Construction and Building Materials: Chamila Gunasekara, David Law, Shamir Bhuiyan, Sujeeva Setunge, Liam Ward

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

Construction and Building Materials 200 (2019) 502–513

Contents lists available at ScienceDirect

Construction and Building Materials


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

Chloride induced corrosion in different fly ash based geopolymer


concretes
Chamila Gunasekara a,⇑, David Law a, Shamir Bhuiyan a, Sujeeva Setunge a, Liam Ward b
a
Civil & Infrastructure Engineering, School of Engineering, RMIT University, Australia
b
Environmental & Chemical Engineering, School of Engineering, RMIT University, Australia

h i g h l i g h t s

 Fly ash geopolymer shows higher corrosion rate than PC concrete with cast-in chlorides.
 Higher corrosion rate in cast-in specimens is attributed to a lower pH in geopolymer.
 Chloride diffusion rate is dependent on the reactivity of the precursor fly ash.
 Chloride binding capacity of fly ash geopolymer is dependent on CaO in precursor.
 Hematite, akageneite & lepidocrocite are main corrosion products of geopolymer.

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

Article history: This study investigates the corrosion of reinforcement in geopolymer concrete manufactured from three
Received 3 October 2018 different low calcium fly ashes from Australian power plants. The long term corrosion condition of
Received in revised form 21 December 2018 embedded rebar in fly ash geopolymer concretes containing cast in chlorides (0–5%) subjected to wet-
Accepted 22 December 2018
dry cycles, together with specimens exposed to ponding in 3% NaCl were examined. Half-cell potential
and linear polarisation resistance techniques were used to measure corrosion up to 540 days of age,
and compared with results of a similar binder content PC concrete. Increased levels of corrosion were
Keywords:
observed in the cast-in chloride geopolymer specimens compared with the equivalent PC concretes.
Fly ash
Geopolymer
However, in the case of the ponded specimens the reinforcement in the geopolymer concrete specimens
Corrosion rate displayed lower corrosion levels than the PC concrete. The higher corrosion rate in the cast-in specimens
Polarization is attributed to a lower pH in the geopolymer specimens resulting in a higher Cl/OH ratio. In the
Chloride concentration ponded specimens the formation of three-dimensional N-A-S-H and C-A-S-H cross linking in the gel
Chloride binding matrix reduces chloride diffusion to rebar depth, resulting in a lower corrosion rate being observed for
ponded geopolymer specimens compared to the PC concrete.
Ó 2018 Elsevier Ltd. All rights reserved.

1. Introduction fly ash based geopolymer concrete has received significant atten-
tion in the last two decades as an alternative to PC concrete for pro-
Portland Cement (PC) is the second most used product in the duction of structural/non-structural elements. The most
world after water, and also the second largest source of anthro- emphasized advantage of geopolymer production is the potential
pogenic carbon dioxide (CO2) emissions, after power generation. reduction of CO2 emission by 26–45% with the replacement of PC
Current cement production contributes 4–7% of the CO2 emission with no associated adverse economic impact [8–10].
worldwide [1–3], with the production of 1 tonne of cement pro- The durability of concrete is of great importance as it governs
ducing from 600 to 800 kg of CO2 [4–6]. In the worst case scenario the performance of reinforced concrete structures during their
rapid urban growth in developing countries may significantly operational life. Chloride induced corrosion in infrastructure is
increase cement demand with estimates that cement production one of the major durability concerns, causing billions of dollars
could represent 10–15% of global CO2 emissions in next decade to be spent globally in repairs and maintenance of reinforced con-
[7]. Due to reduced carbon emission and low embodied energy, crete structures [11]. Concrete provides an alkaline pH that pro-
motes the passivation of steel reinforcement. The ingress of
⇑ Corresponding author. chloride ions results in the depassivation of steel and, once the
E-mail address: chamila.gunasekara@rmit.edu.au (C. Gunasekara).
chloride ions have reached a critical concentration at the rebar,

https://doi.org/10.1016/j.conbuildmat.2018.12.168
0950-0618/Ó 2018 Elsevier Ltd. All rights reserved.
C. Gunasekara et al. / Construction and Building Materials 200 (2019) 502–513 503

the initiation of corrosion. The Cl/OH ratio has been identified as examining the performance of geopolymer concrete up to 540 days
one of the principal factors determining the onset of corrosion, using three different low calcium fly ashes with the same mixing
with the OH ion concentration being dependant on the pH [12]. process, provides a systematic long term durability study of a
Badar et al. [13] noted that due to the high pH of the pore system, range of geopolymer concretes. The reported geopolymer research
fly ash geopolymer concrete can passivate reinforcement similar to data can thus provide a clearer understanding of the behaviour of
PC concrete. However, Lloyd et al. [14] showed that the presence of this novel concrete against chloride ingress, made with a range of
CaO is important in lowering the permeability and inhibiting a low calcium fly ashes, representative of those that are available
drop in the pH due to leaching of alkali ions, which may result in worldwide.
depassivation of the reinforcement in geopolymer concrete.
It has been reported that fly ash geopolymer concrete has a high 3. Experimental procedure
resistance to chloride ingress due to low chloride ion permeability
[15,16]. However, Ganesan et al. [17] stated that chloride penetra- 3.1. Materials used
tion of both PC and fly ash geopolymer concretes are identical.
Babaee and Castel [18] also noted that low-calcium fly ash-based Low calcium, Class F fly ash for this study was provided from
GPC exhibit comparable electrochemical performance to similar Gladstone, Pt.Augusta and Collie power plants in Australia [26].
strength PCC during the propagation phase of corrosion. However, Tables 1 and 2 report the chemical composition and physical and
Olivia et al. [19] observed a higher chloride penetration than PC mineralogical properties of the three fly ashes. X-ray fluorescence
concrete and Pasupathy et al. [20] reported significantly higher and X-ray diffraction techniques were used to measure the chem-
chloride ingress in low Calcium fly ash based geopolymer concrete ical oxide composition and crystallinity of the material. Commer-
following 6 years exposure in a saline environment, while Munt- cially available liquid Na2SiO3 (Grade D, Na2O = 14.7% and
ingh [21] concluded that the precursor fly ash used and mix design SiO2 = 29.4% by mass) and 15 M liquid NaOH were used as alkaline
may result in varying long term performance in geopolymer con- activator. River sand was used as fine aggregate (2600 kg/m3 of
crete. Monticelli et al. [22] studied chloride induced corrosion in density) and crushed granite aggregate used as coarse aggregate
contaminated fly ash geopolymer concrete when adding 2% chlo- (2650 kg/m3 of density). The coarse aggregate was prepared in a
ride by mass of binder in the mixing process. Test results indicated saturated surface dry aggregates condition. The coarse aggregate
that silicate ions in the pore solution in geopolymer concrete pro- consists of 7 and 10 mm grain sizes. Demineralized water was used
vide extra protection for the passivation of the reinforcement. in the concrete production.
However, Miranda et al. [23] reported that embedded steel did
not have a steel shielding passive layer and showed an identical
3.2. Mix designs
level of corrosion to PC concrete. Bastidas et al. [24] worked with
three different fly ash based mortars and investigated the corrosion
The mix design used is an optimised concrete mix developed in
stability of the passive state with 0%, 0.2%, 0.4% and 2% chloride
previous research for each of the three fly ash [27], Table 3. The
additions in relation to the binder material. The authors reported
mix proportions were calculated based on the absolute volume
that the addition of 2% (by binder weight) chlorides increases the
method [28] and water to solid (w/s) ratio (Eq. (1)) is fixed at 0.37.
reinforcement corrosion rate by 100 times. Hence, the corrosion
resistance performance of steel in geopolymer concrete is not w mass of water in alakline activatorþmass of added water
ratio ¼
clearly understood. One of the reasons for the differences in s mass of solid in alkaline activatorþmass of fly ash
reported literature is due to the varied composition of fly ash from ð1Þ
different sources which affects the material properties of the
geopolymer concrete produced [25]. The question of whether rein-
forcing steel can be protected in different fly ash based geopolymer 3.3. Mixing, casting and curing
concretes, as effectively as with PC concrete, is examined in this
study. To that end, a range of geopolymer specimens were pre- A 120 L concrete mixer was used to mix the geopolymer and PC
pared using three different low calcium fly ashes containing a concrete. For the geopolymer concrete mix, the fly ash, sand and
range of cast in chloride contents, varying from 0% to 5%. The cor- coarse aggregates were mixed initially for 4 min. The activator,
rosion state of embedded rebar was evaluated using electrochem- which was pre-mixed, and water were added to the dry mix and
ical techniques, namely half-cell potential and corrosion rate mixed another 8 min. A similar mixing procedure was conducted
(linear polarisation resistance technique), and then compared with for PC concrete, except for the addition of the alkaline activator
results of a similar binder content PC concrete. The study further to the dry materials (i.e. PC, fine and coarse aggregates). For each
evaluates the long-term corrosion state of the embedded rebar in concrete type, a total of 7 cubes (150 mm3) were cast with ribbed
fly ash geopolymer and PC concretes while being subjected to bars (10 mm diameter and 120 mm length). For the ‘‘cast in chlo-
wet-dry cycles and chloride ponding, which was adopted to simu- ride” specimen, the percentages of admixed NaCl percentage are
late natural chloride induced corrosion. A detailed microstructural 0, 0.25, 0.5, 1, 2, 3 and 5% of NaCl by weight of binder. An additional
and chemical analysis was also conducted to determine the effects set of concrete cubes of 150 mm3 were cast with ribbed bars
of cast in chloride and chloride ponding corrosion rate, pH and the (10 mm diameter and 120 mm length) without chloride additions
condition of the rebar. The observed test results up to 540 days of for the ponding tests. A copper wire was welded to each bar prior
age are reported. to casting to enable the steel to be used as the working electrode.
The weld was sealed with epoxy to prevent the onset of unin-
tended corrosion.
2. Significance of research Specimens were placed on a vibration table and vibrated to
remove air bubbles. All geopolymer specimens were initially
To date, few studies have been conducted in examining the stored at room temperature for 1 day and then heat cured in a
long-term durability of 100% fly ash geopolymer concrete subject dry oven for 24 h at 80 °C temperature. After heat curing, the
to chloride induced reinforcement corrosion. This lack of knowl- geopolymer specimens were removed from the oven and left to
edge is one of the main limitations associated with the confident cool to room temperature before demoulding. Following casting
adoption of this concrete for industrial applications. This research, all PC concrete specimens were demoulded and kept at room
504 C. Gunasekara et al. / Construction and Building Materials 200 (2019) 502–513

Table 1
Chemical composition.

Fly ash By weight (%)


SiO2 Al2O3 Fe2O3 CaO K2O TiO2 P2O5 MgO Na2O SO3 MnO
Gladstone 47.87 28.0 14.09 3.81 1.81 1.99 0.93 0.62 0.27 0.21 0.41
Pt.Augusta 49.37 31.25 4.47 4.80 1.65 2.94 1.28 2.21 0.24 0.04 1.30
Collie 52.67 29.60 11.27 0.64 0.65 1.83 1.13 0.72 0 0.48 0.05
PC 22.5 4.5 0.4 66.3 0.15 0.20 0.67 0.51 0.17 2.8 0.10

Table 2 of the reference electrode housing, coated with a permeable foam


Physical and mineralogical properties. material (moistened prior to test) to ensure good contact with the
Properties investigated Gladstone Pt.Augusta Collie concrete. The corrosion current, Icorr, is then calculated [30]. Tafel
Specific Gravity 2.26 2.23 2.42
constants were not directly determined as this can cause irre-
BET Surface Area, (kg/m2) 2363 1228 1095 versible damage to the system due to the large applied overpoten-
Fineness (%) at 10 mm 43.1 46.7 40.9 tials. A value of 26 mV for active steel and 52 mV for passive steel
at 20 mm 61.9 62.1 54.6 were used for the Stern-Geary constant calculations, B based upon
at 45 mm 82.7 80.2 70.0
the values derived for PC concrete [30–32].
Amorphous (%) 71.8 59.5 72.5
Crystalline (%) Quartz 6.8 29.2 18.2 The suspension method was used to obtain the pH profile of the
Mullite 17.9 7.5 8.7 concrete [33]. Concrete powder samples were taken at rebar depth
Others 3.5 3.8 0.6 and mixed in de-ionised water at a 1:1 ratio. The suspensions were
Al coordination Tetrahedral (AlIV) 54.3 40.2 50.7 stirred using a magnetic stirrer for 6 min. A pH meter was used to
Octahedral (AlIV) 45.7 59.8 49.3
Unburnt carbon content (%) 0.42 0.53 0.61
obtain pH readings. In order to verify the pH values, the pore solu-
tion from a respective mortar sample (similar mix design as Table 3,
without coarse aggregate) was extracted and the pH measured. A
pressure of 500 MPa was applied with the loading rate of 1 MPa/
temperature. All chloride contaminated concrete specimens (PC &
s [34]. The measured pH difference was 0.68 (correction) and this
geopolymer) were kept in a spray cabinet which was set to spray
was subtracted from each data point along the profile in order to
water twice a day (every 12 h), with each spray cycle lasting
obtain the calibrated profile.
30 min. The non-chloride contaminated concrete specimens were
Powder samples taken from 540-day matured concrete speci-
stored under laboratory condition for 14 days (24 °C temperature
mens, close to the reinforcement bar, were used for the chloride
and 50% Relative Humidity), prior to ponding in 3% NaCl solution
analysis, Fourier transform infrared (FT-IR) and X-ray diffraction
for one week. After ponding, all specimens were subjected to a
(XRD) tests. Samples were sent to a NATA accredited laboratory
two-week dry cycle. This one-week ponding and two-week drying
to determine the total and free chloride content. FT-IR absorption
cycle process was continued up to 540 days.
spectrometry was used to identify bond formations within the
range 4000–450 cm1. X-ray diffraction (XRD) data was obtained
3.4. Testing using a Bruker AXS D4 Endeavor wide angle X-ray diffractometer
with copper anode at 40 kV and 35 mA.
Compressive strength testing was performed at 28 and 540 days Scanning electron microscopy in backscattered electron (BSE)
using a MTS machine with a loading rate of 20 MPa/min according imaging mode (BSE detector energy setting of 15 eV) was used to
to AS 1012.9 [29]. Four standard cylindrical specimens were tested examine the microstructure of the geopolymer concretes. Speci-
for each data point and the mean value reported. The corrosion men with dimensions 10 mm diameter and 5 mm thickness, were
rates were measured using linear polarisation resistance (LPR), carbon coated prior to insertion in the microscope chamber. A mer-
utilising a 3-electrode arrangement [30]. LPR measurements were cury intrusion porosimeter (V1.09 of Auto pore IV 9500) was used
taken potentiostatically by applying a  10/+10 mv shift, DE, to to measure pore-structure development in the geopolymer con-
the steel (working electrode) using an ACM Field machine. This crete. Geopolymer monoliths were initially dried in an oven at
result is a current, DI, passing between the working electrode 105 °C for 48 h and then vacuum saturated before testing.
and auxiliary electrode, which is measured after a 30 s delay time,
tequilib, to allow equilibrium to be established. The solution resis-
tance, Rs, is automatically measured by applying a high frequency 4. Experimental results
300 Hz input signal between the auxiliary electrode and reinforc-
ing bar and measuring the resulting current. The polarisation resis- Fig. 1 shows the compressive strength development for the dif-
tance, Rp is determined from the potential to current density slope ferent low calcium fly ash based geopolymer and PC concretes after
obtained. A Cu/CuSO4 electrode (CSE) was used as the reference 28 and 540 days with 0%, 2% and 5% NaCl. The Gladstone geopoly-
electrode and the auxiliary electrode was a brass ring on the base mer achieved the highest strength, with the Pt August and PC

Table 3
Optimum mix design (kg/m3).
*
Concrete Fly ash (kg) Cement (kg) Aggregates (kg) Activator (kg) Added Water (kg) Water/solid ratio
Sand 7 mm 10 mm Na2SiO3 (Liquid) NaOH (15 M)
Gladstone 417 – 698 308 619 293 66 7 0.37
Pt.Augusta 417 – 698 308 619 293 66 7 0.37
Collie 420 – 706 312 624 241 92 15 0.37
PC – 410 615 550 550 – – 220 –
*
Water/Cement ratio of 0.54 is used for PC concrete.
C. Gunasekara et al. / Construction and Building Materials 200 (2019) 502–513 505

concretes having a similar strength, whilst the Collie was signifi- geopolymers, the only mix that achieved this condition was the
cantly lower (approx. 50% less than the PC concrete). All mixes control mix, i.e. concrete with no chloride.
increased by a similar percentage from 28 to 540 days. The addi- It is noted that the larger negative readings are not indicative of
tion of NaCl had no observable effect on the 28 day strength of increased susceptibility to corrosion due to many reasons. These
any of the mixes. In contrast, all geopolymer and PC concretes with can include factors such as a lower pH in the geopolymer than
‘‘cast-in chloride” recorded a loss in strength between 28 and the PC concrete, binding of the chloride ions or oxygen reduction
540 days. The loss in strength increased with increasing chloride at the reinforcement area due to oxygen depletion. Oxygen deple-
concentration. Strength reductions were of similar magnitude, tion may occur due to (i) saturation of the concrete associated with
13.2–17.7% for the 2% chloride and 19.8–23.2% for the 5% chloride the wet/drying cycle or (ii) decreased permeability of the concrete
specimens. The Gladstone again yielded the highest strength while due to the fineness of the fly ash. Such factors could limit oxygen
the Collie provided the lowest strength. It is well established that diffusion to cathodic areas on the rebar and result in more negative
additions of CaCl2 can increase initial strength development in PC corrosion potential values [24,36,37].
concrete, being a common set accelerator through the 1960’s and The alkalinity of concrete the pore solution is critical for pro-
1970’s, but that it can have a negative impact on the long term tecting the reinforcement against chloride induced corrosion.
strength gain due to the formation of chloroaluminate hydrates. Hence, the pH profile of each concrete was measured using the sus-
These results show a similar long term reduction in strength for pension method [33] and calibrated by direct measurement of the
geopolymer concretes with cast in chloride. extracted pore solution. The calibrated pH profiles for geopolymer
Corrosion potential (Ecorr) measurements of the rebars in the and PC concretes at different chloride concentrations are shown in
geopolymer and PC concrete specimens are shown in Fig. 2. As Fig. 4. The pH of Gladstone, Pt.Augusta and Collie fly ash based
expected, the potential became more negative as the chloride per- geopolymer concretes varied between 11.58 and 10.72, 11.62–
centage increased in the concrete specimen [35]. The PC concrete 10.88 and 11.58–10.95, respectively, compared to a pH variation
potential was generally more positive than for the geopolymer of 12.48–11.67 for PC concrete. It was noted that all geopolymer
concretes. Although there was no clear pattern within the geopoly- concretes containing no chloride displayed significantly lower pH
mer concretes themselves, the Gladstone samples were the most values (11.6) than PC concrete (12.5) at 28 days, and this was
negative, especially at NaCl concentrations greater than 1%. Based further reduced to  11.3 at 540 days, compared to the pH
on the van der Veer criteria, the 2, 3 and 5% data all showed a of  11.9 for PC concrete. All three geopolymer concretes showed
90% chance of corrosion for all three geopolymer concretes at all a pH drop between 28 and 540 days with increasing chloride con-
ages, while only the 5% PC concrete had a potential consistently centration, in particular in the 2–5% sodium chloride concentra-
more negative than  350 mV CSE. It was noted that PC concrete tions. Gladstone showed the largest pH reduction, followed by Pt.
mixes containing less than 1% of sodium chloride showed corro- Augusta and Collie. The PC concrete followed a similar trend to
sion potentials more positive than  200 mV CSE throughout the the geopolymer concrete at the higher chloride concentrations,
entire testing period. However, from all of the three fly ash but showed a lower pH reduction than the fly ash geopolymer. It

60 60
(a) 28-day 540-day (b) 28-day 540-day
+14.1%a 50
Compressive Strength (MPa)

50
Compressive Strength (MPa)

(+6.7MPa) b
-19.8%
-17.7%
(-9.5MPa) 40
40 (-8.0MPa)
+20.3%
(+6.4MPa) -16.9%
30 30 -23.0%
54.3 (-5.4MPa) (-7.3MPa)
47.6 45.3 48.1
20 38.6 20 37.9
37.3
31.5 31.9 31.7
26.5 24.4
10 10

0 0
Gladstone 0 Gladstone 2 Gladstone 5 Pt.Augusta 0 Pt.Augusta 2 Pt.Augusta 5
60 60
(c) 28-day 540-day (d) 28-day 540-day
50
Compressive Strength (MPa)

50
Compressive Strength (MPa)

40 +20.2%
40
(+7.4MPa)
-13.2%
-19.8%
(-4.7MPa)
30 30 (-7.0MPa)

+16.4% 44.1
20 -23.2% 20
(+3.0MPa) -15.7% 36.7 35.5 35.4
(-2.8MPa) (-4.2MPa) 30.8 28.4
10 21.3 10
18.3 17.8 18.1
15.0 13.9

0 0
Collie 0 Collie 2 Collie 5 PC 0 PC 2 PC 5

a
Fig. 1. Compressive strength development of geopolymer and PC concretes, contained with different percentages of chlorides. Percentage of strength gain/loss
[(540 day  28 day)/28 day]. bAbsolute value of strength gain/loss (MPa).
506 C. Gunasekara et al. / Construction and Building Materials 200 (2019) 502–513

Duraon (Days) Duraon (Days)


0 100 200 300 400 500 600 0 100 200 300 400 500 600
0 0

-100 -100

-200 -200

Ecorr (mV)
Ecorr (mV)

-300 -300

-400 -400

-500 -500

High Corrosion probability Zone High Corrosion probability Zone


-600 -600
Gladstone 0 Gladstone 0.25 Gladstone 0.5 Gladstone 1 Pt.Augusta 0 Pt.Augusta 0.25 Pt.Augusta 0.5 Pt.Augusta 1
Gladstone 2 Gladstone 3 Gladstone 5 Pt.Augusta 2 Pt.Augusta 3 Pt.Augusta 5

Duraon (Days) Duraon (Days)


0 100 200 300 400 500 600 0 100 200 300 400 500 600
0 0

-100 -100

-200 -200
Ecorr (mV)
Ecorr (mV)

-300 -300

-400 -400

-500 -500

High Corrosion probability Zone High Corrosion probability Zone


-600 -600
Collie 0 Collie 0.25 Collie 0.5 Collie 1
PC 0 PC 0.25 PC 0.5 PC 1 PC 2 PC 3 PC 5
Collie 2 Collie 3 Collie 5

Fig. 2. Corrosion potential vs. Cl concentration in geopolymer and PC concretes.

is hypothesized that the reduction in pH can be attributed to a the corrosion potential readings. The Gladstone concrete consis-
number of factors. Firstly, continuing geopolymerisation results tently displayed the highest Icorr for 5% sodium chloride concentra-
in the consumption of hydroxyl ions which results in a lowering tions compared with the other concretes. The Pt Augusta generally
of the pH. Secondly, the pH will be reduced by the formation of had a higher Icorr than the Collie for all chloride concentrations
acid within in the pits due to corrosion of the steel. Finally there except 1%, which was considerably lower than Collie geopolymer
may be some leaching of excess activator during the wet/dry concrete. The Icorr values of the fly ash geopolymer concretes were
cycles. Further it is hypothesized that the more negative corrosion significantly greater than the PC concrete, apart from the low chlo-
potential values obtained by fly ash geopolymer concretes can be ride concentration samples (0–0.5%) where a minimal difference
attributed to the lower pH in the geopolymer concretes compared was observed. The Icorr values for fly ash geopolymer concrete spec-
to the PC concrete. imens ranged from 0.5 to 3.5 lA/cm2, whereas values of 0.5–
The corrosion potential (Ecorr) should be considered more as a 2.0 lA/cm2 were recorded for PC concrete. Based on Icorr values,
qualitative parameter and cannot be used to determine corrosion two distinct groupings can be identified for fly ash geopolymer
rates alone, as there is no direct correlation between Ecorr and Icorr and PC concretes in regard to chloride concentration. After 90 days,
(corrosion current density) [23,38,39]. Thus, corrosion current den- all the 0% and 0.25% chloride samples (except 0.25% Pt Augusta)
sity (Icorr) was measured to provide an indication of the corrosion displayed corrosion rates in the moderate range. All the other spec-
rate. The Icorr for the rebars in the four concretes is shown in imens, i.e. 0.5% chloride and above, had corrosion rates classified in
Fig. 3. As a guide the Icorr values are quantified as low, medium the high range [30]. However, in the geopolymer specimens there
or high as defined elsewhere for PC concrete [30]. Fluctuations in are two distinct groupings evident between the 0 to 0.5% chloride
Icorr were observed over the first 90 days, with most samples show- specimens and those with a chloride concentration above 0.5%.
ing a decrease in corrosion rate, although a small number showed Given that the control specimens with no chloride present display
an increase. This is attributed to the activation and hydration reac- moderate corrosion rates, this would suggest that the calculated
tions and the passivation of the rebars, during this period. These corrosion rates are overestimated.
are dependent upon the materials (FA or PC) and the chloride con- As stated, B values of 26 mV for active and 52 mV for passive
centration. However, beyond 90 days almost all the specimens dis- steel were adopted based upon the values derived for PC concrete.
played an overall increase in corrosion rates with time, though However, Babaee and Castel have reported values of 13 to 20 mV
some fluctuation was observed. for passive steel and 43 to 65 mV for active steel in geopolymer
All concretes with 0% chlorides had a corrosion rate of less than concrete [18] attributing this to the formation of iron oxide layers
1 lA/cm2 during the entire test period, which correlates well with around the rebar. Adopting these values for the geopolymer
C. Gunasekara et al. / Construction and Building Materials 200 (2019) 502–513 507

Gladstone 0 Gladstone 0.25 Gladstone 0.5 Gladstone 1 Pt.Augusta 0 Pt.Augusta 0.25 Pt.Augusta 0.5
Gladstone 2 Gladstone 3 Gladstone 5 Pt.Augusta 1 Pt.Augusta 2 Pt.Augusta 3
3.5 3.5
(a) (b)
3.0 3.0
Icorr (μA/cm2)

Icorr (μA/cm2)
2.5 2.5

2.0 2.0

1.5 1.5

High High
1.0 1.0

Moderate Moderate
0.5 0.5
Low Low
0.0 0.0
0 100 200 300 400 500 600 0 100 200 300 400 500 600
Duraon (Days) Duraon (Days)

Collie 0 Collie 0.25 Collie 0.5 Collie 1


PC 0 PC 0.25 PC 0.5 PC 1 PC 2 PC 3 PC 5
Collie 2 Collie 3 Collie 5
3.5 3.5
(c) (d)
3.0 3.0

Icorr (μA/cm2)
Icorr (μA/cm2)

2.5 2.5

2.0 2.0

1.5 1.5

High High
1.0 1.0

Moderate Moderate
0.5 0.5
Low Low
0.0 0.0
0 100 200 300 400 500 600 0 100 200 300 400 500 600
Duraon (Days) Duraon (Days)

Fig. 3. Corrosion current density (Icorr) vs. Cl concentration in geopolymer and PC concretes.

12.6 12.6
Gladstone 0 Gladstone 1 Gladstone 2 (a) Pt.Augusta 0 Pt.Augusta 1 Pt.Augusta 2 (b)
12.4 12.4
Gladstone 3 Gladstone 5 Pt.Augusta 3 Pt.Augusta 5
12.2 12.2

12.0 12.0
pH

pH

11.8 11.8

11.6 11.6

11.4 11.4

11.2 11.2

11.0 11.0
pH range: 11.58 10.72 pH range: 11.62 10.88
10.8 10.8

10.6 10.6
0 100 200 300 400 500 600 0 100 200 300 400 500 600
Duraon (Days) Duraon (Days)
12.6 12.6
Collie 0 Collie 1 Collie 2 (c) (d)
12.4 12.4
Collie 3 Collie 5
12.2 12.2

12.0 12.0
pH

pH

11.8 11.8
pH range: 12.48 11.67
11.6 11.6

11.4 11.4

11.2 11.2

11.0 11.0
pH range: 11.58 10.95 PC 0 PC 1 PC 2 PC 3 PC 5
10.8 10.8

10.6 10.6
0 100 200 300 400 500 600 0 100 200 300 400 500 600
Duraon (Days) Duraon (Days)

Fig. 4. pH vs. Cl concentration in geopolymer and PC concretes.


508 C. Gunasekara et al. / Construction and Building Materials 200 (2019) 502–513

concretes would increase the corrosion rates for the active speci- pH value of Gladstone, Pt.Augusta, Collie and PC concretes at
mens (1% chloride and above) and reduce the corrosion rates for 540 days were 11.57, 11.51, 10.92 and 11.42, respectively. Collie
the passive specimens (0.5% chloride and below). geopolymer had a significant pH drop between 28 and 540 days,
Images of the steel rebars after 540 days of exposure are shown compared to the other two geopolymer concretes. Both Gladstone
in Fig. 5. No corrosion induced cracking was noted around the steel and Pt.Augusta exhibited slightly higher pH than PC concrete at
reinforcement, but uniform and pitting corrosion can be observed 540 days. Again, the drop in pH observed can be attributed to fur-
on the bars. The pitting attack is concentrated on the rebar surface ther geopolymerisation and to acid formation in the pits due to
and is more prevalent for the higher chloride concentrations (2– corrosion. The magnitude of the pH reductions suggest that acid
5%). It is evident that there is a substantial degree of corrosion of formation may be the more influential, with a greater drop in pH
the reinforcement embedded in Gladstone and Pt.Augusta fly ash observed for the specimens with higher rates of corrosion and
based geopolymer concretes. Conversely in the Collie fly ash more corrosion evident on the rebar.
geopolymer concrete, less rust formation on the surface of the steel Fig. 7 illustrates the corroded surface of the rebars of fly ash
rebars can be observed compared to the other two geopolymers. In geopolymer and PC concretes immersed in 3% sodium chloride
the PC concrete, minimal deposition of steel corrosion products is solution. It was observed that there was a significant amount of
identifiable at 2–5% chloride. These observations correlate well rust formed on the surface of reinforcement in the Collie geopoly-
with the observed potentials and corrosion rate measurements, mer and PC concretes. Both concretes showed uniform and local-
which indicate a higher level of corrosion in the Gladstone and ized corrosion throughout their rebar surface. On the other hand,
Pt. Augusta specimens containing high chloride concentrations, Gladstone and Pt.Augusta geopolymer displayed more isolated
and little corrosion evident on the specimens containing 0.5% chlo- regions of corrosion and overall, less rust formation on the rebar
ride and below, further suggesting that these are in a passive state. surface. This again correlates well with the corrosion rate
The corrosion rates (Icorr) of the geopolymer and PC concretes measurements.
immersed in 3% sodium chloride solution are shown in Fig. 6. The chloride binding capacity in ponded concrete specimens at
The trends observed were similar to those for the cast-in speci- rebar level (i.e. 20–30 mm depth) is calculated by subtracting the
mens. The Gladstone, Pt Augusta and PC specimens showed a measured free chloride content from measured total chloride con-
decrease in Icorr values in the first 90 days, indicative of passivation, tent. The calculated chloride binding capacity using Eq. (2)
though the Collie showed an increase in corrosion rate throughout together with the measured total/free chloride contents in each
having an Icorr above 1.0 lA/cm2 by day 90. All four concretes dis- fly ash geopolymer and PC concrete is tabulated in Table 4.
played a consistent increase between 90 and 540 days, but of dif-  
ferent proportions, indicating corrosion had been initiated. The  ðTotal Cl  Free Cl Þ
Cl binding capacity ¼   100 ð2Þ
Icorr values for Gladstone, Pt.Augusta and Collie geopolymer con- Total Cl
cretes ranged from 0.55 to 1.05 lA/cm2, 0.65–1.10 lA/cm2 and In PC concrete, the chemical binding occurs due to hydrated cal-
0.85–1.65 lA/cm2, respectively, whereas Icorr of PC concrete varied cium aluminate derived from C3A, which produces Friedel’s salt.
between 0.65 and 1.20 lA/cm2. Gladstone geopolymer had the Moreover, chloride ions physically bond with cement hydration
lowest corrosion rate while Collie displayed the highest across products, C-S-H and C-A-H gel [40]. In geopolymer concrete the
the entire testing period. The PC concrete displayed higher corro- chloride ions can react with alumina and calcium oxide in undis-
sion rates than the Gladstone and Pt.Augusta geopolymers. solved fly ash grains, and form Friedel’s salt (3CaOAl2O3CaCl2-
The alkalinity of the four concretes was measured at 28 and 10H2O) [41]. The PC concrete displayed the highest chloride
540 days. The data confirmed that fly ash geopolymer concretes binding capacity, consistent with the chemical composition of
had a lower initial pH (pH = 12.05) than PC concrete (pH = 12.62). the PC and formation of Friedel’s salt. As expected, the highest total
The alkalinity of all concretes reduced with time. The measured chloride content together with the lowest chloride binding

Fig. 5. Corrosion pattern of rebar vs. Cl concentration in geopolymer and PC concretes.
C. Gunasekara et al. / Construction and Building Materials 200 (2019) 502–513 509

1.75 geopolymers (i.e. G0, PA0, C0), the number of unreacted grains
had reduced in all specimens. This was attributed to additional
1.50
High alkali-fly ash reactivity over this period which results in the gel
1.25 produced filling the cracks and voids visible at 28 days, thus
increasing the compressive strength.
Icorr (μA/cm2)

1.00 However, propagated cracks can still be observed after 540 days
for the geopolymers containing chlorides, with the width of these
0.75 Moderate
cracks in Collie being visibly larger than those in the Gladstone and
0.50 Pt.Augusta matrices. Further inspection of the microstructure
reveals that the number of cracks and the width of the cracks in
Gladstone PC Low
0.25 the geopolymer matrix increased with an increase of chloride con-
Collie Pt.Augusta centration in all geopolymer concretes, consistent with the
0.00
0 100 200 300 400 500 600 reduced compressive strengths observed. It is hypothesized that
Duraon (Days) an increase in the corrosion products in rebar would increase the
internal stresses in the adjacent concrete, resulting in formation
Fig. 6. Corrosion rate of geopolymer and PC concrete specimens, immersed in 3% of cracks in the gel matrix and the widening of existing cracks.
NaCl solution.
Fig. 9 illustrates the cumulative macropore volume changes of
the fly ash geopolymer and PC concrete between 50 nm and
25 mm. All concretes showed a porosity reduction between 28
and 365 days, but by different percentages. The degree of geopoly-
merization determines the quantity of gel formation, and this gel
fills the gaps between unreacted fly ash particles and pore spaces,
thus refining the size of these pores. Two major zones were iden-
tified in the examined macropore region, i.e. pore size ranging from
50 nm to 1 mm and 1 mm–25 mm. Collie and Gladstone geopolymer
concretes displayed the highest and lowest macroporosity, respec-
tively, at both 28 and 365 days. Both Gladstone and Pt.Augusta dis-
played a substantial porosity reduction in the range of 50 nm–1 mm
over this period. PC concrete also showed a porosity reduction dur-
ing this period, which was significant in the range of 1 mm–25 mm.
The pore size distribution in Collie varied significantly compared to
the other two fly ash geopolymers, having the lowest and highest
pore volume reduction at 50 nm–1 mm and 1 mm–25 mm regions,
respectively between 28 and 365 days. This variation in pore size
distribution is consistent with the lower compressive strength
Fig. 7. Corrosion pattern of rebar in geopolymer and PC concrete specimens,
immersed in 3% sodium chloride solution. and higher chloride ingress observed for the Collie geopolymer
concrete.
FTIR analysis of three different fly ash based geopolymer con-
capacity was observed for the Collie specimen, given the low, 0.3% cretes at reinforcement-geopolymer binder interface (at 540 days)
calcium oxide content in the ash. The Gladstone and Pt.Augusta is shown in Fig. 10. The corrosion products were examined at the
samples had similar calcium oxide contents and displayed similar rebar/binder interface of 3% cast in chloride geopolymer specimens
binding capacities. This is attributed to the calcium–aluminosili- and compared to the ponding specimens. FTIR analysis identified a
cate (C–A–S–H) gel formed during geopolymerisation physically strong peak of hematite (Fe2O3) and akageneite (FeO(OH)) between
adsorbing chlorides on the surface of the C–A–S–H gel. 770 cm1 and 690 cm1 in all three geopolymers with cast-in chlo-
Fig. 8 shows the microstructure observed in the fly ash geopoly- rides, whereas a small peak of lepidocrocite [c-FeO(OH)] is noted
mer concretes at 28 and 540 days at 0%, 2% and 5% sodium chloride at 1155 cm1. However, for ponded specimens, FTIR analysis did
concentrations. At 0% all geopolymer concretes showed an increase not detect this lepidocrocite peak, and only detected a small peak
in the packing density in the microstructure from 28 to 540 days. identified with hematite and akageneite.
Gladstone has a well compacted, uniform, dense microstructure Peaks associated with A–O–Al/Al–OH and Si–O–Si/Si–OH bonds
with a small number of unreacted/partially reacted fly ash spheres were observed for all geopolymer concretes at 1030 cm1 and
and micro-cracks. Pt.Augusta has a similar microstructure to Glad- 580 cm1, respectively. These bonds are indicators of N–A–S–H
stone, but is heterogeneous in nature, with an increased number of gel [42], which is formed during the formation of the three dimen-
micro-cracks and unreacted grains. At 28 days, the Collie geopoly- sional geopolymer network. It was interesting to note that a broad
mer has a loosely packed microstructure with a large quantity of peak with a small hump occurred at 1030 cm1 for cast-in chloride
unreacted fly ash particles. At 540 days in non-chloride specimens, compared to the narrow, sharp peak observed for the

Table 4
Chloride binding capacity in geopolymer and PC concretes.

Concrete Type Cl (% w/w of Cement) Cl binding Capacity (%)


Total Cl Free Cl Bound Cl
Gladstone GP 0.80166 0.30715 0.49451 61.7
Pt.Augusta GP 0.82659 0.30001 0.52658 63.7
Collie GP 1.33667 0.65001 0.68666 51.4
PC Concrete 1.21260 0.42716 0.78544 64.8
510 C. Gunasekara et al. / Construction and Building Materials 200 (2019) 502–513

Fig. 8. Microstructure development of different fly ash based geopolymer concretes.

0.105
50nm < < 1μm
is shown in Fig. 11. All fly ash geopolymer concretes exhibited
Gladstone-28d
Pt.Augusta-28d
peaks of quartz (SiO2) and mullite (Al6SiO2O13), which are indica-
0.095
Collie-28d tive of the formation of zeolytes during geopolymerisation [43].
Cumulave Pore Volume (ml/g)

0.085 PC-28d In addition to these phases, all geopolymers displayed peaks of cor-
Gladstone-365d rosion products in the form of hematite (Fe2O3) and akageneite
0.075
Pt.Augusta-365d (FeO(OH)), as noted in the FTIR spectra. Both Gladstone and Pt.
0.065 Collie-365d Augusta geopolymer concretes showed a higher intensity of hema-
PC-365d tite and akageneite in the 3% cast-in chloride specimens, compared
0.055
to chloride immersion specimens. However, Collie displayed a
0.045 higher quantity of corrosion products in the immersion specimens.
The XRD results are consistent with the observed corrosion levels
0.035
on the rebar in the specific geopolymer concrete. It is hypothesized
0.025 that with an increase in the formation of three dimensional N–A–
S–H and C–A–S–H cross linking phases in the gel matrix reduces
0.015
the porosity and increases the resistance of the hardened matrix
0.005 to chloride ingress. Hence, the lower ingress of chloride observed
10 100 1000 10000 100000
in the Gladstone and Pt. Augusta concretes.
Pore Diameter, (nm)

Fig. 9. Pore size distribution of geopolymer and PC concretes between 28 and 5. Discussion
365 days.

PC concrete develops compressive strength through the forma-


tion of calcium silicate hydrates, which are produced by the hydra-
ponding specimens, in particular for Gladstone and Pt.Augusta. tion process of water, reacting with the PC [28,44]. However, fly
With increasing degree of geopolymerisation, the A–O–Al/Al–OH ash based geopolymer concrete follows a different reaction path-
bonds tend to become sharper and have an increase in peak height way and gain compressive strength through the formation of alu-
[42]. Hence, the observed FTIR data correlate well with the minosilicate gel. The gel is produced by dissolving alumina and
observed higher degree of geopolymerisation in the ponded silica oxides of the fly ash in a highly alkaline medium, typically
geopolymer specimens compared with the cast-in chloride provided by the sodium hydroxide and sodium silicate solutions.
specimens. The rapid reaction of Gladstone fly ash in alkaline solution, due
The XRD analysis of the different fly ash based geopolymer con- to the higher surface area coupled with a large quantity of reactive
crete specimens at the reinforcement-geopolymer binder interface amorphous Al2O3 and SiO2 in the range of 10–45 mm, is expected to
C. Gunasekara et al. / Construction and Building Materials 200 (2019) 502–513 511

1 [Lepidocrocite] 2 [N-A-S-H /Al-OH gel] 3 [Akaganeite/FeO(OH)] 4 [Hemate /Fe2O3] 5 [N-A-S-H /Si-OH gel]

(a) 3% cast in Chloride specimes (b) Ponded specimes

1 2 3 4 5 2 3 4 5

Collie Collie
Pt.Augusta

Pt.Augusta
Gladstone
Gladstone

1400 1300 1200 1100 1000 900 800 700 600 500 1400 1300 1200 1100 1000 900 800 700 600 500
Wave number (cm-1) Wave number (cm-1)

Fig. 10. FTIR analysis of three different fly ash geopolymer concretes at 540 days.

60 60 60
G3 Q : Quartz (siO2) PA3 Q : Quartz (siO2) C3 Q : Quartz (siO2)
50 M : Mullite (Al6SiO2O13) 50 M : Mullite (Al6SiO2O13) 50 M : Mullite (Al6SiO2O13)
H : Hemate (Fe2O3) H : Hemate (Fe2O3) H : Hemate (Fe2O3)
Q Q A : Akaganeite [FeO(OH)]
Intensity

40 A : Akaganeite [FeO(OH)] 40 A : Akaganeite [FeO(OH)] 40


Q
30 30 30

20 20 20
Q M M Q
A
10 H 10 Q M H 10 H
H M A H H M M
M Q A A A H A M H A H A
H M H Q H A Q
0 0 0
20 30 40 50 60 70 20 30 40 50 60 70 20 30 40 50 60 70
60 60 60
Q : Quartz (siO2) Q Q : Quartz (siO2) Q
Gimm PAimm Cimm Q : Quartz (siO2)
50 M : Mullite (Al6SiO2O13) 50 M : Mullite (Al6SiO2O13) 50 M : Mullite (Al6SiO2O13)
H : Hemate (Fe2O3) H : Hemate (Fe2O3) H : Hemate (Fe2O3)
Q A : Akaganeite [FeO(OH)] A : Akaganeite [FeO(OH)]
Intensity

40 40 40 A : Akaganeite [FeO(OH)]

30 30 30

20 20 20

Q Q A
10 Q M 10 10 H
M H M
H A H Q A M M H M
A H A M H M
A H A Q M A H Q A
0 0 0
20 30 40 50 60 70 20 30 40 50 60 70 20 30 40 50 60 70
2 Theta (degrees) 2 Theta (degrees) 2 Theta (degrees)

Fig. 11. XRD analysis of three different fly ash geopolymer concretes at 540 days.

produce large quantities of sodium–aluminosilicate (N–A–S–H) meric reaction because of its high reactivity, thus, the minimal
and calcium–aluminosilicate (C–A–S–H) gel [45] and form a uni- strength development (14.1%) observed between 28 and 540 days.
form condensed gel matrix at an early age, Fig. 8. This dense Pt.Augusta contains a higher calcium oxide percentage, which
microstructure together with a lower macroporosity, Fig. 9, results in addition to the reactive amorphous Al2O3 and SiO2 in 10–
in the high strength of Gladstone at 28 days compared to the PC 20 mm range, produces additional C–A–S–H gel as well as the N–
concrete. However, little fly ash remains for subsequent geopoly- A–S–H gel. The combination of the N–A–S–H and increased
512 C. Gunasekara et al. / Construction and Building Materials 200 (2019) 502–513

amounts of C–A–S–H gels provides additional rigidity to the three- into the geopolymeric backbone, resulting in a reduction of
dimensional geopolymeric matrix. However, due to the lower sur- pH in the pore solution compared to PC concrete. This leads
face area, the alkali–fly ash dissolution rate is expected to be lower to an increase in the Cl/OH ratio in the geopolymer concrete
than Gladstone fly ash. This resulted in the formation of a non- and hence higher rates of corrosion.
uniform, heterogeneous geopolymer gel matrix which in turn 2. Gladstone and Pt.Augusta geopolymer concretes displayed con-
accounts for the lower compressive strength displayed by Pt tradictory corrosion behaviour when specimens were ponded
Augusta. Collie has a lower degree of alkali–fly ash reactivity due as compared with cast-in chlorides, showing a lower corrosion
to its lower surface area and small quantity of reactive amorphous rate than PC concrete. This is attributed to formation of three
Al2O3 and SiO2 available in the range of 10–45 mm. This is hypoth- dimensional N–A–S–H and C–A–S–H cross linking in the gel
esised as the reason that the Collie demonstrated a less compact, matrix which decreases the diffusion of chlorides in the Glad-
non-uniform cracked microstructure, which was observed to be a stone and Pt.Augusta geopolymer.
significantly different microstructure compared to the other two 3. The chloride diffusion rate in geopolymer concrete is dependent
geopolymers. Hence, the higher macroporosity and lower com- on the porosity of the geopolymer matrix, which is governed by
pressive strength of the Collie geopolymer concrete. the reactivity of the precursor fly ash. The reactivity is depen-
During geopolymerization, the glassy Al2O3 and SiO2 in the pre- dent upon surface area, amorphous Al2O3 and SiO2 content
cursor fly ash react with the strong alkalis and produces a complex and CaO content. Higher surface area and amorphous Al2O3
three-dimensional polymeric aluminosilicate gel network. Hence, and SiO2 increase aluminosilicate gel formation and produce a
the greater the geopolymerisation and gel formation, the higher stronger, less porous geopolymer matrix which reduces the rate
the OH agglomeration into the geopolymeric backbone is to be of chloride ingress.
expected. This would result in reduction of pH in the pore solution 4. The chloride binding capacity of low calcium fly ash based
as the activator is consumed by the geopolymerisation reaction. It geopolymer concrete is dependent on the CaO content in pre-
has been noted that the relative concentrations of the Cl/OH ions cursor, which produces C–A–S–H gel in the geopolymerisation,
in the pore electrolyte govern the risk of chloride-induced corro- resulting in the physical adsorption of chloride ions on the gel
sion in cement matrices [28]. Hence, the increase of Cl/OH ratio surface, reducing the rate of chloride ingress.
due to reduction of alkalinity in Gladstone geopolymer results in 5. Corrosion products in the form of hematite [Fe2O3], akageneite
the highest corrosion rate being observed in the cast in chlorides [FeO(OH)] and lepidocrocite [c-FeO(OH)] are identified.
specimens. The increase in the corrosion rates for the Pt.Augusta
and Collie geopolymer concretes is attributed to the on-going Conflict of interest
geopolymerisation lowering the OH concentration in the respec-
tive pore solutions and thus an increase in the Cl/OH ions in None.
the pore solution.
Conversely both Gladstone and Pt.Augusta had lower corrosion Acknowledgements
rates than the PC concrete for the ponded specimens. Kupwade-
Patil et al. [46] observed similar corrosion behavior for ponded Flyash Australia Pty Ltd and Cement Australia Pty Ltd are acknowl-
fly ash geopolymer concrete specimens, with Icorr = 0.166 lA/cm2 edged for providing fly ash for this research. RMIT University pro-
and Icorr = 2.38 lA/cm2 for geopolymer and PC concrete respec- vided the microscopy facility, X-ray facility and microanalysis
tively. This is attributed to the lower rate of ingress of chloride, cor- facility for this study also acknowledged. Australian Research
responding to a lower chloride diffusion coefficient for the Council Industrial Transformation Research Hub for Nano-science
geopolymer concrete. The availability of free chlorides in the con- based construction material manufacturing (IH150100006) is
crete is also a governing factor. Furthermore, these two geopoly- funded for this work.
mers showed similar chloride binding capacity compared to PC
concrete. On the other hand, Collie had the highest chloride con-
Compliance with Ethical Standards
tent, the higher free chloride content in the pore solution and the
lowest chloride binding capacity. This resulted in the highest cor-
The authors declare that they have no conflict of interest.
rosion rate being observed for the Collie geopolymer specimens
under ponded conditions. The cast-in data would indicate that
for an equivalent chloride concentration at the rebar a higher cor- References
rosion rate can be expected for the Gladstone and Pt.Augusta
[1] E. Gartner, Industrially interesting approaches to low-CO2 cements, Cem.
geopolymers, due to their lower pH. However, due to the low chlo- Concr. Res. 34 (9) (2004) 1489–1498.
ride diffusion coefficients of these geopolymers the rate of ingress [2] C. Meyer, The greening of the concrete industry, Cem. Concr. Compos. 31 (8)
will be slower and thus the time until the critical chloride value (2009) 601–605.
[3] C. Chen et al., Environmental impact of cement production: detail of the
required to initiate corrosion will be extended. Understanding different processes and cement plant variability evaluation, J. Cleaner Prod. 18
the relationship between diffusion rate and critical chloride con- (5) (2010) 478–485.
centration for geopolymer concretes will be critical in determining [4] J. Peng et al., Modeling of carbon dioxide measurement on cement plants, Adv.
Mater. Res. 610–613 (1) (2013) 2120–2128.
their long term durability. [5] C. Li et al., CO2 emissions due to cement manufacture, Mater. Sci. Forum 685
(1) (2011) 181–187.
[6] D.N. Huntzinger, T.D. Eatmon, A life-cycle assessment of Portland cement
6. Summary and conclusion manufacturing: comparing the traditional process with alternative
technologies, J. Cleaner Prod. 17 (7) (2009) 668–675.
[7] N. Mahasenan, S. Smith, K. Humphreys, The cement industry and global
The key findings of this research study can be summarized as climate change: current and potential future cement industry CO2 emissions,
follows: in: Greenhouse Gas Control Technologies-6th International Conference, 2003,
Elsevier.
[8] G. Habert, J.B. d’Espinose de Lacaillerie, N. Roussel, An environmental
1. Fly ash geopolymer concretes showed significantly higher cor- evaluation of geopolymer based concrete production: reviewing current
rosion rates than PC concrete with the same concentration of research trends, J. Cleaner Prod. 19 (11) (2011) 1229–1238.
[9] L.K. Turner, F.G. Collins, Carbon dioxide equivalent (CO2-e) emissions: a
cast-in chlorides. This is attributed to geopolymerisation and
comparison between geopolymer and OPC cement concrete, Constr. Build.
aluminosilicate gel formation leading to OH agglomeration Mater. 43 (1) (2013) 125–130.
C. Gunasekara et al. / Construction and Building Materials 200 (2019) 502–513 513

[10] M. Talha Junaid et al., A mix design procedure for low calcium alkali activated [28] A.M. Neville, Properties of Concrete. Fourth and Final Edition ed. Standards
fly ash-based concretes, Constr. Build. Mater. 79 (2015) 301–310. updated to 2002, 1996, Harlow: Pearson Education Limited.
[11] M. Alexander, M. Thomas, Service life prediction and performance testing – [29] AS, Method of testing concrete, Method 9: Determination of the compressive
current developments and practical applications, Cem. Concr. Res. 78 (2015) strength of concrete specimens, in AS (Australian Standards). 1999, Standards
155–164. Australia: Australia. pp. 1–12.
[12] G.K. Glass, N.R. Buenfeld, The presentation of the chloride threshold level for [30] D.W. Law et al., Measurement of loss of steel from reinforcing bars in concrete
corrosion of steel in concrete, Corros. Sci. 39 (5) (1997) 1001–1013. using linear polarisation resistance measurements, NDT & E Int. 37 (5) (2004)
[13] M.S. Badar et al., Corrosion of steel bars induced by accelerated carbonation in 381–388.
low and high calcium fly ash geopolymer concretes, Constr. Build. Mater. 61 [31] C. Andrade, C. Alonso, Corrosion rate monitoring in the laboratory and on-site,
(1) (2014) 79–89. Constr. Build. Mater. 10 (5) (1996) 315–328.
[14] R.R. Lloyd, J.L. Provis, J.S.J. van Deventer, Pore solution composition and alkali [32] M. Stern, A.L. Geary, Electrochemical polarization. 1. A theoretical analysis of
diffusion in inorganic polymer cement, Cem. Concr. Res. 40 (9) (2010) 1386– the shape of polarization curves, J. Electrochem. Soc. 104 (1) (1957) 56–63.
1392. [33] V. Räsänen, V. Penttala, The pH measurement of concrete and smoothing
[15] C. Gunasekara, D.W. Law, S. Setunge, Long term permeation properties of mortar using a concrete powder suspension, Cem. Concr. Res. 34 (5) (2004)
different fly ash geopolymer concretes, Constr. Build. Mater. 124 (2016) 352– 813–820.
362. [34] R.S. Barneyback, S. Diamond, Expression and analysis of pore fluids from
[16] K. Kupwade-Patil, E.N. Allouche, Examination of chloride-induced corrosion in hardened cement pastes and mortars, Cem. Concr. Res. 11 (2) (1981) 279–285.
reinforced geopolymer concretes, J. Mater. Civ. Eng. 25 (10) (2012) 1465–1476. [35] M. Pourbaix, Atlas of Electrochemical Equilibrium in Aqueous Solutions, NACE
[17] N. Ganesan, R. Abraham, S.D. Raj, Durability characteristics of steel fibre International, Houston, TX, 1996.
reinforced geopolymer concrete, Constr. Build. Mater. 93 (2015) 471–476. [36] A. Poursaee, C.M. Hansson, Potential pitfalls in assessing chloride-induced
[18] M. Babaee, A. Castel, Chloride-induced corrosion of reinforcement in low- corrosion of steel in concrete, Cem. Concr. Res. 39 (5) (2009) 391–400.
calcium fly ash-based geopolymer concrete, Cem. Concr. Res. 88 (2016) 96– [37] H.R. Soleymani, M.E. Ismail, Comparing corrosion measurement methods to
107. assess the corrosion activity of laboratory OPC and HPC concrete specimens,
[19] M. Olivia, H. Nikraz Durability of fly ash geopolymer concrete in a seawater Cem. Concr. Res. 34 (11) (2004) 2037–2044.
environment, in: Proceedings of the CONCRETE 2011 Conference, 2011, The [38] M. Olivia, H. Nikraz, Properties of fly ash geopolymer concrete designed by
Concrete Institute of Australia. Taguchi method, Mater. Des. 36 (1) (2012) 191–198.
[20] K. Pasupathy et al., Durability of low-calcium fly ash based geopolymer [39] I. Martinez, C. Andrade, Polarization resistance measurements of bars
concrete culvert in a saline environment, Cem. Concr. Res. 100 (2017) 297– embedded in concrete with different chloride concentrations: EIS and DC
310. comparison, Mater. Corros. 62 (10) (2011) 932–942.
[21] Y. Muntingh, Durability and Diffusive Behaviour Evaluation of Geopolymeric [40] R. Luo et al., Study of chloride binding and diffusion in GGBS concrete, Cem.
Material, University of Stellenbosch, Stellenbosch, 2006. Concr. Res. 33 (1) (2003) 1–7.
[22] C. Monticelli et al., Corrosion behavior of steel in alkali-activated fly ash [41] D.A. Koleva et al., Correlation of microstructure, electrical properties and
mortars in the light of their microstructural, mechanical and chemical electrochemical phenomena in reinforced mortar. Breakdown to multi-phase
characterization, Cem. Concr. Res. 80 (2016) 60–68. interface structures. Part I: Microstructural observations and electrical
[23] J. Miranda et al., Corrosion resistance in activated fly ash mortars, Cem. Concr. properties, Mater. Charact. 59 (3) (2008) 290–300.
Res. 35 (6) (2005) 1210–1217. [42] P. Duxson et al., Geopolymer technology: the current state of the art, J. Mater.
[24] D. Bastidas et al., A study on the passive state stability of steel embedded in Sci. 42 (2007) 2917–2933.
activated fly ash mortars, Corros. Sci. 50 (4) (2008) 1058–1065. [43] M. Criado et al., An XRD study of the effect of the SiO2/Na2O ratio on the alkali
[25] M.P.C.M. Gunasekara, D.W. Law, S. Setunge, Effect of composition of fly ash on activation of fly ash, Cem. Concr. Res. 37 (5) (2007) 671–679.
compressive strength of fly ash based geopolymer mortar, in: 23rd [44] F. Micelli, A. Nanni, Durability of FRP rods for concrete structures, Constr.
Australasian Conference on the Mechanics of Structures and Materials Build. Mater. 18 (7) (2004) 491–503.
(ACMSM23). 2014. Byron Bay, Australia. [45] M.P.C.M. Gunasekara, D.W. Law, S. Setunge, Effect of composition of fly ash on
[26] AS, Supplementary cementitious materials for use with portland and blended compressive strength of fly ash based geopolymer mortar, in: 23rd
cement, Part 1: Fly ash, in AS (Australian Standards). 1998, Standards Australasian Conference on the Mechanics of Structures and Materials
Australia: Australia. pp. 1–13. (ACMSM23), S.T. Smith, Editor. 2014: Byron Bay, Australia, pp. 113–118.
[27] C. Gunasekara et al., Zeta potential, gel formation and compressive strength of [46] K. Kupwade-Patil, E.N. Allouche, Examination of chloride-induced corrosion in
low calcium fly ash geopolymers, Constr. Build. Mater. 95 (1) (2015) 592–599. reinforced geopolymer concretes, J. Mater. Civ. Eng. 25 (10) (2013) 1465–1476.

You might also like