefficients. This activity correction is thought to be applicable up to about 1 M ionic strength [28].

3. Results and discussion

Fig. 3. XRD measurement of C-A-S-H (Al/Si = 0.1) as a function of time (C: C-S-H, K: katoite), CuKα.

3.1. Kinetics of C-A-S-H formation 3.2. Influence of aluminium on C-S-H To ensure that near equilibrium conditions are attained, long equilibration times up to 2 years are used and the evolution of the aqueous concentrations as function of time is followed. The dissolved concentrations in Fig. 2 show that in the first months, calcium and silicon concentrations decrease indicating the formation of a more stable and thus less soluble C-S-H phase. Little variation in the calcium and silicon concentrations is observed after 3 months indicating that a (meta)stable equilibrium is reached. The samples containing aluminium show significant changes in the solution composition up to 6 months indicating that equilibrium is attained much slower. Little difference is observed between 182 days and 2 years. Fig. 3 shows the diffractogram of C-A-S-H after different equilibration times. After 182 days, katoite is present in addition to C-A-S-H. Katoite seems to have formed initially: its quantity decreases with time as illustrated in Fig. 3 for the C-A-S-H with a Al/Si = 0.1. This indicates a destabilisation of the katoite and an uptake of aluminium in a thermodynamically more stable C-A-S-H phase with time. The intensity of the d002 peak at ~7° 2θ, which relates to the interlayer distance of C-S-H, sharpens with time. This could indicate more regular interlayer distances after longer equilibration times.

3.2.1. Solids The amount of aluminium present affects the composition of the C(A−)S-H phase. At Al/Si ratio ≤ 0.05 only C-A-S-H is detected by TGA (Fig. 4). At Al/Si ≥0.1, in addition to C-A-S-H, katoite and stratlingite precipitate. At Al/Si = 0.33 aluminium hydroxide is also formed. Fig. 5 shows the diffractogram of the C-S-H with different aluminium contents, which confirms the presence of stratlingite and katoite also observed by TGA (Fig. 4). As silica is present, katoite could contain some silica in its structure called hydrogrossular (C3ASyH6-2y). In this case, the XRD pattern is shifted higher 2θ values [37,38]. Fig. 5 show that only silica free katoite (C3AH6) precipitates. For low Al/Si ratios (≤ 0.1), a decrease of the position of the d002 of C-A-S-H is observed from ≈ 7.5° 2θ to around 6.9° 2θ in presence of aluminium. This diffraction peak around 7° 2θ is associated with the basal spacing of the C-S-H and indicates an increase of the interlayer distance from 11.9 Å (Al/Si = 0) to 12.7 Å (Al/Si = 0.05) in the presence of aluminium in C-S-H. Renaudin et al. [39] observe for synthetic C-(A)-S-H phases an even more distinct increase of the interlayer distance from 11.7 in the absence of aluminium to 14.4 Å at Al/

0.02

100 Weight loss

Weight loss (%)

0.00 60 Derivative weight loss -0.01

C(-A)-S-H

40

-0.02

Katoite

Al/Si=0 Al/Si=0.05 Al/Si=0.1 Al/Si=0.2 Al/Si=0.33

Al-hydroxide

20

Stratlingite 0

Fig. 2. Evolution of the dissolved aluminium, calcium and silica concentrations as a function of time during the precipitation of C(−A)-S-H. Empty symbols: Al/Si = 0; filled symbols: Al/Si = 0.2. Diamont: calcium, silicon: square, triangle: aluminium. Measurement error ±10%.

50

100

150

200

250

300

350

400

450

500

550

Temperature (°C) Fig. 4. TGA of C-A-S-H after an equilibration time of 182 days.

-0.03

-0.04 600

Derivative weight loss (%/°C)

0.01

80

E. L’Hôpital et al. / Cement and Concrete Research 75 (2015) 91–103

95

Table 1 C-A-S-H composition in molar ratios determined from mass balance (Rietveld analysis, mass balance and TGA (bound water), AH3 was quantified with TGA, H2O/Si at 30%RH). Constant compositions are considered for stratlingite (C2ASH8) and katoite (C3AH6). TAH is considered as a part of C-S-H.

Fig. 5. XRD of C-A-S-H after an equilibration time of 182 days. Si-poor hydrogossular have a composition range from C3AH6 to C3AS0.42H5.16 and Si-rich hydrogrossular from C3AS0.76H4.48 to C3AS0.99H4.02 [37]. C: C-S-H, S: stratlingite, K: katoite.

Si = 0.1, indicating a structural modification of the C-S-H network by the incorporation of aluminium. Based on quantities initially present, the aqueous concentrations (see below) and the amount of crystalline phases determined by Rietveld analysis for katoite and stratlingite and TGA for AH3, the molar ratios in the C-A-S-H are calculated and compiled in Table 1. The water content of C-S-H depends strongly on the drying conditions. For the C-S-H (Ca/Si = 1.0) without aluminium equilibrated at 30% relative humidity, 1.3 ± 0.3 H2O per silica are observed and stay relatively constant regardless the aluminium content (Table 1). A comparable value, 1.7 H2O per silica, has been observed by Renaudin et al. [39] for Ca/Si = 1.0 after drying over silica gel. The addition of aluminium leads to a decrease of Ca/Si and Ca/ (Si + Al) ratio in the C-S-H, consistent with the increase of the silica chain length as observed by 29Si NMR as discussed below. The 29Si NMR data (Fig. 6 and Table 2) indicate for the aluminium free C-S-H four different silica environments. Previous studies assigned the different environments (e.g. [4,7,11,14,40,41]) as the end of chain Q1 at − 79.4 ppm, the bridging silica Q2b at − 83.4 ppm within the “dreierketten”, the pairing site Q2p at −85.3 ppm occurring in synthetic C-S-H and a bridging tetrahedra connected to another bridging silica tetrahedra by hydrogen bonding, Q2u at 88.5 ppm [27,42,43]. At Al/ Si = 0.2, in addition to the Q2p the presence of stratlingite leads to a signal at −85.6 ppm [44]. The presence of aluminium in C-S-H introduces an additional peak, Q2p(1Al) between −82.0 and −80.5 ppm, previously observed by [10,11,16,45,46]. The 29Si NMR signal of pairing silica, Q2p, is shifted by a tetrahedrally coordinated aluminium, Al(IV) present in the neighbouring bridging position of the “dreierketten” chains and forms Q2p(1Al). The signal of Q2p(1Al) increases in intensity in the presence of more aluminium indicating the uptake of more aluminium in bridging position; the relative intensity of the end of chain (Q1) decreases while Q2b and Q2u remain approximately constant as shown in Table 2. The increasing occupancy of bridging tetrahedra by Al(IV) decreases the Ca/ (Si + Al) ratio and leads to a higher degree of polymerisation of the silica tetrahedra and thus a longer silica chain length, as previously reported [11,45]. The 27Al-NMR spectra recorded on C-A-S-H samples (Fig. 7) show the presence of IV-, V- and VI-fold coordinated aluminium, in agreement with literature [11,13,14,47]. The 27Al NMR confirms the uptake

Time

C-S-H composition (molar ratio)

Interlayer

Other solids (wt%) ± 5%

Days

Ca/Si⁎

distance (A) ±0.2

C3AH6

C2ASH8

AH3

-

-

-

Ca/(Si + Al)⁎

Al/Si⁎

H2O/Si ±0.3

Al/Si = 0 28 0.99 91 0.99 182 0.98 546 0.98

0.99 0.99 0.98 0.98

-

1.2 1.2 1.5

Al/Si = 0.01 91 0.98 182 0.98

0.97 0.97

0.010 0.010

1.2 1.4

-

-

-

Al/Si = 0.02 91 0.98 182 0.98

0.96 0.96

0.020 0.020

1.6 1.4

-

-

-

Al/Si = 0.03 91 0.97 182 0.97

0.95 0.95

0.030 0.030

1.2 1.3

-

-

-

Al/Si = 0.04 91 0.97 182 0.97

0.93 0.93

0.040 0.040

1.3 1.5

-

-

-

Al/Si = 0.05 91 0.99 182 0.99

0.94 0.94

0.050 0.050

1.3 1.4

12.7

-

-

-

Al/Si = 0.1 28 91 0.93 182 0.95 546 0.99

0.88 0.88 0.90

0.062 0.078 0.100

1.2 1.5 1.5

12.4

5 2.8 -

-

-

Al/Si = 0.2 28 0.92 91 0.87 182 0.94 546 0.80

0.85 0.83 0.85 0.75

0.080 0.052 0.106 0.059

1.7 1.0 1.6 1.5

-

1.7 4.9 5.9 14.1

14.1 15.2 13 3.1

-

Al/Si = 0.33 28 0.79 91 0.83 182 0.88 546 0.79

0.77 0.73 0.84 0.76

0.023 0.14 0.048 0.030

1.1 1.1 1.5 1.4

-

4.4 7.1 7.2 4.8

30.5 16.1 25 28.6

1.8 1.3 2.4 2.3

11.9 12.3

12.4

⁎ Error ±0.01 in the absence of other phases, ±0.1 in the presence of the other phases. -: not observable, “no value”: not determined.

of aluminium in the silica chain by the presence of Al(IV) at −60 ppm. Little Al(V) characteristic of dissolved aluminium in the interlayer and Al(VI) are observed. At a target Al/Si = 0.1 the AlIV/Si from 29Si NMR data is with 0.06 somewhat lower than the total Al/Si ratio from mass balance of 0.1, consistent with the presence of some Table 2 Relative fractions (±2) of Qn, AlIV/Si ratio and mean chain length (MCL) obtained by 29Si MAS NMR. Total Al/Si ratio in C-S-H (from mass balance). Al/Si Q1 total −79.4 ±0.5

0 0.01 0.02 0.03 0.05 0.1

Q2p

Q2b

Q2u

Q2(1Al)

−85.3 ±0.5

−83.4 ±0.5

−88.5 ±0.5

−81.7 ±0.5

ppm

ppm

ppm

ppm

ppm

38 32 23 26 24 24

42 43 51 45 44 43

21 20 21 19 19 18

2 5 4 3 4

3 5 6 10 12

−: not observed.

AlIV/Si Total ±0.01 Al/Si ±0.01

MCL ±2

0 0.01 0.02 0.03 0.05 0.06

5 6 9 8 9 9

0 0.010 0.020 0.030 0.050 0.1

96

E. L’Hôpital et al. / Cement and Concrete Research 75 (2015) 91–103

Fig. 6. 29Si NMR of C-A-S-H (a: 91 days, b: 1.5 years): A. with different Al/Si ratios, B. deconvolution of the sample with Al/Si = 0.1.

Al (IV)

Al(V) and Al(VI). No clear indications of the presence of stratlingite (chemical shift at 61 and 18 ppm [44]) are observed, in agreement with the XRD data, although minor amounts might be present. Traces of katoite are observed for Al/Si = 0.05 (chemical shift at 12.4 ppm [48]) but the amount is probably below the detection limit of TGA and XRD which is the raison why it was not observed previously.

TAH Al (VI) Al (V) Al/Si=0.1

Al/Si=0.05 140

120

100

80

60

40

20

ppm

0

-20

-40

-60

-80

b

a

-100

Fig. 7. 27Al NMR of C-A-S-H equilibrated for 1 year days for Al/Si = 0.1 and 91 days for Al/ Si = 0.04 and 0.05.

3.2.2. Aqueous concentrations In Fig. 8, the dissolved concentrations are shown as a function of the total Al/Si ratio. An increase of the Al/Si ratio leads to higher dissolved aluminium concentrations up to an Al/Si ratio of 0.1 (Fig. 8 and Appendix B). The dissolved aluminium concentrations represent less than 0.2% of the total aluminium in the sample meaning that the aluminium is mainly bound in the C-A-S-H. Above Al/Si = 0.1, no further increase of the aluminium concentration is observed. At Al/Si = 0.33 the aluminium concentration is lower due to the consumption of aluminium by aluminium containing phases such as katoite, stratlingite and AH3 observed previously by XRD and TGA. The presence of aluminium also increases the silicon concentrations and decreases calcium and hydroxide concentrations. The measured

E. L’Hôpital et al. / Cement and Concrete Research 75 (2015) 91–103

Fig. 8. Measured concentrations as a function of the total Al/Si ratio in the experiments after 91 days (empty symbols), 182 days (grey symbols) and 1.5 years (black symbols) equilibration at 20 °C.

dissolved concentrations cannot yet be directly compared to the results of thermodynamic modelling, as presently we are missing a model that describes aluminium uptake in C-S-H considering the different aluminium binding sites (this is the work in progress). Thus, the measured data are compared to thermodynamic predictions using a simplified solid solution model for C-S-H (CSHQ model from [35]) as shown in Fig. 9. At Ca/Si = 1.0, aluminium is taken up mainly in the “dreierketten” chains, thus the uptake of aluminium is expected to have in a first approximation a similar effect as the presence of more silica. Hence, the aqueous concentrations at different Ca/Si ratios are calculated and plotted as function of pH to ease comparison with the measured data. In the C-SH model, an increase of silica corresponds to longer silica chains due to an increase of the fraction of bridging tetrahedra, to lower pH values and calcium concentrations and higher silicon concentrations. As the

Fig. 9. Evolution of the dissolved concentrations and Ca/(Si + Al) ratio of the solid as a function of pH, compared with the changes predicted by thermodynamic modelling for aluminium-free C-S-H upon variation of the Ca/Si ratio (CSHQ model from [35]). Empty symbols: C-S-H without aluminium, filled symbols: aluminium-containing C-S-H: after 28, 91, 182 and 540 days: calcium (diamonds) and silicon (squares). Modelling: calcium: black line, silicon: light grey line and Ca/Si: dotted black line.

97

samples with more silica (modelling) and more aluminium (experimental data points) show the same trends, it corrobates the idea that aluminium is taken up mainly in the “dreierketten” chains, which leads to longer mean chain length, as observed by the decrease of the Q1 intensity by 29Si NMR, and a lower Ca/(Si + Al). The good agreement between the measured effect of aluminium and the calculated effect of silica shows that the uptake of aluminium in the bridging tetrahedra stabilises C-S-H; this effect is comparable to the stabilising effect of silica. Based on the measured dissolved concentrations, saturation indices (SI) (Appendix C) were calculated. The saturation indices show that the solutions are slightly undersaturated with respect to low Ca/Si C-S-H (0.67 CaO·SiO2·1.5H2O) and clearly undersaturated with respect to portlandite. At higher Al/Si ratios, the SI of low Ca/Si C-S-H increase slightly indicating that the presence of aluminium affects the C-S-H solubility and lowers the Ca/(Si + Al) in the C-S-H. At Al/Si N 0.1 the solutions are less undersaturated with respect to stratlingite and microcrystalline AH3. Stratlingite is observed by XRD and TGA at Al/ Si ≥ 0.2, AH3 is observed only at Al/Si = 0.33. The solutions are clearly undersaturated with respect to hydrogrossular (endmembers: C3AS0.84H4.32 and C3AS0.41H5.18) and even stronger with respect to katoite (C3AH6), although katoite has been identified by XRD and TGA at Al/Si ≥ 0.1. It seems that katoite precipitates initially at Al/Si ≥ 0.1, and dissolves slowly at Al/Si = 0.1 between 182 days and 1.5 years, but not at higher Al/Si ratio. The persistence of katoite in highly undersaturated solution over a long period of time could indicate a kinetic hindrance of dissolution in the presence of silica.

3.2.3. Correlation between dissolved aluminium and aluminium uptake in C-S-H Comparison of the Al/Si ratio in C-S-H and the dissolved aluminium concentrations (Fig. 10) shows a clear increase of aluminium uptake in C-S-H with the aluminium concentrations. Equilibration time between 91 and 546 days had no significant influence. A similar correlation between dissolved aluminium and aluminium uptake in C-S-H is observed for much higher aluminium concentrations (up to 3.5 mmol/L for Ca/Si = 0.95) under metastable conditions by Pardal et al. [17]. This continuous increase of the dissolved aluminium concentrations and the variation of Al/Si ratio in C-S-H would be consistent with the formation of a continuous solid solution between C-S-H and C-A-S-H end-members.

Fig. 10. Aluminium uptake in C-S-H (determined by mass balance) as a function of the aluminium concentrations. 91 days (empty square), 182 days (grey triangle), 546 days (dark circle). The experimental error based on the error of the different methods (including the error associated with Rietveld quantification) is indicated. The high errors are due to the presence of stratlingite and katoite in these samples.

98

E. L’Hôpital et al. / Cement and Concrete Research 75 (2015) 91–103 100

A) 28 days

80 0.00

Weight loss (%)

70 60

-0.01 50

Portlandite 40

Katoite

-0.02

30

Al/Si=0 Al/Si=0.1 -0.03 Al/Si=0.2 Al/Si=0.33

20 10 0 100

200

300

400

500

Derivative weight loss (1/°C)

0.01

90

-0.04 600

Temperature (°C) 100

B) 182 days

80

Weight loss (%)

0.00

60

-0.01

Portlandite

-0.02

40

Al/Si=0 Al/Si=0.05 Al/Si=0.1 Al/Si=0.2 Al/Si=0.33

Katoite 20

-0.03

-0.04

0 100

200

300

400

500

600

Temperature (°C) Fig. 11. TGA of C-A-S-H after an equilibration times of 28 (A) and 182 days (B).

Derivative weight loss (1/°C)

0.01

Table 3 C-A-S-H composition in molar ratios in presence of 0.5 mol/L of potassium hydroxide determined from Rietveld analysis, mass balance and TGA. Fixed compositions are considered for strat: stratlingite, kat: katoite and mono: monocarboaluminate. Portlandite quantified be TGA. TAH is considered as a part of C-S-H. Time C-S-H composition (molar ratio) days

Ca/Si⁎ Ca/(Si + Al)⁎

Al/Si⁎ H2O/Si ±0.3

Interlayer distance (Å) ±0.2

Other solids (%wt) ±5% Kat

Strat

CH

Mono

Al/Si = 0 182 1.00

1.00

-

1.32

11.6

-

-

-

-

Al/Si = 0.05 182 0.99

0.95

0.048 1.24

12.5

-

-

-

trace

Al/Si = 0.1 182 0.87

0.83

0.059 1.20

13.0

3.30

-

3.45 2.60

Al/Si = 0.2 182 0.95 364 0.97

0.84 0.83

0.13 0.17

1.29 1.26

12.7 -

5.00 2.80

trace trace -

Al/Si = 0.33 182 0.84 364 0.94

0.69 0.75

0.22 0.25

1.26 1.59

13.2 -

13.20 trace 3.90 3.40 -

2.60 -

-

⁎ Error ±0.01 in the absence of other phases, ±0.1 in the presence of the other phases. −: not observed.

3.3. Influence of KOH 3.3.1. Solids The addition of 0.5 mol/L potassium hydroxide leads to the initial formation of portlandite as shown in Fig. 11. The portlandite precipitated initially dissolves slowly over time while C-(A −)S-H with higher Ca/Si ratio forms. After the equilibration time of 182 days or longer, only C-A-S-H is detected for Al/Si = 0.05. At higher Al/Si ratio, katoite precipitates in addition. The presence of potassium hydroxide increases the interlayer distance of C-S-H from 11.9 (at Al/Si = 0, no alkali) to 12.6 Å (at Al/Si = 0, [KOH] = 0.5 mol/L). The uptake of aluminium results also in the presence of potassium hydroxide in a clear widening of the C-S-H interlayer distance. For an Al/Si = 0.1, the interlayer distance increase from 12.4 in

Fig. 12. XRD of C-A-S-H in presence of 0.5 mol/L of potassium hydroxide after an equilibration time of 182 days. C: C-S-H, S: stratlingite, K: katoïte, A: monocarboaluminate due to CO2 contamination, P: portlandite.

E. L’Hôpital et al. / Cement and Concrete Research 75 (2015) 91–103

alkali free system (Table 1) to 13.0 Å in presence of 0.5 mol/L potassium hydroxide (Fig. 12 and Table 3). This increase of the interlayer distances points towards the presence of alumnium and alkali ions in the interlayer. The ζ-potential of C-S-H decreases in the presence of potassium hydroxide as shown in Fig. 13. The ζ-potential measurement determines the difference in electric potential between the liquid and a stationary layer of fluid around the particles at a certain distance to the surface, the slipping plane. The deprotonation of silanol groups Si-OH0 to N SiO- at the C-S-H surface or in the interlayer results in excess negative charge, which is compensated by calcium ions. At higher calcium concentrations and/or high Ca/Si ratios, sorbed calcium overcompensates the negative charge such that an apparent positive charge is measured (for more detailed explanations see [49]). Labbez et al. [50] reported a decrease of the measured ζ-potential of C-S-H in the presence of NaOH and Na2SO4 in the pH range 10.5–13, as the partial replacement of the divalent calcium ions at the surface by monovalent ions (sodium in [50]) takes place and lowers the apparent positive surface charge. This phenomenon is more distinct at higher alkali to calcium ratios [49,50]. Thus the continuing decrease of the ζ-potential of C-S-H in the presence of increasing potassium hydroxide concentrations (Fig. 13) indicates the partial replacement of calcium ions in the interlayer and on the surface by potassium ions. In the presence of potassium, the 29Si NMR measurements indicate a decrease of the “dreierketten” chain length as visible in the increase of the relative Q1 intensity (Fig. 14 and Table 4). In addition, the presence of potassium hydroxide results in a 1 to 3 ppm chemical shift to less negative values. In agreement with our data, a decrease of the silica chain length [51, 52] and of the chemical shift [52,53] has been observed by different authors. Typically the shielding of silica sites by protons or alkalis is weaker than by calcium, leading to less negative chemical shifts [41,54,55]. Changes in the chemical shift indicate that in the presence of potassium hydroxide, part of the deprotonated silica interacts with the potassium in the interlayer, while the silica fraction bound to calcium ions (Ca2+) in the interlayer decreases in agreement with the decrease of the ζ-potential. The replacement of a part of the calcium in the interlayer by potassium ions leads to more calcium available to form the main calcium oxide layer as the total Ca/Si ratio in the experiments is constant. This relative increase of the calcium

Fig. 13. Zeta potential of C-S-H (Ca/Si = 1.0–Al/Si = 0) at different potassium hydroxide concentrations.

2

Q p(1Al) 1

2

QbQp 2

Q

2

Qu

Al/Si=0.1

[KOH]=0.5M [KOH]=0.25M

Al/Si=0

No alkali

[KOH]=0.5M No alkali

-70

-75

-80

-85

-90 ppm

-95

99

-100 -105 -110

Fig. 14. 29Si NMR of C-A-S-H with Ca/Si = 1 and Al/Si = 0 and 0.1 after 1 year. Dash line CA-S-H (Ca/Si = 1, Al/Si = 0.1) without alkali.

Table 4 Peak position (±0.5), relative fractions of Qn, AlIV/Si ratio (±0.01) and mean chain length (MCL, ±2) obtained by 29Si MAS NMR. Al/Si: total Al/Si ratio in C-S-H (from mass balance). Alkali concentration

Q1

Q2p

Q2b

Q2u

Q2(1Al)

AlIV/Si ±0.01

Al/Si⁎

MCL ±2

ppm

%

ppm

%

ppm

%

ppm

%

ppm

%

Al/Si = 0 No alkali 0.5 mol/L KOH

−79.6 −78.6

38 25

−85.3 −84.7

42 62

−83.4 −82.2

21 13

-

-

-

-

0 0

0 0

5 3

Al/Si = 0.1 No alkali 0.25 mol/L KOH 0.5 mol/L KOH

−79.1 −78.8 −78.4

24 29 46

−84.9 −85.0 −83.8

43 34 28

−83.3 −83.5 −82.2

18 12 10

−88.3 −87.7 −85.5

4 5 4

−81.5 −81.8 −80.9

12 21 12

0.06 0.10 0.06

0.1 0.93 0.059

9 8 5

⁎ Error ± 0.01 in the absence of other phases, ±0.1 in the presence of the other phases, −: not observed.

Absence of alkali

Presence of alkali

Fig. 15. Shortening of the silica chain length caused by the partial replacement of calcium ions by potassium ions in the interlayer at constant Ca/Si ratio. Grey circle: calcium ion; empty circle: alkali in the interlayer; grey tetrahedra: protonated silicate.

100

E. L’Hôpital et al. / Cement and Concrete Research 75 (2015) 91–103

Katoite and Monocarboaluminate

Al (VI)

Al (IV) TAH Al (VI) Al (V) [KOH]=0.5 mol/L

No alkali

140

120

100

80

60

40

20

ppm

0

-20

-40

-60

-80

-100

Fig. 16. 27Al NMR of C-A-S-H with Ca/Si = 1 and Al/Si = 0.1 after 1.5 year for the alkali free sample and 1 year for the sample with 0.5 mol/L KOH. TAH: third aluminium hydrate.

oxide to silica ratio in the main layer of C-S-H results in shorter silica chains as illustrated in Fig. 15. The 29Si NMR data estimate the aluminium in the bridging position over silica ratio (AlIV/Si). A comparison between the AlIV/Si and the total Al/Si in C-S-H obtained by mass balance is shown in Table 4. No significant difference is observed in presence of potassium hydroxide meaning that the aluminium is still mainly taken up in the bridging position. The 27Al NMR data indicate the presence of Al(VI) between 10 and 13 ppm that could be assigned to katoite and monocarboaluminate in agreement with the XRD data (see Fig. 12) and confirms the presence of mainly Al(IV) in C-S-H (Fig. 16), while only minor quantities of Al(VI) in TAH and of Al(V) are observed. This result agrees with the 29 Si NMR results showing that the AlIV/Si ratio is similar to the total Al/Si ratio. The amounts of Al(IV) (bridging position) in C-S-H in the absence and presence of potassium hydroxide are comparable, while slightly less Al(V) and Al(VI) in C-S-H could be present in the samples with alkali. The presence of little Al(V) and Al(VI) in the presence of potassium hydroxide is also illustrated by the similarity of the AlIV/Si ratio obtained from the 29Si NMR data (0.06 and Table 4) and the total Al/Si ratio of 0.06 in C-S-H from the mass balance calculations, which includes Al(IV), Al(V) and Al(VI).

3.3.2. Aqueous concentrations Fig. 17 shows that the presence of alkali hydroxides raises the pH value, a key parameter in the system. The pH increases from 10.5-12 for alkali free system to 13.5 in the system containing potassium hydroxide (Appendix B). This high pH increases the concentrations of silicon and aluminium while the calcium concentration decreases (the measured data are summarised in Appendix B). The saturation indices SI (Appendix C) calcaluted from the measured concentrations show that the solutions are strongly undersaturated with respect to all the possible phases including katoite which has been observed to persist. The present C-S-H model [35] predicts the same changes as observed experimentally but clearly overestimates the silica concentrations at high pH values as shown in Fig. 17. The calculated SI indicates undersaturation with respect to C-S-H at high pH values. This apparent undersaturation with respect to C-S-H indicates that the presence of alkalis affects (lowers) the C-S-H solubility and that further development of thermodynamic

Fig. 17. Evolution of the dissolved concentrations (after 91 days: empty symbols, 182 days: grey symbols and 540 days: dark symbols) as a function of pH compared with the changes predicted by thermodynamic modelling for aluminium free C-S-H upon potassium hydroxide addition (CSHQ model from [35]). Calcium (diamond) and silicon (square). Modelling: calcium: black line, silicon: light grey line and Ca/Si (dotted black line) in C-S-H.

models for C-S-H will have to consider the effect of alkalis more explicitly. 3.3.3. Influence of high pH values on the aluminium uptake The addition of potassium hydroxide clearly increases the aluminium uptake in C-S-H as shown in Fig. 18. In alkali free C-S-H at Al/ Si ≤ 0.05, all aluminium is taken up, while at Al/Si ≥ 0.1, a fraction of the aluminium precipitates as stratlingite and katoite, thus limiting the Al/Si ratio in C-S-H to a maximum of ~ 0.06. In the presence of 0.5 mol/L potassium hydroxide, the maximum Al/Si in C-S-H increased to ~ 0.23. This significant increase of aluminium uptake at higher pH agrees with findings in Portland cements, where a strong increase of the aluminium content in the bridging position of the dreierketten chains in the presence of more alkalis is reported [16]. A plot of the aluminium uptake in C-S-H versus the dissolved aluminium concentrations in Fig. 19 underlines the strong positive correlation between dissolved aluminium and aluminium in C-S-H. The presence of alkalis leads to much higher dissolved aluminium and destabilises

Fig. 18. Al/Si ratio in C-S-H (Ca/Si = 1) as a function of the total Al/Si ratio present in the system at different pH (diamond: alkali free samples, square: samples with potassium hydroxide) at 182 days. Empty symbol: AlIV/Si ratio (from 29Si NMR), filled symbol Al/Si (from mass balance).

E. L’Hôpital et al. / Cement and Concrete Research 75 (2015) 91–103

101

4. Conclusions

Fig. 19. Al/Si ratio in C-S-H as a function of the aluminium present in the solution at 182 days. Diamond: alkali free samples, square: samples with potassium hydroxide). Empty symbol: AlIV/Si ratio (from 29Si NMR), filled symbol Al/Si (from mass balance).

stratlingite and aluminium hydroxide. Thus, the higher aluminium uptake in the presence of alkalis seems to be related rather to higher dissolved aluminium concentrations than to a charge balancing effect of the alkali cations. However, the aluminium uptake in C-S-H is slightly supressed compared to the dissolved concentrations: more dissolved aluminium is needed to reach the same Al/Si ratio as in alkali free C-A-S-H. As explained in Section 3.3.1, the zeta potential measurements have shown that a replacement of calcium ions by potassium ions in the interlayer increases the negative charge of the C-S-H. As the negatively charged alumininate ion AlO-2 predominates at pH N7 [56], the aluminium uptake is not favoured in presence of alkali compared to C-S-H without alkali at similar dissolved aluminium concentration. 3.3.4. Comparison of synthetic C-S-H with C-S-H in cement The aluminium uptake in synthetic C-S-H (this study and [17]) and in Portland cements (also blended with fly ash or silica fume) with Ca/ Si ratios from 0.68 [57] to 2.0 [58] shows a strong correlation with the dissolved aluminium concentrations as shown in Fig. 20. The relation between dissolved aluminium and Al/Si ratio of the C-S-H seems to be quite independent of the Ca/Si ratio as a similar relation is observed for portlandite containing cements, portlandite-free blends and Portland cement with silica fume. Fig. 20 also indicates that the same mechanisms dominate the aluminium uptake both in cements and in synthetic C-S-H.

The incorporation of aluminium leads to modification of the C-S-H structure at Ca/Si = 1.0. Up to Al/Si = 0.05, all aluminium is taken up in C-S-H; less than 0.2% of the total aluminium remains in solution. For higher Al/Si ratios, in addition to C-A-S-H, katoite, stratlingite, AH3 or other AFm phases precipitate, which limit the aluminium uptake in C-S-H to Al/Si ≈ 0.1. The uptake of aluminium in C-S-H results in an increase of the interlayer distance observed by XRD. 29Si NMR and 27AlNMR indicate that the aluminium is mainly taken up in the bridging position of the silica chains as Al(IV), although also minor amounts of Al(V) and Al(VI) are observed in other structural sites. The addition of aluminium influences the solution composition by increasing the silicon and aluminium concentrations, while calcium and hydroxide concentrations decrease. Thermodynamic modelling shows that the variations of measured concentrations in aqueous solution due to the presence of aluminium follow the same trends as C-S-H if more silica is added as aluminium is taken up in the “dreierketten” chains. This study demonstrates that also the presence of alkali hydroxide modifies the C-S-H structure. The increase of interlayer distances and the lowering of zeta potential in the presence of potassium hydroxide show that potassium ions (K+) can replace a fraction of the calcium ions (Ca2 +) in the interlayer and on the surface of the C-S-H. This leads to more calcium in the calcium oxide layers of the C-S-H and thus shorter silica chains. In addition, the presence of potassium hydroxide and the resulting higher pH values destabilise stratlingite, leading to more aluminium uptake in C-S-H and to higher dissolved aluminium concentrations. Acknowledgments The financial support of Swiss National Foundation grand n°130419 is gratefully acknowledged. The authors would like to thank Daniel Rentsch, Salaheddine Alahrache, Angela Steffen, Luigi Brunetti, Rupert Myers, Gilles Plusquellec and Boris Ingold for the helpful advice and support during the measurements.

Appendix A. Samples studied Al/Si ratio studied and mixing proportions used to prepare C-A-S-H. Al/Si

0 0.01 0.02 0.03 0.04 0.05 0.1 0.2 0.33

Initial composition CaO (g)

SiO2 (g)

CaO.Al2O3 (g)

0.966 0.957 0.948 0.939 0.930 0.921 0.879 0.799 0.704

1.034 1.030 1.025 1.021 1.017 1.012 0.991 0.951 0.904

0.000 0.014 0.027 0.040 0.053 0.067 0.130 0.250 0.392

Appendix B. Dissolved concentrations Measured dissolved concentrations after different equilibration time at 20 °C as a function of time and the Al/Si ratio in the presence of C-(A−)S-H (target Ca/Si = 1.0). Target Al/Si ratio

0

Fig. 20. Relationship between dissolved aluminium and Al/Si ratio of C-S-H in Portland cement (PC) [58,59], in Portland-fly ash cement [59,60], in portlandite-free blends [57,59,61,62] and in synthetic C-S-H [10]. Adapted from [62].

Time

Si

Days

(mM)

28 91 182 546

0.106 0.083 0.083 0.059

Ca

Al

OH

pH -

2.8 3.0 3.2 3.6

-

6.5 5.2 3.9 5.3

11.9 11.9 11.7 11.9

(continued on next page)

102

E. L’Hôpital et al. / Cement and Concrete Research 75 (2015) 91–103

Appendix B (continued) (continued) Target Al/Si ratio

0.01 0.02 0.03 0.04 0.05 0.1

0.2

0.33

Appendix C (continued) (continued) Time

Si

Days

(mM)

91 182 91 182 91 182 91 182 91 182 28 91 182 546 28 91 182 546 28 91 182 546

0.13 0.13 0.13 0.13 0.15 0.14 0.14 0.14 0.17 0.15 0.16 0.22 0.11 0.17 1.4 0.43 0.34 0.39 1.8 1.3 0.64 0.7

Ca

2.3 2.3 2.1 2.2 2.3 2.1 2.2 2.2 2.2 2.1 2.5 2.1 2.9 2.0 1.7 1.5 1.1 1.1 1.2 1.3 1.2 1.0

Al

OH

≤0.0037 ≤0.0037 0.0068 0.0061 0.011 0.01 0.016 0.016 0.021 0.02 0.033 0.031 0.031 0.0039 0.0087 0.038 0.023 0.041 ≤0.0037 0.0071 0.053 0.014

4.6 6.0 4.1 5.2 3.9 5.2 3.3 5.0 3.2 5.2 5.8 5.2 4.2 3.0 3.3 3.5 2.0 1.3 0.28 1.3 1.4 1.0

pH

Days

TobH*

TobD*

CH

AH3

Strat

Kat

Hydro

11.8 11.9 11.8 11.8 11.8 11.8 11.7 11.8 11.7 11.8 11.9 11.9 11.7 11.6 11.7 11.7 11.4 11.3 10.6 11.3 11.3 11.2

182 546 28 91 182 546 28 91 182 546

−0.84 −0.61 0.28 −0.23 −0.37 −0.61 0.25 0.18 −0.10 −0.11

−0.66 −0.70 −0.42 −0.64 −0.87 −0.85 −0.83 −0.68 −0.74 −0.84

−2.2 −2.7 −3.5 −3.2 −3.5 −3.6 −4.5 −4.0 −3.7 −3.9

−1.6 −2.3 −1.6 −1.1 −1.2 −0.9 n.a. −1.4 −0.79 −1.3

−1.5 −3.4 −2.2 −1.3 −2.1 −1.6 n.a. −2.6 −1.2 −2.4

−6.0 −9.0 −9.9 −8.1 −9.3 −8.9 n.a. −11.1 −8.9 −10.5

−1.6 −4.1 −3.9 −2.6 −3.7 −3.2 n.a. −4.8 −3.0 −4.5

Measurement error: aqueous silicon, calcium and aluminium concentrations ±10%; pH ±0.1 unit, −: no aluminium in the sample.

0

0.05 0.1

0.2

0.33

Time

Si

Days

(mM)

Ca

91 182 546 91 182 91 182 546 91 182 546 91 182 546

0.36 0.59 0.24 0.76 1.8 0.91 1.6 0.68 0.90 1.0 0.79 1.6 2.1 1.2

Al

K

OH

0.37 0.35 0.75 0.70 0.63 2.7 1.9 2.6 4.8 4.7 5.0

482 415 461 338 463 484 424 453 439 413 447 478 421 449

450 419 311 405 396 434 403 336 434 388 336 434 403 336

13.7 13.6 13.5 13.6 13.7 13.6 13.6 13.6 13.6 13.6 13.6 13.6 13.6 13.6

Measurement error: aqueous silicon, calcium and aluminium concentrations ±10%; pH ±0.1 unit, −: no aluminium in the sample.

Appendix C. Saturation indices Saturation indices after different equilibration time at 20 °C as a function of time and the Al/Si ratio in the presence of C-(A−)S-H (target Ca/Si = 1.0). Target Al/Si ratio Time SI

0

0.01 0.02 0.03 0.04 0.05 0.1

0.33

n.a.: not applicable. TobH: 0.67CaO · SiO2 · 1.5H2O, TobD: 0.83CaO · 0.67SiO2 · 1.83H2O, CH: Portlandite, AH3: microcrystalline AH3, Strat: Stratlingite, Kat: Katoite, Hydro: C3AS0.84H4.32. Saturation indices (Ca/Si = 1, [KOH] = 0.5 mol/L) as a function of time. Target Ca/Si ratio

0

0.05 0.1

pH -

≤0.0025 0.083 0.073 0.032 0.055 ≤0.0025 0.12 0.06 ≤0.0025 ≤0.0025 0.03 ≤ 0.0025 0.01 0.025

0.2

0.2

Dissolved concentrations (Ca/Si = 1, [KOH] = 0.5 mol/L) as a function of time. Target Al/Si ratio

Target Al/Si ratio Time SI

-

Days

TobH*

TobD*

CH

AH3

Strat

Kat

Hydro

28 91 182 546 91 182 91 182 91 182 91 182 91 182 28 91

−0.84 −0.94 −0.96 −1.1 −0.72 −0.74 −0.71 −0.72 −0.68 −0.71 −0.68 −0.68 −0.60 −0.67 −0.65 −0.50

−0.68 −0.73 −0.70 −0.76 −0.69 −0.71 −0.72 −0.71 −0.68 −0.72 −0.68 −0.70 −0.65 −0.71 −0.61 −0.63

−2.2 −2.2 −2.1 −1.9 −2.5 −2.5 −2.6 −2.5 −2.5 −2.6 −2.5 −2.6 −2.6 −2.6 −2.4 −2.7

n.a. n.a. n.a. n.a. n.a. n.a. −2.1 −2.2 −1.9 −1.9 −1.7 −1.7 −1.6 −1.6 −1.5 −1.3

n.a. n.a. n.a. n.a. n.a. n.a. −2.9 −3.0 −2.5 −2.6 −2.1 −2.2 −1.9 −2.0 −1.4 −1.3

n.a. n.a. n.a. n.a. n.a. n.a. −8.2 −8.2 −7.7 −7.9 −7.4 −7.5 −7.3 −7.4 −6.4 −7.0

n.a. n.a. n.a. n.a. n.a. n.a. −3.5 −3.5 −2.9 −3.1 −2.6 −2.7 −2.4 −2.5 −1.7 −2.0

0.33

Time

SI

Days

TobH*

TobD*

CH

AH3

Strat

Kat

Hydr

91 182 546 91 182 91 182 546 91 182 546 91 182 546

n.a. −1.6. −2.0 −1.7 −1.4 n.a. −1.2 −1.7 n.a. n.a. −1.9 n.a. n.a. −1.7

n.a. −0.8 −1.0 −1.1 −0.8 n.a. −0.5 −0.9 n.a. n.a. −1.1 n.a. n.a. −1.1

n.a. −1.2 −1.1 −1.8 −1.5 n.a. −1.2 −1.3 n.a. n.a. −1.7 n.a. n.a. −1.8

n.a. n.a. n.a. −2.3 −2.5 n.a. −2.1 −2.2 n.a. n.a. −1.6 n.a. n.a. −1.3

n.a. n.a. n.a. −3.2 −3.0 n.a. −1.6 −2.4 n.a. n.a. −1.8 n.a. n.a. −1.3

n.a. n.a. n.a. −6.2 −5.8 n.a. −4.1 −4.5 n.a. n.a. −4.4 n.a. n.a. −4.1

n.a. n.a. n.a. −2.7 −2.2 n.a. −0.5 −1.3 n.a. n.a. −1.1 n.a. n.a. −0.7

n.a.: not applicable. TobH: 0.67CaO·SiO·21.5H2O, TobD: 0.83CaO·0.67SiO·21.83H2O, AH3: microcrystalline AH3, Strat: Stratlingite, Kat: Katoite, Hydr: hydrogrossular: C3AS0.84H4.32.

References [1] B. Lothenbach, K. Scrivener, R.D. Hooton, Supplementary cementitious materials, Cem. Concr. Res. 41 (2011) 1244–1256. [2] R. Taylor, I.G. Richardson, R.M.D. Brydson, Composition and microstructure of 20year-old ordinary Portland cement-ground granulated blast-furnace slag blends containing 0 to 100% slag, Cem. Concr. Res. 40 (2010) 971–983. [3] C.A. Love, I.G. Richardson, A.R. Brough, Composition and structure of C-S-H in white Portland cement-20% metakaolin pastes hydrated at 25 degrees C, Cem. Concr. Res. 37 (2007) 109–117. [4] A.V. Girão, I.G. Richardson, R. Taylor, R.M.D. Brydson, Composition, morphology and nanostructure of C–S–H in 70% white Portland cement–30% fly ash blends hydrated at 55 °C, Cem. Concr. Res. 40 (2010) 1350–1359. [5] I.G. Richardson, G.W. Groves, The incorporation of minor and trace elements into calcium silicate hydrates (C-S-H) gel in hardened cement pastes, Cem. Concr. Res. 23 (1993) 131–138. [6] I.G. Richardson, The calcium silicate hydrates, Cem. Concr. Res. 38 (2008) 137–158. [7] J.J. Chen, J.J. Thomas, H.F.W. Taylor, H.M. Jennings, Solubility and structure of calcium silicate hydrate, Cem. Concr. Res. 34 (2004) 1499–1519. [8] P. Faucon, T. Charpentier, A. Nonat, J.C. Petit, Triple-quantum two-dimensional Al-27 magic angle nuclear magnetic resonance study of the aluminum incorporation in calcium silicate hydrates, J. Am. Chem. Soc. 120 (1998) 12075–12082. [9] S. Komarneni, R. Roy, D.M. Roy, C.A. Fyfe, G.J. Kennedy, A.A. Bothnerby, J. Dadok, A.S. Chesnick, Al-27 and Si-29 magic angle spinning nuclear magnetic-resonance spectroscopy of Al-substituted tobermorites, J. Mater. Sci. 20 (1985) 4209–4214. [10] X. Pardal, F. Brunet, T. Charpentier, I. Pochard, A. Nonat, Al-27 and Si-29 solid-state NMR characterization of calcium-cluminosilicate-hydrate, Inorg. Chem. 51 (2012) 1827–1836. [11] G.K. Sun, J.F. Young, R.J. Kirkpatrick, The role of Al in C-S-H: NMR, XRD, and compositional results for precipitated samples, Cem. Concr. Res. 36 (2006) 18–29. [12] I.G. Richardson, A.R. Brough, R. Brydson, G.W. Groves, C.M. Dobson, Location of aluminum in substituted calcium silicate hydrate (C-S-H) gels as determined by 29Si and 27Al NMR and EELS, J. Am. Ceram. Soc. 76 (1993) 2285–2288. [13] G. Renaudin, J. Russias, F. Leroux, C. Cau-dit-Coumes, F. Frizon, Structural characterization of C-S-H and C-A-S-H samples-Part II: Local environment investigated by spectroscopic analyses, J. Solid State Chem. 182 (2009) 3320–3329. [14] P. Faucon, A. Delagrave, J.C. Petit, C. Richet, J. Marchand, H. Zanni, Aluminium incorporation in calcium silicate hydrates (C-S-H) depending on their Ca/Si ratio, J. Phys. Chem. B 103 (1999) 7796–7802.

E. L’Hôpital et al. / Cement and Concrete Research 75 (2015) 91–103 [15] M.D. Andersen, H.J. Jakobsen, J. Skibsted, A new aluminium-hydrate species in hydrated Portland cements characterized by 27Al and 29Si MAS NMR spectroscopy, Cem. Concr. Res. 36 (2006) 3–17. [16] J. Skibsted, M.D. Andersen, The effect of alkali ions on the incorporation of aluminum in the calcium silicate hydrate (C–S–H) phase resulting from Portland cement hydration studied by 29Si MAS NMR, J. Am. Ceram. Soc. 96 (2013) 651–656. [17] X. Pardal, I. Pochard, A. Nonat, Experimental study of Si-Al substitution in calciumsilicate-hydrate (C-S-H) prepared under equilibrium conditions, Cem. Concr. Res. 39 (2009) 637–643. [18] H.F.W. Taylor, Cement chemistry, Thomas Telford Publishing, London, 1997. [19] B. Lothenbach, F. Winnefeld, Thermodynamic modelling of the hydration of Portland cement, Cem. Concr. Res. 36 (2006) 209–226. [20] B. Lothenbach, G. Le Saout, E. Gallucci, K. Scrivener, Influence of limestone on the hydration of Portland cements, Cem. Concr. Res. 38 (2008) 848–860. [21] B.H. O'Connor, M.D. Raven, Application of the Rietveld refinement procedure in assaying powdered mixtures, Powder Diffract. 3 (1988) 2–6. [22] D. Jansen, F. Goetz-Neunhoeffer, C. Stabler, J. Neubauer, A remastered external standard method applied to the quantification of early OPC hydration, Cem. Concr. Res. 41 (2011) 602–608. [23] G.A. Lager, R.T. Downs, M. Origlieri, R. Garoutte, High-pressure single-crystal X-ray diffraction study of katoite hydrogarnet: evidence for a phase transition from Ia3d → I43d symmetry at 5 GPa, Am. Mineral. 87 (2002) 642–647. [24] R. Rinaldi, M. Sacerdoti, E. Passaglia, Strätlingite: crystal structure, chemistry, and a reexamination of its polytpye vertumnite, Eur. J. Mineral. (1990) 841–850. [25] D. Massiot, F. Fayon, M. Capron, I. King, S. Le Calve, B. Alonso, J.O. Durand, B. Bujoli, Z. Gan, G. Hoatson, Modelling one-and two-dimensional solid-state NMR spectra, Magn. Reson. Chem. 40 (2002) 70–76. [26] R.J. Myers, E. L'Hôpital, J.L. Provis, B. Lothenbach, Effect of temperature and aluminium on calcium (alumino) silicate hydrate chemistry under equilibrium conditions, Cem. Concr. Res. 68 (2015) 83–93. [27] H. Sato, M. Grutzeck, Effect of starting materials on the synthesis of tobermorite, MRS Proceedings, Cambridge Univ Press, 1991. 235. [28] D. Kulik, GEMS-PSI 3.0, PSI-Villigen, Switzerland, 2011. (available at http://gems. web.psi.ch/). [29] D. Kulik, T. Wagner, S. Dmytrieva, G. Kosakowski, F. Hingerl, K. Chudnenko, U. Berner, GEM-Selektor geochemical modeling package: revised algorithm and GEMS3K numerical kernel for coupled simulation codes, Comput. Geosci. 17 (2013) 1–24. [30] T. Thoenen, D. Kulik, Nagra/PSI chemical thermodynamic database 01/01 for the GEM-Selektor (V.2-PSI) geochemical modeling code, PSI, Villigen, 2003. (available at http://gems.web.psi.ch/doc/pdf/TM-44-03-04-web.pdf). [31] W. Hummel, U. Berner, E. Curti, F.J. Pearson, T. Thoenen, Nagra/PSI chemical thermodynamic data base 01/01, also published as Nagra Technical Report NTB 02-16, Wettingen, Switzerland, Universal Publishers/uPUBLISH.com, USA, 2002. [32] B. Lothenbach, T. Matschei, G. Moschner, F.P. Glasser, Thermodynamic modelling of the effect of temperature on the hydration and porosity of Portland cement, Cem. Concr. Res. 38 (2008) 1–18. [33] T. Matschei, B. Lothenbach, F.P. Glasser, Thermodynamic properties of Portland cement hydrates in the system CaO-Al2O3-SiO2-CaSO4-CaCO3-H2O, Cem. Concr. Res. 37 (2007) 1379–1410. [34] B. Lothenbach, L. Pelletier-Chaignat, F. Winnefeld, Stability in the system CaO– Al2O3–H2O, Cem. Concr. Res. 42 (2012) 1621–1634. [35] D.A. Kulik, Improving the structural consistency of C-S-H solid solution thermodynamic models, Cem. Concr. Res. 41 (2011) 477–495. [36] B.Z. Dilnesa, B. Lothenbach, G. Renaudin, A. Wichser, D. Kulik, Synthesis and characterization of hydrogarnet Ca 3 (Al x Fe 1− x) 2 (SiO 4) y (OH) 4 (3− y), Cem. Concr. Res. 59 (2014) 96–111. [37] T.G. Jappy, F.P. Glasser, Synthesis and stability of silica-substituted hydrogarnet Ca3Al2Si3-xO12-4x(OH)4x, Adv. Cem. Res. 4 (1991) 1–8. [38] E. Passaglia, R. Rinaldi, Katoite, a new nember of the Ca3Al2(SiO4)3-Ca3Al2(OH)12 series and a new nomenclature for the hydrogrossular group of minerals, Bull. Mineral. 107 (1984) 605–618. [39] G. Renaudin, J. Russias, F. Leroux, F. Frizon, C. Cau-dit-Coumes, Structural characterization of C-S-H and C-A-S-H samples – Part I: Long-range order investigated by Rietveld analyses, J. Solid State Chem. 182 (2009) 3312–3319.

103

[40] F. Brunet, P. Bertani, T. Charpentier, A. Nonat, J. Virlet, Application of 29Si homonuclear and 1H-29Si heteronuclear NMR correlation to structural studies of calcium silicate hydrates, J. Phys. Chem. B 108 (2004) 15494–15502. [41] I. Klur, B. Pollet, J. Virlet, A. Nonat, C-S-H structure evolution with calcium content by multinuclear NMR, in: P. Colombet, A.-R. Grimmer, H. Zanni, P. Soozzani (Eds.),Nuclear magnetic resonance spectroscopy of cement-based materials, Springer, Berlin 1998, pp. 119–141. [42] G. Le Saout, E. Lécolier, A. Rivereau, H. Zanni, Chemical structure of cement aged at normal and elevated temperatures and pressures: Part I. Class G oilwell cement, Cem. Concr. Res. 36 (2006) 71–78. [43] G. Le Saoût, E. Lécolier, A. Rivereau, H. Zanni, Chemical structure of cement aged at normal and elevated temperatures and pressures, Part II: Low permeability class G oilwell cement, Cem. Concr. Res. 36 (2006) 428–433. [44] S. Kwan, J. LaRosa, M.W. Grutzeck, 29Si and 27Al MAS NMR study of stratlingite, J. Am. Ceram. Soc. 78 (1995) 1921–1926. [45] M.D. Andersen, H.J. Jakobsen, J. Skibsted, Incorporation of aluminum in the calcium silicate hydrate (C-S-H) of hydrated Portland cements: A high-field Al-27 and Si-29 MAS NMR Investigation, Inorg. Chem. 42 (2003) 2280–2287. [46] I.G. Richardson, A.R. Brough, R. Brydson, G.W. Groves, C.M. Dobson, Location of aluminum in substituted calcium silicate hydrate (C-S-H) gels as determined by Si-29 and Al-27 Nmr and eels, J. Am. Ceram. Soc. 76 (1993) 2285–2288. [47] X. Pardal, F. Brunet, T. Charpentier, I. Pochard, A. Nonat, 27Al and 29Si solid-state NMR characterization of calcium-aluminosilicate-hydrate, Inorg. Chem. 51 (2012) 1827–1836. [48] J. Rivas Mercury, P. Pena, A. De Aza, X. Turrillas, I. Sobrados, J. Sanz, Solidstate b sup N 27b/sup N Al and b sup N 29 b/sup N Si NMR investigations on Sisubstituted hydrogarnets, Acta Mater. 55 (2007) 1183–1191. [49] C. Labbez, I. Pochard, B. Jönsson, A. Nonat, CSH/solution interface: Experimental and Monte Carlo studies, Cem. Concr. Res. 41 (2011) 161–168. [50] C. Labbez, B. Jönsson, I. Pochard, A. Nonat, B. Cabane, Surface charge density and electrokinetic potential of highly charged minerals: experiments and Monte Carlo simulations on calcium silicate hydrate, J. Phys. Chem. B 110 (2006) 9219–9230. [51] W. Kunther, B. Lothenbach, J. Skibsted, Influence of the Ca/Si ratio of the C-S-H phase on the interaction with sulfate ions and its impact on the ettringite crystallization pressure, Cem. Concr. Res. 69 (2015) 37–49. [52] I. Lognot, I. Klur, A. Nonat, NMR and infrared spectroscopies of C-S-H and Alsubstituted C-S-H synthesised in alkaline solutions, in: P. Colombet, H. Zanni, A.-R. Grimmer, P. Sozzani (Eds.),Nuclear magnetic resonance spectroscopy of cementbased materials, Springer, Berlin Heidelberg 1998, pp. 189–196. [53] H. Viallis, P. Faucon, J.-C. Petit, A. Nonat, Interaction between salts (NaCl, CsCl) and calcium silicate hydrates (CSH), J. Phys. Chem. B 103 (1999) 5212–5219. [54] P. Rejmak, J.S. Dolado, M.J. Stott, A. Ayuela, 29Si chemical shift anisotropies in hydrated calcium silicates: a computational study, J. Phys. Chem. C 117 (2013) 8374–8380. [55] M. Magi, E. Lippmaa, A. Samoson, G. Engelhardt, A. Grimmer, Solid-state highresolution silicon-29 chemical shifts in silicates, J. Phys. Chem. 88 (1984) 1518–1522. [56] W. Stumm, J.J. Morgan, Aquatic chemistry: Chemical equilibria and rates in natural waters, John Wiley & Sons, 2012. [57] T. Bach, C. Coumes, I. Pochard, C. Mercier, B. Revel, A. Nonat, Influence of temperature on the hydration products of low pH cements, Cem. Concr. Res. 42 (2012) 805–817. [58] G. Le Saoût, B. Lothenbach, A. Hori, T. Higuchi, F. Winnefeld, Hydration of Portland cement with additions of calcium sulfoaluminates, Cem. Concr. Res. 43 (2013) 81–94. [59] F. Deschner, B. Lothenbach, F. Winnefeld, J. Neubauer, Effect of temperature on the hydration of Portland cement blended with siliceous fly ash, Cem. Concr. Res. 52 (2013) 169–181. [60] K. De Weerdt, M.B. Haha, G. Le Saout, K. Kjellsen, H. Justnes, B. Lothenbach, Hydration mechanisms of ternary Portland cements containing limestone powder and fly ash, Cem. Concr. Res. 41 (2011) 279–291. [61] B. Lothenbach, G. Le Saout, M. Ben Haha, R. Figi, E. Wieland, Hydration of a low-alkali CEM III/B–SiO2 cement (LAC), Cem. Concr. Res. 42 (2012) 410–423. [62] B. Lothenbach, D. Rentsch, E. Wieland, Hydration of a silica fume blended low-alkali shotcrete cement, Phys. Chem. Earth Parts A/B/C 70 (2014) 3–16.