Construction and Building Materials 230 (2020) 116934

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Degradation mechanisms of Portland cement mortar under seawater

attack and drying-wetting cycles
Shukai Cheng a,b, Zhonghe Shui a, Xu Gao c,⇑, Rui Yu a, Tao Sun a,c, Cheng Guo a,b, Yun Huang b
a
State Key Laboratory of Silicate Materials for Architectures, Wuhan University of Technology, Wuhan 430070, China
b
School of Materials Science and Engineering, Wuhan University of Technology, Wuhan 430070, China
c
School of Civil Engineering and Architecture, Wuhan University of Technology, Wuhan 430070, China

h i g h l i g h t s

The degradation mechanisms of mortar exposed to multiple ion attack and drying-wetting cycle are investigated.

The chloride, sulfate and magnesium profiles in marine tidal zone are identified.

The effect of multiple ion on the physical damages is investigated.

The synergetic effect of multiple ion on microstructure characteristics is addressed.

a r t i c l e i n f o a b s t r a c t

Article history: This study investigates the degradation mechanisms of Portland cement mortar under multiple salts
Received 24 March 2019 attack and drying-wetting cycle condition. The results indicate that the sulfate combined with chloride
Received in revised form 8 September 2019 salts retard the chloride diffusion and mitigate the deterioration of mortar under drying-wetting cycles.
Accepted 9 September 2019
However, the presence of sulfate, magnesium and chloride ion result in increased consumption of port-
landite and causes the decalcification of C-S-H, thus aggravating the degradations. Microstructure anal-
ysis shows that the sulfate ion in chloride solution refines the pore structure, while the presence of
Keywords:
sulfate and magnesium ion in chloride environment increases the porosity and pore volume, which accel-
Degradation
Corrosive ions
erates the ingress of corrosive ions.
Drying-wetting cycles Ó 2019 Elsevier Ltd. All rights reserved.
Microstructure
Porosity

1. Introduction Concerning the chemical attacks, the degradation of reinforced

Reinforced concrete is one of the most widely used construction
materials that are applied in marine and offshore engineering.
However, the degradation of concrete structures in harsh marine
environment has been frequently reported to cause serious safety
issues and massive repairing costs [1–3]. The degradation is a com-
plicated process that combines both chemical and physical actions.
The chemical corrosion mainly includes chloride induced rein-
forcement corrosion [4–7], exfoliation caused by magnesium salt
[8,9] and volume expansion due to sulfate attack [10–12]. Mean-
while, the physical processes such as freeze-thaw cycle, drying-
wetting cycle and salt crystallization lead to dissolution, expansion
and cracks of concrete [13–16].
⇑ Corresponding author.
E-mail address: x.gao@whut.edu.cn (X. Gao).
concrete induced by chloride is one of the most severe problems
[7,17]. Numerous studies have explored the erosion of concrete
by chlorides. Moreover, the simultaneous presence of corrosive
ions such as chloride, sulfate and magnesium in seawater even
shows a combined effect on chemical corrosion [18–20]. It has
been found that the sulfate ion obviously decreases the chloride
binding capacity, resulting in the conversion of part of AFm phases
to ettringite and the increase of chloride penetration in concrete
[21,22]. The sulfate ions can inhibit the formation of Friedel’s salt
and release bound chloride, which increase the free chloride con-
tent of pore solution. The magnesium ion in chloride environment
decreases the pH of pore solution and increases the chloride bind-
ing capacity of cement matrix [23]. In addition, the presence of
chloride also alters the mechanism of sulfate attack, while the
presence of magnesium aggravates the sulfate erosion under the
attack of compound salt solution [9,12]. Since the chemical degra-
dations of concrete under multiple salts condition are complex, it is

https://doi.org/10.1016/j.conbuildmat.2019.116934
0950-0618/Ó 2019 Elsevier Ltd. All rights reserved.
2 S. Cheng et al. / Construction and Building Materials 230 (2020) 116934
necessary to carry out systematical investigation regarding this
issue.
On the other hand, the physical degradation of concrete in mar-
ine environment is a non-ignorable factor as it can accelerate the
chemical attacks. Typical physical degradations such as leaching,
wave erosion, repeated drying and wetting usually lead to the dis-
solution of hydration products, the increase of porosity and the
occurrence of surface crack [16,24,25]. For instance, it is reported
that the chloride content of concrete is higher under wetting-
drying cycles compared with that under the immersion condition,
which increases the risk of steel corrosion [13,14]. The sulfate
attack in concrete under drying-wetting cycle shows a higher
degradation rate than that without drying-wetting cycle [15].
The tidal zone in marine environment is the most serious damage
area of concrete structures, where usually presents the most com-
plicated degradations including repeated drying-wetting and mul-
tiple salts attacks [18,20,26–29]. However, the degradation
mechanisms of cementitious materials under multiple ions attack
and drying-wetting cycles have not been studied in detail.
This work highlights the degradation mechanisms of Portland
cement mortar exposed to multiple salts solutions under the
drying-wetting zone. First of all, the degradation behaviors includ-
ing mass changes and relative dynamic modulus of elasticity of
specimens exposed to multiple salt solutions are investigated.
Then the multiple corrosive ion induced phase changes under
drying-wetting cycles are evaluated. Lastly, the pore structure
characteristics of mortar samples under different corrosive envi-
ronments are identified by X-ray computed tomography (X-ray
CT) and mercury intrusion porosimetry (MIP).
2. Materials and methods
2.1. Materials and mix proportions
The Portland cement (strength class 42.5) was used to produce
the mortar specimens according to European standard EN 197-1
[30], its chemical composition was determined by X-ray fluores-
cence (XRF), shown in Table 1. The mineral composition of cement
was measured by quantitative XRD and presented in Table 2. The
analytical reagents sodium chloride, hexahydrate magnesium
chloride and anhydrous sodium sulfate were utilized to simulate
the major elements in seawater.
Mortar mixtures with water to cement (w/c) ratio of 0.5 and
sand to binder ratio of 3 were prepared according to Chinese stan-
dard GB/T 29756-2013 [31]. Specimens were cast and placed in
steel molds of 40  40  160 mm3, and then demoulded after
1 day and cured under the condition of 20 ± 2 °C and 98 ± 2% rela-
tive humidity for 7 days. Mortar specimens were also prepared and
filled into 50 mL centrifuge tube, then stored in standard curing
chamber. The cement pastes with the same w/c ratio were pre-
pared for microstructure analysis.
Table 1
Chemical composition of Portland cement by X-ray
fluorescence.
Constituents (wt.%)
SiO2 21.76
Al2O3 5.83
CaO 62.69
MgO 1.72
Fe2O3 3.41
SO3 1.09
K2O 1.22
Na2O 0.07
LOI 1.11
Table 2
The main mineral composition of
cement by quantitative XRD analysis.
Mineral composition (%)
C3S 57.63
C2S 13.97
C3A 8.48
C4AF 9.84
2.2. Experiment programs
The physical and chemical degradation of samples in the marine
tidal zone was simulated by using a self-made drying-wetting
cycle test chamber. Four corrosive solutions were used: NaCl solu-
tion (N), NaCl + MgCl2 solution (N + M), NaCl + Na2SO4 solution (N
+ S) and NaCl + MgCl2 + Na2SO4 solution (N + M + S). Compared
with the seawater, the detailed composition of each corrosive solu-
tion was presented in Table 3. All solutions were having a fixed
chloride concentration of 0.56 mol/L, and they were stored in four
self-regulating polymethyl methacrylate boxes.
The drying-wetting cycle was carried out as follows: Firstly,
samples were immersed in corrosive solution for 18 h, and then
dried for 0.5 h in the air. Afterwards, the samples were dried for
5 h at temperature of 60 °C and followed by cooling down at the
room temperature for 0.5 h. Thus, the 24 h represents one
drying-wetting cycle. The scheme of drying-wetting cycle chamber
is shown in the Fig. 1. In order to maintain concentration stability
of all ions, the exposed solutions were replaced every two weeks.
2.3. Testing methods
2.3.1. Mass changes
The mass changes after different exposure times can be used to
characterize the surface changes of specimen, which can be
regarded as an indicator of physical degradation. The mass change
of specimen at different exposure time can be calculated by Eq. (1)
according to the Chinese standard GB/T50082-2009 [32].
Gt  G0
DW ¼  100% ð1Þ
G0
where DW is the mass change rate of specimens after corrosion (%);
G0 is the mass of specimen after erosion (g); Gt is the mass of spec-
imen before erosion (g).
2.3.2. Relative dynamic modulus of elasticity
A nonmetal ultrasonic analyzer was used to determine the
modulus of elasticity of mortar samples after different erosion
cycles. The measurement length in this study is 160 mm. The
acoustic time value is converted into the relative dynamic modulus
of elasticity (Erd) by Eq. (2) according to the Chinese standard GB/T
50082-2009.
Ed ¼ ð1þv Þð1
12v Þqv
¼ ð1þvðÞ1
ð12v ÞqL
2 2
v v Þt2
2 t2
ð2Þ
Erd ¼ EEdn ¼ VV n2 ¼ t02
d0 0 n
where Ed and Erd are the dynamic modulus of elasticity (GPa) and
the relative dynamic modulus of elasticity, respectively; V repre-
sents the ultrasonic velocity of specimen (m/s); q denotes the den-
sity of specimen (kg/m3); v is the Poisson’s ratio; V 0 is initial
ultrasonic velocity of specimen before erosion (m/s); V n is the ultra-
sonic velocity of specimen after erosion (m/s); t 0 denotes the initial
ultrasonic time of specimen before erosion (ls); t n represents the
ultrasonic time of specimen after erosion (ls).
 S. Cheng et al. / Construction and Building Materials 230 (2020) 116934 3

Table 3
Detailed composition of the exposure solutions.

Conc. [M] NaCl MgCl2 Na2SO4 Cl conc. [M]

NaCl 0.56 – – 0.56
NaCl + MgCl2 0.46 0.05 – 0.56
NaCl + Na2SO4 0.56 – 0.028 0.56
NaCl + MgCl2 + Na2SO4 0.56 0.05 0.028 0.56
Seawater solution [21] 0.546 0.05 0.028 0.56

2 2.3.5. Phase composition

1
3
The exposed surface (0–5 mm) form the cement paste was char-
4 1. Test chamber acterized by X-Ray diffraction (XRD). The Empyrean X-ray diffrac-
2. Vent hole tometer produced by PANalytical B.V. (Netherlands) was
5 performed for the test experiment using Cu Ka = 1.5406 Å, step
3. Inject hole
size of 0.019°, measuring time 141.804 s/step, start position 5°
6
4. Filter net and end position 70°. The thermogravimetric analysis (TG) was
5. Fixed trestle applied by using a Netzsch STA449F3. The samples were heated
7
10
9 6. Solenoid valve from 30 °C to 1000 °C at a rate of 10 °C/min while the oven was
7. Reservoir chamber purged under an inert atmosphere with N2 flow of 30 mL/min.
8. Drain hole
9. Draught fan 2.3.6. Pore characteristics
10. Water pump X-ray CT is a nondestructive testing technology, which can
8 11. Power supply detect cracks and pore distribution of cement matrix and obtain
the internal damage of specimen by rotating the sample at the
range of 0-360°. The pore distribution of mortar specimens was
11
analyzed by Zeiss Xradia 510 versa X-ray CT. The sample size is
Fig. 1. The schematic of drying-wetting cycles chamber. 40 mm  40 mm  30 mm. The effective capillary pore volume
and pore size distribution of mortar samples was measured by
MIP. Specimens (1–2 g) were soaked in absolute ethyl alcohol,
2.3.3. Total chloride content and then dried in a vacuum atmosphere at 40 ± 2 °C for 12 h before
The specimens in the centrifuge tubes were analyzed after 30, testing.
60, 90 and 120 days of exposure. The exposed surfaces were used
for profile grinding according to the following intervals: 0–5, 5–
10, 10–15 and 15–20 mm. Then, the ground samples were passed 3. Results and discussion
through a sieve of 0.15 mm and stored in plastic sample sack and
placed in a desiccator until analysis. Following the technical spec- 3.1. Mass changes
ification for chloride ion content test in concrete (JGJ/T 322-2013),
the collected powder (5 g) was dissolved in 50 mL (V3) diluted Fig. 2 shows the mass changes of mortar specimens exposed to
nitric acid (water: nitric acid = 15:85), vibrated for 1–2 min and different corrosive solutions under drying-wetting cycle. It is clear
then stored for 24 h. Afterwards, 10 mL of the supernatant was that the mass changes of mixtures exposed to the N solution are
taken by a pipette and placed in a conical flask, and mixed with increased with the increase of exposure time. For the specimens
10 mL silver nitrate solutions. Finally, the mixtures were titrated exposed to the N + M solutions, the mass changes firstly show an
with potassium thiocyanate. The total chloride content of speci- increase and reach the maximum until the exposure age of 90 days,
men can be calculated by Eq. (3). and then it shows a decrease after 120 days of exposure. It is obvi-
ous that the mass change curves of samples exposed to the N + S
0:3545ðC  V  C 0  V 1 Þ
P¼  100% ð3Þ
G  VV 23

where P denotes the total chloride content (%);C is the concentra-

tion of silver nitrate solution (mol/L); C 0 is the potassium thio-
cyanate solution (mol/L); V and V 1 respectively denotes the
volume of the added silver nitrate solution and the consumed vol-
ume of potassium thiocyanate solution at each test (mL); G is the
sample mass of exposed mortar (g); V 2 and V 3 are the selected
amount for titration and the amount of dilute nitric acid for dis-
solved sample (mL).

2.3.4. Magnesium and sulfate content

The filtrate obtained from the acid dissolved ground mortar
powder was also used for determining the total content of magne-
sium (Mg) and sulfate (S) using a Inductively Coupled Plasma-
Optical Emission Spectr (ICP-OES) Prodigy 7 from Teledyne Leeman
Labs. The applied measuring waves ranged from 165 to 1100 nm,
with the optical resolution of 0.007 nm and analytical error less Fig. 2. Mass change vs. time for mortar specimens exposed to different solution
than 2%. under drying-wetting cycles.
4 S. Cheng et al. / Construction and Building Materials 230 (2020) 116934

and N + S + M solution are similar to the group of N + M. The spec- which also implies that the presence of sulfate ion in chloride solu-
imens exposed to the N solution show the maximum mass incre- tion reduces the damage of mortar under drying-wetting cycles.
ment of 1.10%, while the mass increment of mixes exposed to However, the coexistence of sulfate, magnesium and chloride ion
the combined solution N + M, N + S, and N + M + S reach the maxi- intensifies the degradation of mortar. This indicates that the gener-
mum of 2.74, 2.34% and 2.51%, respectively. The samples exposed ated corrosion products such as ettringite and Friedel’s salt (Fig. 6)
to combined solutions exhibit more obvious mass changes than densify the cement matrix when exposed to sulfate and chloride
those exposed to the N solution under drying-wetting cycles. The attack. The corrosion products also refine the pore structure at
mass of specimens in the N + M solution firstly increases to 2.74% early stage of corrosion, but multiple ions attacks lead to more
until 90 days, and follows by a slight decrease down to 2.68%. obvious corrosion at later stages of exposure.
The mass changes of specimens exposed to the N + S and N + S
+ M solutions show a similar tendency. The mass change of Port-
3.3. Chloride profiles
land cement mortar can be associated with the water imbibition
due to the microstructure change during hydration, which is con-
Fig. 4 shows the total chloride profiles of mortar specimens
sumed the water and is induced in pores an unsaturated state.
exposed to different solutions under drying-wetting cycles. With
The continuous hydration of cement and more formation of C-S-
the increase of mortar depths, the total chloride content decreases,
H gels provide increased surface area to absorb water [19,33].
and the total chloride content exhibits a distinctive increase with
Hence, the mass of mortar specimens is increased until the maxi-
the increase of exposure time. In contrast to the N solution, the
mum saturation.
specimens exposed to the N + M and N + M + S solution show
When exposed to corrosive solutions, the presence of sulfate
higher total chloride contents. However, the total chloride content
and magnesium ion in the chloride environment leads to a higher
for mixes exposed to the N + S solutions is lower than the N solu-
mass increment than the sodium chloride solution. This is caused
tion with the increased exposure time. It can be revealed that the
by the precipitation of brucite and AFt, which could adhere to
presence of sulfate ion in chloride solution inhibits the chloride
the surface of pore walls and inside the pore space in cement
ingress. This phenomenon can be attributed to two aspects: 1)
matrix [34,35]. Moreover, the corrosive solutions with sulfate,
According to the charges-equilibrium and mass-equilibrium prin-
magnesium and chloride ions result in a larger mass change than
ciple, the diffusion rate of sulfate and chloride ion for mortar
these with sulfate and chloride. The results are in accordance with
a previous finding, which revealed that concrete immersed in the
magnesium sulfate and sodium chloride solution showed a larger
mass decrease than the magnesium sulfate solution alone [35].
The magnesium ion does not cause obvious mass loss in the chlo-
ride solution, this is because the magnesium ion only reacts with
hydroxyl ions at the outmost surface and limits the internal ero-
sion, resulting in indistinctive degradation of mortar.

3.2. Relative dynamic modulus of elasticity

The dynamic modulus of elasticity (Erd) of mortar specimens

exposed to different exposure solutions under drying-wetting
cycles are presented in Fig. 3. It can be noticed that the curves of
Erd are increased with the increase of exposure age when exposed
to the N, N + M and N + S solution, respectively. The Erd of mortars
in the N + M + S solution firstly increases until 90 days of exposure
and then slightly decreases until 120 days. The presence of sulfate
and magnesium in chloride environment influences the Erd value,

Fig. 4. Total chloride content mortar specimens after (a) 30 days, (b) 60 days, (c)
Fig. 3. Variation of dynamic modulus of elasticity vs. time for mortar specimens 90 days, and (d) 120 days of exposure under drying-wetting cycles exposed to
exposed to different solution under drying-wetting cycles. different solutions.
 S. Cheng et al. / Construction and Building Materials 230 (2020) 116934 5

exposed to the combined solution is less than their individual dif-

fusion [27,36]; 2) The formation of ettringite refines the pore struc-
ture and leads to a densification of the matrix, which blocks the
channels of chloride ion penetration [22,37,38]. The simultaneous
presences of sulfate and magnesium ion in chloride solution exhi-
bit a higher chloride ingress. This is because the magnesium ion
can react with hydroxyl ions and result in a reduction pH in the
pore solution, promoting the chloride ingress based on the
charges-equilibrium [21,35].

3.4. Magnesium and sulfur profiles

Fig. 5 shows the elemental concentration of magnesium (Mg)

and sulfur (S) determined by ICP-OES for the exposed mortar spec-
imens at different exposure times. Since the NaCl solution does not
contain sulfur and magnesium, the specimens exposed to solutions
of N + M and N + M + S are selected for analysis. It can be observed Fig. 6. XRD patterns for cement paste samples exposed to different solutions after
that the magnesium content in the outermost layer for mortar 120 days drying-wetting cycles.
exposed to the N + M + S solution is increased with the increased
exposure time. The outermost layers (0–5 mm) of the mortar spec-
imens exposed to the N + M + S solution are enriched in magne-
sium. As for the sulfate ion, although the sulfur content shows
variation from 30 to 120 days, the sulfur content at the outmost
layer remain stable for the mixes exposed to the N + M + S solution.
The magnesium profiles in the mortar samples exposed to the N
+ M reveal that there is no clear increase in the magnesium content
at the outermost layer except for 120 days of exposure. For mix-
tures exposed to the N + S, the sulfur content also does not further

Fig. 7. DTG curves of for cement paste samples exposed to different solution after
120 days drying-wetting cycles.

Fig. 8. Portlandite and Friedel’s salt determined by DTG in the cement paste
samples exposed to different solution after 120 days drying-wetting cycles.

increase at the outmost layer. Compared with the chloride ingress

Fig. 5. Total magnesium (Mg) and sulfur (S) content determined by ICP-MS for
shown in Fig. 4, it is clearly evident that the penetration depths of
profile ground mortar samples after 30 days, 60 days, 90 days and 120 days exposed magnesium ion and sulfate ion are much smaller. The precipitation
to different solutions. of brucite occurs at higher pH values (pH > 10) and reduces the
6 S. Cheng et al. / Construction and Building Materials 230 (2020) 116934

Fig. 9. X-CT patterns for mortar samples exposed to (a) N, (b) N + M, (c) N + S, and (d) N + M + S after 120 days drying-wetting cycles.

magnesium concentrations, resulting in limited mobility of magne-
sium in cementitious systems [39]. Indeed, the formation of bru-
cite is observed on the surface of mortar but its content in the
mortar does not increase with the increase of exposure time. Sim-
ilar as the results presented in [23,40,41], the magnesium and sul-
fur was enriched in the outmost section (0–1 mm).
Based on the results shown in Sections 3.1 to 3.4, it can be
noticed that the presence of sulfate and magnesium ion has dis-
parate influence on the degradation of Portland cement mortar.
The sulfate ion in chloride solution delays the degradation of mor-
tar and inhibits the chloride ingress. However, the simultaneous
presence of sulfate and magnesium ion in chloride solution accel-
erates the chloride ingress and aggravates the damage of mortar,
which imply that there are interactions between corrosive ions.
In contrast to mortar exposed to N solution, the amount of Fs
decreases in the one exposed to N + S solution, indicating that
the formation of Fs is limited by sulfate ions. This is attributed to
the fact that C-S-H gels can absorb more sulfates instead of chlo-
rides when exposed to the sulfate + chloride solution. The sulfate
presented in combined solution (NaCl + Na2SO4) reduces the chlo-
ride binding and part of Fs is converted to ettringite [21]. Similarly,
the presence of sulfate and magnesium ion in chloride solution
declines the content of Fs, which suggests that the combined sul-
fate and magnesium ion decreased the chloride uptake in the C-
S-H and induced increased formation of corrosion products [21].
Moreover, it is obvious that the CH content in the N + M + S solu-

3.5. Phase assemblages

The hydration products and corrosion products of cement paste

in different solution after 120 days of exposure are analyzed by
XRD and the results are illustrated in Fig. 6. The patterns show that
the hydration products contain hemicarbonate (Hc) and monocar-
bonate (Mc) before exposure and Friedel’s salt (Fs) is observed in
all specimens after exposed to corrosive solutions. Both Hc and
Mc could directly convert to Fs under the environment of high
chloride concentrations. The TG analysis of exposed hydrated
cement for 120 days of exposure (0–5 mm) shows the phase
assemblages containing C-S-H, ettringite (AFt), portlandite (CH)
as well as calcium carbonate (Cc), as shown in Fig. 7. The peak
ranging from 230 to 410 °C are referred to the decomposition of
Fs [42] and the interval of 410–500 °C is mainly ascribed to the
dehydration of portlandite. The amount of CH and Fs are deter- Fig. 10. Porosity statistics in the X-CT patterns for mortar samples exposed to
mined by DTG and the results are presented in Fig. 8. different solutions.
 S. Cheng et al. / Construction and Building Materials 230 (2020) 116934 7

Table 4
Pore structure information obtained from X-CT and MIP analysis.

Pore X-CT MIP

information
Connected Non-connected Volume <20 nm volume 20–100 nm 100–200 nm >200 nm Porosity
volume (%) volume (%) (%) (%) volume (%) volume (%) volume (%) (%)
N 3.850 0.907 4.757 55.606 22.564 6.519 15.310 12.977
N+M 3.492 1.380 4.872 57.308 25.267 6.368 11.057 9.006
N+S 4.047 0.616 4.663 58.391 22.819 5.965 12.825 13.681
N+M+S 4.332 1.978 6.310 51.797 21.212 5.987 21.004 13.031

Fig. 12. Different pore size distribution percentages for mortar samples exposed to
solutions after 120 days drying-wetting cycles.

magnesium ion in NaCl solution increase the pore structure. This

Fig. 11. Pore size distribution for mortar samples exposed to different solutions
after 120 days drying-wetting cycles.
tion is lower than that exposed to the N solution. The precipitation
of brucite reduced the content of CH. A previous study also indi-
cated that the magnesium and sulfate in chloride solution reduce
the formation of Fs at the outmost layer [38].
result is consistent with the ones from Sections 3.1 to 3.3. In order
to further clarify the pore parameters in the 3D structure, the
detailed compositions of pore structures are illustrated in Fig. 10.
The pore structures of samples can be divided into the connected
pore and the non-connected one. It is obvious that the number of
connected pore in all specimens is less than the non-connected
pore one. The detailed information concerning the pore character-
istics in different solutions is presented in Table 4. The connected
pore volume, non-connected pore volume and total porosity in
the specimens exposed to the N + M + S solution are the largest,
while mixes exposed to the N + S solution shows the lowest con-
nected pore volume. For example, when exposed to the N + M + S
solution, the connected pore volume and porosity are increased
by 118.1% and 32.6% than in the N solution, respectively. Mean-
while, when exposed to the N + S solution, the non-connected pore
volume reduces by 67.9%. Based on the X-ray CT observations, the
results indicate that the simultaneous presence of sulfate and mag-
nesium ion in chloride solution increases the number and volume
of harmful pore, while the presence of sulfate and chloride
decreases the total porosity and harmful pore size.

3.6. Pore structure 3.6.2. MIP analysis

3.6.1. Micro-tomography analysis
The 3D defects analyses of mortars exposed to different solu-
tions after 120 days drying-wetting cycles is shown in Fig. 9. The
color changes from blue to red represent a gradual increase of
the pore volume. It can be seen that after 120 days of exposure,
the pore volume of specimens exposed to the N + M + S solution
increases compared to the N solution, while the specimens in the
N + S solution show a less pore volume. It is evident that the sulfate
ion refines the pore structures, while the presences of sulfate and
The MIP analyses for pore parameters and pore size distribution
of specimens exposed to different solutions are illustrated Fig. 11
and Fig. 12, respectively. In addition, the detailed distribution of
pore structure is summarized in Table 4. It can be observed that
the incremental intrusion follows a similar pattern and there are
two main incremental intrusion distribution peaks of the exposed
mortars. One peak is ranged between 10 and 50 nm, which is usu-
ally considered as less harmful pore (20–50 nm). For the specimens
in N + S solutions, the pore volume (<20 nm) is higher than that
exposed to the N solution, while the volume of pore size between
8 S. Cheng et al. / Construction and Building Materials 230 (2020) 116934
100 and 200 nm is lower. The pore size volume (<20 nm) of mor-
tars exposed to the N + M + S solutions is less than the N solution,
whereas the pore size volume (>200 nm) is higher. The results also
agreed with the X-CT results, which demonstrate that the presence
of sulfate ion in chloride solution decreases the less harmful pore
size, whereas the presence of sulfate and magnesium ion in NaCl
solution increases the harmful pore size. The formation of AFt by
sulfate ion precipitates inside the pores and refines the pores size,
which declines the chloride penetration channels and retards the
chloride ingress [38]. However, the degradations caused by magne-
sium and sulfate result in the increase of porosity and harmful pore
size and thus promote the chloride ingress.
4. Conclusions
This paper addresses the degradation mechanisms of Portland
cement mortar subjected to multiple ion attack and drying-
wetting cycles. Based on the experimental results, the following
conclusions can be drawn:
(1) The multiple corrosive ions have significant effects on the
mass changes and relative dynamic modulus of elasticity
of Portland cement mortar. The presence of sulfate ion in
the chloride solution slows down the degradation of mortar.
However, the simultaneous presence of sulfate and magne-
sium ion in chloride solution aggravates the damages of
mortar.
(2) The presence of sulfate ion in the chloride solution firstly
inhibits and then promotes the chloride ingress, while the
magnesium and sulfate in chloride environment increases
the chloride ingress of mortar. In addition, the ingress of
magnesium ion is limited to the outmost (0–10 mm) at the
surface.
(3) The sulfate ion in chloride environment reduces the content
of Friedel’s salt and CH. In addition, the sulfate and magne-
sium ion in chloride solution can further decline the content
of Fs and CH and cause decalcification of C-S-H.
(4) The synergy effects of sulfate in chloride environment refine
the pore structure and decrease the total porosity. The coex-
istence of magnesium and sulfate ions in chloride solution
increases the porosity and harmful pore volume, thus aggra-
vates the ingress of corrosive ions into the matrix.
Declaration of Competing Interest
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared
to influence the work reported in this paper.
Acknowledgements
This study was financially supported by the National Natural
Science Foundation Project of China (No. 51679179), National
Key Research and Development Program of China (No.
2017YFB0310905), YangFan Innovative & Entrepreneurial Research
Team Project (No. 201312C12) and Open research project of
Advanced Engineering Technology Research Institute of Wuhan
University of technology in Zhongshan city (WUT201902).
References
[1] G.M. Ziad, Deterioration of concrete structures in the Arabian Gulf, Concr. Int.
15 (1993).
[2] M. Maaga, S. Helland, J.E. Caflsen, Exposure to Marine Environment of Chloride
Penetration of High Performance Concrete, Science Press, Beijing, 1998.
[3] M. Santhanam, M. Otieno, 5 – Deterioration of concrete in the marine
environment, in: M.G. Alexander (Ed.) Marine Concrete Structures,
Woodhead Publishing2016, pp. 137–149.
[4] S.-J. Kwon, H.-S. Lee, S. Karthick, V. Saraswathy, H.-M. Yang, Long-term
corrosion performance of blended cement concrete in the marine environment
– a real-time study, Constr. Build. Mater. 154 (2017) 349–360.
[5] J. Hu, P. Deng, X. Li, J. Zhang, G. Wang, The vertical non-uniform corrosion of
reinforced concrete exposed to the marine environments, Constr. Build. Mater.
183 (2018) 180–188.
[6] J. Kim, W.J. McCarter, B. Suryanto, Performance assessment of reinforced
concrete after long-term exposure to a marine environment, Constr. Build.
Mater. 192 (2018) 569–583.
[7] K. Li, D. Zhang, Q. Li, Z. Fan, Durability for concrete structures in marine
environments of HZM project: design, assessment and beyond, Cem. Concr.
Res. 115 (2019) 545–558.
[8] E.E. Hekal, E. Kishar, H. Mostafa, Magnesium sulfate attack on hardened
blended cement pastes under different circumstances, Cem. Concr. Res. 32
(2002) 1421–1427.
[9] A. Benli, M. Karatasß, E. Gurses, Effect of sea water and MgSO4 solution on the
mechanical properties and durability of self-compacting mortars with fly ash/
silica fume, Constr. Build. Mater. 146 (2017) 464–474.
[10] M. Santhanam, M.D. Cohen, J. Olek, Mechanism of sulfate attack: a fresh look
Part 2. Proposed mechanisms, Cem. Concr. Res. 33 (2003) 915–921.
[11] T. Sibbick, D. Fenn, N. Crammond, The occurrence of thaumasite as a product of
seawater attack, Cem. Concr. Compos. 25 (2003) 1059–1066.
[12] A. Skaropoulou, S. Tsivilis, G. Kakali, J.H. Sharp, R.N. Swamy, Long term
behavior of Portland limestone cement mortars exposed to magnesium sulfate
attack, Cem. Concr. Compos. 31 (2009) 628–636.
[13] M.H.F. Medeiros, A. Gobbi, G.C. Réus, P. Helene, Reinforced concrete in marine
environment: Effect of wetting and drying cycles, height and positioning in
relation to the sea shore, Constr. Build. Mater. 44 (2013) 452–457.
[14] T. Simčič, S. Pejovnik, G. De Schutter, V.B. Bosiljkov, Chloride ion penetration
into fly ash modified concrete during wetting–drying cycles, Constr. Build.
Mater. 93 (2015) 1216–1223.
[15] Y. Chen, J. Gao, L. Tang, X. Li, Resistance of concrete against combined attack of
chloride and sulfate under drying–wetting cycles, Constr. Build. Mater. 106
(2016) 650–658.
[16] J. Yuan, Y. Liu, Z. Tan, B. Zhang, Investigating the failure process of concrete
under the coupled actions between sulfate attack and drying–wetting cycles
by using X-ray CT, Constr. Build. Mater. 108 (2016) 129–138.
[17] S.-W. Pack, M.-S. Jung, H.-W. Song, S.-H. Kim, K.Y. Ann, Prediction of time
dependent chloride transport in concrete structures exposed to a marine
environment, Cem. Concr. Res. 40 (2010) 302–312.
[18] M. Santhanam, M. Cohen, J. Olek, Differentiating seawater and groundwater
sulfate attack in Portland cement mortars, Cem. Concr. Res. 36 (2006) 2132–
2137.
[19] K. Wang, D.E. Nelsen, W.A. Nixon, Damaging effects of deicing chemicals on
concrete materials, Cem. Concr. Compos. 28 (2006) 173–188.
[20] F.P. Glasser, J. Marchand, E. Samson, Durability of concrete – degradation
phenomena involving detrimental chemical reactions, Cem. Concr. Res. 38
(2008) 226–246.
[21] K. De Weerdt, D. Orsáková, M.R. Geiker, The impact of sulphate and
magnesium on chloride binding in Portland cement paste, Cem. Concr. Res.
65 (2014) 30–40.
[22] M. Maes, N. De Belie, Resistance of concrete and mortar against combined
attack of chloride and sodium sulphate, Cem. Concr. Compos. 53 (2014) 59–72.
[23] K. De Weerdt, A. Colombo, L. Coppola, H. Justnes, M.R. Geiker, Impact of the
associated cation on chloride binding of Portland cement paste, Cem. Concr.
Res. 68 (2015) 196–202.
[24] M.F. Najjar, M.L. Nehdi, A.M. Soliman, T.M. Azabi, Damage mechanisms of two-
stage concrete exposed to chemical and physical sulfate attack, Constr. Build.
Mater. 137 (2017) 141–152.
[25] O. Helson, J. Eslami, A.-L. Beaucour, A. Noumowe, P. Gotteland, Durability of
soil mix material subjected to wetting/drying cycles and external sulfate
attacks, Constr. Build. Mater. 192 (2018) 416–428.
[26] P.W. Brown, S. Badger, The distributions of bound sulfates and chlorides in
concrete subjected to mixed NaCl, MgSO4, Na2SO4 attack, Cem. Concr. Res. 30
(2000) 1535–1542.
[27] M. Frias, S. Goñi, R. García, R. Vigil de La Villa, Seawater effect on durability of
ternary cements. Synergy of chloride and sulphate ions, Compos. Part B: Eng.
46 (2013) 173–178.
[28] K. De Weerdt, H. Justnes, M.R. Geiker, Changes in the phase assemblage of
concrete exposed to sea water, Cem. Concr. Compos. 47 (2014) 53–63.
[29] J. Geng, D. Easterbrook, L.-Y. Li, L.-W. Mo, The stability of bound chlorides in
cement paste with sulfate attack, Cem. Concr. Res. 68 (2015) 211–222.
[30] European standard EN 197-1 Cement. Composition, Specifications and
conformity criteria for common cements, 2011.
[31] Test methods of physical properties for dry-mix mortar, Beijing, 2013.
[32] Standard for test methods of long-term performance and durability of
ordinary concrete, Beijing, 2009.
[33] K. Sotiriadis, E. Nikolopoulou, S. Tsivilis, Sulfate resistance of limestone cement
concrete exposed to combined chloride and sulfate environment at low
temperature, Cem. Concr. Compos. 34 (2012) 903–910.
[34] T. Chiker, S. Aggoun, H. Houari, R. Siddique, Sodium sulfate and alternative
combined sulfate/chloride action on ordinary and self-consolidating PLC-
based concretes, Constr. Build. Mater. 106 (2016) 342–348.
 S. Cheng et al. / Construction and Building Materials 230 (2020) 116934
[35] M. Maes, N. De Belie, Influence of chlorides on magnesium sulphate attack for
mortars with Portland cement and slag based binders, Constr. Build. Mater.
155 (2017) 630–642.
[36] S. Goñi, M.P. Lorenzo, J.L. Sagrera, Durability of hydrated portland cement with
copper slag addition in NaCl + Na2SO4 medium, Cem. Concr. Res. 24 (1994)
1403–1412.
[37] Y. Cao, L. Guo, B. Chen, Influence of sulfate on the chloride diffusion
mechanism in mortar, Constr. Build. Mater. 197 (2019) 398–405.
[38] S. Cheng, Z. Shui, T. Sun, X. Gao, C. Guo, Effects of sulfate and magnesium ion
on the chloride transportation behavior and binding capacity of Portland
cement mortar, Constr. Build. Mater. 204 (2019) 265–275.
9
[39] N.R. Buenfeld, J.B. Newman, The development and stability of surface layers on
concrete exposed to sea-water, Cem. Concr. Res. 16 (1986) 721–732.
[40] U.H. Jakobsen, K. De Weerdt, M.R. Geiker, Elemental zonation in marine
concrete, Cem. Concr. Res. 85 (2016) 12–27.
[41] K. De Weerdt, B. Lothenbach, M.R. Geiker, Comparing chloride ingress from
seawater and NaCl solution in Portland cement mortar, Cem. Concr. Res. 115
(2019) 80–89.
[42] Z. Shi, M.R. Geiker, B. Lothenbach, K. De Weerdt, S.F. Garzón, K. Enemark-
Rasmussen, J. Skibsted, Friedel’s salt profiles from thermogravimetric analysis
and thermodynamic modelling of Portland cement-based mortars exposed to
sodium chloride solution, Cem. Concr. Compos. 78 (2017) 73–83.