Barbara Lothenbach 2019, Cemdata18 - A Chemical Thermodynamic Database For Hydrated Portland Cements and Alkali-Activated Materials

&HPHQWDQG&RQFUHWH5HVHDUFK²
Contents lists available at ScienceDirect
Cement and Concrete Research

journal homepage: www.elsevier.com/locate/cemconres
Cemdata18: A chemical thermodynamic database for hydrated Portland 7

cements and alkali-activated materials
Barbara Lothenbacha, , Dmitrii A. Kulikb, Thomas Matscheic, Magdalena Balonisd,
⁎
Luis Baquerizoe, Belay Dilnesaf, George D. Mironb, Rupert J. Myersg,1

a
Empa, Laboratory for Concrete & Construction Chemistry, CH-8600 Dübendorf, Switzerland
b
Paul Scherrer Institut, Laboratory for Waste Management, 5232 Villigen PSI, Switzerland
c
HTW Dresden University of Applied Sciences, Department of Civil Engineering, 01069 Dresden, Germany
d
Department of Materials Science and Engineering, University of California Los Angeles, Los Angeles, CA, USA
e
Lafarge Centre de Recherche, 38291 Saint-Quentin Fallavier, France
f
BASF Schweiz AG, 5082 Kaisten, Switzerland
g
University of Sheffield, Department of Materials Science and Engineering, Sheffield S1 3JD, UK
A R TIC L E INFO A B S TR A C T
Keywords: Thermodynamic modelling can reliably predict hydrated cement phase assemblages and chemical compositions,
Thermodynamic modelling including their interactions with prevailing service environments, provided an accurate and complete thermo-
Cement dynamic database is used. Here, we summarise the Cemdata18 database, which has been developed specifically
Database for hydrated Portland, calcium aluminate, calcium sulfoaluminate and blended cements, as well as for alkali-
Solubility
activated materials. It is available in GEMS and PHREEQC computer program formats, and includes thermo-
C-S-H
dynamic properties determined from various experimental data published in recent years. Cemdata18 contains
thermodynamic data for common cement hydrates such as C-S-H, AFm and AFt phases, hydrogarnet, hydro-
talcite, zeolites, and M-S-H that are valid over temperatures ranging from 0 to at least 100 °C. Solid solution
models for AFm, AFt, C-S-H, and M-S-H are also included in the Cemdata18 database.
1. Introduction The quality of thermodynamic modelling results depends directly on

the accuracy and completeness of the input thermodynamic properties
Numerous studies have shown that chemical thermodynamic mod- of substances and phases, which are usually supplied from a thermo-
elling, coupled with accurate and complete thermodynamic databases, dynamic database. Relevant thermodynamic data for solid cementitious
can reliably predict hydrated cement phase assemblages and chemical substances, such as the solubility products of ettringite or hydrogarnet,
compositions. One of the most interesting aspects of applying thermo- have been compiled in several specific “cement databases” such as (1)
dynamics to hydrated cements has been the discovery that the chemical the Cemdata07 and Cemdata14 databases [1,7–12] (http://www.empa.
compositions of Al2O3-Fe2O3 mono (AFm) and Al2O3-Fe2O3 tri (AFt) ch/cemdata), which are available for GEMS [13,14], (2) the Thermo-
phases are very sensitive to the presence of carbonate [1–3] and tem- ddem (http://thermoddem.brgm.fr/) database [15,16] available for the
perature [4–6], thus demonstrating that these factors may significantly Geochemists Workbench® [17](https://www.gwb.com/) and
modify hydrated cement phase assemblages. Experiments have shown PHREEQC [18] or (3) HATCHES database [19] available for PHREEQC
that compositions of hydrate cement phase assemblages can alter ra- [18]. Data in the first two databases are generally comparable, although
pidly, often within weeks or months, reflecting changing system com- some differences exist, as discussed in more detail in Damidot et al.
positions and temperatures. Thus, thermodynamic calculations and [20]. Our experience applying Cemdata in thermodynamic modelling
experiments support each other: on the one hand, calculations enable applications underlines the importance of a careful data selection and
more complete interpretations of limited experimental datasets and evaluation process, and of including sensitivity analyses into the ana-
help to identify key experiments to perform; and on the other hand, lysis and discussion of results.
experiments provide the data that are needed to validate calculation Additional experimental data, and thermodynamic properties de-
results and model parameters. rived from these data, have become available since the first compilation
⁎
Corresponding author.
E-mail address: barbara.lothenbach@empa.ch (B. Lothenbach).
1
Current address: University of Edinburgh, School of Engineering, Edinburgh, EH9 3FB, UK.
https://doi.org/10.1016/j.cemconres.2018.04.018
Received 31 October 2017; Received in revised form 19 April 2018; Accepted 20 April 2018
$YDLODEOHRQOLQH2FWREHU
(OVHYLHU/WG$OOULJKWVUHVHUYHG
B. Lothenbach et al. &HPHQWDQG&RQFUHWH5HVHDUFK²
of Cemdata07 in 2007/2008 and subsequent compilation of Cemdata14 cementitious systems. The data for hydrotalcite-like phases and detailed
in 2013/2014 [1,7,21]. Cemdata18 provides a significant update to discussions of the different models for C-S-H are given in Sections 2.6
both Cemdata07 and Cemdata14. Cemdata18 is written into a format and 2.7. Standard thermodynamic data for minerals such as calcite,
supporting the GEM-Selektor code [13,14] and is fully compatible with brucite and aqueous and gaseous species already documented in the
the freely available GEMS-Selektor version of the PSI-Nagra 12/07 TDB PSI-Nagra chemical thermodynamic database [22] are not repeated in
[22,23] (http://gems.web.psi.ch/). PSI/Nagra 12/07 TDB [22] contains these tables, but given only in summary tables in Appendices B and D.
the same entries for aqueous species/complexes relevant to cement To enable users to model cementitious systems using the Cemdata18
systems as the PSI/Nagra 01/01 [24], with only slight changes: the dataset with the law of mass action (LMA) geochemical modelling
thermodynamic properties of Si4O8(OH)44− and AlSiO3(OH)43− were package PHREEQC [18], a variant of the Cemdata18 dataset has been
added, while the complex AlSiO(OH)6− was removed. The GEMS ver- generated as documented in Appendix B.
sion of the PSI/Nagra 12/07 TDB includes further changes to the
thermodynamic properties of Al bearing species/complexes and the 2.1. Solubility of Al(OH)3 and its effect on calcium aluminate and calcium
addition of Helgeson-Kirkham-Flowers equation of state parameters to sulfoaluminate cements
account for changes in temperature and pressure [25,26]. Cemdata18
includes a comprehensive selection of cement hydrates commonly en- The solubility of precipitated Al(OH)3 decreases with time. Initially
countered in Portland cement (PC) systems in the temperature range of “amorphous” or poorly ordered Al(OH)3 precipitates with a solubility
0 to 100 °C, including calcium silicate hydrate (C-S-H), magnesium si- product of approximately 0 ± 0.2. With time, the degree of ordering
licate hydrate (M-S-H), hydrogarnet, hydrotalcite-like phases, some increases, and microcrystalline Al(OH)3 forms, while the solubility
zeolites, AFm and AFt phases, and various solid solutions used to de- product decreases to −0.7 after 2 years. The solubility of hydro-
scribe the solubility of these phases. Solubility constants have generally thermally prepared gibbsite is with −1.1 lower as illustrated in Fig. 1,
been calculated based on critical reviews of all available experimental however its formation is not expected within the timeframe of months
data and from additional experiments made either to obtain missing to years generally considered for hydrating cements. At 60 °C and
data or to verify existing data. Additional solubility data were measured above, it is expected that microcrystalline Al(OH)3 does not persist, but
and compiled using temperatures ranging from 0 to 100 °C in many that gibbsite forms relatively fast (Fig. 1). The solubility of Al(OH)3
instances, as documented in [9,12,27,28]. Numerous solid solutions determines whether CAH10 (as in the presence of Al(OH)3 with log
among AFm and AFt phases, siliceous hydrogarnets, hydrotalcite-like KS0 ≥ −0.6 at 25 °C) is formed initially in calcium aluminate cements
phases, C-S-H, and M-S-H have been observed and are included in or whether it converts to C3AH6 and microcrystalline Al(OH)3 [12]. The
Cemdata18. decrease of the solubility of Al(OH)3 with time is also responsible for
Several C-S-H solid solution models, as well as two models for hy- the initial occurrence of CAH10 and ettringite instead of monosulfate
droxide-hydrotalcite are available in Cemdata18. The CSHQ model plus microcrystalline Al(OH)3 in some calcium sulfoaluminate cements,
from [11] and the OH-hydrotalcite end member with Mg/Al = 2 are as discussed in more detail in [53].
well adapted for PC. Although the CSHQ model is able to describe the Which Al(OH)3 modification (see Table 1) should be taken into
entire range of Ca/Si ratios encountered, it is best used for high Ca/Si C- account depends mainly on the timeframe and the temperature con-
S-H, as it still lacks the ability to predict aluminium uptake, which is of sidered. While gibbsite should be allowed to form at temperatures
less importance for Portland cements than for blended cements. For above 60 °C, its precipitation should be suppressed for calculations at
alkali activated binders, the calcium (alkali) aluminosilicate hydrate (C- ambient temperatures, where microcrystalline Al(OH)3 will form in-
(N-)A-S-H) gel model, with lower calcium but higher aluminium and stead. Within very short timeframes (minutes to hour), possibly only
alkali content than in the C-S-H type phase which exists in hydrated PC, amorphous Al(OH)3 should be allowed to precipitate. Similarly, also the
and a Mg-Al layered double hydroxide with variable Mg/Al ratio, are formation of some other stable phases such as goethite (FeOOH), he-
available. matite (Fe2O3) and quartz (SiO2) should be suppressed in calculations
This paper summarises Cemdata18, which includes the most im- of hydrated cements in favour of their more disperse counterparts:
portant additions to the Cemdata07 and Cemdata14 databases in recent microcrystalline FeOOH (or microcrystalline or amorphous Fe(OH)3,
years. It also discusses the relevance and implications of these addi- depending on the timeframe considered), and amorphous SiO2.
tions, and compares Cemdata07 and Cemdata18, accounting for their
main differences. Summaries of the thermodynamic data compiled in 2.2. Thaumasite
the Cemdata18 database are available in formats supported (readable)
by the computer programs GEM-Selektor [13,14] and PHREEQC [18]. Damidot et al. [54] obtained solubility data to derive a solubility
Both of these Cemdata18 variants can be freely downloaded from constant for thaumasite at 25 °C, at which temperature thaumasite was
http://www.empa.ch/cemdata. considered to be stable. Invariant points were calculated for phase as-
semblages including thaumasite in the system CaO–A-
2. Thermodynamic data for cements l2O3–SiO2–CaSO4–CaCO3–H2O. Schmidt et al. [55] used the solubility
data of Macphee and Barnett [56] to derive thermodynamic data for
Recent experimental data has enabled the Cemdata07 and thaumasite over the temperature range 1 to 30 °C to confirm experi-
Cemdata14 databases to be extended and refined [1,7,21]. We report mental data showing formation of thaumasite in mortars at 8 and 20 °C
this more comprehensive and refined dataset here as Cemdata18, as shown in Fig. 2. Another set of solubility data at 8 °C for natural
compiled in several tables. Cemdata18 has been developed to predict thaumasite was reported by Bellmann [57] who also highlighted the
changes in chemistry that occur during the hydration of Portland, potential pathways of formation of thaumasite at this temperature.
blended and alkali activated cements, and also their interactions with Macphee and Barnett [56] obtained the solubility data of ettringite-
service environments during use. thaumasite solid solutions in the temperature range between 5 °C and
Table 1 reports the thermodynamic properties of minerals important 30 °C; no apparent decomposition of thaumasite and related solid so-
for cementitious systems, while Table 2 reports their solubility products lutions occurred after 6 months storage at 30 °C, which suggests the
referring to the dominate species present at the high pH values of persistence of thaumasite at temperatures at least up to ~30 °C. A

Table 1
Cemdata18 database: Standard thermodynamic properties at 25 °C and 1 bar. Update of Cemdata07 [1,7,29]. The data are compatible with the GEMS version of the
PSI/Nagra 12/07 TDB [22,23]. Standard properties of master species and properties of reactions of forming product species out of master species, commonly used in
LMA programs such as PHREEQC, are compiled in the Appendix B.
ΔfG° [kJ/mol] ΔfH° [kJ/mol] S° [J/K/ a0 [J/K/ a1 [J/mol/ a2 [J K/mol] a3 [J/K0.5/ V° [cm3/ Ref
mol] mol] K2] mol] mol]
AFt-phases
(Al-)ettringitea,b,c −15,205.94 −17,535 1900 1939 0.789 – – 707 [1,7]
C6As3H30c −14,728.1 −16,950.2 1792.4 1452 2.156 – – 708 [30]
C6As3H13 −10,540.6 −11,530.3 1960.4 970.7 1.483 – – 411 [30]
C6As3H9 −9540.4 −10,643.7 646.6 764.3 1.638 – – 361 [30]
Tricarboaluminatea −14,565.64 −16,792 1858 2042 0.559 −7.78·106 – 650 [1,7]
Fe-ettringiteb −14,282.36 −16,600 1937 1922 0.855 2.02·106 – 717 [1,21]
Thaumasite −7564.52 −8700 897.1 1031 0.263 −3.40·106 – 330 [28]
Hydrogarnet
C3AH6d −5008.2 −5537.3 422 290 0.644 −3.25·106 – 150 [9,12]
C3AS0.41H5.18⁎,d −5192.9 −5699 399 310 0.566 −4.37·106 – 146 [9]
C3AS0.84H4.32⁎,e −5365.2 −5847 375 331 0.484 −5.55·106 – 142 [9]
C3FH6⁎⁎f −4122.8 −4518 870 330 1.237 −4.74·106 – 155 [9]
Al-Fe siliceous hydrogarnet (solid
solution)
C3FS0.84H4.32e,f −4479.9 −4823 840 371 0.478 −7.03·106 – 149 [9]
C3A0.5F0.5S0.84H4.32e −4926.0 −5335 619 367 0.471 −8.10·106 – 146 [9]
C3FS1.34H3.32 −4681.1 −4994 820 395 0.383 −8.39·106 – 145 [9]
AFm-phases
C4AH19 −8749.9 −10,017.9 1120 1163 1.047 – −1600 369 [12,31]
C4AH13g −7325.7 −8262.4 831.5 208.3 3.13 – – 274 [31]
C4AH11 −6841.4 −7656.6 772.7 0.0119 3.56 1.34·10−7 – 257 [31]
C2AH7.5 −4695.5 −5277.5 450 323 0.728 – – 180 [12]
CAH10 −4623.0 −5288.2 610 151 1.113 – 3200 193 [12]
C4Ac0.5H12 −7335.97 −8270 713 664 1.014 −1.30·106 −800 285 [1,7]
C4Ac0.5H10.5 −6970.3 −7813.3 668.3 0.0095 2.836 1.07·10−7 – 261 [31]
C4Ac0.5H9 −6597.4 −7349.7 622.5 0.0088 2.635 9.94·10−8 – 249 [31]
C4AcH11 −7337.46 −8250 657 618 0.982 −2.59·106 – 262 [1,7]
C4AcH9 −6840.3 −7618.6 640.6 192.4 2.042 – – 234 [31]
C4AsH16 −8726.8 −9930.5 975.0 636 1.606 – – 351 [31,32]
C4AsH14 −8252.9 −9321.8 960.9 1028.5 – – – 332 [31,32]
C4AsH12g,h −7778.4 −8758.6 791.6 175 2.594 – – 310 [31,32]
C4AsH10.5 −7414.9 −8311.9 721 172 2.402 – – 282 [31,32]
C4AsH9 −7047.6 −7845.5 703.6 169 2.211 – – 275 [31,32]
C2ASH8i −5705.15 −6360 546 438 0.749 −1.13·106 −800 216 [1,7]
C2ASH7i −5464.0 −6066.8 487.6 0.0063 1.887 7.12·10−8 – 215 [31]
C2ASH5.5 −5095.2 −5603.4 454.8 0.0057 1.685 6.36·10−8 – 213 [31]
C4As0.5ClH12 −7533.4 −8472j 820 557 1.141 −1.02·106 751 289 [27,33]
C4ACl2H10k −6810.9 −7604 731 498 0.895 −2.04·106 1503 272 [33,34]
C4A(NO3)2H10 −6778.1 −7719.3 821 580 1.02 −2.77·106 872 296 [34,35]
C4A(NO2)2H10 −6606.8 −7493.1 799 565 0.99 −2.24·106 703 275 [34–36]
C4FH13⁎⁎ −6438.6 −7435 630 694 1.113 2.02·106 1600 286 [9]
C4Fc0.5H10 −5952.9 −6581 1270 308 1.201 −9.08·105 3200 273 [8]
C4FcH12 −6674.0 −7485 1230 612 1.157 −5.73·105 – 292 [8]
C4FsH12h −6873.2 −7663 1430 577 1.234 2.02·106 – 321 [10]
C2FSH8 Not stable [37]
C4FCl2H10k −5900.1 −6528l 1286 481 0.961 −1.61·104 1503 278l [37]
Sulfates
Cs (anhydrite) −1322.12 −1434.60 106.7 70.2 −0.099 – – 46 [22,23]
CsH2 (gypsum) −1797.76 −2023.36 193.8 91.4 −0.318 – – 75 [22,23]
CsH0.5(hemihydrate) −1436.34m −1575.3m 134.3 124.1 – – – 62 [38]
Syngenite −2884.91 −3172 326 201 0.308 −1.78·106 – 128n [29]
(Hydr)oxides
Al(OH)3(am) −1143.2 - Not defined 32 [1]
Al(OH)3(mic) −1148.4 −1265.3o 140o 36 0.191 – – 32 [12]
Al(OH)3(gibbsite)⁎ −1151.0 −1288.7 70.1 36.2 0.191 – – 32 [22,23]
Fe(OH)3(am) −700.1 Not defined [22,23]
Fe(OH)3(mic) −711.6 Not defined [22,23]
FeOOH(mic) −480.14 −551.1 60 1.25 −0.233 −3.14·105 – 21 [9,22]
FeOOH(goethite)⁎ −497.26 −568.2 60 1.25 −0.233 −3.14·105 – 21 [22,23]
CH (portlandite) −897.01 −985 83 187 −0.022 – −1600 33 [22,23]
SiO2 (am) −848.90 −903 41 47 0.034 −1.13·106 – 29 [1,7]
SiO2 (quartz)⁎ −854.79 −909 41 47 0.034 −1.13·106 – 29 [22,23]
Hydrotalcite-pyroaurite (solid
solution)
½M6AcH13p −4339.85 −4875.9 411 512.6 – – – 115 [39]
½M6FcH13p −3882.60 −4415.1 423 521.7 – – – 119 [39]
(continued on next page)

Table 1 (continued)
ΔfG° [kJ/mol] ΔfH° [kJ/mol] S° [J/K/ a0 [J/K/ a1 [J/mol/ a2 [J K/mol] a3 [J/K0.5/ V° [cm3/ Ref
mol] mol] K2] mol] mol]
M-S-H (solid solution)

M1.5S2H2.5, Mg/Si = 0.75q −3218.43 −3507.52 270r 318r – – – 95 [40]
M1.5SH2.5, Mg/Si = 1.5q −2355.66 −2594.22 216r 250r – – – 74 [40]
Zeolites
Zeolite P(Ca)⁎ −5057.8 −5423 779 753 – – – 153s [41]
Natrolite⁎ −5325.7 −5728 360 359 – – – 169s [41]
Chabazite −7111.8 −7774 581 617 – – – 251s [41]
Zeolite X(Na) −5847.5 −6447 566 586 – – – 214t [41]
Zeolite Y(Na) −7552.5 −8327 734 739 – – – 283u [41]
Clinkers
C3S −2784.33 −2931 169 209 0.036 −4.25·106 – 73 [1,7,42]
C2S −2193.21 −2308 128 152 0.037 −3.03·106 – 52 [1,7,42]
C3A −3382.35 −3561 205 261 0.019 −5.06·106 – 89 [1,7,42]
C12A7 −18,451.44 −19,414 1045 1263 0.274 −2.31·107 – 518v [42]
CA −2207.90 −2327 114 151 0.042 −3.33·106 – 54w [42]
CA2 −3795.31 −4004 178 277 0.023 −7.45·106 – 89x [42]
C4AF −4786.50 −5080 326 374 0.073 – – 130 [1,7,42]
C (lime) −604.03 −635 39.7 48.8 0.0045 −6.53·105 – 17 [43]
Ks (K2SO4 arcanite) −1319.60 −1438 176 120 0.100 −1.78·106 – 66 [44]
K (K2O) −322.40 −363 94 77 0.036 −3.68·105 – 40 [43]
Ns (Na2SO4 thenardite) −1269.80 −1387 150 58 0.023 – – 53 [44]
N (Na2O) −376.07 −415 75 76 0.020 −1.21·106 – 25 [43]
a0, a1, a2, a3 are the empirical coefficients of the heat capacity function: C°p = a0 + a1T + a2T −2 + a3T −0.5; heat capacity functions for cement hydrates are
typically valid up to 100 °C only; ”–” = 0. Cement shorthand notation is used: A = Al2O3; C = CaO; F = Fe2O3; H = H2O; M = MgO; S = SiO2; c = CO2; s = SO3;
⁎
Precipitates very slowly at 20 °C, generally not included in calculations.
⁎⁎
Tentative value.
a
Non-ideal solid solutions; miscibility gap: XCO3,solid = 0.45–0.90 reproduced with the dimensionless Guggenheim interaction parameters α0 = 1.67 and
α1 = 0.946; downscaled in this paper to 1CO2:: 1SO3 replacement, instead of the 3CO2:: 3SO3 used in [4,7].
b
Non-ideal solid solution; miscibility gap: XAl,solid = 0.25–0.65 reproduced with the dimensionless Guggenheim interaction parameters α0 = 2.1 and α1 = −0.169
[45].
c
Ideal solid solutions c.f. [9,11,30,39].
d
e
f
g
Non-ideal solid solutions; miscibility gap: XOH,solid = 0.50–0.97 reproduced with the dimensionless Guggenheim interaction parameters α0 = 0.188 and
α1 = 2.49 [7]
h
Non-ideal solid solutions; miscibility gap: XAl,solid = 0.45–0.95 reproduced with the dimensionless Guggenheim interaction parameters α0 = 1.26 and α1 = 1.57
[10].
i
j
Typing error in [27], recalculated from Gf° and S from [27].
k
l
Typing error in [37], recalculated from Gf° and S from [37]. Volume calculated from XRD data [37].
m
Recalculated from ΔGr° of −20,500 J/mol [38].
n
Calculated from density data from [33,46].
o
Valid up to 60 °C only, estimated to describe solubility of microcrystalline Al(OH)3 aged for 19 months between 5 and 60 °C [12].
p
q
r
Estimated from Cp and S of talc, chrysotile and H2O using data from [43].
s
Volume from [47].
t
Calculated from XRD data: pdf 00-038-0237 [48].
u
Calculated from XRD data; pdf 00-039-1380 [49].
v
[50].
w
[51].
x
[52]
complete solubility dataset representative for the stability range of phase synthetic thaumasite. It was shown that pure thaumasite was
thaumasite was missing, as [56] reported the solubility data for thau- thermally stable up to 68 ± 5 °C. The obtained new data agreed well,
masite-ettringite solid-solutions but not for pure thaumasite. Hence, within limits of error, with those obtained by Macphee and Barnett
due to a lack of experimental data, no thermodynamic data for thau- [56], but differs significantly from the data for natural thaumasite re-
masite were included in the Cemdata07 database, but were added in a ported by Bellmann [57] at 8 °C. Experiments done by [28,56] excluded
first update using the data derived in Schmidt et al. [55] based on the atmospheric carbon dioxide, whereas the solubility determinations re-
solubility data given by Macphee and Barnett [49]. In 2015, Matschei ported in [57] were made in the presence of air containing carbon di-
and Glasser [28] published a new dataset obtained on apparently pure- oxide. The contact with the air may lead to the decomposition of

Table 2
Equilibrium solubility products of solids and formation constants for calcium-silica complexes at 1 bar, 25 °C in Cemdata18 (as given in Table 1).
Mineral log KS0 Dissolution reactions used to calculate solubility products.
Solids
(Al-)ettringite −44.9 Ca6Al2(SO4)3(OH)12·26H2O → 6Ca2+ + 2Al(OH)4− + 3SO42− + 4OH− + 26H2O
tricarboaluminate −46.5 Ca6Al2(CO3)3(OH)12·26H2O → 6Ca2+ + 2Al(OH)4−+ 3CO32− + 4OH− + 26H2O
Fe-ettringite −44.0 Ca6Fe2(SO4)3(OH)12·26H2O → 6Ca2+ + 2Fe(OH)4− + 3SO42− + 4OH− + 26H2O
thaumasite −24.75 Ca3(SiO3)(SO4)(CO3)·15H2O → 3Ca2+ + H3SiO4− + SO42− + CO32– + OH– + 13H2O
C3AH6 −20.50 Ca3Al2(OH)12 → 3Ca2+ + 2Al(OH)4− + 4OH−
C3AS0.41H5.18⁎ −25.35 Ca3Al2(SiO4)0.41(OH)10.36 → 3Ca2+ + 2Al(OH)4− + 0.41 SiO(OH)3− + 3.59OH− − 1.23H2O
C3AS0.84H4.32⁎ −26.70 Ca3Al2(SiO4)0.84(OH)8.64 → 3Ca2+ + 2Al(OH)4− + 0.84 SiO(OH)3− + 3.16OH− − 2.52H2O
C3FH6 −26.30⁎⁎ Ca3Fe2(OH)12 → 3Ca2+ + 2Fe(OH)4− + 4OH−
C3FS0.84H4.32 −32.50 Ca3Fe2(SiO4)0.84 (OH)8.64 → 3Ca2+ + 2Fe(OH)4− + 0.84 SiO(OH)3− + 3.16OH− − 2.52H2O
C3(F,A)S0.84H4.32 −30.20 Ca3FeAl(SiO4)0.84 (OH)8.64 → 3Ca2+ + Al(OH)4− + Fe(OH)4− + 0.84 SiO(OH)3− + 3.16OH− − 2.52H2O
C3FS1.34H3.32 −34.20 Ca3Fe2(SiO4)1.34 (OH)6.64 → 3Ca2+ + 2Fe(OH)4− + 1.34 SiO(OH)3− + 2.66OH− − 4.02H2O
C4AH19 −25.45 Ca4Al2(OH)14·12H2O → 4Ca2+ + 2Al(OH)4− + 6OH− + 12H2O
C4AH13 −25.25 ⁎⁎⁎ Ca4Al2(OH)14·6H2O → 4Ca2+ + 2Al(OH)4− + 6OH− + 6H2O
C2AH7.5 −13.80 Ca2Al2(OH)10·2.5H2O → 2Ca2+ + 2Al(OH)4− + 2OH− + 2.5H2O
CAH10 −7.60 CaAl2(OH)8·6H2O → Ca2+ + 2Al(OH)4− + 6H2O
C4Ac0.5H12 −29.13 Ca4Al2(CO3)0.5(OH)13·7H2O → 4Ca2+ + 2Al(OH)4− + 0.5CO32− + 5OH− + 7H2O
C4AcH11 −31.47 Ca4Al2(CO3)(OH)12·5H2O → 4Ca2+ + 2Al(OH)4− + CO32– + 4OH− + 5H2O
C4AsH14 −29.26 Ca4Al2(SO4)(OH)12·6H2O → 4Ca2+ + 2Al(OH)4− + SO42− + 4OH− + 6H2O
C4AsH12 −29.23 ⁎⁎⁎ Ca4Al2(SO4)(OH)12·6H2O → 4Ca2+ + 2Al(OH)4− + SO42− + 4OH− + 6H2O
C2ASH8 −19.70 Ca2Al2SiO2(OH)10⋅3H2O → 2Ca2+ + 2Al(OH)4− + SiO(OH)3− + OH– + 2H2O
Friedel's salt −27,27 Ca4Al2Cl2(OH)12·4H2O → 4Ca2+ + 2Al(OH)4− + 2Cl− + 4OH− + 4H2O
Kuzel's salt −28,53 Ca4Al2Cl(SO4)0.5(OH)12·6H2O → 4Ca2+ + 2Al(OH)4− + Cl− + 0.5SO42− + 4OH− + 6H2O
Nitrate-AFm −28.67 Ca4Al2(OH)12(NO3)2·4H2O → 4Ca2+ + 2 Al(OH)4− + 2 NO3− + 4OH− + 4H2O
Nitrite-AFm −26.24 Ca4Al2(OH)12(NO2)2·4H2O → 4Ca2+ + 2 Al(OH)4− + 2 NO2− + 4OH− + 4H2O
C4FH13 −30.75⁎⁎ Ca4Fe2(OH)14·6H2O → 4Ca2+ + 2Fe(OH)4− + 6OH− + 6H2O
Fe-hemicarbonate −30.83 Ca4Fe2(CO3)0.5(OH)13·3.5H2O → 4Ca2+ + 2Fe(OH)4− + 0.5CO32− + 5OH− + 3.5H2O
Fe-monocarbonate −34.59 Ca4Fe2(CO3)(OH)12·6H2O → 4Ca2+ + 2Fe(OH)4− + CO32– + 4OH− + 6H2O
Fe-monosulfate −31.57 Ca4Fe2(SO4)(OH)12·6H2O → 4Ca2+ + 2Fe(OH)4− + SO42− + 4OH− + 6H2O
Fe-Friedel's salt −28.62 Ca4Fe2Cl2(OH)12·4H2O → 4Ca2+ + 2Fe(OH)4− + 2Cl− + 4OH− + 4H2O
Cs (anhydrite) −4.357 CaSO4 → Ca2+ + SO42−
CsH2 (gypsum) −4.581 CaSO4·2H2O → Ca2+ + SO42− + 2H2O
CsH0.5(hemihydrate) −3.59 CaSO4·0.5H2O → Ca2+ + SO42− + 0.5H2O
syngenite −7.20 K2Ca(SO4)2·H2O → 2 K+ + Ca2+ + 2SO42− + H2O
Al(OH)3(am) 0.24 Al(OH)3(am) → Al(OH)4− − OH–
Al(OH)3(mic) −0.67 Al(OH)3(mic) → Al(OH)4− − OH–
Al(OH)3(gibbsite) ⁎ −1.12 Al(OH)3(gibbsite) → Al(OH)4− − OH–
Fe(OH)3(am) −2.6 Fe(OH)3(am) → Fe(OH)4− − OH−
Fe(OH)3(mic) −4.6 Fe(OH)3(mic) → Fe(OH)4− − OH–
FeOOH(mic) −5.6 FeOOH(mic) → Fe(OH)4− − OH– − H2O
FeOOH(goethite) ⁎ −8.6 FeOOH(goethite) → Fe(OH)4− − OH– − H2O
CH −5.2 Ca(OH)2 → Ca2+ + 2OH−
SiO2(am) −2.714 SiO2 (am) → SiO20
SiO2(quartz) ⁎ −3.746 SiO2(quartz) → SiO20
1
/2M6AcH13 −33.29⁎⁎⁎ Mg3Al(OH)8(CO3)0.5·2.5H2O → 3Mg2+ + Al(OH)4− + 0.5CO32−. + 4OH− + 2.5H2O
1
/2M6FcH13 −33.64⁎⁎⁎ Mg3Fe(OH)8(CO3)0.5·2.5H2O → 3Mg2+ + Fe(OH)4− + 0.5CO32−. + 4OH− + 2.5H2O
M1.5S2H2.5 −28.80 (MgO)1.5(SiO2)2(H2O)2.5 → 1.5Mg2+ + 2SiO20 + 3OH− + H2O
M1.5SH2.5 −23.57 (MgO)1.5SiO2(H2O)2.5 → 1.5Mg2+ + SiO20 + 3OH− + H2O
Zeolite P(Ca) ⁎ −20.3 CaAl2Si2O8·4.5H2O → Ca2+ + 2Al(OH)4− + 2SiO20 + 0.5H2O
Natrolite ⁎ −30.2 Na2Al2Si3O10·2H2O → 2Na+ + 2Al(OH)4− + 3SiO20 - 2H2O
Chabazite −25.8 CaAl2Si4O12·6H2O → Ca2+ + 2Al(OH)4− + 4SiO20 + 2H2O
Zeolite X(Na) −20.1 Na2Al2Si2.5O9·6.2H2O → 2Na+ + 2Al(OH)4− + 2.5SiO20 + 2.2H2O
Zeolite Y(Na) −25.0 Na2Al2Si4O12·8H2O → 2Na+ + 2Al(OH)4− + 4SiO20 + 4H2O
Calcium silicate complexes

CaHSiO3+ 1.2⁎v Ca2+ + HSiO32− → CaHSiO3+
CaSiO30 4.6⁎v Ca2+ + SiO32− → CaSiO30
⁎
Precipitates very slowly at 20 °C, generally not included in calculations.
⁎⁎
Tentative value.
⁎⁎⁎
Recalculated in this paper from ΔGf° values.
⁎v
The formation of less strong calcium silicate complexes have been recently suggested (log K(CaHSiO3+) = 0.5 and log K(CaSiO30) = 2.9. Within Cemdata18,
however, the listed values for calcium silicate complexes have to be used to maintain compatibility with the C-S-H models.
thaumasite, which would make the interpretation of the solubility data recalculate solubilities under the chosen experimental conditions with
invalid. the thermodynamic data of the Cemdata18 dataset. As illustrated in
The heat capacities were estimated using a reference reaction with a Fig. 3, the calculated solubility data for thaumasite show generally
solid having a known heat capacity and similar structure, as discussed good agreement with the experimentally-derived dataset. Despite an
in more detail in [28,55]. As shown by Helgeson et al. [43], this underestimation of the calculated silicon concentrations at 1 °C and
principle can be successfully applied to estimate the heat capacity of 5 °C, both datasets, experimental and calculated, generally agree,
silicate minerals by formulating a reaction involving a structurally-re- proving the internal consistency of the data. Especially in the tem-
lated mineral of known heat capacity. perature range from 1 to ~40 °C, where the solid phase assemblage
Finally, it is possible to do an internal consistency check and consists mainly of thaumasite and traces of calcite, differences between

Fig. 1. Logarithm of the solubility product of Al(OH)3

(referring to Al(OH)4− and OH−) as a function of time
and temperature calculated from the literature, adapted
from [12]. Gibbsite solubility (dotted line) was calculated
using data from the GEMS version of the PSI/Nagra 12/
07 TDB [22,23], whereas the solubility of microcrystal-
line Al(OH)3 (black line) and amorphous Al(OH)3 (black
hyphen) was calculated based on the data given in
Table 1.
-22 2.3. Chloride-, nitrate-, and nitrite-AFm phases

Calc. solubility product log Ksp
synthetic thaumasite
-23 Binding of chloride and the formation of chloride bearing cement
natural thaumasite
hydrates has been widely studied due to its impact on the corrosion of
-24 steel in reinforced concrete. The first comprehensive solubility data for
Friedel's salt (Ca4Al2Cl2(OH)12·4H2O) and Kuzel's salt (Ca4Al2Cl
1)
-25 (SO4)0.5(OH)12·6H2O) were provided in the late nineties. Birnin-Yauri
[58] has described the dissolution of Friedel's salt as congruent and
-26 1) 1) provided values of log KS0 as −27.1 and −24.8 (KS0 = {Ca2+}4{Al
(OH)4−}2{Cl−}2{OH−}4{H2O}4). Hobbs [59] estimated log KS0 as
-27 1) Macphee and Barnett 2004 −27.6 ± 0.9 and Bothe [60] has estimated via geochemical modelling
2) Bellmann 2004 that the solubility product of Friedel's salt should fall within the range
2)
-28 −28.8 < log KS0 < −27.6. Balonis et al. [27] provided solubility
0 10 20 30 40 50 60 data for Friedel's salt as a function of time and temperature with an
estimated value of solubility product for an ideal composition and at
Temperature [°C]
room temperature to be −27.27 [34,36]. Compilation of the available
Fig. 2. Calculated solubility products referring to Ca2+, SiO(OH)3−, SO42−, solubility data is shown by triangles on Fig. 4.
CO32−, OH– and H2O of synthetic and natural thaumasite samples from solu- The estimated thermodynamic data [36] (ΔfG0 ∼ −6810.9 kJ/mol,
bility experiments. The curve shows the calculated best fit using a three-term ΔfH0 ∼ −7604 kJ/mol, S0 731 J/mol K) have similar values (except the
temperature extrapolation. Reproduced from [28]. entropy) to the dataset published by Blanc et al. [16]
(ΔfG0 ∼ −6815.44 kJ/mol, ΔfH0 ∼ −7670.04 kJ/mol, S0 527.70 J/
experimental calcium and sulfate concentrations are within analytical mol K), and agree reasonably well with the data obtained by Grish-
errors. In the temperature range 1 °C to ~40 °C, concentrations of cal- chenko et al. [61] (ΔfG0 estimated in a range between 6800 and
cium, sulfate and silicon increase with rising temperature, whereas 6860 kJ/mol, S0 ∼ 680 J/mol K), though it should be kept in mind that
calculated carbonate concentrations show a continuous decrease. At Grishchenko‘s composition is reported to be slightly contaminated with
temperatures > ~40 °C, calcium and sulfate concentrations increase carbonate ions. Attempts to synthesize Cl-AFt at temperatures above
significantly, whereas silicon concentrations decrease due to the for- 0 °C were unsuccessful [34], hence no thermodynamic data are avail-
mation of C-S-H. Thaumasite is absent at temperatures above 70 °C. able that can be used.
Th.+ Calc. Calc.+

Th.+ Calc. Calc+
Thaumasite + Calcite Thaumasite + Calcite + C-S-H C-S-H
+ C-S-H C-S-H
16 10
Concentration [mM]
14
Concentration [mM]
12 1
10 silicon
calcium
8 0.1
6
4 sulfate 0.01
carbonate
2
0 0.001
0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80
Temperature [°C] Temperature [°C]
Fig. 3. Experimentally measured (markers) and re-calculated (lines) solubility data for thaumasite; (filled markers represent the experimental data for synthetic
thaumasite, open markers – the data for natural thaumasite from [28]). Calculations are based on the new thermodynamic data for thaumasite complemented with
the CSHQ data from Cemdata18 [1,7]. Predicted solid phases/phase assemblages are shown along the top.

monocarbonate, siliceous hydrogarnet can precipitate, as well as form

solid solutions with their Al-containing analogues [8–10,21].
The stability of Fe-containing phases generally is only moderately
affected by temperature, as shown in Fig. 5. At ambient temperature,
Fe-ettringite (C6Fs3H32), Fe-monosulfate (C4FsH12), Fe-monocarbonate
(C4FcH12), Fe-Friedel's salt (C4FCl2H10), and Fe-siliceous hydrogarnet
(C3FS0.95H4.1, C3FS1.52H2.96) are stable, while Fe-katoite (C3FH6) and
Fe-hemicarbonate (C4Fc0.5H10) are metastable [8–10,21,37]. Attempts
to synthesize Fe-strätlingite (C2FSH8) failed, as only portlandite, C-S-H
and iron hydroxide formed, indicating the instability of Fe-strätlingite
at ambient conditions. C4FsH12, C4FcH12, and C4FCl2H10 are also stable
at 50° but not at 80 °C, while Fe-siliceous hydrogarnet is stable at up to
110 °C. The limited stability field of the Fe-containing AFm and AFt
hydrates is related to the very high stability of goethite (FeOOH) and
Fig. 4. Solubility products of Friedel's salt, Kuzel's salt, NO3-AFm and NO2-AFm
hematite (Fe2O3), which form at 50 °C within several months and at
(referring to reactions using Ca2+, Cl−, SO42−, NO3−, NO2−, OH– and H2O as 80 °C within days [9]. Although hematite and portlandite would be
indicated in Table 2) as a function of temperature. Data for Friedel's salt from more stable than the Fe-katoite, AFt and AFm phases between 0 and
[27,58–60,62], data for other AFm are from Balonis and co-workers 100 °C, the formation of goethite and hematite at ambient temperatures
[27,34–36]. is very slow, such that Fe-containing siliceous hydrogarnet, AFt and
AFm phases can be synthesized instead. Fig. 3 shows the solubility
Glasser et al. [62] first measured the solubility of Kuzel's salt and products of Fe-containing phases calculated based on the measured
noted that its dissolution is strongly incongruent, with ettringite pre- composition of the liquid phase at 20, 50 and 80 °C; those data were
cipitating as a secondary phase. From the solubility data given by used to derive the thermodynamic data for standard conditions (25 °C,
Glasser et al. a log KS0 of Kuzel's salt −28.54 (KS0 = {Ca2+}4{Al 1 atm) given in Table 1. The formation of solid solutions between Al
(OH)4−}2{Cl−}{SO42−}0.5{OH−}4{H2O}6) was estimated [27]. Ba- and Fe-containing endmembers has been observed for ettringite, silic-
lonis et al. [27] has also experimentally derived the solubility data and eous hydrogarnet, monosulfate, and Friedel's salt, while no solid solu-
calculated solubility products for Kuzel's salt at different temperatures tion formed between the rhombohedral Fe-monocarbonate with the
ranging from 5 to 85 °C for the period between 1 and 12 months, with triclinic Al-monocarbonate due to the structural differences
the solubility product at room temperature determined to be log [8–10,21,37].
KS0 = −28.53. Data for 12 months are shown by the filled circles in While different Fe-containing hydrates could be synthesized, only
Fig. 4. Fe-siliceous hydrogarnet is expected to occur in hydrated cements. The
In recent years, the impact of soluble nitrate and nitrite corrosion solubility product of Fe-siliceous hydrogarnet (given in Table 1) is 5 to
inhibitors on the mineralogy of cement pastes has been studied 7 log units lower than that of Al-siliceous hydrogarnet indicating a high
[34,36,63], and it has been demonstrated that the AFm phase has the stabilization of Fe-siliceous hydrogarnet, while the solubility products
ability to accommodate NO3− and NO2− ions in the interlayer position. of the Fe-containing hydrates are comparable or only somewhat more
Solubility data along with thermodynamic parameters for the nitrate stable than their Al-containing analogues. In fact, in hydrated PC, Fe
AFm (NO3-AFm) and nitrite AFm (NO2-AFm) published by Balonis et al. (III) precipitates as iron hydroxide during the first hours and as silic-
[34,35] are shown in Fig. 4. Similarly, as in the case of Cl-AFt, an at- eous hydrogarnet (C3(A,F)S0.84H4.32) after 1 day and longer [64–66].
tempted synthesis of NO3- or NO2-AFt at room temperature was not The data for the C3FS0.84H4.32 and for the mixed Al- and Fe-containing
successful [34]. C3A0.5F0.5S0.84H4.32 determined by Dilnesa et al. [9] are included in
Cemdata18, but not the data for the Al-based C3AS0.84H4.32 due to its
formation being kinetically hindered at ambient conditions [9].
2.4. Iron containing hydrates
2.5. Effect of relative humidity
The main source of iron in cements is 5–15% ferrite clinker in
Portland cements and slag in blended cements. In synthetic systems
Cement hydrates are known to show varying water content as
containing only water, C2F, calcium sulfate, calcium carbonate or silica, functions of temperature and relative humidity (RH). Some of these
different Fe-containing phases like ettringite, monosulfate,
hydrates are crystalline phases with layered structure such as the AFm-
phases or ettringite-type structures. The AFm and AFt phases have
different hydration states (i.e. varying molar water content) depending
on the exposure conditions, which can impact the volume stability,
porosity and density of cement paste. The molar volume of some AFm
phases can decrease by as much as 20% during drying [31], which may
strongly influence the porosity and performance of some cementitious
systems.
In gel-like phases such as C-S-H, water can be present within the
intrinsic gel porosity, as well as in its interlayer. Unfortunately, until
recently there was no thermodynamic model capable of assessing this
varying water content.
The crystalline AFm phases have a layered structure and are known
for their varying water content in the interlayer, which can be of two
types. Firstly, the “space filling”, loosely integrated zeolitic water mo-
Fig. 5. Solubility product (KS0) of Fe-containing hydrogarnet and AFm-phases lecules, which are easily removed from the structure upon increase of
at different temperatures, referring to reactions using Ca2+, Fe(OH)4−, SiO temperature or at an initial small decrease of RH and have thermo-
(OH)3−, SO42−, CO32−, Cl−, OH– and H2O as indicated in Table 2. Data from dynamic properties close to liquid water. Secondly, the “structural
Dilnesa and co-workers [8–10,37]. water” molecules, which are strongly bound to calcium cations of the

Ettringite, C6As3H32, is also known to have varying water content.

This hydrate is a common phase occurring during the hydration of PC.
It is also the main hydration product in calcium sulfoaluminate cements
and calcium aluminate cement blended with gypsum. Understanding
the stability of ettringite during hydration and under different drying
conditions is of great importance to assess the performance of systems
containing large amounts of this phase. In general, ettringite contains
32 H2O molecules per formula unit: 30 fixed in the columns and 2 H2O
of zeolitic water loosely bound in the channels. Removal of the two
inter-channel water molecules takes place with decreasing relative
humidity (RH) without any significant change of the structure.
Nevertheless, a series of structural changes are observed when the
water content is below 30 H2O, resulting in an amorphous phase
commonly known as metaettringite. The thermodynamic properties of
crystalline ettringite, having 32 and 30 H2O, and amorphous ettringite
(or metaettringite) having 13H2O and 9H2O were recently derived by
Baquerizo et al. [30] and are listed in Table 1. Something interesting to
notice is that decomposition and reformation of ettringite takes place
reversibly but with a marked hysteresis, which makes the estimation of
Fig. 6. Volume changes of the AFm phases studied as function of RH at 25 °C. thermodynamic properties difficult. The values presented in Table 1
100% volume corresponds to the higher hydration state of each phase. corresponds to those derived using the desorption equilibrium proper-
ties. Fig. 7 shows the stability of ettringite at 25°, presenting three
different zones:
- The zone of decomposition, which has to be reached in order to

decompose ettringite into metaettringite.
- The hysteresis loop, where crystalline ettringite will not undergo
decomposition unless the zone of decomposition is reached and
amorphous metaettringite will not reform unless the zone of re-
formation is reached.
- The zone of reformation, which has to be reached in order to be
convert metaettringite back to crystalline ettringite.
2.6. Mg-Al layered double hydroxide (hydrotalcite-like phase)
Mg-Al layered double hydroxide (LDH) type phases are structurally

similar to hydrotalcite and typically occur as secondary reaction pro-
ducts in hydrated Portland cements [67] and in alkali-activated
granulated blast furnace slag (GBFS) [68,69]. In hydrated or alkali-
activated cementitious materials free from carbonation, Mg-Al LDH
phases normally exhibit poor long-range structural order and are
Fig. 7. Stability of ettringite as a function of relative humidity and temperature.
thought to significantly occur along the solid solution series
Mg(1−x)Alx(OH)(2+x)(H2O)4, where 0.2 ≤ x ≤ 0.33 [70,71] due to the
main layer and can only be removed at low water activities and/or high deficiency of CO2 in the system. Mg-Al LDH formation is thus often
temperatures, typically accompanied by high enthalpies values. difficult to observe by conventional X-ray diffraction, particularly at
Recently, the thermodynamic properties of the different hydration low MgO content.
states of the most important AFm phases were determined by Baquerizo Few solubility data for hydroxide containing hydrotalcite like Mg-Al
et al. [31,32] and are listed in Table 1. A summary of the volume sta- LDH phases have been measured; the data at 25 °C are summarised in
bility of AFm phases at 25 °C is shown in Fig. 6. Fig. 8A and B. The samples studied by Bennet et al. [72] were
Fig. 8. Solubility of A) M4AH10 (from Cemdata07 + 18) and B) of the MgAl-OH-LDH solid solution compared to the solubility of microcrystalline Al(OH)3 and brucite
(dotted lines) and to the experimental data (Mg: circles, Al: triangles) determined by Bennet et al. [72] and Gao and Li [73].

to 1 month [72]). We recommend using MgAl-OH-LDH_ss for alkali

activated materials.
Usage of the MgAl-OH-LDH_ss model (describing hydrotalcite-like
phases with variable Mg/Al ratio, and recommended for use in alkali
activated material systems) does not lead to hydrotalcite formation
under typical PC conditions due to the low aluminium concentrations in
the pore solution [29] of PCs, for which brucite would be calculated to
precipitate instead. As the formation of hydrotalcite like phases is re-
ported in well hydrated PCs with dolomite [75], the use of a single
phase, M4AH10, with a lower solubility product (see Table 3, Fig. 9)
derived from the long-term experiments in [72] only, is recommended
for hydrated PC. The necessity to use presently two different datasets
and the large differences in the available data indicates that the solu-
bility data selected for M4AH10 and for MgAl-OH-LDH_ss are tentative
and may require updating as more data become available. Therefore,
we believe that additional solubility measurements for Mg-Al LDH
phases are needed.
2.7. C-S-H solid solution models
Fig. 9. Measured and calculated solubility products of M4AH10 (reactions refer The C-S-H gel-like phase is the major hydrate in PC and blended PC
to Mg2+, Al(OH)4−, OH– and H2O as indicated in Table 3) at different tem-
pastes. C-S-H is also the main “sorbent” of alkali, alkali-earth, and ha-
peratures. Adapted from Myers et al. [74].
zardous cations (Sr2+, UO22+, Zn2+, etc.) in hydrated cements used as
waste matrices, including engineered barriers in nuclear waste re-
synthesized for 2 days at 80 °C, dried, and then re-dispersed in water for positories.
4 weeks at 25 °C. This procedure resulted in a solubility product of C-S-H phases have a variable composition that depends on the
10−47 for M4AH10. Further re-dispersion steps lowered the solubility prevailing Ca/Si ratio in the system that can change by pozzolanic re-
product of M4AH10 to 10−56. This lower solubility product of 10−56 for action, leaching caused by the ingress of water and/or chemical attack,
M4AH10 was selected for use in Cemdata07 [1,29] (see Figs. 8A and 9), such as carbonation. There are differences between properties of C-S-H
and by Bennet et al. [72]. samples prepared by (a) C3S or C2S hydration; (b) co-precipitation
Based on the solubility data of Gao and Li [73] for samples pre- (double-decomposition) methods [76]. C-S-H has a ‘defect-tobermorite’
cipitated from oversaturated solutions (equilibration time 2 days), so- structure with a mean silicate chain length depending on the Ca/Si
lubility data for hydrotalcite like Mg-Al LDH phases intercalated with ratio, pH and the presence of aluminium [77]. It has variable “non-gel”
OH– (MgAl-OH-LDH) were recently recompiled and recalculated [74], water content (i.e. structural water and water present in the interlayer
as shown in Figs. 8B and 9. Solubility products for the end members of [78,79]), also depending on the Ca/Si ratio and the synthesis route,
MgAl-OH-LDH solid solution model were defined using the available variable particle morphology, stacking, and “gel” water content, i.e.
data [72,73] and guided using experimental observations in alkali ac- water present between C-S-H particles. Many C-S-H experimental so-
tivated slag cements with the high stability of MgAl-OH-LDH and ab- lubility data sets available to date have been critically analyzed [80],
sence of brucite in uncarbonated alkali-activated slag cements is widely including C-S-H type phases with variable aluminium and alkali con-
documented and provides a reliable proxy for this task. An ideal (simple tents [76,81–84].
mixing) solid solution thermodynamic model (MA-OH-LDH_ss) was C-S-H solubility can be reliably modelled using either solid solution
provisionally defined using these data for Mg/Al molar ratios between 2 models [11,80,85] or (to a limited extent) using a surface complexation
and 4. The use of independent experimental observations to derive the approach [86,87]. Quantitative knowledge of C-S-H solubility is needed
solid solution model is important because solubility products derived in essentially all studies of cement hydration and of waste-cement in-
from the available solubility data are scattered by up to ~10 log10 units teractions, which explains why measuring and modelling the C-S-H
at 25 °C, possibly due to the varied equilibration times used (2 days [73] solubility and water content is a major topic in cement chemistry [76].
Table 3
Standard thermodynamic properties at 25 °C and 1 atm for hydrotalcite-like phases (provided in separate modules of Cemdata18 database). The data are consistent
with the GEMS version of the PSI/Nagra 12/07 TDB [22,23] and the data detailed in Tables 1 and 4.
ΔfG° [kJ/mol] ΔfH° [kJ/mol] S° [J/K/mol] a0 [J/K/mol] a1 [J/mol/K2] a2 [J K/mol] a3 [J/K0.5/mol] V° [cm3/mol] Ref
6
M4AH10 ⁎
−6394.6 −7196 549 −364 4.21 3.75·10 629 220 [1,29]
MgAl-OH-LDH (ideal ternary solid solution)⁎⁎

M4AH10 −6358.5 −7160.2 548.9 547.6 – – – 219.1 [74]
M6AH12 −8022.9 −9006.7 675.2 803.1 – – – 305.4 [74]
M8AH14 −9687.4 −10,853.3 801.5 957.7 – – – 392.4 [74]
Mineral log KS0 Dissolution reactions used to calculate solubility products.

M4AH10⁎ −56.02⁎ Mg4Al2(OH)14·3H2O → 4Mg2+ + 2Al(OH)4− + 6OH− + 3H2O
M4AH10⁎⁎ −49.7 Mg4Al2(OH)14·3H2O → 4Mg2+ + 2Al(OH)4− + 6OH− + 3H2O
−2 −0.5
a0, a1, a2, a3 are the empirical coefficients of the heat capacity function: C°p = a0 + a1T + a2T + a3T .
⁎
Tentative value; recommended for PC based systems.
⁎⁎
Tentative values; recommended for alkali activated materials.

Table 4
Solid solution models of C-S-H (provided in separate modules of Cemdata18 database).a
Phase, ΔfG° ΔfH° S° a0 a1 a2 V° Ref
2 3
End member [kJ/mol] [kJ/mol] [J/K/mol] [J/K/mol] [J/mol/K ] [J K/mol] [cm /mol]
C-S-H (CSH-II solid solution),

Tob: C0.83SH1.3 -1744.36 -1916 80 85 0.160 59 [1]
Jen: C1.67SH2.1 -2480.81 -2723 140 210 0.120 -3.07·106 78 [1]
C-S-H-K-N (ECSH-1 solid solution)

TobCa-1: C0.83SH1.83 -1863.62 -2059.5 114.6 170.4 68 [85]
SH: SH (SiO2H2O) -1085.45 -1188.6 111.3 119.8 34 [85]
NaSH-1: N0.5S0.2H0.45 -433.57 -480.4 41.2 37.9 10.5 [88]
KSH-1: K0.5S0.2H0.45 -443.35 -490.0 48.4 40.6 12.4 [88]
SrSH-1: SrSH2 -2020.89 -2231.6 141.9 174.8 64 [88]
(-2017.47b) (-2228b)
C-S-H-K-N (ECSH-2 solid solution)

TobCa-2: C0.83SH1.83 -1863.62 -2059.5 114.6 170.4 68 [85]
JenCa: CS0.6H1.1 -1569.05 -1741.6 73.0 114.5 36 [85]
NaSH-2: N0.5S0.2H0.45 -430.72 -477.6 41.2 37.9 10.5 [88]
KSH-2: K0.5S0.2H0.45 -440.49 -487.2 48.4 40.6 12.4 [88]
SrSH-2: SrSH2 -2019.75 -2230.5 141.9 174.8 64 [88]
(-2016.33b) (-2227b)
C-S-H (CSHQ solid solution)

TobH Ca/Si=0.67: C2/3SH1.5 -1668.56 -1841.5 89.9 141.6 55 [11]
TobD Ca/Si =1.25:C5/6S2/3H1.83 -1570.89 -1742.4 121.8 166.9 48 [11]
JenH Ca/Si =1.33: C1.33SH2.17 -2273.99 -2506.3 142.5 207.9 76 [11]
JenD Ca/Si=2.25: C1.5S0.67H2.5 -2169.56 -2400.7 173.4 232.8 81 [11]
NaSH: N0.5S0.2H0.45 -431.20 478.0 41.2 37.9 10.5 [88,89]
KSH: K0.5S0.2H0.45 -440.80 -489.6 48.4 40.6 12.4 [88,89]
C-S-H (CSH3T solid solution)

TobH Ca/Si=0.67: C1S3/2H5/2 -2561.53 -2832.97 152.8 231.2 85 [11]
T5C Ca/Si=1.0: C5/4S5/4H5/2 -2518.66 -2782.03 159.9 234.1 79 [11]
T2C Ca/Si=1.5: C3/2S1H5/2 -2467.08 -2722.40 167.0 237.0 81 [11]
C-(N-)A-S-H (CNASH solid solution)

TobH c: C1S3/2H5/2 -2560.00 -2831.4 152.8 231.2 85.0 [90]
INFCA: C1A5/32S38/32H53/32 -2342.90 -2551.3 154.5 180.9 59.3 [90]
INFCN: C1N5/16S3/2H19/16 -2452.46 -2642.0 185.6 183.7 71.1 [90]
INFCNA: C1A5/32N11/32S38/32H42/32 -2474.28 -2666.7 198.4 179.7 69.3 [90]
T5C c: C5/4S5/4H5/2 -2516.90 -2780.3 159.9 234.1 79.3 [90]
5CA: C5/4A1/8S1H13/8 -2292.82 -2491.3 163.1 177.1 57.3 [90]
5CNA: C5/4N1/4A1/8S1H11/8 -2381.81 -2568.7 195.0 176.2 64.5 [90]
T2C c: C3/2S1H5/2 -2465.40 -2720.7 167.0 237.0 80.6 [90]
a0, a1, a2, are the empirical coefficients of the heat capacity equation: C°p = a0 + a1T + a2T−2; no value = 0.
a
Only CSH-II solid solution included in Cemdata’07.03 database.
b
For the ACW conditions.
c
Thermodynamic properties were slightly modified relative to the T2C, T5C, and TobH end members of the downscaled CSH3T thermodynamic model [11].
In Table 4, five alternative C-S-H solid solution models are re- ratios = 0.4–0.8. The water content in this C-S-H II is lower than in the
presented, in part for backward compatibility with previous versions of other models discussed below, but corresponds well to the water pre-
Cemdata (Cemdata07 and Cemdata14); they are provided in the Cem- sent in the interlayer of C-S-H as measured by 1H NMR [78,79]. In
data18 database. Here we provide a brief overview of those models Cemdata18, we provide the CSH-II solid solution model only, covering
with some recommendations for their use. the range of Ca/Si ratios from 0.83 to 1.67, for backward compatibility
with the Cemdata07 database and as an alternative to the newer
models.
2.7.1. CSH-II model
This simple ideal C-S-H solid solution model [85] has been used for
many years, and was included (with a modified stability to better de- 2.7.2. ECSH-1 and ECSH-2 models
scribe the changes in the calcium concentrations with pH and less water ECSH-1 and ECSH-2 models extend both CSH-I and CSH-II models
to correspond to the composition of C-S-H present in cements) into with Na-, K- and Sr- containing end members. Aimed at pragmatic de-
Cemdata07 database [1,29]. The original model [85] consisted of two scription of uptake of minor cations, these provisional ideal solid so-
binary ideal solutions CSH-I and CSH-II. CSH-I used end-members of lution models [88] were constructed with help of the statistical dual-
amorphous silica (SH; SiO2) and a tobermorite-like C-S-H gel phase thermodynamic method [91] based on GEM-Selektor calculations. With
(Tob-I; (Ca(OH)2)2(SiO2)2.4·2H2O). CSH-II used end-members of to- this method, one can retrieve both the unknown stoichiometry and the
bermorite-like (Tob-II; (Ca(OH)2)0.8333SiO2:0.8333·H2O) and jennite-like standard molar Gibbs energy ΔfG°298 of ideal solid solution end mem-
(Jen; (Ca(OH)2)1.6666SiO2·H2O) C-S-H gel phases. The CSH-II phase co- bers from the experimental bulk compositions of the aqueous solution
exists with CH (portlandite) at Ca/Si ratios above 1.5 to 1.7. The CSH-I and co-existing solid solution. In total, 13 possible end member stoi-
solid solution has been shown to be unrealistic ([80] and references chiometries with the general formula [(Ca(OH)2)nCa(Sr
therein) and amorphous SiO2 co-exists with C-S-H gel of Ca/Si (OH)2)nSr(KOH)nK(NaOH)nNaSiO2H2O]nSi were considered for these

Dissolved K Dissolved Ca
0 0
C:S=0.85
-1 -1
log[K],dissolved C:S=1.2
log[Ca],dissolved
C:S=1.5
-2 C:S=1.8 -2
exp.0.85
-3 exp.1.2 -3
exp.1.5
-4 exp.1.8 -4
-5 -5
-4 -3 -2 -1 0 -4 -3 -2 -1 0
log[K],added log[K],added
Dissolved Na Dissolved Ca
0 0
C:S=0.85
-1 -1
C:S=1.2
log[Na],di ssol ved
log[Ca],di ssol ved

C:S=1.5
-2 C:S=1.8
-2
exp.0.85
-3 exp.1.2 -3
exp.1.5
-4 exp.1.8 -4
-5 -5
-4 -3 -2 -1 0 -4 -3 -2 -1 0
log[Na],added log[Na],added
Fig. 10. Comparison of sorption isotherms for K or Na calculated using the ECSH-II Aq-SS model (curves) with the data for K and Na sorption [81] (scattered
symbols). Abscissa: log10 moles of added K or Na per 1 kg H2O; ordinate: log10 molar.
models. To develop these models, the nCa, nSr, … coefficients were With no known thermodynamic properties of structural analogues
adjusted in order to minimize the standard deviations of estimated Go298 available, the standard entropy and heat capacity of the ECSH end
values for model end members in trial GEM calculations for a number of members were estimated assuming linear dependencies of entropy and
experimental data points. These trial GEM calculations employed: (1) heat capacity effects of reactions on the Ca/Si ratio in C-S-H [11]:
(CaO) x (SiO2 ) y (H2 O)z = ySiO2 + xCa(OH) 2 + (z

the Nagra-PSI database [24]; (2) many experimental data points at
different Ca/Si, Sr/Si, Na/Si, K/Si ratios; and (3) varying stoichiometry
x)H2 O (1)
H end member) = y(silica) + x(portlandite) + (z

coefficients of solid solution end members within the ranges of
(C S x)(water)
0.1 < nSi < 2, 0 < nCa < 1.6, 0 < nSr < 2, 0 < nK < 2, and
= y (61.054 + 5.357 x/y)
0 < nNa < 2. o
r S 298 (2a)
Followed by ‘forward GEM modelling’ of Sr uptake data in pure
= y (31.881
water and in artificial cement water (ACW), this procedure resulted in o
r Cp 298 11.905 x/y) (2b)
ideal ECSH-II and ECSH-I solid solution models that provided the op-
timal description of data (over 96 experiments published in [92] and Using Eqs. (2a) and (2b), and were calculated and ΔrSo298 ΔrCpo298
additional in-house Sr uptake data on C-S-H in water and in ACW). rounded off to the nearest whole numbers. The ΔrHo298 values were
ΔfG°298 values for Na- and K-containing end members were also fine- calculated from log10Ko298 and ΔrSo298 values together with the So, Cpo,
tuned using literature data [81] on Na and K uptake isotherms in C-S-H ΔfH° and ΔfG° values at Tr = 298.15 K using the ReacDC module of
(Fig. 10). The ECSH-1 and ECSH-2 models can realistically describe the GEM-Selektor code and thermodynamic properties of water, portlandite
uptake of cations and the decrease of (maximum) Ca/Si ratios in and amorphous silica from the GEMS version of the PSI/Nagra 12/07
equilibrium with portlandite upon increasing alkali concentration in TDB [22,23] and Cemdata18 databases. The resulting thermodynamic
aqueous solution. However, it was not possible to use the same ΔfG°298 properties (Table 4) are expected to suffice for temperatures between 0
of SrSH end member to model isotherms of Sr uptake in C-S-H prepared and 90 °C within 0.5 pK units uncertainty.
in water and in the artificial cement pore water (ACW with pH ≈13.3 Next, the So298, Cpo298 and values of the SrSH, NaSH and KSH end
at 25 °C, containing 0.18 M KOH, 0.114 M NaOH and 1.2 mM Ca(OH)2). members of ECSH phases were evaluated. This was done by taking the
We believe that the ΔfG°298(SrSH) difference (up to 3.4 kJ/mol) can properties of their reference calcium hydroxide counterpart C1S1H2 and
probably be explained by different silica polymerization and cation then either subtracting or adding respective properties of solid por-
exchange capacity of C-S-H due to the presence of alkali. We anticipate tlandite Ca(OH)2, as well as solid Sr(OH)2, solid NaOH or solid KOH
that ECSH-1 and ECSH-2 will be replaced by more accurate C-S-H-K-Na from Wagman et al. [93]. This is equivalent to assuming ΔrSo298 = 0 and
models in the near future. ΔrCpo298 = 0 for the reactions:

CSH2 + SrH = SrSH2 + CH (3a) with U(VI) end members [97] and with Al and Na end members [90].
0.2 CSH2 + 0.5 KOH = (KOH)0.5 (SiO2 )0.2 (H2 O) 0.2 + 0.2 CH
The ideal CSH3T SS model [11] produces quite realistic curves for so-
(3b)
lubility of the synthetic C-S-H co-precipitation (double decomposition)
0.2 CSH2 + 0.5 NaOH = (NaOH)0.5 (SiO2 )0.2 (H2 O) + 0.2 CH. (3c) data. More accurate C-S-H multi-site solid solution models are in de-
velopment.
For these calculations, ∆rSo298, and ΔrCpo298, So298, Cpo298
of the re-
ference compound CSH2 were computed using Eqs. (1), (2a) and (2b). 2.7.5. CNASH_ss model
Note that the stoichiometries of the K and Na C-S-H end members de- CNASH_ss model [90] includes Al and Na and represents an exten-
fined by reactions 3b and 3c correspond to N0.25S0.2H0.45 or sion of the CSH3T model that was optimised for alkali activated sys-
K0.25S0.2H0.45, but not N0.25S0.2H0.3 or K0.25S0.2H0.3 as defined in Kulik tems. The calcium (alkali) aluminosilicate hydrate (C-(N-)A-S-H) gel-
et al. [88]. The respective values for ΔfG°298, and ΔfHo298 are sum- like phase that precipitates in alkali-activated cements contains sig-
marised in Table 4. nificantly less Ca, more Al and alkali and has a more densely packed
structure than the C-(A-)S-H which forms in hydrated PC-based mate-
2.7.3. CSHQ model rials [98,99]. However, both phases are based on the same defect-to-
CSHQ model [11] was developed in order to address some known bermorite structure. In alkali-activated slag cements (an exemplary
shortcomings of the earlier CSH-I and CSH-II models [29,85], namely ‘high-Ca’ alkali-activated material [100]), the C-(N-)A-S-H phase typi-
insufficient connection to the C-S-H structure and the unrealistic as- cally has a Ca/Si ≈ 1 and an Al/Si ≤ 0.25 [90].
sumption of ideal mixing between tobermorite-like and amorphous si- Many solubility and chemical composition data for the C-(N-)A-S-H
lica end members. It was based on structural data supporting the defect- system have been published. Much of this data was used to develop an
tobermorite model [94–96], represented as a solid solution model with ideal solid solution thermodynamic model (CNASH_ss), including con-
four different structural sites (sublattices) [11]: [BTI+2]1:[TU−]2: figurational entropy terms, which explicitly includes mixing of Al and
[CU0]2:[IW0]5. The main assumption was that in BTI sites, the in- Na [90]. The CNASH_ss model enables Al incorporation into C-(N-)A-S-
corporation of Ca2+ ion in the interlayer occurs simultaneously with H gel to be explicitly considered in thermodynamic modelling simula-
the removal of a bridging tetrahedron in the silica “dreierketten” chain, tions. The CNASH_ss model has been applied to simulate phase as-
and this process is reversible. Excess calcium can also be incorporated semblages in NaOH, sodium silicate, Na2CO3, and Na2SO4-activated
as a Ca(OH)2 moiety, either interstitially in the tobermorite interlayer, slag systems [74,101]. This model is also applicable to thermodynamic
or forming domains of jennite-like structure. This was accounted for by modelling of PC-based materials; however, it less closely represents the
an exchange of a vacancy with Ca(OH)2 in CU sites. The occupation of full body of available solubility data for the C-S-H phase [102] at Ca/
TU and IW sublattices was fixed as 2CaSiO3.5− and 4H2O + vacancy, Si > ~1.3 than other C-S-H thermodynamic models, e.g. [11,80].
respectively. This led to four end members with stoichiometries de- CNASH_ss closely represents the full set of solubility data for the C-(N-
pending on the assumed Ca2+/H+ ratio in BTI sites. )A-S-H gel phase down to Ca/Si = 0.67. Therefore, we recommend
This solid solution model has a correct built-in dependence of the using CNASH_ss for alkali activated systems rather than hydrated PC
mean silica chain length (MCL) on Ca/Si ratios. By downscaling the systems, where we recommend the use of CSHQ or C-S-H-II.
end-member stoichiometries to Si = 1.0 and adjusting the Go298 values Additional solubility data for C-(N-)A-S-H gel not used to validate
of end members, the CSHQ model could be fine-tuned to various C-S-H CNASH_ss were recently published, including for C-(N-)A-S-H gels at
solubility data sets [11]. The downscaled ideal CSHQ model (Table 4) synthesis temperatures of 7 °C, 50 °C and 80 °C [103,104] and using K
provides a reasonable fit to the variety of C-S-H solubility data in the rather than Na [84,105]. Future refinement to the CNASH_ss thermo-
[Ca]-[Si], [Ca]-C/S and [Si]-Ca/Si spaces as discussed in more detail in dynamic model should include these data and formally extend the
[11]. model to different temperatures and alkali type.
In this Cemdata18 database, two end members for K and Na (similar
to those from the ECSH model) were provisionally added to improve 2.7.6. General remarks
predictions of pH and composition of the PC porewater. The extension During the last 20 years, ideal solid solution models of C-S-H have
to cover the uptake of alkalis by C-S-H was based on an ideal solid evolved starting from simple ideal solid solutions using full end-
solution model between jennite, tobermorite, [(KOH)2.5SiO2H2O]0.2 member mixing up to recent truly multi-site mixing models consistent
and [(NaOH)2.5SiO2H2O]0.2 as proposed by Kulik et al. [88] and using with both solubility data and structural/spectroscopic data. Because
the thermodynamic data reported in [89]: ΔfG° = −440′800 J/mol end members in multi-site solid solutions are constructed of moieties
and − 431′200 J/mol (at 20 °C) for [(KOH)2.5SiO2H2O]0.2 and substituting each other on different sublattices, such models have the
[(NaOH)2.5SiO2H2O]0.2, respectively. best potential for: (1) extension by adding moieties for other elements
of interest (e.g. K, Na, Al, U, Sr) in their respective sites; (2) generating
2.7.4. CSH3T model all possible end members; and (3) parameterizing end members based
CSH3T model [11] was aimed at more consistency with the to- on available solubility, element uptake, and spectroscopic data (e.g.
bermorite-like structure of C-S-H phases at Ca/Si < 1.5. The evidence using the GEMSFITS code [106]) and are the subjects of ongoing re-
of interlayer ordering in tobermorite-like C-S-H with 0.9 < C/ search.
S < 1.25 [96] has led to setting the CU sites always vacant, and to For the calcium silicate hydrate complexes, CaH3SiO4+ (CaHSiO3+
splitting the BTI sublattice into two ([BTI1+]1:[BTI2+]1:[TU−]2: +2H2O) and CaH2SiO40 (CaSiO30 + 2H2O), the reported complex for-
[IW0]4) with substitutions of Si0.5OH+ by HO0.5Ca0.5+. This yielded a mation data show a significant scatter. In particular, complex formation
solid solution model with end members TobH (C2S3H5), T5C constants for CaH2SiO40 vary by more than one log unit. While the PSI/
(C2.5S2.5H5), T2C (C3S2H5), connected by an ordering reaction Nagra TDB [22,23] reports a complex formation constant of 104.6 for
½TobH + ½T2C = T5C. The model has a built-in dependence of the the reaction Ca2+ + SiO32− → CaSiO30 (see Table 2), which has a large
mean chain length on the composition, consistent with measured values effect on the silicon concentrations in presence of C-S-H at Ca/Si > 1
[95] for the co-precipitated tobermorite-like C-S-H. The CSH3T model [80], no such constant is defined in the PHREEQC database [18].
[11] in its downscaled form (Table 4) can be computed just using a Walker et al. [80] recommended to use a constant of 104.0, making the
simple ideal mixing model. The CSH3T model has been later extended complex less important, while recently an even lower complex

Monosulfoaluminate
Ettringite M and/or H Monocarbonate
Hemicarbonate
C4AF
1 year
7 days
1 day
PC4
PC
8 9 10 11 12
Angles 2q ((degrees)) CuKa
Fig. 12. Experimentally observed phase assemblage in a PC without additional
limestone (PC) and with 4 wt% of limestone (PC4); reproduced from [123].
Fig. 11. Evolution of the solubility product (KS0) of magnesium silicate hy-
drates at room temperature as a function of the total Mg/Si; referring to reac- aluminosilicate minerals from adjacent rocks and the formation of
tions using Mg2+, SiO2 and H2O as indicated in Table 2. Adapted from [118]. secondary zeolite minerals [119] in the context of deep underground
nuclear waste repositories. Zeolite formation also occurs in alkali acti-
formation constant of 102.9 has been derived based on titration ex- vated cement systems. These zeolites are often related to the poorly
periments [107]. This large scatter of data results in very diverging crystalline N-A-S-H (sodium-aluminium-silicate-hydrate) and K-A-S-H
assessment of the importance of the CaSiO30 complex at Ca/Si > 1 and (potassium-aluminium-silicate-hydrate) gels that form in these systems
has a significant impact on the C-S-H solubility as this complex accounts [74,103]; the type of gel formed depends on the presence of Na+ or K+,
for about 90% of aqueous dissolved silicon in equilibrium with both C- cation concentrations, the relative degree of saturation of the liquid
S-H and portlandite. Dedicated investigations not only of calcium sili- phase with respect to silica, pH and temperature [120]. Several papers
cate hydrate complexes but also of other possible complexes between in recent years estimated solubility data for different zeolites, based
aluminium, calcium and silicate at high pH values are urgently needed. mainly on heat capacity and enthalpy measurements [47,74,121]. This
may lead to considerable bias in the estimated solubility data in the
2.8. Magnesium silicate hydrates range of several log units due to uncertainties associated with the
measurements of enthalpy data. The determination of solubility data for
The formation of magnesium silicates hydrate (M-S-H) has been zeolites has been hindered by variability in cation composition (Ca, Na,
observed at the interfacial zone of cement paste with clays K), Al/Si ratios, H2O contents and atomic structure, and also their slow
[67,108,109] and/or as secondary products from the degradation of reaction kinetics.
cement pastes by groundwater or seawater [110–112]. The combina- In 2017, two independent studies [41,103] reported very similar
tion of leaching and carbonation of the cement paste decreases pH at solubility products for zeolite Y and X (or for N-A-S-H gel with Al/
the surface of the cement, decalcifies C-S-H and leads the formation of a Si = 0.5 and Al/Si = 0.8) based on experimental data. The data for
Mg-enriched phase, M-S-H. M-S-H phases are poorly ordered but have a zeolite X(Na), zeolite Y(Na) and chabazite [41] make it possible to
layered structure with tetrahedral silica arranged in sheets similar to predict zeolite formation in sodium activated cements; data for po-
clay minerals, have variable Mg/Si from ≈0.8 to Mg/Si ≈ 1.2 and are tassium-based zeolites are still missing in the Cemdata18 database. Also
stable at pH values between 7.5 and 11.5 [40,113–115]. Given the data for natrolite and zeolite P(Ca) have been included [41]. In ex-
difference in structure and pH domains, most studies [114–117] ob- periments with high pH values their formation was kinetically hindered
served the precipitation of distinct C-S-H and M-S-H phases and not of a (although natrolite and zeolite P(Ca) were more stable than zeolite X
mixed magnesium calcium silicate hydrate phase. Solubility measure- (Na), zeolite Y(Na) and chabazite). Thus we recommend that natrolite
ments [40,113,118] indicated an only slightly higher solubility of the and zeolite P(Ca) should be considered in modelling the interface be-
poorly ordered M-S-H in comparison to crystalline magnesium-silicates tween cement and adjacent rocks. However, their formation may be
such as talc, antigorite or chrysotile as shown in Fig. 11. The ideal solid supressed in models for alkali activated systems, where zeolite X(Na),
solution model for M-S-H published by Nied et al. [40] has been se- zeolite Y(Na) and chabazite or their amorphous or nanocrystalline
lected for the present version of the database. As several groups precursors are formed [122].
[113,114,118] are currently working on thermodynamic data for M-S-
H, we expect that more sophisticated models will be published in the 3. Comparison Cemdata07 with Cemdata18
coming years.
The updates since the first cemdata version, cemdata07 (published
2.9. Zeolites in 2008), are significant. In particular, the distribution of iron and
aluminium, the volume and Ca/Si in C-S-H as well as the alkali con-
Interactions of highly alkaline solutions in hydrated PC systems with centrations in the pore solution in PC can significantly affect thermo-
service environments will likely result in the partial dissolution of dynamic modelling results. To illustrate these differences, the effect of

A) Cemdata07 B) Cemdata18
monosulfate
1.70 hemicarbonate monocarbonate 1.70
Volume [cm3/100 g unhydrated cement]

Volume [cm3/100 g unhydrated cement]
60 60
55 55
hemicarbonate ettringite
50 monocarbonate 50
monosulfate 1.65 C3(A,F)S0.84H4.32 1.65
45 45
C3(A,F)H6 ettringite
40 40
C-S-H
35 35
Ca/Si
Ca/Si
1.60 30 1.60
30
C-S-H Ca/Si
25 25
Ca/Si
20 20
15 1.55 15 1.55
portlandite
10 10
portlandite hydrotalcite
5 hydrotalcite 5
0 1.50 0 1.50
calcite 10 0 2 4 6 8
0 2 4 6 8 calcite 10
Calcite [wt.%] Calcite [wt.%]
Fig. 13. Comparison of calculated solid phase assemblage using A) Cemdata07 and B) Cemdata18 assuming complete hydration of PC using the composition reported
in [123].
Fig. 14. Effect of the amount of limestone on the phase assemblage and the distribution of aluminium and iron in hydrated PC calculated using Cemdata07 (A, C) and
Cemdata18 (B, D).
limestone on the same PC was calculated with Cemdata07 and 3.1. Effect of limestone on solid and liquid phase composition
Cemdata18 and compared below. The effect of relative humidity on
calculated hydrates is used below as a second example. These com- The influence of limestone on cement hydration has been widely
parisons concentrate on PC, as compiled specific data for alkali acti- studied and was the subject of several publications by the authors
vated materials are only now available (in this paper). [2,20,123]. Experimental investigations showed that the presence of
calcium carbonate prevents the destabilisation of ettringite to

Fig. 15. Effect of the amount of limestone on the phase assemblage and the distribution of aluminium and iron in hydrated PC calculated using A) Cemdata07 and B)
Cemdata18.
monosulfate at long hydrations times and stabilises monocarbonate monocarboaluminate thus theoretically preventing their presence.
together with ettringite (see e.g. [123–125] and Fig. 12). Since monosulfate, hemi- and monocarboaluminate are experimentally
Also thermodynamic modelling [2,20,123] (mainly using the Cem- observed in hydrated PC, it was assumed that this was due to a kinetic
data2007 database) showed that the presence of small amounts of hindrance in the formation of siliceous hydrogarnet and that possibly a
limestone significantly impacted the mineralogy of hydrated cements. later conversion of hemi- and monocarboaluminate to siliceous hy-
In the absence of any limestone no ettringite but only monosulfate as drogarnet could occur.
well as of a small amount of katoite (C3(A,F)H6) was predicted as shown The new data for (C3A0.5F0.5S0.84H4.32) by Dilnesa et al. [9], in-
in Fig. 13A. The presence of a small amount of limestone was calculated cluded in Cemdata18, suggest that mixed Al- and Fe-containing silic-
to stabilise hemicarbonate and at higher dosages monocarbonate plus eous hydrogarnet can coexist with monosulfate, hemi- and mono-
ettringite, resulting in an increase of the total volume. The higher vo- carboaluminate at ambient conditions, which is in better agreement
lume in the presence of a small amount of limestone due to the stabi- with the observed experimental data presented in Fig. 12 and elsewhere
lization of ettringite has been found to have a positive effect on the [123–125]. Fig. 13B displays the predicted phase assemblage of a hy-
mechanical properties of PC and blended cements [20,124]. drated PC with limestone using Cemdata18 as given in Tables 1-4;
The stability of siliceous hydrogarnet was a matter of debate during employing CSHQ and M4AH10. The formation of hemi- and mono-
the development of Cemdata07 and in most calculations with carboaluminate accompanied by a stabilization of ettringite instead of
Cemdata07 the formation of siliceous hydrogarnet C3AS0.8H4.4 had monosulfoaluminate was correctly predicted by both datasets. As
been suppressed assuming kinetic hindrance. Based on the data com- shown in Fig. 13 the biggest difference between the two datasets is the
piled in Cemdata07, which originated from measurements from prediction of a katoite-type siliceous hydrogarnet phase
[7,72,126], ettringite and siliceous hydrogarnet were calculated to be (C3A0.5F0.5S0.84H4.32), modelled as solid solution with a varying alu-
significantly more stable than monosulfate, hemi- or mina and iron by using Cemdata18, together with hemi- and
Fig. 16. Calculated specific volume changes of a hydrated model mixture consisting of C3A, portlandite and with fixed sulfate ratio (SO3/Al2O3 = 1, molar bulk ratio)
in dependence of changing calcite content at 25 °C; calculated using Cemdata18.

Fig. 17. Calculated specific volume changes of a hydrated model mixture consisting of C3A, portlandite and with fixed sulfate ratio (SO3/Al2O3 = 1, molar bulk ratio)
in dependence of changing calcite content at 25 °C., as shown in Fig. 16 for the Systems A, B and C; calculated using Cemdata18.
monocarboaluminate and ettringite throughout the modelled compo- [1,29,123,127], Fig. 15) although there are differences between mea-
sition range independently of the CaCO3 content. sured and calculated values, in particular for Ca and Al for Cemdata07
The consideration of the siliceous hydrogarnet solid solution in and for sulfate and silicon for Cemdata18.
Cemdata18 led to a quite significant redistribution of alumina and iron
within the phase assemblage. Whereas with Cemdata07 around 70% of
3.2. Effect of relative humidity on hydrated cements
the available alumina was bound in AFm phases (see Fig. 14A) the
predictions based on Cemdata18 suggest that only about 25% of alu-
Using the thermodynamic properties of phases with different water
mina is bound in AFm phases and ~30% in the hydrogarnet phase
contents described in Section 2.5 and Table 1 it was possible to predict
(Fig. 14B). For iron, the difference is even more drastic. The predictions
the drying behaviour of hydrated systems.
with Cemdata18 suggest that close to 100% of the iron is bound by the
Drying of the CaO-Al2O3-SO3-CO2-H2O was simulated because it is
siliceous hydrogarnet solid solution (Fig. 14D) which is also in agree-
directly relevant to PC and limestone blended cements. The initial
ment with experimental observations [64–66], where predominantly
model mixture contained C3A, portlandite (CH), calcium sulfate (SO3/
the formation of mixed aluminium and iron containing hydrogarnet
Al2O3 = 1 M bulk ratio), and varying amounts of calcite at 25 °C. The
phases in close proximity to the original ferrite phases was observed in
amount of solids was kept constant at 100 g and reacted with 90 g
hydrated cements.
water. A diagram of the specific volume changes of the hydrated mix-
The binding of alkalis in C-S-H lowers the alkali and hydroxide
ture with respect to calcite content is shown in Fig. 16.
concentrations [81,84,88] in the pore solution of hydrated PC and thus
Due to their differing AFm-AFt mineralogy hydrate phase assem-
the pH values from above 14 to ~13 to 13.5 [1,29,123,127]. The dis-
blages A, B and C in Fig. 16, with 0%, 7% and 13.2% of calcite re-
regard of alkali binding by C-S-H would result in very high predicted pH
spectively, were selected as initial hydrated systems for the drying
values of 14 and above, which does not agree with measurements of the
modelling. Drying was simulated by continuously removing water from
pore solution composition [5,29]. As in 2007 no thermodynamic models
the assemblages until a RH of zero was reached. The investigated sys-
to describe the uptake of alkali in C-S-H were available, distribution
tems were:
coefficients (Kd values) were used together with Cemdata07 in most
calculations of hydrated cements as described in details e.g. in
- System A: monosulfoaluminate (Ms14) and portlandite (CH)
[1,29,123]. The use of distribution coefficient allowed predicting the
- System B: ettringite (Ett32), hemicarboaluminate (Hc12) and por-
alkali concentrations in PC relatively well as shown in Fig. 15A, but the
tlandite (CH)
approach was not adequate to predict alkali uptake in low Ca/Si C-S-H
- System C: ettringite (Ett32), monocarboaluminate (Mc11) and por-
present in blended cements. Kd values do not account for competitive
tlandite (CH)
sorption on specific sites as would be expected for the C-S-H gel, and also
tend to be experiment-specific and so cannot generally be applied to
Fig. 17a, b and c present the evolution of specific solid volume as a
other systems under different conditions. In the Cemdata18, the uptake
function of RH. We can see that dehydration happens stepwise at cri-
of alkalis by C-S-H is modelled by introducing additional Na- and K-
tical RH stability limits of the phase assemblages, representing in-
endmembers ([(NaOH)2.5SiO2H2O]0.2 and [(KOH)2.5SiO2H2O]0.2) in the
variant points where the RH is fixed due to phase rule restrictions. At
CSHQ model, as described above (Section 2.7). The introduction of these
this critical RH two hydration states of the same cement hydrate coexist
provisional data simplify the modelling, as no additional Kd values have
and buffer the humidity in a similar manner as conventional drying
to be introduced in the models, and allows the calculation of alkali up-
agents. Another important finding is that the addition of calcite and the
take over the whole range of Ca/Si ratios, although the agreement be-
formation of carboaluminates and ettringite will enhance the dimen-
tween measured and calculated alkali concentrations is only satisfacto-
sional stability of hydrated cement paste and makes it less sensitive to
rily, as shown in Fig. 15B. Due to the lack of appropriate models for
humidity fluctuations, which appears to be relevant for limestone
sodium and potassium uptake in C-S-H valid over the complete range of
blended cements. Due to the presence of monocarboaluminate and et-
Ca/Si, the modelling of alkali and hydroxide concentrations in the pore
tringite system C is the most stable phase assemblage, which only de-
solution remains a challenge.
composes at very low humidities (below 2% RH) whereas mono-
The trends in the concentrations of calcium, sulfate, silicon and
sulfoaluminate quickly loses part of its interlayer water at < 99% RH.
aluminium are generally correctly reproduced by both models (see e.g.
Something important to keep in mind is that, although

experimentally we observe the changes shown in Fig. 17, several of These additions improve the reliability of thermodynamic model-
these dehydration processes are metastable with respect to other phase ling of cement systems, in particular for alkali activated materials and
assemblages. This has to be considered when predicting the drying for processes at cement/environment interfaces, where hydrates such as
behaviour of cementitious systems. thaumasite, Friedel's salt, M-S-H, and zeolites may form.
The consideration of siliceous hydrogarnet solid solution in
4. Conclusions Cemdata18 leads to a quite significant redistribution of alumina and
iron within the phase assemblage in PC; the predictions based on
The Cemdata18 database summarised in this paper can reliably Cemdata18 suggest that alumina is bound not only in AFt, AFm phases
calculate the type, composition, amount and volume of hydrates formed and hydrotalcite but also in siliceous hydrogarnet phase while all hy-
and the pH and composition of the pore solution during hydration and drated iron is present in siliceous hydrogarnet.
degradation of cementitious systems. The Cemdata18 database, as Several C-S-H solubility models as well two models for hydroxide-
compiled in Table 1 to Table 4, includes carefully selected thermo- hydrotalcite are available (Table 4, Appendices A and B). The CSHQ
dynamic data published in the literature based on critical reviews and the OH-hydrotalcite with Mg/Al = 2 are well adapted for PC sys-
supplemented with new experimental data. Data for solids commonly tems. Although CSHQ is able to describe the entire range of Ca/Si ratios
encountered in cement systems in the temperature range 0–100 °C, in- encountered, it is best used for high Ca/Si C-S-H as it lacks the ability to
cluding C-S-H, M-S-H, hydrogarnet, hydrotalcite-like phases, some predict aluminium uptake, however, this is less important in PC where
zeolite, AFm and AFt phases and their respective solid solutions has the aluminium content is relatively low. For alkali activated binders,
been compiled. The Cemdata18 database is an update of the Cemdata07 the CNASH model has been developed for C-S-H type calcium (alkali)
and Cemdata14 databases, and is compatible with the GEMS version of aluminosilicate hydrate gels with lower calcium but higher aluminium
the PSI/Nagra 12/07 TDB [22,23]. Cemdata18 TDB is freely down- and alkali content. An Mg-Al layered double hydroxide model with
loadable (http://www.empa.ch/cemdata) in formats supporting the variable Mg/Al ratio is also available for use in alkali activated cement
computer programs GEM-Selektor [13,14] and PHREEQC [18]. Further systems.
details are available in Appendices A and B. Despite significant additions to the Cemdata18 TDB, several im-
The most important additions to the Cemdata18 TDB include: portant gaps still exist in the database. In particular, reliable thermo-
dynamic data for alkali, aluminium and water uptake in C-S-H applic-
• C-S-H: able to high and low Ca/Si C-S-H and M-S-H, data for hydrotalcite-like
- CSHQ model for Portland and blended cements, the uptake of al- phases of variable composition and for different interlayer ions, data for
kalis by C-S-H is modelled by additional Na- and K-containing end further zeolites derived from experimental solubility measurements,
members data for aqueous complexes which possibly form at high pH values as
- CSH3T model that corresponds to pure defect-tobermorite struc- well as data for the reaction products of alkali silica reaction are
ture with ordering at Ca/Si ratio close to 1.0, and forms the basis needed. However, these data gaps should be viewed as possible future
for CNASH-ss model improvements rather than barriers to use thermodynamic modelling:
- C-(N-)A-S-H model for alkali activated materials (CNASH-ss), Cemdata18 database has already been successfully applied to model
which calculates the uptake of aluminium and sodium in low Ca/Si hydrated PC, calcium aluminate, calcium sulfoaluminate and blended
C-S-H cements, and also alkali activated materials. Cemdata18, therefore,
• iron-containing hydrates, in particular for the mixed Fe-Al-hydro- enables improved characterisation and understanding of the chemistry
garnet solid solution, C3FS0.84H4.32-C3A0.5F0.5S0.84H4.32, which takes and related in-service performance properties of a wide range of cement
up iron and a part of the aluminium in hydrated cements systems, including the most common types.
• AFm and AFt-phases with different water contents to describe the
effect of water activity and drying on hydrates
• amorphous, microcrystalline AH3 and gibbsite to study the effect of Acknowledgements
AH3 solubility on the hydrates in calcium aluminate and calcium
sulfoaluminate cements The partial financial support from the NANOCEM consortium
• chloride, nitrate and nitrate-containing AFm phases (www.nanocem.org), the Swiss National Foundation (SNF grants No.
• thaumasite and the uptake of carbonates in SO4-ettringite. 117605, 132559, 130419 and 200021_169014), from Nagra,
• description of the variation in Mg/Al in layered double hydroxides Wettingen, Switzerland, and from the BMBF ThermAc3 Verbundprojekt
(hydrotalcite-like phases) observed in alkali activated materials (Germany) are gratefully acknowledged. The authors thank also Tres
• data for M-S-H and some Na- and Ca-based zeolites, which can form Thoenen, Ravi Patel and Andres Idiart for their support on the
at the interaction zone of cement with clays, rocks or seawater and PHREEQC version.
in alkali activated materials.
Appendix A. Cemdata18 dataset in GEMS format
Cemdata18 database in GEM-Selektor v.3 format can be freely downloaded (http://www.empa.ch/cemdata) and is fully compatible with the
GEMS version of the PSI/Nagra 12/07 TDB [22,23] (http://gems.web.psi.ch). As several alternative C-S-H models, as well as two models for
hydroxide-hydrotalcite are available, the user needs to select the appropriate models during the generation of new projects, as illustrated in Fig. A.1.
The CSHQ and the OH-hydrotalcite with Mg/Al = 2 are well adapted for Portland cement systems (select cemdata, pc, ht. and cshq as indicated at
the left hand side of Fig. A.1).
For alkali activated binders, the CNASH model has been developed for C-S-H type calcium (alkali) aluminosilicate hydrate gels with lower
calcium but higher aluminium and alkali content. An Mg-Al layered double hydroxide model with variable Mg/Al ratio is also available for use in
alkali activated cement systems. For alkali activated binders, the selection of cemdata and aam and deselection of pc, including ht and csh is
recommended as illustrated at the right hand side of Fig. A.1.

B. Lothenbach et al.

Fig. A.1. Selection of modules of Cemdata18 and related databases in GEM-Selektor to model PC (Portland-cement) systems (left) and to model AAM (alkali-activated materials). For PC systems, one of four alternative

solid solution models of C-S-H should be selected (see Section 2.7); selection of Fe-containing solid solutions (“ss-fe3” module) is also optional.
Appendix B. Cemdata18 dataset in PHREEQC format
To enable users to model cementitious systems using the Cemdata18 dataset with the popular PHREEQC geochemical speciation code [18], a
PHREEQC “.dat” format database of the Cemdata18 dataset (CEMDATA18.dat) is provided for download from http://www.empa.ch/cemdata. This
LMA (Law of Mass Action) type dataset has been generated using the reaction generator module of the ThermoMatch code (Miron et al. in pre-
paration) and exported into the PHREEQC format “.dat” file using the ThermoMatch database export module. The reaction generator algorithm is
based on the matrix “row reduce” method described by Smith and Missen [128]. In this process, all aqueous and solid species from the Cemdata18
GEM-Selektor database were considered. The supplementary data for aqueous, gaseous and solid species corresponding to the list of elements
covered by Cemdata18 were selected from the GEMS version of the PSI/Nagra TDB [22,23]. The latter and the Cemdata18 GEM database are
mutually consistent, and should be used together in GEMS codes for modelling cementitious systems.
To generate PHREEQC-style reactions for product species, firstly the following master species were selected based on their generic predominance:
Ca+2, Mg+2, Sr+2, Na+, K+, H+, CO3−2, SO4−2, Cl−, NO3−, AlO2−, FeO2−, SiO20, H2O0, and e−(electron). Using selected master species, the
reactions were automatically generated for the remaining (product) species, and their properties at 25 °C and 1 bar were calculated. Formation
reactions were generated for aqueous product species, and dissolution reactions - for gaseous and solid product species. The LMA dataset of reactions
was then exported into a PHREEQC “dat” file (CEMDATA18.dat) using the ThermoMatch database export module. Parameters for the logKo = f(T)
analytical expressions were calculated for the 3-term extrapolation method that assumes the ΔrCpo to be not zero and independent of temperature.
These reported parameters are used by PHREEQC for calculating the log10Ko as a function of temperature. Such temperature extrapolations of
log10Ko should be valid at least up to 100 °C.
Table B.1 contains the generated formation reactions for the aqueous product species, together with the values for reaction standard effects at
25 °C and 1 bar. Table B.2 contains the generated dissolution reactions for gaseous and solid product species, together with the reaction standard
effects at 25 °C and 1 bar. Table B.2 contains, in addition to the Cemdata18 database as detailed in Table 1 to Table 4, also the thermodynamic data of
all solids composed of Al, C, Ca, Cl, Fe, H, K, Mg, N, Na, S, Si or Sr compiled in the GEMS version of the PSI/Nagra 12/07 TDB [22,23], needed to
allow the generation of a compatible dataset in PHREEQC. Figs. B.1, B.2, and B.3 show comparisons of cement-related modelling problems between
GEM-Selektor (using GEM-type Cemdata18) and PHREEQC (using LMA-type Cemdata18 CEMDATA18.dat). For the PHREEQC calculations,
PHREEQC for Windows version 2.18.00 (uses PHREEQC-2 source version 2.18.3-5570) was used. In all three cases, the considered solid solutions
were modelled in PHREEQC using the simple ideal mixing model.

Table B.1
Product aqueous species reactions from master species, together with their reaction properties at 25 °C and 1 bar.
Product Substance Reaction log10Ko298 ΔrGo298 ΔrHo298 ΔrSo298 ΔrCpo298 ΔrVo298

[J/mol] [J/mol] [J/K/mol] [J/K/mol] [J/bar]
Al(SO4)+ S|6|O4–2 + AlO2- + 4H + = Al(SO4)+ + 2H2O@ 26.8 −152,857 −159,164 −21.2 261.8 0.8
Al(SO4)2- 2S|6|O4–2 + AlO2- + 4H + = Al(SO4)2- + 2H2O@ 28.8 −164,272 −165,197 −3.1 463.6 3.2
Al+3 AlO2- + 4H + = Al+3 + 2H2O@ 22.9 −130,595 −176,821 −155.0 71.1 −1.9
AlHSiO3+2 AlO2- + 3H+ + SiO2 = AlHSiO3+2 + H2O 20.5 −116,839 −106,764 33.8 −136 3.3
AlO+ AlO2- + 2H+ = AlO+ + H2O@ 12.3 −70,124 −73,951 −12.8 −0.7 0.9
AlO2H@ AlO2- + H + = AlO2H@ 6.4 −36,798 −21,554 51.1 −160.2 0.4
AlOH+2 AlO2- + 3H+ = Al(OH) + 2 + H2O@ 17.9 −102,299 −127,582 −84.8 180.4 0.6
AlSiO5–3 AlO2- + H2O@ + SiO2@ = AlSiO5–3 + 2H+ −22.6 129,059 71,975 −191.5 0.0 −7.8
Ca(CO3)@ CO3–2 + Ca+2 = CaCO3@ 3.2 −18,405 16,462 116.9 196.4 0.9
Ca(HCO3)+ CO3–2 + Ca+2 + H+ = Ca(HCO3)+ 11.4 −65,269 −13,562 173.4 554.0 3.8
Ca(HSiO3)+ Ca+2 + H2O@ + SiO2@ = Ca(HSiO3)+ + H+ −8.6 49,143 30,328 −63.1 48.9 −2.2
Ca(SO4)@ Ca+2 + S|6|O4–2 = CaSO4@ 2.3 −13,128 4336 58.6 192.4 1.0
CaOH+ Ca+2 + H2O@ = Ca(OH)+ + H+ −12.8 72,950 77,301 14.6 −38.4 0.6
CaSiO3@ Ca+2 + H2O@ + SiO2@ = CaSiO3@ + 2H+ −18.5 105,827 48,743 −191.5 0.0 −1.6
CH4@ CO3–2 + 8e- + 10H+ = C|-4|H4@ + 3H2O@ 38.2 −217,923 −270,138 −175.1 677.3 9.8
ClO4- Cl- + 4H2O@ = Cl|7|O4– + 8e- + 8H+ −187.7 1,071,500 1,181,308 368.3 −87.6 −4.6
CO2@ CO3–2 + 2H + = CO2@ + H2O@ 16.7 −95,216 −24,408 237.5 607.8 5.7
Fe(CO3)@ CO3–2 + e- + 4H+ + Fe|3|O2– = FeCO3@ + 2H2O@ 39.0 −222,615 −216,144 21.7 537.5 2.5
Fe(HCO3)+ CO3–2 + e- + 5H+ + Fe|3|O2– = FeHCO3+ + 2H2O@ 46.9 −267,988 −246,735 71.3 892.0 5.0

Fe(HSO4)+ S|6|O4–2 + e- + 5H+ + Fe|3|O2– = FeHSO4+ + 2H2O@ 37.7 −215,125 −208,704 21.5 975.6 4.2
Fe(HSO4)+2 S|6|O4–2 + 5H+ + Fe|3|O2– = Fe|3|HSO4+2 + 2H2O@ 26.1 −148,798 −200,163 −172.3 1078.5 2.5
Fe(SO4)@ S|6|O4–2 + e- + 4H+ + Fe|3|O2– = Fe(SO4)@ + 2H2O@ 36.9 −210,456 −212,106 −5.5 535.7 2.4
Fe(SO4)+ S|6|O4–2 + 4H+ + Fe|3|O2– = Fe|3|(SO4)+ + 2H2O@ 25.6 −146,355 −160,671 −48.0 505.8 2.0
Fe(SO4)2- 2S|6|O4–2 + 4H+ + Fe|3|O2– = Fe|3|(SO4)2- + 2H2O@ 27.0 −154,003 −162,930 −29.9 707.5 4.0
Fe+2 e- + 4H+ + Fe|3|O2– = Fe+2 + 2H2O@ 34.6 −197,613 −220,183 −75.7 338.8 1.3
Fe+3 4H+ + Fe|3|O2– = Fe|3|+3 + 2H2O@ 21.6 −123,295 −177,529 −181.9 308.9 −0.2
FeHSiO3+2 FeO2- + 3H+ + SiO2 = Fe|3|HSiO3+2 + H2O@ 21.5 −122,665 −148,472 −86.5 101.9 3.5
FeCl+ Cl- + e- + 4H+ + Fe|3|O2– = FeCl+ + 2H2O@ 34.8 −198,413 −218,887 −68.7 580.2 1.9
FeCl+2 Cl- + 4H+ + Fe|3|O2– = Fe|3|Cl+2 + 2H2O@ 23.1 −131,742 −173,502 −140.1 523.0 −0.5
FeCl2+ 2Cl- + 4H+ + Fe|3|O2– = Fe|3|Cl2+ + 2H2O@ 23.7 −135,453 −179,471 −147.6 931.3 1.1
FeCl3@ 3Cl- + 4H+ + Fe|3|O2– = Fe|3|Cl3@ + 2H2O@ 22.7 −129,649 −190,999 −205.8 1121.3 2.0

FeO+ 2H+ + Fe|3|O2– = Fe|3|O+ + H2O@ 15.9 −90,930 −97,152 −20.9 109.3 −2.4
FeO2H@ H+ + Fe|3|O2– = Fe|3|O2H@ 9.0 −51,601 −37,130 48.5 −77.2 0.7
FeOH+ e- + 3H+ + Fe|3|O2– = FeOH+ + H2O@ 25.1 −143,387 −167,718 −81.6 358.9 0.1
FeOH+2 3H+ + Fe|3|O2– = Fe|3|(OH)+2 + H2O@ 19.4 −110,794 −134,855 −80.7 276.6 −0.8
Fe2(OH)2+4 2FeO2- + 6H + = Fe|3|2(OH)2+4 + 2H2O@ 40.3 −229,747 −298,572 −230.8 617.9 7.5
Fe3(OH)4+5 3FeO2- + 8H + = Fe|3|3(OH)4+5 + 2H2O@ 58.5 −333,918 −472,753 −465.6 926.8 7.6
Table B.1 (continued)
Product Substance Reaction log10Ko298 ΔrGo298 ΔrHo298 ΔrSo298 ΔrCpo298 ΔrVo298

[J/mol] [J/mol] [J/K/mol] [J/K/mol] [J/bar]
H2@ 2e- + 2H+ = H|0|2@ −3.1 17,729 −4018 −72.9 138.0 2.5
H2S@ S|6|O4–2 + 8e- + 10H+ = H2S|-2|@ + 4H2O@ 40.7 −232,203 −272,852 −136.3 631.4 9.4
HCN@ CO3–2 + NO3- + 10e- + 13H+ = HC|-1|N|0|@ + 6H2O@ 117.4 −669,854 −729,336 −199.5 664.1 8.6
HCO3- CO3–2 + H+ = HCO3- 10.3 −58,959 −14,699 148.4 254.5 3.0
HS- S|6|O4–2 + 8e- + 9H+ = HS|-2|- + 4H2O@ 33.7 −192,305 −250,044 −193.7 358.3 8.0
HSiO3- H2O@ + SiO2@ = HSiO3- + H+ −9.8 55,992 29,057 −90.3 −207.0 −3.0
HSO3- S|6|O4–2 + 2e- + 3H+ = HS|4|O3– + H2O@ 3.8 −21,822 −3885 60.2 307.2 3.8
HSO4- S|6|O4–2 + H+ = HS|6|O4- 2.0 −11,346 20,464 106.7 288.8 2.2
K(SO4)- S|6|O4–2 + K+ = KSO4- 0.9 −4851 3070 26.6 212.6 0.6
KOH@ H2O@ + K+ = KOH@ + H+ −14.5 82,538 63,874 −62.6 −168.8 −1.2
Mg(CO3)@ CO3–2 + Mg+2 = Mg(CO3)@ 3.0 −17,009 9124 87.7 194.5 1.1
Mg(HCO3)+ CO3–2 + Mg + 2 + H + = Mg(HCO3)+ 11.4 −65,056 −12,725 175.5 565.4 3.7
Mg(HSiO3)+ Mg+2 + H2O@ + SiO2@ = Mg(HSiO3)+ + H+ −8.3 47,429 25,758 −72.7 60.5 −2.3
MgOH+ Mg+2 + H2O@ = Mg(OH)+ + H+ −11.4 65,300 61,792 −11.8 75.5 0.6
MgSiO3@ Mg+2 + H2O + SiO2 = MgSiO3@ + 2H+ −17.4 99,548 85,125 −48.4 −363.0 1.0
MgSO4@ S|6|O4–2 + Mg + 2 = Mg(SO4)@ 2.4 −13,529 6855 68.4 197.4 1.1
N2@ 2NO3- + 10e- + 12H + = N|0|2@ + 6H2O@ 207.3 −1.183,095 −1.311,876 −431.9 675.7 8.4
Na(CO3)- CO3–2 + Na+ = NaCO3- 1.3 −7251 −22,969 −52.7 199.9 0.7
Na(HCO3)@ CO3–2 + Na+ + H+ = NaHCO3@ 10.1 −57,533 −13,909 146.3 451.5 4.0
Na(SO4)- S|6|O4–2 + Na + = Na(SO4)- 0.7 −3996 3314 24.5 197.9 0.7

NaOH@ Na+ + H2O@ = NaOH@ + H+ −14.2 80,940 56,026 −83.6 −126.9 −1.3
NH3@ NO3- + 8e- + 9H + = N|-3|H3@ + 3H2O@ 109.9 −627,314 −732,284 −352.1 254.4 5.0
NH4+ NO3- + 8e- + 10H + = N|-3|H4+ + 3H2O@ 119.1 −680,039 −784,011 −348.7 244.6 4.4
O2@ 2H2O@ = O|0|2@ + 4e- + 4H+ −86.0 490,812 559,525 230.5 141.1 −0.6
OH- H2O@ = OH– + H+ −14.0 79,913 55,872 −80.6 −211.7 −2.3
S-2 S|6|O4–2 + 8e- + 8H + = S|-2|-2 + 4H2O@ 14.7 −83,851 −250,044 −557.4 358.3 5.9
S2O3–2 2S|6|O4–2 + 8e- + 10H + = S|2|2O3–2 + 5H2O@ 38.0 −216,987 −259,866 −143.8 555.2 9.2
SCN- CO3–2 + NO3- + S|6|O4–2 + 16e- + 20H + = S|0|C|0|N|-1|- + 10H2O@ 156.9 −895,711 −990,513 −318.0 1105.5 18.1
Si4O10–4 2H2O@ + 4SiO2@ = Si4O10–4 + 4H+ −36.3 207,202 207,202 0.0 0.0 −10.0
SiO3–2 H2O@ + SiO2@ = SiO3–2 + 2H+ −23.1 132,084 75,000 −191.5 0.0 −3.4
SO3–2 S|6|O4–2 + 2e- + 2H+ = S|4|O3–2 + H2O@ −3.4 19,390 −13,070 −108.9 31.6 0.1
Sr(CO3)@ CO3–2 + Sr+2 = Sr(CO3)@ 2.8 −16,013 18,891 117.1 196.6 0.9

Sr(HCO3)+ CO3–2 + Sr+2 + H + = SrHCO3+ 11.5 −65,721 −12,816 177.4 541.0 3.8
Sr(SO4)@ S|6|O4–2 + Sr+2 = Sr(SO4)@ 2.3 −13,072 9071 74.3 197.1 1.0
SrOH+ Sr+2 + H2O@ = Sr(OH)+ + H+ −13.3 75,860 82,619 22.7 −65.5 0.7
SrSiO3@ Sr+2 + H2O@ + SiO2@ = SrSiO3@ + 2H+ −18.8 107,136 107,184 0.2 0.1 0.0
e- - represents one electron.

|| - is used for specifying a different valence for the element.
@ - is used to represent a neutral aqueous species.
Table B.2
Product solid and gaseous species reactions from master species, together with their reaction properties at 25 °C and 1 bar.
log A G A H A A Cp
Product Substance Reaction 10^° r 298 r 298 r$ 298 r 298 Ar^°
298 [J/mol] [J/mol] [J/K/ [J/K/ 298
mol] mol] [J/
bar]
5CA (CaO)1.25(SiO2)1(Al2O3)0.125(H2O)1.625 + 2.25H+ = 0.25AlO2- + 1.25Ca+2 + 2.75H2O@ + SiO2@ 15.9 -90709 -92987 -7.6 23.7 -1.2
5CNA (CaO)1.25(SiO2)1(Al2O3)0.125(Na2O)0.25(H2O)1.375 + 2.75H+ = 0.25AlO2- + 1.25Ca+2 + 0.5Na+ + 23.2 -132663 -135750 -10.4 43.6 -2.0
2.75H2O@ + SiO2@
AlOHam Al(OH)3 = AlO2- + H+ + H2O@ -13.8 78536 69482 -30.4 -66.8 -0.4
AlOHmic Al(OH)3 = AlO2- + H+ + H2O@ -14.7 83730 53830 -100.3 -66.8 -0.4
Amor-Sl SiO2 = SiO2@ -2.7 15492 15492 0.0 0.0 -1.3
Anh CaSO4 = Ca+2 + S|6|O4-2 -4.4 24872 -18165 -144.3 -396.7 -5.1
Arg CaCO3 = CO3-2 + Ca+2 -8.3 47583 -11060 -196.7 -401.5 -5.9
Brc Mg(OH)2 + 2H+ = Mg+2 + 2H2O@ 16.8 -96124 -114419 -61.4 51.8 -1.1
C12A7 (CaO)12(Al2O3)7 + 10H+ = 14AlO2- + 12Ca+2 + 5H2O@ 167.2 -954361 -1489798 -1795.9 -1765.7 -51.6
C2AClH5 Ca2AlCl(OH)6(H2O)2 + 2H+ = AlO2- + 2Ca+2 + Cl- + 6H2O@ 14.4 -82445 -92106 -32.4 -195.7 -3.8
C2AH65 Ca2Al(OH)7(H2O)3 + 3H+ = AlO2- + 2Ca+2 + 8H2O@ 29.4 -167699 -167567 0.4 -79.0 -2.0
C2AH7.5 Ca2Al2(OH)10(H2O)2.5 + 2H+ = 2AlO2- + 2Ca+2 + 8.5H2O@ 14.2 -81085 -89743 -29.0 -55.2 -4.4
C2S (CaO)2SiO2 + 4H+ = 2Ca+2 + 2H2O@ + SiO2@ 38.5 -219567 -237276 -59.4 4.7 -3.6
C3A (CaO)3Al2O3 + 4H+ = 2AlO2- + 3Ca+2 + 2H2O@ 70.7 -403436 -491527 -295.5 -249.5 -8.9
C3AFS0.84H4.32 (AlFe|3|O3)[Ca3O3(SiO2)0.84(H2O)4.32] + 4H+ = AlO2- + 3Ca+2 + 6.32H2O@ + 0.84SiO2@ + 22.3 -127203 -215980 -297.8 -286.9 -6.3
Fe|3|O2-

C3AH6 Ca3Al2O6(H2O)6 + 4H+ = 2AlO2- + 3Ca+2 + 8H2O@ 35.5 -202666 -230152 -92.2 -33.6 -4.2
C3AS0.41H5.18 Ca3Al2O6(SiO2)0.41(H2O)5.18 + 4H+ = 2AlO2- + 3Ca+2 + 7.18H2O@ + 0.41SiO2@ 28.9 -165175 -197973 -110.0 -61.6 -4.6
C3AS0.84H4.32 Ca3Al2O6(SiO2)0.84(H2O)4.32 + 4H+ = 2AlO2- + 3Ca+2 + 6.32H2O@ + 0.84SiO2@ 25.8 -147182 -185475 -128.4 -90.9 -5.1
C3FH6 Ca3Fe 13| 2O6(H2O)6 + 4H+ = 3Ca+2 + 8H2O@ + 2Fe|3|O2- 29.7 -169558 -286244 -391.4 -426.6 -6.5
C3FS0.84H4.32 (Fe|3|Fe|3|O3)[Ca3O3(SiO2)0.84(H2O)4.32] + 4H+ = 3Ca+2 + 6.32H2O@ + 0.84SiO2@ + 2Fe|3|O2- 20.0 -114074 -246486 -444.1 -482.9 -7.5
C3FS1.34H3.32 Ca3Fe|3|2O6(SiO2)1.34(H2O)3.32 + 4H+ = 3Ca+2 + 5.32H2O@ + 1.34SiO2@ + 2Fe|3|O2- 16.2 -92409 -233544 -473.4 -516.4 -8.5
C3S (CaO)3SiO2 + 6H+ = 3Ca+2 + 3H2O@ + SiO2@ 73.3 -418180 -444107 -87.0 6.2 -5.8
C4AClH10 Ca4Al2Cl2(OH)12(H2O)4 + 4H+ = 2AlO2- + 4Ca+2 + 2Cl- + 12H2O@ 28.9 -164890 -184212 -64.8 -391.5 -7.6
C4AF (CaO)4(Al2O3)(Fe|3|2O3) + 4H+ = 2AlO2- + 4Ca+2 + 2H2O@ + 2Fe|3|O2- 50.5 -288060 -402597 -384.2 -937.0 -14.8
C4AH11 Ca4Al2(OH)14(H2O)4 + 6H+ = 2AlO2- + 4Ca+2 + 14H2O@ 60.5 -345302 -369182 -80.1 -228.6 -5.9
C4AH13 Ca4Al2(OH)14(H2O)6 + 6H+ = 2AlO2- + 4Ca+2 + 16H2O@ 58.8 -335407 -335144 0.9 -158.0 -4.0
C4AH19 Ca4Al2(OH)14(H2O)12 + 6H+ = 2AlO2- + 4Ca+2 + 22H2O@ 58.6 -334271 -294930 132.0 53.7 -2.6

C4AsClH12 Ca4Al2Cl(SO4)0.5(OH)12(H2O)6 + 4H+ = 2AlO2- + 4Ca+2 + Cl- + 0.5S|6|O4-2 + 14H2O@ 27.6 -157430 -175702 -61.3 -351.3 -6.7
C4FH13 Ca4Fe|3|2(OH)14(H2O)6 + 6H+ = 4Ca+2 + 16H2O@ + 2Fe|3|O2- 53.3 -304002 -199193 351.5 -343.7 -7.0
CA CaOAl2O3 = 2AlO2- + Ca+2 -0.3 1756 -67154 -231.1 -254.7 -5.3
CA2 CaO(Al2O3)2 + H2O@ = 4AlO2- + Ca+2 + 2H+ -30.1 171600 44867 -425.1 -502.0 -8.8
CAH10 CaOAl2O3(H2O)10 = 2AlO2- + Ca+2 + 10H2O@ -7.6 43348 35098 -27.7 -43.2 -1.3
Cal CaCO3 = CO3-2 + Ca+2 -8.5 48404 -10975 -199.2 -402.1 -6.1
CH4 C|-4|H4 + 3H2O@ = CO3-2 + 10H+ + 8e- -41.0 234277 257142 76.7 -435.8 -
Cls SrSO4 = S|6|O4-2 + Sr+2 -6.6 37855 -854 -129.8 -415.5 -5.1
log A G A H A A Cp
298 [J/mol] [J/mol] [J/K/ [J/K/ 298
mol] mol] [J/
bar]
CO2 CO2 + H2O@ = CO3-2 + 2H+ -18.1 103560 4079 -333.7 -401.8 -
CSH3T-T2C ((CaO)0.75(SiO2)0.5(H2O)1.25)2 + 3H+ = 1.5Ca+2 + 4H2O@ + SiO2@ 25.3 -144257 -123574 69.4 62.5 -2.0
CSH3T-T5C ((CaO)1(SiO2)1(H2O)2)1.25 + 2.5H+ = 1.25Ca+2 + 3.75H2O@ + 1.25SiO2@ 18.1 -103538 -78676 83.4 65.4 -1.4
CSH3T-TobH (CaO)1(SiO2)1.5(H2O)2.5 + 2H+ = Ca+2 + 3.5H2O@ + 1.5SiO2@ 12.5 -71524 -42463 97.5 68.4 -1.6
CSHQ-JenD (CaO)1.5(SiO2)0.6667(H2O)2.5 + 3H+ = 1.5Ca+2 + 4H2O@ + 0.6667SiO2@ 28.7 -164004 -149357 49.1 51.9 -2.6
CSHQ-JenH (CaO)1.3333(SiO2)1(H2O)2.1667 + 2.6666H+ = 1.3333Ca+2 + 3.5H2O@ + SiO2@ 22.2 -126609 -106258 68.3 59.1 -2.1
CSHQ-TobD ((CaO)1.25(SiO2)1(H2O)2.75)0.6667 + 1.66675H+ = 0.833375Ca+2 + 2.6668H2O@ + 0.6667SiO2@ 13.7 -77952 -64488 45.2 37.9 -0.4
CSHQ-TobH (CaO)0.6667(SiO2)1(H2O)1.5 + 1.3334H+ = 0.6667Ca+2 + 2.1667H2O@ + SiO2@ 8.3 -47306 -27842 65.3 45.5 -1.2
Dis-Dol CaMg(CO3)2 = 2CO3-2 + Ca+2 + Mg+2 -16.5 94412 -43108 -461.2 -789.0 -11.7
ECSH1-KSH ((KOH)2.5SiO2H2O)0.2 + 0.5H+ = 0.7H2O@ + 0.2SiO2@ + 0.5K+ 5.5 -31394 -13703 59.3 25.2 0.8
ECSH1-NaSH ((NaOH)2.5SiO2H2O)0.2 + 0.5H+ = 0.5Na+ + 0.7H2O@ + 0.2SiO2@ 5.4 -30883 -17398 45.2 42.8 0.5
ECSH1-SH (SiO2H2O)1 = H2O@ + SiO2@ -2.6 14839 14839 0.0 0.0 0.0
ECSH1-SrSH ((Sr(OH)2)1SiO2H2O)1 + 2H+ = Sr+2 + 3H2O@ + SiO2@ 15.4 -87911 -64744 77.7 54.2 -1.1
ECSH1-TobCa ((Ca(OH)2)0.8333SiO2H2O)1 + 1.6666H+ = 0.8333Ca+2 + 2.6666H2O@ + SiO2@ 11.0 -62909 -43194 66.1 49.3 -1.9
ECSH2-JenCa ((Ca(OH)2)1.6667SiO2H2O)0.6 + 2.00004H+ = 1.00002Ca+2 + 2.60004H2O@ + 0.6SiO2@ 17.6 -100489 -77495 77.1 77.2 0.2
ECSH2-KSH ((KOH)2.5SiO2H2O)0.2 + 0.5H+ = 0.7H2O@ + 0.2SiO2@ + 0.5K+ 6.0 -34250 -16559 59.3 25.2 0.8
ECSH2-NaSH ((NaOH)2.5SiO2H2O)0.2 + 0.5H+ = 0.5Na+ + 0.7H2O@ + 0.2SiO2@ 5.9 -33732 -20247 45.2 42.8 0.5
ECSH2-SrSH ((Sr(OH)2)1SiO2H2O)1 + 2H+ = Sr+2 + 3H2O@ + SiO2@ 16.2 -92473 -69306 77.7 54.2 -1.1

ECSH2-TobCa ((Ca(OH)2)0.8333SiO2H2O)1 + 1.6666H+ = 0.8333Ca+2 + 2.6666H2O@ + SiO2@ 11.0 -62909 -43194 66.1 49.3 -1.9
ettringite ((H2O)2)Ca6Al2(SO4)3(OH)12(H2O)24 + 4H+ = 2AlO2- + 6Ca+2 + 3S|6|O4-2 + 34H2O@ 11.2 -63708 -23594 134.5 -694.0 -14.6
ettringite03_ss (SO4)Ca2Al0.6666667(OH)4(H2O)8.6666667 + 1.3333332H+ = 0.6666667AlO2- + 2Ca+2 + S|6|O4-2 + 3.7 -21231 -7860 44.8 -231.3 -4.9
11. 3333333H2O@
ettringite05 Ca3Al(SO4)1.5(OH)6(H2O)13 + 2H+ = AlO2- + 3Ca+2 + 1.5S|6|O4-2 + 17H2O@ 5.6 -31852 -11795 67.3 -347.0 -7.3
ettringite13 Ca6Al2(SO4)3(OH)12(H2O)7 + 4H+ = 2AlO2- + 6Ca+2 + 3S|6|O4-2 + 15H2O@ 39.0 -222527 -596572 -1254.6 -1364.4 -19.3
Ettringite13_des Ca6Al2(SO4)3(OH)12(H2O)7 + 4H+ = 2AlO2- + 6Ca+2 + 3S|6|O4-2 + 15H2O@ 39.0 -222527 -596572 -1254.6 -1364.4 -19.3
ettringite30 Ca6Al2(SO4)3(OH)12(H2O)24 + 4H+ = 2AlO2- + 6Ca+2 + 3S|6|O4-2 + 32H2O@ 11.8 -67136 -36633 102.3 -764.6 -18.3
ettringite9 Ca6Al2(SO4)3(OH)12(H2O)3 + 4H+ = 2AlO2- + 6Ca+2 + 3S|6|O4-2 + 11H2O@ 48.0 -273943 -339631 -220.3 -1505.6 -21.5
Ettringite9_des Ca6Al2(SO4)3(OH)12(H2O)3 + 4H+ = 2AlO2- + 6Ca+2 + 3S|6|O4-2 + 11H2O@ 48.0 -273943 -339631 -220.3 -1505.6 -21.5
Fe Fe|0| + 2H2O@ = 4H+ + 3e- + Fe|3|O2- -18.6 106109 127947 73.2 -367.4 -4.3
Fe-ettringite Ca6Fe|3|2(SO4)3(OH)12(H2O)26 + 4H+ = 6Ca+2 + 3S|6|O4-2 + 34H2O@ + 2Fe|3|O2- 12.1 -68843 4803 247.0 -1091.4 -17.4

Fe-ettringite05 Ca3Fe|3|(SO4)1.5(OH)6(H2O)13 + 2H+ =3Ca+2 + 1.5S|6|O4-2 + 17H2O@ + Fe|3|O2- 6.0 -34420 2403 123.5 -545.7 -8.7
Fe-hemicarbonate Ca3O3Fe|3|2O3(CaCO3)0.5(CaO2H2)0.5(H2O)9.5 + 5H+ = 0.5CO3-2 + 4Ca+2 + 12.5H2O@ + 39.2 -223627 -390054 -558.2 -637.4 -12.3
2Fe|3|O2-
Femonocarbonate Ca4O4Fe|3|2O3CO2(H2O)12 + 4H+ = CO3-2 + 4Ca+2 + 14H2O@ + 2Fe|3|O2- 21.4 -122258 -252941 -438.3 -778.2 -11.8
Fe-monosulph05 Ca2Fe|3|S0.5O5(H2O)6 + 2H+ = 2Ca+2 + 0.5S|6|O4-2 + 7H2O@ + Fe|3|O2- 12.2 -69792 -154688 -284.7 -386.3 -6.4
Fe-monosulphate Ca4Fe|3|2SO10(H2O)12 + 4H+ = 4Ca+2 + S|6|O4-2 + 14H2O@ + 2Fe|3|O2- 24.5 -139576 -309368 -569.5 -772.7 -12.8
FeOOHmic Fe|3|OOH = H+ + Fe|3|O2- -19.6 111875 65468 -155.7 -309.3 -3.4
Gbs Al(OH)3 = AlO2- + H+ + H2O@ -15.1 86323 77269 -30.4 -66.8 -0.4
log A G A H A A Cp
298 [J/mol] [J/mol] [J/K/ [J/K/ 298
mol] mol] [J/
bar]
Gp CaSO4(H2O)2 = Ca+2 + S|6|O4-2 + 2H2O@ -4.6 26147 -1167 -91.6 -332.5 -4.4
Gr C|0| + 3H2O@ = CO3-2 + 6H+ + 4e- -32.2 183569 182332 -4.1 -466.4 -6.6
Gt Fe|3|O(OH) = H+ + Fe|3|O2- -22.6 128995 124419 -15.4 -309.2 -2.0
H2 H|0|2 = 2H+ + 2e- 0.0 0 0 0.0 0.0 -
H2O H2O = H2O@ 1.5 -8500 -43481 -117.3 35.3 -
H2S H2S|-2| + 4H2O@ = S|6|O4-2 + 10H+ + 8e- -41.7 237742 254458 56.1 -486.4 -
Hem Fe|3|2O3 + H2O@ = 2H+ + 2Fe|3|O2- -42.1 240195 219672 -68.8 -650.0 -4.7
hemicarbonat10.5 (CaO)3Al2O3(CaCO3)0.5(CaO2H2)0.5(H2O)10 + 5H+ = 2AlO2- + 0.5CO3-2 + 4Ca+2 + 13H2O@ 42.6 -243220 -264276 -70.6 -232.4 -8.4
hemicarbonate (CaO)3Al2O3(CaCO3)0.5(CaO2H2)0.5(H2O)11.5 + 5H+ = 2AlO2- + 0.5CO3-2 + 4Ca+2 + 14.5H2O@ 40.9 -233337 -236348 -10.1 -179.5 -8.0
hemicarbonate9 (CaO)3Al2O3(CaCO3)0.5(CaO2H2)0.5(H2O)8.5 + 5H+ = 2AlO2- + 0.5CO3-2 + 4Ca+2 + 11.5H2O@ 45.6 -260338 -299005 -129.7 -285.4 -9.9
hemihydrate CaSO4(H2O)0.5 = Ca+2 + S|6|O4-2 + 0.5H2O@ -3.6 20413 -20432 -137.0 -383.4 -5.8
hydrotalcite Mg4Al2O7(H2O)10 + 6H+ = 2AlO2- + 4Mg+2 + 13H2O@ 28.0 -159755 -235072 -252.6 146.4 -5.4
INFCA (CaO)1(SiO2)1.1875(Al2O3)0.15625(H2O)1.65625 + 1.6875H+ = 0.3125AlO2- + Ca+2 + 2.5H2O@ + 9.0 -51116 -50080 3.5 14.1 -1.1
1.1875SiO2@
INFCN (CaO)1(SiO2)1.5(Na2O)0.3125(H2O)1.1875 + 2.625H+ = Ca+2 + 0.625Na+ + 2.5H2O@ + 1.5SiO2@ 18.8 -107089 -97770 31.3 64.3 -2.1
INFCNA (CaO)1.25(SiO2)1(Al2O3)0.125(Na2O)0.25(H2O)1.375 + 2.75H+ = 0.25AlO2- + 1.25Ca+2 + 0.5Na+ + 23.2 -132663 -135750 -10.4 43.6 -2.0
2.75H2O@ + SiO2@
Jennite (SiO2)1(CaO)1.666667(H2O)2.1 + 3.333334H+ = 1.666667Ca+2 + 3.766667H2O@ + SiO2@ 29.3 -167347 -146306 70.6 66.0 -2.5

K2O K2O + 2H+ = H2O@ + 2K+ 84.1 -480020 -426988 177.9 8.0 -0.4
K2SO4 K2SO4 = S|6|O4-2 + 2K+ -1.8 10214 23725 45.3 -379.3 -3.5
Kln Al2Si2O5(OH)4 = 2AlO2- + 2H+ + H2O@ + 2SiO2@ -38.3 218758 185703 -110.9 -173.4 -3.0
*KSiOH ((KOH)2.5SiO2H2O)0.2 + 0.5H+ = 0.7H2O@ + 0.2SiO2@ + 0.5K+ 5.8 -32604 -13463 65.3 20.7 0.8
Lim CaO + 2H+ = Ca+2 + H2O@ 32.6 -186017 -193861 -26.3 1.6 -1.7
M4A-OH-LDH Mg4Al2(OH)14(H2O)3 + 6H+ = 2AlO2- + 4Mg+2 + 13H2O@ 34.3 -195826 -271137 -252.6 147.4 -5.3
Mag FeFe|3|2O4 + 2H2O@ = 4H+ + e- + 3Fe|3|O2- -67.8 387007 361013 -87.2 -992.8 -7.9
Melanterite FeSO4(H2O)7 = S|6|O4-2 + 4H+ + e- + 5H2O@ + Fe|3|O2- -36.8 210222 230774 68.9 -109.8 -4.3
Mg2AlC0.5OH Mg2Al(OH)6(CO3)0.5(H2O)2 + 2H+ = AlO2- + 0.5CO3-2 + 2Mg+2 + 6H2O@ 5.9 -33755 -100890 -225.2 -182.5 -4.0
Mg2FeC0.5OH Mg2Fe|3|(OH)6(CO3)0.5(H2O)2 + 2H+ = 0.5CO3-2 + 2Mg+2 + 6H2O@ + Fe|3|O2- 5.8 -33224 -81424 -161.7 -377.1 -5.6

Mg3AlC0.5OH Mg3Al(OH)8(CO3)0.5(H2O)2.5 + 4H+ = AlO2- + 0.5CO3-2 + 3Mg+2 + 8.5H2O@ 22.7 -129680 -215112 -286.5 -130.7 -2.1
Mg3FeC0.5OH Mg3Fe|3|(OH)8(CO3)0.5(H2O)2.5 + 4H+ = 0.5CO3-2 + 3Mg+2 + 8.5H2O@ + Fe|3|O2- 22.4 -127660 -194156 -223.0 -325.3 -3.4
Mgs MgCO3 = CO3-2 + Mg+2 -8.3 47310 -28349 -253.8 -386.8 -5.6
monocarbonate Ca4Al2CO9(H2O)11 + 4H+ = 2AlO2- + CO3-2 + 4Ca+2 + 13H2O@ 24.5 -140067 -165182 -84.2 -412.8 -8.8
monocarbonate05 Ca2AlC0.5O4.5(H2O)5.5 + 2H+ = AlO2- + 0.5CO3-2 + 2Ca+2 + 6.5H2O@ 12.3 -70033 -82591 -42.1 -206.4 -4.4
monocarbonate9 Ca4Al2CO9(H2O)9 + 4H+ = 2AlO2- + CO3-2 + 4Ca+2 + 11H2O@ 28.5 -162891 -224860 -207.8 -483.4 -9.6
mononitrate Ca4Al2(OH)12N|5|2O6(H2O)4 + 4H+ = 2AlO2- + 4Ca+2 + 2NO3- + 12H2O@ 27.3 -156023 -148389 25.6 -356.1 -7.7
mononitrite Ca4Al2(OH)12N|3|2O4(H2O)4 = 2AlO2- + 4Ca+2 + 2NO3- + 4e- + 10H2O@ -25.7 146749 197173 169.1 -420.1 -9.2
log A G A H A A Cp
298 [J/mol] [J/mol] [J/K/ [J/K/ 298
mol] mol] [J/
bar]
monosulphate10.5 Ca4Al2SO10(H2O)10.5 + 4H+ = 2AlO2- + 4Ca+2 + S|6|O4-2 + 12.5H2O@ 28.1 -160590 -194728 -114.5 -434.2 -9.8
monosulphate12 Ca4Al2SO10(H2O)12 + 4H+ = 2AlO2- + 4Ca+2 + S|6|O4-2 + 14H2O@ 26.8 -152912 -176816 -80.2 -381.2 -7.1
monosulphate1205 Ca2AlS0.5O5(H2O)6 + 2H+ = AlO2- + 2Ca+2 + 0.5S|6|O4-2 + 7H2O@ 13.4 -76454 -88406 -40.1 -190.6 -6.0
monosulphate16 Ca4Al2SO10(H2O)16 + 4H+ = 2AlO2- + 4Ca+2 + S|6|O4-2 + 18H2O@ 26.9 -153259 -148473 16.1 -246.2 -6.7
N2 N|0|2 + 6H2O@ = 2NO3- + 12H+ + 10e- -210.5 1201289 1301508 336.1 -470.7 -
Na2O Na2O + 2H+ = 2Na+ + H2O@ 67.4 -384943 -351639 111.7 82.7 -0.9
Na2SO4 Na2SO4 = S|6|O4-2 + 2Na+ -0.3 1703 -2457 -14.0 -254.9 -4.3
*NaSiOH ((NaOH)2.5SiO2H2O)0.2 + 0.5H+ = 0.5Na+ + 0.7H2O@ + 0.2SiO2@ 5.7 -32021 -19201 43.7 34.7 0.5
O2 O|0|2 + 4H+ + 4e- = 2H2O@ 83.1 -474371 -571762 -326.7 63.7 -
Ord-Dol CaMg(CO3)2 = 2CO3-2 + Ca+2 + Mg+2 -17.1 97552 -36538 -449.7 -789.0 -11.7
Portlandite Ca(OH)2 + 2H+ = Ca+2 + 2H2O@ 22.8 -130145 -130156 0.0 32.3 -1.5
Py FeS|0|S|-2| + 10H2O@ = 2S|6|O4-2 + 20H+ + 15e- + Fe|3|O2- -120.5 687822 780233 309.9 -1366.5 -17.8
Qtz SiO2 = SiO2@ -3.7 21386 21386 0.0 0.0 -0.7
Sd FeCO3 + 2H2O@ = CO3-2 + 4H+ + e- + Fe|3|O2- -45.5 259775 204566 -185.2 -742.7 -7.1
Str SrCO3 = CO3-2 + Sr+2 -9.3 52918 -324 -178.6 -413.3 -6.3
straetlingite Ca2Al2SiO7(H2O)8 + 2H+ = 2AlO2- + 2Ca+2 + 9H2O@ + SiO2@ 4.1 -23479 -38065 -48.9 -39.9 -5.5

straetlingite5.5 Ca2Al2SiO7(H2O)5.5 + 2H+ = 2AlO2- + 2Ca+2 + 6.5H2O@ + SiO2@ 7.1 -40503 -79963 -132.3 -128.1 -9.7
straetlingite7 Ca2Al2SiO7(H2O)7 + 2H+ = 2AlO2- + 2Ca+2 + 8H2O@ + SiO2@ 4.8 -27477 -45434 -60.2 -75.2 -7.3
Sulfur S|0| + 4H2O@ = S|6|O4-2 + 8H+ + 6e- -35.8 204198 233827 99.4 -503.8 -7.5
syngenite K2Ca(SO4)2H2O = Ca+2 + 2S|6|O4-2 + H2O@ + 2K+ -7.2 41078 19412 -72.7 -743.9 -8.4
T2C-CNASHss (CaO)1.5(SiO2)1(H2O)2.5 + 3H+ = 1.5Ca+2 + 4H2O@ + SiO2@ 25.6 -145938 -125275 69.3 62.5 -2.0
T5C-CNASHss (CaO)1.25(SiO2)1.25(H2O)2.5 + 2.5H+ = 1.25Ca+2 + 3.75H2O@ + 1.25SiO2@ 18.4 -105297 -80438 83.4 65.4 -1.5
thaumasite (CaSiO3)(CaSO4)(CaCO3)(H2O)15 + 2H+ = CO3-2 + 3Ca+2 + S|6|O4-2 + 16H2O@ + SiO2@ -0.9 5236 23833 62.4 -468.7 -7.3
TobH-CNASHss (CaO)1(SiO2)1.5(H2O)2.5 + 2H+ = Ca+2 + 3.5H2O@ + 1.5SiO2@ 12.8 -73056 -44000 97.5 68.3 -1.6
Tob-I (SiO2)2.4(CaO)2(H2O)3.2 + 4H+ = 2Ca+2 + 5.2H2O@ + 2.4SiO2@ 26.7 -152677 -105616 157.8 119.1 -4.5
Tob-II (SiO2)1(CaO)0.833333(H2O)1.333333 + 1.666666H+ = 0.833333Ca+2 + 2.166666H2O@ + SiO2@ 11.1 -63617 -44009 65.8 49.6 -1.9
tricarboalu03 (CO3)Ca2Al0.6666667(OH)4(H2O)8.6666667 + 1.3333332H+ = 0.6666667AlO2- + CO3-2 + 2Ca+2 + 3.2 -18115 -21090 -10.0 -236.8 -4.9
11.3333333H2O@

Tro Fe|2|S|-2| + 6H2O@ = S|6|O4-2 + 12H+ + 9e- + Fe|3|O2- -73.6 420228 471331 171.4 -873.9 -11.3
* - reaction properties at 20 °C 1 bar.

e- - represents one electron.
|| - is used for specifying a different valence for the element.
@ - is used to represent a neutral aqueous species.
Standard molar volumes for ideal gases are not listed because they are all equal to 2479 J/bar at 1 bar, 298.15 K.
Th.+ Calc. Calc+ Th.+ Calc. Calc.+
Thaumasite + Calcite Thaumasite + Calcite + C-S-H C-S-H
+ C-S-H C-S-H
16 10
14
Concentration [mM]
Concentration [mM]
12 1
10 silicon
calcium
8 0.1
6
4 sulfate 0.01
carbonate
2

0 0.001
0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80
Temperature [°C] Temperature [°C]
Fig. B.1. Calculated (curves) solubility data for thaumasite, based on the new thermodynamic data for thaumasite complemented with the CSHQ data from Cemdata18 [1,7] in GEM format; Calculated (dotted lines)
solubility data for thaumasite, based on data Cemdata18 [1,7] in PHREEQC format.

Fig. B.2. Effect of the amount of limestone on the phase assemblage and the distribution of aluminium and iron in hydrated Portland cement calculated using
Cemdata18 GEM format (dashed lines) and Cemdata18 PHREEQC format (dotted lines), in both cases using ideal solid solutions.



Fig. B.3. Calculated aqueous composition in equilibrium with CO3-SO4-ettringite solid solution as a function of SO4 in the solid. Solid lines calculated using the Cemdata18 GEM format using non ideal solid solution;
Dashed lines calculated using the Cemdata18 GEM format using ideal solid solution; Dotted lines calculated using the Cemdata18 PHREEQC format using ideal solid solution; Circles: experimental data [7,129].
Appendix C. Thermodynamic equations and assumptions
The solubility products compiled in Cemdata18 have generally been derived from solutions composition measured at different temperatures, as
documented in detail in [1,7–10,12,27,28,30,31,34–37,39–41]. The activity of a species i, ai, has been calculated with GEMS from the measured
concentrations considering the formation of aqueous complexes. By definition ai = γi*mi, where γi is the activity coefficient and mi the concentration
in mol/kg H2O. Activity coefficients of aqueous species γi were computed using the built-in extended Debye-Hückel equation with the common ion-
size parameter ai of 3.67 Å for KOH and 3.31 Å for NaOH solutions and the common third parameter by according to the Eq. (C.1):
log i = + by I
Ay z i2 I
1 + B y ai I (C.1)
where zi denotes the charge of species i, I is the effective molal ionic strength, by is a semi-empirical parameter (~0.123 for KOH and ~0.098 for
NaOH electrolyte at 25 °C), and Ay and By are P,T-dependent coefficients. For uncharged species, Eq. (C.1) reduces to logγi = byI. This extended
Debye-Hückel activity correction is applicable up to approx. 1 m ionic strength [130].
From the solubility products K of solids calculated at different temperatures T, the Gibbs free energy of reaction, ΔrG°, the Gibbs free energy of
formation, ΔfG°, and the absolute entropy, S°, at T0 = 298.15 K were obtained according to Eqs. (C.2) and (C.3):
r G° = =
° °
i fG RT ln K
i (C.2)
= T0) +
T T
o o
Cpo
a GT f GT0 STo0 (T Cp0 dT dT
T0 T0
T (C.3)
Using C°p = a0 + a1T + a2T−2 + a3T-0.5 [131], where a0–3 are the empirical parameters defined for each mineral, the two integral terms of Eq.
(C.3) can be solved to give Eq. (C.4):
= T + T0
o o T (T T0 ) 2 2( T T0 ) 2
a GT f GT0 STo0 (T T0) a 0 T ln 0.5a1 (T T0 )2 a2 a3
T0 2T T0 2 T0 (C.4)
where νi are the stoichiometric reaction coefficients, R = 8.31451 J/mol/K, T is the temperature in K, and C°p is the heat capacity at constant
pressure. The apparent Gibbs free energy of formation, ΔaG°T, refers to standard Gibbs energies of elements at 298.15 K. A more detailed description
of the derivation of the dependence of the Gibbs free energy on temperature is available in [131,132].
Dependence of the solubility product on temperature, consistent to Eq. (C.4) can be expressed as:
log KT = A0 + A1 T + + A3 ln T + 24 + A5 T 2 + A6 T
A2 A
T T (C.5)
[131], where A0, … A6 are empirical coefficients. If the entropy (S°), the enthalpy (ΔfH°), and the coefficients (a0, a1, …) of the heat capacity
equation (C°p = a0 + a1T + a2T−2 + a3T-0.5 + a4T2) of the species are available, the coefficients A0, … A6 can be calculated directly (see [131]).
These calculations involving Eqs. (C.4) and (C.5) are all implemented in the GEM-Selektor.
The heat capacity function, Cp = f(T) is usually obtained from calorimetry experiments. In many cases, the heat capacity has to be estimated by using
a reference reaction with a solid having a known heat capacity and similar structure, as described in publications [1,7–10,12,27,28,30,31,34–37,39–41].
Helgeson et al. [43] applied this principle successfully to estimate heat capacities of silicate minerals by formulating reactions involving structurally-
related minerals with known heat capacity functions. This method has limitations due to the differing thermodynamic properties of “water” varieties,
bound loosely as a hydration water, or structurally as OH-groups. To minimize errors associated with the varying strengths of bonding for “water”,
reference reactions had been formulated to involve no “free” water as a substituent in reactions, wherever appropriate.
The value of ΔrCp0 has little influence on the calculated log K value in the temperature range 0–100 °C and is thus often assumed to be constant in
a narrow temperature range: ΔrCp0T = ΔrCp0T0 = Δao. This simplifies Eq. (C.5) to the so called 3-term approximation of the temperature dependence,
see Eq. (C.6), which can be used to compute the standard thermodynamic properties of each solid [132] to obtain a temperature-dependent “log K"
function using Eqs. (C.6)-(C.12) (implemented in GEMS).
log KT = A0 + A2 T 1 + A3 ln T (C.6)
and
A0 = + ln T0 )]
0.4343 0 0
[ r ST0 r CpT 0 (1
R (C.7)
A2 =
0.4343 0 0
( r HT0 r CpT 0 T0 )
R (C.8)
A3 =
0.4343 0
r CpT 0
R (C.9)
= +
0 0 0 T
r ST r ST0 r CpT 0 ln
T0 (C.10)
0
r HT = 0
r HT 0 + 0
r CpT 0 (T T0) (C.11)
0
r GT = 0
r HT +T 0
r ST (C.12)
Within the relatively narrow temperature range of 0 to 100°C, where the Cemdata18 database is valid, this simplification has a negligible
influence on the resulting solubility products, also for non-isoelectric reactions as exemplified for ettringite in [20].

Appendix D. Thermodynamic data for aqueous and gaseous species
The thermodynamic data for aqueous and gaseous species compatible with Cemdata18 are summarised in Tables D.1 and D.2.
Table D.1
Standard (partial molal) thermodynamic properties and equation of state parameters of aqueous species at 25 °C, 1 bar used in GEM calculations, as detailed in the GEMS version of the PSI/Nagra 12/
07 TDB [22,23]. Numbers referring to the charge of aqueous species are written after the plus or minus signs to avoid any ambiguity; “@” is used to represent a neutral aqueous species.
Species ΔG0 ΔH0 S0 Cp0 V0 a1·10⁎ a2·10–2⁎ a3⁎ a4·10−4⁎ c1⁎ c2·10−4⁎ ω0·10−5⁎
(kJ/mol) (kJ/mol) (J/ (J/mol∙K) (J/bar) (cal/mol/ (cal/mol) (cal∙K/mol/ (cal∙K/ (cal/mol/ (cal∙K/ (cal/mol)
mol∙K) bar) bar) mol) K) mol)
Al(SO4)+ −1250.43 −1422.67 −172.38 −204.01 −6.02 1.3869 −4.3920 7.4693 −2.5974 −11.6742 −12.9914 1.1729
Al(SO4)2− −2006.30 −2338.40 −135.50 −268.37 31.11 6.8275 8.8925 2.2479 −3.1466 −12.0220 −16.1447 2.1199
Al+3 −483.71 −530.63 −325.10 −128.70 −45.24 −3.3802 −17.0071 14.5185 −2.0758 10.7000 −8.0600 2.7530
AlO+ −660.42 −713.64 −112.97 −125.11 0.31 2.1705 −2.4811 6.7241 −2.6763 −2.5983 −9.1455 0.9570
AlO2− −827.48 −925.57 −30.21 −49.04 9.47 3.7221 3.9954 −1.5879 −2.9441 15.2391 −5.4585 1.7418
AlO2H@ −864.28 −947.13 20.92 −209.21 13.01 3.5338 0.8485 5.4132 −2.8140 −23.4129 −13.2195 −0.0300
AlOH+2 −692.60 −767.27 −184.93 55.97 −2.73 2.0469 −2.7813 6.8376 −2.6639 29.7923 −0.3457 1.7247
Ca(CO3)@ −1099.18 −1201.92 10.46 −123.86 −15.65 −0.3907 −8.7325 9.1753 −2.4179 −11.5309 −9.0641 −0.0380
Ca(HCO3)+ −1146.04 −1231.94 66.94 233.70 13.33 3.7060 1.2670 5.2520 −2.8310 41.7220 8.3360 0.3080
Ca(HSiO3)+ −1574.24 −1686.48 −8.33 137.80 −6.74 1.0647 −5.1787 7.7785 −2.5649 30.8048 3.6619 0.5831
Ca(SO4)@ −1310.38 −1448.43 20.92 −104.60 4.70 2.4079 −1.8992 6.4895 −2.7004 −8.4942 −8.1271 −0.0010
Ca+2 −552.79 −543.07 −56.48 −30.92 −18.44 −0.1947 −7.2520 5.2966 −2.4792 9.0000 −2.5220 1.2366

CaOH+ −717.02 −751.65 28.03 6.05 5.76 2.7243 −1.1303 6.1958 −2.7322 11.1286 −2.7493 0.4496
CH4@ −34.35 −87.81 87.82 277.26 37.40 6.7617 8.7279 2.3212 −3.1397 42.0941 10.4707 −0.3179
Cl− −131.29 −167.11 56.74 −122.49 17.34 4.0320 4.8010 5.5630 −2.8470 −4.4000 −5.7140 1.4560
ClO4− −8.54 −129.33 182.00 −24.00 43.90 8.1411 15.5654 −7.8077 −3.4224 16.4500 −6.5700 0.9699
CO2@ −386.02 −413.84 117.57 243.08 32.81 6.2466 7.4711 2.8136 −3.0879 40.0325 8.8004 −0.0200
CO3−2 −527.98 −675.31 −50.00 −289.33 −6.06 2.8524 −3.9844 6.4142 −2.6143 −3.3206 −17.1917 3.3914
e− 0 0 65.34 14.42 0 0 0 0 0 0 0 0
Fe(CO3)@ −644.49 −763.51 −58.45 −123.03 −17.23 −0.6069 −9.2604 9.3828 −2.3961 −11.4137 −9.0233 −0.0380
Fe(HCO3)+ −689.86 −794.10 −8.87 231.41 8.18 3.1064 −0.1934 5.8191 −2.7710 43.9175 8.2195 0.5831
Fe(HSO4)+ −853.48 −990.45 10.21 338.23 18.81 4.5330 3.2897 4.4500 −2.9149 58.2305 13.4217 0.5121
Fe(HSO4)+2 −787.15 −981.91 −248.95 426.71 2.32 2.8251 −0.8804 6.0891 −2.7426 83.8315 17.6994 1.9551
Fe(SO4)@ −848.81 −993.86 −16.86 −101.60 1.67 1.9794 −2.9454 6.9007 −2.6572 −8.4131 −7.9804 −0.0380

Fe(SO4)+ −784.71 −942.42 −124.68 −145.93 −2.64 1.7837 −3.4232 7.0885 −2.6374 −5.1341 −10.1600 0.9986
Fe(SO4)2− −1536.81 −1854.38 −87.78 −210.37 30.49 6.6756 8.5215 2.3937 −3.1312 −5.4923 −13.3173 1.9457
Fe+2 −91.50 −92.24 −105.86 −32.44 −22.64 −0.7867 −9.6969 9.5479 −2.3780 14.7860 −4.6437 1.4382
Fe+3 −17.19 −49.58 −277.40 −76.71 −37.79 −2.4256 −13.6961 11.1141 −2.2127 19.0459 −6.8233 2.5812
FeCl+ −223.59 −258.05 −42.09 86.49 0.85 2.1468 −2.5367 6.7401 −2.6741 24.6912 1.1617 0.7003
FeCl+2 −156.92 −212.67 −178.82 14.83 −22.86 −0.7164 −9.5277 9.4878 −2.3851 23.8149 −2.3482 1.7013
FeCl2+ −291.92 −385.75 −129.66 300.72 10.27 3.5610 0.9165 5.3828 −2.8168 57.6940 11.5846 1.0276
FeCl3@ −417.51 −564.39 −131.06 368.22 35.94 6.6686 8.5038 2.4024 −3.1304 57.3959 14.8930 −0.0380
Table D.1 (continued)
FeO+ −222.00 −255.09 −46.44 −200.94 −42.02 −3.7118 −16.8408 12.3595 −2.0827 −15.3982 −12.8325 0.7191
FeO2− −368.26 −443.82 44.35 −234.93 0.45 2.3837 −1.9602 6.5182 −2.6979 −13.3207 −14.5028 1.4662
FeO2H@ −419.86 −480.95 92.88 −312.14 7.21 2.7401 −1.0905 6.1776 −2.7338 −37.8300 −18.2305 −0.0300
FeOH+ −274.46 −325.65 −41.84 63.06 −16.71 −0.2561 −8.4029 9.0457 −2.4315 21.4093 0.0209 0.7003
FeOH+2 −241.87 −292.79 −106.27 −33.69 −25.34 −1.1562 −10.6009 9.9077 −2.3407 14.6102 −4.7048 1.4382
H+ 0 0 0 0 0 0 0 0 0 0 0 0
H2@ 17.73 −4.02 57.74 166.85 25.26 5.1427 4.7758 3.8729 −2.9764 27.6251 5.0930 −0.2090
H2O@ −237.18 −285.88 69.92 75.36 18.07 0 0 0 0 0 0 0
H2S@ −27.93 −39.03 125.52 179.17 34.95 6.5097 6.7724 5.9646 −3.0590 32.3000 4.7300 −0.1000
HCN@ 114.37 103.75 131.30 0 0 0 0 0 0 0 0 0
HCO3– −586.94 −690.01 98.45 −34.85 24.21 7.5621 1.1505 1.2346 −2.8266 12.9395 −4.7579 1.2733
HS− 11.97 −16.22 68.20 −93.93 20.21 5.0119 4.9799 3.4765 −2.9849 3.4200 −6.2700 1.4410
HSiO3− −1014.60 −1144.68 20.92 −87.20 4.53 2.9735 −0.5181 5.9467 −2.7575 8.1489 −7.3123 1.5511
HSO3− −529.10 −627.70 139.75 −5.38 32.96 6.7014 8.5816 2.3771 −3.1338 15.6949 −3.3198 1.1233
HSO4− −755.81 −889.23 125.52 22.68 34.84 6.9788 9.2590 2.1108 −3.1618 20.0961 −1.9550 1.1748
K(SO4)− −1031.77 −1158.77 146.44 −45.13 27.46 5.9408 6.7274 3.0989 −3.0571 9.9089 −5.2549 1.0996
K+ −282.46 −252.14 101.04 8.39 9.01 3.5590 −1.4730 5.4350 −2.7120 7.4000 −1.7910 0.1927
KOH@ −437.11 −474.15 108.37 −85.02 14.96 3.7938 1.4839 5.1619 −2.8402 −6.1240 −7.2104 −0.0500
Mg(CO3)@

−998.98 −1132.12 −100.42 −116.50 −16.78 −0.5450 −9.1130 9.3320 −2.4020 −10.4990 −8.7060 −0.0380
Mg(HCO3)+ −1047.02 −1153.97 −12.55 254.42 9.34 3.2710 0.2060 5.6690 −2.7880 47.2840 9.3400 0.5990
Mg(HSiO3)+ −1477.15 −1613.91 −99.50 158.65 −10.85 0.6289 −6.2428 8.1967 −2.5209 36.7882 4.6702 0.9177
Mg+2 −453.99 −465.93 −138.07 −21.66 −22.01 −0.8217 −8.5990 8.3900 −2.3900 20.8000 −5.8920 1.5372
MgOH+ −625.87 −690.02 −79.91 129.23 1.64 2.3105 −2.1365 6.5827 −2.6906 32.0008 3.2394 0.8449
MgSO4@ −1211.97 −1368.77 −50.88 −90.31 1.81 1.9985 −2.8987 6.8823 −2.6591 −6.8307 −7.4304 −0.0380
N2@ 18.19 −10.37 95.81 234.16 33.41 6.2046 7.3685 2.8539 −3.0836 35.7911 8.3726 −0.3468
Na(CO3)− −797.11 −938.56 −44.31 −51.28 −0.42 2.3862 −1.9521 6.5103 −2.6982 15.3395 −5.5686 1.7870
Na(HCO3)@ −847.39 −929.50 154.72 200.33 32.32 6.1730 7.2943 2.8760 −3.0805 33.8790 6.7193 −0.0380
Na(SO4)- −1010.34 −1146.66 101.75 −30.09 18.64 4.7945 3.9284 4.1990 −2.9414 13.4899 −4.5256 1.2606
Na+ −261.88 −240.28 58.41 38.12 −1.21 1.8390 −2.2850 3.2560 −2.7260 18.1800 −2.9810 0.3306
NaOH@ −418.12 −470.14 44.77 −13.40 3.51 2.2338 −2.3287 6.6683 −2.6826 4.0146 −3.6863 −0.0300

NH3@ −26.67 −81.53 107.82 76.89 24.45 5.0911 2.7970 8.6248 −2.8946 20.3000 −1.1700 −0.0500
NH4+ −79.40 −133.26 111.17 67.11 18.08 3.8763 2.3448 8.5605 −2.8759 17.4500 −0.0210 0.1502
NO3− −110.91 −206.89 146.94 −66.80 28.66 7.3161 6.7824 −4.6838 −3.0594 7.7000 −6.7250 1.0977
O2@ 16.45 −12.24 108.95 234.13 30.50 5.7889 6.3536 3.2528 −3.0417 35.3530 8.3726 −0.3943
OH– −157.27 −230.01 −10.71 −136.34 −4.71 1.2527 0.0738 1.8423 −2.7821 4.1500 −10.3460 1.7246
S2O3−2 −519.99 −649.86 66.94 −238.47 27.59 6.6685 12.4951 −7.7281 −3.2955 −0.0577 −14.7066 2.9694
SCN− 92.70 76.40 144.01 −39.69 35.36 7.0244 9.3687 2.0708 −3.1662 10.7414 −4.9900 1.1073
Table D.1 (continued)
SO3−2 −487.89 −636.89 −29.29 −280.99 −4.12 2.4632 −1.7691 6.4494 −2.7058 −2.7967 −16.7843 3.3210
SO4−2 −744.46 −909.70 18.83 −266.09 12.92 8.3014 −1.9846 −6.2122 −2.6970 1.6400 −17.9980 3.1463
Sr(CO3)@ −1107.83 −1207.29 35.56 −134.32 −15.23 −0.3332 −8.5922 9.1201 −2.4237 −12.9961 −9.5733 −0.0380
Sr(HCO3)+ −1157.54 −1239.00 95.94 210.07 14.08 3.7702 1.4274 5.1820 −2.8380 37.4746 7.1883 0.2058
Sr(SO4)@ −1321.37 −1451.50 61.59 −110.60 5.02 2.4382 −1.8251 6.4604 −2.7035 −9.6731 −8.4183 −0.0380
Sr+2 −563.84 −550.87 −31.51 −41.56 −17.76 0.7071 −10.1508 7.0027 −2.3594 10.7452 −5.0818 1.1363
SrOH+ −725.16 −754.14 61.09 −31.66 7.10 2.8620 −0.7922 6.0586 −2.7462 4.7576 −4.5826 0.3306
Temperature correction using Cp(T) integration a0 a1 a2 (Cp0 = a0 + a1T + a2T−2)
SiO2@⁎⁎ −833.41 −887.86⁎ 41.34 44.47 1.61 46.94 0.034 −1.13E + 06
Temperature correction using logK(T) A0 A1 A2 (logKT = A0 + A1T + A2T−1)
SiO3–2⁎⁎ −938.51 −1098.74 −80.20 119.83 0 −10.0006 0 −3917.5
Si4O10–4⁎⁎⁎ −3600.81 −3915.99 305.20 328.58 0 0 0 −10,822.8
CaSiO3@⁎⁎ −1517.56 −1668.06 −136.68 88.90 0 0 0 1371.49
MgSiO3@ −1425.03 −1554.54 −75.17 −264.79 0 5.7 0 0
AlSiO5−3 ⁎⁎⁎ −1769.01 −2027.33 −110.41 70.78 −3.41 0 0 158.02
AlHSiO3+2 ⁎v −1540.55 −1634.31 −24.99 −215.896 0 14.5828 0 −2141.57
FeHSiO3+2 ⁎v −1087.15 −1194.26 −70.77 −163.91 0 9.7 0 0
Fe2(OH)2+4 ⁎v −491.9 −614.44 −281.97 −2.71 0 6.94586 0 −2950.45

Fe3(OH)4+5 ⁎v −964.33 −1232.44 −472.43 71.30 0 4.1824 0 −3125.33

SrSiO3@⁎⁎⁎ −1527.29 −1617.43 79.92 78.39 1.64 0 0 1302.92
S−2 120.42 −16.22 −295.55 −93.93 0 −19 0 0
⁎
Parameters of the HKF-equation of state; given in original calorimetric units (see [25,26,133]) as used in GEM.
Calculated in Matschei et al. [7] assuming ΔrS° = ΔrC°p = 0 using S° and C°p from SiO2 (quartz) for the reactions:
SiO2 0 > SiO2 (quartz) r G° = r H° = 21.386; SiO32 + 2H+ > SiO2 0 + H2 O r G° = 132.08, r H° = 75, r S° = 191.46, r C° p .
⁎⁎
= 0; SiO32 + Ca2 + > CaSiO30 r G° = r H° = 26.257, r S° = 0, r C° p = 0

Calculated in this paper assuming ΔrS° = ΔrC°p = 0 using S° and C°p from SiO2 (quartz) for the reactions:
SiO32 +AlO2 > AlSiO53 r G° = r G° = 3.025, r S° = 0, r C° p = 0; Si 4 O10 4 +4H+ > 4SiO2 0 + 2H2 O = = 207.2, = 0,
⁎⁎⁎
° ° ° °
rG rH rS rC p .
= 0; SiO32 +Mg2 + > MgSiO30 rG
° = 32.54, rH
° = 0, rS
° = 109.126, °
rC p = 0; SiO32 + Sr 2 + > SrSiO30 rG
° = rH
° = 29.944, rS
°
= 0, C° =0

r p
From the GEMS version of the PSI/Nagra 12/07 TDB [22,23]:
Al+3 + HSiO3 > AlHSiO3+2 = r H = 41, = 279.19, r C° p = 0; Fe+3 + HSiO3 > FeHSiO3+2 rG = = 0, r S° = 185.7,
⁎v
° ° ° ° ° °
rG 42.24, rS 55.37, rH r C p.
= 0; 2Fe+3+2 H2 O > Fe2 (OH)2+4 + 2H+ rG

° = 16.84, rH
° = 56.486, rS
° = 132.98, °
rC p
= 0; 3Fe+3 + 4H2 O > Fe3 (OH)4+5 + 4H+ r G = 35.96,

°
r H = 59.834,
°
r S = 80.07,
° °
rC p =0
Table D.2
Standard (partial molal) thermodynamic properties and heat capacity coefficients (Cp0 = a0 + a1T + a2T−2) of gaseous species at 25 °C, 1 bar used
in GEM calculations, as used in the GEMS version of the PSI/Nagra 12/07 TDB [22,23]. Standard molar volumes for ideal gases are not listed because
they are all equal to 2479 J/bar at 1 bar, 298.15 K.
Species ΔG0 ΔH0 S0 Cp0 a0 a1 a2

2
(kJ/mol) (kJ/mol) (J/mol K) (J/mol K) (J/mol/K) (J/mol/K ) (J∙K/mol)
CH4 −50.66 −74.81 186.26 35.75 23.64 0.0479 −192,464

CO2 −394.39 −393.51 213.74 37.15 44.22 0.0088 −861,904
H2 0 0 130.68 28.82 27.28 0.0033 50,208
H2O −228.68 −242.40 187.25 40.07 52.99 −0.0435 5472
H2S −33.75 −20.63 205.79 34.20 32.68 0.0124 −192,464
N2 0 0 191.61 29.13 28.58 0.0038 −50,208
O2 0 0 205.14 29.32 29.96 0.0042 −167,360
References (2008) 1–18.

[22] T. Thoenen, W. Hummel, U. Berner, E. Curti, The PSI/Nagra Chemical
Thermodynamic Data Base 12/07, PSI Report 14-04, Villigen PSI, Switzerland,
[1] B. Lothenbach, T. Matschei, G. Möschner, F.P. Glasser, Thermodynamic modelling 2014.
of the effect of temperature on the hydration and porosity of Portland cement, [23] T. Thoenen, D.A. Kulik, Nagra/PSI Chemical Thermodynamic Data Base 01/01 for
Cem. Concr. Res. 38 (2008) 1–18. the GEM-Selektor (V.2- PSI) Geochemical Modeling Code: Release 28-02-03.
[2] T. Matschei, B. Lothenbach, F.P. Glasser, The role of calcium carbonate in cement Internal Report TM-44-03-04, available from http://gems.web.psi.ch/TDB/doc/
hydration, Cem. Concr. Res. 37 (2007) 551–558. pdf/TM-44-03-04-web.pdf, (2003) (checked 2018-04-19).
[3] M. Moesgaard, D. Herfort, M. Steenberg, L.F. Kirkegard, Y. Yue, Physical perfor- [24] W. Hummel, U. Berner, E. Curti, F.J. Pearson, T. Thoenen, Nagra/PSI Chemical
mance of blended cements containing calcium aluminosilicate glass powder and Thermodynamic Data Base 01/01, Universal Publishers/uPUBLISH.com, USA,
limestone, Cem. Concr. Res. 41 (2011) 359–364. Also Published as Nagra Technical Report NTB 02-16, Wettingen, Switzerland,
[4] T. Matschei, F.P. Glasser, Temperature dependence, 0 to 40 °C, of the mineralogy 2002.
of Portland cement paste in the presence of calcium carbonate, Cem. Concr. Res. [25] E. Shock, D. Sassani, M. Willis, D. Sverjensky, Inorganic species in geologic fluids:
40 (2010) 763–777. Correlations among standard molal thermodynamic properties of aqueous ions
[5] B. Lothenbach, F. Winnefeld, C. Alder, E. Wieland, P. Lunk, Effect of temperature and hydroxide complexes, Geochim. Cosmochim. Acta 61 (1997) 907–950.
on the pore solution, microstructure and hydration products of Portland cement [26] D. Sverjensky, E. Shock, H. Helgeson, Prediction of the thermodynamic properties
pastes, Cem. Concr. Res. 37 (2007) 483–491. of aqueous metal complexes to 1000 C and 5 kb, Geochim. Cosmochim. Acta 61
[6] F. Deschner, B. Lothenbach, F. Winnefeld, J. Neubauer, Effect of temperature on (1997) 1359–1412.
the hydration Portland cement blended with siliceous fly ash, Cem. Concr. Res. 52 [27] M. Balonis, B. Lothenbach, G. Le Saout, F.P. Glasser, Impact of chloride on the
(2013) 169–181. mineralogy of hydrated Portland cement systems, Cem. Concr. Res. 40 (2010)
[7] T. Matschei, B. Lothenbach, F.P. Glasser, Thermodynamic properties of Portland 1009–1022.
cement hydrates in the system CaO-Al2O3-SiO2-CaSO4-CaCO3-H2O, Cem. Concr. [28] T. Matschei, F.P. Glasser, The thermal stability of thaumasite, Mater. Struct. 48
Res. 37 (2007) 1379–1410. (2015) 2277–2289.
[8] B.Z. Dilnesa, B. Lothenbach, G. Le Saout, G. Renaudin, A. Mesbah, Y. Filinchuk, [29] B. Lothenbach, F. Winnefeld, Thermodynamic modelling of the hydration of
A. Wichser, E. Wieland, Iron in carbonate containing AFm phases, Cem. Concr. Portland cement, Cem. Concr. Res. 36 (2006) 209–226.
Res. 41 (2011) 311–323. [30] L.G. Baquerizo, T. Matschei, K.L. Scrivener, Impact of water activity on the sta-
[9] B.Z. Dilnesa, B. Lothenbach, G. Renaudin, A. Wichser, D. Kulik, Synthesis and bility of ettringite, Cem. Concr. Res. 76 (2016) 31–44.
characterization of hydrogarnet Ca3(AlxFe1−x)2(SiO4)y(OH)4(3−y), Cem. Concr. [31] L.G. Baquerizo, T. Matschei, K.L. Scrivener, Hydration states of AFm cement
Res. 59 (2014) 96–111. phases, Cem. Concr. Res. 73 (2015) 143–157.
[10] B.Z. Dilnesa, B. Lothenbach, G. Renaudin, A. Wichser, E. Wieland, Stability of [32] L.G. Baquerizo, T. Matschei, K.L. Scrivener, M. Saeidpour, A. Thorell, L. Wadsö,
monosulfate in the presence of iron, J. Am. Ceram. Soc. 95 (2012) 3305–3316. Methods to determine hydration states of minerals and cement hydrates, Cem.
[11] D.A. Kulik, Improving the structural consistency of C-S-H solid solution thermo- Concr. Res. 65 (2014) 85–95.
dynamic models, Cem. Concr. Res. 41 (2011) 477–495. [33] M. Balonis, F.P. Glasser, The density of cement phases, Cem. Concr. Res. 39 (2009)
[12] B. Lothenbach, L. Pelletier-Chaignat, F. Winnefeld, Stability in the system 733–739.
CaO–Al2O3–H2O, Cem. Concr. Res. 42 (2012) 1621–1634. [34] M. Balonis, The Influence of Inorganic Chemical Accelerators and Corrosion
[13] D. Kulik, T. Wagner, S. Dmytrieva, G. Kosakowski, F. Hingerl, K. Chudnenko, Inhibitors on the Mineralogy of Hydrated Portland Cement Systems, Thesis
U. Berner, GEM-Selektor geochemical modeling package: revised algorithm and University of Aberdeen, Aberdeen, UK, 2010.
GEMS3K numerical kernel for coupled simulation codes, Comput. Geosci. 17 [35] M. Balonis, M. Medala, F.P. Glasser, Influence of calcium nitrate and nitrite on the
(2013) 1–24. constitution of AFm and AFt cement hydrates, Adv. Cem. Res. 23 (2011) 129–143.
[14] T. Wagner, D.A. Kulik, F.F. Hingerl, S.V. Dmytrieva, GEM-Selektor geochemical [36] M. Balonis, F.P. Glasser, Calcium nitrite corrosion inhibitor in portland cement:
modeling package: TSolMod library and data interface for multicomponent phase influence of nitrite on chloride binding and mineralogy, J. Am. Ceram. Soc. 94
models, Can. Mineral. 50 (2012) 1173–1195. (2011) 2230–2241.
[15] P. Blanc, X. Bourbon, A. Lassin, E. Gaucher, Chemical model for cement-based [37] B.Z. Dilnesa, Fe-containing Hydrates and Their Fate During Cement Hydration:
materials: Temperature dependence of thermodynamic functions for nanocrys- Thermodynamic Data and Experimental Study, Thesis EPFL, Lausanne, 2012.
talline and crystalline C-S-H phases, Cem. Concr. Res. 40 (2010) 851–866. [38] D. Garvin, V.B. Parker, H.J. White, CODATA Thermodynamic Tables. Selections
[16] P. Blanc, X. Bourbon, A. Lassin, E. Gaucher, Chemical model for cement-based for Some Compounds of Calcium and Related Mixtures: A Prototype Set of Tables,
materials: thermodynamic data assessment for phases other than CSH, Cem. Springer Verlag, Berlin, 1987.
Concr. Res. 40 (2010) 1360–1374. [39] K.B. Rozov, U. Berner, D.A. Kulik, L.W. Diamond, Solubility and thermodynamic
[17] C.M. Bethke, Geochemical and Biogeochemical Reaction Modeling, 2nd Ed., properties of carbonate-bearing hydrotalcite-pyroaurite solid solutions with a 3:1
Cambridge University Press, New York, NY, USA, 2008. Mg/(Al + Fe) mole ratio, Clay Clay Miner. 59 (2011) 215–232.
[18] D.J. Parkhurst, C.A.J. Appelo, Description of Input and Examples for PHREEQC [40] D. Nied, K. Enemark-Rasmussen, E. L'Hôpital, J. Skibsted, B. Lothenbach,
Version 3 - A Computer Program for Speciation, Batch-reaction, One-dimensional Properties of magnesium silicate hydrates (M-S-H), Cem. Concr. Res. 79 (2016)
Transport, and Inverse Geochemical Calculations, 6 USGS, Denver, CO, USA, 323–332.
2013. [41] B. Lothenbach, E. Bernard, U. Mäder, Zeolite formation in the presence of cement
[19] J.E. Cross, F.T. Ewart, HATCHES - a thermodynamic database and management hydrates and albite, Phys. Chem. Earth 99 (2017) 77–94.
system, Radiochim. Acta 52/53 (1991) 421–422. [42] V.J. Babushkin, G.M. Matveyev, O.P. Mchedlov-Petrossyan, Thermodynamics of
[20] D. Damidot, B. Lothenbach, D. Herfort, F.P. Glasser, Thermodynamics and cement Silicates, Springer-Verlag, Berlin, H, 1985.
science, Cem. Concr. Res. 41 (2011) 679–695. [43] H.C. Helgeson, J.M. Delany, H.W. Nesbitt, D.K. Bird, Summary and critique of the
[21] G. Möschner, B. Lothenbach, J. Rose, A. Ulrich, R. Figi, R. Kretzschmar, Solubility thermodynamic properties of rock-forming minerals, Am. J. Sci. 278-A (1978)
of Fe-ettringite (Ca6[Fe(OH)6]2(SO4)3⋯26H2O), Geochim. Cosmochim. Acta 72 1–229.

[44] R.A. Robie, B.S. Hemingway, Thermodynamic properties of minerals and related silicate, β-dicalcium silicate, Portland cement, and blends of Portland cement with
substances at 298.15 K and 1 bar (105 Pascals) pressures and at higher tempera- blast-furnace slag, metakaolin, or silica fume, Cem. Concr. Res. 34 (2004)
tures, U.S. Geol. Surv. Bull. 2131 (1995) 1–461. 1733–1777.
[45] G. Möschner, B. Lothenbach, A. Ulrich, R. Figi, R. Kretschmar, Solid solution be- [78] A. Muller, K. Scrivener, A. Gajewicz, P. McDonald, Use of bench-top NMR to
tween Al-ettringite and Fe-ettringite (Ca6[Al1−xFex(OH)6]2(SO4)3·26H2O), Cem. measure the density, composition and desorption isotherm of C–S–H in cement
Concr. Res. 39 (2009) 482–489. paste, Microporous Mesoporous Mater. 178 (2013) 99–103.
[46] E. Corazza, C. Sabelli, The crystal structure of syngenite, K2Ca(SO4)2·H2O, Z [79] A. Muller, K. Scrivener, J. Skibsted, A. Gajewicz, P. McDonald, Influence of silica
Kristallog 124 (1967) 398–408. fume on the microstructure of cement pastes: new insights from 1H NMR re-
[47] P. Blanc, P. Vieillard, H. Gailhanou, S. Gaboreau, N. Marty, F. Claret, B. Made, laxometry, Cem. Concr. Res. 74 (2015) 116–125.
E. Giffaut, ThermoChimie database developments in the framework of cement/ [80] C.S. Walker, S. Sutou, C. Oda, M. Mihara, A. Honda, Calcium silicate hydrate (C-S-
clay interactions, Appl. Geochem. 55 (2015) 95–107. H) gel solubility data and a discrete solid phase model at 25 °C based on two
[48] R.M. Milton, Molecular sieve adsorbents, US Patent No (1959) 2,882,244. binary non-ideal solid solutions, Cem. Concr. Res. 79 (2016) 1–30.
[49] G. Gottardi, E. Galli, Natural Zeolites, Mineral and Rocks, 18 (1985). [81] S.-Y. Hong, F.P. Glasser, Alkali binding in cement pastes: part I. The C-S-H phase,
[50] H. Boysen, M. Lerch, A. Stys, A. Senyshyn, Structure and oxygen mobility in Cem. Concr. Res. 29 (1999) 1893–1903.
mayenite (Ca12Al14O33): a high-temperature neutron powder diffraction study, [82] E. L'Hôpital, B. Lothenbach, D. Kulik, K. Scrivener, Influence of calcium to silica
Acta Crystallogr. B63 (2007) 675–682. ratio on aluminium uptake in calcium silicate hydrate, Cem. Concr. Res. 85 (2016)
[51] W. Hörkner, H. Müller-Buschbaum, Zur Kristallstruktur von CaAl2O4, J. Inorg. 111–121.
Nucl. Chem. 38 (1976) 983–984. [83] E. L'Hôpital, B. Lothenbach, G. Le Saout, D.A. Kulik, K. Scrivener, Incorporation of
[52] D.W. Goodwin, A.J. Lindop, The crystal structure of CaO·2Al2O3, Acta Crystallogr aluminium in calcium-silicate hydrate, Cem. Concr. Res. 75 (2015) 91–103.
B 26 (1970) 1230–1235. [84] E. L'Hôpital, B. Lothenbach, K. Scrivener, D.A. Kulik, Alkali uptake in calcium
[53] F. Winnefeld, B. Lothenbach, Phase equilibria in the system Ca4Al6O12SO4 – alumina silicate hydrate (C-A-S-H), Cem. Concr. Res. 85 (2016) 122–136.
Ca2SiO4 – CaSO4 – H2O referring to the hydration of calcium sulfoaluminate ce- [85] D.A. Kulik, M. Kersten, Aqueous solubility diagrams for cementitious waste sta-
ments, RILEM Tech. Lett. 1 (2016) 10–16. bilization systems: II, end-member stoichiometries of ideal calcium silicates hy-
[54] D. Damidot, S.J. Barnett, F.P. Glasser, D.E. Macphee, Investigation of the CaO- drate solid solutions, J. Am. Ceram. Soc. 84 (2001) 3017–3026.
Al2O3-SiO2-CaSO4-CaCO3-H2O system at 25 °C by thermodynamic calculation, [86] J. Haas, A. Nonat, From C–S–H to C–A–S–H: experimental study and thermo-
Adv. Cem. Res. 16 (2004) 69–76. dynamic modelling, Cem. Concr. Res. 68 (2015) 124–138.
[55] T. Schmidt, B. Lothenbach, M. Romer, K.L. Scrivener, D. Rentsch, R. Figi, A [87] S.V. Churakov, C. Labbez, Thermodynamics and molecular mechanism of Al in-
thermodynamic and experimental study of the conditions of thaumasite formation, corporation in calcium silicate hydrates, J. Phys. Chem. C 121 (2017) 4412–4419.
Cem. Concr. Res. 38 (2008) 337–349. [88] D. Kulik, J. Tits, E. Wieland, Aqueous-solid solution model of strontium uptake in
[56] D.E. Macphee, S.J. Barnett, Solution properties of solids in the ettringite-thau- C-S-H phases, Geochim. Cosmochim. Acta 71 (2007) A530.
masite solid solution series, Cem. Concr. Res. 34 (2004) 1591–1598. [89] B. Lothenbach, G. Le Saout, M. Ben Haha, R. Figi, E. Wieland, Hydration of a low-
[57] F. Bellmann, On the formation of thaumasite CaSiO3·CaSO4·CaCO3·15H2O: part I, alkali CEM III/B-SiO2 cement (LAC), Cem. Concr. Res. 42 (2012) 410–423.
Adv. Cem. Res. 16 (2004) 55–60. [90] R. Myers, S.A. Bernal, J.L. Provis, A thermodynamic model for C-(N-)A-S-H gel:
[58] U.A. Birnin-Yauri, F.P. Glasser, Friedel's salt, Ca2Al(OH)6(Cl,OH)·2H2O: its solid CNASH_ss. Derivation and validation, Cem. Concr. Res. 66 (2014) 27–47.
solutions and their role in chloride binding, Cem. Concr. Res. 28 (1998) [91] D. Kulik, Dual-thermodynamic estimation of stoichiometry and stability of solid
1713–1723. solution end members in aqueous-solid solution systems, Chem. Geol. 225 (2006)
[59] M.Y. Hobbs, Solubilities and Ion Exchange Properties of Solid Solutions Between 189–212.
OH, Cl and CO3 End Members of the Monocalcium Aluminate Hydrates, Thesis [92] J. Tits, E. Wieland, C.J. Müller, C. Landesman, M.H. Bradbury, Strontium binding
University of Waterloo, Ontario, Canada, 2001. by calcium silicate hydrates, J. Colloid Interface Sci. 300 (2006) 78–87.
[60] J.V. Bothe Jr, P.W. Brown, PhreeqC modeling of Friedel's salt equilibria at [93] D.D. Wagman, E.H. Evans, V.B. Parker, R.H. Schumm, I. Halow, S.M. Bailey,
23 ± 1 °C, Cem. Concr. Res. 34 (2004) 1057–1063. K.L. Churney, L. N.R., The NBS tables of chemical thermodynamic properties.
[61] R.O. Grishchenko, A.L. Emelina, P.Y. Makarov, Thermodynamic properties and Selected values for inorganic and C1 and C2 organic substances in SI units, J. Phys.
thermal behavior of Friedel's salt, Thermochim. Acta 570 (2013) 74–79. Chem. Ref. Data 11 (Suppl. 2) (1982) 1–392.
[62] F.P. Glasser, A. Kindness, S.A. Stronach, Stability and solubility relationships in [94] I.G. Richardson, The calcium silicate hydrates, Cem. Concr. Res. 38 (2008)
AFm phases. Part I. Chloride, sulfate and hydroxide, Cem. Concr. Res. 29 (1999) 137–158.
861–866. [95] J.J. Chen, J.J. Thomas, H.F.W. Taylor, H.M. Jennings, Solubility and structure of
[63] G. Falzone, M. Balonis, G. Sant, X-AFm stabilization as a mechanism of bypassing calcium silicate hydrate, Cem. Concr. Res. 34 (2004) 1499–1519.
conversion phenomena in calcium aluminate cements, Cem. Concr. Res. 72 (2015) [96] K. Garbev, M. Bornefeld, G. Beuchle, P. Stemmermann, Cell dimensions and
54–68. composition of nanocrystalline calcium silicate hydrate solid solutions. part 2: X-
[64] B.Z. Dilnesa, E. Wieland, B. Lothenbach, R. Dähn, K. Scrivener, Fe-containing Ray and thermogravimetry study, J. Am. Ceram. Soc. 91 (2008) 3015–3023.
phases in hydrated cements, Cem. Concr. Res. 58 (2014) 45–55. [97] X. Gaona, D.A. Kulik, N. Macé, E. Wieland, Aqueous–solid solution thermo-
[65] M. Vespa, E. Wieland, R. Dähn, B. Lothenbach, Identification of the thermodynamic model of U(VI) uptake in C–S–H phases, Appl. Geochem. 27 (2012)
dynamically stable Fe-containing phase in aged cement pastes, J. Am. Ceram. Soc. 81–95.
98 (2015) 2286–2294. [98] A.J. Allen, J.J. Thomas, H.M. Jennings, Composition and density of nanoscale
[66] H.F.W. Taylor, D.E. Newbury, An electron microporbe study of a mature cement calcium–silicate–hydrate in cement, Nat. Mater. 6 (2007) 311–316.
paste, Cem. Concr. Res. 14 (1984) 565–573. [99] J.J. Thomas, A.J. Allen, H.M. Jennings, Density and water content of nanoscale
[67] U. Mäder, A. Jenni, C. Lerouge, S. Gaboreau, S. Miyoshi, Y. Kimura, V. Cloet, solid C–S–H formed in alkali-activated slag (AAS) paste and implications for
M. Fukaya, F. Claret, T. Otake, M. Shibata, B. Lothenbach, 5-year chemico-phy- chemical shrinkage, Cem. Concr. Res. 42 (2012) 377–383.
sical evolution of concrete–claystone interfaces, Mont Terri rock laboratory [100] J.L. Provis, S.A. Bernal, Geopolymers and related alkali-activated materials, Annu.
(Switzerland), Swiss J. Geosci. 110 (2017) 307–327. Rev. Mater. Res. 44 (2014) 299–327.
[68] S.A. Bernal, R. San Nicols, R.J. Myers, R. Mejia de Gutierrez, F. Puertas, J.S.J. Van [101] R.J. Myers, S.A. Bernal, J.L. Provis, Phase diagrams for alkali-activated slag bin-
Deventer, J.L. Provis, MgO content of slag controls phase evolution and structural ders, Cem. Concr. Res. 95 (2017) 30–38.
changes induced by accelerated carbonation in alkali-activated binders, Cem. [102] C.S. Walker, D. Savage, M. Tyrer, K.V. Ragnarsdottir, Non-ideal solid solution
Concr. Res. 57 (2014) 33–43. aqueous solution modeling of synthetic calcium silicate hydrate, Cem. Concr. Res.
[69] I.G. Richardson, A.R. Brough, G.W. Groves, C.M. Dobson, The characterization of 37 (2007) 502–511.
hardened alkali-activated blast-furnace slag pastes and the nature of the calcium [103] L. Gomez-Zamorano, M. Balonis, B. Erdemli, N. Neithalath, G. Sant, C–(N)–S–H
silicate hydrate (C-S-H) phase, Cem. Concr. Res. 24 (1994) 813–829. and N–A–S–H gels: Compositions and solubility data at 25 °C and 50 °C, J. Am.
[70] H. Taylor, Crystal structures of some double hydroxide minerals, Mineral. Mag. 39 Ceram. Soc. 100 (2017) 2700–2711.
(1973) 377–389. [104] R.J. Myers, E. L'Hôpital, J.L. Provis, B. Lothenbach, Effect of temperature and
[71] I. Richardson, Clarification of possible ordered distributions of trivalent cations in aluminium on calcium (alumino)silicate hydrate chemistry under equilibrium
layered double hydroxides and an explanation for the observed variation in the conditions, Cem. Concr. Res. 68 (2015) 83–93.
lower solid-solution limit, Acta Crystallogr. B 69 (2013) 629–633. [105] R.J. Myers, E. L'Hôpital, J.L. Provis, B. Lothenbach, Composition-solubility-struc-
[72] D.G. Bennett, D. Read, M. Atkins, F.P. Glasser, A thermodynamic model for ture relationships in calcium (alkali) aluminosilicate hydrate (C-(N,K-)A-S-H),
blended cements. II: cement hydrate phases; thermodynamic values and modelling Dalton Trans. 44 (2015) 13530–13544.
studies, J. Nucl. Mater. 190 (1992) 315–325. [106] G.D. Miron, D.A. Kulik, S.V. Dmytrieva, T. Wagner, GEMSFITS: code package for
[73] W. Gao, Z. Li, Solubility and K SP of Mg4Al2(OH)14·3H2O at the various ionic optimization of geochemical model parameters and inverse modeling, Appl.
strengths, Hydrometallurgy 117–118 (2012) 36–46. Geochem. 55 (2015) 28–45.
[74] R.J. Myers, B. Lothenbach, S. Bernal, J.L. Provis, Thermodynamic modelling of [107] L. Nicoleau, E. Schreiner, Determination of Ca2+ complexation constants by
alkali-activated slag-based cements, Appl. Geochem. 61 (2015) 233–247. monomeric silicate species at 25 °C with a Ca2+ ion selective electrode, Cem.
[75] M. Zajac, S.K. Bremseth, M. Whitehead, M. Ben Haha, Effect of CaMg(CO3)2 on Concr. Res. 98 (2017) 36–43.
hydrate assemblages and mechanical properties of hydrated cement pastes at 40 °C [108] A. Jenni, U. Mäder, C. Lerouge, S. Gaboreau, B. Schwyn, In situ interaction be-
and 60 °C, Cem. Concr. Res. 65 (2014) 21–29. tween different concretes and Opalinus clay, Phys. Chem. Earth, 70–71 (2014)
[76] B. Lothenbach, A. Nonat, Calcium silicate hydrates: solid and liquid phase com- 71–83.
position, Cem. Concr. Res. 78 (2015) 57–70. [109] A. Dauzères, G. Achiedo, D. Nied, E. Bernard, S. Alahrache, B. Lothenbach,
[77] I. Richardson, Tobermorite/jennite-and tobermorite/calcium hydroxide-based Magnesium perturbation in low-pH concretes placed in clayey environment—solid
models for the structure of CSH: applicability to hardened pastes of tricalcium characterizations and modeling, Cem. Concr. Res. 79 (2016) 137–150.

[110] D. Bonen, M.D. Cohen, Magnesium sulfate attack on portland cement paste—II. [123] B. Lothenbach, G. Le Saout, E. Gallucci, K. Scrivener, Influence of limestone on the
Chemical and mineralogical analyses, Cem. Concr. Res. 22 (1992) 707–718. hydration of Portland cements, Cem. Concr. Res. 38 (2008) 848–860.
[111] M. Santhanam, M.D. Cohen, J. Olek, Mechanism of sulfate attack: a fresh look: [124] K. De Weerdt, M. Ben Haha, G. Le Saout, K.O. Kjellsen, H. Justnes, B. Lothenbach,
part 1: summary of experimental results, Cem. Concr. Res. 32 (2002) 915–921. Hydration mechanisms of ternary Portland cements containing limestone powder
[112] U.H. Jakobsen, K. De Weerdt, M.R. Geiker, Elemental zonation in marine concrete, and fly ash, Cem. Concr. Res. 41 (2011) 279–291.
Cem. Concr. Res. 85 (2016) 12–27. [125] F. Deschner, F. Winnefeld, B. Lothenbach, S. Seufert, P. Schwesig, S. Dittrich,
[113] C. Roosz, S. Grangeon, P. Blanc, V. Montouillout, B. Lothenbach, P. Henocq, F. Goetz-Neunhoeffer, J. Neubauer, Hydration of a Portland cement with high
E. Giffaut, P. Vieillard, S. Gaboreau, Crystal structure of magnesium silicate hy- replacement by siliceous fly ash, Cem. Concr. Res. 42 (2012) 1389–1400.
drates (M S H): the relation with 2:1 Mg–Si phyllosilicates, Cem. Concr. Res. 73 [126] T.G. Jappy, F.P. Glasser, Synthesis and stability of silica-substituted hydrogarnet
(2015) 228–237. Ca3Al2Si3−xO12−4x(OH)4x, Adv. Cem. Res. 4 (1991) 1–8.
[114] E. Bernard, B. Lothenbach, F. Le Goff, I. Pochard, A. Dauzères, Effect of magne- [127] A. Vollpracht, B. Lothenbach, R. Snellings, J. Haufe, The pore solution of blended
sium on calcium silicate hydrates (C-S-H), Cem. Concr. Res. 97 (2017) 61–72. cements: a review, Mater. Struct. 49 (2016) 3341–3367.
[115] D.R.M. Brew, F.P. Glasser, Synthesis and characterisation of magnesium silicate [128] W.R. Smith, R.W. Missen, Chemical Reaction Equilibrium Analysis: Theory and
hydrate gels, Cem. Concr. Res. 35 (2005) 85–98. Algorithms, Wiley-Interscience, New York 1982. Reprinted With Corrections,
[116] B. Lothenbach, D. Nied, E. L'Hôpital, G. Achiedo, A. Dauzères, Magnesium and Krieger, Malabar, FL, (1991).
calcium silicate hydrates, Cem. Concr. Res. 77 (2015) 60–68. [129] T. Matschei, F.P. Glasser, New approaches to quantification of cement hydration,
[117] W.-S. Chiang, G. Ferraro, E. Fratini, F. Ridi, Y.-Q. Yeh, U. Jeng, S.-H. Chen, in: J. Stark (Ed.), 16 Internationale Baustofftagung (ibausil), Weimar, Germany,
P. Baglioni, Multiscale structure of calcium-and magnesium-silicate-hydrate gels, 2006, pp. 390–400.
J. Mater. Chem. A 2 (2014) 12991–12998. [130] B.J. Merkel, B. Planer-Friederich, Groundwater Geochemistry. A Practical Guide to
[118] E. Bernard, B. Lothenbach, D. Rentsch, I. Pochard, A. Dauzères, Formation of Modeling of Natural and Contaminated Aquatic Systems, Springer Berlin, 2008.
magnesium silicate hydrates (M-S-H), Phys. Chem. Earth 99 (2017) 142–157. [131] G.M. Anderson, D.A. Crerar, Thermodynamics in Geochemistry: The Equilibrium
[119] J.A.T. Smellie, Maqarin Natural Analogue Study: Phase III, SKB, SKB Technical Model, Oxford University Press, Oxford, 1993.
Report 98-04, Stockholm, Sweden, 1998. [132] D. Kulik, Minimising uncertainty induced by temperature extrapolations of ther-
[120] S.J. Chipera, J.A. Apps, Geochemical stability of natural zeolites, Rev. Mineral. modynamic data: a pragmatic view on the integration of thermodynamic data-
Geochem. 45 (2001) 117–161. bases into geochemical computer codes, The Use of Thermodynamic Databases in
[121] R. Arthur, H. Sasamoto, C. Walker, M. Yui, Polymer model of zeolite thermo- Performance Assessment, OECD, Barcelona, 2002, pp. 125–137.
chemical stability, Clays Clay Miner. 59 (2011) 626–639. [133] J.W. Johnson, E.H. Oelkers, H.C. Helgeson, SUPCRT92: a software package for
[122] J.L. Provis, G.C. Lukey, J.S. van Deventer, Do geopolymers actually contain na- calculating the standard molal thermodynamic properties of minerals, gases,
nocrystalline zeolites? A reexamination of existing results, Chem. Mater. 17 (2005) aqueous species, and reactions from 1 to 5000 bar and 0 to 1000 °C, Comput.
3075–3085. Geosci. 18 (1992) 899–947.

Barbara Lothenbach 2019, Cemdata18 - A Chemical Thermodynamic Database For Hydrated Portland Cements and Alkali-Activated Materials

Uploaded by

Copyright:

Available Formats

Barbara Lothenbach 2019, Cemdata18 - A Chemical Thermodynamic Database For Hydrated Portland Cements and Alkali-Activated Materials

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Barbara Lothenbach 2019, Cemdata18 - A Chemical Thermodynamic Database For Hydrated Portland Cements and Alkali-Activated Materials

Uploaded by

Copyright:

Available Formats

&HPHQWDQG&RQFUHWH5HVHDUFK  ²

Contents lists available at ScienceDirect

Cement and Concrete Research

Cemdata18: A chemical thermodynamic database for hydrated Portland 7

Luis Baquerizoe, Belay Dilnesaf, George D. Mironb, Rupert J. Myersg,1

1. Introduction The quality of thermodynamic modelling results depends directly on

M-S-H (solid solution)

Calcium silicate complexes

Fig. 1. Logarithm of the solubility product of Al(OH)3

-22 2.3. Chloride-, nitrate-, and nitrite-AFm phases

Th.+ Calc. Calc.+

monocarbonate, siliceous hydrogarnet can precipitate, as well as form

Ettringite, C6As3H32, is also known to have varying water content.

- The zone of decomposition, which has to be reached in order to

2.6. Mg-Al layered double hydroxide (hydrotalcite-like phase)

Mg-Al layered double hydroxide (LDH) type phases are structurally

to 1 month [72]). We recommend using MgAl-OH-LDH_ss for alkali

2.7. C-S-H solid solution models

MgAl-OH-LDH (ideal ternary solid solution)⁎⁎

Mineral log KS0 Dissolution reactions used to calculate solubility products.

C-S-H (CSH-II solid solution),

C-S-H-K-N (ECSH-1 solid solution)

C-S-H-K-N (ECSH-2 solid solution)

C-S-H (CSHQ solid solution)

C-S-H (CSH3T solid solution)

C-(N-)A-S-H (CNASH solid solution)

log[Ca],di ssol ved

(CaO) x (SiO2 ) y (H2 O)z = ySiO2 + xCa(OH) 2 + (z

H end member) = y(silica) + x(portlandite) + (z

Ettringite M and/or H Monocarbonate

Volume [cm3/100 g unhydrated cement]

Appendix A. Cemdata18 dataset in GEMS format

&HPHQWDQG&RQFUHWH5HVHDUFK  ²

Appendix B. Cemdata18 dataset in PHREEQC format

Product Substance Reaction log10Ko298 ΔrGo298 ΔrHo298 ΔrSo298 ΔrCpo298 ΔrVo298

&HPHQWDQG&RQFUHWH5HVHDUFK  ²

Product Substance Reaction log10Ko298 ΔrGo298 ΔrHo298 ΔrSo298 ΔrCpo298 ΔrVo298

&HPHQWDQG&RQFUHWH5HVHDUFK  ²

e- - represents one electron.

&HPHQWDQG&RQFUHWH5HVHDUFK  ²

&HPHQWDQG&RQFUHWH5HVHDUFK  ²

&HPHQWDQG&RQFUHWH5HVHDUFK  ²

&HPHQWDQG&RQFUHWH5HVHDUFK  ²

* - reaction properties at 20 °C 1 bar.

&HPHQWDQG&RQFUHWH5HVHDUFK  ²

&HPHQWDQG&RQFUHWH5HVHDUFK  ²

Appendix C. Thermodynamic equations and assumptions

&HPHQWDQG&RQFUHWH5HVHDUFK  ²

&HPHQWDQG&RQFUHWH5HVHDUFK  ²

Fe3(OH)4+5 ⁎v −964.33 −1232.44 −472.43 71.30 0 4.1824 0 −3125.33

= 0; SiO32 + Ca2 + > CaSiO30 r G° = r H° = 26.257, r S° = 0, r C° p = 0

&HPHQWDQG&RQFUHWH5HVHDUFK  ²

= 0; 2Fe+3+2 H2 O > Fe2 (OH)2+4 + 2H+ rG

= 0; 3Fe+3 + 4H2 O > Fe3 (OH)4+5 + 4H+ r G = 35.96,

Species ΔG0 ΔH0 S0 Cp0 a0 a1 a2

CH4 −50.66 −74.81 186.26 35.75 23.64 0.0479 −192,464

References (2008) 1–18.

You might also like

&HPHQWDQG&RQFUHWH5HVHDUFK²

&HPHQWDQG&RQFUHWH5HVHDUFK²

&HPHQWDQG&RQFUHWH5HVHDUFK²

&HPHQWDQG&RQFUHWH5HVHDUFK²

&HPHQWDQG&RQFUHWH5HVHDUFK²

&HPHQWDQG&RQFUHWH5HVHDUFK²

&HPHQWDQG&RQFUHWH5HVHDUFK²

&HPHQWDQG&RQFUHWH5HVHDUFK²

&HPHQWDQG&RQFUHWH5HVHDUFK²

&HPHQWDQG&RQFUHWH5HVHDUFK²

&HPHQWDQG&RQFUHWH5HVHDUFK²

&HPHQWDQG&RQFUHWH5HVHDUFK²

&HPHQWDQG&RQFUHWH5HVHDUFK²