Review

Cement and Concrete Composites 121 (2021) 104100
Contents lists available at ScienceDirect
Cement and Concrete Composites

journal homepage: www.elsevier.com/locate/cemconcomp
Development of sustainable concrete incorporating seawater: A critical

review on cement hydration, microstructure and mechanical strength
Peiran Li a, Wengui Li a, *, Zhihui Sun b, Luming Shen c, Daichao Sheng a
a
School of Civil and Environmental Engineering, University of Technology Sydney, NSW, 2007, Australia
b
Department of Civil and Environmental Engineering, University of Louisville, Louisville, KY, 40292, USA
c
School of Civil Engineering, The University of Sydney, NSW, 2006, Australia
A R T I C L E I N F O A B S T R A C T
Keywords: Many countries are experiencing freshwater crises due to the increasing growth of the population together with
Seawater the infrastructure construction that is aligned with the needs of freshwater for concrete production. There are
Chloride ion also deficiencies in freshwater in many coastal areas where seawater is more accessible. To reduce unnecessary
Cementitious concrete
resource-wasting and meanwhile drive sustainable development in the construction industry, great efforts have
Hydration
Corrosion
been made to utilize seawater as the alternative mixing water for concrete casting, which presents potential
Sustainability economical and environmental benefits in the coastal and island regions. This paper comprehensively reviews
the current studies on the predominant performance differences between seawater-mixed and conventional
concretes with freshwater. Particular attention is paid to the chloride-induced hydration mechanism due to the
chloride ions in seawater. The main findings of this review reveal that although harmful ingredients in seawater
may weaken some of the concrete performances, applying proper curing conditions and adding moderate ad
ditives and admixtures could significantly and effectively mitigate these defects in properties. However, the
unstable chloride binding ability in cement hydrates cannot eliminate the risk of rebar corrosion caused by
chlorides in seawater, resulting in a limited scope of practical application. Finally, some trade-offs are recom
mended in using seawater in concrete, suggesting prospects of applications in the future construction industry.
This study guides for the safer use of seawater in sustainable concrete through reviewing the advanced research
progress.
1. Introduction significantly. In this case, utilizing seawater as an alternative to fresh

water for concrete mixing is recommended. While in practice, the
With the world hurtling towards the urban future accompanying presence of chloride ions in seawater can render serious corrosion in
increased demand for infrastructure, concrete production has experi reinforced concrete and subsequently cause deterioration, undoubtedly
enced unprecedented large-scale popularity worldwide [1,2]. Mean limiting the practical implementation severely. Therefore, appropriate
while, the tremendous amount of concrete production results in high measures have to be taken to alleviate negative effects caused by
greenhouse gas emissions [3], and great consumption of freshwater, seawater used in concrete mixtures.
mineral, and energy [4]. It has been reported that in 2012, over 2 gig Although studies in this area have been elaborated and progressed
atons of freshwater was consumed only used as mixing water in concrete greatly, divergent opinions still exist against using seawater in concrete
production [5], while there could be 10 times more freshwater involved [7]. Otsuki [8] has proposed that the negative effect induced by
during the whole manufacturing process if the water for cooling, seawater mixing decreases with age. Therefore a comparable long-term
washing, energy, and transportation is considered [6]. The unsustain performance can be presented in seawater concrete [9]. While Neville
ability induced by such considerable resource consumption inevitably [10] pointed out using seawater could cause a significant safety issue
drives the revolution of the concrete industry. In some island and coastal that should not be ignored. Meanwhile, it should be noted that chloride
regions that have limited access to freshwater, long-distance trans corrosion is unlikely to occur in plain concrete or concrete incorporated
portation of freshwater may increase the cost and energy consumption with noncorrosive reinforcement bars, which enhances both the
* Corresponding author. School of Civil and Environmental Engineering, University of Technology Sydney, NSW 2007, Australia.
E-mail address: wengui.li@uts.edu.au (W. Li).
https://doi.org/10.1016/j.cemconcomp.2021.104100
Received 8 February 2021; Received in revised form 3 May 2021; Accepted 5 May 2021
Available online 12 May 2021
0958-9465/© 2021 Elsevier Ltd. All rights reserved.
P. Li et al. Cement and Concrete Composites 121 (2021) 104100
economical and environmental benefits to use seawater in concrete hydration. Therefore, Wang et al. [18] stated that seawater mixing in
[11]. creases the total heat of hydration seems questionable. Besides, differ
This paper reviews current and previous studies on the predominant ences are also found in the clinkers hydrated in seawater. Li et al. [17]
performance differences between seawater-mixed and conventional reported that an increasing water/binder (w/b) ratio in seawater mix
concretes. Particular attention has been paid to the chloride-induced tures does not further stimulate the hydration of C3S, while the hydra
hydration mechanism to provide a scientific and unbiased foundation tion of C3A is observed promoted (see Fig. 1). The inconsistent results
for a better understanding of the applicability of seawater in construc indicate that the hydration of C3S is enhanced by raising the concen
tion. Through evaluating both the benefits and limitations of seawater tration of the influential ions, not by increasing their total content. While
mixtures, this work hopes to guide the safer use of seawater as a con as for C3A, the hydration can simply be promoted by adding more
struction material. seawater-related ions. Based on these results, a conclusion can be drawn
that during the accelerating hydration period, some special hydration
2. Properties of cementitious composites with seawater products were formed by the reaction of C3A and some ions in seawater,
for instance, the Friedel’s salt [20]. However, the stimulating effect of
Due to the long-term transfer from rainwater to the ocean, a great seawater on the hydration rate of C3S is limited in a certain range.
number of alkaline minerals are migrated constantly from the inland
rock stratum to seawater. The salinity varies in different marine regions 2.2. Fresh properties
that, for instance, the North Sea has a salinity of 3.3%, the Red Sea has
that of 4%, while the Baltic Sea has only 0.7% [10]. The main ions It is well understood that in a cast-in-situ construction, the fresh
concentration of seawater in different marine regions are listed in properties of concrete are required to be tested at the time of placement.
Table 1. Typically, seawater has a total salinity of 3.5%, which mainly As an important evaluating indicator, slump value is widely applied to
consists of chloride, sodium, magnesium, sulfur, etc. The concentration determine the workability of fresh concrete. Generally, a higher slump
of predominant components needed in freshwater to prepare substitute represents better workability, assuming that there is no segregation
seawater is specified in ASTM D1141-98 (see Table 1). In this section, a occurs. Regarding the seawater-mixed concrete, a consensus has been
comprehensive review is provided to summarize the advanced progress proposed among the existing studies that seawater mixtures presented a
of seawater-mixed cement composite, followed by a detailed illustration higher cohesive, viscous, and compact behavior, compared with the
of the mechanisms. conventional concrete [18,21]. The workability, however, becomes
worse, along with a shortened setting period [22]. According to the
2.1. Hydration kinetics studies by Bachtiar et al. [23] and Ting et al. [24], using seawater
instead of freshwater mixing would cause an approximately 30%
The exotherm of cement hydration changes greatly when mixed with reduction in slump values. It should, however, be noted that the influ
seawater, in particular, at the early stage. As proposed in the existing ence of seawater on slump value is reduced when the w/b ratio in
literature, a consensus has been suggested that seawater increases the creases, as Otsuki et al. [25] suggested that the slump of seawater-mixed
heat flow rate, due to the enhanced kinetics of hydration [12–15]. In concrete is only 15% lower at a w/b ratio of 0.53. Similar observations
seawater mixtures, the accelerated hydration is mainly dominated by were reported on ordinary Portland cement (OPC) concrete blended
the abundant chloride ions [16], which dramatically promotes the hy with supplementary cementitious materials (SCMs) in the studies by
dration of tricalcium silicate (C3S) and tricalcium aluminate (C3A) [17]. Sena-Cruz et al. [21], Younis et al. [12], and Katano et al. [26], in which
Also, the hydration of cement can be affected by the other ions that the reduced slump by seawater ranges from 10 to 30% at the decreasing
existed in seawater, such as sulfate, magnesium, and calcium [18,19]. w/b ratio from 0.5 to 0.26. Besides, a 20–25% shortened initial and final
The specific effects of these ions on cement hydration will be discussed setting time was observed in seawater-mixed concrete [27], which
in Section 3. Based on the aforementioned studies, in general, the showed a negative effect on the workability. In Wang et al. [18], the
exothermic peak of seawater mix presents 35–40% higher and occurs rheological behaviors were measured in cement pastes mixed with
approximately 17–30% earlier, compared to the conventional concrete. seawater and deionized water. Under the same shear rate, the shear
Furthermore, Younis et al. [12] found that the cumulative heat of stress and viscosity were observed remarkably higher in seawater mix
seawater mixtures is higher at the early age, while it becomes almost tures at 5 min and 15 min, indicating the stiffer feature in the fresh state.
identical to that of freshwater mixtures after 7 days. It seems that the The higher viscosity can be attributed to the larger solid content and the
heat of cement hydration in seawater is released in advance, subse stronger flocculations between the suspended particles from the accel
quently being caught up by freshwater mixtures. The exothermic feature erated hydration by seawater. Moreover, there is a consensus that the air
in seawater mix is very well accordant with the characteristic of content in seawater mixtures is typically 5–20% lower than that of
compressive strength growth, which both are dominated by cement freshwater mixtures [13,25,26], probably due to the higher viscosity in
Table 1
Concentration of the main ions in seawater from various seas.
Conc. [mg/l] Cl− Na2+ SO2-
4 Mg2+ Ca2+ K+ pH Refs.
South China Sea 33400 21700 3840 1040 579 417 7.8 [28]
Arabian Gulf 26000 15000 3700 2300 500 520 8.2 [12,18]
Red Sea 22660 11350 3051 1867 531 1350 − * [134]
Mediterranean Sea 20800 11640 2820 1360 490 420 8.0 [27,135]
Melbourne 20700 11940 3420 1430 − * 622 − * [136]
Norway Sea 19420 9460 2583 1130 351 348 − * [137]
Yellow Sea 19360 10780 2702 1297 408 388 − * [138]
Atlantic Sea 19920 10735 2697 1416 424 380 − * [16]
North Sea 16550 12200 2220 1110 430 500 − * [139]
Marmara Sea 14390 8100 2034 1035 328 340 7.9 [140]
Black Sea 9500 4900 1362 640 236 230 7.4 [140]
Baltic Sea 3000 1800 410 240 98 67 − * [141]
ASTM D1141-98 19437 10833 2769 1292 419 398 7 [142]
*
Not reported.
2
Table 2
Initial slump flow in seawater/cementitious composite.
Concrete type W/b ratio Initial slump flow Refs.
(mm)
FW SW SW/
FW
OPC paste 0.25 (with SP 54 46 0.85 [28]

1%)
OPC concrete 0.32 (with SP 121 82 0.68 [24]
1%)
OPC concrete 0.35 (with SP 122 80 0.66 [23]
1%)
OPC concrete 0.53 66 56 0.85 [25]
OPC 80% + Fly ash 20% 0.26 (with SP 610 420 0.69 [21]
concrete 1%)
OPC 35% + Slag 65% 0.34 600 480 0.80 [12]
concrete
OPC 50% + Slag 50% 0.5 165 150 0.91 [26]
concrete
Note: SP is superplasticizer; FW denoted freshwater mixtures; SW represents

seawater mixtures.
instance, the strength of cement composites can be varied after being

cured in water or moist condition. Fig. 2 summarizes the relative
compressive strength of seawater mixtures versus those of freshwater
mixtures under various curing conditions.
In moist curing where cement composite has been constantly stored
in the standard curing chamber, a higher compressive strength could be
retained in the seawater mixtures for 28 days or perhaps longer [24].
Similar results were also identified in the studies by Li et al. [17,28,29]
and Erniati et al. [30]. Specifically, mixing with seawater improves the
compressive strength by nearly up to 60% at 1 day, 7–28% at 3 days, and
0–5% at 28 days, respectively. In a long-term of over 28 days, it has been
demonstrated that the compressive strength between freshwater-mixed
composite and seawater-mixed composite displays almost no difference.
Slower development of strength can be observed in seawater mix
tures when immersed in freshwater or seawater, compared to conven
tional concrete under the same curing condition. In Islam et al. [31], an
approximately 5% lower value in strength was measured in
seawater-mixed specimens from 7 to 180 days in the water-curing
conditions, which was further reduced with the w/b ratio increases.
According to Wegian et al. [32], when the w/b ratio is 0.45, the
compressive strength was found to be 0–10% higher in seawater con
crete within 7 days. However, the increase of strength was slowed down
Fig. 1. Heat evolution of OPC hydrated in: (a) Freshwater (FW); (b) over time such that the 90-day compressive strength has only reached
Seawater (SW). about 85% of that in conventional concrete. Similar results could also be
verified by Guo et al. [33] that at a higher w/b ratio of 0.5, comparable
the paste matrix. Besides, the water film thickness was calculated a early strength and lower long-term strength could be obtained in
slightly smaller in seawater mixtures by Li et al. [28], which also led the seawater mixtures. These interesting findings indicate that the effect of
adverse effects on flowability of the fresh paste. According to these seawater on the strength development of cementitious composite is
studies, the associated experimental results are summarized in Table 2. complex. The seawater can accelerate the hydration of cement to pro
mote early-stage strength. On the other hand, the growth rate of strength
in seawater mixtures is not as high as that in freshwater mixtures, which
2.3. Mechanical properties
is probably due to the stability of microstructure affected by some spe
cific hydration products formed (i.e., brucite, calcium oxychloride,
Numerous studies have been done to evaluate the effectiveness of
M-S-H, etc. [34,35]). Based on the principle of chemical equilibrium,
compressive strength for seawater-mixed composite, as shown in
hydration products formed previously become unstable to decompose
Table 3. According to the existing reports, it is generally believed that
when pore solution changes. Therefore, for cement mixture, a sealed or
concrete made with seawater exhibits higher early strength by 5–30%
stable curing condition is beneficial for steady structure development.
than freshwater mixtures within 3 days [29,30]. However, it appears
When the pore solution is much closer to the matrix, the decomposition
that conflicting results have been reported in the studies regarding the
and leaching effect of cement hydrates will diminish. The conclusion can
late-age strength of seawater-mixed concrete. Based on the experimental
also well illustrate that seawater-mixed concrete achieves better
results, using seawater might produce nearly no impact or, inconsis
strength development when being cured in seawater than in freshwater
tently, cause serious adverse effects on the late-age strength, which
(see Fig. 2).
seems contradictory and, thus causing two opposite views on seawater
Mohammed et al. [36] investigated the seawater-mixed concrete
mixtures [10,25]. Nevertheless, through the perusal of the related
cured in tidal condition for 20 years and concluded that there was no
studies, it should be noted that the effect of curing conditions should be
significant difference in compressive strength between seawater-mixed
considered as a very influential factor in strength development. For
3
Table 3
Recent studies on the compressive strength in seawater-mixed composite.
Concrete type W/b ratio Curing condition Compressive strength (MPa) Refs.
FW SW SW/FW (%)
OPC paste 0.20 Moist curing 84.7 (3 d) 87.3 (3 d) 1.03 [43]

0.25 (with SP 0–2%) 110.9–123.0 (28 d) 111.7–132.5 (28 d) 1.02–1.08 [28]
0.35 (with SP 0–2%) 89.0–99.1 (28 d) 90.8–105.5 (28 d) 1.02–1.05
OPC mortar 0.40 Moist curing 42.1 (3 d) 49.1 (3 d) 1.17 [17]

51.9 (14 d) 53.7 (14 d) 1.03
54.2 (28 d) 54.5 (28 d) 1.01
OPC concrete 0.45 Moist curing 35.46 (3 d) 45.41 (3 d) 1.28 [29]
40.47 (7 d) 51.85 (7 d) 1.28
46.17 (28 d) 55.98 (28 d) 1.21
50.63 (56 d) 60.18 (56 d) 1.19
0.32 (with SP 1%) Moist curing 60.7 (7 d) 64.9 (7 d) 1.07 [24]
75.2 (28 d) 74.7 (28 d) 0.99
91.2 (90 d) 90.1 (90 d) 0.99
0.35 (with SP 1%) Moist curing 9.1 (1 d) 14.6 (1 d) 1.60 [30]
25.5 (3 d) 31.3 (3 d) 1.23
35.2 (7 d) 38.0 (7 d) 1.08
50.6 (28 d) 51.9 (28 d) 1.03
52.7 (56 d) 53.7 (56 d) 1.02
0.40 Freshwater curing 18.3 (7 d) 17.3 (7 d) 0.95 [31]
24.5 (28 d) 23.5 (28 d) 0.96
30.0 (90 d) 27.5 (90 d) 0.92
31.5 (180 d) 28.9 (180 d) 0.92
Seawater curing 17.9 (7 d) 17.0 (7 d) 0.95
24.2 (28 d) 23.1 (28 d) 0.95
28.5 (90 d) 27.4 (90 d) 0.96
29.6 (180 d) 28.7 (180 d) 0.97
0.45 Seawater curing 27.6 (7 d) 30 (7 d) 1.09 [32]
32.1 (14 d) 33.8 (14 d) 1.05
34.9 (28 d) 34.7 (28 d) 0.99
30.6 (90 d) 25.5 (90 d) 0.83
Freshwater curing 41.2 (7 d) 43.7 (7 d) 1.06 [33]
46.1 (14 d) 46.5 (14 d) 1.01
54.3 (28 d) 45.4 (28 d) 0.84
65.3 (90 d) 57.1 (90 d) 0.88
46.5 (14 d) 45.5 (14 d) 0.98
50.0 (28 d) 48.3 (28 d) 0.97
62.5 (90 d) 54.6 (90 d) 0.87
45.2 (14 d) 41.5 (14 d) 0.92
47.2 (28 d) 41.4 (28 d) 0.88
59.3 (90 d) 45.4 (90 d) 0.77
46.2 (14 d) 43.6 (14 d) 0.94
49.8 (28 d) 46.1 (28 d) 0.93
55.0 (90 d) 53.3 (90 d) 0.97
14.0 (14 d) 16.0 (14 d) 1.14
15.8 (21 d) 17.0 (21 d) 1.08
20.0 (28 d) 20.3 (28 d) 1.02
14.5 (14 d) 17.9 (14 d) 1.23
13.1 (21 d) 19.8 (21 d) 1.51
18.7 (28 d) 21.9 (28 d) 1.17
Note: SP: Superplasticizer, FW: Freshwater-mixed composite, SW: Seawater-mixed composites.
concrete and freshwater-mixed concrete. For the tidal seawater curing, 2.4. Transport properties
it should, however, be noted that the external condition was an enor
mous environment relative to the concrete itself along with constant In general, concrete deterioration is mainly due to the mass transfer
dry-wet alternated, where the higher water absorption and outward of chemicals that can be transmitted inside and outside through the pore
leaching effect of alkalis and calcium hydroxide occurred to cause severe solution. The transport properties are accordingly valued to predict the
deterioration [37,38]. It should be noted that the expansive product, service life of the concrete. Pore structure, as an important character for
calcium oxychloride, may cause serious deterioration to reduce concrete evaluating the quality of hardened cement mixtures, dominating the
properties [34,39] (detailed discussion in Section 3). Therefore, the permeability that affects moisture and ions diffusion in concrete. Ac
adverse effect of long-term seawater curing is quite significant than cording to the available research, using seawater in concrete shows a
those only with seawater as a mixture on their strength development. great effect on the early-stage pore structure, while the difference of
long-term pore size distribution is hardly observed. Li et al. [29] re
ported that the ratio of pores smaller than 10 nm is significantly
4
Fig. 4. The differential pore volume of concrete at 3, 7 and 28 days between

Fig. 2. Time-dependent changes of compressive strength ratios between mixed
FW and SW (FW: concrete mixed with seawater, SW: concrete mixed with
with seawater and mixed with freshwater cured at different conditions [17,24,
freshwater) [40].
28–33,43,143].
days [13], which is well accordant with the slightly higher compressive
increased by using seawater at 3 days (see Fig. 3), while the porosity is
strength in seawater-mixed concrete.
5% higher than that of conventional concrete. During the curing time
In Wang et al. [18], the C–S–H gels with a high surface area were
which followed, the porosity decreased remarkably over time and,
detected in seawater paste by Brunauer-Emmett-Teller (BET) surface
gradually, the amount of small pores (<10 nm) overtakes the large pores
area measurements. It has been found that the cement paste with
(10–5000 nm) at 7 days. However, it appears that the porosity has not
seawater possessed a surface area of up to 27.4 m2/g, while that of
markedly altered at 28 days when using seawater mixing instead of
freshwater paste was only 14.3 m2/g. In the meantime, the volume of
freshwater, indicating that the two composites have formed a similar
pores ranged between 2 and 60 nm is 0.0434 cm3/g in seawater paste,
pore structure at the late stage. The result has also been verified by
which was 70% higher than that of freshwater paste. The result shows a
Montanari et al. [16] that an identical total porosity ranged between
finer structure of C–S–H gels with a higher volume of pores below 5 nm
0.166 and 0.167 ml/g is determined at 91 days, in both cement pastes
formed after seawater mixing. Likewise, Younis et al., [12] measured the
mixed with seawater and freshwater. Shi et al. [40] investigated the
permeability performance of seawater concrete through rapid chloride
effect of seawater on the evolution of pore structure with the aid of
permeability test (RCPT), chloride migration test (CMT), and water
mercury intrusion porosimetry (MIP), as shown in Fig. 4. The results
absorption test (WAT). The results demonstrate that the seawater mix
demonstrate that in the presence of seawater, most of the pores over 10
tures were found to achieve higher impermeability at 28 days. However,
nm are converted to finer pores, exhibiting a refined structure at the
the results of the sea mixtures and freshwater mixtures were comparable
early stage, after which the distinction between the two mixtures rela
over 56 days, revealing that the late-stage permeability is less affected
tively decreases with age. Still, the volumes of pores ranged between 50
by the mixed seawater.
and 1000 nm were slightly higher in seawater mixtures at 28 days, and
The migration behavior of conductive ions in the pore solution, such
more fine pores less than 30 nm formed in the seawater mixtures at 60
as chloride, sulfate, or possibly hydroxyl ions dominates the electrical
resistance of the concrete. When an electric field is applied, electro
migration occurred as the only mechanism of concrete conductivity
[41]. Montanari et al. [16] demonstrated that using seawater as the
mixing water decreased the electrical resistivity of concrete by
approximately 40–50%, due to the considerable amount of ions in pore
solution, as shown in Fig. 5. Moreover, due to the concentration of the
pore solution increasing with hydration age, cement pastes with fresh
water showed a 40% reduction in the electrical resistivity at 28 days,
while only a 25% reduction in the resistivity was observed in seawater
paste within the same time duration.
2.5. Shrinkage
During the cement hydration or drying process of concrete,

shrinkage occurs due to the moisture decrease, which may, in some
cases, result in structural cracking or deterioration in the service life.
According to the distinction of water consumption, the shrinkage
behavior can be divided into autogenous shrinkage and drying
shrinkage [42]. Li et al. [43] revealed that the autogenous shrinkage of
OPC paste was increased by nearly 30% when seawater was used
Fig. 3. Evolution of porosity and the ratios of large and small pores between instead, attributing to the remarkably refined microstructure from a
mixed with seawater and mixed with freshwater [29]. quicker self-desiccation of cement in seawater [40]. Furthermore, the
5
Fig. 7. Drying shrinkage and mass loss of mortar mixtures [13].

Fig. 5. Evolution of electrical resistivity in cement paste mixed with seawater
or seawater [16].
autogenous shrinkage strains were reducing with the w/b ratio increases
[13]. Khatibmasjedi et al. [13] showed that using seawater instead
increased the 63-day autogenous shrinkage from 213 to 387 microstrain
at a w/b ratio of 0.36, and that from 149 to 314 μs at a w/b ratio of 0.45,
as illustrated in Fig. 6. It should be noted that the shrinkage by seawater
is significantly higher in 3 days, which might be due to the acceleration
in the early-age hydration by seawater [40].
On the other hand, drying shrinkage induced by vapor diffusion also
occurs in concrete during the external drying process, which inevitably
happened after the moist-curing period [44]. Unlike the autogenous
shrinkage which was scarcely changed at the late stage, the drying
shrinkage has grown continuously throughout 65 days [13]. In Fig. 7,
the drying shrinkage values are quite close in both seawater and fresh
water mixtures at a w/b ratio of 0.36, approaching 600 μs and 560 μs,
respectively. With the increase of the w/b ratio, the drying shrinkage
becomes more significant in seawater mixtures, which reaches 980 μs
and exhibits 17% higher than the freshwater mixtures at the w/b ratio of
0.45. Also, the use of seawater has resulted in a reduction of mass loss
during the drying process, and thus the slope of the drying
shrinkage-mass loss curve of seawater mixtures becomes larger, as seen
Fig. 8. Drying shrinkage versus mass loss for mortar mixtures [13].
in Fig. 8. The higher shrinkage can be due to the denser microstructure
of hardened cement paste formed in seawater [16,45], where the

capillary induced stress caused by menisci becomes higher in smaller
pores, leading to greater internal shrinkage stress. On the other hand,
the chloride-based salts in seawater such as sodium chloride are hy
groscopic [46], resulting in slower moisture escaping. Therefore, it can
be drawn that seawater mixtures are featured with a lower mass loss but
a higher shrinkage instead, compared to conventional concrete.
3. Effect of seawater-related ions on hydration
Based on the existing studies, the effect of some seawater-related ions

on cement hydration is discussed separately. This section focuses mainly
on the existing ions that play important roles in cement hydration at
their specific concentration in seawater; however, it should be noted
that multiple chemical reactions occur simultaneously in practice, and
the effects are closely related to their actual concentration.
3.1. Chloride ion
As the most influential ion in seawater on concrete properties,

Fig. 6. Autogenous shrinkage of mortar mixtures with w/b 0.36 and 0.45 [13]. chlorides should be discussed in detail first. The effect of chlorides on
6
hydration can be divided into three parts: 1) the interaction with AFm in the AFm-related adsorption as well. Hence, the formation of Friedel’s
(aluminate-ferrite-mono) phase; 2) the formation of calcium oxy salt also occurs by an ionic interaction from other AFm phases [51,53,
chloride (CAOXY); and 3) the adsorption by C–S–H, as discussed below 54]. Additionally, the various bound anions can be partially miscible
separately. with each other in AFm phases [55] (e.g., Kuzel’s salt:
3CaO⋅Al2O3⋅1/2CaCl2⋅1/2CaSO4⋅10H2O; Hemicarboaluminate: Ca4Al2
3.1.1. Interaction with AFm (CO3)0⋅5(OH)13⋅5.5H2O, etc. [49]). Therefore, the thermodynamic sta
In Portland cement, chloride can be chemically bound with the bilization of the AFm hydrates is highly associated with the formation
layered calcium-aluminate hydrates as AFm hydrates [47], such as mechanism of chloride-related hydrates.
calcium chloroaluminate, commonly known as Friedel’s salt: At temperature of 25 ◦ C and the standard atmosphere (100,000 Pa),
3CaO⋅Al2O3⋅CaCl2⋅10H2O [20]. Suryavanshi et al. [48] proposed that monocarboaluminate (CO3-AFm) is more stable than mono
there are two mechanisms for the formation of Friedel’s salt, namely, the sulfoaluminate (SO3-AFm) in Portland cement, which promotes a pro
adsorption mechanism and the anion-exchange mechanism. Firstly, the gressive substitution of CO2− 3 in SO3-AFm to convert into hemi
bulk of free-chloride ions present in the pore solution can be adsorbed carboaluminate or, more likely, CO3-AFm [49]. Similarly to CO2− 3 , Cl
−
directly in the interlayers of the layered AFm structures, to balance the can also displace sulfate in SO3-AFm, converting to Friedel’s salt as the
charge arising from the displacement of a Ca2+ ion by an Al3+ ion, as more stable AFm hydrates. Numerous studies have investigated the
shown in Fig. 9. Secondly, the free-chloride ions can displace the OH− chloride effect on the mineralogy of hydrated Portland cement [56–59].
ions from hydroxy-AFm hydrates (OH-AFm), such as C4AH13, C4AH11, Due to the significant competitiveness of carbonate in binding AFm
C4AH19, etc., forming Friedel’s salt coincided with a comparable content hydrates, the impact mechanism of chloride has to be discussed
of OH− ions released in the pore solution. The two separate mechanisms respectively in the cementitious system with and without carbonates as
can effectively explain the phenomenon that an increase of OH− con follows.
centration occurs during the chloride ions binding, and also that the
concentration of bound Cl− is much higher than that of the displaced
OH− from the interlayers.
However, it seems not necessary to strictly distinguish the two
mechanisms aforementioned. Intending to balance charges, both chlo
ride ions, and hydroxyl ions can be adsorbed by the principal layer of the
AFm structure, and the two adsorptions occur simultaneously. During
the process, the adsorbed OH− ions are more easily displaced by Cl−
ions, due to the fact that OH− hardly competes with Cl− in balancing
charge. Meanwhile, the free OH− ions are retained in the pore solution,
leading to an increase of pH. Therefore, the increase of alkalinity in pore
solution is due to the significant substitution of adsorbed OH− ions by
Cl− ions, and it also means that the metastable hydroxy-AFm hydrates
hardly coexist with aqueous Cl− [49]. The same opinion has been put
forward by Galan and Glasser [50] that the ion exchange mechanism is
dominant in the formation of Friedel’s salt [51]. Suryavanshi et al. [48]
also proposed theoretical support but lacks strong evidence. Due to the
competitive relationship between hydroxyl ions and chloride ions to
bind with AFm, the formation of Friedel’s salt is reduced in high alkaline
conditions. In other words, chloride binding is decreased with high al
kali content, which was verified by Rasheeduzzafar et al. [52].
Additionally, the formation of Friedel’s salt is not limited to the Fig. 10. The relative amount of solid hydrate phases of a hydrated model
substitution reaction between free Cl− and OH− , which means that the mixture consisting of initially 1 mol C3A, excess of portlandite, and with fixed
reaction of Cl− and AFm is more than just one way on a single AFm initial sulfate ratio (SO3/Al2O3 = 1), showing phase development and its
phase. Some other anions, such as SO2− 2−
3 and CO3 , can compete with Cl
−
dependence on changing chloride ratios (2Cl/Al2O3) at 25 ◦ C [60].
Fig. 9. Structural representations of AFm phase: hydrogen bond network within the interlayer region [48,145].
7
In the carbonate-free system, as depicted in Fig. 10 by Balonis et al. products in laboratory conditions. Damidot et al. [81] have investigated
[60], Cl− substitute SO2− 3 from SO3-AFm hydrates, forming an inter the CaO–Al2O3–CaCl2–H2O system at 25 ◦ C and defined the stable
mediate compound Kuzel’s salt at lower chloride concentration (molar invariant points that Friedel’s salt can coexist with CH and 3-1-15
ration of 2Cl/Al2O3 < 0.70) and Friedel’s salt at higher chloride con compound with a Cl− concentration of 3.285 mol/kg, and Friedel’s
centration (molar ratio 2Cl/Al2O3 > 0.70). Meanwhile, the substituted salt can coexist with both 3-1-15 compound and 1-1-2 compound with a
SO2−
3 are released to the aqueous solution, forming ettringite subse Cl− concentration of 6.427 mol/kg. Besides, CAOXY also can be
quently when the required concentration is reached [50,61–63]. The observed in the presence of MgCl2 solution [82,83], and the concen
ettringite is hard to decompose due to the higher stability and thus the tration is also required. Wang et al. [34] indicated no CAOXY formed in
content of ettringite increases with the presence of chloride, as proposed solution with 4% MgCl2, but it was detected in solution with 20% MgCl2.
by Hirao et al. [64] and Cao et al. [57]. Additionally, some studies are However, the existing studies indicated no sign of CAOXY formation
consistent that ettringite shows no binding capacity to chlorides [64, when cement paste was exposed to NaCl solution [82,84,85], with an
65], and thus can coexist steadily with Friedel’s salt at high chloride exception in carbonated concrete in one study [86]. The studies agree
concentration. In Zibara [66] and Ekolu et al. [62], however, ettringite well with the conclusion that CaCl2 and MgCl2 solutions might exhibit
was observed decomposed and gradually converted to Friedel’s salt in expansion and crack in cement paste [78], but NaCl is relatively benign
NaCl solution of 3 M and 2.8 M, respectively. [82,84,85], since the expansive product CAOXY does not exist in
In practice, however, many modern types of cement contain added Ref. [87]. Therefore, CAOXY would hardly form in cement paste hy
calcite, leading to a portlandite-calcium carbonate-carbon dioxide drated in seawater, owing to the fact that the required concentration of
equilibrium in the hydration system [50]. Therefore, hemi CaCl2 and MgCl2 is not reached. Moreover, in practice cases, there is no
carboaluminate or, more likely, monocarboaluminate will form in the clear evidence in the existing studies that CAOXY was observed in the
carbonate-contained Portland cement. In this hydration system, it can be OPC hydrated in seawater at standard concentration.
observed that carbonate ions in AFm phases are gradually substituted by
chloride with a rising chloride concentration [55,60]. The released 3.1.3. Chloride adsorped by C–S–H gels
carbonate ions are subsequently bound as calcite. It should, however, be In addition to chemical combinations, some researchers found that
noted that there is no Kuzel’s salt appeared even as the intermediate chloride can be physically adsorbed in C–S–H gels due to its large spe
products during the substitution in the carbonate-containing system, cific area [88,89]. It is found that the chloride binding capacity of C–S–H
which can be attributed to the more stability of CO3-AFm than that of increases with the increase of chloride concentration [64,89,90], and
Kuzel’s salt [67]. The result reflects that Kuzel’s salt is incompatible also more chloride can be found held in C–S–H with a higher
with CO3-Afm, and it is destabilized when encountering even small calcium-silicate (Ca/Si) ratio [66,91,92]. Therefore, C–S–H gels with a
amounts of carbonate in the hydration system. Furthermore, Chang et al. higher Ca/Si ratio are seemingly more beneficial to reduce the free
[68] observed most of Friedel’s salt and Kuzel’s salt decomposed after chlorine in concrete under marine environments. Experimental and
experiencing a 28-day carbonization period with CO2 concentration modeling results [93] suggest that the chloride ions associated with the
being 20%. The results indicate that the bound chlorides in AFm were C–S–H are mainly present in the diffuse layer of the C–S–H surface,
unstable and prone to be set free again in the carbonization which is positively charged in the presence of high calcium concentra
environment. tions. The finding can well explain that C–S–H exposed in CaCl2 solution
Some other studies have been conducted to observe the binding of has higher chloride adsorption than that exposed in NaCl with the same
chloride in hydrated calcium aluminoferrite to find that the rate of hy chloride concentration [90]. However, despite the adsorption proposed
dration reaction is slower than C3A and decreases with increasing iron in most previous studies, an opposite view has been put forward by
content as the order C6A2F > C4AF > C6AF2 [69]. The studies show that Plusquellec and Nonat [94] that C–S–H particles cannot adsorb chlo
both of these aluminoferrites can bind chloride and develop hydrates rides, based on the zeta potential measurement. The research observed
with similar structures to Friedel’s salt but with iron instead of that the zeta potential was well linearly increasing with the increase of
aluminum (e.g., 3CaO⋅Fe2O3⋅CaCl2⋅10H2O) [70,71], and the concentration, and thus no adsorption occurred in the process to reduce
iron-containing Friedel’s salt may form solid solutions with other AFm the zeta potential. This finding questions the common experimental
phases. method of calculating the concentration of chlorides that some
electrical-attracted chlorides by C–S–H particles are inevitably removed
3.1.2. The formation of CAOXY during the filtration process, and thus the reduction of chloride in so
Apart from the aluminate hydrates, chloride can also be chemically lution is wrongly attributed to adsorption by C–S–H. As a result, further
bound during the cement hydration process through the formation of
calcium oxychloride compounds: xCa(OH)2⋅yCaCl2⋅zH2O (CAOXY) [58,
72–75]. As calculated by Qiao et al. [39], the volume of CAOXY is 303% Table 4
as large as Ca(OH)2 and thus leads to a volumetric expansion. In hard Main interactions between chloride ions and cement hydrates.
ened cement paste, the expansive pressure caused by CAOXY is un Mechanism Cement hydrates Comment
doubtedly responsible for the deterioration of mechanical properties. involved
The most common compositions range from the dehydrated phase with Chloride AFm converts to Friedel’s Anions in AFm such as hydroxyl,
x:y:z of 1:1:0 compound to the hydrated phase with the ratio of 3:1:15 exchange salt sulfate, and carbonate can be
(3-1-15 compound) and 4:1:10 (4-1-10 compound), depending on the displaced by chloride to form
various chloride concentrations and test conditions [74,76]. In most Friedel’s salt. The stability of related
hydrates follows the order of
cases, however, the formation of CAOXY can be observed in CaCl2 so
Ettringite > Friedel’s salt > CO3-AFm
lutions instead of in NaCl2 solutions [34,39,58,77–79], and a certain > SO3-AFm > hydroxy-AFm in the
concentration is required. Farnam et al. [58] suggested that the mini range 0–70 ◦ C [50].
mum concentration of CaCl2 is 11.3% when CAOXY is formed. Qiao Phase change Destabilization of CH and Portlandite is converted to CAOXY of
et al. [80] determined the required Cl− concentration for CAOXY for formation of CAOXY 3-1-15 compound or, perhaps, 1-1-2
compound. The reaction occurs in a
mation, which is 1.20 M at 5 ◦ C and 3.64 at 23 ◦ C. Similar results were high concentration of aqueous CaCl2
reported by Brown and Bothe [76] that the formation of 3:1:15 com or MgCl2, and not occurs in aqueous
pound requires a chloride concentration approaching 4 M, and the for NaCl.
mation of 1-1-2 compound (hydrated phase with the ratio of 1:1:2) Chloride C–S–H More chloride can be adsorbed in
Adsorption C–S–H with a high Ca/Si ratio.
needs about 9 M of chloride concentration, based on the synthetic
8
studies are needed to clarify the chloride-adsorption capacity of C–S–H. leading to the formation of higher-Ca/Si C–S–H and M-S-H with higher
Based on the aforementioned discussion, Table 4 summarizes the main chloride content [56,74,100]. The second cause can be the formation of
effect of chloride on cement hydrates. CAOXY or magnesium oxychloride (MAOXY) [34,35,74,83,104]. The
third one is that in cement blended with alumina-rich SCMs, higher
3.1.4. Effect of chloride ions in seawater calcium concentration can enhance the formation of Friedel’s salt,
The change in properties of cement composites caused by seawater which increases the chloride binding capacity of the composite system
(Section 2), can be well explained through the chloride effects afore [101,105]. The schematic diagram of the chloride distribution in cement
mentioned. Because the chloride concentration in seawater is close to hydrates is summarized in Fig. 11. Furthermore, De Weerdt et al. [61]
0.55 M, which is lower than the threshold of that in the formation of observed an approximately 0.25 M higher chloride bound in MgCl2 than
CAOXY (1.2–9 M [76,80]), it is still questionable whether the reaction CaCl2, which probably indicates a higher binding chloride enhanced by
between chloride and CH occurs in seawater composites. Therefore, the Mg2+.
property differences between seawater composite and freshwater com
posite can be considered to be dominated by the interaction between
3.4. Chloride binding in hydrates
chloride and AFm phases. During the seawater-mixed hydration process,
especially in the early stage, the calcium, aluminate, and ferrite ions are
Chloride binding happens during the cement hydration process,
consumed to form Friedel’s salt, thus enhancing the hydration kinetics
which significantly limited the free chloride penetration. Therefore, the
and accelerating the hydration rate of C3S, C3A, and C4AF. The enhanced
threshold concentration of chloride-induced corrosion can be affected
hydration in the early stage is responsible for the rapid heat evolution
by the chloride binding capacity of cement hydrates, and it further in
and higher autogenous shrinkage; the ratio of solid phase is increased to
fluences the service life of the concrete. The chloride binding of cement
reduce the setting time; the cement hardens faster with higher
hydrates can be affected by various factors, such as hydration degree,
compressive strength and refined pore structure.
cation of chloride salt, carbonation, etc. [67,106]. Among them, chlo
ride binding is significantly associated with chloride concentration and
3.2. Sulfate and carbonate ions
different binding mechanisms. In general, the relationship between the
adsorbed chloride and free chloride in OPC is considered as non-linear
Sulfates and carbonates, as the other major anions in seawater, can
isotherms:
also affect cement hydration. Although the content involved in seawater
is much less than that from gypsum and calcite in cement itself, the effect Freundlich equation: ⋅ Cb = α⋅Cfβ (1)
can not be ignored. Similar to the chlorides, the sulfate ions can be
bound in AFm and C–S–H, with similar mechanisms aforementioned. α⋅Cf
Langmuir equation: ⋅ Cb = (2)
Therefore, chloride and sulfate would compete and displace mutually 1 + β⋅Cf
for the binding space, and the chloride binding ability of hydrates is thus
reduced [57,61]. For example, more ettringite would be developed in where α and β are binding constants at a given temperature, Cb is the
the presence of sulfate in seawater [55,95], including that delayed absorbed amount of chloride per gram of the solid at equilibrium, and Cf
forming by the displaced sulfate by chloride. The ettringite possesses is the chloride concentration in solution at equilibrium.
higher stability, consuming AFm hydrates, and presenting no chloride Tang and Nilsson [107] investigated the chloride binding capacity of
capacity as well [64,65]. Analogously, carbonate also presents negative OPC concrete and found Freundlich isotherm is better suited for high
effects on the chloride binding of cement hydrates. Chang et al. [68] chloride concentration (>0.01 M) and Langmuir is more applicable for
determined that when cement is exposed to a carbonation environment low chloride concentration (<0.05 M). However, it should be noted that
with a CO2 concentration of 20%, all chloride bound previously in AFm because the two equations are empirical and designed to fit data, the
and nearly 90% of that in C–S–H would be released into pore solution constants α and β serve no practical concern and may vary in a wide
after 28 days and 56 days, respectively. While even worse, the chloride range.
binding ability will no longer exist in carbonated cement. Therefore, if
seawater is used as mixing water instead, there is only a small amount of 3.4.1. Chloride binding capacity of AFm
cement hydrates changed from the limited additional sulfates and car The chloride binding capacity of sulfo-AFm has been investigated by
bonates. However, if being long-term exposed to seawater, sulfate attack Hirao et al. [64]. They found that the relationship between bound
or carbonation, hydration products such as Friedel’s salt and C–S–H chloride and sulfo-AFm phase follows Freundlich isotherm, in mg of
would become more vulnerable to convert and release bound chloride. bound Cl/g of sulfo-AFm:
Cb, sulfo− = 51.89⋅Cf0.58 (3)

3.3. Sodium, magnesium and calcium ions AFm
where the molecular formula of sulfo-AFm is C4 ASH10 at 11% relative

The cations in seawater mainly include sodium, magnesium, calcium
humidity (RH).
and other trace amounts of heavy metal ions. Among them, the content
While as for another AFm phase hydroxy-AFm, the chloride binding
of sodium, magnesium, and calcium accounts for approximately 30.6%,
capacity is higher since hydroxyl ions are more easily substituted by
3.7%, and 1.2% of the total mass in seawater, respectively [96]. It
chloride in AFm (see section 3.1). As the study by Birnin-Yauri and
suggested that sodium ion has no specific interaction with the cement
Glasser [108], pure Friedel’s salt can be observed stable at 20 ◦ C when
hydrates [97]. To maintain charge neutrality, some sodium ions have to
chloride concentration ranges over 14.5 mM, indicating that all
be removed from the solution when chloride is bound. Therefore, the
hydroxy-AFm phase has been bound chloride and converted to Friedel’s
unstable sodium ions can be also absorbed in C–S–H by reaction with
salt. The binding capacity, in mg of bound Cl/g of hydroxy-AFm, is thus
silanol groups [48,51] or substituting calcium [92,98,99]. Furthermore,
a constant when chloride concentration is over 14.5 mM:
one thing should be noted that the associated cations such as sodium,
calcium, and magnesium can affect chloride binding. Specifically, cal Cb, hydroxy− AFm = 126.5 (4)
cium and magnesium ions significantly increase the chloride binding
capacity of hydrates while sodium presents no remarkable effect where the molecular formula of sulfo-AFm is C4AH13 at 11% RH.
[100–103]. One of the main reasons is the calcium and magnesium ions,
which cause the precipitation of Ca(OH)2 or Mg(OH)2; the pH of the 3.4.2. Chloride binding capacity of C–S–H
solution decreases and the solubility of Ca2+ increase consequently, Some studies investigated the chloride binding capacity of the C–S–H
9
Fig. 11. Schematic diagram of chloride distribution in hydrates of OPC and alumina-riched SCMs exposed to NaCl and CaCl2 solutions [58,80,92,101].
phase and, however, the suggested isotherms vary in a wide range, with
Cb, Friedel′ s salt = 0.31⋅Cf0.46 (11)
poor fit compared to experimental data [64,66,109]. Hirao et al. [64]
described that the relationship between chloride concentration and
where the molecular formula of Friedel’s salt is 3CaO⋅Al2O3⋅CaCl2⋅6H2O
bound chloride follows a Langmuir isotherm in C–S–H, while a
at 11% RH [111].
Freundlich isotherm was considered by Elakneswaran et al. [109], in mg
As mentioned by Elakneswaran et al. [109], the sorption capacity of
of bound Cl/g of C–S–H:
ettringite was lower than portlandite and Friedel’s salt, but higher than
2.65⋅Cf tobermorite. However, the capacity of tobermorite described in Eq. (7) is
Cb, C− = 21.84⋅ (5)
S− H, H
1 + 2.65⋅Cf higher than those of portlandite and Friedel’s salt in Eqs. (10) and (11),
which seems contradictory. Hence, the binding capacity of AFt is still not
Cb, C− S− H, E = 12⋅Cf0.63 (6) clear. Assuming that the minimum binding capacity of AFt is as same as
that of C3S, Florea et al. [110] suggested the total bound chloride by AFt,
Zibara [66] also investigated the chloride binding capacity of C–S–H. which also takes into account the effect of the delay formed ettringite
Although the specific equation is not given, the binding isotherm of from chloride substitution (section 3.2):
cement minerals of C2S and C3S has been determined, in mg of bound
Cl/g of solid: mAFt + mdelayed AFt mC AS H + 0.3⋅mFriedel′ s salt
mcl, AFt = Cb, AFt ⋅ = 6.65⋅Cf0.334 ⋅ 6 3 12
msample msample
Cb, C3 S = 6.65⋅Cf0.334 (7)
(12)
Cb, C2 S = 7.89⋅Cf0.136 (8) where the molecular formula of AFt is C6 AS3 H12 at 11% RH [111]; mAFt ,
By considering the almost same hydration products for C2S and C3S mdelayed AFt , mFriedel′ s salt and msample refer to the mass of initial ettringite,
and the amount of C–S–H generated, Florea et al. [110] combined the delayed formed ettringite, formed Friedel’s salt and the sample,
two isotherms and the ratio of two minerals as the isotherm of C–S–H respectively.
provided by Zibara [66]: Table 5 and Fig. 12 summarise the chloride binding capacity of
Cb, C− S− H, Z = 6.65⋅Cf0.334 ⋅δC3 S + 7.89⋅Cf0.136 ⋅δC2 S (9)

Table 5
where δC3 S and⋅δC2 S is the mass fraction of C3S and C2S. Florea et al. Chloride binding capacity of cementitious phase.
[110] evaluated the three binding isotherms of C–S–H to their experi
Phase Isotherm Chloride binding capacity Refs.
mental data. The Cb, C− S− H, Z isotherm in Eq. (9) is considered as the best
Hydroxy- Constant Cb, hydroxy− = 126.5 [108]
to express the chloride binding capacity of C–S–H, which presents the AFm
AFm
lowest relative error and standard deviations to experimental results. Sulfo-AFm Freundlich Cb, sulfo− = 51.89⋅C0.58 [64]
AFm f
C–S–H Langmuir 2.65⋅Cf [64]
3.4.3. Chloride binding capacity of other cement hydrates Cb, C− S− H, H = 21.84⋅
1 + 2.65⋅Cf
As for other phases, ettringite and portlandite, the chloride binding C–S–H Freundlich Cb, C− S− H, E = 12⋅C0.63
f
[109]
ability is highly debated as most of the existing studies determined no C–S–H Freundlich Cb, C− S− H, Z = 6.65⋅C0.334
f ⋅δC3 S + 7.89⋅ [66,110]
chemical binding exhibited, even when exposed to a high concentration C0.136
f ⋅δC2 S
of chloride solution [63,64,66]; while the physical binding capacity, CH Freundlich Cb, CH = 0.087⋅C0.62
f
[109]
although insignificant, has also confirmed its existence and even Friedel’s salt Freundlich Cb, Friedel′ s salt = 0.31⋅C0.46
f
[109]
detected in Friedel’s salt [109]. Elakneswaran et al. [109] concluded C3S Freundlich Cb, C3 S = 6.65⋅C0.334
f
[66]
isotherms of portlandite and Friedel’s salt, in mg of bound Cl/g of solid: C2S Freundlich Cb, C2 S = 7.89⋅C0.136
f
[66]
AFt Freundlich not clear [109,
Cb, CH = 0.087⋅Cf0.62 (10)
110]
10
cement hydrates, the chloride binding capacity is very low and can be
ignored.
3.4.4. Chloride binding in seawater

Sodium chloride is the main chloride source in seawater, and the
magnesium and calcium content is too low to cause any significant
enhancement of chloride binding [61]. Therefore, the chloride binding
of Portland cement hydrated in seawater is similar to that in NaCl so
lution, which performs lower than that in calcium chloride or magne
sium chloride. According to the aforementioned binding capacity, in
seawater concentration (0.546 M), the total bound chloride by cement
hydrates can be divided as 49% to hydroxy-AFm, 20% to sulfo-AFm,
28% to C–S–H, and 3% to portlandite and Friedel’s salt [110], as
described in Fig. 13. In Fig. 14, a schematic diagram is given to sum
marize the preferable chloride distribution in seawater mixtures. It is
proposed that the conversion of the AFm phase to Friedel’s salt seems
the only significant mechanism for chemically binding chloride; most of
the binding mechanisms are either lower effective or else do not work in
low chloride concentration. It should also be noted that most of the
binding mechanism is chloride absorption and, perhaps, reversible that
Fig. 12. Chloride binding capacity of cement hydrates including hydroxy-AFm, the free chlorides and bound chlorides are hard to distinguish rigorously
sulfo-AFm, C–S–H, CH, and Friedel’salt at 11 RH% [64,66,108–110]. [50].
various cementitious phases at 11% RH based on the fitted isotherms, 4. Performance of seawater-mixed concrete
determined by existing studies. It can be concluded that the hydroxy-
AFm phase, exhibiting a constant binding capacity versus chloride Based on the existing studies, the negative effects and specific
concentration, provides the greatest contribution to chloride binding, properties of seawater-mixed concrete are discussed. To alleviate the
which can be completely converted to Friedel’s salt in chloride con potential defects, various methods have been proposed and adopted to
centration over 0.015 M. The binding capacity of sulfo-AFm increases improve concrete performance, as discussed separately.
with the increase of chloride concentration, from 13% at 0.3 M to over
33% at 3 M [110]. Therefore, the AFm phase provides the greatest 4.1. Superplasticizer and retarder
contribution to chloride binding, accounting for around 70% of the total
chloride binding capacity of OPC. The second strongest chloride binding It is generally believed that seawater cement paste has lower work
is C–S–H that the binding ability remains stable and effective in a wide ability and flowability due to the rapid hydration effect at an early age
range of chloride concentrations, which comprises 25–28% of total [12]. Therefore, using a superplasticizer (SP) and retarder is recom
bound chloride [110]. Although the content of CH is higher in OPC mended in seawater-mixed concrete. Li et al. [28] compared the SP ef
hydrates, the corresponding chloride binding ability of CH and Friedel’s fect of cement paste with freshwater (FW) and seawater (SW) in Fig. 15.
salt is only 2–5%, due to the insignificant adsorption. While as for other The result showed that the slump increases significantly when SP is
added over 1%, though the slump of SW remains a slightly lower value
Fig. 13. Distribution of bound chloride in OPC hydrates in seawater [110].
11
Fig. 14. Chloride binding mechanism and distribution versus chloride threshold for steel corrosion in a range of concentration (1–104 mM), assuming aqueous pH
are around 12.5 and the temperature is around 25 ◦ C. Data from the study of Chalhoub et al. [146], Jones et al. [74], Damidot et al. [147], Stark et al. [62], Zibara
[66], and Hou et al. [148].
than that of FW. The difference between the two samples is gradually demonstrate that the hydration degree of BFS in the seawater-mixed
narrowed in slump value if SP dosage reaches 2% or more, indicating the composite is 5–30% higher than that in the freshwater-mixed compos
lower workability in the cement-seawater mixtures can be thoroughly ite. Li et al. [43] also observed enhanced hydration stimulated by
improved by SP. Furthermore, the improvement was about 12–18% of seawater, which presents as an early occurrence of the main hydration
compressive strength in FW, but even 16–23% of that in SW, which peak for nearly 2 h. Due to the stimulated hydration in seawater-mixed
exhibits the increase of compressive strength by SP is even greater in composite, a slightly lower slump is presented in the BFS-mixed mix
SW. Younis et al. [12] improved the fresh properties of seawater-mixed tures than OPC [25], while a small increase of slump value is determined
concrete to achieve a comparable degree to freshwater-mixed concrete, in some other studies [7,120]. The differences vary only over narrow
by adding a retarder with a concentration of 0.25 L/m3 and an addi limits, proving that the effect of using BFS on the workability of seawater
tional 15% more superplasticizer. Meanwhile, if the excessively rapid mixtures is not significant. Li et al. [29] studied the seawater mixtures
hydration in seawater was eliminated through using the chemical ad incorporated with MK. The result showed that the porosity was
ditives, the mechanical performance can also be effectively improved decreased with a refined pore structure. Analogously, Cheng et al. [121]
[12]. investigated the seawater mixtures incorporated with MK and BFS and
measured that the permeability of concrete was reduced and the me
chanical properties of ITZ were improved, due to the filling effect and
4.2. Supplementary cementing materials
pozzolanic reaction of SCMs [121]. Khatibmasjedi [13] found that
adding 20% FA significantly increased the drying shrinkage from 738
Generally, the use of supplementary cementing materials (SCMs),
μm to 1370 μm in seawater-mixed composite after 60 days, which
such as blast furnace slag (BFS), fly ash (FA), silica fume (SF), and
further aggravates the cracking risk of seawater mixtures. Therefore,
metakaolin (MK) that can modify binder mineralogy has shown a great
attention should be taken when combining use seawater and FA in
influence on durability [112,113], by improving mechanical properties
cement mixtures.
[9,114], reducing chloride diffusivity [115], increasing chloride binding
[105,116,117] and refining pore structure [118,119]. The main mineral
4.2.2. Effect on chloride binding capacity
composition as silica, alumina, and calcite oxide content of SCMs used in
SCMs alter the composition of cement hydrates and, thus, the chlo
related studies is summarized in Fig. 16. For the potential defects in
ride binding capacity [105,116,117]. For example, MK, which com
seawater mixtures, the effect of SCMs has been investigated, which can
prises approximately 40% alumina presents a superior capacity of
be divided into two parts and are discussed separately herein.
binding chloride as the formation of AFm hydrates [29,40]. Shi et al.
[40] added MK (0–6%) in seawater mixtures to determine the chloride
4.2.1. Effect on fresh and hardened properties
binding effect. The result indicates 48–55% of total chloride bound by 3
Numerous studies have investigated the effect of BFS in seawater
days and 67% of that by 28 days. Due to the negligible calcium oxide
composites and are consistent with the finding that BFS increases the
content in MK, it is also suggested that calcium is necessary to stimulate
mechanical properties of concrete mixed with seawater. Table 6 sum
the binding ability of MK, which reveals the MK to lime ratio as the
marized the SCMs effect on the compressive strength in the seawater-
influential factor [66]. Lesser chloride binding, by contrast, can be
mixed composite. Katano et al. [26] used BFS to partly replace OPC
detected in BFS and FA, which contains alumina (10–25%). Dhir et al.
with a ratio ranging from 30% to 70%. The result shows that the
[122] found that at a high BFS replacement level of 67%, the chloride
compressive strength decreases with the increase of BFS. However, a
binding is 5 times that of OPC measured at 5 M chloride in 14 days. The
higher strength can be observed in seawater-mixed composite relative to
high chloride binding for BFS also was verified by Li et al. [43] that
freshwater-mixed composite, and the enhancing effect is even greater at
chloride concentration of 1.75 mg/g can be decreased to 1.46 mg/g and
the early stage, despite the lower strength value. For example, when the
1.33 mg/g for specimens with 25% BFS and 50% BFS, respectively in 3
blending ratio of BFS is 30%, the compressive strength is 37% higher in
days. Nevertheless, it should be noted that there is an optimum BFS
seawater mixtures at 7 days, and the figure reaches 50% or more when
replacement for binding capacity, and when adding BFS to excess, the
the blended ratio is over 50%. The result indicates that seawater can
chloride binding becomes lower than OPC, which is supported by Zibara
promote the hydration of BFS, due to the higher content of alumina than
[66] who even studied the 100% BFS paste. By contrast with BFS, FA
OPC to develop more AFm hydrates. Similar observations have also been
seems too inert to bind chloride at the early stage. Otsuki et al. [120]
identified by Otsuki et al. [25] that the compressive strength ratio has
observed a similar chloride binding capacity in OPC and OPC with 30%
the largest value of 1.62 on the first day of hydration. The test results
12
Fig. 16. Ternary diagram of silica, alumina, and calcium oxide content in
SCMs. Data from the SCMs used in the related studies.
third effect, the decrease in pH resulted from the substitution of OPC,

corresponding to a diluted chloride binding of OPC. However, adding SF
leads to the formation of C–S–H with a lower Ca/Si ratio, which has been
proved to reduce the amount of chloride adsorbed on the surface of
C–S–H [66,91,92]. In FA and BFS, the formation of low Ca/Si C–S–H also
occurs, while the shortages are filled up by higher content of alumina
and results in increased binding [117]. Therefore, it shows a negative in
SF but positive in FA and BFS on enhancing chloride binding of OPC.
5. Mechanism discussion
The seawater-mixed concrete has been studied for over a century

[126], while the general view on using seawater in concrete structure
remains conservative. Still, construction specifications in various
countries can be summarized as mentioned by Neville [10] that
seawater must never be used in reinforced concrete; never, never be
used in prestressed concrete; and preferably not be used in plain con
Fig. 15. Effect of SP on the slump and compressive strength in SW and crete, which often contains some steel. The high chloride content has
FW [28]. always been the trickiest problem in seawater-mixed concrete that
significantly increases the risk of corrosion in reinforcement and struc
FA-replaced blends exposed to seawater for 4 months. Shi et al. [114] ture failure. In this section, the most concerned research problems of
found that however, adding FA reduced the chloride binding instead of seawater mixtures will be discussed in detail, based on the current
increase as also proposed by Cheewaket et al. [123]. Therefore, there is studies.
no clear generic relevance in the chloride binding capacity of FA,
probably due to the large diversity in mineral composition (see Fig. 16) 5.1. Characteristics and deterioration of seawater concrete
and the degree of hydration [66]. Furthermore, it should be noted that
not all of the alumina in fly ash possesses chloride binding ability. For The effect of chloride is the main factor leading to the discrepancy
instance, much of alumina in fly ash is presented in crystalline compo between seawater and conventional mixtures. Due to the chloride pro
nents such as mullite, which cannot form AFm phases during further moting hydration, the rapid exotherm, shorter setting period, and higher
hydration [117]. Another reason is that strätlingite is formed when FA early strength are presented in seawater mixtures undoubtedly. How
or BFS is added as the AFm phase with aluminosilicate anion [AlSi ever, it seems that seawater mixtures are more sensitive to the envi
(OH)8]− , which also presents no chloride binding capacity [50]. ronment: no matter in freshwater curing, seawater curing, or tidal
SF contains almost no aluminate and, hence, has the lowest binding curing, seawater mixtures present lower mechanical properties than
ability and results in greatly reduced chloride binding capacity in OPC those of freshwater mixtures at the late stage, except for the moist curing
[114,117,124]. Nilsson et al. [125] propose the effect of SF on chloride condition. The lower strength development in seawater mixtures may be
binding in three ways: (1) increase of the specific area of pores and explained by the higher water absorption and the significant leaching
C–S–H which increase chloride binding; (2) reactions with C3A effect of alkalis and calcium hydroxide [37]. Furthermore, the shrinkage
consuming the AFm phase which may decrease chloride binding; (3) of seawater mixtures, especially the drying shrinkage, is aggravated by
reactions with alkalis reducing pH, which may increase binding. For the its hygroscopic property, which should be noted: a 17% higher drying
13
Table 6 shrinkage can be observed in seawater-mixed concrete and, more seri

The SCMs effect on the compressive strength in seawater-mixed concretes. ously, over 80% higher drying shrinkage in seawater mixtures with 20%
Concrete W/b Curing Compressive strength (MPa) Refs. of FA, compared to the freshwater-mixed concrete [13]. Hence, the high
type ratio condition risk of shrinkage cracking can also cause deteriorations in seawater
FW SW SW/
FW mixtures.
(%)
OPC 70%+ 0.5 Moist curing 21.52 29.40 1.37 [26]

5.2. Limited chloride binding effect in seawater concrete
BFS 30% (7 d) (7 d) 1.20
31.76 38.09 1.19 It seems feasible to reduce chloride content by cementitious mate
(28 d) (28 d) rials themselves through chemical binding or physical adsorption.
39.94 47.50
However, it is doubtful that the low chlorine required for reinforcement
(91 d) (91 d)
OPC 50%+ 0.5 Moist curing 15.33 23.00 1.50 [26] protection can be reached through the practical binding effect. First,
BFS 50% (7 d) (7 d) 1.13 assuming that all chloride ions have been bound by OPC, there will be
26.09 29.49 0.85 more than 30% of the chloride bound by C–S–H and other cement hy
(28 d) (28 d) drates by physical adsorption (see section 3.4). Compared to the
36.97 31.51
(91 d) (91 d)
chemical binding of chloride forming Friedel’s salt, the physical
OPC 30%+ 0.5 Moist curing 10.76 17.16 1.59 [26] adsorption is reversible [50] that the bound chloride can be released
BFS 70% (7 d) (7 d) 1.01 easily when the external concentration changes. Second, during the
20.23 20.42 0.83 long-term carbonization, the C–S–H phase decomposes inevitably and
(28 d) (28 d)
the chloride binding capacity is significantly reduced. Also, the Friedel’s
27.01 22.61
(91 d) (91 d) salt converts to carbo-AFm gradually, with the chloride substituted no
Ground BFS 0.5 Moist curing 45.35 48.27 1.06 [144]. longer bound afterward. In these two aspects, the chloride binding effect
cement (7 d) (7 d) 1.12 of OPC appears to be temporary and limited, because chloride ions failed
57.51 64.32 to be weeded out from the mixtures throughout the process; even if
(28 d) (28 d)
OPC 80%+ 0.5 Moist curing 20.56 23.55 1.15 [26]
applying some effective SCMs, the risk of chloride corrosion cannot be
FA 20% (7 d) (7 d) 1.06 eliminated in seawater mixtures. Therefore, seawater should not be used
30.43 32.33 1.09 in concrete contains steel reinforcement or components. In other words,
(28 d) (28 d) the research on anti-chloride-corrosion materials is the more significant
35.79 38.95
development trend of seawater concrete.
(91 d) (91 d)
OPC 30%+ 0.5 Moist curing 4.62 (1 7.48 (1 1.62 [25]
BFS 70% d) d) 1.21 5.3. Application of seawater concrete
28.95 35.00 1.08
(7 d) (7 d) 1.05 Now that chloride corrosion has always been a risk in seawater
45.79 49.42
(28 d) (28 d)
mixtures, concrete without steel has a great potential for application,
60.21 63.51 such as plain concrete and fiber-reinforced polymer (FRP) composites
(91 d) (91 d) [21,127–130]. Because no component can be corroded by chloride ions,
OPC 80%+ 0.26 Moist curing 66.5 58.0 0.87 [21] special consideration on corrosion is no longer required in seawater
FA 20% (with SP (28 d) (28 d)
mixtures. The use of seawater can be regarded as adding an accelerator
1%)
OPC 35%+ 0.34 Cured in 38.04 39.22 1.03 [12] of sodium chloride, calcium chloride, and magnesium chloride that
BFS 65% freshwater (3 d) (3 d) 1.05 leads to a rapid development of early strength. Three aspects should also
50.72 41.24 0.90 be noted in seawater mixtures. Firstly, the main chloride source in
(7 d) (7 d) 0.93 seawater is sodium chloride, which is alkaline in a reactive state. Hence,
68.76 62.22
(28 d) (28 d)
the risk of alkali-aggregate reaction in seawater mixtures is increased
73.46 68.10 [131,132]. Secondly, considering the leaching effect on strength
(56 d) (56 d) reduction, it is not advised to use seawater mixtures as long-used
OPC 35%+ 0.34 Cured in 40.0 (3 39.22 0.98 [12] structures in water-level fluctuation zones, or those exposed to rain.
BFS 65% seawater d) (3 d) 0.99
Thirdly, continuous dampness and efflorescence occur on the surface of
53.59 53.07 0.92
(7 d) (7 d) 0.96 the seawater mixtures, due to the moisture absorption in higher salinity.
67.19 61.70 Therefore, in the structures where the appearance is considered or a
(28 d) (28 d) plaster finish is to be applied, the seawater concrete mixtures are not
71.11 68.10 recommended [133].
(56 d) (56 d)
OPC 85%+ 0.2 Moist curing 89.31 91.35 1.02 [43]
SF 15% (with SP (3 d) (3 d) 5.4. Prospects of seawater concrete
2%)
OPC 70%+ 0.2 Moist curing 89.97 91.42 1.02 [43] Since there are many shortages observed in seawater mixtures, it
SF 30% (with SP (3 d) (3 d)
should, cautiously, be worth considering the prospects of the applica
2%)
OPC 75%+ 0.2 Moist curing 84.40 88.45 1.05 [43] tion. The most significant advantage of seawater concrete is to obtain
BFS 25% (with SP (3 d) (3 d) the water easily, rapidly, and locally, which is prominent in some coastal
2%) areas and island areas where the freshwater resource is scarce. Unfor
OPC 50%+ 0.2 Moist curing 80.75 84.59 1.05 [43] tunately, since convenience has been chosen, the performance of con
BFS 50% (with SP (3 d) (3 d)
2%)
crete has to be sacrificed, manifesting seawater as a two-edged sword.
Due to the risk of corrosion and deterioration that cannot be eliminated,
seawater mixtures should never be used in concrete components that are
crucial to structures. The long-term safety of structures should not be
belittled, although using seawater saves manpower and material re
sources. Besides, many researchers believe that using seawater in
14
concrete is helpful to relieve the stress of freshwater shortages, while it is Council (DE150101751, IH200100010 and IH180100010), University
too idealistic. Considering the water used in concrete manufacturing, of Technology Sydney Research Academic Program at Tech Lab (UTS
such as water for cooling clinker, water for washing aggregates, water RAPT), and the University of Technology Sydney Tech Lab Blue Sky
for energy, and water for transportation et al. The mixing water accounts Research Scheme.
only for 10–15% of total water consumption and 2.5–4% of water
withdrawal for concrete production [6]. Therefore, improving the References
durability of concrete for prolonging its service life is an effective way to
alleviate water shortages, not just using seawater directly as a mixture. [1] M.W. Doyle, D.G. Havlick, Infrastructure and the environment, Annu. Rev.
Environ. Resour. 34 (2009) 349–373.
However, seawater concrete is a promising research topic, because [2] P.J. Monteiro, S.A. Miller, A. Horvath, Towards sustainable concrete, Nat. Mater.
technology is progressing through intrepid exploration. Research needs 16 (2017) 698.
are still promoted by the trends of development in seawater mixtures, [3] S.A. Miller, F.C. Moore, Climate and health damages from global concrete
production, Nat. Clim. Change 10 (2020) 439–443.
summarized as follows: firstly, seawater mixtures are available to cover [4] T. Dhondy, A. Remennikov, M.N. Shiekh, Benefits of using sea sand and seawater
the demand of non-crucial and short-term structures, and meanwhile, in concrete: a comprehensive review, Aust. J. Struct. Eng. 20 (2019) 280–289.
specification is needed to standardize the use of seawater, due to the [5] S.A. Miller, A. Horvath, P.J. Monteiro, Readily implementable techniques can cut
annual CO2 emissions from the production of concrete by over 20%, Environ.
varying salinity in different marine areas; secondly, seawater mixtures Res. Lett. 11 (2016), 074029.
can be used in some structural concrete, where FRP is applied instead of [6] S.A. Miller, A. Horvath, P.J. Monteiro, Impacts of booming concrete production
steel to avoid corrosion. Once the problem of the poor modulus of on water resources worldwide, Nature Sustainability 1 (2018) 69–76.
[7] T. Nishida, N. Otsuki, H. Ohara, Z.M. Garba-Say, T. Nagata, Some considerations
elasticity, shear strength, and durability of FRP can be solved, a great
for applicability of seawater as mixing water in concrete, J. Mater. Civ. Eng. 27
potential would be exhibited in the use of seawater mixtures; thirdly, in (2015) B4014004.
areas where freshwater is significantly scarce and seawater has to be [8] N. Otsuki, Research on the influence of chloride on corrosion of the embedded
used in construction, it is feasible to use seawater even in reinforced steel bars in concrete, Report of the port and harbour research institute, Ministry
of Transport 24 (1985) 183–185.
concrete, as long as the chloride concentration is reduced as far as [9] Japan Concrete Institute (JCI-TC-121A), JCI Technical Committee Report on the
possible through desalination combined with using SCMs. Use of Seawater in Concrete, 2015.
[10] A. Neville, Seawater in the mixture, Concr. Int. 23 (2001) 48–51.
[11] A. Younis, U. Ebead, S. Judd, Life cycle cost analysis of structural concrete using
6. Conclusion seawater, recycled concrete aggregate, and GFRP reinforcement, Construct. Build.
Mater. 175 (2018) 152–160.
Massive resource consumption in concrete production has been an [12] A. Younis, U. Ebead, P. Suraneni, A. Nanni, Fresh and hardened properties of
seawater-mixed concrete, Construct. Build. Mater. 190 (2018) 276–286.
increasing environmental concern. Utilizing seawater instead of fresh [13] M. Khatibmasjedi, S. Ramanthan, P. Suraneni, A. Nanni, Shrinkage behavior of
water as mixing water is a feasible measure to alleviate water shortages cementitious mortars mixed with seawater, Adv. Civil Eng. Mater. 8 (2019)
in the coastal and island regions. This paper has critically reviewed the 64–78.
[14] M. Guo, B. Hu, F. Xing, X. Zhou, M. Sun, L. Sui, Y. Zhou, Characterization of the
seawater-induced changes in performance of OPC, specifically in the mechanical properties of eco-friendly concrete made with untreated sea sand and
terms of fresh properties, mechanical properties, transport properties, seawater based on statistical analysis, Construct. Build. Mater. 234 (2020)
shrinkage properties and hydration kinetics. Due to the accelerated early 117339.
[15] P. Sikora, K. Cendrowski, M. Abd Elrahman, S.-Y. Chung, E. Mijowska,
hydration, concrete mixtures with seawater present lower workability,
D. Stephan, The effects of seawater on the hydration, microstructure and strength
higher early strength, denser early microstructure and greater development of Portland cement pastes incorporating colloidal silica, Appl.
shrinkage. The long-term properties of seawater mixtures are more Nanosci. (2019) 1–12.
vulnerable to environmental factors, for example, the slow properties [16] L. Montanari, P. Suraneni, M. Tsui-Chang, M. Khatibmasjedi, U. Ebead, J. Weiss,
A. Nanni, Hydration, pore solution, and porosity of cementitious pastes made
development in underwater or tidal conditions due to the leaching effect with seawater, J. Mater. Civ. Eng. 31 (2019), 04019154.
of alkalis and calcium hydrates. Additionally, according to the effects of [17] P. Li, W. Li, T. Yu, F. Qu, V.W.Y. Tam, Investigation on early-age hydration,
various seawater-related ions, the hydration mechanism in seawater is mechanical properties and microstructure of seawater sea sand cement mortar,
Construct. Build. Mater. 249 (2020) 118776.
dominated by chloride ions. The AFm phase, such as hydroxy-AFm and [18] J. Wang, E. Liu, L. Li, Multiscale investigations on hydration mechanisms in
sulfo-AFm, is the most effective chloride binding phase with the highest seawater OPC paste, Construct. Build. Mater. 191 (2018) 891–903.
binding capacity. The physical adsorption of chloride by C–S–H and [19] Y. Farnam, A. Wiese, D. Bentz, J. Davis, J. Weiss, Damage development in
cementitious materials exposed to magnesium chloride deicing salt, Construct.
other cement hydrates accounts for only 30% of total bound chloride. Build. Mater. 93 (2015) 384–392.
The chloride binding effect of cement hydrates is irreversibly deterio [20] P. Friedel, Sur un chloro-aluminate de calcium hydraté se maclant par
rated when the carbonation occurs, indicating the chloride binding is compression, Bulletin Soc. Franc. Minéral 19 (1897) 122–136.
[21] J. Sena-Cruz, E. Pereira, N. Freitas, E. Pereira, S.R.R. Soares, Bond behaviour
temporary and limited. Although the performance of seawater concrete between GFRP rods and concrete produced with seawater: an experimental
mixtures can be improved through the additive and admixture, the risk research, in: Proceedings of 9th International Conference on Fibre-Reinforced
of chloride corrosion cannot be eliminated and it would significantly Polymer (FRP) Composites in Civil Engineering, 2018, pp. 1–8. CICE2018.
[22] S. Jain, B. Pradhan, Fresh, mechanical, and corrosion performance of self-
limit the scope of its practical applications. Therefore, seawater concrete
compacting concrete in the presence of chloride ions, Construct. Build. Mater.
should not be used in concrete components that are crucial to in 247 (2020) 118517.
frastructures, unless anti-chloride-corrosion materials are applied [23] E. Bachtiar, M. Tjaronge, R. Djamaluddin, V. Sampebulu, Compressive strength
instead or desalination of seawater is conducted. In short, this review and slump flow of self compacting concrete uses fresh water and sea water, ARPN
J. Eng. Appl. Sci. 10 (2015) 2373–2377.
summarizes and compares the sustainable but defective seawater con [24] M.Z.Y. Ting, K.S. Wong, M.E. Rahman, M. Selowara Joo, Mechanical and
crete mixtures and proposes the viewpoints of application, based on the durability performance of marine sand and seawater concrete incorporating
advanced research progress. silicomanganese slag as coarse aggregate, Construct. Build. Mater. 254 (2020)
119195.
[25] N. Otsuki, D. Furuya, T. Saito, Y. Tadokoro, Possibility of sea water as mixing
Declaration of competing interest water in concrete, in: Proceedings of Conference on Our World in Concrete &
Structures, Tokyo Institute of Technology, Japan, 2011.
[26] K. Katano, N. Takeda, Y. Ishizeki, K. Iriya, Properties and Application of Concrete
The authors declare that they have no known competing financial Made with Sea Water and Un-washed Sea Sand, 2013.
interests or personal relationships that could have appeared to influence [27] M. Etxeberria, J.M. Fernandez, J. Limeira, Secondary aggregates and seawater
the work reported in this paper. employment for sustainable concrete dyke blocks production: case study,
Construct. Build. Mater. 113 (2016) 586–595.
[28] L.G. Li, X.Q. Chen, S.H. Chu, Y. Ouyang, A.K.H. Kwan, Seawater cement paste:
Acknowledgment effects of seawater and roles of water film thickness and superplasticizer dosage,
Construct. Build. Mater. 229 (2019) 116862.
All the authors appreciate the supports from the Australian Research
15
[29] Q. Li, H. Geng, Z. Shui, Y. Huang, Effect of metakaolin addition and seawater [61] K. De Weerdt, D. Orsáková, M.R. Geiker, The impact of sulphate and magnesium
mixing on the properties and hydration of concrete, Appl. Clay Sci. 115 (2015) on chloride binding in Portland cement paste, Cement Concr. Res. 65 (2014)
51–60. 30–40.
[30] M.W.T. Erniati, R. Djamaluddin, V. Sampebulu, Compressive Strength and Slump [62] J. Stark, K. Bollmann, Delayed ettringite formation in concrete, Nordic Concrete
Flow of Self Compacting Concrete Uses Fresh Water and Sea Water, 2015. Research -Publications 23 (2000) 4–28.
[31] M.M. Islam, M.S. Islam, M. Al-Amin, M.M. Islam, Suitability of sea water on [63] S.O. Ekolu, M.D.A. Thomas, R.D. Hooton, Pessimum effect of externally applied
curing and compressive strength of structural concrete, J. Civ. Eng. 40 (2012) chlorides on expansion due to delayed ettringite formation: proposed mechanism,
37–45. Cement Concr. Res. 36 (2006) 688–696.
[32] F.M. Wegian, Effect of seawater for mixing and curing on structural concrete, IES [64] H. Hirao, K. Yamada, H. Takahashi, H. Zibara, Chloride binding of cement
J. Part A Civ. Struct. Eng. 3 (2010) 235–243. estimated by binding isotherms of hydrates, J. Adv. Concr. Technol. 3 (2005)
[33] Q. Guo, L. Chen, H. Zhao, J. Admilson, W. Zhang, The effect of mixing and curing 77–84.
sea water on concrete strength at different ages, in: Proceedings of MATEC Web of [65] H.G. Midgley, J.M. Illston, The penetration of chlorides into hardened cement
Conferences, EDP Sciences, 2018, 02004. pastes, Cement Concr. Res. 14 (1984) 546–558.
[34] X. Wang, S. Sadati, P. Taylor, C. Li, X. Wang, A. Sha, Material characterization to [66] H. Zibara, Binding of External Chlorides by Cement Pastes, National Library of
assess effectiveness of surface treatment to prevent joint deterioration from Canada, 2001.
oxychloride formation mechanism, Cement Concr. Compos. 104 (2019) 103394. [67] R. Loser, B. Lothenbach, A. Leemann, M. Tuchschmid, Chloride resistance of
[35] C. Qiao, P. Suraneni, M.T. Chang, J. Weiss, Damage in cement pastes exposed to concrete and its binding capacity – comparison between experimental results and
MgCl2 solutions, Mater. Struct. 51 (2018) 74. thermodynamic modeling, Cement Concr. Compos. 32 (2010) 34–42.
[36] T.U. Mohammed, H. Hamada, T. Yamaji, Performance of seawater-mixed [68] H. Chang, P. Feng, K. Lyu, J. Liu, A novel method for assessing C-S-H chloride
concrete in the tidal environment, Cement Concr. Res. 34 (2004) 593–601. adsorption in cement pastes, Construct. Build. Mater. 225 (2019) 324–331.
[37] M. Khatibmasjedi, Sustainable Concrete Using Seawater and Glass Fiber [69] J. Csizmadia, G. Balázs, F.D. Tamás, Chloride ion binding capacity of
Reinforced Polymer Bars, University of Miami, 2018. aluminoferrites, Cement Concr. Res. 31 (2001) 577–588.
[38] R. Spragg, S. Jones, Y. Bu, Y. Lu, D. Bentz, K. Snyder, J. Weiss, Leaching of [70] T. Sumranwanich, S. Tangtermsirikul, A model for predicting time-dependent
conductive species: implications to measurements of electrical resistivity, Cement chloride binding capacity of cement-fly ash cementitious system, Mater. Struct.
Concr. Compos. 79 (2017) 94–105. 37 (2004) 387.
[39] C. Qiao, P. Suraneni, J. Weiss, Flexural strength reduction of cement pastes [71] A.K. Suryavanshi, J.D. Scantlebury, S.B. Lyon, The binding of chloride ions by
exposed to CaCl2 solutions, Cement Concr. Compos. 86 (2018) 297–305. sulphate resistant portland cement, Cement Concr. Res. 25 (1995) 581–592.
[40] Z. Shi, Z. Shui, Q. Li, H. Geng, Combined effect of metakaolin and sea water on [72] F.P. Glasser, J. Marchand, E. Samson, Durability of concrete—degradation
performance and microstructures of concrete, Construct. Build. Mater. 74 (2015) phenomena involving detrimental chemical reactions, Cement Concr. Res. 38
57–64. (2008) 226–246.
[41] P. Claisse, Transport properties of concrete, Concr. Int. 27 (2005) 43–48. [73] P. Suraneni, V.J. Azad, B.O. Isgor, W.J. Weiss, Calcium oxychloride formation in
[42] A. Radlinska, F. Rajabipour, B. Bucher, R. Henkensiefken, G. Sant, J. Weiss, pastes containing supplementary cementitious materials: thoughts on the role of
Shrinkage mitigation strategies in cementitious systems: a closer look at cement and supplementary cementitious materials reactivity, RILEM Technical
differences in sealed and unsealed behavior, Transport. Res. Rec. 2070 (2008) Letters 1 (2016) 24–30.
59–67. [74] C. Jones, S. Ramanathan, P. Suraneni, W.M. Hale, Calcium oxychloride: a critical
[43] H. Li, N. Farzadnia, C. Shi, The role of seawater in interaction of slag and silica review of the literature surrounding the formation, deterioration, testing
fume with cement in low water-to-binder ratio pastes at the early age of procedures, and recommended mitigation techniques, Cement Concr. Compos.
hydration, Construct. Build. Mater. 185 (2018) 508–518. (2020) 103663.
[44] A. Hajibabaee, M.T. Ley, Impact of wet and sealed curing on curling in cement [75] F.D. Tamas, Acceleration and Retardation of Portland Cement Hydration by
paste beams from drying shrinkage, ACI Mater. J. 112 (2015). Additives, Highway Research Board Special Report, 1966.
[45] S.S. Park, S.-J. Kwon, H.-W. Song, Analysis technique for restrained shrinkage of [76] P. Brown, J. Bothe, The system CaO-Al2O3-CaCl2-H2O at 23±2 ◦ C and the
concrete containing chlorides, Mater. Struct. 44 (2011) 475–486. mechanisms of chloride binding in concrete, Cement Concr. Res. 34 (2004)
[46] F. Autelitano, M. Rinaldi, F. Giuliani, Winter highway maintenance strategies: are 1549–1553.
all the sodium chloride salts the same? Construct. Build. Mater. 226 (2019) [77] Y. Farnam, B. Zhang, J. Weiss, Evaluating the use of supplementary cementitious
945–952. materials to mitigate damage in cementitious materials exposed to calcium
[47] T. Matschei, B. Lothenbach, F. Glasser, The AFm phase in Portland cement, chloride deicing salt, Cement Concr. Compos. 81 (2017) 77–86.
Cement Concr. Res. 37 (2007) 118–130. [78] K. Peterson, G. Julio-Betancourt, L. Sutter, R.D. Hooton, D. Johnston,
[48] A.K. Suryavanshi, J.D. Scantlebury, S.B. Lyon, Mechanism of Friedel’s salt Observations of chloride ingress and calcium oxychloride formation in laboratory
formation in cements rich in tri-calcium aluminate, Cement Concr. Res. 26 (1996) concrete and mortar at 5◦ C, Cement Concr. Res. 45 (2013) 79–90.
717–727. [79] C. Qiao, P. Suraneni, J. Weiss, Damage in cement pastes exposed to NaCl
[49] T. Matschei, B. Lothenbach, F.P. Glasser, The AFm phase in Portland cement, solutions, Construct. Build. Mater. 171 (2018) 120–127.
Cement Concr. Res. 37 (2007) 118–130. [80] C. Qiao, P. Suraneni, T. Nathalene Wei Ying, A. Choudhary, J. Weiss, Chloride
[50] I. Galan, F.P. Glasser, Chloride in cement, Adv. Cement Res. 27 (2015) 63–97. binding of cement pastes with fly ash exposed to CaCl2 solutions at 5 and 23 ◦ C,
[51] M.R. Jones, D.E. Macphee, J.A. Chudek, G. Hunter, R. Lannegrand, R. Talero, S. Cement Concr. Compos. 97 (2019) 43–53.
N. Scrimgeour, Studies using 27Al MAS NMR of AFm and AFt phases and the [81] D. Damidot, U. Birnin-Yauri, F. Glasser, Thermodynamic investigation of the CaO-
formation of Friedel’s salt, Cement Concr. Res. 33 (2003) 177–182. Al2O3-CaCl 2-H2O system at 25 ◦ C and the influence of Na2O, Cemento 91 (1994)
[52] S. Rasheeduzzafar, Ehtesham Hussain, S.S. Al-Saadoun, Effect of cement 243, 243.
composition on chloride binding and corrosion of reinforcing steel in concrete, [82] G.A. Julio-Betancourt, Effect of De-icer and Anti-icer Chemicals on the Durability,
Cement Concr. Res. 21 (1991) 777–794. Microstructure, and Properties of Cement-Based Materials, University of Toronto,
[53] U.A. Birnin-Yauri, F.P. Glasser, Friedel’s salt, Ca2Al(OH)6(Cl,OH)⋅2H2O: its solid 2009.
solutions and their role in chloride binding, Cement Concr. Res. 28 (1998) [83] L. Sutter, K. Peterson, S. Touton, T. Van Dam, D. Johnston, Petrographic evidence
1713–1723. of calcium oxychloride formation in mortars exposed to magnesium chloride
[54] Y. Hosokawa, K. Yamada, B. Johannesson, L.-O. Nilsson, Models for chloride ion solution, Cement Concr. Res. 36 (2006) 1533–1541.
bindings in hardened cement paste using thermodynamic equilibrium [84] A. Ghazy, M.T. Bassuoni, Resistance of concrete to different exposures with
calculations, in: 2nd International RILEM Symposium on Advances in Concrete chloride-based salts, Cement Concr. Res. 101 (2017) 144–158.
through Science and Engineering, 2006. [85] I. Galan, L. Perron, F.P. Glasser, Impact of chloride-rich environments on cement
[55] M. Frias, S. Goñi, R. García, R. Vigil de La Villa, Seawater effect on durability of paste mineralogy, Cement Concr. Res. 68 (2015) 174–183.
ternary cements, Synergy of chloride and sulphate ions, Composites Part B: [86] H. Egüez Álava, N. De Belie, G. De Schutter, Proposed mechanism for the
Engineering 46 (2013) 173–178. formation of oxychloride crystals during sodium chloride application as a deicer
[56] K. De Weerdt, H. Justnes, The effect of sea water on the phase assemblage of salt in carbonated concrete, Construct. Build. Mater. 109 (2016) 188–197.
hydrated cement paste, Cement Concr. Compos. 55 (2015) 215–222. [87] H.G. Smolczyk, Chemical reactions of strong chloride-solution with concrete, in:
[57] Y. Cao, L. Guo, B. Chen, J. Wu, Thermodynamic modelling and experimental Proceedings of the 5th International Congress on the Chemistry of Cement 1968,
investigation on chloride binding in cement exposed to chloride and chloride- 1968, pp. 274–280.
sulfate solution, Construct. Build. Mater. 246 (2020) 118398. [88] E. Theissing, The combining of sodium chloride and calcium chloride by the
[58] Y. Farnam, S. Dick, A. Wiese, J. Davis, D. Bentz, J. Weiss, The influence of calcium hardened portland cement compounds C3S, C2S, C3A and C4AF, in: Proceedings
chloride deicing salt on phase changes and damage development in cementitious of the 7th International Congress on the Chemistry of Cement 1980, 1980,
materials, Cement Concr. Compos. 64 (2015) 1–15. pp. 823–828.
[59] Z. Shi, M.R. Geiker, B. Lothenbach, K. De Weerdt, S.F. Garzón, K. Enemark- [89] E. Yogarajah, T. Nawa, K. Kurumisawa, Influence of surface electrical properties
Rasmussen, J. Skibsted, Friedel’s salt profiles from thermogravimetric analysis of C–S–H on chloride binding in slag-blended cementitious materials, J. Mater.
and thermodynamic modelling of Portland cement-based mortars exposed to Civ. Eng. 30 (2018), 04018064.
sodium chloride solution, Cement Concr. Compos. 78 (2017) 73–83. [90] Y. Zhou, D. Hou, J. Jiang, P. Wang, Chloride ions transport and adsorption in the
[60] M. Balonis, B. Lothenbach, G. Le Saout, F.P. Glasser, Impact of chloride on the nano-pores of silicate calcium hydrate: experimental and molecular dynamics
mineralogy of hydrated Portland cement systems, Cement Concr. Res. 40 (2010) studies, Construct. Build. Mater. 126 (2016) 991–1001.
1009–1022. [91] J.J. Beaudoin, V.S. Ramachandran, R.F. Feldman, Interaction of chloride and C-S-
H, Cement Concr. Res. 20 (1990) 875–883.
16
[92] S. Yoon, J. Ha, S.R. Chae, D.A. Kilcoyne, P.J.M. Monteiro, X-ray [121] S. Cheng, Z. Shui, T. Sun, Y. Huang, K. Liu, Effects of seawater and supplementary
spectromicroscopic study of interactions between NaCl and calcium silicate cementitious materials on the durability and microstructure of lightweight
hydrates, Mag. Concr. Res. 66 (2014) 141–149. aggregate concrete, Construct. Build. Mater. 190 (2018) 1081–1090.
[93] H. Friedmann, O. Amiri, A. Aït-Mokhtar, Physical modeling of the electrical [122] R.K. Dhir, M.A.K. El-Mohr, T.D. Dyer, Chloride binding in GGBS concrete, Cement
double layer effects on multispecies ions transport in cement-based materials, Concr. Res. 26 (1996) 1767–1773.
Cement Concr. Res. 38 (2008) 1394–1400. [123] T. Cheewaket, C. Jaturapitakkul, W. Chalee, Long term performance of chloride
[94] G. Plusquellec, A. Nonat, Interactions between calcium silicate hydrate (C-S-H) binding capacity in fly ash concrete in a marine environment, Construct. Build.
and calcium chloride, bromide and nitrate, Cement Concr. Res. 90 (2016) 89–96. Mater. 24 (2010) 1352–1357.
[95] S. Yoon, J. Ha, S.R. Chae, D.A. Kilcoyne, Y. Jun, J.E. Oh, P.J. Monteiro, Phase [124] P. Mangat, B. Molloy, Chloride binding in concrete containing PFA, gbs or silica
changes of monosulfoaluminate in NaCl aqueous solution, Materials 9 (2016) fume under sea water exposure, Mag. Concr. Res. 47 (1995) 129–141.
401. [125] L. Nilsson, E. Poulsen, P. Sandberg, H.E. Sørensen, O. Klinghoffer, HETEK,
[96] S. Levitus, R.A. Locarnini, T.P. Boyer, A.V. Mishonov, J.I. Antonov, H.E. Garcia, Chloride Penetration into Concrete, State-Of-The-Art, Transport Processes,
O.K. Baranova, M.M. Zweng, D.R. Johnson, D. Seidov, World Ocean Atlas, vol. Corrosion Initiation, Test Methods and Prediction Models, 1996, pp. 909–4288.
2009, 2010. Denmark.
[97] F. Barberon, V. Baroghel-Bouny, H. Zanni, B. Bresson, L. Malosse, Z. Gan, [126] D.A. Abrams, Tests of Impure Waters for Mixing Concrete, Structural Materials
Interactions between chloride and cement-paste materials, Magn. Reson. Imag. 23 Research Laboratory, 1924.
(2005) 267–272. [127] J. Xiao, C. Qiang, A. Nanni, K. Zhang, Use of sea-sand and seawater in concrete
[98] D. Sugiyama, Chemical alteration of calcium silicate hydrate (C–S–H) in sodium construction: current status and future opportunities, Construct. Build. Mater. 155
chloride solution, Cement Concr. Res. 38 (2008) 1270–1275. (2017) 1101–1111.
[99] J. Tritthart, Chloride binding in cement II. The influence of the hydroxide [128] Y. Li, X. Zhao, R.S. Raman, Mechanical properties of seawater and sea sand
concentration in the pore solution of hardened cement paste on chloride binding, concrete-filled FRP tubes in artificial seawater, Construct. Build. Mater. 191
Cement Concr. Res. 19 (1989) 683–691. (2018) 977–993.
[100] K. De Weerdt, A. Colombo, L. Coppola, H. Justnes, M.R. Geiker, Impact of the [129] A. Ahmed, S. Guo, Z. Zhang, C. Shi, D. Zhu, A review on durability of fiber
associated cation on chloride binding of Portland cement paste, Cement Concr. reinforced polymer (FRP) bars reinforced seawater sea sand concrete, Construct.
Res. 68 (2015) 196–202. Build. Mater. 256 (2020) 119484.
[101] Z. Shi, M.R. Geiker, K. De Weerdt, T.A. Østnor, B. Lothenbach, F. Winnefeld, [130] S. Soares, N. Freitas, E. Pereira, E. Nepomuceno, E. Pereira, J. Sena-Cruz,
J. Skibsted, Role of calcium on chloride binding in hydrated Portland Assessment of GFRP bond behaviour for the design of sustainable reinforced
cement–metakaolin–limestone blends, Cement Concr. Res. 95 (2017) 205–216. seawater concrete structures, Construct. Build. Mater. 231 (2020) 117277.
[102] P. Lambert, C.L. Page, N.R. Short, Pore solution chemistry of the hydrated system [131] M. Kawamura, K. Takeuchi, Alkali-silica reaction and pore solution composition
tricalcium silicate/sodium chloride/water, Cement Concr. Res. 15 (1985) in mortars in sea water, Cement Concr. Res. 26 (1996) 1809–1819.
675–680. [132] P. Nixon, C. Page, I. Canham, R. Bollinghaus, Influence of sodium chloride on
[103] A. Machner, M. Zajac, M. Ben Haha, K.O. Kjellsen, M.R. Geiker, K. De Weerdt, alkali-silica reaction, Adv. Cement Res. 1 (1988) 99–106.
Chloride-binding capacity of hydrotalcite in cement pastes containing dolomite [133] A.M. Neville, Properties of Concrete, Longman, London, 1995.
and metakaolin, Cement Concr. Res. 107 (2018) 163–181. [134] J. Hamad, C. Ha, M.D. Kennedy, G.L. Amy, Application of ceramic membranes for
[104] L. Sutter, T. Van Dam, K.R. Peterson, D.P. Johnston, Long-term effects of seawater reverse osmosis (SWRO) pre-treatment, Desalination and Water
magnesium chloride and other concentrated salt solutions on pavement and Treatment 51 (2013) 4881–4891.
structural Portland cement concrete, Transport. Res. Rec. 1979 (2006) 60–68. [135] C. Fritzmann, J. Löwenberg, T. Wintgens, T. Melin, State-of-the-art of reverse
[105] H. Justnes, A review of chloride binding in cementitious systems, 21, Nordic osmosis desalination, Desalination 216 (2007) 1–76.
Concrete Research-Publications, 1998, pp. 48–63. [136] Y. Li, X. Zhao, R.R. Singh, S. Al-Saadi, Experimental study on seawater and sea
[106] Q. Yuan, C. Shi, G. De Schutter, K. Audenaert, D. Deng, Chloride binding of sand concrete filled GFRP and stainless steel tubular stub columns, Thin-Walled
cement-based materials subjected to external chloride environment – a review, Struct. 106 (2016) 390–406.
Construct. Build. Mater. 23 (2009) 1–13. [137] K. De Weerdt, B. Lothenbach, M.R. Geiker, Comparing chloride ingress from
[107] T. Luping, L.-O. Nilsson, Chloride binding capacity and binding isotherms of OPC seawater and NaCl solution in Portland cement mortar, Cement Concr. Res. 115
pastes and mortars, Cement Concr. Res. 23 (1993) 247–253. (2019) 80–89.
[108] U. Birnin-Yauri, F. Glasser, Friedel’s salt, Ca2Al (OH)6 (Cl, OH)⋅2H2O: its solid [138] C. Yu, Q. Wu, J. Yang, Effect of seawater for mixing on properties of potassium
solutions and their role in chloride binding, Cement Concr. Res. 28 (1998) magnesium phosphate cement paste, Construct. Build. Mater. 155 (2017)
1713–1723. 217–227.
[109] Y. Elakneswaran, T. Nawa, K. Kurumisawa, Electrokinetic potential of hydrated [139] A. Yeginobali, Sulfate resistance of mortars mixed with sea waters, in:
cement in relation to adsorption of chlorides, Cement Concr. Res. 39 (2009) Proceedings of the 3rd International Conference on Durability of Building
340–344. Materials and Components, 1984, pp. 55–65.
[110] M.V.A. Florea, H.J.H. Brouwers, Chloride binding related to hydration products: [140] R.D. Browne, A.F. Baker, The reinforcement of structural concrete in a marine
Part I: ordinary Portland Cement, Cement Concr. Res. 42 (2012) 282–290. environment, in: F.D. Lydon (Ed.), Developments in Concrete Technology-I,
[111] H.F.W. Taylor, Cement Chemistry, Academic Press, 1992. Applied Science Publishers, London, 1979.
[112] K.A. Gruber, T. Ramlochan, A. Boddy, R.D. Hooton, M.D.A. Thomas, Increasing [141] A. Cwirzen, P. Sztermen, K. Habermehl-Cwirzen, Effect of Baltic seawater and
concrete durability with high-reactivity metakaolin, Cement Concr. Compos. 23 binder type on frost durability of concrete, J. Mater. Civ. Eng. 26 (2014) 283–287.
(2001) 479–484. [142] ASTM D1141 - 98, Standard Practice for the Preparation of Substitute Ocean
[113] Y.-S. Choi, J.-G. Kim, K.-M. Lee, Corrosion behavior of steel bar embedded in fly Water, ASTM International, 2013.
ash concrete, Corrosion Sci. 48 (2006) 1733–1745. [143] O. Akinkurolere, C. Jiang, O. Shobola, The influence of salt water on the
[114] X. Shi, Z. Yang, Y. Liu, D. Cross, Strength and corrosion properties of Portland compressive strength of concrete, J. Eng. Appl. Sci. 2 (2007) 412–415.
cement mortar and concrete with mineral admixtures, Construct. Build. Mater. 25 [144] M. Etxeberria, A. Gonzalez-Corominas, P. Pardo, Influence of seawater and blast
(2011) 3245–3256. furnace cement employment on recycled aggregate concretes’ properties,
[115] M.D.A. Thomas, P.B. Bamforth, Modelling chloride diffusion in concrete: effect of Construct. Build. Mater. 115 (2016) 496–505.
fly ash and slag, Cement Concr. Res. 29 (1999) 487–495. [145] A. Mesbah, J.P. Rapin, M. François, C. Cau-dit-Coumes, F. Frizon, F. Leroux,
[116] X. Shi, N. Xie, K. Fortune, J. Gong, Durability of steel reinforced concrete in G. Renaudin, Crystal structures and phase transition of cementitious Bi-anionic
chloride environments: an overview, Construct. Build. Mater. 30 (2012) 125–138. AFm-(Cl− , CO32− ) compounds, J. Am. Ceram. Soc. 94 (2011) 261–268.
[117] M.D.A. Thomas, R.D. Hooton, A. Scott, H. Zibara, The effect of supplementary [146] C. Chalhoub, R. François, M. Carcasses, Determination of chloride threshold
cementitious materials on chloride binding in hardened cement paste, Cement initiating corrosion: a new set-up taking the localized aspect of corrosion into
Concr. Res. 42 (2012) 1–7. account, Cement Concr. Res. 124 (2019) 105825.
[118] J. Skibsted, R. Snellings, Reactivity of supplementary cementitious materials [147] D. Damidot, F.P. Glasser, Thermodynamic investigation of the
(SCMs) in cement blends, Cement Concr. Res. 124 (2019) 105799. CaO⋅Al2O3⋅CaSO4⋅H2O system at 50 ◦ C and 85 ◦ C, Cement Concr. Res. 22 (1992)
[119] O. Linderoth, P. Johansson, L. Wadsö, Development of pore structure, moisture 1179–1191.
sorption and transport properties in fly ash blended cement-based materials, [148] D. Hou, T. Li, Q. Han, J. Zhang, Insight on the sodium and chloride ions
Construct. Build. Mater. 261 (2020) 120007. adsorption mechanism on the ettringite crystal: structure, dynamics and
[120] N. Otsuki, T. Nishida, C. Yi, T. Nagata, H. Ohara, Effect of Blast Furnace Slag interfacial interaction, Comput. Mater. Sci. 153 (2018) 479–492.
Powder and Fly Ash on Durability of Concrete Mixed with Seawater, 2014.
17

Review

Uploaded by

Copyright:

Available Formats

Review

Uploaded by

Document Information

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

Review

Uploaded by

Copyright:

Available Formats

Cement and Concrete Composites 121 (2021) 104100

Contents lists available at ScienceDirect

Cement and Concrete Composites

Development of sustainable concrete incorporating seawater: A critical

1. Introduction significantly. In this case, utilizing seawater as an alternative to fresh­

OPC paste 0.25 (with SP 54 46 0.85 [28]

Note: SP is superplasticizer; FW denoted freshwater mixtures; SW represents

instance, the strength of cement composites can be varied after being

OPC paste 0.20 Moist curing 84.7 (3 d) 87.3 (3 d) 1.03 [43]

0.35 (with SP 0–2%) 89.0–99.1 (28 d) 90.8–105.5 (28 d) 1.02–1.05

OPC mortar 0.40 Moist curing 42.1 (3 d) 49.1 (3 d) 1.17 [17]

Note: SP: Superplasticizer, FW: Freshwater-mixed composite, SW: Seawater-mixed composites.

Fig. 4. The differential pore volume of concrete at 3, 7 and 28 days between

During the cement hydration or drying process of concrete,

Fig. 7. Drying shrinkage and mass loss of mortar mixtures [13].

of hardened cement paste formed in seawater [16,45], where the

3. Effect of seawater-related ions on hydration

Based on the existing studies, the effect of some seawater-related ions

3.1. Chloride ion

As the most influential ion in seawater on concrete properties,

Cb, sulfo− = 51.89⋅Cf0.58 (3)

where the molecular formula of sulfo-AFm is C4 ASH10 at 11% relative

Cb, C− S− H, Z = 6.65⋅Cf0.334 ⋅δC3 S + 7.89⋅Cf0.136 ⋅δC2 S (9)

3.4.4. Chloride binding in seawater

Fig. 13. Distribution of bound chloride in OPC hydrates in seawater [110].

third effect, the decrease in pH resulted from the substitution of OPC,

The seawater-mixed concrete has been studied for over a century

Table 6 shrinkage can be observed in seawater-mixed concrete and, more seri­

OPC 70%+ 0.5 Moist curing 21.52 29.40 1.37 [26]

You might also like

1. Introduction significantly. In this case, utilizing seawater as an alternative to fresh

Table 6 shrinkage can be observed in seawater-mixed concrete and, more seri