See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/321377141

Phase Diagram and Volume Change of the Ca(OH)2 - CaCl2 - H2O System for
Varying Ca(OH)2/CaCl2 Molar Ratios

Article in Journal of Materials in Civil Engineering · February 2018

DOI: 10.1061/(ASCE)MT.1943-5533.0002145

CITATIONS READS

29 2,402

3 authors:

Chunyu Qiao Prannoy Suraneni

Wiss, Janney, Elstner Associates, Inc. University of Miami
42 PUBLICATIONS 1,060 CITATIONS 143 PUBLICATIONS 3,878 CITATIONS

SEE PROFILE SEE PROFILE

William Jason Weiss

Oregon State University
572 PUBLICATIONS 18,758 CITATIONS

SEE PROFILE

All content following this page was uploaded by Chunyu Qiao on 12 December 2017.

The user has requested enhancement of the downloaded file.

 Phase Diagram and Volume Change of the
Ca
OH2─CaCl2─H2O System for Varying
Ca
OH2=CaCl2 Molar Ratios
Chunyu Qiao, Ph.D. 1; Prannoy Suraneni, Ph.D. 2; and Jason Weiss, Ph.D., M.ASCE 3

Abstract: Calcium chloride (CaCl2 ) from deicing salts can react chemically with calcium hydroxide [CaðOHÞ2 ] from hydrated cement paste
Downloaded from ascelibrary.org by CHUNYU QIAO on 11/29/17. Copyright ASCE. For personal use only; all rights reserved.

to form an expansive product, calcium oxychloride, that can damage concrete. This paper uses two experimental techniques, volume change
measurement and low temperature differential scanning calorimetry (LT-DSC), to characterize the phase transitions associated with the for-
mation of calcium oxychloride in the CaðOHÞ2 ─CaCl2 ─H2 O system. The temperature at which calcium oxychloride begins to form increases
as the CaCl2 concentration increases, and it is not influenced by the [CaðOHÞ2 =CaCl2 ] molar ratio. Within the testing temperature range, the
extent of reaction [the mass percentage of CaðOHÞ2 consumed to form calcium oxychloride] increases as the CaCl2 concentration increases;
the extent of reaction decreases as the [CaðOHÞ2 =CaCl2 ] molar ratio increases. The temperatures associated with the phase transitions are
used to develop phase isopleths for the CaðOHÞ2 ─CaCl2 ─H2 O system for varying concentrations and molar ratios. These phase isopleths are
integrated to form a ternary phase diagram of the CaðOHÞ2 ─CaCl2 ─H2 O system. The phase diagram provides a powerful tool to determine
the phase transitions in the CaðOHÞ2 ─CaCl2 ─H2 O system as the [CaðOHÞ2 =CaCl2 ] molar ratio varies. DOI: 10.1061/(ASCE)MT.1943-
5533.0002145. © 2017 American Society of Civil Engineers.
Author keywords: Calcium chloride; Calcium oxychloride; Volume change; Phase diagram.

Introduction phases to form Friedel’s salt (3CaO · Al2 O3 · CaCl2 · 10H2 O)

(Birnin-Yauri and Glasser 1998) and Kuzel’s salt (3CaO · Al2 O3 ·
Deicing salts are widely used on pavement surfaces to melt ice 0.5CaSO4 · 0.5CaCl2 · 10H2 O) (Mesbah et al. 2011). In addition
and snow by lowering the freezing temperature of water (British to these forms of chloride binding, concrete exposed to deicing
Columbia Ministry of Transportation and Highways 1995; Fischel salts such as CaCl2 has an additional form of chloride binding that
2001; Mussato et al. 2004; Yehia and Tuan 1998). Commonly used occurs due to the reaction between calcium hydroxide [CaðOHÞ2 ]
deicing salts are chloride-based [sodium chloride (NaCl), calcium and CaCl2 . This reaction leads to the formation of calcium oxy-
chloride (CaCl2 ), and magnesium chloride (MgCl2 )]. Chloride ions chloride (Brown and Bothe 2004; Collepardi et al. 1994; Shi 2001)
can be transported into the pores of concrete via processes such as as shown in Eq. (1)
diffusion (Collepardi et al. 1972; Di Bella et al. 2012; Page et al.
1981; Xi and Bazant 1999), absorption (Ababneh et al. 2003; 3CaðOHÞ2 þ CaCl2 þ 12H2 O → 3CaðOHÞ2 · CaCl2 · 12H2 O
Nielsen and Geiker 2003; Samson and Marchand 2007), and wick- ð1Þ
ing (Buenfeld et al. 1995; Puyate et al. 1998; Qiao et al. 2017a).
The chloride ions either remain free in the pore solution, or are Although calcium oxychloride also has other forms, depend-
bound physically and chemically into the cementitious matrix ing on humidity and temperature (Demediuk et al. 1955; Shi
(Delagrave et al. 1997), the latter of which helps mitigate the chlo- 2001), it is 3CaðOHÞ2 · CaCl2 · 12H2 O that has been reported most
ride ingress (Glass and Buenfeld 2000; Qiao et al. 2015). frequently to damage concrete (Berntsson and Chandra 1982;
When concrete is exposed to NaCl deicing salt solution, Chatterji 1978). Few studies have measured either the properties of
chloride binding occurs primarily in two forms. One form of bind- the calcium oxychloride or damage caused by calcium oxychloride
ing is the physical adsorption of chloride ions onto the surface directly (Farnam et al. 2015; Monosi and Collepardi 1989, 1990).
of C─S─H (Beaudoin et al. 1990), and the other form of binding The amount of calcium oxychloride that forms can be quantified
is the chemical reaction between chloride ions and the aluminate using low temperature differential scanning calorimetry (LT-DSC)
and it has been shown that this amount can be reduced through
1 the use of supplementary cementitious materials (Suraneni et al.
Postdoctoral Researcher, School of Civil and Construction Engineer-
ing, Oregon State Univ., Corvallis, OR 97331 (corresponding 2016, 2017).
author). ORCID: https://orcid.org/0000-0002-2583-5296. E-mail: qiaoc@ The phase transition associated with calcium oxychloride has
oregonstate.edu also been characterized by measuring the volume change of the
2
Assistant Professor, Dept. of Civil, Architectural, and Environmental CaðOHÞ2 ─CaCl2 ─H2 O system with different CaCl2 solutions at
Engineering, Univ. of Miami, Coral Gables, FL 33146. a fixed [CaðOHÞ2 =CaCl2 ] molar ratio (Qiao et al. 2017b). Calcium
3
Professor, School of Civil and Construction Engineering, Oregon
oxychloride forms at temperatures above the freezing temperature
State Univ., Corvallis, OR 97331.
Note. This manuscript was submitted on December 15, 2016; approved of water, and the temperature at which it forms depends on the con-
on July 31, 2017; published online on November 30, 2017. Discussion centration of the CaCl2 solution (Farnam et al. 2015; Qiao et al.
period open until April 30, 2018; separate discussions must be submitted 2017b). LT-DSC has been used to develop phase isopleths in
for individual papers. This paper is part of the Journal of Materials in Civil the CaðOHÞ2 ─CaCl2 ─H2 O system with varying [CaðOHÞ2 =
Engineering, © ASCE, ISSN 0899-1561. CaCl2 ] molar ratios (Farnam et al. 2015). While the phase transition

© ASCE 04017281-1 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2018, 30(2): 04017281

 Table 1. Mass of CaðOHÞ2 and CaCl2 Solutions in the Volume Measurement Test (in Grams)
[CaðOHÞ2 =CaCl2 ]
molar ratio Materials DI water 5% CaCl2 10% CaCl2 15% CaCl2 20% CaCl2 25% CaCl2 30% CaCl2
0.5 Solution — — 25.59 26.67 27.64 — —
CaðOHÞ2 — — 0.87 1.33 1.83 — —
1.0 Solution 22.78 24.65 25.34 25.95 26.40 27.14 27.48
CaðOHÞ2 3.00 0.86 1.73 2.62 3.56 4.57 5.57
2.0 Solution — — 24.73 24.92 26.84 — —
CaðOHÞ2 — — 3.27 4.93 6.63 — —
3.0 Solution — — 23.87 23.28 25.81 — —
CaðOHÞ2 — — 4.77 7.08 9.33 — —

has been briefly discussed in the CaðOHÞ2 ─CaCl2 ─H2 O system Volume Change Setup
Downloaded from ascelibrary.org by CHUNYU QIAO on 11/29/17. Copyright ASCE. For personal use only; all rights reserved.

with [CaðOHÞ2 =CaCl2 ] molar ratios of 1 and 3 (Farnam et al. A method to monitor the volume change associated with the reaction
2015), more work needs to be done to extend the results for the was developed to describe the calcium oxychloride reaction, and
full range of [CaðOHÞ2 =CaCl2 ] molar ratios. It is important to de- to quantify the extent of reaction (Qiao et al. 2017b). A program-
termine the phase transitions in the CaðOHÞ2 ─CaCl2 ─H2 O system mable water bath (VWR, Radnor, Pennsylvania) was used to control
with varying CaCl2 concentrations and [CaðOHÞ2 =CaCl2 ] molar the temperature of the water bath during the test. Based on the
ratios, which can be used to quantify the formation of calcium oxy- [CaðOHÞ2 =CaCl2 ] molar ratios of 0.5, 1, 2, and 3, the amounts
chloride in a broad range of environmental conditions. of CaðOHÞ2 and CaCl2 solutions were calculated to fit the glass vials
In this paper, the phase transition associated with calcium (i.e., 24 mL, VWR, Radnor, Pennsylvania) and to have a measurable
oxychloride in the CaðOHÞ2 ─CaCl2 ─H2 O system for varying displacement of the red mark in the pipettes, as listed in Table 1.
CaCl2 concentrations and [CaðOHÞ2 =CaCl2 ] molar ratios is char- After filling the glass vials with CaðOHÞ2 and CaCl2 solutions, a
acterized using LT-DSC and volume change measurements. The rubber stopper was placed in the container with a color-coded reus-
development of the ternary phase diagram of the CaðOHÞ2 ─ able glass volumetric pipette embedded in it. The pipette has a
CaCl2 ─H2 O system in the range of 0 to 30% CaCl2 solution capacity of 1 or 2 mL and an accuracy of 0.01 mL (VWR, Radnor,
concentration is also illustrated. Pennsylvania). Care was taken to ensure that no air bubbles were
entrapped in the sealed glass vials. Water-insoluble red liquid dye
was added onto the top of the solution meniscus in the glass pipettes
Materials and Experimental Procedures to enable the volume change measurable. The sample preparation
and mixing was performed at a high temperature (50°C), at which
no calcium oxychloride is expected to form (Farnam et al. 2015).
Materials The sealed glass vials were placed on a sample holder in the
Granular calcium chloride dihydrate (CaCl2 · 2H2 O, EMD water bath at an initial temperature of 50°C. The water bath was
Millipore, reagent grade, Burlington, Massachusetts) was dissolved programmed to go through a cooling (50 to 0°C)–heating (0 to
in deionized (DI) water to prepare CaCl2 solutions with concentra- 50°C) cycle following the specific time and temperature steps de-
tions of 5, 10, 15, 20, 25, and 30% by mass, respectively. Calcium scribed elsewhere (Qiao et al. 2017b). Each step was maintained for
hydroxide powder, CaðOHÞ2 , (Avantor Performance Materials, 8 h after the temperature was stabilized, which enables the volume
reagent grade, Center Valley, Pennsylvania) was mixed with CaCl2 change to approach equilibrium. A digital thermometer (Testo Inc.)
solutions at different [CaðOHÞ2 =CaCl2 ] molar ratios of 0.5, 1, 2, 3, was used to record the temperature in the water bath. A camera was
and 4 for LT-DSC and volume change measurements. used to automatically record the volume of the fluid in the contain-
ers by tracking the location of the red marks in the glass pipettes
every 5 min. Each test required 11 days to complete a cooling-
Experimental Procedures heating cycle. The change in the location of the red marks in
the pipettes is converted to the total volume change in each
CaCl2 ─CaðOHÞ2 ─H2 O system (Fu et al. 2012). The process to
Low Temperature Differential Scanning Calorimetry
determine the volume change associated with the formation of cal-
A LT-DSC instrument (TA Instruments) was used to measure the
cium oxychloride is described in the following sections.
heat flow, which can be used to detect phase transitions in
the CaðOHÞ2 ─CaCl2 ─H2 O system (Monical et al. 2016). The
CaðOHÞ2 powder (0.6–12 mg) was mixed with 15 to 36 mg Results and Discussion
CaCl2 solution with different concentrations (5, 10, 15, 20, 25,
and 30% by mass) to develop [CaðOHÞ2 =CaCl2 ] molar ratios of
0.5, 1, 2, 3, and 4. The mixing was done in high volume stainless Concentration Based Phase Isopleths in the
steel pans, which were then sealed placed in the LT-DSC. LT-DSC Ca
OH2 ─CaCl 2 ─H 2 O System
ran through the following temperature cycle: isothermal at 25°C for Fig. 1 shows the phase isopleth sections of the CaðOHÞ2 ─
one hour; 3°C=min cooling until −90°C; low temperature loop CaCl2 ─H2 O system with different CaCl2 solution concentrations
from −90 to −70°C and back to −90°C at 3°C=min; 0.25°C=min for a [CaðOHÞ2 =CaCl2 ] molar ratio of 1 and 3. These results in
heating until 50°C (Monical et al. 2016). Because undercooling and Fig. 1(a) are similar to the phase isopleth sections reported else-
other effects occur on cooling, they complicate the determination of where (Farnam et al. 2015; Vol’nov and Latysheva 1957) for the
phase transition temperatures on cooling (Scherer 1999; Stillinger system with a [CaðOHÞ2 =CaCl2 ] molar ratio of 1. As the CaCl2
1995); therefore, the temperatures reported in this paper are deter- concentration increases, the temperature at which calcium oxy-
mined based on the heating curve. chloride begins to form increases, and the temperature at which

© ASCE 04017281-2 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2018, 30(2): 04017281

 80
T 1 ¼ −0.02c20 þ 2.10c0 ð0 ≤ c0 ≤ 30Þ ð2Þ
60 Ca(OH)2+Solution (CaCl2+Ca(OH)2+H2O)

T1 where T 1 = calcium oxychloride liquidus temperature (°C); and c0 =

40
initial concentration of CaCl2 solution (% by mass).
T2
Although the [CaðOHÞ2 =CaCl2 ] molar ratio has no influence on
Temperature/ C

20 L1
L2 Oxychloride the temperature at which calcium oxychloride begins to form, it
0 CaCl2 Ca(OH)2 12H2O+ Liquidus
Solution (CaCl2+H2O) Eutectic greatly influences the temperature at which the reaction is com-
-20
CaCl2 Ca(OH)2 12H2O+
pleted because different amounts of calcium oxychloride form in
Ice+Solution (CaCl2+H2O)
L3 systems with different [CaðOHÞ2 =CaCl2 ] molar ratios.
-40
As shown in Fig. 1(a), when the temperature decreases and
-60 L4 passes below T 1 on the calcium oxychloride liquidus line, calcium
CaCl2 Ca(OH)2 12H2O+Ice+CaCl2 6H2O
oxychloride starts to precipitate. The CaCl2 concentration in the
-80
0 5 10 15 20 25 30 35 liquid phase (cL ) at a specific temperature lower than T 1 can be
(a) CaCl2 content/ % w.t.
determined using Eq. (2) because the composition of the liquid
Downloaded from ascelibrary.org by CHUNYU QIAO on 11/29/17. Copyright ASCE. For personal use only; all rights reserved.

changes following the liquidus line, L1. For a molar ratio of

80 ½CaðOHÞ2 =CaCl2  ≤ 3, there is CaCl2 remaining in the liquid phase
with a lower concentration cfinal (% by mass), while the CaðOHÞ2
60
Ca(OH)2+Solution(CaCl2+Ca(OH)2+H2O) would have completely reacted.
40 Based on the stoichiometry in Eq. (1), the value of cfinal can be
L1 calculated. If one assumes that the solution mass is msol (g), the
Temperature/ C

CaCl2 Ca(OH)2 12H2O+

initial mass of CaCl2 salt is c0100
msol
20
Solution (CaCl2+Ca(OH)2+H2O) Oxychloride
(g), when the [CaðOHÞ2 =CaCl2 ]
0
L5
Liquidus molar ratio rm is less than 3, CaðOHÞ2 will be completely con-
Eutectic rm c0 msol
sumed. Because the number of moles of CaðOHÞ2 is 100M
-20 CaCl 2
[M CaCl2 = molar mass of CaCl2 (g=mol)], the consumed moles
-40 CaCl2 Ca(OH)2 12H2O + Ice rm c0 msol rm c0 msol
of CaCl2 and H2 O are 13 100M CaCl
and 4 100M CaCl
, respectively. So
2 2
-60
the value of cfinal is given by
-80
c 1 rm c0 msol
0 5 10 15 20 25 30 35 100 msol − 3 · 100M CaCl2 M CaCl2
(b) CaCl2 content/ % w.t. cfinal ¼ 1 rm c0 msol rm c0 msol · 100
msol − 3 · 100MCaCl2 M CaCl2 − 4 · 100M CaCl
M H2O
2

Fig. 1. Phase isopleth sections of the CaðOHÞ2 ─CaCl2 ─H2 O system 1 − 13 rm

with varying CaCl2 solution concentrations: (a) molar ratio of ¼ M c0 rm < 3 ð3aÞ
CaðOHÞ2 =CaCl2 ¼ 1; (b) molar ratio of CaðOHÞ2 =CaCl2 ¼ 3 1 − ð4rm MCaCl
H2 O
þ 13 rm Þc0 =100
2

ice begins to form decreases. As expected, the eutectic temperature cfinal ¼ 0 rm ≥ 3 ð3bÞ
is constant and independent of the CaCl2 concentration (Askeland
et al. 2011) because the formation of eutectic compounds is a where M H2 O = molar mass (g=mol) of H2 O. Thus, the temperature
phase transition process with zero degree of freedom in the phase T 2 (°C) at which the formation of calcium oxychloride is com-
diagram (Askeland et al. 2011). pleted, represented by the dashed line labeled as L2 in Fig. 1(a),
For the system with the [CaðOHÞ2 =CaCl2 ] molar ratio of 3, as can be expressed as
the CaCl2 concentration increases, the temperature at which cal- T 2 ¼ −0.02cfinal 2 þ 2.10cfinal rm < 3 ð4Þ
cium oxychloride begins to form increases, as shown in Fig. 1(b).
Because both CaCl2 and CaðOHÞ2 should completely react in the For the mixtures with a molar ratio of ½CaðOHÞ2 =CaCl2  ≥ 3,
systems with a [CaðOHÞ2 =CaCl2 ] molar ratio of 3, only water and cfinal equals 0 because there is only water remaining in the liquid
calcium oxychloride should exist after the reaction is complete. phase after CaCl2 in the liquid phase is consumed up by the reac-
Thus, in these systems, the ice liquidus line should be the same tion. Thus, the formation of calcium oxychloride is completed
as that of ice for pure water and should not vary with the initial at 0 °C, and T 2 at rm ≥ 3 coincides with the temperature of ice
CaCl2 concentration [Fig. 1(b)]. formation (0°C).
Note that eutectic compounds are observed in Fig. 1(b), which is Initially in the CaðOHÞ2 ─CaCl2 ─H2 O system, the liquid phase
believed to be a result of the incomplete reaction in Eq. 1. However, has a chemical composition of [CaðOHÞ2 þ CaCl2 þ H2 O] until
in the case of complete reaction, a eutectic melting peak should not the temperature decreases to T 2 . For the system with rm < 3,
be observed. The hypothesis for the discrepancy is that the cooling CaðOHÞ2 will be completely consumed leaving the liquid with a
rate during LT-DSC is not slow enough and the reaction cannot be composition of (CaCl2 þ H2 O) when the temperature is below
completed before the formation of ice. Low temperature differential T 2 . The liquid composition remains constant until the temperature
scanning calorimetry data, not shown in this paper, indicates that decreases to the freezing point. For the liquid phase in these sys-
the eutectic melting peak is quite small. tems, the freezing point is lower than that of water (0°C), and the
By comparing Figs. 1(a and b), it can be observed that the tem- temperature at which ice begins to form is the freezing temperature
perature at which calcium oxychloride begins to form depends on of CaCl2 solution with a concentration of cfinal .
the CaCl2 concentration, while it does not depend on the [CaðOHÞ2 = Based on the conventional CaCl2 ─H2 O binary phase diagram,
CaCl2 ] molar ratio. Thus, the liquidus line [L1 in Figs. 1(a and b)] the ice liquidus line can be fitted using a cubic function in the range
representing the temperature at which calcium oxychloride begins of 0 to 30% CaCl2 by mass (Farnam et al. 2015). Considering
to form can be fitted using the equation as follows: the value of cfinal , the temperature T 3 (°C) on the ice liquidus line

© ASCE 04017281-3 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2018, 30(2): 04017281

 [Line L3 in Fig. 1(a)] in the CaðOHÞ2 ─CaCl2 ─H2 O system, can be T 4 ¼ −52.1 rm < 3 ð6Þ
expressed as (0 ≤ c0 ≤ 30)
In the systems with rm ¼ 3, both CaðOHÞ2 and CaCl2 will be
T 3 ¼ −0.002cfinal 3 þ 0.041cfinal 2 − 0.887cfinal rm < 3 ð5Þ completely consumed, thus the remaining liquid phase will be only
water. In the systems with rm > 3, the formation of calcium oxy-
The eutectic temperature T 4 (°C), represented by Line L4 in chloride will completely deplete the CaCl2 in the liquid phase,
Fig. 1(a), is a constant in the systems with rm < 3 which will have a chemical composition of [CaðOHÞ2 þ H2 O].

0.0 0.0

-0.3 -0.3

Total volume change (ml)

Total volume change (ml)
Downloaded from ascelibrary.org by CHUNYU QIAO on 11/29/17. Copyright ASCE. For personal use only; all rights reserved.

-0.6 -0.6

-0.9 -0.9

-1.2 Cooling-experiment -1.2 Cooling-experiment

Heating-experiment Heating-experiment
Heating-modeling Heating-modeling
-1.5 -1.5
0 10 20 30 40 50 0 10 20 30 40 50
(a) Temperature ( C) (b) Temperature ( C)

0.0 0.0

-0.3 -0.3
Total volume change (ml)
Total volume change (ml)

-0.6 -0.6

-0.9 -0.9

-1.2 Cooling-experiment -1.2 Cooling-experiment

Heating-experiment Heating-experiment
Heating-modeling Heating-modeling
-1.5 -1.5
0 10 20 30 40 50 0 10 20 30 40 50
(c) Temperature ( C) (d) Temperature ( C)

0.0 0.0

-0.3 -0.3 Cooling-experiment

Total volume change (ml)

Total volume change (ml)

Heating-experiment
Heating-modeling
-0.6 -0.6

-0.9 -0.9

-1.2 Cooling-experiment -1.2

Heating-experiment
Heating-modeling
-1.5 -1.5
0 10 20 30 40 50 0 10 20 30 40 50
(e) Temperature ( C) (f) Temperature ( C)

Fig. 2. Experimental and modeled total volume change for CaðOHÞ2 −CaCl2 −H2 O systems with varying CaCl2 concentrations at a molar ratio
of ½CaðOHÞ2 =CaCl2  ¼ 1: (a) 5% CaCl2 ; (b) 10% CaCl2 ; (c) 15% CaCl2 ; (d) 20% CaCl2 ; (e) 25% CaCl2 ; (f) 30% CaCl2

© ASCE 04017281-4 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2018, 30(2): 04017281

 Thus, the liquid phase should freeze at the freezing temperature
of saturated CaðOHÞ2 solution, which has an approximate value
of 0°C. Thus, the ice liquidus temperature T 5 (°C) on Line L5 in
Fig. 1(b) in the systems with rm ≥ 3 can be expressed as

T5 ¼ 0 rm ≥ 3 ð7Þ

As shown in Fig. 1, the calculated phase transition lines in

the phase isopleth sections fit the measured data from LT-DSC
well. This indicates the feasibility and the accuracy of the afore-
mentioned equations to characterize phase transitions in the
CaðOHÞ2 ─CaCl2 ─H2 O system with a CaCl2 concentration in
the range of 0 to 30% by mass.
Downloaded from ascelibrary.org by CHUNYU QIAO on 11/29/17. Copyright ASCE. For personal use only; all rights reserved.

Volume Change in Ca
OH2 ─CaCl 2 ─H 2 O Systems

with Varying CaCl 2 Concentrations
Fig. 3. Illustration of CL and Cs where Ci = initial concentration of
Fig. 2 shows the total volume change of the mixtures with different CaCl2 solution; CL = concentration of the CaCl2 solution on the liqui-
CaCl2 solutions with rm ¼ 1. As the CaCl2 concentration in- dus line L1 at a given temperature below T1; Cs ¼ CaCl2 content in
creases, the total volume change increases and there is a larger drop calcium oxychloride with a value of 33.94%
at a higher temperature due to the formation of calcium oxychlor-
ide. Additionally, there is a hysteresis of volume change during
the cooling-heating cycle. The temperature at which calcium
oxychloride begins to form on cooling is 5–10°C lower than that the case and the solubility varies with temperature]. The cL and cs
on heating, which may be attributable to the undercooling needed are shown in Fig. 3. Thus, the value of x can be calculated
for the nucleation of calcium oxychloride (Scherer 1999; Stillinger c0 − cL
1995). x¼ ð10Þ
cS − cL
For the mixture of CaðOHÞ2 þ 5% CaCl2 solution, calcium
oxychloride is reported to form at a temperature of 5.59°C from Table 2 lists the values of xmax , cfinal , and the temperature
LT-DSC; however no significant volume change was detected that T 2 at which cL reaches cfinal . Based on the molar ratio of
would be associated with the formation of calcium oxychloride CaðOHÞ2 =CaCl2 ¼ 1 and mixture proportions (Table 1), the total
[Fig. 2(a)]. One reason for this inconsistency is that the lowest tem- volume V T (mL) at a specific temperature T can be written as
perature (2.10°C) in the volume change measurement may still be
msol m
higher than the temperature at which calcium oxychloride begins to VT ¼ þ CH T 1 < T < T 0 ð11aÞ
form due to undercooling. ρsol ðci ; TÞ ρCH
When the temperature changes, the total volume change in a  
CaðOHÞ2 ─CaCl2 ─H2 O system is composed of three parts: ð1 − xÞcL msol mCH xc m M
VT ¼ þ − 3 s sol CH
(1) the liquid volume change due to the density change; (2) the ρsol ðcL ; TÞ ρCH MCaCl2 ρCH
liquid volume change due to the reaction to form calcium oxychlor-
xcs msol M oxy
ide; and (3) the solid [CaðOHÞ2 and CaCl2 · 3CaðOHÞ2 · 12H2 O] þ T2 ≤ T < T1 ð11bÞ
volume change. Based on the fitted oxychloride liquidus line M CaCl2 ρoxy
in Fig. 1, it is possible to calculate the total volume change on  
heating. ð1 − xmax Þcfinal
i msol mCH xmax cs msol M CH
VT ¼ þ − 3
At the initial temperature T 0 (51.2°C), the total volume of the ρsol ðcfinal
i ; TÞ ρCH M CaCl2 ρCH
mixture can be expressed as xmax cs msol M oxy
þ T < T2 ð11cÞ
msol m M CaCl2 ρoxy
V ini ¼ þ CH ð8Þ
ρsol ðci ; T 0 Þ ρCH
where mCH and msol = mass of CaðOHÞ2 and CaCl2 solution (g),
where V ini = initial total volume of the mixture (mL); mCH and respectively; M CH and M oxy = molar mass of CaðOHÞ2 and calcium
msol = initial mass (g) of CaðOHÞ2 powder and CaCl2 solution, oxychloride (g=mol), respectively; and ρoxy and ρCH are parameters
respectively; and ρCH and ρsol ðci ; T 0 Þ (g=cm3 ) denote the density that denote the densities of calcium oxychloride and calcium
of CaðOHÞ2 and CaCl2 solution, respectively. hydroxide with the values of 1.805 and 2.21 g=cm3 , respectively
Assuming x portion of the CaCl2 solution is consumed to form (Qiao et al. 2017b). The density of CaCl2 solutions (g=cm3 )
calcium oxychloride at a temperature lower than T 1 , the mass bal-
ance for CaCl2 can be written as
Table 2. Values of xmax , cfinal and T 2
ð1 − xÞ · msol · cL þ x · msol · cS ¼ msol · c0 ð9Þ
CaCl2 content/% by mass xmax cfinal /% by mass T 2 =°C
where cL = concentration of the CaCl2 solution on the liquidus line, 5 0.049 3.51 7.1
Line L1, at a given temperature below T1; and cs ¼ CaCl2 content 10 0.098 7.39 14.4
in calcium oxychloride with a value of 33.94% by mass [the solu- 15 0.147 11.73 21.9
bility of calcium oxychloride is assumed to be constant over a range 20 0.196 16.59 29.3
of temperatures (0–50°C) to simplify the calculations. While this 25 0.246 22.09 36.6
30 0.290 27.69 43.4
is a reasonable approximation, strictly speaking this is not exactly

© ASCE 04017281-5 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2018, 30(2): 04017281

 ρsol ðci ; TÞ as a function of temperature and concentration is known change associated with the formation of calcium oxychloride is cal-
(Qiao et al. 2017b). Thus, the total volume change ΔV (mL) can be culated by subtracting the volume change of the liquid due to the
calculated as density change from the total volume change based on the method
described elsewhere (Qiao et al. 2017b). There is a larger hysteresis
ΔV ¼ V T − V ini ð12Þ of volume change in the system with a higher [CaðOHÞ2 =CaCl2 ]
molar ratio. As the [CaðOHÞ2 =CaCl2 ] molar ratio increases, the
Fig. 2 shows the modeling results, which compare well to the value of cfinal decreases based on Eq. (3). A lower temperature
experimental data. For the mixtures with 5 and 10% CaCl2 solution T 2 is needed to achieve a lower cfinal according to Eq. (4). Thus,
[Figs. 2(a and b)], the inconsistency between the experimental and a larger temperature range (T 1 –T 2 ) is needed on the calcium oxy-
modeling result may be because of the incomplete formation of chloride liquidus line (Line L1 in Fig. 1) to complete the reaction.
calcium oxychloride on cooling (due to undercooling). As shown in Fig. 5, the [CaðOHÞ2 =CaCl2 ] molar ratio has little
Eq. (11a) illustrates that the total volume of the system includes influence on the temperatures at which calcium oxychloride begins
the volume of liquid (the first term on the right side) and the volume to form on both cooling and heating. This is in accordance with the
of the CaðOHÞ2 (the second term on the right side) at a given tem- observations in Fig. 1.
perature above T 1 . Eqs. (11b) and (11c) illustrate that the volume of
Downloaded from ascelibrary.org by CHUNYU QIAO on 11/29/17. Copyright ASCE. For personal use only; all rights reserved.

Figs. 5(a and b) show that the maximum normalized

liquid (the first term on the right side), the volume of the CaðOHÞ2 volume change is similar to the theoretical maximum value of
(the middle two terms on the right side), and the volume of formed −0.184 mL=g CaðOHÞ2 (Qiao et al. 2017a) in the systems with
calcium oxychloride (the last term on the right side) contribute to [CaðOHÞ2 =CaCl2 ] molar ratios of 0.5 and 1, indicating the com-
the total volume at a given temperature below T 1. Thus, based on plete phase transition associated with calcium oxychloride. The
the calculated total volume change of the CaðOHÞ2 ─CaCl2 ─H2 O incomplete hysteresis and the smaller values of the maximum
system in Eq. (11), it is possible to quantify the partial volume of normalized volume change [Figs. 5(c and d)] indicate incomplete
each phase on heating, as shown in Fig. 4. As the temperature de- phase transition associated with calcium oxychloride in the systems
creases, the total volume decreases; however, the partial volume of with the [CaðOHÞ2 =CaCl2 ] molar ratios of 2 and 3.
solid phases dramatically increases due to the formation of calcium
oxychloride. This increase in volume caused by the crystallization
of calcium oxychloride could potentially explain damage observed Extent of Reaction
in concrete exposed to CaCl2 solutions. This mode of damage due The extent of reaction with respect to the amount of CaðOHÞ2 can
to calcium oxychloride formation has been tentatively suggested be computed based on the normalized volume change correspond-
elsewhere (Chatterji 1978). ing to the phase transition associated with calcium oxychloride
Note that the composition of the liquid phase changes from a (Qiao et al. 2017b). Another method to calculate the extent of
mixed solution of [CaCl2 þ CaðOHÞ2 ] to a CaCl2 solution when reaction was developed based on normalizing the heat released
the temperature is reduced below the temperature at which in the calcium oxychloride phase transition from LT-DSC with that
CaðOHÞ2 completely reacts, as illustrated in Fig. 4. This change from the heat release for pure calcium oxychloride (Monical et al.
in the composition of the liquid phase happens in systems with 2016). Fig. 6 shows that the extents of reaction determined by
a molar ratio of ½CaðOHÞ2 =CaCl2  < 3. both methods are comparable although some variations exist. This
indicates the feasibility of the volume change measurement to
quantify the phase transition associated with calcium oxychloride.
Volume Change in Ca
OH2 ─CaCl 2 ─H 2 O Systems In the mixtures containing CaCl2 solutions with high concen-
with Varying [Ca
OH2 =CaCl 2 ] Molar Ratios trations (20, 25, and 30% by mass), the reaction to form calcium
Fig. 5 shows the normalized volume change with respect to the oxychloride is almost complete, as shown in Fig. 6. However, at
amount of CaðOHÞ2 in the mixtures with 20% CaCl2 solution at lower concentrations (5 and 10% CaCl2 by mass), the reaction
different [CaðOHÞ2 =CaCl2 ] molar ratios. The normalized volume is incomplete using both methods, and the reaction extents from
LT-DSC are higher than those from the volume change measure-
ment. This may be because more calcium oxychloride forms be-
tween the lowest temperature (2.1°C) in the volume change
25.0 measurement and that in the LT-DSC test (−90°C).
MCa(OH)2 : MCaCl2 =1 The extent of reaction is plotted against the molar ratio of
CaðOHÞ2 =CaCl2 in Fig. 7. As the molar ratio of CaðOHÞ2 =CaCl2
22.5 increases, the mixture needs a larger temperature range (on Line L1
in Fig. 1) to complete the formation of calcium oxychloride. How-
Total volume (ml)

CaCl2 solution CaCl2+Ca(OH)2 solution

ever, as calcium oxychloride begins to form at the same tempera-
20.0 ture on cooling, the temperature range from this point to the lower
5.0 limit of the cooling-heating cycle decreases as the molar ratio in-
creases. Thus, a decreasing extent of reaction is expected as the
molar ratio of CaðOHÞ2 =CaCl2 increases, as shown in Fig. 7.
2.5 Caclium oxychloride

Ca(OH)2 Molar Ratio Based Phase Isopleths in the

0.0 Ca
OH2 ─CaCl 2 ─H 2 O System
0 10 20 30 40 50 The effect of the CaCl2 concentration and temperature on the differ-
Temperature ( C) ent phases that form and the temperature and CaCl2 concentration
ranges in which they form is indicated clearly in Figs. 1 and 4.
Fig. 4. Partial volume of different phases in a CaðOHÞ2 ─CaCl2 ─H2 O
This data may be analyzed in another manner, namely the effect
system {molar ratio of [CaðOHÞ2 =CaCl2  ¼ 1, 20% CaCl2 solution}
of the molar ratio on the different phases that form. Because a

© ASCE 04017281-6 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2018, 30(2): 04017281

 0.00 0.00

Normalized volume change(ml/g CH)

Normalized volume change(ml/g CH)
Heating Heating
-0.06 -0.06

-0.12 Cooling
-0.12
Cooling

-0.18 -0.18

-0.24 Cooling-CH+20% CaCl2 -0.24 Cooling-CH+20% CaCl2

Heating-CH+20% CaCl2 Heating-CH+20% CaCl2
Heating-modeling Heating-modeling
-0.30 -0.30
Downloaded from ascelibrary.org by CHUNYU QIAO on 11/29/17. Copyright ASCE. For personal use only; all rights reserved.

0 10 20 30 40 50 0 10 20 30 40 50
(a) Temperature( C) (b) Temperature( C)

0.00 0.00
Normalized volume change(ml/g CH)

Normalized volume change(ml/g CH)

Heating Heating
-0.06 -0.06
Cooling
Cooling
-0.12 -0.12

-0.18 -0.18

-0.24 Cooling-CH+20% CaCl2 -0.24 Cooling-CH+20% CaCl2

Heating-CH+20% CaCl2 Heating-CH+20% CaCl2
Heating-modeling Heating-modeling
-0.30 -0.30
0 10 20 30 40 50 0 10 20 30 40 50
(c) Temperature( C) (d) Temperature( C)

Fig. 5. Normalized volume change for CaðOHÞ2 ─CaCl2 ─H2 O systems with 20% CaCl2 solution at different molar ratios of [CaðOHÞ2 =CaCl2 ]: 0.5,
1, 2, and 3

120
10% CaCl2
100
15% CaCl2
Volume change 100
LT-DSC 20% CaCl2
80
Extent of reaction

80
Extent of reaction

60
60

40
40

20
20

0
0 5 10 15 20 25 30 0
CaCl2 content/ % w.t. 0.5 1 2 3
Molar ratio of Ca(OH)2:CaCl2
Fig. 6. Comparison of reaction extent measured by the volume
change measurement and LT-DSC test [the molar ratio of CaðOHÞ2 = Fig. 7. Extent of reaction measured from volume change measure-
CaCl2 ¼ 1] ments in various CaðOHÞ2 ─CaCl2 ─H2 O systems

© ASCE 04017281-7 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2018, 30(2): 04017281

 80
by four lines: (1) the calcium oxychloride liquidus line, L1, repre-
60
Ca(OH)2+Solution(CaCl2+Ca(OH)2+H2O)
senting the temperature at which calcium oxychloride begins to
40
form; (2) Line L2 representing the temperature at which the forma-
L1
tion of calcium oxychloride finishes; (3) the ice liquidus line, L3,
Temperature/ C

Ca(OH)2+Ca-Oxy+Solution Oxychlolride
20 L2
starts forming
representing the temperature at which ice begins to form; and
(CaCl2+Ca(OH)2+H2O)
0
Ca-Oxy+Solution(CaCl2+H2O) Oxychloride (4) the eutectic line, Line L4. In the isopleth section with a molar
L3
L5 finishes forming
Liquidus
ratio of ½CaðOHÞ2 =CaCl2  ≥ 3, CaCl2 would be completely con-
-20
Ca-Oxy+Ice+Solution(CaCl2+H2O)
Eutectic sumed and thus there are three areas divided by two lines: (1) the
Ca(OH)2+
-40 Ca-Oxy +
calcium oxychloride liquidus line, Line L1; and (2) the ice liquidus
L4
Ice line, Line L5, representing the ice formation temperature in saturated
-60
Ca-Oxy+Ice+CaCl2 6H2O CaðOHÞ2 solution.
-80 The data from LT-DSC is also plotted in Fig. 8, and it fits the
0 1 2 3 4
theoretically constructed phase isopleths. For the section with the
(a) Molar ratio of Ca(OH)2/CaCl2
molar ratio of ½CaðOHÞ2 =CaCl2  ≥ 3, peaks at the eutectic temper-
Downloaded from ascelibrary.org by CHUNYU QIAO on 11/29/17. Copyright ASCE. For personal use only; all rights reserved.

80 ature are still observed due to the incomplete reaction during

LT-DSC tests. The temperature at which the formation of calcium
60 Ca(OH)2+Solution(CaCl2+Ca(OH)2+H2O)
oxychloride finishes is higher than the theoretical value in some
L1
40 cases.
Temperature/ C

Ca(OH)2+Ca-Oxy+Solution
20 Oxychloride
L2 (CaCl2+Ca(OH)2+H2O) starts forming
0
Ca-Oxy+Solution(CaCl2+H2O) Oxychloride Phase Diagram of Ca
OH2 ─CaCl 2 ─H 2 O System
L5 finishes forming
L3 Liquidus Figs. 9(a and b) show the locations of different mixtures in the
-20
Eutectic
Ca-Oxy+Ice+Solution(CaCl2+H2O) Ca(OH)2+ CaðOHÞ2 ─CaCl2 ─H2 O ternary phase diagram. The points repre-
-40 Ca-Oxy +
Ice
senting the mixtures with varying CaCl2 solutions {at a fixed molar
-60 L4
Ca-Oxy+Ice+CaCl2 6H2O
ratio of [CaðOHÞ2 =CaCl2 ] are located on the line starting from the
-80 vertex for H2 O [Fig. 9(a)]}, while the points representing the mix-
0 1 2 3 4 tures with varying molar ratios of [CaðOHÞ2 =CaCl2 ] are located on
(b) Molar ratio of Ca(OH)2/CaCl2 the line starting from the vertex for CaðOHÞ2 [Fig. 9(b)]. Based on
the obtained phase isopleths from the both directions, it is possible
Fig. 8. Phase isopleth section of the CaðOHÞ2 ─CaCl2 ─H2 O system to construct the partial ternary phase diagram in the range of 0–30%
with varying molar ratios of [CaðOHÞ2 =CaCl2 ]: (a) 15% CaCl2 CaCl2 concentration.
solution; (b) 20% CaCl2 solution At the given initial CaCl2 solution c0 and mixture proportion
(mCH and msol ), the location of the mixture (x, y, z, x þ y þ z ¼ 1)
in the ternary system can be computed
master calcium oxychloride liquidus line has already been de- c0 msol =100
scribed [Eq. (2)], it is possible to construct the phase isopleths x¼ ð13aÞ
mCH þ msol
of the CaðOHÞ2 ─CaCl2 ─H2 O system with different molar ratios
of CaðOHÞ2 =CaCl2 , as shown in Fig. 8. The stoichiometry in mCH
Eq. (1) separates the isopleth into two sections: the section with y¼ ð13bÞ
mCH þ msol
a molar ratio of ½CaðOHÞ2 =CaCl2  < 3 and the other with a molar
ratio of ½CaðOHÞ2 =CaCl2  ≥ 3.
ð1 − c0 =100Þmsol
In the isopleth section with a molar ratio of ½CaðOHÞ2 =CaCl2  < 3 z¼ ð13cÞ
(Fig. 8) there are five areas as the temperature decreases divided mCH þ msol

0.00 0.00
1.00 1.00

0.25 0.25
0.75 0.75
Ca

Ca
O

O
(O
2

(O
H

2
H

0.50 0.50
H) 2

H) 2

0.50 5% CaCl2 0.50

10% CaCl2
0.75 Molar ratio=0.5 0.75 15% CaCl2
Molar ratio=1.0 0.25 0.25
20% CaCl2
Molar ratio=2.0
Molar ratio=3.0 25% CaCl2
1.00 Molar ratio=4.0 1.00 30% CaCl2
0.00 0.00
0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00
(a) CaCl2 (b) CaCl2

Fig. 9. Location of the CaðOHÞ2 ─CaCl2 ─H2 O systems with (a) the same molar ratios and (b) the same CaCl2 concentrations in the ternary phase
diagram, respectively (note that no phases are traced out in the gray regions)

© ASCE 04017281-8 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2018, 30(2): 04017281

 0.0 0.0
1.0 1.0

0.2 0.2

Ca

Ca
0.8 0.8

)
(O

(O
i on

i on
H) 2

H) 2
or t

or t
0.4 0.4

r op

r op
(

(
ma

ma
0.6 0.6

p
ss

ss
ss

ss
pr o

pro
ma

ma
0.6 0.6

po

po
O(

O(
44.5 C 0.4 0 C 0.4

r tio

r tio
2

2
H

H
40.0 C 40 C

n)

n)
0.8 30.0 C 0.8 30 C
20.0 C 0.2 20 C 0.2
10.0 C 10 C
1.0 0.0 C 1.0 0 C
0.0 0.0
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
Downloaded from ascelibrary.org by CHUNYU QIAO on 11/29/17. Copyright ASCE. For personal use only; all rights reserved.

(a) CaCl2 (mass proportion) (b) CaCl2 (mass proportion)

0.0 0.0
1.0 1.0

0.2 Ca 0.2

Ca
0.8 0.8

n)
n)

(O

(O
r t io
r t io

H) 2

H) 2
po
po

0.4 0.4
(m

(m
pr o
pr o

0.6 0.6
a ss

as s
ss
ss

p ro

p ro
ma
ma

0.6 0.6
po

po
O(
O(

0 C 0.4 0.4
r ti o

r ti o
2
2

H
H

-10 C
n)

n)
0.8 -20 C 0.8
-30 C 0.2 0.2
-40 C
1.0 1.0 -51.1 C
-50 C
0.0 0.0
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
(c) CaCl2 (mass proportion) (d) CaCl2 (mass proportion)

Fig. 10. Contour maps of the CaðOHÞ2 ─CaCl2 ─H2 O ternary system: (a) contour lines for temperature at which calcium oxychloride begins to form;
(b) counter lines for temperature at which calcium oxychloride finishes forming; (c) counter lines for temperature for the formation of ice; (d) counter
lines for temperature for the formation of eutectic compounds; the star represents the example case

The molar ratio of [CaðOHÞ2 =CaCl2 ] rm can be determined solution. If the transport of the solution into the concrete is ignored
using the equation and it is assumed that the concrete is fully saturated with the CaCl2
solution, mCH and msol can be calculated to be 92 kg=m3 and
mCH M CaCl2
rm ¼ · ð14Þ 164.92 kg=m3 . Based on Eq. (13), the system can then be located
c0 msol =100 MCH at (0.128, 0.358, and 0.514) (the mass portions of CaCl2 , CaðOHÞ2 ,
With the initial concentration c0 and the molar ratio rm , and H2 O) in Fig. 10. When the point for this system is plotted in
Eqs. (2)–(7) can be used to calculate the temperature associated with Fig. 10, the temperatures for phase transition can be determined.
the phase transitions in the ternary phase diagram. Moreover, the As the temperature decreases, theoretically, calcium oxychloride
mixtures with the same phase transition temperature can be linked would start forming at 34°C and finish forming at 0°C, ice would
to plot the contour line map in the phase diagram. Figs. 10(a–d) form at 0°C, and there would be no eutectic compounds.
illustrate the contour maps representing the temperature at which
calcium oxychloride begins to form (L1), the temperature at which
its formation completes (L2), the temperature at which ice begins Conclusions
to form (L3 and L5), and the temperature at which the eutectic
compounds forms (L4), respectively. Note that the phase diagram This paper examined the phase transitions associated with the
in the grey region (Figs. 9 and 10) was not determined in this study formation of calcium oxychloride in the CaðOHÞ2 ─CaCl2 ─H2 O
because the CaCl2 concentration discussed in this paper is within system. These transitions were studied using LT-DSC and volume
30% by mass. The contour maps in Fig. 10 provide a powerful tool change measurements.
to determine the temperature range associated with the phase tran- The temperature at which calcium oxychloride begins to form
sition of calcium oxychloride, which could be useful in predicting increases as the CaCl2 concentration increases; however, it is
the damage to concrete caused by calcium oxychloride under a not affected by the [CaðOHÞ2 =CaCl2 ] molar ratio. After the reac-
wide range of environmental conditions. tion to form calcium oxychloride completes, the remaining solution
As an example, it is assumed that a normal weight plain concrete has a chemical composition of either CaCl2 þ CaðOHÞ2 þ H2 O,
(ρconcrete ¼ 2;300 kg=m3 ) with a CaðOHÞ2 amount of 4% and a CaðOHÞ2 þ H2 O, or CaCl2 þ H2 O, depending on the molar ratio
porosity of 14% is exposed to 20% CaCl2 (ρsol ¼ 1;178 kg=m3 ) and temperature.

© ASCE 04017281-9 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2018, 30(2): 04017281

 The phase transition lines in the CaðOHÞ2 ─CaCl2 ─H2 O system Collepardi, M., Marcialis, A., and Turriziani, R. (1972). “Penetration of
for varying CaCl2 concentrations and [CaðOHÞ2 =CaCl2 ] molar ra- chloride ions into cement pastes and concretes.” J. Am. Ceram. Soc.,
tios are based on the LT-DSC data (determined from the heating 55(10), 534–535.
cycle). The phase isopleth from LT-DSC can be used to compute Delagrave, A., Marchand, J., Ollivier, J. P., Julien, S., and Hazrati, K.
the total volume change on heating. The extent of reaction from (1997). “Chloride binding capacity of various hydrated cement paste
systems.” Adv. Cem. Based Mater., 6(1), 28–35.
the volume change measurement is comparable to that from
Demediuk, T., Cole, W., and Hueber, H. V. (1955). “Studies on
LT-DSC; it increases with CaCl2 concentration and decreases as
magnesium and calcium oxychlorides.” Aust. J. Chem., 8(2),
the [CaðOHÞ2 =CaCl2 ] molar ratio increases. 215–233.
Based on the data, the ternary phase diagram of the Di Bella, C., Villani, C., Hausheer, E., and Jason, W. (2012). “Chloride
CaðOHÞ2 ─CaCl2 ─H2 O system was quantitatively constructed in transport measurements for a plain and internally cured concrete
the CaCl2 concentration range of 0 to 30% by mass. The phase dia- mixture.” ACI Spec. Publ., 290, 1–16.
gram can provide contour line maps representing the temperatures Farnam, Y., Dick, S., Wiese, A., Davis, J., Bentz, D., and Weiss, J. (2015).
associated with the phase transitions in the CaðOHÞ2 ─CaCl2 ─H2 O “The influence of calcium chloride deicing salt on phase changes and
system. This provides a powerful tool to determine the phases damage development in cementitious materials.” Cem. Concr. Compos.,
Downloaded from ascelibrary.org by CHUNYU QIAO on 11/29/17. Copyright ASCE. For personal use only; all rights reserved.

that form for any solution concentration, molar ratio, and tempera- 64, 1–15.
ture for the CaðOHÞ2 ─CaCl2 ─H2 O system. Fischel, M. (2001). “Evaluation of selected deicers based on a review of the
literature.” CDOT-DTD-R-2001-15, Colorado Dept. of Transportation,
Denver, 9–13.
Acknowledgments Fu, T., Deboodt, T., and Ideker, J. H. (2012). “Simple procedure for deter-
mining long-term chemical shrinkage for cementitious systems using
The authors gratefully acknowledge the financial support from improved standard chemical shrinkage test.” J. Mater. Civ. Eng.,
the National Ready Mix Concrete Association (NRMCA), the 10.1061/(ASCE)MT.1943-5533.0000486, 989–995.
Portland Cement Association (PCA), and a pooled fund by Glass, G. K., and Buenfeld, N. R. (2000). “The influence of chloride
the Oklahoma Department of Transportation [TP-5(297)] binding on the chloride induced corrosion risk in reinforced concrete.”
“Improving Specifications to Resist Frost Damage in Modern Corros. Sci., 42(2), 329–344.
Mesbah, A., et al. (2011). “Crystal structure of Kuzel’s salt 3CaO · Al2 O3 ·
Concrete.” The authors also acknowledge insightful discussions
1=2CaSO4 · 1=2CaCl2 · 11H2 O determined by synchrotron powder
with Dr. Vahid Jafari Azad and Professor Burkan Isgor at Oregon diffraction.” Cem. Concr. Res., 41(5), 504–509.
State University. The contents of this paper reflect the perspec- Monical, J., Unal, E., Barrett, T., Farnam, Y., and Weiss, W. J. (2016).
tives of the authors, who are responsible for the facts and the “Reducing joint damage in concrete pavements.” Transp. Res. Rec.,
accuracy of the data presented herein. The contents do not ne- 2577, 17–24.
cessarily reflect the official views or policies of a Department of Monosi, S., and Collepardi, L. (1989). “Chemical attack of calcium
Transportation or the Federal Highway Administration at the time chloride on the portland cement paste.” Cemento, 86(2),
of publication. 97–104.
Monosi, S., and Collepardi, M. (1990). “Research on 3CaO · CaCl2 ·
15H2 O identified in concretes damaged by CaCl2 attack.” Cemento,
References 87(1), 3–8.
Mussato, B. T., Gepraegs, O. K., and Farnden, G. (2004). “Relative effects
Ababneh, A., Benboudjema, F., and Xi, Y. P. (2003). “Chloride penetration of sodium chloride and magnesium chloride on reinforced concrete:
in nonsaturated concrete.” J. Mater. Civ. Eng., 10.1061/(ASCE)0899- State of the art.” Transp. Res. Rec., 1866, 59–66.
1561(2003)15:2(183), 183–191. Nielsen, E. P., and Geiker, M. R. (2003). “Chloride diffusion in
Askeland, D. R., Fulay, P. P., and Wright, W. J. (2011). The science and partially saturated cementitious material.” Cem. Concr. Res., 33(1),
engineering of materials, Cengage Learning, Stamford, CT. 133–138.
Beaudoin, J. J., Ramachandran, V. S., and Feldman, R. F. (1990). Page, C. L., Short, N. R., and El Tarras, A. (1981). “Diffusion of
“Interaction of chloride and C─S─H.” Cem. Concr. Res., 20(6), chloride ions in hardened cement pastes.” Cem. Concr. Res., 11(3),
875–883. 395–406.
Berntsson, L., and Chandra, S. (1982). “Damage of concrete sleepers by
Puyate, Y. T., Lawrence, C. J., Buenfeld, N. R., and McLoughlin, I. M.
calcium chloride.” Cem. Concr. Res., 12(1), 87–92.
(1998). “Chloride transport models for wick action in concrete at large
Birnin-Yauri, U. A., and Glasser, F. P. (1998). “Friedel’s salt,
Peclet number.” Phys. Fluids, 10(3), 566–575.
Ca2 AlðOHÞ6 ðCl; OHÞ · 2H2 O: Its solid solutions and their role in
Qiao, C., Ni, W., and Weiss, J. (2015). “Influence of shrinkage reducing
chloride binding.” Cem. Concr. Res., 28(12), 1713–1723.
admixtures on chloride binding and diffusion.” Proc., 14th Int.
British Columbia Ministry of Transportation and Highways. (1995).
Congress on the Chemistry of Cement, China National Building
“Round VI (1995-6) road and bridge maintenance contracts, mainte-
nance services manual, standards for road and bridge maintenance Materials Group Corporation, Beijing.
services.” Victoria, BC, Canada. Qiao, C., Ni, W., and Weiss, J. (2017a). “Transport due to diffusion,
Brown, P., and Bothe, J., Jr. (2004). “The system CaO─Al2 O3 ─CaCl2 ─H2 O drying, and wicking in concrete containing a shrinkage reducing
at 23  2°C and the mechanisms of chloride binding in concrete.” admixture.” J. Mater. Civ. Eng., 10.1061/(ASCE)MT.1943-5533
Cem. Concr. Res., 34(9), 1549–1553. .0001983, 04017146.
Buenfeld, N. R., Shurafa-Daoudi, M.-T., and McLoughlin, I. M. (1995). Qiao, C., Suraneni, P., and Weiss, J. (2017b). “Measuring volume
“Chloride transport due to wick action in concrete.” 1st RILEM change due to calcium oxychloride phase transformation in A
Workshop on Chloride Penetration into Concrete, L. O. Nilsson CaðOHÞ2 ─CaCl2 ─H2 O system.” Adv. Civ. Eng. Mater., 6(1),
and J. P. Ollivier, eds., RILEM, Saint-Remy-les-Chevreuse, France, 157–169.
315–324. Samson, E., and Marchand, J. (2007). “Modeling the transport of ions in
Chatterji, S. (1978). “Mechanism of the CaCl2 attack on portland cement unsaturated cement-based materials.” Comput. Struct., 85(23–24),
concrete.” Cem. Concr. Res., 8(4), 461–467. 1740–1756.
Collepardi, M., Coppola, L., and Pistolesi, C. (1994). “Durability of Scherer, G. W. (1999). “Crystallization in pores.” Cem. Concr. Res., 29(8),
concrete structures exposed to CaCl2 based deicing salts.” Proc., 3rd 1347–1358.
CANMENT/ACI Int. Conf., American Concrete Institute, Farmington Shi, C. (2001). “Formation and stability of 3CaO · CaCl2 · 12H2 O.” Cem.
Hills, MI, 107–120. Concr. Res., 31(9), 1373–1375.

© ASCE 04017281-10 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2018, 30(2): 04017281

 Stillinger, F. H. (1995). “A topographic view of supercooled liquids and Vol’nov, I. I., and Latysheva, E. I. (1957). “Separation of calcium chloride
glass formation.” Science, 267(5206), 1935–1939. from solvay spent liquor through calcium hydroxichloride.” J. Appl.
Suraneni, P., Azad, V. J., Isgor, O. B., and Weiss, J. (2016). “Deicing Chem. USSR, 30, 1039–1046.
salts and durability of concrete pavements and joints.” Concr. Int., Xi, Y., and Bazant, Z. P. (1999). “Modeling chloride penetration in satu-
38(4), 48–54. rated concrete.” J. Mater. Civ. Eng., 10.1061/(ASCE)0899-1561(1999)
Suraneni, P., Monical, J., Unal, E., Farnam, Y., and Weiss, J. (2017). 11:1(58), 58–65.
“Calcium oxychloride formation potential in cementitious pastes Yehia, S., and Tuan, C. Y. (1998). “Bridge deck deicing.” Transportation
exposed to blends of deicing salt.” ACI Mater. J., 114(4), 631–641. Conf. Proc., Iowa State Univ., Ames, IA, 51–57.
Downloaded from ascelibrary.org by CHUNYU QIAO on 11/29/17. Copyright ASCE. For personal use only; all rights reserved.

© ASCE 04017281-11 J. Mater. Civ. Eng.

View publication stats J. Mater. Civ. Eng., 2018, 30(2): 04017281