Powder Technology 338 (2018) 725–733

Contents lists available at ScienceDirect

Powder Technology

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

Influences of limestone powder on the resistance of concretes to the

chloride ion penetration and sulfate attack
Jianwei Sun ⁎, Zhonghui Chen
School of Mechanics and Civil Engineering, China University of Mining and Technology, Beijing 100083, China

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

Article history: Influences of limestone powder on the resistance of concretes to the chloride ion penetration and sulfate attack
Received 22 May 2018 with a constant water/binder ratio and a constant 28-day compressive strength were studied. The sensitivity of
Received in revised form 5 July 2018 the properties of concrete to the initial moist curing time was also explored. The results indicate that, under a
Accepted 13 July 2018
constant water/binder ratio condition, the resistance to sulfate attack of concrete deteriorates with the increasing
Available online xxxx
of limestone powder content, and the resistance to chloride ion penetration decreases when the replacement
Keywords:
ratio of limestone powder is 24%. As the initial moist curing time declines, the reducing magnitude of the prop-
Limestone powder erties of concrete containing limestone powder is larger than that of plain cement concrete. Nevertheless, lower-
Initial moist curing time ing the water/binder ratio of concrete containing limestone powder can significantly reduce the sensitivity of the
Compressive strength properties of the concrete to the initial moist curing time. No matter how long the initial moist curing time is,
Chloride ion penetration replacing cement with 8% limestone powder can improve the resistance of concrete to sulfate attack. Concrete
Sulfate attack with up to 24% limestone powder addition can still obtain a resistance to chloride ion penetration and sulfate
attack similar to plain cement concrete on the premise of a constant 28-day compressive strength. Moreover, it
was found that both the crystallization of sodium sulfate and the formation of ettringite collectively result in
the deterioration of concrete subjected to the sulfate solution.
© 2018 Elsevier B.V. All rights reserved.

1. Introduction later-age strength is decreased due to the dilution effect of limestone

Limestone powder, produced by grinding limestone into powder, is
a by-product of the limestone quarry industry. Limestone powder has
been recognized as a kind of mineral admixture in cement and concrete
for many years due to its technical and economic advantages [1–3]. The
main component of limestone powder is calcium carbonate (CaCO3),
and the content of CaCO3 is typically N95% [4]. SiO2, Fe3O4, MgO and
Al2O3 are other constituents of limestone powder [5–7]. Limestone
powder is used primarily as a filler material because of its small particle
size in comparison to the particle size of Portland cement. Numerous
studies have proved that the effect of limestone powder is neither poz-
zolanic nor inert since calcite has a low activity [5, 8–11]. Early in 1977,
Soroka et al. [12] determined that limestone powder could act as nuclei
for the formation of C-S-H gel. When limestone powder exists, a series
of reactions occurs between calcium carbonate and all kinds of calcium
aluminate hydrates, finally forming a stable AFm phase, i.e., calcium
monocarboaluminate [9, 13, 14]. The results lead to an increase in the
total solid volume and a decrease in the overall porosity [15]. Therefore,
the early-age mechanical properties of concrete can be improved by the
promoting effect of limestone powder on cement hydration, while the
⁎ Corresponding author.
E-mail address: tbp1600602045@student.cumtb.edu.cn (J. Sun).
powder, resulting from the increase of the effective water to cement
ratio [16–18].
Ingress of chloride into concrete is considered one of the most dan-
gerous mechanisms influencing the durability of the structure since
chloride plays the main role in accelerating the initiation of corrosion
in the reinforcing steel [19]. Good chloride resistance derives from the
low ion diffusivity of concrete, resulting from hindering ion penetration.
The presence of mineral admixtures is acknowledged to affect the pore
structures of concrete because the mineral admixtures change the
microstructures of the pastes and the interfacial transition zone (ITZ)
between pastes and aggregates [20]. According to previous studies,
limestone powder has significant effects on pore structure, thus affect-
ing the durability of concrete at a later stage. Elgalhud et al. [21] found
that the pore structures of cementitious mixtures remained unimpaired
when the replacement ratio of limestone powder is b25%. According to
Hornain et al. [22], the addition of limestone powder slightly reduces
the chloride ion diffusion coefficient, which is attributed to the effect
of the limestone powder filler particles on the tortuosity of the matrix.
Nevertheless, Wang et al. [23] found that the long-term (up to 5
years) chloride penetration of the concrete containing limestone pow-
der is higher than that of plain cement concrete because of the coarser
pore structure and less C-S-H production, leading to the reduction of
chloride binding capacity.

https://doi.org/10.1016/j.powtec.2018.07.041
0032-5910/© 2018 Elsevier B.V. All rights reserved.
726 J. Sun, Z. Chen / Powder Technology 338 (2018) 725–733

Concrete structures may suffer from sulfate attack, which causes

serious deterioration of the concrete and affects the service life of the
structures. Sulfate ions penetrate the pores and thus form an aggressive
solution, which react with certain components of hydrated cement
(calcium hydroxide, hydrated calcium aluminate and AFm phases) to
form ettringite (AFt) and gypsum that induce expansion [16]. As a
result, the concrete begins to crack, soften and lose strength. In addition,
thaumasite, another product of sulfate attack, is formed from the pore
solution of the pastes in the presence of calcium silicate ions and
carbonate ions (CO2− 3 ) in damp atmosphere [24–28]. Hence, thaumasite
sulfate attack (TSA) may occur in concrete containing limestone powder
since it contains plenty of CO2–3 [29]. This deterioration will be acceler-
ated in cold environments (below 15 °C) [16, 30]. Proper content of
limestone powder can improve the pore structure and decrease the
porosity of the hardened paste because of its filling effect, thus improv-
ing the performance of concrete [31]. When limestone powder is pres-
ent in cementitious materials, both the physical dilution effect that
decreases the C3A content and the chemical reaction between calcite
and C3A that forms some different AFm phases will affect the alumina-
hydrated compounds. Therefore, when limestone powder replacement
is b10%, the expansion is reduced and there is no impairment to the
sulfate attack [32, 33]. Liu et al. [34] reported that the addition of
limestone powder could improve the magnesium sulfate resistance of
mortars. Torres et al. [35] found that mortars containing 5% limestone
powder tended to be more tolerant to sulfate attack when compared
to ordinary Portland cement mortar due to the formation of thaumasite.
Nevertheless, Lee et al. [36] found that higher replacement levels of
limestone powder led to more serious sulfate attack, which was also
strongly associated with the formation of thaumasite.
The studies discussed above were carried out mostly under the
condition of an equal w/b ratio and standard curing for 28 days. How-
ever, the 28-day compressive strength of concrete is often defined as
the design strength in many engineering applications. Therefore, it is
necessary to adjust the w/b ratio to obtain the design strength when
limestone powder is used as a partial substitute for cement. Further-
more, the initial moist curing time affects the microstructures and
thus affects the durability of concrete. A large number of concrete
structures cannot be cured sufficiently at early ages, especially are
built in the salty soil region that lack water and have poor construction
conditions. Hence, it is necessary to investigate that the sulfate attack
and chloride ion penetration of the concrete with high content of
mineral admixture in the case of insufficient moist curing.
Therefore, influences of limestone powder matching with initial
moisture curing time (3 days, 7 days and 28 days) on chloride ion
penetration and sulfate attack of concretes were studied under different
conditions including an equal w/b ratio and an equal 28-day compres-
sive strength.
2. Experimental
2.1. Raw materials
P·I 42.5 Portland cement conforming to Chinese National Standard
GB 175 and limestone powder obtained by grinding limestone into
powder were used as cementitious materials. The chemical composi-
tions determined by X-Ray Fluorescence (XRF), and particle size
distributions of cement and limestone powder are presented in Table
1 and Fig. 1, respectively. Natural river sand with the maximum size of
5 mm was used as fine aggregate, while crushed limestone with a size
Fig. 1. Particle size distributions of cement and limestone powder.
between 5 and 25 mm was used as coarse aggregate. The
polycarboxylate-based superplasticizer was used to adjust the
flowability of fresh concrete.
2.2. Mix proportions
The mix proportions of the concretes were provided in Table 2. The
concretes were divided into two groups. A plain cement concrete
sample with a w/b ratio of 0.45 (Sample C-0.45) was used as the
reference sample. Three replacement ratios (8, 16, and 24% by mass)
of limestone powder in each group were chosen. Compared to sample
C-0.45, concretes in Group I have the same w/b ratio whereas concretes
in Group II have the similar 28-day compressive strength under
standard curing condition (20 ± 1 °C, 95% RH). In Group II, the w/b
ratio decreases with the increasing replacement ratio of limestone
powder to obtain similar compressive strength at the age of 28 days,
which is about 53 MPa.
2.3. Curing and testing
Concrete samples with the size of 100 mm × 100 mm × 100 mm
were prepared. During the first 28 days, three kinds of early curing
conditions were adopted and provided in Table 3.
After curing, the 28-day compressive strengths of the concretes
were determined according to the Chinese National Standard GB/T
50081–2002. The penetration of the concrete to chloride ion was tested
at the age of 28 and 90 days according to ASTM C1202 “Standard Test
Method for Electrical Indication of Concretes Ability to Resist Chloride
Ion Penetration”.
The resistance to sulfate attack of concrete was tested using the
drying-wetting cycles method. According to Thaulow et al. [37], sodium
sulfate solution (5% Na2SO4) was selected as the corrosive solution since
Na2SO4 is the most commonly found salt on concrete surfaces exposed
to sulfate attack environments. Each cycle lasted 24 h: At the age of
28 days, the concrete was immersed in Na2SO4 solution for 15 h, dried
Table 2
Mix proportions of the concretes.
Samples Mix proportions (kg/m3)
Cement Limestone Fine Coarse Water
powder aggregates aggregates

Reference C-0.45 400 0 788 1044 180

Table 1 Group I L1–0.45 368 32 788 1044 180
Chemical compositions of raw materials (%). L2–0.45 336 64 788 1044 180
L3–0.45 304 96 788 1044 180
SiO2 Al2O3 Fe2O3 CaO MgO SO3 CaCO3
Group II L1–0.43 368 32 792 1051 172
Cement 21.86 4.25 2.66 63.59 2.19 2.42 – L2–0.39 336 64 799 1060 156
Limestone powder 7.25 1.88 1.18 – 2.96 0.29 84.28 L3–0.35 304 96 806 1069 140
 J. Sun, Z. Chen / Powder Technology 338 (2018) 725–733
Table 3
Curing regimes within the initial 28 days.
Curing period Condition 1 Condition 2 Condition 3
0–3 days 20 ± 1 °C, 95% RH 20 ± 1 °C, 95% RH 20 ± 1 °C, 95% RH
4–7 days 20 ± 1 °C, 95% RH 20 ± 1 °C, 95% RH 20 °C, 40–50% RH
8–28 days 20 ± 1 °C, 95% RH 20 °C, 40–50% RH 20 °C, 40–50% RH
in an electric vacuum drying oven at (80 ± 5) °C for 6 h, and then cooled
for 3 h. The compressive strength of the concrete was tested after 120
and 150 cycles. Meanwhile, the reference specimens that were further
water-cured for 120 and 150 days were tested. The sulfate attack resis-
tance of concrete was evaluated by the compressive strength loss ratio
which was defined as:

5.1%, 11.3% and 18.6%, respectively. For 7-day initial moist curing, the
F ¼ C Ref −C  C Ref  100%
where F is the compressive strength loss ratio, CRef is the compressive
strength of the reference sample (the concrete further cured in water)
and C is the compressive strength of the concrete after sulfate attack.
The pore structure of paste was measured by mercury injection
porosimetry (MIP). The morphology and microstructure of the concrete
containing limestone powder which had been exposed to the sulfate
salt solutions under drying-wetting cycles for 150 days were investi-
gated by scanning electron microscope (SEM) and energy dispersive
X-ray spectrums (EDS).
3. Results and discussion
3.1. Pore structure
Fig. 2 shows the pore structures of the pastes cured at the standard
condition for 3 and 7 days. The cumulative pore volume of Sample
L3–0.35 is obviously smaller than that of Sample C-0.45 at all ages,
while the cumulative pore volume of Sample L2–0.39 is similar to that
of Sample C-0.45 at 7 days. It is obvious that reducing the w/b ratio
and replacing cement with 24% limestone powder can make the micro-
structure of the pastes more compact, less porous and denser. Although
the total porosity of Sample L1–0.43 is almost identical to that of Sample
C-0.45 at 3 days, the cumulative pore volume with diameters larger
than 50 nm of Sample C-0.45 is obviously smaller than that of Sample
L1–0.43 at 7 days. It is normally accepted the addition of limestone
powder would increase the early porosity at the same w/b ratio.
However, lowering the w/b ratio can significantly decrease the early
porosity. Thus the porosity of the hardened paste containing 24% lime-
stone powder with lower w/b ratio is smaller than reference sample.
Fig. 2. Pore structures of hardened pastes. (a) at the age of 3 days; (b) at the age of 7 days.
727
Therefore, lowering the w/b ratio in case of largely mixing limestone
powder can greatly enhance and refine the microstructure of the hard-
ened paste.
3.2. Compressive strength
3.2.1. Constant w/b ratio
Fig. 3 shows the 28-day compressive strength of all the concrete
with a constant w/b ratio under different curing conditions. Overall,
the compressive strength tends to decrease as the initial moist curing
time declines and also decreases with the increase of limestone powder
content. Compared to the compressive strength of plain cement
concrete for 28-day initial moist curing, the compressive strengths of
concretes containing 8%, 16% and 24% limestone powder decrease
ð1Þ compressive strengths of concretes decrease 1.4%, 11.6% and 21.5%,
respectively. For 3-day initial moist curing, the compressive strengths
decrease 4.2%, 12.9% and 24.7%, respectively. On the whole, compressive
strength losses are significantly lower than replacement ratios. CaCO3 in
limestone powder can react with C3A and C4AF to form an unstable
hemicarboaluminate phase at early ages. Along with the hydration,
hemicarboaluminate converts into monocarboaluminate, which
belongs to a stable AFm phase [14, 38, 39]. However, these reactions
occupy a long period and the degree of these reactions is very low due
to kinetic restrictions [10, 15]. Consequently, the effect of the reactions
of calcite on the compressive strength at the age of 28 days is negligible.
Nevertheless, replacing part of the cement with limestone powder
increases the effective water to cement ratio and provides the growing
space for hydration products, thus changing the hydrating environment
and increasing the hydration degree of the binders. In addition, the
filling effect and heterogeneous nucleation of limestone powder also
contribute to the compressive strength [40].
Compared with 28-day initial moist curing, the compressive
strengths of plain cement concretes for 7-day and 3-day initial moist
curing decrease by 2.8% and 5.5%, respectively. In terms of containing
8% limestone powder, the compressive strengths of concretes for
7-day and 3-day initial moist curing decrease by −1% and 4.7%, respec-
tively. The compressive strengths of concretes containing 16% limestone
powder decrease by 3.2% and 7.2%, respectively. However, the compres-
sive strengths decrease by 6.3% and 12.6% when the replacement ratio
of limestone powder is 24%. It indicates that the sensitivity of 28-day
compressive strength of concretes containing 8% and 16% limestone
powder to the initial moist curing time is close to that of plain cement
concrete. However, as the replacement ratio increased to 24%, the
28-day strength of concrete is obviously more sensitive to the initial
moist curing time.
728 J. Sun, Z. Chen / Powder Technology 338 (2018) 725–733
Fig. 3. The compressive strength of Group I with different initial curing conditions at the
age of 28 days.
3.2.2. Constant 28-day compressive strength
Fig. 4 shows the 28-day compressive strength of Group II under
different curing conditions. Compared to the 28-day compressive
strength of plain cement concrete, the 28-day compressive strengths
of concretes containing different limestone powder content varied
from −1.7% to 2.3% under standard curing condition. When the initial
moist curing time is reduced to 7 days and 3 days, the compressive
strength variation ranges of concretes containing limestone powder
compared to the reference concrete are 1.9%–4.1% and 0.4%–2.2%,
respectively. It can be seen that the 28-day compressive strengths of
these four concretes present similar sensitivities to the initial moist
curing time. This result is different from the circumstances where
concrete samples have the same w/b ratio since lowering the w/b
ratio has a significant effect on improving the early-age density of the
structures. Even with up to 24% limestone powder, decreasing the w/b
ratio from 0.45 to 0.35 can attain the 28-day compressive strength
equivalent to plain cement concrete regardless of the initial curing
conditions. Moreover, compressive strengths have a small difference
at 3-day initial moist curing, but compressive strength containing 24%
limestone powder is the highest at 7-day initial moist curing. The results
of compressive strength under curing condition 2 and 3 correspond to
the results of pore structure analysis.
3.3. Chloride ion penetration
3.3.1. Constant w/b ratio
Fig. 5 shows the chloride ion penetrability grades of the four
concretes with a constant w/b ratio. Fig. 5a shows that the penetrability
grade of Samples L1–0.45 and L2–0.45 falls in the “Moderate” level at
Fig. 4. The compressive strength of Group II with different initial curing conditions at the
age of 28 days.
the age of 28 days under standard curing condition, which is the same
as Sample C-0.45. Chloride ion penetration is related to connected
porosity, and the compressive strength of concrete is closely associated
with pore structure [41]. Though replacing cement with limestone
powder decreases the compressive strength of concrete, the filling effect
of limestone powder can block the connectivity of some pores [22, 42].
However, the addition of 24% limestone powder (Sample L3–0.45)
results in a significant decrease of hydration products, leading to a
great increase in total porosity and connected porosity. Consequently,
the chloride ion penetration grade of Sample L3–0.45 falls in the
“High” level, and it is higher than that of Sample C-0.45. In addition,
the chloride ion penetration grade of Samples C-0.45, L1–0.45 and
L2–0.45 has not changed as the initial moist curing time declines to 7
days, indicating that decreasing initial moist curing time from 28 days
to 7 days does little harm to the chloride ion penetration at the age of
28 days. But for 3 days of initial moist curing, chloride ion penetration
grade of Sample L2–0.45 changes from “moderate” to “high” level.
Therefore, when the replacement ratio of limestone powder is increased
to 16%, shortening the initial moist curing time has an obviously detri-
mental effect on resistance to chloride ion penetration.
It can be referred from Fig. 5b that the chloride ion penetration grade
of only Sample L3–0.45 under standard moist curing condition changes
from “High” to “Moderate” level, indicating that the pore structures of
concrete with 24% limestone powder replacement is greatly improved
when the curing age is prolonged from 28 days to 90 days. This may
be explained by two reasons. On one hand, the addition of limestone
powder results in an increase of effective water to cement ratio and
provides a larger space for the formation of hydration products. On
the other hand, the reaction degree of limestone powder increases at
later ages, and the monocarboaluminate phase continues to form in
the pores [39], which also contributes to optimization of the chloride
ion penetration. However, the chloride ion penetration grade of
Samples C-0.45 and L1–0.45 does not change, indicating that shortening
of the initial moist curing time has an inconspicuous influence on the
90-day chloride ion penetration of plain cement concrete and concrete
containing a small amount of limestone powder.
Although replacing cement by limestone powder will reduce the
active components of the binders, the addition of an appropriate
amount of limestone powder can promote the early-age hydration of
cement [43]. Therefore, the activity of binders with 8% limestone pow-
der replacement is close to the activity of plain cement. However,
when a large amount of limestone powder is used to replace cement,
the early-age activity of binders decreases and it is necessary to keep
28-initial moist curing time. Thus, the chloride ion penetration of
concrete with 24% limestone powder is sensitive to initial moist curing
time.
3.3.2. Constant 28-day compressive strength
The chloride ion penetration of concretes with an equal 28-day com-
pressive strength is presented in Fig. 6. It can be seen in Fig. 6a that the
chloride ion penetration grades of all concretes fall in the “Moderate”
level at the age of 28 days. The reduction of initial moist curing time
has no influence on chloride ion penetration grades, indicating that re-
ducing the w/b ratio can reduce the sensitivity of 28-day chloride ion
penetration of concrete containing limestone powder compared to
Group I. A lower w/b ratio makes denser early-age microstructures
that can hinder the loss of inner moisture to guarantee later-age hydra-
tion, more than the reduced activity of the binders due to the increment
of limestone powder.
As can be seen from Fig. 6b, the chloride ion penetration grades of all
concrete samples containing limestone powder at the age of 90 days are
lower than that of plain cement concrete by one level only under stan-
dard curing condition, indicating that the improving magnitude of the
later-age microstructures of concrete with limestone powder replace-
ment is greater than that of plain cement concrete under constant 28-
day compressive strength condition.
 J. Sun, Z. Chen / Powder Technology 338 (2018) 725–733
Fig. 5. Chloride ion penetration of Group I with different initial curing conditions (a) at the age of 28 days; (b) at the age of 90 days.
It can be obtained from Fig. 6 that, reducing the w/b ratio can reduce
the sensitivity of chloride ion penetration of concrete containing
limestone powder to the initial moist curing time. When the concretes
containing limestone powder gain the same 28-day compressive
strength as plain cement concrete under standard curing condition,
the chloride ion penetration of concrete containing limestone powder
and plain cement concrete can reach the same level even if initial
moist curing time is insufficient (even 3 days).
3.4. Sulfate attack resistance
3.4.1. Constant w/b ratio
Fig. 7 shows the compressive strength of Group I with different ini-
tial curing conditions at the age of 120 and 150 days. Fig. 8 shows the
loss of compressive strength. Strength losses of all concretes increase
with the increase of the times of the drying-wetting cycles and the
shortening of the initial moist curing time, which reflects the decrease
of the concrete in the capability to resist sulfate attack. In general,
when external sulfate ion enters the pore solution of the concrete,
salty crystallization of sulfate and a series of complex chemical reactions
between sulfate ions and some of the hydrates will occur inside the
pores, both of which cause expansion of the solid volume and then
cracking and strength loss will occur. Therefore, the degree of difficulty
of the sulfate ion entering interior concrete through connected pores is a
key factor of resistance to sulfate attack [44]. It can be deduced from the
effect of the initial moist curing time on compressive strength and chlo-
ride ion penetration of concrete that the porosity of concrete and
connectivity of pores are increased with the decrease in the initial
moist curing time.
Furthermore, it can be concluded that the compressive strength loss
of concrete shows an increasing trend as the replacement ratio of lime-
stone powder increases regardless of the initial curing conditions. Thus,
Fig. 6. Chloride ion penetration of Group II with different initial curing conditions (a) at the age of 28 days; (b) at the age of 90 days.
729
it is suggested that the addition of limestone powder under the condi-
tion of constant w/b ratio causes deterioration of capability to resist
sulfate attack. On one hand, the addition of limestone powder makes
the pores of matrix coarser. On the other hand, the addition of limestone
powder provides a carbonate source for the formation of TSA. TSA
occurs more frequently at low temperatures (b15 °C) and the formation
of is so slow at ambient temperature that TSA rarely occurs [16, 30].
However, the external source of CO2– 3 will promote the reaction kinetics
of TSA and more carbonate source results in a greater risk of thaumasite
formation [45–47]. Thus it makes the corrosion more severe.
3.4.2. Constant 28-day compressive strength
Fig. 9 shows the compressive strength of Group II with different ini-
tial curing conditions at the age of 120 and 150 days. Fig. 10 shows the
compressive strength loss. Similar to the trend under constant w/b
ratio condition, the compressive strength losses of all concretes increase
with the increase of the times of the drying-wetting cycles and the
shortening of the initial moist curing time. The compressive strength
loss of concrete with 8% limestone powder addition is lower than that
of plain cement concrete, whereas the compressive strength losses of
concretes containing 16% and 24% limestone powder are close to the
strength loss of plain cement concrete. It indicates that the resistance
to sulfate attack of concrete containing 8% limestone powder is still
superior to that of plain cement concrete, and concretes containing
16% and 24% limestone powder keep similar sulfate attack resistance
to plain cement concrete.
Hence, lowering the w/b ratio in the case of adding limestone
powder, similar or even better resistance to sulfate attack of concrete
containing limestone powder compared to that of plain cement
concrete can be achieved. The reason is that lowering the w/b ratio
makes the hydration products better fill in and even block the
connected pores in spite of the low activity of limestone powder. Thus
730 J. Sun, Z. Chen / Powder Technology 338 (2018) 725–733
Fig. 7. The compressive strength of Group I with different initial curing conditions. (a) at the age of 120 days; (b) at the age of 150 days.
a denser initial microstructure can be formed to hinder the invasion of
aggressive ions and enhance the capability to resist sulfate attack.
Therefore, reducing the w/b ratio of concrete when adding limestone
powder can significantly lower the sensitivity of the sulfate attack resis-
tance to initial moist curing time. In this way, similar and even better
capability to resist sulfate attack of concrete containing limestone pow-
der can be achieved under insufficient moist curing conditions.
3.5. Mechanism analysis
Sulfate attack of concrete is a complex physical and chemical
process. Concerning the deterioration mechanism, researchers attrib-
uted the failure of concrete to physical attack and chemical attack [48,
49]. The physical attack generally means that salt crystallization inside
the porous material leads to the collapse of concrete. For example,
thenardite (Na2SO4), mirabilite (Na2SO4·10H2O) or epsomite
(MgSO4·7H2O) can be identified in pores [50–53]. Thaulow et al. [54]
have classified the damage mechanism caused by the salt crystallization
into three groups: the solid volume change hypothesis; the salt hydra-
tion pressure hypothesis; the salt crystallization pressure hypothesis.
The sulfate attack of concrete is commonly combined with other
damage mechanisms, especially the cyclic environment conditions,
e.g., freezing-thawing, drying-wetting and heating-cooling cycles
[55, 56]. The coupled effect of sulfate attack and cyclic environment con-
ditions tends to aggravate the damage of concrete due to the accelerated
sulfate salt crystallization.
The chemical reactions in the pore solution can occur at any sulfate
concentration. The chemical attack means that the complex
Fig. 8. The compressive strength loss of Group I.
physicochemical process of the chemical reactions leads to the
collapse of concrete, forming the harmful minerals and causing vol-
ume expansion [57]. When cement paste is in contact with sulfate
ions, the harmful mineral phases refer to the gypsum from the reac-
tion of Ca(OH)2 and sulfate ions as well as AFt from the reaction of
sulfate ions and aluminate. Meanwhile, the pH value of the pore solu-
tion is influenced by the Ca(OH)2 concentration and controls the sta-
bility of AFt and C-S-H gel [58]. Gypsum and AFt, the chemical
reaction products, could be clearly identified in previous papers.
When concrete is exposed to a sulfate environment, the salt crystalli-
zation, the formation and the growth of the harmful mineral phases
inside the pores provide the internal stress and aggravate to destroy
matrix. Therefore, it is often taken for granted that physical attack
and chemical attack collectively results in the degradation of the in-
herent mechanical properties.
Fig. 11 shows the morphology of the concrete after sulfate attack for
150 days observed by SEM. The EDS results of different mineral phases
in Fig. 11 are shown in Fig. 12. These figures clearly indicate the mech-
anisms of the sulfate attack of concrete with the accelerating test in
this study. As is shown in Fig. 11a and b, a large amount of sodium
sulfate and a very small amount of AFt crystals are co-existed inside
the matrix of the concrete. Only the rod-like AFt is found on the fracture
surfaces in this research. The phenomenon illustrates that the expansive
force inside the concrete is mainly caused by crystallization of sodium
sulfate and the delayed AFt formation.
As can be seen from Fig. 11c, the cracks inside the concrete matrix
are filled up by massive sodium sulfate crystals. According to the
research of Aye and Oguchi [59], the crystallization of sodium sulfate
occurs in cycling environment conditions. Flatt have found the same
result: Na2SO4 crystallized and accumulated in the porous materials
during the drying process and Na2SO4·10H2O precipitation occurred
while the samples was exposed to the sulfate solutions again accompa-
nied with high crystallization pressure [60].
The migration of the sulfate ion is a very important process during
the sulfate attack of concrete. The sulfate attack process may be divided
into seven main classes: (1) sulfate ions penetrate into the concrete, (2)
sulfate ions react with the aluminate to produce the AFt, (3) the growth
of the AFt crystals provides the expansive force, (4) cracks appear due to
expansion at the early stages, (5) in turn cracks increase the permeating
rate of the sulfate ion, (6) the AFt is less produced due to the rapid
consumption of aluminate at the late stages, (7) the further crystalliza-
tion due to sodium sulfate continuously provides the expansive force
under drying-wetting cycles. As is discussed above, both the crystalliza-
tion of sodium sulfate and the formation of AFt inside the pores collec-
tively result in the deterioration of the concrete subjected to the
sulfate solution in this research.
 J. Sun, Z. Chen / Powder Technology 338 (2018) 725–733 731

Fig. 9. The compressive strength of Group II with different initial curing conditions. (a) at the age of 120 days; (b) at the age of 150 days.

4. Conclusions

(1) Under the condition of a constant w/b ratio, compressive

strength and resistance to sulfate attack of concrete decline
with the increase of limestone powder content. When the re-
placement ratio of limestone powder is relatively high, resis-
tance to the penetration of chloride ions decreases. In
addition, with the increment of limestone powder, properties
of concrete become more sensitive to the initial moist curing
time: shortening the initial moist curing time will cause sig-
nificant deterioration in the resistance to chloride ion pene-
tration and sulfate attack.
(2) By lowering the w/b ratio, the properties of concrete containing
limestone powder can be made less sensitive to initial moist cur-
ing time. With a constant 28-day compressive strength under
standard curing condition, the resistance to sulfate attack of con-
Fig. 10. The compressive strength loss of Group II.

Fig. 11. SEM figures of the concrete containing limestone under drying-wetting cycles for 150 days.
732 J. Sun, Z. Chen / Powder Technology 338 (2018) 725–733
Fig. 12. EDS spectrums of different mineral phases in Fig. 11. a-e: scanning positions 1 to 5.
crete containing 8% limestone powder is superior to that of plain
cement concrete, and concrete with up to 24% limestone powder
replacement can still achieve the chloride ion penetration and re-
sistance to sulfate attack similar to plain cement concrete under
each moist curing condition.
(3) Both the crystallization of sodium sulfate and the formation of
AFt inside the pores collectively provide the inner expansive
force, resulting in the deterioration of the concrete subjected to
the sulfate solution.
Acknowledgment
Authors acknowledge the support from the National Key Research
and Development Program of China (2017YFC1503100).
References
[1] Gritsada Sua-Iam, N. Makul, Utilization of limestone powder to improve the proper-
ties of self-compacting concrete incorporating high volumes of untreated rice husk
ash as fine aggregate, Constr. Build. Mater. 156 (2013) 455–464.
[2] Gritsada Sua-Iam, N. Makul, Use of increasing amounts of bagasse ash waste to pro-
duce self-compacting concrete by adding limestone powder waste, J. Clean. Prod. 20
(2013) 308–319.
[3] V. Bonavetti, H. Donza, V. Rahhal, et al., Influence of initial curing on the properties
of concrete containing limestone blended cement, Cem. Concr. Res. (5) (2000)
703–708.
[4] A. Skaropoulou, S. Tsivilis, G. Kakali, et al., Long term behavior of Portland limestone
cement mortars exposed to magnesium sulfate attack, Cem. Concr. Compos. 31
(2009) 628–636.
[5] S. Tsivilis, J. Tsantilas, G. Kakali, et al., The permeability of Portland limestone cement
concrete, Cem. Concr. Res. 33 (2003) 1465–1471.
[6] H. Siad, A. Alyousif, O.K. Keskin, et al., Influence of limestone powder on mechanical,
physical and self-healing behavior of engineered cementitious composites, Constr.
Build. Mater. 99 (2015) 1–10.
[7] K.De. Weerdt, M.B. Haha, G.Le. Saout, et al., Hydration mechanisms of ternary
Portland cements containing limestone powder and fly ash, Cem. Concr. Res. 41
(2011) 279–291.
[8] J.J. Chen, A.K.H. Kwan, Y. Jiang, Adding limestone fines as cement paste replacement
to reduce water permeability and sorptivity of concrete, Constr. Build. Mater. 56
(2014) 87–93.
[9] T. Matschei, B. Lothenbach, F.P. Glasser, The role of calcium carbonate in cement hy-
dration, Cem. Concr. Res. 37 (2007) 551–558.
[10] S. Tsivilis, E. Chaniotakis, G. Kakali, et al., An analysis of the properties of Portland
limestone cements and concrete, Cem. Concr. Compos. 24 (2002) 371–378.
 J. Sun, Z. Chen / Powder Technology 338 (2018) 725–733
[11] N. Voglis, G. Kakali, E. Chaniotakis, et al., Portland-limestone cements. Their proper-
ties and hydration compared to those of other composite cements, Cem. Concr.
Compos. 27 (2005) 191–196.
[12] I. Soroka, N. Setter, The effect of fillers on strength of cement mortars, Cem. Concr.
Res. 7 (1977) 449–456.
[13] K. Celik, M.D. Jackson, M. Mancio, et al., High-volume natural volcanic pozzolan and
limestone powder as partial replacements for Portland cement in self-compacting
and sustainable concrete, Cem. Concr. Compos. 45 (2014) 136–147.
[14] L. Pelletier-Chaignat, F. Winnefeld, B. Lothenbach, et al., Beneficial use of limestone
filler with calcium sulphoaluminate cement, Constr. Build. Mater. 26 (2012)
619–627.
[15] B. Lothenbach, G. Le Saout, E. Gallucci, et al., Influence of limestone on the hydration
of Portland cements, Cem. Concr. Res. 38 (2008) 848–860.
[16] E.F. Irassar, Sulfate attack on cementitious materials containing limestone filler-A re-
view, Cem. Concr. Res. 39 (2009) 241–254.
[17] Z. Zhang, Q. Wang, H. Chen, Properties of high-volume limestone powder concrete
under standard curing and steam-curing conditions, Powder Technol. 301 (2016)
16–25.
[18] S. Liu, Q. Li, J. Song, Study on the grinding kinetics of copper tailing powder, Powder
Technol. 330 (2018) 105–113.
[19] Q. Wang, P. Yan, J. Yang, Influence of steel slag on mechanical properties and dura-
bility of concrete, Constr. Build. Mater. 47 (2013) 1414–1420.
[20] J. Sun, Z. Wang, Z. Chen, Hydration mechanism of composite binders containing
blast furnace ferronickel slag at different curing temperatures, J. Therm. Anal.
Calorim. 131 (2018) 2291–2301.
[21] A.A. Elgalhud, R.K. Dhir, G. Ghataora, Limestone addition effects on concrete poros-
ity, Cem. Concr. Compos. 72 (2016) 222–234.
[22] H. Hornain, J. Marchand, V. Dohot, et al., Diffusion of chloride ions in limestone filler
blended cement pastes and mortars, Cem. Concr. Res. 25 (1995) 1667–1678.
[23] Q. Wang, J. Yang, H. Chen, Long-term properties of concrete containing limestone
powder, Mater. Struct. 50 (2017) 168.
[24] A.P. Barker, D.W. Hobbs, Performance of Portland limestone cements in mortar
prisms immersed in sulfate solutions at 5 °C, Cem. Concr. Compos. (2) (1999)
129–137.
[25] S. Tsivilis, G. Kakali, A. Skaropoulou, et al., Use of mineral admixtures to prevent
thaumasite formation in limestone cement mortar, Cem. Concr. Compos. 25
(2003) 969–976.
[26] G. Kakali, S. Tsivilis, A. Skaropoulou, et al., Parameters affecting thaumasite forma-
tion in limestone cement mortar, Cem. Concr. Compos. 25 (2003) 977–981.
[27] S.A. Hartshorn, J.H. Sharp, R.N. Swamy, Thaumasite formation in Portland-limestone
cement pastes, Cem. Concr. Res. 29 (1999) 1331–1340.
[28] M.E. Gaze, N.J. Crammond, The formation of thaumasite in a cement: lime: sand
mortar exposed to cold magnesium and potassium sulfate solutions, Cem. Concr.
Compos. 22 (2000) 209–222.
[29] J. Bensted, Thaumasite-background and nature in deterioration of cements, mortars
and concretes, Cem. Concr. Compos. 21 (1999) 117–121.
[30] N.J. Crammond, The thaumasite form of sulfate attack in the UK, Cem. Concr.
Compos. 25 (2003) 809–818.
[31] S. Liu, P. Yan, Effect of limestone powder on microstructure of concrete, J. Wuhan
Univ. Technol. 25 (2010) 328–331.
[32] M.A. González, E.F. Irassar, Effect of limestone filler on the sulfate resistance of low
C3A Portland cement, Cem. Concr. Res. 28 (1998) 1655–1667.
[33] Q. Wang, D. Wang, H. Chen, The role of fly ash microsphere in the microstructure
and macroscopic properties of high-strength concrete, Cem. Concr. Compos. 83
(2017) 125–137.
[34] S. Liu, P. Yan, J. Feng, Effect of limestone powder and fly ash on magnesium sulfate
resistance of mortar, J. Wuhan Univ. Technol. 25 (2010) 700–703.
[35] S.M. Torres, J.H. Sharp, R.N. Swamy, et al., Long term durability of Portland-lime-
stone cement mortars exposed to magnesium sulfate attack, Cem. Concr. Compos.
25 (2003) 947–954.
733
[36] S.T. Lee, R.D. Hooton, H. Jung, et al., Effect of limestone filler on the deterioration of
mortars and pastes exposed to sulfate solutions at ambient temperature, Cem.
Concr. Res. 38 (2008) 68–76.
[37] N. Thaulow, S. Sahu, Mechanism of concrete deterioration due to salt crystallization,
Mater. Charact. 53 (2004) 123–127.
[38] L.H.J. Martin, F. Winnefeld, C.J. Müller, et al., Contribution of limestone to the hydra-
tion of calcium sulfoaluminate cement, Cem. Concr. Compos. 62 (2015) 204–211.
[39] A. Ipavec, R. Gabrovšek, T. Vuk, et al., Carboaluminate phases formation during the
hydration of calcite-containing Portland cement, J. Am. Chem. Soc. 94 (2011)
1238–1242.
[40] S. Liu, T. Zhang, Y. Guo, et al., Effects of SCMs particles on the compressive strength
of micro-structurally designed cement paste: inherent characteristic effect, particle
size refinement effect, and hydration effect, Powder Technol. 330 (2018) 1–11.
[41] P. Theoulakis, A. Moropoulou, Microstructural and mechanical parameters deter-
mining the susceptibility of porous building stones to salt decay, Constr. Build.
Mater. 11 (1997) 65–71.
[42] V. Bonavetti, H. Donza, G. Menéndez, et al., Limestone filler cement in low w/c con-
crete: a rational use of energy, Cem. Concr. Res. 33 (2003) 865–871.
[43] D.P. Bentz, T. Sato, I. de la Varga, et al., Fine limestone additions to regulate setting in
high volume fly ash mixtures, Cem. Concr. Compos. 34 (2012) 11–17.
[44] M.L. Nehdi, A.R. Suleiman, A.M. Soliman, Investigation of concrete exposed to dual
sulfate attack, Cem. Concr. Res. 64 (2014) 42–53.
[45] E.F. Irassar, V.L. Bonavetti, M. González, Microstructural study of sulfate attack on or-
dinary and limestone Portland cements at ambient temperature, Cem. Concr. Res. 33
(2003) 31–41.
[46] T. Schmidt, B. Lothenbach, M. Romer, et al., A thermodynamic and experimental
study of the conditions of thaumasite formation, Cem. Concr. Res. 38 (2008)
337–349.
[47] K. Sotiriadis, E. Nikolopoulou, S. Tsivilis, Sulfate resistance of limestone cement con-
crete exposed to combined chloride and sulfate environment at low temperature,
Cem. Concr. Compos. 34 (2012) 903–910.
[48] Z. Liu, D. Deng, G.D. Schutter, Does concrete suffer sulfate salt weathering, Constr.
Build. Mater. 66 (2014) 692–701.
[49] J. Amarchand, Sulfate Attack on Concrete, Spon, London, 2001.
[50] D. Benavente, M.A. García Del Cura, J. García-Guinea, et al., Role of pore structure in
salt crystallisation in unsaturated porous stone, J. Cryst. Growth 260 (2004)
532–544.
[51] D. Benavente, J. Martínez-Martínez, N. Cueto, et al., Salt weathering in dual-porosity
building dolostones, Eng. Geol. 94 (2007) 215–226.
[52] E. Ruiz-Agudo, F. Mees, P. Jacobs, et al., The role of saline solution properties on po-
rous limestone salt weathering by magnesium and sodium sulfates, Environ. Geol.
52 (2007) 305–317.
[53] C. Cardella, D. Benavente, J. Rodríguez-Gordillo, Weathering of limestone building
material by mixed sulfate solutions. Characterization of stone microstructure, reac-
tion products and decay forms, Mater. Charact. 59 (2008) 1371–1385.
[54] N. Thaulow, S. Sahu, Mechanism of concrete deterioration due to salt crystallization,
Mater. Charact. 53 (2004) 123–127.
[55] W.T. Kuo, C.C. Liu, J.Y. Wang, Evaluation of the sulfate resistance of fly ash and slag
concrete by using modified ACMT, Constr. Build. Mater. 49 (2013) 40–45.
[56] H. Siad, S. Kamali-Bernard, H.A. Mesbah, et al., Characterization of the degradation of
self-consolidating concretes in sodium sulfate environment: influence of different
mineral admixtures, Constr. Build. Mater. 47 (2013) 1188–1200.
[57] A. Bonakdar, B. Mobasher, Multiparameter study of external sulfate attack in
blended cement materials, Constr. Build. Mater. 24 (2010) 61–70.
[58] Z. Zhang, Q. Wang, H. Chen, et al., Influence of the initial moist curing time on the
sulfate attack resistance of concretes with different binders, Constr. Build. Mater.
144 (2017) 541–551.
[59] T. Aye, C.T. Oguchi, Resistance of plain and blended cement mortars exposed to se-
vere sulfate attacks, Constr. Build. Mater. 25 (2011) 2988–2996.
[60] R.J. Flatt, Salt damage in porous materials: how high supersaturations are generated,
J. Cryst. Growth 242 (2002) 435–454.