Construction and Building Materials: Meysam Najimi, Nader Ghafoori, Mohammadreza Sharbaf

Construction and Building Materials 164 (2018) 625–643
Contents lists available at ScienceDirect
Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat
Alkali-activated natural pozzolan/slag mortars: A parametric study

Meysam Najimi a,b,⇑, Nader Ghafoori a, Mohammadreza Sharbaf a
a
Department of Civil and Environmental Engineering and Construction, University of Nevada Las Vegas, Las Vegas, NV, USA
b
Department of Civil, Construction and Environmental Engineering, Iowa State University, Ames, IA, USA
h i g h l i g h t s
Various properties of alkali-activated natural pozzolan/slag mortars are provided.

Effects of different variables on properties of alkali-activated mortars are studied.
Alkali-activated mortars with proper properties are produced.
a r t i c l e i n f o a b s t r a c t
Article history: This study was devoted to assess the influence of binder’ aluminosilicate precursor combination (natural
Received 21 August 2017 pozzolan/slag combinations), activator combination (sodium silicate/sodium hydroxide combinations),
Received in revised form 7 December 2017 sodium hydroxide concentration, and alkaline activator solution-to-binder ratio (S/B) on properties of
Accepted 27 December 2017
alkali-activated natural pozzolan/slag mortars. The evaluated properties included flow, setting time,
Available online 6 January 2018
hydration heat, compressive strength, absorption, rapid chloride permeability, and drying shrinkage.
The performance of alkali-activated natural pozzolan/slag mortars were also compared with the perfor-
Keywords:
mance of Portland cement mortars. The results of this study revealed that in comparison with the
Alkali activation
Natural pozzolan
Portland cement mortars, alkali-activated natural pozzolan/slag mortars generated significantly lower
Slag heat, thus offering a great benefit to mass concreting. Their hydration activity accelerated and setting
Fresh properties time shortened with increases in sodium hydroxide concentration and dosage within the studied ranges,
Strength increases in slag portion of binder, and decreases in S/B. The compressive strengths of alkali-activated
Transport properties mortars were lower than those of the Portland cement mortars. It was shown, however, doable to pro-
Sodium hydroxide concentration duce alkali-activated natural pozzolan/slag mortars reaching the strength of portland cement mortars.
Activator combination The compressive strength improved with increases in sodium silicate dosage and reduction in S/B.
Binder combination
Regarding compressive strength, the optimum binder combination and sodium hydroxide concentration
Alkaline solution-to-binder ratio
were found to be interdependent of one another. While absorption of alkali-activated mortars was
slightly higher than that of the portland cement mortars, their chloride penetration depths were signif-
icantly lower than those of the portland cement mortars. The chloride penetration depth reduced with
decreases in natural pozzolan portion of binder, sodium silicate dosage, NaOH concentration, and
solution-to-binder ratio. A similar trend was seen for the drying shrinkage as reduction of these variables
also decreased the drying shrinkage. The drying shrinkage of the studied alkali-activated natural poz-
zolan/slag mortars was considerably higher than those of the portland cement mortars.
Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction
Concrete, as a foundational component to infrastructure in any

modern society, faces two major concerns. The first concern stems
from its main component, portland cement (PC) which its produc-
Abbreviations: AANPSL, Alkali-activated natural pozzolan/slag; S/B, Alkaline
activator solution-to-binder ratio; PC, Portland cement; NaOH, Sodium hydroxide; tion is responsible for approximately 7% of world’s total CO2 emis-
RCPT, Rapid chloride penetration test; RMT, Rapid chloride migration test. sion. Long-term material durability is another concern as PC
⇑ Corresponding author at: Department of Civil, Construction and Environmental concrete is not as long lasting as many believe. Past reports have
Engineering, Iowa State University, Ames, IA, USA.
E-mail address: [email protected] (M. Najimi).
https://doi.org/10.1016/j.conbuildmat.2017.12.222
0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.
626 M. Najimi et al. / Construction and Building Materials 164 (2018) 625–643
frequently documented failures of PC concrete infrastructures due tor combination (sodium hydroxide and sodium silicate combina-
to aging and deterioration [1–3]. tions of 80/20, 75/25 and 70/30), NaOH concentration (0.5, 1 and
The desire to reduce CO2 emissions and to produce more dur- 2 M), and alkaline activator solution-to-binder ratio (S/B of 0.52,
able concrete has given impetus to search for new binders. In 0.56 and 0.6). Additionally, two portland cement mortars were
recent years, several studies have been conducted to find alterna- made for the purpose of comparison. The properties studied
tive binders for PC [1–3]. These research investigations have included fresh properties (flow, setting time and heat of hydra-
resulted in introducing different types of binders such as alkali- tion), compressive strength, transport properties (rapid chloride
activated binders, calcium aluminate cements, calcium carbonate penetration test, rapid chloride migration test, and absorption),
cements, magnesium oxy-carbonate cements, Bellite-calcium and volume stability (drying shrinkage).
sulphoaluminate-ferrite cements, and partially pre-hydrated C-S-
H based binders; from which several researchers suggested
2. Literature on alkali-activated natural pozzolan/ slag binders
alkali-activated binders have the potential to lead the transition
to the cement of the future [1–3]. Low CO2 emission and adequate
Fig. 1 presents the number of publications on fly ash/slag and
durability of alkali-activated binders have been proven in various
natural pozzolan/slag combinations. It can be seen that there has
studies [4–7]. These binders are not only a more sustainable and
been a growing research activity on the activation of combined
economical choice due to reducing PC consumption and providing
binders mainly using fly ash/slag combinations. There have been
adequate durability, but also they (1) incorporate industrial by-
a total of 33 publications on fly ash/slag binders as listed in Table 1.
products such as slag and/or fly ash, thus resulting in reduction
On the other hand, there have been only three research studies
of disposal sites, and (2) are suitable for stabilization/solidification
reported on alkaline activation of combined natural pozzolan and
of hazardous and radioactive wastes [8–10].
slag. A brief review of the studies on natural pozzolan/slag binders
There are two major types of alkali-activated binders; low cal-
is presented below:
cium and high calcium binders. For several decades, a great deal
Allahverdi et al. [13] replaced 5, 15 and 25% of natural pozzolan
of research has been dedicated to alkali activation of blast furnace
with blast-furnace slag. In addition to SEM and XRD tests, they
slag (high-calcium binders). In recent decades, the research has
assessed effects of these replacements on compressive strength
been shifted towards activation of low-calcium binders including
and setting time of the produced pastes. They reported that the
fly ash, to a larger extent, and metakaolin and natural pozzolans,
setting times of mixtures were long and use of slag didn’t consid-
to a lesser extent. Each of these alkali-activated binders provide
erably affect the setting times. By analyzing the individual values
certain benefits, in particular in terms of durability. The N-A-S-H
in their study, setting times even slightly increased by replacing
gel (Na2O-Al2O3-SiO2-H2O) as the main reaction product of low-
a portion of natural pozzolan with slag. They also observed that
calcium binders offers excellent chemical and thermal resistances,
the strength of pastes having natural pozzolan and slag was similar
while C-(N)-A-S-H (Na2O-Cao-Al2O3-SiO2-H2O) or C-A-S-H (Cao-
to or slightly lower than that of pastes containing only natural poz-
Al2O3-SiO2-H2O) as the primary reaction products of high-
zolan. The observation of this study was in contrast with the
calcium binders provide chemical binding of water which reduces
expectation driven from the literature on use of blast-furnace slag.
permeability. ‘‘The N-A-S-H gel may also offer anion-binding
The authors attributed this finding to the increases in Si/Al of the
mechanisms that retard chloride ingress, thus increasing service
binder by use of slag which served to inhibit strength develop-
life of reinforced concrete by prolonging the time taken to initiate
ment. The observed behavior, however, could also be related to
corrosion of embedded steel” [11]. Aside from their benefits, these
unsuitability of the used slag for the purpose of alkaline activation.
binders face two major concerns which can potentially limit their
Robayo et al. [14] did a similar study by replacing 10, 20, 30% of
viability for construction materials including (1) the need of low
natural pozzolan with slag. They measured setting time, hydration
calcium-based alkali-activated binders (such as alkali-activated
heat, and compressive strength of the produced pastes. In contrast
fly ash and natural pozzolan) for curing at elevated temperatures,
to the results reported by Allahverdi et al. [13], they noticed signif-
which limits their applications to precast concrete products, and
icant increase in compressive strength and considerable reduction
(2) the fast setting time of high calcium-based alkali-activated bin-
in setting times through partial replacement of natural pozzolan
ders (e.g. alkali-activated slag).
with slag. These replacements also led to a considerable increase
In order to address these concerns and take advantage of bene-
in heat of hydration.
fits offered by each of low- and high-calcium binders, authors com-
In the most recent study, Jafari Nadoushan and Ramezanian-
bined low calcium and high calcium binders [12]. It was found that
pour [15] replaced 5, 10, 15, 20, 50, 75 and 100% of slag with nat-
a suitable combination of low calcium with high calcium binders
ural pozzolans (pumice) and assessed flow and strength of the
along with a proper selection of alkaline concentration mitigated
produced pastes. For the purpose of activation, they used sodium
the need for elevated temperatures as well as the concern for fast
or potassium hydroxides (6 and 8 M) combined with sodium sili-
setting times [12]. While the suggested combination showed the
cate and provided a constant sodium silicate-to-hydroxide ratio
capability to alleviate the abovementioned concerns, further inves-
of 0.4. In general, flow of alkali-activated pastes reduced by inclu-
tigation was needed to assess properties of the resulting binder as
sion of natural pozzolans. This observation was related to the
influenced by mixture ingredients and proportions. The literature
higher specific surface area of pumice in comparison with that of
on combination of low- and high- calcium binders; in particular
slag. They also noticed improvement in 91-day strength of the
on combination of natural pozzolan and slag, is limited. Fig. 1
studied pastes by replacing 5–25% of slag with natural pozzolan.
shows that while there is a growing attention to the combined bin-
The higher replacement level resulted in a significant strength
ders (Fig. 1(a)), combination of natural pozzolan (low-calcium bin-
reduction. The optimum replacement level was found to be 5–10%.
der) and slag (high-calcium binder) has been rarely investigated.
As shown in Fig. 1(b), there are only 3 studies which dealt with
fresh and strength properties of natural pozzolan/slag pastes. 3. Research significance
The study presented herein is devoted to a parametric study on
the properties of alkali-activated natural pozzolan/slag (AANPSL) Lack of prior research activity on alkali-activated natural poz-
mortars. The studied parameters were binder’ aluminosilicate pre- zolan/slag binders was the impetus for this study to assess the
cursor combination hereafter called binder combination (natural effects of a number of influential parameters; i.e. binder combina-
pozzolan and slag combinations of 30/70, 50/50 and 70/30), activation, activator combination, S/B and alkaline concentration, on var-
M. Najimi et al. / Construction and Building Materials 164 (2018) 625–643 627
Fig. 1. (a) Total number of research on fly ash/slag and natural pozzolan/slag binders from 1996 to 2015, (b) Number of research on fly ash/slag and natural pozzolan/slag
binders in different categories of paste, mortar and concrete till 2016; obtained from Scopus database searched by different keywords in the ‘‘article title, abstract, keywords”
section.
ious properties of alkali-activated natural pozzolan/slag mortars. solid part was a combination of sodium oxide (Na2O) and silicon
The findings of this study not only provide a set of complete data dioxide (SiO2) with SiO2-to-Na2O ratio (modulus) of almost 2
on alkali-activated natural pozzolan/slag binders, but also con- (1.98). The used sodium silicate had a pH of 12.9 and specific grav-
tribute to and support for development of hybrid low-calcium ity of 1.53.
and high-calcium alkali-activated binders as an alternative to tra- The fine aggregate passing the gradation size requirements of
ditional cementitious materials. ASTM C33 [51] was used in this study. The fine aggregate was nat-
ural siliceous river sand with a specific gravity of 2.78, absorption
of 0.81%, and fineness modulus of 2.64.
4. Experimental program
4.1. Materials 4.2. Mixture proportions
The materials used in this study included natural pozzolan and Table 4 documents the mixture proportions of alkali-activated
slag as binders; sodium hydroxide (NaOH) and sodium silicate natural pozzolan/slag mortars. As mentioned earlier, these mix-
solutions as alkaline activators; fine aggregates; and distilled tures were designed with slag contents of 30%, 50% and 70% (nat-
water. ural pozzolan contents of 70%, 50% and 30%, respectively); sodium
The chemical properties of natural pozzolan and slag are docu- silicate contents of 20%, 25% and 30% (sodium hydroxide contents
mented in Table 2. Table 3 presents the chemico-physical require- of 80%, 75% and 70%, respectively); NaOH concentrations of 0.5, 1
ments of natural pozzolan and slag based on ASTM standards. It and 2 M; and alkaline activator solution-to-blinder ratios (S/B) of
can be seen that the summation of silica, alumina and iron oxide 0.52, 0.56 and 0.60. A constant fine aggregate-to-binder ratio of 2
was 79.6% for the used natural pozzolan, exceeding the 70% mini- was used for all alkali-activated natural pozzolan/slag mortars
mum level for class N raw and calcined natural pozzolans specified which was selected from a preliminary research on different ratios
in ASTM C618 [49]. The calcium oxide content of natural pozzolan of 1.5, 2, 2.5 and 3 [12]. In this Table, mixture id. XM-SLY-SSZ-W
was low (2.3%), indicating that this material was not cementitious represents that NaOH solution had X mole, slag was Y% of the total
on its own. The 7-, and 28-day compressive strengths of mortars binder ((100-Y)% was for natural pozzolan), sodium silicate consti-
having 20% natural pozzolan were 84 and 91% of the prepared tuted Z% of the total activator ((100-Z)% was sodium hydroxide),
standard reference mortar, respectively, which were well above and W was the solution-to-binder ratio.
the minimum requirement of 75% specified in ASTM C618 [49]. It For the purpose of comparison, two control portland cement
can also be seen that the used slag contained 43.6% calcium oxide, mortars were also made; Table 5 reports their mixture proportions.
offering cementitious properties. Its other major compounds were These mixtures were designed with fine aggregate-to-binder ratios
silica (31%) and alumina (11.5%). The 7-, and 28-day compressive of 2.75 (standard mortar) and 2 (similar to the alkali-activated
strengths of the mortars having 50% slag were 90 and 107% of mortars). A water-to-cement ratio of 0.485 was used for portland
the prepared standard reference mortar, respectively; well above cement mortars in accordance with ASTM C109 [52].
the minimum requirement for Grade 100 slag specified in ASTM
C989 [50].
Sodium hydroxide pellets were used to make NaOH solutions. 4.3. Testing
The sodium hydroxide pellets were composed of 97 to 98.2%
sodium hydroxide, up to 1.2% sodium chloride and 0.4 to 1.0% Flowability of alkali-activated and PC mortars was measured in
sodium carbonate. NaOH solutions were made with various con- accordance with ASTM C1437 [53] and upon batching. Setting
centrations including 0.5, 1 and 2 mol (M). A liter of solution with times of mortars were measured by Vicat and modified Vicat
these molarities was prepared by solving 20, 40 and 80 g of sodium (ASTM C807 [54]) upon batching. Two types of needles were used
hydroxide pellets in enough water to reach the volume of one liter, having 1 and 2 mm diameters. The setting time was defined as the
respectively. After conducting a preliminary study on different time that 1 mm- and 2 mm-needles penetrate 25 and 10 mm in the
types of sodium silicate solution [12], Type D sodium silicate with mortar within 30 s, respectively. Heat of hydration of mortars was
modulus of 2 was selected. Type D sodium silicate was in liquid measured by use of 100 100 mm cylindrical samples encased in
form composed of 55.97% water and 44.03% solid by weight. The insulated boxes.
Table 1
Previous works on alkali-activated fly ash/slag systems [12,16–48].
Year References Mixture type Studied properties

2000 Puertas et al. [16] Paste Compressive strength, mineralogical characteristics
2002 Puertas et al. [17] Mortar Sulfate resistance
2003 Puertas and Fernandez-Jimenez [18] Paste Mineralogical and microstructural characteristics, compressive strength
Puertas et al. [19] Mortar Compressive and flexural strengths, freeze-thaw
2006 Escalante Garcia [20] Paste Compressive strength, XRD
2007 Zhao et al. [21] Paste Setting time, SEM, XRD
2010 Kumar et al. [22] Paste Strength properties, setting time, heat of hydration
Izquiredo et al. [23] Paste Leaching
2012 Luo et al. [24] Paste Use of specified design method/ simulation
Wang et al. [25] Paste Addition of metakaolin; mineralogical and microstructural characteristics, compressive strength
2013 Puligilla and Mondal [26] Paste Ultrasonic, compressive strength
Lee and Lee [27] Paste/concrete Setting time and mechanical properties
Kar et al. [28] Concrete Compressive strength, ultrasonic
Chi and Huang [29] Mortar Strength, water absorption, drying shrinkage, SEM, XRD
Ismail et al. [30] Paste Sulfate resistance, mineralogical characteristics
Ismail et al. [31] Mortar/concrete Compressive strength, transport properties
Rashad [32] Concrete Flow, strength, shrinkage
2014 Ismail et al. [33] Paste SEM, XRD, FTIR, TGA
Jang et al. [34] Paste Setting time, flow, heat of hydration
Ghafoori and Najimi [35] Mortar Fresh, mechanical and transport properties
Lee et al. [36] Paste/mortar Shrinkage, XRD
Yang et al. [37] Paste Chloride diffusion
2015 Gao et al. (3 studies) [38–40] Paste Heat of hydration, compressive strength, mineralogical characteristics
Chi et al. [41] Mortar Compressive strength, length change, water absorption, SEM, XRD
Marjanovic et al. (2 studies) [42,43] Mortar Strength properties, shrinkage, SEM
Deb et al. [44] Concrete Compressive strength, shrinkage
Wang et al. [45] Paste Flow, compressive strength, ultrasonic, sulfate attack
Wardhono et al. [46] Mortar Compressive strength
Lee and Lee [47] Paste Microstructure, reaction products
2016 Abdalqader et al. [48] Paste Compressive strength, SEM, XRD, TGA, hydration
Table 2 silver chloride forms when silver nitrate reaches and reacts with
Chemical compositions of natural pozzolan, slag, and portland cement. penetrated chloride).
Compound, % Natural pozzolan Slag Portland cement
Calcium oxide (CaO) 2.3 43.64 63.31 5. Results and discussions
Silica (SiO2) 71.0 31.0 20.42
Alumina (Al2O3) 7.9 11.5 4.25
5.1. Flowability
Iron oxide (Fe2O3) 0.7 0.8 4.05
Sodium oxide (Na2O) 3.2 – 0.04
Potassium oxide (K2O) 4.3 0.84 0.69 Table 6 documents the flow spread of alkali-activated natural
Sulfur trioxide (SO3) 0.1 4.85 2.98 Pozzolan/slag (AANPSL) mortars. Fig. 2 shows overall influence of
Magnesium oxide (MgO) – 4.7 2.0 different variables on the workability of the studied mortars. The
Titanium oxide (TiO2) – 0.57 –
influence of each variable is discussed below.
Potential compound, %
C3S – – 53
C2S – – 18 5.1.1. Effects of binder combination
C3A – – 4 The influence of binder combination on the flow can be seen in
C4AF – – 12 Table 6 and Fig. 2(a) and (b). In general, increases in slag content
(reduction of natural pozzolan dosage) resulted in lower flowabil-
ity. On average, there were about 18 and 33% reduction in the flow
Compressive strength of mortars was measured at ages of 1, 3, of mixtures having S/B of 0.52, 0.56 and 0.60 when slag content
7, 28 and 90 days based on ASTM C109 [52] using 75 by 150 mm was increased from 30 to 50 and 50 to 70%, respectively. The aver-
cylinders. Drying shrinkage was measured in accordance with age reductions were about 24, 20 and 0% for each 20% increase in
ASTM C596 [55] using 25 25285 mm prisms. For the purpose slag content of mortars having sodium hydroxide concentrations
of this test, samples were cured till the age of 7 days before moving of 0.5, 1 and 2 M, respectively. This observation can be related to
to humidity chamber (50 ± 5% relative humidity). the finer particle size distribution of the used slag in comparison
Absorption, density and volume of permeable voids of the stud- with that of the used natural pozzolan. As it was shown in Table 3,
ied mortars were measured based on ASTM C642 [56]. This test used slag had specific surface of 5420 cm2/g and 2.6% retained on
was conducted on 28-day mortar disks having diameter of 100 45 mm sieve, whereas these values were 4505 cm2/g and 8.7% for
mm and height of 50 mm. Similar sized disks and the same curing used natural pozzolan, respectively.
age were used for measurement of rapid chloride penetration
(RCPT) as per ASTM C1202 [57] and rapid chloride migration 5.1.2. Effects of activator combination
(RMT) based on NT BUILD 492 [58]. RCPT measures the charge The effect of activator combination on the flow spread of
passed through the concrete sample and relates it to chloride ion AANPSL mortars can be seen in Fig. 2(a) and Table 6 (comparing
penetration. On the other hand, the measurement in RMT is by the results presented in rows 4 to 6, 16 to 18, and 19 to 21). It
spraying silver nitrate on the split sample and observing the depth was found that increases in sodium silicate content (decreases in
of penetrated chloride through formation of silver chloride (white sodium hydroxide dosage) reduced the workability of the studied
Table 3
Chemico-physical properties of natural pozzolan and slag according to ASTM standards.
Requirements Natural pozzolan Slag

Requirements (ASTM C618 [49]) Results Requirements (ASTM C989 [50]) Results
Chemical SiO2+Al2O3+Fe2O3, % Min, 70.0 79.6 – 43.3
Sulfur trioxide (SO3), % Max, 4.0 0.1 – 4.85
Sulfide sulfur, % as SO3 – – Max, 2.5 0.66
Sulfate Ion, % as SO3 – – Max, 4 3.2
Moisture content, % Max, 3.0 0.2 – –
Loss on ignition, % Max, 10.0 3.4 Max, 10.0 0.3
Physical Amount retained when wet-sieved on 45 lm sieve, % Max, 34 8.7 Max, 20 2.6
Strength activity index, at 7 days, percent of control Min, 75 84 Min, 75* 90
Strength activity index, at 28 days, percent of control Min, 75 91 Min, 95* 107
Water requirement, percent of control Max, 115 98 – –
Air content of mortar, % – – Max, 12 5.8
Autoclave expansion or contraction, % Max, 0.8 0 Max, 0.5 0
Specific gravity – 2.33 – 2.87
Specific surface, cm2/g – 4505 – 5420
*
For Grade 100 slag based on ASTM C989[50].
Table 4
Mixture proportions of alkali-activated natural pozzolan/slag mortars.
Mix ID. Mixture description Materials, g

Slag Natural pozzolan NaOH solution Sodium silicate
2M-SL30-SS30-0.60 30% slag, 30% sodium silicate, 2M sodium hydroxide, S/B = 0.60 1583 3694 2217 950
0.5M-SL30-SS30-0.60 30% slag, 30% sodium silicate, 0.5M sodium hydroxide, S/B = 0.60 1583 3694 2217 950
*Fine aggregate was 10,556 g for all the mixtures.
Table 5
Mixture proportions of portland cement mortar.
Mix ID. Fine aggregate-to-cement ratio Water-to-cement ratio Materials, g

Portland cement Water Fine aggregate
Control 1 (standard PC mortar) 2.75 0.485 4486 2176 12,338
Control 2 2 5278 2560 10,556
mixtures. The flow reduced by averagely 11% for each 5% increases AANPSL mortars via use of higher dosages of sodium silicate can
in sodium silicate content. It should, however, be noted that the be attributed to higher solid-to-liquid ratio of sodium silicate in
flow table limited measurement of above 15.24 cm, which in turn, comparison with that of sodium hydroxide. The solid-to-liquid
didn’t allow for exact comparisons. It’s also worth mentioning that ratio was 44.03/55.97 = 0.787 for sodium silicate solution, whereas
the average reduction can’t be extrapolated to sodium silicate con- it was about 0.02, 0.04 and 0.08 for sodium hydroxide solutions
tents of above 30%. In preliminary study on AANPSL pastes [12], it with concentrations of 0.5, 1 and 2 M, respectively.
was seen that the flow significantly reduced when sodium silicate
increased from 30 to 40%. To confirm these findings, two additional 5.1.3. Effects of alkaline concentration
mixtures were made with 35 and 40% sodium silicate contents Table 6 (rows 1–9) and Fig. 2(b) present the flow spreads of
(having slag dosage of 30%). Their flow spreads were 14 and 81% AANPSL mortars with different NaOH concentrations. It can be
lower than that of 30% sodium silicate contained mortar, respec- seen that the flowability increased as sodium hydroxide molarity
tively. The reduction in flow increased drastically for sodium sili- increased from 0.5 to 1 M, after which it reduced with increasing
cate content of above 35%. The reduction in workability of molarity from 1 to 2 M. When NaOH molarity increased from 0.5
Table 6
Fresh properties.
Row No. Mix ID. Flow (cm) Setting times Heat of hydration
(minutes)
Initial Final Peak temperature (°C) Time to reach peak temperature (hr)
1 2M-SL30-SS30-0.60 11.1 146 220 30.50 10.5
2 2M-SL50-SS30-0.60 11.1 121 169 36.05 8.3
3 2M-SL70-SS30-0.60 11.1 109 141 38.60 6.0
4 1M-SL30-SS30-0.60 14.4 245 403 28.45 20.4
5 1M-SL50-SS30-0.60 12.4 230 313 33.05 10.1
6 1M-SL70-SS30-0.60 8.6 129 180 36.30 6.8
7 0.5M-SL30-SS30-0.60 12.7 325 523 28.85 25.6
8 0.5M-SL50-SS30-0.60 8.9 264 349 31.50 11.1
9 0.5M-SL70-SS30-0.60 6.7 150 210 34.25 7.4
10 1M-SL30-SS30-0.56 11.6 150 256 29.30 17.3
11 1M-SL50-SS30-0.56 9.5 168 244 33.10 8.8
12 1M-SL70-SS30-0.56 7.6 99 135 36.85 6.1
13 1M-SL30-SS30-0.52 10.2 118 219 29.65 14.1
14 1M-SL50-SS30-0.52 7.9 104 178 34.10 8.3
15 1M-SL70-SS30-0.52 4.1 58 98 37.05 5.7
16 1M-SL30-SS20-0.60 15.2 293 325 29.85 13.5
17 1M-SL50-SS20-0.60 15.2 138 193 33.35 6.8
18 1M-SL70-SS20-0.60 15.1 73 98 34.60 5.0
19 1M-SL30-SS25-0.60 15.2 270 425 29.40 18.2
20 1M-SL50-SS25-0.60 15.2 200 285 33.40 9.4
21 1M-SL70-SS25-0.60 13.0 118 150 35.75 6.0
22 Control 1 15.2 155 243 43.25 10.6
23 Control 2 15.2* 176 255 47.90 10.0
*This flow was achieved in 18 drops instead of 25 times. In fact, the flow was higher than 15.2 cm but the flow Table didn’t allow for further measurement.
**Mix ID. XM-SLY-SSZ-W means that NaOH solution had X mole, slag was Y% of the total binder, sodium silicate constituted Z% of the total activator, and W was the solution-
to-binder ratio.
tively. The observed behavior was similar to the observations of

the preliminary study on AANPSL pastes [12] where overall use
of higher sodium hydroxide reduced the workability, but mixtures
activated with 0.5 M sodium hydroxide and sodium silicate had
higher workability than mixtures activated with water and sodium
silicate. These trends were similar to the observations of Jolicouer
et al. [59] where using 0.5, 1 and 2% Na2O in the system of pastes
resulted in lower workability than inclusion of 4% Na2O. It was also
observed by Smaoui et al. [60] that addition of NaOH to portland
cement concrete resulted in higher workability. In addition, it
was reported by Li et al. [61] that the flow of high performance
cementitious mortars increased with increasing alkaline content
to an optimum level, after which it significantly reduced for higher
alkaline dosages. Therefore, it can be inferred that a certain amount
of alkalis can increase the workability, after which increasing
NaOH concentration results in reduction of flow due to the reduc-
tion of water content. For a given volume, increases in NaOH
molarity reduce water portion of the activator solution.
5.1.4. Effects of alkaline activator solution-to-binder ratio

Table 6 also shows the flow of AANPSL mortars having different
solution-to-binder ratios (rows 4 to 6, 10 to 12, and 13 to 15).
These results are also presented in Fig. 2(a). It was seen that the
workability reduced with decreases in S/B. The flow reduced by
averagely 18 and 25% when S/B decreased from 0.60 to 0.56 and
Fig. 2. Flow of alkali-activated mortars; (a) effects of binder combination, activator
0.56 to 0.52, respectively. This trend was similar to the typical
combination and S/B on the flow of mixtures having NaOH concentration of 1 M; (b)
Effects of NaOH concentration on mixtures with S/B = 0.60. behavior of portland cement. ‘‘As the mixtures with higher S/B con-
tained more water, they became more plastic and workable, allow-
ing for ease of mixing and placement” [62].
to 1 M, the flow of the studied mortars increased by 1.7, 3.5 and
1.9 cm for mortars having slag contents of 30, 50 and 70%, respec- 5.2. Setting time
tively. For 70% slag contained mortars, the increases in molarity
from 1 to 2 M resulted in 2.5 cm further increases in the flow. By The setting times of alkali-activated natural pozzolan/slag mor-
increasing the molarity from 1 to 2 M, the flow reduced by 3.3 tars are reported in Table 6. The AANPSL mortars had a wide range
and 1.3 cm for mixtures with slag contents of 30 and 50%, respec- of setting times which were dependent on combinations of binders
ing sodium silicate contents of 20, 25 and 30%, the peak

temperature increased by averagely 2.38, 3.18 and 3.93 °C and
time of peak shortened by averagely 4.23, 6.09 and 6.80 h for each
20% increases in slag contents, respectively. Similarly, each 20%
increases in slag content resulted in averagely 3.70, 3.78 and
3.93 °C increases in peak temperature of mortars having S/B of
0.52, 0.56 and 0.60, respectively. For similar solution-to-binder
ratios, the time of peak temperature decreased by averagely 4.18,
5.59 and 6.80 h for each 20% increase in slag content, respectively.
It can also be seen that for 0.5 M mortars, the time to reach maxi-
mum temperature reduced from 25.6 to 7.4 h (71% reduction)
when slag content was increased from 30 to 70%. This gap was
13.6 h (from 20.4 to 6.8 h with 67% reduction) in case of 1 M mor-
Fig. 3. Effects of slag content, molarity and S/B on initial setting time of alkali- tars, and 4.5 h (from 10.5 to 6 h with 43% reduction) for 2 M mor-
activated mortars. tars. These observations follow a similar trend found in the
preliminary study on AANPSL pastes [12] indicating that as molar-
ity of sodium hydroxide solution increased, the differences
between setting times of the mixtures having different slag con-
tents narrowed.
The higher reactivity of slag compared to natural pozzolan can
be attributed to its higher CaO content than that of natural poz-
zolan as lime is the main chemical component of slag. It was
reported in different studies that higher lime content accelerated
the hydration reaction of the mixture [27,63]. It should also be
noted that while available lime in slag could react with water,
sodium hydroxide and sodium silicate, the reactions of natural
pozzolan only occur with strong alkaline solutions and with a
slower reactivity rate (longer reaction times). The higher reactivity
Fig. 4. Initial setting time of alkali-activated mortars having different sodium
of slag than natural pozzolan can also be easily concluded by com-
silicate contents. paring the behavior of alkali-activated slag and alkali-activated
natural pozzolan binders in the literature. It was shown that while
alkali-activated slag binders set quickly as a sign of fast reactions,
and activators used. Effects of these parameters on the setting natural pozzolan based geopolymers set late without use of ele-
time, which are also shown in Figs. 3 and 4, are discussed below. vated temperatures [12].
5.2.1. Effects of binder combination 5.2.2. Effects of activator combination

The influence of binder combination on the setting times of Table 6 and Fig. 4 present the setting times of AANPSL mortars
AANPSL mortars can be seen in Table 6 and Figs. 3 and 4. Overall, having different activator combinations. It can be seen that overall,
the setting times reduced with reduction of natural pozzolan con- increases in sodium silicate content in the range of 20 to 30%
tents (increases in slag-to-total binder ratio). In the case of AANPSL delayed the setting times of AANPSL mortars.
mortars having S/B of 0.6, an increase in slag content from 30 to The initial and final setting times increased by averagely 33 and
50% resulted in 61, 15 and 25 min reduction in the initial setting 44% when sodium silicate content was increased from 20 to 25%,
time and 175, 90 and 51 min reduction in the final setting time respectively. Further increases in sodium silicate from 25 to 30%
for sodium hydroxide concentrations of 0.5, 1 and 2 M, respec- resulted in averagely 5 and 8% increases in the initial and final set-
tively. When slag content was increased from 50 to 70%, the initial ting times, respectively. These results were similar to the observa-
setting times decreased by 114, 101 and 13 min, and final setting tions of preliminary study on AANPSL pastes [12] that the setting
time reduced by 139, 133 and 28 min for similar NaOH molarities, time increased when sodium silicate content increased from 20
respectively. Similarly, an overall reduction in the setting times to 30%. It was, however, observed in preliminary study that further
was observed with increases in slag content for mortars having dif- increases in sodium silicate dosage (from 30 to 40%) led to shorter
ferent sodium silicate contents and S/B. For sodium silicate con- setting times for AANPSL pastes. To confirm this finding, the setting
tents of 20, 25 and 30%, the setting times of mortars having 70% times of mortars having 35 and 40% sodium silicate were measured
natural pozzolan (30% slag) were averagely 30.2 and 61.5% higher for 30% slag contained mixtures. Use of 35 and 40% sodium silicate
than those of mortars containing 50 and 30% natural pozzolan (50 significantly reduced the setting times. The setting times of mor-
and 70% slag), respectively. In the case of mixtures with S/B of 0.52, tars containing 35 and 40% sodium silicate were averagely 62
0.56 and 0.60, when slag content was increased from 30 to 50 and and 80% lower than those of mortars having 30% sodium silicate,
50 to 70%, the setting times reduced by averagely 9 and 44%, respectively. Therefore, the conclusion of delaying setting times
respectively. through increases in sodium silicate content was only valid for
The observed behaviors can be related to the higher reactivity of sodium silicate contents of 20 and 30%. Further increases in
slag in comparison with that of natural pozzolan as confirmed with sodium silicate content showed reversing trends for both AANPSL
the results of hydration heat. It was seen in Fig. 5(a) that by pastes and mortars.
increasing the slag dosage, AANPSL mortars generated more heat These results are well supported with the findings of hydration
and reached maximum temperature in a shorter period of time heat. Fig. 5(b) and Table 6 show the time that mixtures with differ-
which points to their faster reactivity. Table 6 shows the maximum ent sodium silicate contents reached their peak temperature. It can
generated temperatures and their occurrence time for mixtures be seen that similar to the setting time results, as sodium silicate
having different slag contents. These results confirm the above- content was increased, it took longer time for AANPSL mortars to
mentioned conclusion for all the studied mortars. For mortars hav- reach their maximum temperature. It can also be seen that the heat
Fig. 5. Effect of different variables on hydration heat of alkali-activated mortars; (a) binder combination, (b) activator combination, (c) solution-to-binder ratio, (d) NaOH
molarity.
generation happened sooner for mixtures with lower sodium sili- S/B contained more water, evaporation of the extra water took
cate contents (higher sodium hydroxide dosages). For the mixtures longer time, resulting in slower hardening of the mortars.
having 30% slag, the maximum temperature happened 4.74 and It can be deduced from these observations that increasing S/B
2.20 h later as sodium silicate was increased from 20 to 25% and delayed the reaction activities which were also observed by Shi
25 to 30%, respectively. In the case of mortars having 50 and 70% and Day [68] when water-to-slag ratio of alkali-activated slag
slag, the peak temperature delayed by 2.58 and 1.02 h when pastes increased. The delayed reactivity, specially seen for mix-
sodium silicate content was increased from 20 to 25%, and retarded tures having high natural pozzolan content, are likely due to the
by 0.68 and 0.78 h when silicate content was increased from 25 to delay in geopolymerization stage and product formation by
30%, respectively. increasing water content of the mixture.
These observations can be related to the effects of initial pH on
early hydration which is the dominant factor for initial reactivity.
In a number of publications, it was stated that the early hydration 5.2.4. Effects of sodium hydroxide concentrations
of slag was accelerated as the pH of activator increased [64-66]. Fig. 3 and Table 6 show the influence of NaOH concentration on
The main reason for this explanation was the effect of pH on disso- the setting times of AANPSL mortars. It can be seen that increases
lution of slag. It was shown that percentage of soluble silica from in molarity of NaOH solution reduced the setting times. The
slag almost doubled when pH increased from 13 to 14 [10,67]. Sim- observed reduction narrowed as slag content was increased.
ilarly, increases in pH elevates silica dissolution of aluminosilicates The setting times of AANPSL mortars having 30% slag contents
[10]. In case of this study, the used sodium silicate had pH of 12.9, reduced by averagely 23.8 and 42.8% when sodium hydroxide con-
whereas the pH of sodium hydroxide solution was about 14. There- centrations increased from 0.5 to 1 and 1 to 2 M, respectively. Sim-
fore, the accelerated reactions with increases in sodium hydroxide ilar increases in molarity of NaOH solution reduced the setting
portion of activator can be related to the increases in the pH of the times by averagely 11.6 and 46.6% for mortars having 50% slag,
whole activator. and 14.2 and 18.5% for mortars containing 70% slag, respectively.
This observation can be related to the accelerated activation by
increasing the alkaline concentration as dissolution of aluminosil-
5.2.3. Effects of alkaline activator solution-to-binder ratio icates occurred faster with increases in alkaline concentrations,
The influence of S/B on the initial and final setting times of thus leading to accelerated reactions.
AANPSL mortars are presented in Fig. 3 and Table 6. It was found Fig. 5(d) and Table 6 also present the time to reach the peak
that the setting times reduced as S/B decreased. For each 0.04 temperature for AANPSL mortars having different sodium hydrox-
reduction in S/B, the initial setting time reduced by averagely ide concentrations. The results shown in Table 6 are in agreement
26.0, 27.4 and 27.7% for mixtures having slag contents of 30, 50 with the observations reported for setting time. For mixtures with
and 70%, respectively. For similar slag dosages, the final setting slag contents of 30, 50 and 70%, the final setting times of 0.5 M
time shortened by averagely 22.8, 21.6 and 22.9% for each 0.04 mixtures were 2.38, 2.07 and 1.49 times of that of 2 M mixtures,
reduction in S/B, respectively. This behavior was similar to the typ- respectively. For similar slag contents, time to reach the peak tem-
ical trend seen for portland cement. This observation is also in line perature for 0.5 M mixtures were 2.44, 1.33 and 1.22 times more
with the results of hydration heat (time to reach the maximum than those obtained for 2 M mixtures, respectively. A similar
temperature) presented in Fig. 5(c). As the mixtures with a higher behavior was observed by Shi and Day [68] and Song et al. [66]
reporting an increased cumulative hydration heat by increasing oxide within the alkali-activated fly ash/slag binder is essential
NaOH dosage of alkali-activated slag pastes. In particular, Song to the high early strength development.
et al. [66] found a linear relationship between dq/dtmax (heat evo- As the curing age was extended, the reaction rate of natural
lution) and NaOH concentration and an inverse relationship pozzolan increased. The compressive strength of mixtures with
between tmax (time to reach peak temperature) and NaOH concen- higher pozzolan contents increased at a higher rate than mixtures
tration. This observation was related to faster hydration of slag in with lower natural pozzolan dosages. This narrowed the large gap
higher NaOH concentrations. A similar outcome took place for nat- observed at early ages between mixtures having different slag con-
ural pozzolan as dissolution of silica increased with an increase in tents. The ratios of 7-, 28- and 90-day compressive strengths of
NaOH molarity [69]. mortars with 70% slag to those of mixtures with 30% slag were
1.28, 1.04 and 1.05 for 0.5 M mixtures; 1.24, 1.05 and 1.12 for 1
5.3. Compressive strength M mixtures; and 1.69, 1.36 and 1.45 for 2 M mixtures, respectively.
The 7-, 28- and 90-day compressive strengths of mortars with
The compressive strengths of alkali-activated natural pozzolan/ 70% slag were 8, 10 and 6% lower than those of mortars with 50%
slag mortars were measured at the ages of 1, 3, 7, 28 and 90 days, slag for sodium hydroxide concentration of 0.5 M, respectively. In
and their results are presented in Table 7. Overall, the 28-day com- the case of 1 M, 70% slag contained mixtures had 4, 7 and 9% lower
pressive strengths of AANPSL mortars were in a broad range of 20 7-, 28- and 90-day compressive strengths than 50% slag contained
to 47 MPa depending on the binder combination, activator combi- mixtures, respectively. At similar ages, the compressive strengths
nation, sodium hydroxide concentration, and S/B. The influence of of 70% slag mortars were averagely about 13% higher than those
these factors is discussed in the sections to follow. of 50% slag mortars for sodium hydroxide concentration of 2 M.
From these observations, it can be concluded that the optimum
slag content increased with increases in sodium hydroxide concen-
5.3.1. Effects of binder combination
tration. It increased from about 50% for sodium hydroxide concen-
Influences of binder combination (slag and natural pozzolan
tration of 0.5 and 1 M to 70% for 2 M mixtures.
contents) on the compressive strength of mortars are shown in
The overall increases in compressive strength through increases
Table 7. At early age of curing (1 and 3 days), increases in slag
in slag content can be related to the presence of CaO which
dosage showed a positive influence on the compressive strength.
improved strength of the binder by forming an amorphously struc-
At these ages, the studied mortars with 70% slag developed higher
tured Ca–Al–Si gel [27], resulting in a denser microstructure. This
strength than mortars having 50 and 30% slag with the latter
was also reported in a study by Yip [63]. Kumar et al. [22] also sug-
gained the lowest compressive strengths. The 1-day compressive
gested that the increase in compressive strength with the addition
strengths of mixtures with 70% slag were 5.58, 4.11 and 2.79 times
of slag could result from the formation of gel phases (C-S–H and A-
of those of mixtures having 30% slag, and 1.17, 1.17 and 1.28 times
S–H) and compactness of the microstructure.
of those of 50% slag contained mixtures for molarities of 0.5, 1 and
2 M, respectively. For similar molarities, the 3-day compressive
strengths of mortars having 70% slag were 1.77, 1.81, and 1.83 5.3.2. Effects of activator combination
times of those of mortars containing 30% slag, and 0.98, 1.05 and Fig. 6 presents the influence of activator combination on the
1.21 times of those of 50% slag contained mixtures, respectively. compressive strength of AANPSL mortars. It can be seen that
This observation can be related to the higher reactivity of slag than increases in sodium silicate dosage considerably improved the
that of natural pozzolan at early ages as also corroborated with the compressive strength of AANPSL mortars. The 1-, 3-, 7-, 28- and
results reported for hydration heat and setting time in Subsection 90-day compressive strengths increased by averagely 28.1, 48.6,
5.2.1. This finding also correlates well with the results of Lloyd 55.1, 51.9 and 40.0% when sodium silicate content was increased
et al. [70] who showed from a micro-analysis that the calcium from 20 to 25%, respectively. At the same curing ages, increases
Table 7
Results of compressive strength and ultimate drying shrinkage.
Mix ID. Compressive strength (MPa) Ultimate drying shrinkage (%)

1 day 3 days 7 days 28 days 90 days
2M-SL30-SS30-0.60 9.2 17.2 22.0 34.3 38.8 0.4423
2M-SL50-SS30-0.60 20.1 26.1 32.5 40.0 51.7 0.3542
2M-SL70-SS30-0.60 25.7 31.5 37.1 46.5 56.4 0.3458
1M-SL30-SS30-0.60 5.6 15.9 25.5 38.2 43.2 0.4069
1M-SL50-SS30-0.60 19.7 27.4 33.1 43.1 53.1 0.3075
1M-SL70-SS30-0.60 23.0 28.7 31.7 40.0 48.3 0.3042
0.5 M-SL30-SS30-0.60 3.8 14.0 22.6 35.5 43.5 0.3984
0.5 M-SL50-SS30-0.60 18.1 25.2 31.4 41.0 48.5 0.3329
0.5 M-SL70-SS30-0.60 21.2 24.8 28.9 36.8 45.7 0.3185
1M-SL30-SS30-0.56 7.1 18.4 28.1 39.7 46.2 0.3806
1M-SL50-SS30-0.56 21.2 28.4 33.7 43.2 53.6 0.3078
1M-SL70-SS30-0.56 24.9 29.9 35.1 43.5 51.9 0.3014
1M-SL30-SS30-0.52 11.8 22.2 30.9 41.9 47.9 0.3320
1M-SL50-SS30-0.52 22.8 28.5 34.0 43.2 53.0 0.2824
1M-SL70-SS30-0.52 25.4 29.4 34.7 42.2 49.3 0.2727
1M-SL30-SS20-0.60 8.5 13.4 16.3 21.1 25.5 0.2297
1M-SL50-SS20-0.60 10.7 14.2 16.4 20.9 25.9 0.1982
1M-SL70-SS20-0.60 11.2 12.9 15.1 20.4 26.4 0.1947
1M-SL30-SS25-0.60 6.1 16.4 23.0 29.2 33.1 0.3068
1M-SL50-SS25-0.60 16.9 22.6 25.7 35.1 40.9 0.2449
1M-SL70-SS25-0.60 17.3 21.2 25.3 30.5 34.9 0.2480
Control 1 24.9 37.8 46.9 59.3 71.9 0.0952
Control 2 28.8 42.8 52.2 71.5 82.7 0.1207
increased by averagely 14.2, 7.9, 7.6, 4.3 and 5.1% when S/B
reduced from 0.60 to 0.56, respectively. Further reduction in S/B
from 0.56 to 0.52 led to 25.3, 6.4, 3.2, 0.9 and -0.8% increases in
1-, 3-, 7-, 28- and 90-day compressive strengths, respectively.
It was also observed that the improvements were less signifi-
cant for mortars with higher slag contents (lower natural pozzolan
dosages); in particular when S/B was changed from 0.56 to 0.52.
This observation could be related to the lower workability of mix-
tures 1 M-SL70-SS30-0.52 (70% slag and S/B of 0.52) and 1 M-SL50-
SS30-0.52 (50% slag and S/B of 0.52) which may have caused insuf-
ficient densification.
The overall reduction of compressive strength by increases in S/
B can be related to the increased porosity. Ravikumar [76] found a
reliable relationship between porosity, compressive strength and
Fig. 6. Compressive strength of alkali-activated mortars with different activator
combination.
activator-to-binder ratio. He stated that an increase in activator-
to-binder ratio increases the porosity of alkali-activated mixture,
which in turn, reduces the compressive strength. It can be said that
in sodium silicate content from 25 to 30% led to 13.8, 17.9, 21.7, as solution-to-binder ratio increases, the amount of evaporable
28.3 and 32.9% improvements, respectively. water in the paste also increases. As evaporable water moves
This observation can be related to the reactions of sodium sili- towards surface, it vacates capillary pores. The higher amounts of
cate with calcium oxide of slag. The onions of sodium silicate react pores results in higher porosity and lower strength and mechanical
with Ca2+ dissolving from the surface of the slag grains and form properties. The results presented in this study also validate this
the primary C-S–H [71]. As more sodium silicate content becomes process as documented by the results of volume of voids which
available, the increased onion of sodium silicate elevates C-S-H for- increased with increases in solution-to-binder ratio (presented in
mation. These observations are in agreement with the findings of Subsection 5.4.3).
different studies presented in the literature. Atis et al. [72] and
Shi and Li [73] observed that in all their studied cases, sodium 5.3.4. Effects of sodium hydroxide concentrations
silicate-activated slags developed higher strengths than sodium Table 7 presents the results of compressive strength for AANPSL
hydroxide-activated slags which indicate more positive effects of mortars with different sodium hydroxide concentrations. It can be
sodium silicate than sodium hydroxide. Also, in various studies seen that the influences of NaOH molarity on the compressive
on low calcium binders, it was observed that activation of fly ash strength was dependent on slag content as the optimum sodium
and natural pozzolan was more effective by combination of sodium hydroxide concentration increased with increases in slag content
silicate and sodium hydroxide than when only sodium hydroxide of the studied mixtures.
was used [69,74]. Lloyd et al. [70] also found from microstructural In the case of AANPSL mortars having 30% slag, the late-age
observations that geopolymer activated with a silicate-based solu- compressive strengths (28 and 90 days) increased as molarity of
tion showed a more homogeneous microstructure than that acti- solution was increased from 0.5 to 1 M, after which they reduced
vated with a NaOH-based solution, which may improve by further increases in NaOH concentration. For these mixtures,
compressive strength. In addition to improvement of microstruc- the 28-day and 90-day compressive strengths increased by aver-
ture, Aydin and Baradan [75] related the strength improvement agely 3.5% when NaOH molarity was increased from 0.5 to 1 M,
to the decreases in Ca/Si ratio of C-S–H with increases in silicate and reduced by 10.2% when it was increased from 1 to 2 M. In fact,
content of activator which improved the binding ability of CAS–H. 2 M mortars developed lower strengths than 0.5 M mortars. The
higher molarity only resulted in higher strengths at early ages.
5.3.3. Effects of alkaline activator solution-to-binder ratio The 1- and 3-day compressive strengths of 2 M mixtures (with
The influence of S/B on the compressive strength of AANPSL 30% slag) were 64.3 and 8.2% higher than those of 1 M mixtures;
mortars was also studied and its typical trend is presented in and 142.1 and 22.9% higher than those of 0.5 M mixtures,
Fig. 7. In general, the compressive strength of studied mortars respectively.
slightly improved with reduction of S/B, displaying a similar A similar trend was observed for mixtures having 50% slag and
behavior to that of portland cement mixtures. The improvements, 50% natural pozzolan; the late-age compressive strength increased
however, reduced as the curing age was increased. The 1-, 3-, 7-, as sodium hydroxide concentration increased from 0.5 to 1 M and
28- and 90-day compressive strengths of AANPSL mortars reduced by further increases in NaOH molarity. The 2 M mixtures,
however, developed higher compressive strengths than 0.5 M mix-
tures. For these mixtures, the 90-day compressive strengths of 1 M
mixtures were 2.7 and 9.5% higher than those of 2 and 0.5 M mix-
tures, respectively.
A different trend was observed for mixtures containing 70% slag
and 30% natural pozzolan as the higher molarity led to higher
strengths for these mixtures. The 1-day, 3-day, 7-day, 28-day and
90-day compressive strengths of 2 M mixtures were 11.7, 9.8,
17.0, 16.3 and 16.8% higher than those of 1 M mixtures; and
21.2, 27.0, 28.4, 26.4 and 23.4% higher than those of 0.5 M mix-
tures, respectively.
While increases in compressive strength of 70% slag contained
mixtures can be correlated to the increases in solubility of the bin-
ders, the strength reduction for mixture having high natural poz-
zolan content (and low slag content) can be corroborated via
Fig. 7. Effects of S/B on the compressive strength of alkali-activated mortars. explanations given by different researchers who studied the effects
of NaOH concentration on the strength of alkali-activated fly ash 5.3.5. Effects of curing age
and metakaolin (aluminosilicates). Palomo et al. [1], Memon Fig. 8 shows the strength development rate of AANPSL mortars
et al. [77] and Alonso and Palomo [78] observed that excessive at different ages in reference to 90-day compressive strength. It can
OH concentration in the system of aluminosilicate-based binders be seen in Fig. 8(a) that as slag content increased, a higher rate of
caused strength reductions. It was stated by Memon et al. [77] that strength development took place. AANPSL mortars with 30, 50 and
excessive hydroxyl ions led to aluminosilicate gel precipitation at 70% slag developed 20, 40 and 48% of their 90-day compressive
the early stages, hindering subsequent geopolymerization to result strength in 1 day. They nearly reached the same level of strength
in lower strength. Alonso and Palomo [78] related the strength development at 7 and 28 days as more than 60 and 80% of strength
reduction to the delay in formation of alkaline polymer as a result were attained for all mixtures at these ages, respectively. It can be
of (1) ‘‘the stability of ion species: polymerization is favored when seen in Fig. 8(b) that for majority of the studied ages, mixtures with
dissolved components are in their molecular form, and an increase 25% silicate developed a higher percentage of their ultimate
of pH provokes higher stability of ion species”, and (2) ‘‘ion mobil- strength compared to mixtures having 20 and 30% silicate. At early
ity: a higher sodium hydroxide concentration provokes an increase age (1-day), 20% sodium silicate contained mortars developed the
in the dissolved species amount, and a greater difficulty in the spe- highest percentage of their 90-day strength.
cies diffusion in the aqueous phase.” Because of the similarity in Fig. 8(c) indicates that as molarity of NaOH solution increased,
chemical compositions amongst fly ash, metakaolin, and natural the studied mortars produced strength at slightly faster rate.
pozzolan; the above explanations can be considered equally valid AANPSL mortars with NaOH concentrations of 0.5, 1 and 2 M
for the studied alkali-activated mixtures having high amounts of gained 31, 33 and 36% of their 90-day strength in 1 day. They
natural pozzolan. nearly reached the same level of strength development at 7 and
28 days as 60 to 62% and 82 to 84% of strength were attained for
all mixtures at these ages, respectively. Variation of S/B also chan-
ged the rate of compressive strength development as shown in
Fig. 8. Compressive strength development within curing age for (a) mixtures with different binder combination, (b) mixtures with different activator combination, (c)
mixtures with different NaOH concentration, (d) mixtures with different S/B.
Table 8
Results of absorption, rapid chloride penetration test and rapid chloride migration test.
Row Mix ID. Absorption after Absorption after immersion Volume of permeable Rapid chloride penetration Rapid chloride migration
No. immersion (%) and boiling (%) voids (%) test (Coulombs) test (mm)
1 2M-SL30- 9.27 9.79 19.73 6613 18.68
SS30-0.60
2 2M-SL50- 8.58 9.25 18.95 1522 4.64
SS30-0.60
3 2M-SL70- 8.71 9.57 19.75 1108 3.15
SS30-0.60
4 1M-SL30- 10.40 10.88 21.74 4958 11.97
SS30-0.60
5 1M-SL50- 8.74 9.37 19.06 944 3.79
SS30-0.60
6 1M-SL70- 8.92 9.93 20.19 743 3.23
SS30-0.60
7 0.5M-SL30- 10.80 11.55 22.77 3511 9.90
SS30-0.60
8 0.5M-SL50- 8.95 9.72 19.63 791 4.09
SS30-0.60
9 0.5M-SL70- 9.74 10.97 22.00 656 2.91
SS30-0.60
10 1M-SL30- 8.74 9.24 18.80 2915 8.51
SS30-0.56
11 1M-SL50- 8.22 8.89 18.24 762 2.76
SS30-0.56
12 1M-SL70- 8.15 9.14 18.81 518 2.43
SS30-0.56
13 1M-SL30- 7.75 8.38 17.20 1601 6.99
SS30-0.52
14 1M-SL50- 7.26 7.93 16.44 516 2.43
SS30-0.52
15 1M-SL70- 8.05 8.84 18.37 467 1.93
SS30-0.52
16 1M-SL30- 10.39 10.66 21.23 652 5.37
SS20-0.60
17 1M-SL50- 8.90 9.19 18.62 443 3.30
SS20-0.60
18 1M-SL70- 10.84 11.27 22.55 448 3.01
SS20-0.60
19 1M-SL30- 10.09 10.41 20.73 1727 6.49
SS25-0.60
20 1M-SL50- 8.98 9.34 19.00 514 3.74
SS25-0.60
21 1M-SL70- 9.65 10.24 20.69 482 3.00
SS25-0.60
22 Control 1 6.87 7.15 15.66 9689 32.95
23 Control 2 8.16 8.33 17.76 10,536 34.04
Fig. 8(d). With increasing S/B, the rate of strength development The observed trend was fairly similar to the trend found for
reduced. At 1, 3 and 7 days, the rate of strength development were compressive strength. For majority of the mixtures, mortars having
in the ranges of 33 to 40%, 49 to 53% and 62 to 66%, respectively. 50% slag performed superior to mortars with 30 and 70% slag in
The lowest boundaries (33, 49 and 62%) were for S/B = 0.6, whereas both absorption and compressive strength tests. Also, for both
the highest boundaries (40, 49 and 66%) were for S/B = 0.52. properties, the differences in performance of mixtures with differ-
ent slag contents narrowed as NaOH molarity was increased.
5.4. Absorption These observations are close to the findings of Ismail et al. [31]
on 28-day absorption of slag/fly ash activated mortars. At this age
Table 8 reports the results of absorption test for all the studied of curing, the performance of their studied mortars was in order of
mortars. The effects of different variables on absorption results are 50/50–75/25–25/75 slag/fly ash combination. A similar trend was
presented below. observed by Chi and Huang [29] where they produced slag/fly
ash activated mortars with different combinations of fly ash and
5.4.1. Effects of binder combinations slag. In case of their studied mortars, the absorption of 50/50 com-
It can be seen in Table 8 that the absorption and permeable bination of slag/fly ash was slightly superior to 75/25 combination;
voids of mortars having 50% slag were lower than those of mortars both were significantly better than 25/75 and 0/100 combinations.
containing 30 and 70% slag. While the absorption results of mor- They related this behavior to the difference of the formed gels and
tars with 30 and 70% slag contents were fairly similar to one finer particle size of slag in comparison with fly ash. ‘‘The slag-rich
another, the 70% slag contained mortars displayed a slightly lower systems predominantly form a C–(A)–S-H gel with a significant
absorption than that of 30% slag contained mortars for most of the bound water content, while the fly ash-rich systems form N–A–
studied cases. The void contents of mortars with 50 and 70% slag S–(H) with a lower bound water content” [29]. Therefore, fly ash-
were lower than those of mortars having 30% slag by 13.8 and rich systems (having more N-A-S-H gels) absorbed more water.
3.4% for 0.5 M mixtures; 12.3 and 7.1% for 1 M mixtures; and 4.0 Furthermore, their used slag had a much finer particle size to fill
and 0% for 2 M mixtures, respectively. the pores resulting in lower water absorption. Due to similarity
between the binder used in this study and that used in the studies
reported by Chi and Huang [29] and Ismail et al. [31], both of the
above explanations are seemingly valid for the studied alkali-
activated natural pozzolan/slag mortars.
5.4.2. Effects of activator combination

The influence of sodium silicate contents on the absorption
properties of AANPSL mortars are presented in Table 8 (rows 4 to
6, 16 to 18, and 19 to 21). As can be seen, while the absorption
results were in the same ranges for majority of the studied natural
pozzolan/slag mortars, the influence of sodium silicate content on
absorption was dependent on the slag content of the mixture. In
case of mortars with 30% slag, absorption of mortars having 20
and 25% sodium silicate content was almost similar. For this slag Fig. 9. Results of RCPT and RMT for alkali-activated natural pozzolan/slag mortars
dosage, increases in sodium silicate content from 25 to 30% with different slag contents and sodium hydroxide concentrations.
resulted in slight increases in absorption and volume of permeable
voids. For mortars having 50% slag, the absorption of mixtures
were almost the same. For these mixtures, absorption increased
slightly as sodium silicate content increased. An opposite trend
was observed for mixtures having 70% slag. For these mixtures,
there were obvious improvements in absorption due to increases
in sodium silicate contents. This observation can be attributed to
the modification of C-S-H gel in the mixtures having high slag
dosage through increases in silicate content. As alluded in discus-
sion for compressive strength, use of higher sodium silicate in
the mortars containing high amounts of slag improved microstruc-
ture and binding capacity of C-S-H gel through reduction of Ca/Si
ratio.

Table 8 shows the results of absorption test for AANPSL mortars Fig. 10. Results of RCPT and RMT for alkali-activated natural pozzolan/slag mortars
having different solution-to-binder ratios. As expected, reduction having different sodium silicate and slag contents.
of S/B improved absorption properties. On average, each 0.04

reduction in S/B resulted in 11.5, 7.7 and 5.5% reduction in absorp-
tion of mortars having 30, 50 and 70% slag, respectively. For similar
slag contents, volume of permeable voids reduced by averagely
10.4, 6.9 and 4.5% for each 0.04 reduction in S/B, respectively.
This behavior is consistent with the results reported for com-
pressive strength and is similar to the behavior seen in portland
cement. The observed behavior may be related to entrapment of
more water within the alkali-activated paste which allowed for
additional capillary voids and more porous microstructure upon
evaporation of migrated pore water to the surface of the studied
specimens.
5.4.4. Effects of sodium hydroxide concentration

The results of absorption test for AANPSL mortars having differ-
ent NaOH concentrations are reported in rows 1–9 of Table 8. It can Fig. 11. Results of RCPT and RMT for alkali-activated natural pozzolan/slag mortars
having different S/B and slag contents.
be seen that absorption reduced as the molarity of NaOH was
increased. The absorption properties (absorption and volume of
permeable voids) of 30% slag contained mortars reduced by aver- 5.5.1. Effects of binder combination
agely 4.7 and 10.0% when the NaOH concentration was increased The influence of binder combination on rapid chloride penetra-
from 0.5 to 1 and 1 to 2 M, respectively. These improvements were tion of alkali-activated natural pozzolan/slag mortars are shown in
3.0 and 1.2% for 50% slag contained mortars, and 8.7 and 2.7% for Figs. 9, 10 and 11 for mixtures having different NaOH molarities,
70% slag contained mortars, respectively. These findings follow sodium silicate contents and S/B, respectively. It can be seen that
the observations made by Melo Neto et al. [79] and Chi and Huang increases in slag content resulted in reduction of passing charges.
[29] where reductions in porosity of alkali-activated slag systems There was a significant reduction in the passing charge, when
with increases in Na2O content were reported. slag content was increased from 30 to 50%. For mixtures with
NaOH concentrations of 0.5, 1 and 2 M, this increase in slag content
5.5. Rapid chloride penetration test led to 77.5, 81.0 and 77.0% reduction in RCPT passing charges,
respectively. For sodium silicate contents of 20, 25 and 30%, the
Table 8 documents the results of rapid chloride penetration test RCPT results of mortars having 50% slag were 32.1, 70.2 and
(RCPT). The general trend for effects of binder combination, activa- 81.0% lower than those of mortars containing 30% slag, respec-
tor combination, sodium hydroxide concentration and S/B on the tively. Similarly, increases in slag content from 30 to 50% resulted
results of RCPT are presented in Figs. 9 through 11. in 81.0, 73.9 and 67.8% reductions in the passing charge of mortars
with S/B of 0.6, 0.56 and 0.52, respectively. Further increases of slag The observed increases can be related to the increases in hydro-
content also reduced the passing charges, but to a lesser extent. On xyl ions. As molarity of sodium hydroxide increased, the hydroxyl
average, the passing charges of mortars with 70% slag were 16% ions increased, resulting in increased passing charges.
lower than those of mortars having 50% slag.
As the amounts of hydroxide ions in pore solution has been
known as a major factor contributing to the charges passing in 5.6. Rapid chloride migration test
rapid chloride penetration test, it can be inferred from these obser-
vations that there were more free hydroxyls available in mixtures Table 8 also reports the results of rapid chloride migration test
with lower slag dosages. While chloride permeability improved (RMT) for all alkali-activated natural pozzolan/slag mortars. The
through increases in slag content, a considerable portion of the effect of different factors on the results is discussed below.
observed reductions could be due to the higher availability of
hydroxyls in the mixtures having lower slag contents. This obser- 5.6.1. Effects of binder combination
vations is consistent with the results of compressive strength Figs. 9, 10 and 11 show effects of slag and natural pozzolan con-
where increases in NaOH concentration was ineffective for mix- tents on the chloride penetration depth of AANPSL mortars with
tures made with 30% slag, whereas use of higher NaOH molarity different NaOH concentrations, sodium silicate contents and S/B,
considerably improved the strength of mixtures having 70% slag. respectively.
It is a suspect that sodium hydroxide solution reacted more effi- Similar to the RCPT results, the higher slag contents led to lower
ciently with slag in comparison with natural pozzolan, thus more chloride penetration. The chloride penetration depth into AANPSL
hydroxyl ions left in the mortars containing high amount of natural mortars considerably reduced when slag content was increased
pozzolan (i.e. 70% by weight of total binder). from 30 to 50%. The penetration depths of mortars having 50% slag
were 58.7, 68.3 and 75.2% lower than those of mortars having 30%
5.5.2. Effects of activator combination slag for NaOH molarities of 0.5, 1 and 2 M, respectively. Increases
Fig. 10 shows the effect of sodium silicate contents on the in slag content from 30 to 50% resulted in 38.5, 42.4 and 68.3%
results of rapid chloride penetration test. It can be seen that the improvements (reductions) in the chloride penetration depths of
passing charges increased with increases in sodium silicate con- mortars having sodium silicate contents of 20, 25 and 30%; and
tent. As sodium silicate was increased from 20 to 25%, the passing 68.3, 67.6 and 65.2% reduction in penetration depth of mixtures
charges increased by almost 165, 16 and 7% for mortars having slag with S/B of 0.6, 0.56 and 0.52, respectively. Further increases in
contents of 30, 50 and 70%, respectively. For the similar slag con- slag reduced the chloride penetration depths, but to a lesser extent.
tents, increases in sodium silicate dosage from 25 to 30% resulted The penetration depths reduced by averagely 19.5% when slag con-
in 187, 84 and 54% increases in the passing charges, respectively. tents increased from 50 to 70%.
The increases in the RCPT values through increases in sodium Ismail et al. [31] observed a similar trend in their RMT studies
silicate content can be related to the higher alkalis content of using fly ash/slag-based alkali-activated concretes. They stated
sodium silicate in comparison with that of sodium hydroxide. that the degree of penetration increased with the inclusion of
While 14.8% of sodium silicate was alkalis, sodium hydroxide solu- higher fly ash in the binder. They attributed their findings to ‘‘the
tions were prepared by dissolving 20 g (almost 2%), 40 (almost 4%) change of the microstructure of the system from a dense C-AAS-
and 80 g (almost 8%) NaOH pellets in nearly a liter of water for con- H binding gel to a more porous hybrid (CAN)-AAS-H type gel”
centrations of 0.5, 1 and 2 M, respectively. It is worth noting that [31]. Similarly, Yang et al. [37] reported that ‘‘the incorporation
Douglas et al. [80] also observed increases in RCPT passing charges of slag as a secondary precursor in fly ash-based geopolymers con-
by increasing sodium silicate-to-slag ratio of alkali-activated slag tributed to the refinement of pore structure and thus restricted the
concretes. movement of chloride ions in the paste.” Due to the similarity in
binder chemical compositions, it can be deduced that the studied
alkali-activated natural pozzolan/slag mortars also behaved simi-
5.5.3. Effects of alkaline activator solution-to-binder ratio larly once the natural pozzolan content of the binder varied.
The results of RCPT for AANPSL mortars with different solution-
to-binder ratios are shown in Fig. 11. Considerable reductions in
the passing charges were observed through reduction of S/B. For 5.6.2. Effects of activator combination
slag content of 30%, the passing charges of mortars with S/B of Fig. 10 presents the results of rapid chloride migration test for
0.52 were 45.1 and 67.7% lower than those having S/B of 0.56 AANPSL mortars having different activator combinations. It can
and 0.6, respectively. These improvements reduced to 32.3 and be seen that similar to the results of rapid chloride penetration
45.3% for slag content of 50%; and 9.8 and 37.1% for slag content test, the chloride penetration depth increased with increases in
of 70%, respectively. sodium silicate content. Increases in sodium silicate content from
The observed improvements by use of lower S/B can be related 20 to 25% and 25 to 30% resulted in 20.9 and 84.4% increases in
to the improvement in pore structure (reported for the results of chloride penetration depth of mortars having slag contents of
absorption and compressive strength) as well as reduction of free 30%, respectively. The increases were not as significant for mix-
hydroxyl ions through reduction of alkaline solution content. tures with slag contents of 50 and 70%. Similar increases in sodium
silicate content reduced the chloride penetration depth by 13.3 and
1.3% for 50% slag contained mixtures; and -0.3 and 7.7% for 70%
5.5.4. Effects of sodium hydroxide concentrations slag contained mixtures, respectively.
The influence of NaOH concentrations on the RCPT charges is The significant increases in chloride penetration depth of 30%
presented in Fig. 9. It can be seen that the passing charges slag contained mixtures can be explained through the results
increased as NaOH concentration increased. Increases in NaOH obtained for absorption and RCPT of the studied mortars. The
molarity from 0.5 to 1 M caused 1477, 153 and 187 C increases increases in the results of these two tests with increases in sodium
in the passing charges of mortars having 30, 50 and 70% slag, silicate content suggested that the pore structure of 30% slag con-
respectively. For similar slag contents, the increases in NaOH contained mortars were weakened with the higher sodium silicate
centration from 1 to 2 M resulted in 1655, 578 and 365 C increases content. This, in turn, allowed for higher penetration of chloride
in the passing charges, respectively. ions.

Fig. 11 shows the chloride penetration depths of AANPSL mor-
tars having different solution-to-binder ratios. Decreases in S/B
led to noticeable reductions in the depth of penetration. As S/B
was reduced from 0.6 to 0.56, the penetration depth decreased
by 14.0, 32.5 and 16.5% for mortars containing 30, 50 and 70% slag,
respectively. For similar slag contents, the reduction of S/B from
0.56 to 0.52 led to 17.9, 12.0 and 20.6% reductions in penetration
depths, respectively. The higher amounts of capillary pores as a
result of higher solution-to-binder ratio facilitated movement of
chloride ions through the studied mortars.
5.6.4. Effects of sodium hydroxide concentrations

Fig. 9 presents the results of rapid chloride migration test for Fig. 13. Drying shrinkage of alkali-activated natural pozzolan/slag mortars having
AANPSL mortars having NaOH concentrations of 0.5, 1 and 2 M. different S/B and slag content of 30% and 50%
Overall, increases in sodium hydroxide concentrations resulted in
increased chloride penetration depth. This was particularly seen
in mortars having 30% slag and 70% natural pozzolan. The chloride reduction, however, was more significant when slag content was
penetration depths of 30% slag contained mortars activated with 2 increased from 30 to 50% in comparison with when it was
M sodium hydroxide were 8.8 and 6.7 mm higher than those hav- increased from 50 to 70%. The ultimate drying shrinkages of 0.5
ing 0.5 and 1 M NaOH, respectively. In case of mortars having 50 M mortars having 50 and 70% slag were 16.4 and 20.1% lower than
and 70% slag, there was a slight difference between penetration those of mortars containing 30% slag, respectively. In case of 1 and
depths of mixtures having different sodium hydroxide concentra- 2 M mortars, the mortars containing 50% slag shrank 24.4 and
tions. For these slag contents, the penetration depths of 2 M mix- 19.9% lower than mortars having 30% slag, respectively. Similar
tures were averagely 0.75 and 0.05 mm more than those of to the 0.5 M mortars, 1 M and 2 M mortars having 50 and 70% slag
mixtures with lower NaOH concentrations of 0.5 and 1 M, shrank at almost the same levels.
respectively. Similar trends were seen in mixtures prepared with different
The observed increases in penetration depth by increasing sodium silicate contents and S/B. For mixtures with sodium silicate
NaOH concentration from 0.5 (and 1) to 2 M for 30% slag contained dosages of 20, 25 and 30%, the ultimate drying shrinkages of 30%
mixtures may be related to its microstructure which also produced slag contained mortars were averagely 19.4 and 19.9% higher than
lower compressive strength. those of mortars having 50 and 70% slag, respectively. For S/B of
0.60, 0.56 and 0.52, the ultimate drying shrinkage improved by
24.4, 19.1 and 14.9%, when slag content was increased from 30
5.7. Drying shrinkage
to 50%, respectively. For similar S/B, the increases in slag content
from 50 to 70% led to only 1.1, 2.1 and 3.4% reduction in the ulti-
The alkali-activated natural pozzolan/slag mortars were trans-
mate drying shrinkage, respectively.
ferred to the shrinkage room (with relative humidity of 50 ± 5%
It can also be seen in Fig. 12 (1M mixtures) that 30% slag con-
and a temperature of 24 °C) after 7 days of curing in the moist-
tained mortar reached its ultimate shrinkage within two weeks
curing room. The drying shrinkage of samples was measured for
to a month. In case of mixture containing 50% slag, it took almost
a period of about 5 months. The ultimate drying shrinkages of
1.5 to 2.5 months to reach the ultimate shrinkage. This time was
alkali-activated mortars are reported in Table 7. Figs. 12 through
between 2.5 and 3 months for 70% slag contained mixture. This
14 show the typical trends for different studied variables affecting
trend was also seen for mixtures with sodium hydroxide concen-
drying shrinkage of the studied alkali-activated natural pozzolan/
trations of 0.5 and 2 M.
slag mortars.
The reduced shrinkage due to increases in slag content was
opposite to the trend reported by Rashad [32], Chi and Huang
5.7.1. Effects of binder combination [29], Chi et al. [41], and Deb et al. [44] on fly ash/slag based mix-
The typical influence of binder combination on the drying tures who observed higher shrinkage through increases in slag por-
shrinkage is presented in Fig. 12 (mixtures with sodium silicate tion of the binder. In these studies, two reasons were used to
dosage of 30% and different slag contents). It can be seen that the explain this observation: (1) Chi et al. [41] related shrinkage
drying shrinkage reduced with increases in slag content. The increases to chemical shrinkage (reduction in volume resulting
from the chemical reaction between the reagent and water) of
the produced gels. The main reaction products of alkali-activated
binders having high slag content are hydrotalcite and C-S-H gel
with high chemical shrinkage, which lead to an increase in the
total amount of shrinkage. (2) Rashad [32] attributed the reduction
in drying shrinkage of alkali-activated fly ash/slag mortars by
increasing fly ash portion of binder to the spherical shape of fly
ash particles which acted as micro-aggregate increasing volume
stability of the mixture. From this point of view, unlike fly ash, nat-
ural pozzolan has angular shape particles.
There were also two studies showing a nearly similar trend to
the observations of this investigation. In the study conducted by
Marjanovic et al. [43], drying shrinkage of mortar with 25/75 com-
bination of fly ash/slag was almost similar to or slightly more than
Fig. 12. Effect of binder and activator combinations on the drying shrinkage of that of 0/100 combination. Yusuf et al. [81] observed that the dry-
alkali-activated natural pozzolan/slag mortars.
subsequent increases in the drying shrinkage of the studied alkali-

activated natural pozzolan/slag mortars.

Influence of S/B on the drying shrinkage of AANPSL mortars is
shown in Figure 13. As expected, the drying shrinkage reduced
with decreases in S/B. The ultimate drying shrinkage of mortars
with S/B of 0.52 was 12.8 and 18.4% lower than that of mortars
having S/B of 0.56 and 0.60 for slag content of 30%, respectively.
These improvements narrowed to 8.3 and 8.2% for 50% slag con-
tent; and 9.5 and 10.4% for 70% slag content, respectively.
The increases in drying shrinkage, due to increases in solution-
to-binder ratio, can be related to the paste weakened microstruc-
ture and higher availability of evaporable moisture which is
Fig. 14. Drying shrinkage of alkali-activated natural pozzolan/slag mortars having responsible for volume instability. Chi et al. [84] also observed a
different NaOH concentrations. similar trend with increases in liquid-to-slag ratio.
ing shrinkage of ground steel slag/ultrafine palm ash pastes and 5.7.4. Effects of sodium hydroxide concentrations
mortars reduced by increasing slag portion of the binder. They Influence of NaOH concentrations on the drying shrinkage of
related the significant improvement in shrinkage by increasing slag AANPSL mortars are shown in Fig. 14. Overall, the drying shrink-
content in preventing the internal micro-cracks and enhancing ages of 2 M mixtures were higher than those of 0.5 and 1 M mix-
pore-filling and pore refinement. The reason was that hydration tures. The 1 M mixtures shrank almost similar to that of 0.5 M
of slag generated additional C-S-H that changed the macrospores mixtures.
into the mesopores through pore filling or pore refinement. This The ultimate drying shrinkage of 2 M mixtures were 8.0, 13.2
consequently reduced the shrinkage of the mortars and corrobo- and 12.0% higher than that of 1 M mixtures for slag contents of
rated the pore filling capability of slag. The role of Ca2+ as a pore 30, 50 and 70%, respectively. For the same amounts of slag, 1 M
filler and a network modifier in reducing drying shrinkage was also mixtures shrank averagely 3.6% less than 0.5 M mixtures,
reported by Yip et al. [82]. Due to similarity in trend, coupled with respectively.
the results of mercury intrusion porosimetry (porosity) observed A similar trend was observed by Atis el al. [72] where drying
by authors [12], it is reasonable to suspect that the explanations shrinkage of alkali-activated slag mortars increased by increasing
given by Yusuf et al. [81] are also valid for the studied alkali- sodium hydroxide concentrations. Aydin and Baradan [75] also
activated natural pozzolan/slag mortars. observed increases in drying shrinkage with an increase in Na2O
content of activator. Similar to the results of this study, they found
the effect of Na2O content on drying shrinkage to be less significant
as compared to the influence of silicate-to-Na2O content (silicate
5.7.2. Effects of activator combination
content of activator) on drying shrinkage of alkali-activated
The typical trend for effects of activator combination on the
mortars.
drying shrinkage of AANPSL mortars is shown in Fig. 12. It was
seen that the drying shrinkage increased with increases in sodium
silicate dosage (decreases in sodium hydroxide content). In case of 6. Comparison with portland cement mortars
30% slag content, the mortars with 30% sodium silicate shrank 77.1
and 32.6% more than mortars containing 20 and 25% sodium sili- As discussed earlier, for the purpose of comparison, two refer-
cate, respectively. The ultimate drying shrinkage of mortars having ence mixtures were designed using portland cement as the sole
50 and 70% slag increased by 23.6 and 27.4% when sodium silicate binder. Tables 5 and 6 report their mixture proportions and fresh
was increased from 20 to 25%; and by 25.6 and 22.7% as sodium sil- properties, respectively. Their flow was 15.24 cm, thus they were
icate content was increased from 25 to 30%, respectively. densified and casted easily. No sign of bleeding was observed. Their
A similar trend was observed by Atis et al. [72] studying alkali- initial and final setting times were in the ranges of 155 to 176 and
activated pastes. They observed higher drying shrinkage by 242 to 255 min, respectively. Compared to the portland cement
increasing the sodium silicate content. Cincotto et al. [83] also mortars, the alkali-activated natural pozzolan/slag mortars had a
observed a similar trend for autogenous shrinkage. Similarly, Deb wider range of setting times which were dependent on combina-
et al. [44] found increases in drying shrinkage of fly ash/slag based tions of binders and activators used. The results of flow and setting
mixtures by increasing sodium silicate content. Aydin and Baradan time confirmed that acceptable fresh properties can be attained
[75] observed increases in drying shrinkage by increasing silica-to- using proper combinations of binders and appropriate activator
Na2O ratio of activator (modulus of sodium silicate). They related dosage and combination.
the increases in drying shrinkage through increasing silicate con- Table 6 also shows the peak temperature and time to reach
tent of activator to the change in microstructure of mortars. They maximum temperature of portland cement mortars. Control 2 with
stated that ‘‘shrinkage under various conditions depends on the lower amounts of fine aggregate had higher peak temperature,
loss of water from the mesopores (pore diameter between 2.5 which was due to its higher amounts of cement for a given volume.
and 50 nm) and also the size of the macropores (pore diameter It can be seen that portland cement mortars generated consider-
between 50 and 10,000 nm), which determines easiness of evapo- ably more heat than alkali-activated natural pozzolan/slag mor-
ration of water from the mesopores.” They attributed the increases tars. The peak temperature of Control 1 and 2 were 43.25 and
of shrinkage by increasing the silicate content to the observed 47.9 °C, respectively, whereas the peak temperature of alkali-
increases in volume of mesopores. A similar process may have activated mortars ranged between 28.4 and 38.6 °C. There was a
taken place in this investigation as an increase in sodium silicate considerably wider range (5–26 h) for AANPSL mortars to reach
content and decrease in sodium hydroxide concentration increased the peak temperature, as compared to that of portland cement
silica-to-Na2O ratio, leading to increases in mesopores volume and mortars (10–10.6 h). Lower heat of hydration can offer a great ben-
efit for use of alkali-activated natural pozzolan/slag mortars in restrain volume changes due to drying shrinkage, the gap in drying
mass concrete production. This finding represents an advantage shrinkage between portland cement and alkali-activated concretes
of alkali-activated natural pozzolan/slag binder over portland are expected to be narrowed.
cement.
The compressive strengths of portland cement mortars are 7. Conclusion
shown in Table 7. The reference mortars Control 1 and Control 2
developed 28-day compressive strengths of 59 and 72 MPa, respec- In this study, different properties of alkali-activated natural
tively. These strengths were higher than those of alkali-activated pozzolan/sag mortars as influenced by changing binder combina-
natural pozzolan/slag mortars ranging between 20 and 47 MPa. It tion, activator combination, S/B and sodium hydroxide concentra-
should be noted that higher strengths (more than the abovemen- tion, were investigated. From the findings of this study, the
tioned range) could be produced with alkali-activated natural poz- following conclusions can be drawn:
zolan/slag mortars. As an example, only increases of sodium
silicate to 35% were tried for 1 M mixtures having 50 to 70% slag 1) It was found that adequate fresh properties can be achieved
which increased the strength to about 55 MPa [12]. Also, as a next with proper combinations of mixture proportions. However,
phase of this study, alkali-activated natural pozzolan/slag con- proper design of alkali-activated binders is more compli-
cretes were made which produced a similar compressive strength cated than portland cement as there are many influential
to that of PC concrete [12]. Overall, these results demonstrate that variables affecting outcomes. Nearly for all studied alkali-
with proper selections of binders and activators, there was neither activated natural pozzolan/slag mortars, the flow and setting
need for curing in elevated temperatures, nor there was any issue times were in the acceptable ranges. The flow and setting
with time of setting of the studied AANPSL mortars. times of alkali-activated mortars reduced with decreases in
The results of rapid chloride penetration, rapid chloride migra- natural pozzolan content and solution-to-binder ratio. The
tion and absorption tests of portland cement mortars are docu- setting time of alkali-activated mortars also reduced by
mented in Table 8. The absorption (after boiling), volume of increasing sodium hydroxide dosage and concentration.
permeable voids, passing charge and chloride penetration depth The flow improved by increasing sodium hydroxide portion
were averagely 7.74%, 16.71%, 10,113 C and 33.5 mm, respectively. of activator and NaOH molarity for up to 1 M.
Unlike alkali-activated mortars, portland cement mortars showed a 2) In comparison with the control portland cement mortars,
very high passing charges. The average charges passed through alkali-activated natural pozzolan/slag mortars generated
alkali-activated natural pozzolan/slag mortars were almost 80% significantly lower heat of reactions which can offer a great
lower than those of the studied portland cement mortars. Simi- benefit for mass concrete production. The results of hydra-
larly, the chloride penetration depths of alkali-activated natural tion heat showed that the initial reactivity accelerated as
pozzolan/slag mortars were considerably lower than those of port- slag portion of binder increased, sodium hydroxide portion
land cement mortars. The lowest penetration depth was for alkali- of activator increased, S/B reduced, and sodium hydroxide
activated mortar 1 M-SL70-SS30-0.52 (1.93 mm) which was 94% concentration increased.
lower than that of portland cement mortars. The average chloride 3) It was found achievable to produce alkali-activated mixtures
penetration depth for AANPSL mortars was 85% of that of portland reaching the strength of portland cement mixtures. The
cement mortars. This significant reduction of chloride penetration/ compressive strength of alkali-activated natural pozzolan/
migration offers a major benefit for use of alkali-activated natural slag mortars improved with increases in sodium silicate por-
pozzolan/slag materials, in particular where corrosion of reinforce- tion of activator and decreases in solution-to-binder ratio.
ment and chemical durability of concrete are of concern. This per- 4) The strength development accelerated with increases in slag
formance was further corroborated by evaluating corrosion of portion of binder. Accordingly, at the early ages of curing (1
alkali-activated natural pozzolan/slag concrete [12]. The average and 3 days), the strength increased with increases in slag
total pores of AANPSL mortars were slightly higher than those of content resulting in the highest early strengths in the mix-
portland cement mortars. It should, however, be noted that a num- tures having 70% slag. At the curing ages of 28 and 90 days,
ber of studied alkali-activated natural pozzolan/slag mortars had the optimum binder combination and sodium hydroxide
similar total pore values to those of portland cement mortars. concentration were interdependent. The optimum NaOH
The different trends observed in absorption and chloride pene- concentration increased with an increase in slag portion of
tration and migration tests were explained by Ismail et al. [31] who binder. Similarly, the optimum slag dosage increased with
attributed these observations to the poor reliability of the absorp- increases in sodium hydroxide concentration.
tion and sorptivity tests for alkali-activated materials when the 5) The studied alkali-activated mortars have shown superior
specimens are severely pre-dried before testing. Similar to the transport properties in comparison to portland cement mor-
results of this study, Ismail et al. [31] observed that while the tars. While absorption of alkali-activated natural pozzolan/
migration coefficients (chloride penetration) of portland cement slag mortars were slightly higher than that of portland
concrete were 10 times higher than those of alkali activated slag/ cement mortars, their chloride penetration depths were sig-
fly ash concretes, their absorption and sorptivity results suggested nificantly lower than those of control mortars. The absorp-
that portland cement concrete specimens had a lower water tion of alkali-activated natural pozzolan/slag mortars
uptake. reduced as S/B reduced and sodium hydroxide molarity
The ultimate drying shrinkages of portland cement mortars are increased. A 50/50 combination of natural pozzolan and slag
shown in Table 7. It can be seen that the alkali-activated natural was the optimum combination for absorption results. For
pozzolan/slag mortars shrank considerably more than those of majority of the studied mortars, the absorption didn’t
portland cement mortars. The minimum, average and maximum change noticeably by varying sodium silicate content in
shrinkage observed in different AANPSL mortars were almost 1.7, the range of 20 to 30%.
3 and 4 times of those of portland cement mortars, respectively. 6) The chloride penetration depth reduced with decreases in
The drying shrinkage could become a major concern in need of a natural pozzolan portion of binder, sodium silicate dosage,
viable solution. It should, however, be noted that (1) there were NaOH concentration, and solution-to-binder ratio. A similar
some AANPSL mortars performing closer in drying shrinkage to trend was seen for the drying shrinkage, as reduction of
portland cement mortars, and (2) as coarse aggregates serves to
these variables also decreased the drying shrinkage. The dry- [26] S. Puligilla, P. Mondal, Role of slag in microstructural development and
hardening of fly ash-slag geopolymer, Cem. Concr. Res. 43 (2013) 70–80.
ing shrinkage of the studied alkali-activated natural poz-
[27] N.K. Lee, H.K. Lee, Setting and mechanical properties of alkali-activated fly ash/
zolan/slag mortars were significantly higher than those of slag concrete manufactured at room temperature, Constr. Build. Mater. 47
control portland cement mortars, a matter in need of further (2013) 1201–1209.
investigation. [28] A. Kar, U.B. Halabe, I. Ray, A. Unnikrishnan, Nondestructive characterizations of
alkali activated fly ash and/or slag concrete, Eur. Sci. J. 9 (24) (2013) 52–74.
[29] M. Chi, R. Huang, Binding mechanism and properties of alkali-activated fly
ash/slag mortars, Constr. Build. Mater. 40 (2013) 291–298.
References [30] I. Ismail, S.A. Bernal, J.L. Provis, S. Hamdan, J.S.J. Van Deventer, Microstructural
changes in alkali activated fly ash/slag geopolymers with sulfate exposure,
Mater. Struct. 46 (2013) 361–373.
[1] A. Palomo, M.W. Grutzeck, M.T. Blanco, Alkali-activated fly ashes a cement for
[31] I. Ismail, S.A. Bernal, J.L. Provis, R. San Nicolas, D.G. Brice, A.R. Kilcullen, S.
the future, Cem. Concr. Res. 29 (1999) 1323–1329.
Hamdan, J.S.J. Van Deventer, Influence of fly ash on the water and chloride
[2] F. Pacheco-Torgal, J. Castro-Gomes, S. Jalali, Alkali-activated binders: a review:
permeability of alkali-activated slag mortars and concretes, Constr. Build.
part 1. Historical background, terminology, reaction mechanisms and
Mater. 48 (2013) 1187–1201.
hydration products, Constr. Build. Mater. 22 (7) (2008) 1305–1314.
[32] A.M. Rashad, Properties of alkali-activated fly ash concrete blended with slag,
[3] C. Shi, A. Fernandez-Jimenez, A. Palomo, New cements for the 21st century: the
Iranian J. Mater. Sci. Eng. 10 (1) (2013) 57–64.
pursuit of an alternative to Portland cement, Cem. Concr. Res. 41 (2011) 750–
[33] I. Ismail, S.A. Bernal, J.L. Provis, R. San Nicolas, S. Hamdan, J.S.J. Van Deventer,
763.
Modification of phase evolution in alkali-activated blast furnace slag by the
[4] J. Davidovits, D.C. Comrie, J.H. Paterson, D.J. Ritcey, Geopolymeric concretes for
incorporation of fly ash, Cem. Concr. Compos. 45 (2014) 125–135.
environmental protection, ACI Concr. Int. 12 (1990) 30–40.
[34] J.G. Jang, N.K. Lee, H.K. Lee, Fresh and hardened properties of alkali-activated
[5] P. Duxson, J. Provis, G. Luckey, J.S.J. Van Deventer, The role of inorganic
fly ash/slag pastes with superplasticizers, Constr. Build. Mater. 50 (2014) 169–
polymer technology in the development of ‘‘Green Concrete’’, Cem. Concr. Res.
176.
37 (2007) 1590–1597.
[35] N. Ghafoori, M. Najimi, Properties of alkali-activated fly ash/slag repair
[6] B. McLellan, R. Williams, J. Lay, A. Van Riessen, G. Corder, Costs and carbon
mortars, Concrete Solutions, CRC Press, Belfast, United Kingdom, 2014, p. 77.
emissions for geopolymer pastes in comparison to ordinary Portland cement, J.
[36] N.K. Lee, J.G. Jang, H.K. Lee, Shrinkage characteristics of alkali-activated fly ash/
Cleaner Prod. 19 (2011) 1080–1090.
slag paste and mortar at early ages, Cem. Concr. Compos. 53 (2014) 239–248.
[7] P. Duxson, J.S.J Van Deventer, in: J. Provis, J.S.J Van Deventer, editor.
[37] T. Yang, X. Yao, Z. Zhang, Quantification of chloride diffusion in fly ash-slag-
Geopolymers, structure, processing, properties and applications. Cambridge,
based geopolymers by X-ray fluorescence (XRF), Constr. Build. Mater. 69
UK: Woodhead Publishing Limited Abington Hall, 2009 [ISBN -13: 978 1 84569
(2014) 109–115.
449 4].
[38] X. Gao, Q.L. Yu, H.J.H. Brouwers, Characterization of alkali activated slag–fly
[8] C. Shi, A. Fernandez-Jimenez, Stabilization/solidification of hazardous and
ash blends containing nano-silica, Constr. Build. Mater. 98 (2015) 397–406.
radioactive wastes with alkali-activated cements, J. Hazard. Mater. 137 (3)
[39] X. Gao, Q.L. Yu, H.J.H. Brouwers, Properties of alkali activated slag–fly ash
(2006) 1656–1663.
blends with limestone addition, Cem. Concr. Compos. 59 (2015) 119–128.
[9] J.L. Provis, Immobilization of toxic waste in geopolymers, in: Geopolymers:
[40] X. Gao, Q.L. Yu, H.J.H. Brouwers, Reaction kinetics, gel character and strength of
structure, processing, properties and industrial applications, eds. J.L. Provis,
ambient temperature cured alkali activated slag–fly ash blends, Constr. Build.
and J.S.J. Van Deventer, Woodhead Publishing, Cambridge, 2009, pp. 421–440.
Mater. 80 (2015) 105–115.
[10] F. Pacheco-Torgal, J.A. Labrincha, C. Leonelli, A. Palomo, P. Chindaprasirt,
[41] M. Chi, Y. Liu, R. Huang, Mechanical and microstructural characterization of
Handbook of alkali-activated cements, mortars and concretes, Woodhead
alkali-activated materials based on fly ash and slag, IACSIT Int. J. Eng. Technol.
publishing, 2015.
7 (1) (2015) 59–64.
[11] J.L. Provis, J.S.J. Van Deventer, Alkali activated materials: state-of-the-art
[42] N. Marjanovića, M. Komljenovića, Z. Baščarevića, V. Nikolića, Comparison of
report, RILEM TC 224-AAM, 2014.
two alkali-activated systems: mechanically activated fly ash and fly ash-blast
[12] M. Najimi, Alkali-activated natural pozzolan/slag binder for sustainable
furnace slag blends, Procedia Eng. 108 (2015) 231–238.
concrete, PhD dissertation, University of Nevada Las Vegas, December 2016,
[43] N. Marjanovića, M. Komljenovića, Z. Baščarevića, V. Nikolića, R. Petrović,
p. 431.
Physical–mechanical and microstructural properties of alkali-activated fly
[13] A. Allahverdi, E. Najafi Kani, M. Yazdanipour, Effects of blast-furnace slag on
ash–blast furnace slag blends, Ceram. Int. 41 (2015) 1421–1435.
natural pozzolan-based geopolymer cement, Ceramics-Silikáty 55 (1) (2011)
[44] P.S. Deb, P. Nath, P.K. Sarker, Drying shrinkage of slag blended fly ash
68–78.
geopolymer concrete cured at room temperature, Procedia Eng. 125 (2015)
[14] R.A. Robayo, R.M. de Gutiérrez, M. Gordillo, Natural pozzolan-and granulated
594–600.
blast furnace slag-based binary geopolymers, Materiales de Construcción, 66
[45] W.C. Wang, H.Y. Wang, M.H. Lo, The fresh and engineering properties of alkali-
(321) (2016) 8.
activated slag as a function of fly ash replacement and alkali concentration,
[15] M. Jafari Nadoushan, A.A. Ramezanianpour, The effect of type and
Constr. Build. Mater. 84 (2015) 224–229.
concentration of activators on flowability and compressive strength of
[46] A. Wardhono, D.W. Law, A. Strano, The strength of alkali-activated slag/fly ash
natural pozzolan and slag-based geopolymers, Constr. Build. Mater. 111
mortar blends at ambient temperature, Procedia Eng. 125 (2015) 650–656.
(2016) 337–347.
[47] N.K. Lee, H.K. Lee, Reactivity and reaction products of alkali-activated, fly ash/
[16] F. Puertas, S. Martinez-Ramirez, S. Alonso, E. Vazquez, Alkali-activated fly ash/
slag paste, Constr. Build. Mater. 81 (2015) 303–312.
slag cement: strength behavior and hydration products, Cem. Concr. Res. 30
[48] A.F. Abdalqader, F. Jin, A. Al-Tabbaa, Development of greener alkali-activated
(2000) 1625–1632.
cement: utilisation of sodium carbonate for activating slag and fly ash
[17] F. Puertas, R. de Gutierrez, A. Fernandez-Jimenez, S. Delvasto, J. Maldonado,
mixtures, J. Cleaner Prod. 113 (2016) 66–75.
Alkaline cement mortars: chemical resistance to sulfate and seawater attack,
[49] ASTM C618, Standard Specification for Coal Fly Ash and Raw or Calcined
Materiales de Construcción 52 (267) (2002) 55–71.
Natural Pozzolan for Use in Concrete, ASTM International, West
[18] F. Puertas, A. Fernandez-Jimenez, Mineralogical and microstructural
Conshohocken, PA, 2015.
characterization of alkali-activated fly ash/slag pastes, Cem. Concr. Compos.
[50] ASTM C989, Standard Specification for Slag Cement for Use in Concrete and
25 (2003) 287–292.
Mortars, ASTM International, West Conshohocken, PA, 2016.
[19] F. Puertas, T. Amat, A. Fernandez-Jimenez, T. Vazquez, Mechanical and durable
[51] ASTM C33, Standard Specification for Concrete Aggregates, ASTM
behavior of alkaline cement mortars reinforced with polypropylene fibres,
International, West Conshohocken, PA, 2016.
Cem. Concr. Res. 33 (12) (2003) 2031–2036.
[52] ASTM C109, Standard Test Method for Compressive Strength of Hydraulic
[20] J.I. Escalante Garcia, K. Campos-Venegas, A. Gorokhovsky, A. Fernandez,
Cement Mortars, ASTM International, West Conshohocken, PA, 2016.
Cementitious composites of pulverised fuel ash and blast furnace slag
[53] ASTM C1437, Standard Test Method for Flow of Hydraulic Cement Mortar,
activated by sodium silicate: effect of Na2O concentration and modulus, Adv.
ASTM International, West Conshohocken, PA, 2015.
Appl. Ceram. 105 (4) (2006) 201–208.
[54] ASTM C807, Standard Test Method for Time of Setting of Hydraulic Cement
[21] F.Q. Zhao, W. Ni, H.J. Wang, H.J. Liu, Activated fly ash/slag blended cement,
Mortar by Modified Vicat Needle, ASTM International, West Conshohocken,
Resour. Conserv. Recycl. 52 (2007) 303–313.
PA, 2013.
[22] S. Kumar, R. Kumar, S. Mehrotra, Influence of granulated blast furnace slag on
[55] ASTM C596, Standard Test Method for Drying Shrinkage of Mortar Containing
the reaction, structure and properties of fly ash based geopolymer, J. Mater.
Hydraulic Cement, ASTM International, West Conshohocken, PA, 2009.
Sci. 45 (3) (2010) 607–615.
[56] ASTM C642, Standard Test Method for Density, Absorption, and Voids in
[23] M. Izquierdo, X. Querol, C. Phillipart, D. Antenucci, M. Towler, The role of open
Hardened Concrete, ASTM International, West Conshohocken, PA, 2013.
and closed curing conditions on the leaching properties of fly ash-slag-based
[57] ASTM C1202, Standard Test Method for Electrical Indication of Concrete’s
geopolymers, J. Hazard. Mater. 176 (1–3) (2010) 623–628.
Ability to Resist Chloride Ion Penetration, ASTM International, West
[24] X. Luo, J. Xu, E. Bai, W. Li, Systematic study on the basic characteristics of
Conshohocken, PA, 2010.
alkali-activated slag-fly ash cementitious material system, Constr. Build.
[58] NT BUILD 492, Concrete, Mortar and Cement-Based Repair Materials: Chloride
Mater. 29 (2012) 482–486.
Migration Coefficient from Non-Steady-State Migration Experiments,
[25] J. Wang, X.L. Wu, J.X. Wang, C.Z. Liu, Y.M. Lai, Z.K. Hong, J.P. Zheng,
NORDTEST, Espoo, Finland, 1999.
Hydrothermal synthesis and characterization of alkali-activated slag-fly ash-
[59] C. Jolicoeur, M.A. Simard, J. Sharman, R. Zamojska, M. Dupuis, N. Spiratos, E.
metakaolin cementitious materials, Microporous Mesoporous Mater. 155
Douglas, V.M. Malhotra, Chemical activation of blast-furnace slag, An overview
(2012) 186–191.
and systematic experimental investigation, in: Advances in concrete [72] C.D. Atis, C. Bilim, O. Celik, O. Karahan, Influence of activator on the strength
technology, ministry of supply and services, ed. V.M. Malhotra, Ottawa, and drying shrinkage of alkali-activated slag mortar, Constr. Build. Mater. 23
Canada, 1992, pp. 471–502. (1) (2009) 548–555.
[60] N. Smaoui, M.A. Bérubé, B. Fournier, B. Bissonnette, B. Durand, Effects of alkali [73] C. Shi, Y. Li, Investigation on some factors affecting the characteristics of alkali
addition on the mechanical properties and durability of concrete, Cem. Concr. phosphorus slag cement, Cem. Concr. Res. 19 (4) (1989) 527–533.
Res. 35 (2005) 203–212. [74] N. Ghafoori, K. Sierra, M. Najimi, M.R. Sharbaf, Properties of alkali-activated fly
[61] Z. Li, K. Afshinnia, P.R. Rangaraju, Effect of alkali content of cement on ash mortars made with multiple activators, in: Second International
properties of high performance cementitious mortar, Constr. Build. Mater. 102 Conference on Concrete Sustainability, Madrid, Spain, 2016, p. 1330.
(2016) 631–639. [75] S. Aydin, B. Baradan, Effect of activator type and content on properties of
[62] N. Ghafoori, M. Najimi, B. Radke, Natural Pozzolan- based geopolymers for alkali-activated slag mortars, Compos. Part B 57 (2014) 166–172.
sustainable construction, Environ. Earth Sci. 75 (1110) (2016) 16, https://doi. [76] D. Ravikumar, Property development, microstructure, and performance of
org/10.1007/s12665-016-5898-5. alkali activated fly ash and slag systems, Ph.D. Dissertation, Clarkson
[63] C.K. Yip, The role of calcium in geopolymerisation Ph.D. Thesis, University of University, 2012, p. 217.
Melbourne, Australia, 2004. [77] F.A. Memon, M.F. Nuruddin, S. Khan, N. Shafiq, T. Ayub, Effect of sodium
[64] R. Yuan, Q. Gao, S. Ouyang, Study on structure and latent hydraulic activity of hydroxide concentration on fresh properties and compressive strength of self-
and its activation mechanism, J. Wuhan Univ. Technol. (in Chinese)P. R. China compacting geopolymer concrete, J. Eng. Sci. Technol. 8 (1) (2013) 44–56.
3 (1987) 297–303. [78] S. Alonso, A. Palomo, Alkaline activation of metakaolin and calcium hydroxide
[65] A. Roy, P.J. Schilling, H.C. Eaton, P.G. Malone, W.N. Brabston, L.D. Wakeley, mixtures: influence of temperature, activator concentration and solids ratio,
Activation of ground-blast furnace slag by alkali-metal and alkaline-earth Mater. Lett. 47 (2001) 55–62.
hydroxides, J. Am. Ceram. Soc. 75 (12) (1992) 3233–3240. [79] A.A. Melo Neto, M.A. Cincotto, W. Repette, Drying and autogenous shrinkage of
[66] S. Song, D. Sohn, H.M. Jennings, T.O. Mason, Hydration of alkali-activated pastes and mortars with activated slag cement, Cem. Concr. Res. 38 (4) (2008)
ground granulated blast furnace slag, J. Mater. Sci. 35 (2000) 249–257. 565–574.
[67] A. Fernandez-Jimenez, Cementos de escorias activadas alcalinamante: [80] E. Douglas, A. Bilodeau, V.M. Malhotra, Properties and durability of alkali-
influencia de las variables y modelizacion del proceso, Ph.D. Thesis, activated slag concrete, ACI Mater. J. 89 (5) (1992) 509–516.
Universidad Autonoma de Madrid, 2000. [81] M.O. Yusuf, M.A.M. Johari, Z.A. Ahmad, M. Maslehuddin, Shrinkage and
[68] C. Shi, R.L. Day, Factors affecting early hydration characteristics of alkali-slag strength of alkaline activated ground steel slag/ultrafine palm oil fuel ash
cements, Cem. Concr. Res. 26 (3) (1996) 439–448. pastes and mortars, Mater. Des. 63 (2014) 710–718.
[69] D. Bondar, C.J. Lynsdale, N.B. Milestone, N. Hassani, A.A. Ramezanianpour, [82] C.K. Yip, G.C. Lukey, J.L. Provis, J.S.J. Van Deventer, Effect of calcium silicate
Effect of type, form, and dosage of activators on strength of alkali-activated sources on geopolymerisation, Cem. Concr. Res. 38 (2008) 554–564.
natural pozzolans, Cem. Concr. Compos. 33 (2011) 251–260. [83] M.A. Cincotto, A.A. Melo, W.L. Repette, Effect of different activators type and
[70] R.R. Lloyd, J.L. Provis, J.S.J. Van Deventer, Microscopy and microanalysis of dosages and relation with autogenous shrinkage of activated blast furnace slag
inorganic polymer cements. 1: remnant fly ash particles, J. Mater. Sci. 44 (2) cement, in: Proceedings of the 11th International Congress on the Chemistry of
(2009) 608–619. Cement, eds. G. Grieve & G. Owens, Durban, South African, 2003, p. 1878.
[71] M.N. Qureshi, S. Ghosh, Effect of silicate content on the properties of alkali- [84] M. Chi, J. Chang, R. Huang, Strength and drying shrinkage of alkali-activated
activated blast furnace slag paste, Arabian J. Sci. Eng. 39 (2014) 5905–5916. slag paste and mortar, Adv. Civ. Eng. (2012) Article ID 579732, 7.

Construction and Building Materials: Meysam Najimi, Nader Ghafoori, Mohammadreza Sharbaf

Uploaded by

Copyright:

Available Formats

Construction and Building Materials: Meysam Najimi, Nader Ghafoori, Mohammadreza Sharbaf

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Construction and Building Materials: Meysam Najimi, Nader Ghafoori, Mohammadreza Sharbaf

Uploaded by

Copyright:

Available Formats

Construction and Building Materials 164 (2018) 625–643

Contents lists available at ScienceDirect

Construction and Building Materials

Alkali-activated natural pozzolan/slag mortars: A parametric study

 Various properties of alkali-activated natural pozzolan/slag mortars are provided.

Concrete, as a foundational component to infrastructure in any

4.1. Materials 4.2. Mixture proportions

Year References Mixture type Studied properties

Requirements Natural pozzolan Slag

Mix ID. Mixture description Materials, g

*Fine aggregate was 10,556 g for all the mixtures.

Mix ID. Fine aggregate-to-cement ratio Water-to-cement ratio Materials, g

tively. The observed behavior was similar to the observations of

5.1.4. Effects of alkaline activator solution-to-binder ratio

ing sodium silicate contents of 20, 25 and 30%, the peak

5.2.1. Effects of binder combination 5.2.2. Effects of activator combination

Mix ID. Compressive strength (MPa) Ultimate drying shrinkage (%)

5.4.2. Effects of activator combination

5.4.3. Effects of alkaline activator solution-to-binder ratio

of S/B improved absorption properties. On average, each 0.04

5.4.4. Effects of sodium hydroxide concentration

5.6.3. Effects of alkaline activator solution-to-binder ratio

5.6.4. Effects of sodium hydroxide concentrations

subsequent increases in the drying shrinkage of the studied alkali-

5.7.3. Effects of alkaline activator solution-to-binder ratio

You might also like

Various properties of alkali-activated natural pozzolan/slag mortars are provided.