Influence of sulfate ion and associated cation type on steel reinforcement corrosion in concrete powder aqueous solution in the presence of chloride ions

Cement and Concrete Research 91 (2017) 73–86

Contents lists available at ScienceDirect

Cement and Concrete Research journal homepage: www.elsevier.com/locate/cemconres

Influence of sulfate ion and associated cation type on steel reinforcement corrosion in concrete powder aqueous solution in the presence of chloride ions Fouzia Shaheen, Bulu Pradhan ⁎ Department of Civil Engineering, Indian Institute of Technology Guwahati, Guwahati 781039, India

a r t i c l e

i n f o

Article history: Received 9 February 2016 9 October 2016 Accepted 11 October 2016 Available online 18 October 2016 Keywords: Corrosion (C) Chloride (D) Reinforcement (D) Sulfate (D) Concrete (E)

a b s t r a c t State of rebar corrosion in concrete powder aqueous solution contaminated with chloride and sulfate ions has been determined by conducting a potentiodynamic polarization test. XRD analysis and FTIR spectroscopy were also performed. From the results, different zones of corrosion in terms of potential ranges have been identified. The presence of Na2SO4 has mitigated the effect of chloride ions whereas the presence of MgSO4 has stimulated the effect of chloride ions on reducing the passivity of steel reinforcement in a chloride environment. Ordinary Portland cement performed better against Mg-oriented sulfate attack whereas Portland pozzolana cement performed better against Na-oriented sulfate attack in the presence of chloride ions. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction Concrete is a heterogeneous material with many special characteristics, including high alkalinity of the pore solution, high electrical resistivity and has a structure that acts as a physical barrier for mass transport and crack behaviour [1]. Concrete durability problems are the major cause of concern all over the world especially in the coastal areas where the structures are undergoing deterioration well before their expected life and need proper attention and care [2]. In the marine environment, chloride and sulfate salts do exist concomitantly. The conjoint presence of these salts may cause deterioration of concrete due to reinforcement corrosion and sulfate attack [3]. Chloride ions enter into concrete through two sources such as internal chloride (at the time of preparation of concrete, through chloride contaminated aggregates, chloride-containing admixtures, or mixing water) and external chloride (entering from the external environment into the hardened concrete through deicing salts, sea water, soil and ground water) [4]. In the hardened cement paste, when both chlorides and sulfates exist conjointly, they react with tricalcium aluminate (C3A) to form various compounds. Chloride ions react with C3A to form calcium chloroaluminate (Friedel's salt) [5]. Sulfate ions react with calcium hydroxide and hydrated C3A to form gypsum and ettringite respectively, which lead to expansion and disruption of hardened concrete. The higher proportions of C3A would reduce the level of ⁎ Corresponding author. E-mail address: [email protected] (B. Pradhan).

http://dx.doi.org/10.1016/j.cemconres.2016.10.008 0008-8846/© 2016 Elsevier Ltd. All rights reserved.

reinforcement corrosion by lowering corrosion-inducing free chlorides from the concrete pore solution, but it would pose a serious concrete durability problem in terms of sulfate attack [6]. The electrolytic pore solution of concrete has a pH value between 12.5 and 13.5 due to the presence of calcium hydroxide along with small amounts of Na2O and K2O [1]. In this highly alkaline environment, a thin protective layer known as a passive layer of γ-Fe2O3 is formed on the rebar surface [7– 9]. However the free chloride content (also known as water soluble chloride) in the pore solution of concrete breaks down the passive layer of steel and initiates corrosion [10]. Sulfate ions enter into concrete externally from exposure condition and internally through aggregates, mixing and curing water, and admixtures [3]. Sulfate ions can be associated with magnesium, calcium and sodium cations. Since the solubility of calcium sulfate is very low, magnesium and sodium sulfate are mainly responsible for the deterioration of concrete due to sulfate attack [11]. In the conjoint presence of chloride and sulfate ions, the mechanism of deterioration of concrete becomes complex due to the simultaneous interaction of these ions with hydrated cement phases. Further, the cation type associated with these aggressive ions makes the mechanism even more complex. From the review of literature, it is inferred that different researchers have studied the corrosion behaviour of steel embedded in concrete exposed to chloride, sulfate and mixed chloride-sulfate environment. Zuquan et al. [12] reported that the presence of sulfate ions in a composite solution enhanced the resistance to chloride penetration into the concrete at the early stage of exposure, but the opposite behaviour was observed at a later exposure period. Further from XRD results, the authors observed the formation of a higher amount of

74

F. Shaheen, B. Pradhan / Cement and Concrete Research 91 (2017) 73–86

Nomenclature CCA E G CH MH CC T Q

Calcium chloroaluminate Ettringite Gypsum Calcium hydroxide Magnesium hydroxide Calcium carbonate Thaumasite Quartz

gypsum and ettringite in the concrete exposed to sulfate solution as compared to that exposed to composite sulfate-chloride solution. A study conducted by Al-Amoudi and Maslehuddin [7] on the respective effect of chloride and sulfate ions on corrosion of steel in cement paste indicated that sulfate ions are hardly able to initiate reinforcement corrosion. However considerable reinforcement corrosion was observed in specimens immersed in mixed chloride-sulfate solution. Jarrah et al. [13] reported that the time to initiation of reinforcement corrosion was higher in blended cement than plain cement in all chloride and chloride-sulfate solutions and the sulfate ions alone do not initiate the reinforcement corrosion but they increase the corrosion activity once the corrosion was initiated due to the conjoint presence of chloride and sulfate salts. Dehwah et al. [14] have found that the presence of sulfate ions in a chloride environment did not affect the time to initiation of reinforcement corrosion but corrosion current density increased with an increase in the concentration of sodium sulfate and magnesium sulfate. The presence of chloride and sulfate salts significantly affects the pore solution chemistry of concrete. The pore solution of concrete significantly varies in the concentration of hydroxyl ions, sulfate ions, chloride ions, solubility of metal cations and temperature [15]. Different researchers have investigated the corrosion behaviour of reinforcing steel in simulated concrete pore solutions. Dehwah et al. [9] studied the effect of chloride and sulfate contamination on the pore solution chemistry in plain and blended cements and the results from the study indicated that the chloride binding capacity of both plain and blended cements decreased due to the concomitant presence of chloride and sulfate salts. It was also observed that the increase in chloride and sulfate ions in mixed chloride-sulfate solution enhances the corrosion of steel reinforcement. Zhang et al. [16] have found that the passivity breakdown potential or pitting potential is lower with higher chloride concentrations and they also observed that the higher pH of the simulated concrete pore solution facilitates the passivation of steel, whereas the higher concentration of Cl− leads to a passivity breakdown at relatively lower potential. Ghods et al. [17] have investigated the growth of oxide film on steel surface in saturated calcium hydroxide solution with different amounts of NaOH, KOH and Ca(SO)4. The authors reported that pore solution composition has an effect on the protective properties of passive oxide film and the presence of SO2− ions in the 4 pore solution has a significant negative effect on the protective properties of the passive oxide film. Chen et al. [18] reported that in chloride free saturated calcium hydroxide solution, steel remains in a passive state and the corrosion current of steel was very low. On the other hand in the saturated calcium hydroxide with 0.5 M NaCl, the corrosion current density increased considerably and the steel surface was unstable with chloride attack and localized corrosion appeared with FeCO3 and Fe2O3 as main corrosion products on its surface. Padilla and Alfantazi [19] found that the most severe corrosion damage in terms of corrosion rate and degradation of the corrosion film formed on the surface and was observed when galvanized steel was immersed in NaCl + Na2SO4 solution whereas the best performance was observed in MgCl2 + Na2SO4 solution. Pradhan and Bhattacharjee [20] have

conducted a potentiostatic study on reinforcing steel in chloride contaminated concrete powder solution extracts. The authors reported that chloride content has the strongest influence in governing zones of corrosion of steel reinforcement as compared to other parameters. Aal et al. [21] have reported that the pitting corrosion current density in constant Ca(OH)2 solution increased with the increase in SO24 − and Cl− ions, whereas under the constant concentrations of SO2− and Cl− 4 ions, pitting corrosion current density decreased with an increase in the concentration of Ca(OH)2. From the review of past research work, it is observed that very few studies have been conducted to evaluate the corrosion performance of steel reinforcement in simulated concrete pore solutions contaminated with composite chloride-sulfate salts. Further, in these studies mostly saturated calcium hydroxide solution has been taken as the simulated pore solution. On the other hand, concrete powder aqueous solutions can also be taken as the electrolytic concrete pore solution. Since concrete powder from which the aqueous solution is extracted, is a mixture of cement hydrates, coarse aggregates and fine aggregates, the aqueous solution may represent the electrolytic pore solution of concrete more closely as compared to saturated calcium hydroxide solution. From the literature review, it is inferred that the work on corrosion performance of steel in concrete powder aqueous solutions contaminated with chloride ions and/or composite chloride-sulfate ions is meager. Therefore, in the present research work, an attempt has been made to study the electrochemical behaviour of steel in electrolytic concrete powder aqueous solutions contaminated with chloride and composite chloride-sulfate ions through anodic polarization curves by conducting a potentiodynamic polarization test. Further to analyze the effect of chemical composition of the concrete powder aqueous solutions on corrosion behaviour of steel reinforcement, ionic concentration, pH and conductivity of concrete powder aqueous solutions were determined. In addition, for the purpose of analyzing the changes in phase composition of hardened concrete and for identifying different functional groups associated with different products formed in concrete in the presence of chloride ion and sulfate ions, X-ray diffraction (XRD) analysis and Fourier transform infrared (FTIR) spectroscopy were conducted. 2. Experimental work The experimental program has been designed to evaluate the corrosion performance of steel in ordinary Portland cement (OPC) and Portland pozzolana cement (PPC) concrete powder aqueous solutions contaminated with different concentrations of chloride ions and sulfate ions. The details of materials used, test specimens and tests conducted in the present investigation are presented below. 2.1. Materials used and specimen preparation Concrete cube specimens with a size of 150 mm were prepared using ordinary Portland cement satisfying IS: 12269-1987 [22] and ASTM Type I [23]; and Portland pozzolana cement satisfying IS: 14891991 [24] and ASTM Type IP [25] with a water-cement ratio (w/c) of 0.50. The chemical composition of OPC and PPC determined by XRF (X-ray fluorescence) analysis are presented in Table 1. Locally available river sand was used as fine aggregate. The specific gravity of sand is 2.61 and it is conforming to grading zone II as per IS: 383-1970 [26] and as per ASTM C33/C33M-13 [27]. The coarse aggregate of quartzite origin

Table 1 Chemical composition of cement determined by X-ray fluorescence (XRF) analysis. Compound (wt.%)

CaO

Ordinary Portland cement (OPC) Portland pozzolana cement (PPC)

65.2 19.2

SiO2

Al2O3 Fe2O3 MgO Na2O K2O

SO3 LOI

5.2

2.4

3.4

0.3

0.62 1.5

1.4

64.7 20.52 4.2

3.4

1.55

0.35

1.31 1.6

1.2

F. Shaheen, B. Pradhan / Cement and Concrete Research 91 (2017) 73–86 Table 2 Chemical composition of coarse aggregate in wt%, determined by energy-dispersive X-ray (EDX) analysis. O

Si

Na

K

Ca

Al

42.8

29.2

0.1

15.7

0.2

8.7

is used and its chemical composition determined by energy-dispersive X-ray (EDX) analysis is presented in Table 2. The coarse aggregates of 20 mm MSA (maximum size aggregate) with a specific gravity of 2.64 and 10 mm MSA with a specific gravity of 2.63 were used in the proportion of 1.941:1 by mass of total coarse aggregate content to satisfy the overall grading requirement of coarse aggregate as per IS: 383-1970 [26] and ASTM C33/C33M-13 [27]. Tap water from the laboratory was used as mixing water in the preparation of concrete mixes. The water content in all the concrete mixes was kept at 195 kg/m3. The mix proportion of concrete for both OPC and PPC is presented in Table 3. The concrete mixes were prepared with varying dosages of chloride and sulfate salts as shown in Table 4. The concrete mixes were admixed with only NaCl for chloride contamination. Sodium chloride (NaCl) is normally present in seawater and in contaminated soil and groundwater in higher concentrations [28,12]. Further chloride ions may also enter into fresh concrete at the time of its preparation through chloride contaminated aggregates, mixing water and chloride bearing admixtures. In addition, as reported in literature, different researchers have used sodium chloride in different concentrations as the source of chloride ions in their investigations [4,12,14]. On the basis of these factors, in the present work, lower to very high concentrations of NaCl were used to perform electrochemical measurements. The concentrations of NaCl used in this investigation were 3%, 5% and 7% by weight of cement content. Further in seawater and contaminated groundwater, sulfate salts mostly in the form of sodium sulfate and magnesium sulfate are also present along with chloride salts. In the present work, the concrete mixes were admixed with NaCl plus Na2SO4 and NaCl plus MgSO4 for composite chloride-sulfate contaminations. The admixed concentrations of Na2SO4 and MgSO4 were 3%, 6% and 12% each by weight of cement content. The required quantities of these salts were dissolved in the mixing water during the preparation of concrete mixes. Cube specimens from different concrete mixes were prepared for obtaining concrete powder samples. After 24 h of casting, the cube specimens were demoulded and subjected to moist curing till the age of 28 days. Then the specimens were removed from the curing tank and kept in the laboratory exposure condition till the period of crushing. The steel specimens of 12 mm diameter and 70 mm length were used as steel reinforcement and these were made from commonly used Tempcore TMT (thermomechanically treated) ribbed steel bars. The chemical composition of steel reinforcement determined by energy-dispersive X-ray (EDX) analysis is presented in Table 5. The steel specimens were cleaned with wire brush to remove any surface scale. A hole of 3 mm diameter was drilled and threaded at one end of the steel specimen. Epoxy coating was applied on the surface of a steel specimen leaving an exposed portion of 5 mm length at the opposite face of the drilled end. The schematic diagram of a steel specimen is shown in Fig. 1. 2.2. Preparation of concrete powder aqueous solutions The cubes prepared for obtaining concrete powder were crushed at the age of 56 days from the day of preparation in the compression testing machine and were further crushed in an abrasion testing machine. The collected concrete powder was then sieved through a sieve of

75

Table 4 Concentration of chloride and sulfate salts admixed in concrete mixes (added by weight of cement). Concentration of admixed salts 3% NaCl 5% NaCl 7% NaCl 3% NaCl 5% NaCl 7% NaCl 3% NaCl 5% NaCl 7% NaCl 3% NaCl 5% NaCl 7% NaCl 3% NaCl 5% NaCl 7% NaCl 3% NaCl 5% NaCl 7% NaCl 3% NaCl

Abbreviation Group 3NC 5NC 7NC 3NC + 3NS 5NC + 3NS 7NC + 3NS 3NC + 6NS 5NC + 6NS 7NC + 6NS 3NC + 12NS 5NC + 12NS 7NC + 12NS 3NC + 3MS 5NC + 3MS 7NC + 3MS 3NC + 6MS 5NC + 6MS 7NC + 6MS 3NC + 12MS 5NC + 12MS 7NC + 12MS

+ 3% Na2SO4 + 3% Na2SO4 + 3% Na2SO4 + 6% Na2SO4 + 6% Na2SO4 + 6% Na2SO4 + 12% Na2SO4 + 12% Na2SO4 + 12% Na2SO4 + 3% MgSO4 + 3% MgSO4 + 3% MgSO4 + 6%MgSO4 + 6% MgSO4 + 6% MgSO4 + 12% MgSO4

5% NaCl + 12% MgSO4 7% NaCl + 12% MgSO4

Sodium chloride

Sodium chloride plus sodium sulfate

Sodium chloride plus magnesium sulfate

square opening with a size of 150 μm. The sieved concrete powder was stored in air tight plastic containers. For the purpose of preparing concrete powder aqueous solutions, the stored concrete powder was mixed with distilled water in 1:1 proportion by mass and then stirred for a period of half an hour followed by boiling for 15–20 min. After that, the concrete powder solution was allowed to settle and cool to room temperature. This solution was then filtered through Whatman no. 1 filter paper. The obtained filtered concrete powder aqueous solution nearly represents all the species in the vicinity of the rebar in the contaminated concrete as it is extracted from concrete powder, which is a mixture of cement hydrates, coarse aggregate, fine aggregate and admixed dosages of chloride and composite chloride-sulfate ions. The concrete powder aqueous solution was then used to perform a potentiodynamic polarization test on the prepared steel specimens. 2.3. Test techniques 2.3.1. Chemical analysis of concrete powder aqueous solution The chemical composition of the solution extracted from concrete powder depends on the type of binder (OPC and PPC) used and also on salt concentrations (NaCl, Na2SO4 and MgSO4), which are admixed at the time of preparation of the concrete mix. The extracted concrete powder aqueous solutions are chemically analyzed to determine sodium (Na+), potassium (K+), calcium (Ca++), chloride (Cl−) and sulfate + ++ (SO2− 4 ) ion concentrations. The concentrations of cations i.e. Na , Ca and K+ ions in concrete powder aqueous solutions were determined by using a flame photometer by preparing appropriate dilutions. Before measuring the concentrations of Na+, Ca++ and K+ ions, the flame photometer was calibrated with standard solutions of Na+ (10 ppm), Ca++ (100 ppm) and K+ (10 ppm). The procedure to determine Na+, Ca++ and K+ ions by using the flame photometer was followed as mentioned in the Standard Methods for the examination of water and wastewater Table 5 Chemical composition of steel in wt%, determined by energy-dispersive X-ray (EDX) analysis.

Table 3 Mixture proportion of concrete for OPC and PPC. Ingredients

Cement

Water

Fine aggregate

Coarse aggregate

Mn

Si

Ni

Cr

Cu

S

P

C

Fe

Quantity (kg/m3)

390

195

648

1152

1.2

0.6

0.2

0.2

0.1

0.2

0.1

b0.1

Balance

76

F. Shaheen, B. Pradhan / Cement and Concrete Research 91 (2017) 73–86

Section Y-Y

Epoxy coating Y 70 mm

12 mm

12 mm

3mm dia. threaded hole for connecting to corrosion Y monitoring instrument

5 mm exposed length

Fig. 1. Schematic diagram of a steel specimen.

published by the American Public Health Association (APHA) [29]. The free chloride (Cl−) ion concentration in the concrete powder aqueous solution was determined by argentometric titration. The procedure of argentometric titration to determine free Cl− ion concentration was followed as described in Standard Methods for the examination of water and wastewater published by APHA [29]. The turbidimetric method as mentioned in Standard Methods for the examination of water and wastewater published by APHA [29] was used to determine the sulfate ion (SO2− 4 ion) concentration in the concrete powder aqueous solution. 2.3.2. pH and conductivity measurement The pH and conductivity of concrete powder aqueous solutions contaminated with chloride ions and chloride-sulfate ions were determined. The digital pH meter with pH range 0–14 and digital conductivity meter were used to determine pH and conductivity respectively of the concrete powder aqueous solutions. 2.3.3. Potentiodynamic polarization test A potentiodynamic polarization test was carried out to evaluate the electrochemical behaviour of steel specimens in electrolytic concrete powder aqueous solutions. The electrochemical cell consists of a concrete powder aqueous solution, working electrode, auxiliary electrode and reference electrode. The steel specimen immersed in the aqueous solution was used as the working electrode. A saturated calomel electrode (SCE) and a platinum electrode were used as the reference electrode and auxiliary electrode respectively. The schematic representation of the test set-up is shown in Fig. 2. The potentiodynamic polarization test was performed on steel specimens using a corrosion monitoring instrument (make: ACM, Gill

AC serial no. 1542-sequencer). The guidelines for conducting a potentiodynamic polarization test are mentioned in ASTM G5-12 [30]. In the present work, the potentiodynamic polarization test was conducted on the steel specimen by applying the potential scan from 0 mV to 1500 mV with an offset from equilibrium potential at a sweep rate of 50 mV per minute. The anodic polarization curves obtained from the potentiodynamic polarization test were analyzed to identify the potential ranges of different zones of corrosion of the steel reinforcement. In this investigation, two replicate steel specimens were tested for a given concrete powder aqueous solution to observe the reproducibility. 2.3.4. X-ray diffraction (XRD) analysis X-ray diffraction was conducted on concrete powder samples to analyze the changes in the phase composition of hardened concrete. The analysis was performed on a Bruker D-8 Advance X-ray diffractometer with Cu Kα radiation (λ = 1.5405 Å). Diffraction patterns were generated on a vertical goniometer attached to a board focus X-ray tube with a copper target operating at 40 kV and 40 mÅ. The concrete powder was filled into the sample holder and was scanned from 5° to 55° (2θ) at a sampling interval of 0.05° 2θ per second. After obtaining the XRD patterns, the phase identification process involves the calculation of the most likely match score for a given phase based on peak intensity and peak position, when compared with the database of standard phases. 2.3.5. Fourier transform infrared (FTIR) spectroscopy Fourier transform infrared (FTIR) spectroscopy was conducted to identify different functional groups associated with different products formed in concrete. The FTIR spectrum of a concrete powder sample

Fig. 2. Schematic illustration of an electrochemical cell.

F. Shaheen, B. Pradhan / Cement and Concrete Research 91 (2017) 73–86

was collected in the transmission mode using a Thermo Fisher Scientific Nicolet iS10 FTIR spectrometer. Sample pallets were prepared by mixing 250 mg of potassium bromide (KBr) with 3 mg of concrete powder sample. Fifteen scans were recorded over a range of 4000 cm−1 to 400 cm−1. The background spectrum was collected at ambient atmosphere and then the spectra of the samples were collected.

3. Results and discussion 3.1. Chemical composition of chloride contaminated concrete powder aqueous solution The chemical composition of a concrete powder aqueous solution prepared from the control mix and chloride contaminated concrete mixes made with OPC and PPC are presented in Table 6. In addition in this table, measured pH and conductivity values of solutions are also presented. From Table 6, it is observed that the concentration of free chloride ion i.e. Cl− ion increased with an increase in admixed NaCl dosage in both OPC and PPC. Further chloride ion concentration was lower in OPC as compared to that in PPC in the presence of admixed NaCl. This may be attributed to higher chloride binding with C3A in OPC as compared to that in PPC as a result of higher C3A content in OPC. The concentration of sulfate ion was higher in OPC as compared to that in PPC as observed from Table 6. It may be noted that sulfate ion present in the concrete powder solution (made from the control mix and chloride contaminated concrete mix) is due to the presence of gypsum added in the manufacturing process to control the early setting and hardening behaviour of Portland cement. The higher concentration of sulfate ion in OPC as compared to that in PPC may be due to the preferential reaction of chloride ion than gypsum, with C3A in OPC. The conductivity of the concrete powder aqueous solution made from the control mix was higher in PPC as compared to that in OPC, which may be attributed to higher concentration of cations Na+, Ca++ and K+ in PPC than that in OPC as observed from Table 6. Further, the conductivity of a chloride contaminated concrete powder aqueous solution made from PPC was higher than that made from OPC, which may be due to the higher concentration of Cl− ion in PPC as compared that in OPC. In addition, the concentrations of cations i.e. Ca++ and K+ were also higher in PPC as compared to those made from OPC as observed from Table 6. The higher concentration of Ca++ and K+ ions in PPC may be attributed to an alteration in the extent of hydration of cement compounds due to the pozzolanic reaction in PPC concrete and also due to a change in reactivity of the alkali compounds. The concentration of Na+ ion was higher in OPC as compared to that in PPC, which may be attributed to the reaction of chloride ions with hydrated C3A to a greater extent in OPC in the presence of admixed NaCl, thereby resulting in a higher concentration of Na + ion in the concrete powder aqueous solution. The presence of sodium chloride decreased the pH of the concrete powder aqueous solution as observed from Table 6. However, there was no significant difference in the pH value of the concrete powder aqueous solution with different concentrations of admixed NaCl and also between OPC and PPC.

77

3.2. Chemical composition of chloride-sulfate contaminated concrete powder aqueous solution The measured ionic concentration along with pH and conductivity of concrete powder aqueous solutions prepared from OPC and PPC and contaminated with conjoint chloride-sulfate salts are presented in Table 7 for NaCl plus Na2SO4 and in Table 8 for NaCl plus MgSO4 contaminations. From these tables, it is observed that the concentration of Cl− ion increased with an increase in admixed NaCl concentration for both types of cement and at all admixed concentrations of Na2SO4 and MgSO4. The Cl− ion concentration was higher whereas SO2− ion con4 centration was lower in the concrete powder aqueous solution admixed with NaCl plus MgSO4 as compared to that admixed with NaCl plus Na2SO4 for both OPC and PPC as evident from Tables 7 and 8. This may be attributed to reduced bound chloride and increased sulfate binding with cement hydration products in the presence of MgSO4 in concrete. From Table 7, it is observed that in the conjoint presence of NaCl plus Na2SO4, the concentration of Cl− ion increased whereas that of SO2− 4 ion decreased with an increase in Na2SO4 dosage in OPC. This may be due to lower chloride binding as a result of the preferential reaction of sulfate ions than chloride ions, with C3A hydrates resulting in higher Cl− ion and lower SO24 − ion concentrations in the presence of NaCl plus Na2SO4. In PPC the opposite behaviour was observed i.e. the concentration of Cl− ion decreased whereas that of SO2− 4 ion increased with an increase in Na2SO4 dosage in the presence of NaCl plus Na2SO4. This may be due to the preferential reaction of chloride ions than sulfate ions, with C3A hydrates, thereby resulting in lower Cl− ion concentration and higher SO2− ion concentration in PPC in the conjoint presence of 4 NaCl plus Na2SO4. From Table 8, it is observed that the concentration of SO2− ion de4 creased whereas that of Cl− ion increased with an increase in MgSO4 dosage in both OPC and PPC in the conjoint presence of NaCl plus MgSO4. This may be attributed to higher sulfate binding as compared to chloride binding, with cement hydration products in the presence of MgSO4 in concrete. From Tables 7 and 8, it is observed that the conductivity of the concrete powder aqueous solution increased with an increase in admixed NaCl concentration for OPC and PPC at all levels of admixed Na2SO4 and MgSO4, which may be attributed to an increase in Cl− concentration. The conductivity of the concrete powder aqueous solution was higher in the conjoint presence of NaCl plus MgSO4 as compared to that in the conjoint presence of NaCl plus Na2SO4. This indicates that the presence of sulfate ion when associated with magnesium cation increases the conductivity in the presence of NaCl. Further, the conductivity of the solution made from OPC increased with an increase in admixed Na2SO4 and MgSO4 concentrations. In the case of PPC, the conductivity increased with an increase in admixed MgSO4 concentration whereas the conductivity mostly decreased with an increase in admixed Na2SO4 concentration. These variations in the conductivity of the concrete powder aqueous solution made from OPC and PPC in the conjoint presence of NaCl plus Na2SO4 and NaCl plus MgSO4 may be attributed to the variations in both Cl− ion and SO2− ion concentrations as observed 4 from Tables 7 and 8. The concentrations of cations i.e. Na+ and Ca++ in

Table 6 Chemical composition, pH and conductivity of a concrete powder aqueous solution prepared from control concrete and concrete mix admixed with NaCl for both OPC and PPC. Cement type

Admixed salts concentration (% by weight of cement)

pH

Conductivity (mS/cm)

Cl− (mmol/l)

SO2− 4 (mmol/l)

Na+ (mmol/l)

Ca++ (mmol/l)

K+ (mmol/l)

OPC PPC OPC

Control mix (without salt contamination) Control mix (without salt contamination) 3% NC 5% NC 7% NC 3% NC 5% NC 7% NC

13.4 13.2 12.39 12.36 12.34 12.36 12.34 12.32

5.5 6.6 22.2 27.3 33.8 24.4 28.2 34.6

2.8 5.6 67.7 138.2 169.2 73.3 146.7 172.1

0.22 0.3 1.5 1.7 2.2 1.3 1.5 1.7

12.8 17.1 155.2 174.6 227.4 113.7 138.9 184.2

25.2 88.6 62.4 94.7 126 514.6 791.7 1058.7

11.2 349.1 15.5 16.6 23.5 33.9 35.3 35.8

PPC

78

F. Shaheen, B. Pradhan / Cement and Concrete Research 91 (2017) 73–86

Table 7 Chemical composition, pH and conductivity of a concrete powder aqueous solution prepared from concrete mix admixed with NaCl and Na2SO4 for both types of cement. Cement type

Admixed salts concentration (% by weight of cement)

OPC

3% NaCl 3% NaCl 3% NaCl 3% NaCl 3% NaCl 3% NaCl 5% NaCl 5% NaCl 5% NaCl 5% NaCl 5% NaCl 5% NaCl 7% NaCl 7% NaCl 7% NaCl 7% NaCl 7% NaCl 7% NaCl

PPC

OPC

PPC

OPC

PPC

+ + + + + + + + + + + + + + + + + +

3% Na2SO4 6% Na2SO4 12% Na2SO4 3% Na2SO4 6% Na2SO4 12% Na2SO4 3% Na2SO4 6% Na2SO4 12% Na2SO4 3% Na2SO4 6% Na2SO4 12% Na2SO4 3% Na2SO4 6% Na2SO4 12% Na2SO4 3% Na2SO4 6% Na2SO4 12% Na2SO4

pH

Conductivity (mS/cm)

Cl− (mmol/l)

SO2− 4 (mmol/l)

Na+ (mmol/l)

Ca++ (mmol/l)

K+ (mmol/l)

12.66 12.6 12.57 12.48 12.64 12.72 12.62 12.58 12.55 12.44 12.61 12.68 12.61 12.56 12.44 12.18 12.58 12.65

38.4 41.6 47.4 40 38.7 39.7 39.7 42.3 49.2 42.81 40.2 40.4 40.3 45.7 50.2 47.4 42.4 41.6

42 54 62 71 65 56 124 133 135 121 116 113 144 158 164 138 135 118

51.6 40.7 29.4 40.1 44.8 59.4 30.5 19.1 18.7 23.4 33.2 44.8 13.7 7.0 5.8 13.2 23.4 40.6

224.4 233.7 242.2 150.4 194.8 221.7 230.7 284.6 285.9 159.1 219.0 269.6 298.0 349.1 455.2 192.2 238.3 314.4

130.6 134.4 147.2 1120.3 872.4 717.2 160.6 176.2 196.9 1033.0 627.4 504.5 188.4 199.1 235.5 908.2 593.2 492.2

9.0 9.9 10.1 31.5 28.6 25.6 11.1 11.1 11.3 38.8 37.2 36.2 12.0 12.3 12.3 48.0 42.2 40.5

the concrete powder aqueous solution varied significantly with concentrations of admixed NaCl, Na2SO4 and MgSO4. However, the variation in concentration of K+ ion with an admixed concentration of these salts was not significant. Further the concentrations of Ca++ and K+ ions in the concrete powder aqueous solution made from PPC were higher than that in the solution made from OPC in the conjoint presence of NaCl plus Na2SO4 and NaCl plus MgSO4 as observed from Tables 7 and 8. The higher concentration of Ca++ and K+ ions in PPC may be due to a change in the extent of the hydration reaction due to the pozzolanic activity and also as a result of a change in reactivity of the alkali compounds. The concentration of Na+ ion was higher in the concrete powder aqueous solution contaminated with NaCl plus Na2SO4 in OPC than that in PPC, whereas in NaCl plus MgSO4 contamination, the opposite behaviour was observed, i.e. the Na+ ion concentration was higher in PPC as compared to that in OPC. In NaCl plus Na2SO4 contamination, the higher concentration of Na+ ion in OPC than that in PPC may be attributed to the reaction of sodium sulfate with calcium hydroxide to a greater extent, thereby resulting in a higher concentration of Na+ ion. In PPC, the dominant effect of admixed NaCl has resulted in a higher concentration of Na+ ion in the presence of MgSO4 as compared to OPC. From Tables 7 and 8, it is observed that the concrete powder aqueous solution contaminated with NaCl plus Na2SO4 showed a higher pH as compared to that contaminated with NaCl plus MgSO4 for both OPC and PPC. The higher pH in the presence of sodium sulfate may be

attributed to the formation of sodium hydroxide (as a result of the reaction between calcium hydroxide and sodium sulfate), which increases the pH value. In the presence of magnesium sulfate, magnesium hydroxide that is formed due to the reaction between calcium hydroxide and magnesium sulfate has very low solubility and low pH value [3] that has resulted in a reduction in pH value of the concrete powder aqueous solution contaminated with NaCl plus MgSO4. 3.3. Different zones of corrosion The anodic polarization curves for steel in electrolytic concrete powder aqueous solutions made from different concrete mixes have been obtained by conducting a potentiodynamic polarization test. The anodic polarization curves of steel in concrete powder aqueous solutions made from OPC and PPC at w/c ratio of 0.5 for the control mix are shown in Fig. 3. In these curves, different zones of corrosion of steel namely the active zone, passive zone and transpassivity zone are shown. In the presence of chloride ions, the transpassivity zone is termed as pitting zone. From Fig. 3, it is observed that the range of the passive zone is higher in OPC as compared to that in PPC, which may be attributed to more alkalinity due to the availability of more amounts of calcium hydroxide in the OPC concrete. The corrosion potential (Ecorr), potential at the point of transition from active to passive state i.e. act/pass zone boundary potential and potential at the point of transition from passive to

Table 8 Chemical composition, pH and conductivity of a concrete powder aqueous solution prepared from concrete mix admixed with NaCl and MgSO4 for both types of cement. Cement type

Admixed salts concentration (% by weight of cement)

pH

Conductivity (mS/cm)

Cl− (mmol/l)

SO2− 4 (mmol/l)

Na+ (mmol/l)

Ca++ (mmol/l)

K+ (mmol/l)

OPC

3% NaCl 3% NaCl 3% NaCl 3% NaCl 3% NaCl 3% NaCl 5% NaCl 5% NaCl 5% NaCl 5% NaCl 5% NaCl 5% NaCl 7% NaCl 7% NaCl 7% NaCl 7% NaCl 7% NaCl 7% NaCl

12.36 12.24 12.01 12.27 12.18 12.04 12.32 12.21 11.97 12.23 12.15 11.84 12.28 12.17 11.78 12.21 12.1 11.33

41.4 44.2 48.5 44.9 46.6 51.3 43.6 45.1 50.4 46.2 49.7 51.8 47.8 48.3 52.6 48.7 50.8 54.5

73 76 85 79 85 90 141 147 158 152 161 169 164 175 178 181 183 186

10.26 3.68 2.77 8.2 3.86 1.6 5.13 3.49 1.86 4.25 3.37 1.36 4.17 2.86 1.6 4.01 2.81 0.43

103.5 110.2 135.4 104.8 113.7 146.3 193.5 193.5 198.9 194.8 215.4 221.8 204.4 226.1 247.8 210.0 237.0 243.5

132.1 171.2 172.9 859.3 658.6 592.9 166.9 176.2 194.4 814.2 609.6 492.2 199.2 207.0 221.1 754.0 509.0 404.2

20.7 20.2 19.6 30.7 31.2 30.8 15.9 15.5 14.3 33.3 35.0 35.1 13.6 12.7 12.3 35.4 37.9 38.3

PPC

OPC

PPC

OPC

PPC

+ + + + + + + + + + + + + + + + + +

3% MgSO4 6% MgSO4 12% MgSO4 3% MgSO4 6% MgSO4 12% MgSO4 3% MgSO4 6% MgSO4 12% MgSO4 3% MgSO4 6% MgSO4 12% MgSO4 3% MgSO4 6% MgSO4 12% MgSO4 3% MgSO4 6% MgSO4 12% MgSO4

F. Shaheen, B. Pradhan / Cement and Concrete Research 91 (2017) 73–86

79

Fig. 3. Anodic polarization curve showing zones of corrosion obtained for steel in the concrete powder aqueous solution made from OPC and PPC concrete (control mix).

transpassive state i.e. pass/transpass zone boundary potential are shown in Fig. 3. In the presence of chloride ion, the potential at the point of transition from passive to pitting state is termed as breakdown potential or pitting potential. 3.3.1. Effect of chloride contamination on zones of corrosion The anodic polarization curves of steel in concrete powder aqueous solutions contaminated with 3%, 5% and 7% NaCl concentrations are shown in Fig. 4 for both OPC and PPC. From these obtained polarization curves, corrosion potential (Ecorr), act/pass boundary potential and pass/pitt boundary potential (pitting potential) were obtained and are presented in Table 9. In addition, the potential values for the control mix (i.e. without any admixed salt) are also presented in this table. As already stated, two replicate steel specimens were tested for a given concrete powder aqueous solution to observe the reproducibility. From the obtained polarization curves of replicates, it is observed that, there is not much difference between the profiles of polarization curve of both the replicates. The boundary potential values presented in

Fig. 4. Anodic polarization curves for steel in concrete powder aqueous solutions prepared from OPC and PPC concrete admixed with 3% NaCl, 5% NaCl and 7% NaCl (added by weight of cement).

Table 9 are the average value of two replicate steel specimens for a given concrete powder aqueous solution. From Table 9, it is observed that the corrosion potential (Ecorr) became more negative with an increase in admixed NaCl concentration from 3% to 7% for both OPC and PPC. In addition, the range of a passive zone (calculated from the difference of breakdown potential and act/ pass boundary potential) decreased with an increase in admixed chloride content in both OPC and PPC. The breakdown potential (pitting potential) became more negative with an increase in admixed NaCl dosage, which is attributed to an increase in Cl− ion concentration in the aqueous solution as observed from Table 6. While analyzing the effect of binder type on pitting potential, it is observed that the pitting potential of steel in PPC is more negative than that in OPC. This is attributed to higher Cl− ion concentration in PPC as compared to that in OPC. Further, the range of the passive zone of a steel reinforcement was less in PPC than that in OPC at all levels of chloride contamination. A typical plot showing the variations of boundary potentials with Cl− ion concentration for OPC is presented in Fig. 5. From this figure, it is observed that the range of the passive zone decreased with an increase in Cl− ion concentration. The XRD patterns obtained from concrete powder samples contaminated with 3%, 5% and 7% NaCl concentrations are shown in Fig. 6a, b and c respectively, for both OPC and PPC. The XRD patterns in Fig. 6(a–c) indicate the peaks of gypsum (G), ettringite (E), calcium hydroxide (CH), calcium carbonate (CC), quartz (Q), thaumasite (T) and calcium chloroaluminate (CCA), which are formed in concrete. The calcium chloroaluminate peak was found at 11.2° 2θ and 23° 2θ. The peak intensity of CCA was more in OPC as compared to that in PPC at all dosages of admixed NaCl, as observed from Fig. 6(a–c), which is attributed to more chloride binding with hydrated C3A in OPC than that in PPC. It may be noted that the Cl− ion concentration was lower in OPC as compared to that in PPC as observed from Table 6. Therefore, the steel reinforcement in a chloride contaminated concrete powder aqueous solution made from OPC showed a higher range of passivity as compared to that made from PPC. Gypsum peaks were identified at 32.1° 2θ and 50.5° 2θ and the presence of this primary gypsum may be attributed to its addition to a Portland cement clinker during manufacturing, to prevent flash setting of cement. Gypsum when combined with C3A, forms ettringite. The peaks of ettringite were found at 8.8° 2θ, 15.75° 2θ, 25.6° 2θ, and 27.5° 2θ. It is observed that the peaks of ettringite in OPC

80

F. Shaheen, B. Pradhan / Cement and Concrete Research 91 (2017) 73–86

Table 9 Corrosion potential and boundary potentials of steel reinforcement in a concrete powder aqueous solution prepared from control concrete and concrete admixed with NaCl for OPC and PPC. Cement type

Admixed salts concentration (% by weight of cement)

Corrosion potential (mV/SCE)

Act/pass boundary potential (mV/SCE)

Pass/transpass boundary potential (for control mix) and Pass/pitt boundary potential (pitting potential) (mV/SCE)

OPC

Control mix (without salt contamination) Control mix (without salt contamination) 3% NaCl 5% NaCl 7% NaCl 3% NaCl 5% NaCl 7% NaCl

−346

−216

523

−280

−211

430

−326 −349 −393 −338 −360 −363

−236 −283 −342 −267 −304 −334

−34 −105 −178 −83 −145 −186

PPC OPC

PPC

concrete are less intense than that in PPC concrete at all levels of admixed NaCl. This is due to the preferential reaction of chloride ions with C3A to form more amounts of calcium chloroaluminate in OPC, resulting in lesser availability of C3A to react with gypsum to form a lower amount of ettringite. The OPC concrete showed the precipitation of calcium hydroxide through well-defined peaks at 18.08° 2θ, 34.1° 2θ, and 36.5° 2θ, whereas in PPC, the peaks of calcium hydroxide almost disappeared at 18.08° 2θ and 34.1° 2θ. The peaks of calcium hydroxide in PPC was found only at 36.5° 2θ with reduced intensity as compared to that in OPC as evident from Fig. 6(a–c). The reduced peaks of calcium hydroxide in PPC are due to its consumption in the pozzolanic reaction. Further peaks of quartz were found at 20.85° 2θ, 26.65° 2θ, 39.45° 2θ, 42.5° 2θ and 50.1° 2θ, as shown in Fig. 6(a–c), which is mostly due to the presence of aggregates in the concrete. Similarly, the peaks of thaumasite and calcium carbonate were found at 27.9° 2θ and 29.4° 2θ respectively. As stated earlier, for the purpose of identifying different functional groups associated with different products formed in concrete, Fourier transform infrared (FTIR) spectroscopy was conducted. FTIR spectra of concrete powder samples contaminated with 3%, 5% and 7% NaCl concentrations are shown in Fig. 7a–c respectively, for both OPC and PPC. From Fig. 7(a–c), it is observed that the bands ranging from 3445 cm−1 to 3452 cm−1 cm and 1642 cm−1 to 1649 cm−1 for OPC and those at 3394 cm−1 and ranging from 1594 cm−1 to 1598 cm−1 for PPC at all dosages of admixed chloride are caused by stretching and bending bands of the O\\H group particularly contributed from gypsum [31–33]. Further, the infrared spectra for OPC indicate an additional stretching band of the O\\H group at 3663 cm−1 that is contributed by the presence of more amounts of calcium hydroxide in the OPC concrete [31,33]. The XRD patterns as shown in Fig. 6(a–c) indicate the presence of calcium carbonate in concrete at 29.4° 2θ, which is substantiated by infrared spectra showing CO23 − bands ranging from 1424 cm− 1 to 1432 cm−1 for OPC and ranging from 1388 cm−1 to 1399 cm−1 for PPC

Fig. 5. Potential vs. Cl− ion concentration in OPC concrete admixed with 3%, 5% and 7% NaCl (added by weight of cement).

[33–35] as shown in Fig. 7(a–c). The bands ranging from 996 cm−1 to 1004 cm−1 and 776 cm−1 to 778 cm−1 for OPC and at 940 cm−1 and ranging from 725 cm−1 to 730 cm−1 for PPC represent ν1 and ν4 vibrations of SO2− bands contributed by the presence of gypsum 4 [31–33]. The formation of ettringite as indicated by XRD patterns was corroborated by the presence of Al\\O peak at 872 cm−1 for OPC and 824 cm−1 for PPC [33,35] as shown in Fig. 7(a–c) at all dosages of admixed chloride. 3.3.2. Effect of conjoint chloride-sulfate contamination on zones of corrosion The anodic polarization curves of steel in concrete powder aqueous solutions contaminated with all concentrations (shown in Table 4) of NaCl plus Na2SO4 and NaCl plus MgSO4 were obtained and some typical anodic polarization curves of steel in concrete powder aqueous solutions contaminated with 3%NaCl + 3%Na2SO4, 5%NaCl + 6%Na2SO4, and 7%NaCl + 12%Na2SO4 are shown in Figs. 8 and 9 for OPC and PPC respectively. Similarly Figs. 10 and 11 show the typical anodic polarization curves obtained from OPC and PPC respectively in concrete powder aqueous solutions contaminated with 3%NaCl + 3%MgSO4, 5%NaCl + 6%MgSO4, and 7%NaCl + 12%MgSO4. From the anodic polarization curves, the values of corrosion potential (Ecorr), act/pass boundary potential and pass/pitt boundary potential (pitting potential) were obtained and are presented in Table 10 for NaCl plus Na2SO4 and NaCl plus MgSO4 contaminations. The boundary potential values presented in these tables are the average value of two replicate steel specimens for a given concrete powder aqueous solution. 3.3.2.1. Effect of conjoint NaCl plus Na2SO4 contamination on passivity of steel. From the boundary potential values presented in Table 10, it is observed that the range of the passive zone decreased with an increase in admixed NaCl concentration irrespective of admixed Na2SO4 concentration for both OPC and PPC. However the change in range of the active zone with admixed NaCl concentration is not systematic in the presence of sulfate ions for both types of cement. Further from Table 10, it is observed that the range of the passive zone, which is calculated from the difference of pass/pitt boundary potential and act/pass boundary potential is higher in OPC as compared to that in PPC at 3% Na2SO4 concentration whereas PPC showed higher passivity range at 6% and 12% Na2SO4 concentrations for all dosages of admixed NaCl. From the results presented in Table 7, it is observed that for 3% Na2SO4, the Cl− ion concentration was significantly higher in PPC as compared to that in OPC at 3% NaCl, whereas there was a minor difference in Cl− ion concentration between OPC and PPC at 5% and 7% NaCl concentrations. However at 6% and 12% Na2SO4 concentrations, Cl− ion concentration was mostly higher in the concrete powder aqueous solution made from OPC than that in PPC at all dosages of NaCl. This indicates that the higher passivity range in OPC at a lower Na2SO4 concentration (i.e. 3%) and that in PPC at higher Na2SO4 concentrations (i.e. 6% and 12%) may be attributed to higher chloride binding resulting in a lower Cl− ion concentration at corresponding Na2SO4 concentrations in the presence of NaCl.

F. Shaheen, B. Pradhan / Cement and Concrete Research 91 (2017) 73–86

81

Fig. 7. a: FTIR spectra of OPC and PPC concrete admixed with 3% NaCl (added by weight of cement). b: FTIR spectra of OPC and PPC concrete admixed with 5% NaCl (added by weight of cement). c: FTIR spectra of OPC and PPC concrete admixed with 7% NaCl (added by weight of cement).

Fig. 6. a: XRD patterns of OPC and PPC concrete admixed with 3% NaCl (added by weight of cement). b: XRD patterns of OPC and PPC concrete admixed with 5% NaCl (added by weight of cement). c: XRD patterns of OPC and PPC concrete admixed with 7% NaCl (added by weight of cement).

From Table 10, it is observed that the passivity breakdown potential or pitting potential of steel in aqueous solutions made from OPC concrete decreased (i.e. became more negative) with the increase in Na2SO4 concentration at all dosages of NaCl. This may be attributed to

82

F. Shaheen, B. Pradhan / Cement and Concrete Research 91 (2017) 73–86

Fig. 8. Typical anodic polarization curves obtained for steel in concrete powder aqueous solutions prepared from OPC concrete admixed with different concentrations of NaCl and Na2SO4 (added by weight of cement).

Fig. 10. Typical anodic polarization curves obtained for steel in concrete powder aqueous solutions prepared from OPC concrete admixed with different concentrations of NaCl and MgSO4 (added by weight of cement).

an increase in Cl− ion concentration and a decrease in pH of the solution (Table 7) with an increase in Na2SO4 concentration in the presence of NaCl. For steel in the concrete powder aqueous solution made from PPC, the opposite behaviour was observed, i.e. the pitting potential increased (i.e. became less negative) with the increase in concentration of Na2SO4 at all dosages of NaCl as observed from Table 10. The increase in pitting potential for PPC may be due to a decrease in Cl− ion concentration and an increase in pH of the solution with an increase in Na2SO4 concentration in the presence of NaCl as observed from Table 7. This indicates that the breakdown/pitting potential is a function of Cl− ion concentration in the solution, which in turn depends on the concentration of Na2SO4 and is also influenced by a little change in pH value of the solution. The minimum (more negative) pitting potential was observed in the presence of higher Cl− ion concentration and lower SO2− ion con4 centration (Tables 7 and 10) for both OPC and PPC in the concrete powder aqueous solution contaminated with both NaCl and Na2SO4. The reasons for variations in Cl− ion and SO2− ion concentrations for OPC 4 and PPC are already stated in Section 3.2. The XRD patterns of concrete powder samples contaminated with different concentrations of NaCl plus Na2SO4 were obtained and that of a typical chloride-sulfate contamination is presented in Fig. 12 for 3%NaCl + 3%Na2SO4 for both OPC and PPC. It is to be noted that, the compounds formed in concrete in the presence of composite chloride-

sulfate ions are identified in the XRD patterns at 2θ values the same as that in the case of chloride contaminated concrete. In Fig. 12, the XRD patterns show the peaks of gypsum and ettringite for both OPC and PPC. The formation of gypsum and ettringite is also corroborated from FTIR spectra shown in Fig. 13 for the above mentioned typical combination of NaCl plus Na2SO4 in both OPC and PPC. For OPC, the bands at 3447 cm−1 and 1641 cm−1 as shown in the FTIR spectra (Fig. 13) correspond to stretching and bending modes of the O\\H group contributed from gypsum. The bands at 1011 cm−1 and ranging from 649 cm−1 to 778 cm−1 are due to ν1 and ν4 vibrations of SO2− associated with gyp4 sum. The presence of ettringite is confirmed by the stretching bands of Al\\O at 874 cm−1 and 533 cm−1 in OPC concrete. For PPC concrete, FTIR spectra shown in Fig. 13 indicate the stretching and bending modes of the O\\H group at 3447 cm−1 and 1638 cm−1, which are contributed from gypsum. Further, the bands at 1000 cm−1 and ranging 601 cm−1 to 776 cm−1 due to ν1 and ν4 vibrations of SO2− are also as4 sociated with gypsum. The Al\\O stretching bands associated with ettringite are found at 874 cm− 1 and 534 cm−1 for PPC as shown in Fig. 13. Similar to 3% NaCl plus 3% Na2SO4 contamination, functional groups associated with different products as shown by FTIR spectra were also obtained for other levels of NaCl plus Na2SO4 contamination. The peak intensity of calcium chloroaluminate was more in PPC than that in OPC for 3% NaCl plus 3% Na2SO4 contamination as observed from Fig. 12. In PPC the preferential reaction of chloride ions with C3A has led

Fig. 9. Typical anodic polarization curves obtained for steel in concrete powder aqueous solutions prepared from PPC concrete admixed with different concentrations of NaCl and Na2SO4 (added by weight of cement).

Fig. 11. Typical anodic polarization curves obtained for steel in concrete powder aqueous solutions prepared from PPC concrete admixed with different concentrations of NaCl and MgSO4 (added by weight of cement).

F. Shaheen, B. Pradhan / Cement and Concrete Research 91 (2017) 73–86

83

Table 10 Corrosion potential and boundary potentials of steel reinforcement in a concrete powder aqueous solution prepared from concrete mix admixed with NaCl plus Na2SO4 and NaCl plus MgSO4 for both types of cement. Chloride-sulfate contamination

Cement type

Admixed salts concentration (% by weight of cement)

Sodium chloride plus sodium sulfate

OPC

3% NaCl 3% NaCl 3% NaCl 3% NaCl 3% NaCl 3% NaCl 5% NaCl 5% NaCl 5% NaCl 5% NaCl 5% NaCl 5% NaCl 7% NaCl 7% NaCl 7% NaCl 7% NaCl 7% NaCl 7% NaCl 3% NaCl 3% NaCl 3% NaCl 3% NaCl 3% NaCl 3% NaCl 5% NaCl 5% NaCl 5% NaCl 5% NaCl 5% NaCl 5% NaCl 7% NaCl 7% NaCl 7% NaCl 7% NaCl 7% NaCl 7% NaCl

PPC

OPC

PPC

OPC

PPC

Sodium chloride plus magnesium sulfate

OPC

PPC

OPC

PPC

OPC

PPC

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +

3% Na2SO4 6% Na2SO4 12% Na2SO4 3% Na2SO4 6% Na2SO4 12% Na2SO4 3% Na2SO4 6% Na2SO4 12% Na2SO4 3% Na2SO4 6% Na2SO4 12% Na2SO4 3% Na2SO4 6% Na2SO4 12% Na2SO4 3% Na2SO4 6% Na2SO4 12% Na2SO4 3% MgSO4 6% MgSO4 12% MgSO4 3% MgSO4 6% MgSO4 12% MgSO4 3% MgSO4 6% MgSO4 12% MgSO4 3% MgSO4 6% MgSO4 12% MgSO4 3% MgSO4 6% MgSO4 12% MgSO4 3% MgSO4 6% MgSO4 12% MgSO4

Corrosion potential (mV/SCE)

Act/pass boundary potential (mV/SCE)

Pass/pitt boundary potential (pitting potential) (mV/SCE)

−322 −328 −356 −362 −318 −354 −344 −426 −409 −440 −315 −328 −333 −381 −346 −361 −356 −324 −204 −361 −396 −415 −266 −466 −320 −384 −396 −366 −285 −436 −286 −366 −436 −398 −418 −376

−202 −248 −305 −284 −265 −212 −306 −363 −379 −408 −289 −250 −258 −318 −290 −310 −302 −268 −172 −269 −338 −380 −213 −407 −212 −294 −325 −289 −236 −338 −172 −291 −348 −344 −360 −340

369 27 −80 −54 59 120 −34 −141 −197 −189 −33 18 −83 −167 −168 −152 −83 −42 16 −90 −179 −219 −76 −281 −36 −134 −181 −133 −90 −268 −9 −143 −230 −201 −229 −240

to higher chloride binding, thus resulting in the formation of more amounts calcium chloroaluminate and a similar observation was found for other levels of NaCl plus Na2SO4 contamination. The intensity of a calcium hydroxide peak in XRD patterns is less in PPC as compared to that in OPC at all levels of chloride-sulfate contaminations, which is attributed to the consumption of Ca(OH)2 in pozzolanic reaction in PPC concrete. Gypsum and ettringite peaks were more intense in OPC as compared to that in PPC as observed from Fig. 12 and similar results were found for other combinations of NaCl plus Na2SO4 concentrations.

The availability of a higher amount of Ca(OH)2 in OPC has resulted in the formation of more amounts of gypsum in the presence of sulfate ions, thus showing intense peaks of gypsum in OPC as compared to that in PPC. Further in OPC, the preferential reaction of sulfate ions with C3A has resulted in the formation of a higher amount of ettringite as compared to that in PPC in the conjoint presence of Na2SO4 and NaCl, thus showing intense peaks of ettringite in OPC. The reasons for the formation of lower amounts of gypsum and ettringite in PPC as compared to

Fig. 12. XRD patterns of OPC and PPC concrete admixed with 3% NaCl + 3% Na2SO4 (added by weight of cement).

Fig. 13. FTIR spectra of OPC and PPC concrete admixed with 3% NaCl + 3% Na2SO4 (added by weight of cement).

84

F. Shaheen, B. Pradhan / Cement and Concrete Research 91 (2017) 73–86

OPC in the conjoint presence of Na2SO4 and NaCl are: a) in the first stage, sodium sulfate reacts with calcium hydroxide liberated during the hydration reaction and forms calcium sulfate and sodium hydroxide. This stage is hindered in PPC due to less reserve of calcium hydroxide leading to less formation of gypsum, thus showing its less intense peaks in XRD patterns. In the second stage, gypsum produced (in the first stage) reacts with C3A to form expansive ettringite. The amount of ettringite formation is reduced in PPC due to a lesser availability of gypsum and C3A (diluted due to the replacement of a part of the Portland cement clinker by fly ash in PPC), thus showing less intense peaks of ettringite in PPC. The XRD patterns (Fig. 12) show the peaks of quartz, thaumasite and calcium carbonate for both OPC and PPC at all levels of chloride-sulfate contaminations. Thaumasite may be formed as a result of the reaction between C\\S\\H, calcium carbonate and calcium sulfate in the presence of water. It may also be formed through a woodfordite route due to the reaction between C\\S\\H, ettringite and calcium carbonate in the presence of water [36,37]. The presence of calcium carbonate is also substantiated by FTIR spectra showing CO23 − bands at 1427 cm− 1 for OPC and at 1424 cm− 1 for PPC as observed from Fig. 13. 3.3.2.2. Effect of conjoint NaCl plus MgSO4 contamination on passivity of steel. From the boundary potential values presented in Table 10 for NaCl plus MgSO4 contamination, it is observed that the passivity range decreased with an increase in NaCl concentration for both OPC and PPC at all levels of MgSO4 concentration. The variation in range of the active zone with varying concentrations of NaCl is not systematic in the presence of sulfate ions for both types of cement as observed from boundary potential values presented in Table 10. Further the range of the passive zone decreased with an increase in admixed MgSO4 concentration at all levels of admixed NaCl for both types of cement as observed from Table 10, which may be due to an increase in Cl− ion concentration with an increase in the dosage of MgSO4. The increase in Cl− ion concentration (presented in Table 8) with an increase in added magnesium sulfate concentration may be due to lower chloride binding as a result of the preferential reaction of sulfate ions as compared to chloride ions, with hydrated C3A in concrete. It is also observed that the range of passive zone of steel is higher in concrete powder aqueous solutions made from OPC as compared to that made from PPC at all levels of sodium chloride plus magnesium sulfate concentrations, which indicates that OPC maintains the passivity of reinforcing steel to a greater extent than PPC in the conjoint presence of NaCl and MgSO4. This is attributed to lower Cl− ion concentration observed in OPC as compared to that in PPC at all levels of admixed NaCl plus MgSO4 concentrations. A further decrease in the passivity range of steel in PPC in the conjoint presence of NaCl and MgSO4 can also be attributed to its susceptibility to Mg-oriented attack. Magnesium sulfate reacts with calcium hydroxide liberated in the hydration reaction of calcium silicates (C3S and C2S) to form calcium sulfate (gypsum) and magnesium hydroxide (brucite) [38]. In PPC due to less reserve of calcium hydroxide, which acts as the first defensive material to react with magnesium sulfate, the magnesium sulfate attack is therefore directed extensively towards C\\S\\H gel, ultimately forming more amounts of gypsum and non-cementitious magnesium silicate hydrate (M\\S\\H) [3]. Thus the passivity of steel in the concrete powder aqueous solution made from PPC in the presence of NaCl plus MgSO4 is decreased due to an increase in the conductivity of the concrete powder aqueous solution as a result of more Cl− ion concentration and formation of more amounts of non-cementitious magnesium silicate hydrate. From the pitting potential values presented in Table 10, it is observed that the pitting potential became more negative with an increase in MgSO4 concentration for OPC at all dosages of NaCl. This is attributed to a drop in pH of the solution and also due to an increase in Cl− ion concentration in the solution (Table 8) with an increase in MgSO4 concentration in the conjoint presence of NaCl and MgSO4. For PPC, the pitting potential increased (less negative) with an increase in MgSO4

concentration up to 6% followed by a decrease (more negative) at 12% MgSO4 concentration for 3% and 5% NaCl concentrations as observed from Table 10. However at 7% NaCl, the pitting potential decreased with an increase in MgSO4 concentration. Although, the variation in pitting potential with MgSO4 concentration is not systematic in PPC, the minimum pitting potential was observed at 12% MgSO4 concentration for all dosages of NaCl and this may be due to higher Cl− ion concentration and lower pH of the solution at 12% MgSO4 concentration as observed from Table 8. For both OPC and PPC, the XRD patterns of concrete samples contaminated with different concentrations of NaCl plus MgSO4 were obtained and a typical plot is shown in Fig. 14 for 3%NaCl + 3%MgSO4. The XRD patterns in Fig. 14 show the peaks of calcium hydroxide and magnesium hydroxide (at 33.9° 2θ) for both OPC and PPC. From this figure, it is observed that peak intensities of both calcium hydroxide and magnesium hydroxide (MH) are less in PPC as compared to that in OPC. The similar results were obtained for other levels of NaCl plus MgSO4 concentrations. The lower Ca(OH)2 reserve in PPC concrete is due to its consumption in pozzolanic reaction, thus showing its less intense peaks. The lower amount of Ca(OH)2 in PPC has resulted in the formation of a lower amount of magnesium hydroxide in the presence of magnesium sulfate, thus showing less intense peaks in PPC as compared to that in OPC. Further, the XRD patterns show the peaks of gypsum and ettringite for both types of cement at all levels of sodium chloride plus magnesium sulfate contaminations. The peak intensities of gypsum and ettringite are more in PPC as compared to that in OPC in the conjoint presence of NaCl and MgSO4 as observed from the XRD patterns. The formation of a higher amount of gypsum in PPC may be attributed to a Mg-oriented attack on C\\S\\H because of less reserves of calcium hydroxide in PPC, thus showing more intense peaks of gypsum. The more intense peaks of ettringite in PPC as compared to that in OPC may be due to the reaction of gypsum with hydrated aluminate phases in concrete to a greater extent. The formation of gypsum and ettringite is also confirmed from FTIR spectra shown in Fig. 15 for the typical combination of 3% NaCl plus 3% MgSO4 in both OPC and PPC. For OPC, the FTIR spectra shown in Fig. 15 indicate that the bands at 3446 cm−1 and 1641 cm−1 that correspond to stretching and bending modes of the O\\H group contributed from gypsum and the bands at 1008 cm−1 and ranging from 649 cm−1 to 778 cm−1 are due to ν1 and ν4 vibrations of SO2− associated with gypsum. The presence of ettringite is confirmed 4 by the stretching bands of Al\\O at 874 cm−1 and 535 cm−1 for OPC concrete. For PPC, the FTIR spectra show the stretching and bending modes of the O\\H group at 3456 cm−1 and 1639 cm−1 that correspond to gypsum. Further the bands at 965 cm−1 and ranging from 601 cm−1 to 773 cm−1 due to ν1 and ν4 vibrations of SO2− are also associated with 4

Fig. 14. XRD patterns of OPC and PPC concrete admixed with 3% NaCl + 3% MgSO4 (added by weight of cement).

F. Shaheen, B. Pradhan / Cement and Concrete Research 91 (2017) 73–86

85

Fig. 16. Potential vs. Cl− ion concentration in OPC concrete admixed with different dosages of NaCl with 6% Na2SO4 and 6% MgSO4 (added by weight of cement). Fig. 15. FTIR spectra of OPC and PPC concrete admixed with 3% NaCl + 3% MgSO4 (added by weight of cement).

gypsum. The Al\\O stretching bands associated with ettringite are found at 875 cm− 1 and 535 cm− 1 in PPC. Similarly, functional groups associated with different products as shown by FTIR spectra were also obtained for other concentrations of NaCl plus MgSO 4 contamination. The reaction of chloride ions with C3A has led to the formation of calcium chloroaluminate, as shown by peaks in both OPC and PPC as observed from Fig. 14. The XRD patterns indicate the formation of higher amount of calcium chloroaluminate in PPC in most of the cases as compared to OPC for NaCl plus MgSO4 contaminations. The higher free Cl− ion concentration in PPC than that in OPC for NaCl plus MgSO4 contamination (as observed from Table 8) may be attributed to the dominant effect of less physical binding of chloride ions with C\\S\\H, as C\\S\\H converts to non-cementitious magnesium silicate hydrate (M\\S\\H) in the presence of MgSO4 to a greater extent in PPC. The comparatively higher physical chloride binding in OPC has resulted in lower Cl− ion concentration, although there is less formation of calcium chloroaluminate as indicated by its lower intense peaks in XRD patterns. The peaks of quartz, calcium carbonate and thaumasite were also found in XRD patterns shown in Fig. 14 for both OPC and PPC. The presence of calcium carbonate is also substantiated by FTIR spectra shown in Fig. 15 −1 through the bands of CO2− for OPC and at 1478 cm−1 for 3 at 1424 cm PPC. 3.3.3. Comparison between chloride contamination and conjoint chloridesulfate contamination on passivity In both types of cement, the range of the passive zone of steel in concrete powder aqueous solutions contaminated with only sodium chloride is less as compared to that in concrete powder aqueous solutions contaminated with sodium chloride plus sodium sulfate as observed from Tables 9 and 10. This indicates that the presence of sodium sulfate in a chloride environment has mitigated the effect of chloride ions on reducing the passivity of reinforcing steel. The range of the passive zone of reinforcing steel is more in concrete powder aqueous solutions contaminated with only sodium chloride as compared to that in the conjoint presence of sodium chloride and magnesium sulfate as observed from Tables 9 and 10. This indicates that the presence of magnesium sulfate in a chloride environment has stimulated the effect of chloride ions on reducing the passivity of steel reinforcement. It may also be noted that a lower Cl− ion concentration was found in the concrete mixes contaminated with NaCl plus Na2SO4 whereas a higher Cl− ion concentration was found in the concrete mixes contaminated with NaCl plus MgSO4 as compared to the concrete mixes contaminated with only NaCl. On comparison of the effect of cation type (Na+ and Mg++) associated with sulfate ions, it is observed that the range of the passive zone of

steel is more in a NaCl plus Na2SO4 environment as compared to that in a NaCl plus MgSO4 environment for both types of cement as observed from boundary potential values presented in Table 10. This is also evident from a typical plot shown in Fig. 16, which is showing the variation in boundary potential values with Cl− ion concentration for Na2SO4 and MgSO4 contaminations. This implies that Mg-oriented sulfate attack is more aggressive in the presence of chloride ions in reducing the passivity of steel reinforcement as compared to Na-oriented sulfate attack. Further, the range of the passive zone of steel is mostly higher in PPC as compared to that in OPC in the conjoint presence of sodium chloride and sodium sulfate whereas OPC showed a higher range of the passive zone as compared to PPC in the conjoint presence of sodium chloride plus magnesium sulfate. This indicates OPC performed better against Mg-oriented sulfate attack whereas PPC performed better against Naoriented sulfate attack in the presence of chloride ions. 4. Conclusions Following conclusions have been drawn from the results of the present investigation. i. The concrete powder aqueous solution contaminated with NaCl plus Na2SO4 showed higher pH as compared to that contaminated with NaCl plus MgSO4 for both types of cement. Further, the conductivity of the concrete powder aqueous solution was lower in the conjoint presence of NaCl and Na2SO4 as compared to that in the conjoint presence of NaCl and MgSO4. ii. The passive zone decreased with an increase in admixed NaCl concentration in all the concrete mixes. iii. The passivity range of steel reinforcement in chloride contaminated concrete powder aqueous solutions is higher in OPC as compared to that in PPC. iv. The presence of sodium sulfate in a chloride environment has mitigated the effect of chloride ions whereas the presence of magnesium sulfate in a chloride environment has stimulated the effect of chloride ions, on reducing the passivity of steel reinforcement. v. The steel reinforcement in a concrete powder aqueous solution made from PPC mostly showed a higher range of the passive zone as compared to OPC in the conjoint presence of NaCl and Na2SO4. vi. In the concrete powder aqueous solutions contaminated with NaCl and MgSO4, the range of the passive zone of the steel reinforcement was less in PPC as compared to that in OPC. vii. The XRD patterns indicating the formation of different compounds such as gypsum, calcium carbonate and ettringite in concrete were further substantiated with the FTIR spectra showing the functional groups such O\\H, SO24 −, CO23 − and Al\\O

86

F. Shaheen, B. Pradhan / Cement and Concrete Research 91 (2017) 73–86

associated with these compounds. viii. For both NaCl plus Na2SO4 and NaCl plus MgSO4 contaminations, the minimum (more negative) pitting potential was observed at a higher Cl− ion concentration and lower SO2− 4 ion concentration for both OPC and PPC. ix. On evaluating the effect of cement type on the passivity of steel reinforcement, it is observed that OPC performed better against magnesium-oriented sulfate attack whereas PPC performed better against sodium-oriented sulfate attack in a chloride environment.

Acknowledgement The authors express their gratitude to Department of Science and Technology (DST), Government of India (SR/FTP/ETA-0072/2009) for funding the corrosion monitoring instrument (used in this study) through a project under the Fast Track scheme for Young Scientists. References [1] G.P. Gu, J.J. Beaudoin, V.S. Ramachandran, Techniques for corrosion investigation in reinforced concrete, in: G.V.S. Ramachandran, J.J. Beaudoin (Eds.), Handbook of Analytical Techniques in Concrete Science and Technology, William Andrew/Noyes Publications 2006, pp. 441–504. [2] Performance of different types of Pakistani cements exposed to aggressive environments, in: A.R. Khan, Alexander, et al., (Eds.), Concrete Repair, Rehabilitation and Retrofitting II, Taylor & Francis Group, London 2009, pp. 153–158. [3] O.S.B. Al-Amoudi, Attack on plain and blended cements exposed to aggressive sulfate environments, Cem. Concr. Compos. 24 (2002) 305–316. [4] B. Pradhan, B. Bhattacharjee, Role of steel and cement type on chloride-induced corrosion in concrete, ACI Mater. J. 104 (2007) 612–619. [5] A.M. Neville, J.J. Brooks, Concrete Technology, Pearson Education, 2004 Fourth Indian reprint. [6] O.S.B. Al-Amoudi, Influence of chloride ions on sulphate deterioration in plain and blended cements, Mag. Concr. Res. 46 (1994) 113–123. [7] O.S.B. Al-Amoudi, M. Maslehuddin, The effect of chloride and sulfate ions on reinforcement corrosion, Cem. Concr. Res. 23 (1993) 139–146. [8] B. Pradhan, B. Bhattacharjee, Rebar corrosion in chloride environment, Constr. Build. Mater. 25 (2011) 2565–2575. [9] H.A.F. Dehwah, M. Maslehuddin, S.A. Austin, Effect of sulphate ions and associated cation type on the pore solution chemistry in chloride-contaminated plain and blended cements, Cem. Concr. Compos. 25 (2003) 513–525. [10] N. Asrar, A.U. Malik, S. Ahmad, F.S. Mujahid, Corrosion protection performance of microsilica added concretes in NaCl and seawater environments, Constr. Build. Mater. 13 (1999) 213–219. [11] K.M.A. Hossain, Resistance of scoria-based blended cement concrete against deterioration and corrosion in mixed sulfate environment, J. Mater. Civ. Eng. ASCE 21 (2009) 299–308. [12] J. Zuquan, S. Wei, Z. Yunsheng, J. Jinyang, L. Jianzhong, Interaction between sulfate and chloride solution attack of concretes with and without fly ash, Cem. Concr. Res. 37 (2007) 1223–1232. [13] N.R. Jarrah, O.S.B. Al-Amoudi, M. Maslehuddin, O.A. Ashiru, A.I. Al-Mana, Electrochemical behaviour of steel in plain and blended cement concretes in sulphate and/or chloride environments, Constr. Build. Mater. 9 (1995) 97–103. [14] H.A.F. Dehwah, M. Maslehuddin, S.A. Austin, Long-term effect of sulfate ions and associated cation type on chloride-induced reinforcement corrosion in Portland cement concretes, Cem. Concr. Compos. 24 (2002) 17–25.

[15] A. Alhozaimy, R.R. Hussain, A. Al-Negheimish, R. Al-Zaid, D.D.N. Singh, Effect of simulated concrete pore solution chemistry, chloride ions, and temperature on passive layer formed on steel reinforcement, ACI Mater. J. 111 (2014) 411–422. [16] F. Zhang, J. Pan, C. Lin, Localized corrosion behaviour of reinforcement steel in simulated concrete pore solution, Corros. Sci. 51 (2009) 2130–2138. [17] P. Ghods, O.B. Isgor, G. Mcrae, T. Miller, The effect of concrete pore solution composition on the quality of passive oxide films on black steel reinforcement, Cem. Concr. Compos. 31 (2009) 2–11. [18] W. Chen, R. Du, C. Ye, Y. Zhu, C. Lin, Study on the behaviour of reinforcing steel in simulated concrete pore solutions using in situ Raman spectroscopy assisted by electrochemical techniques, Electrochim. Acta 55 (2010) 5677–5682. [19] V. Padilla, A. Alfantazi, Corrosion film breakdown of galvanized steel in sulphatechloride solutions, Constr. Build. Mater. 66 (2014) 447–457. [20] B. Pradhan, B. Bhattacharjee, Corrosion zones of rebar in chloride contaminated concrete through potentiostatic study in concrete powder solution extracts, Corros. Sci. 49 (2007) 3935–3952. [21] E.E.A. Aal, S.A. Wanees, A. Diab, S.M.A. Haleem, Environmental factors affecting the corrosion behavior of reinforcing steel III. Measurement of pitting corrosion currents of steel in Ca(OH)2 solutions under natural corrosion conditions, Corros. Sci. 51 (2009) 1611–1618. [22] IS 12269–1987, Specifications for 53 Grade Ordinary Portland Cement, Bureau of Indian Standards, New Delhi, reaffirmed 2004. [23] ASTM C150/C150M-12, Standard Specification for Portland Cement, ASTM International, West Conshohocken, PA, 2012. [24] IS 1489–1991, Specifications for Portland Pozzolana Cement, Part 1: Fly Ash Based, Bureau of Indian Standards, New Delhi, reaffirmed 2005. [25] ASTM C595/C595M-12, Standard Specification for Blended Hydraulic Cements, ASTM International, West Conshohocken, PA, 2012. [26] IS: 383–1970, Specification for Coarse and Fine Aggregate from Natural Sources for Concrete, Bureau of Indian Standard, New Delhi, reaffirmed 2002. [27] ASTM C33/C33M-13, Standard Specification for Concrete Aggregates, ASTM International, West Conshohocken, PA, 2013. [28] K. Sotiriadis, E. Nikolopoulou, S. Tsivilis, Sulfate resistance of limestone cement concrete exposed to combined chloride and sulfate environment at low temperature, Cem. Concr. Compos. 34 (2012) 903–910. [29] Standard Methods for the Examination of Water and Wastewater, 21st edition American Public Health Association (APHA), Washington, 2005. [30] ASTM G5-13, Standard Reference Test Method for Making Potentiodynamic Anodic Polarization Measurements, ASTM International, West Conshohocken, PA, 2013. [31] R. Ylmen, U. Jaglid, B. Steenari, I. Panas, Early hydration and setting of Portland cement monitored by IR, SEM, and vicat technique, Cem. Concr. Res. 39 (2009) 433–439. [32] S. Wang, X. Peng, L. Tang, L. Zeng, C. Lan, Influence of inorganic admixtures on the 11 Å-tobermorite formation prepared from steel slags: XRD and FTIR analysis, Constr. Build. Mater. 60 (2014) 42–47. [33] T.L. Hughes, C.M. Methven, T.G.J. Jones, S.E. Pelham, P. Fletcher, C. Hall, Determining cement composition by Fourier transform infrared, Adv. Cem. Based Mater. 2 (1995) 91–104. [34] M.Y.A. Mollah, W. Yu, R. Schennach, D.L. Cocke, A Fourier transform infrared spectroscopic investigation of the early hydration of Portland cement and the influence of sodium lignosulfonate, 30 (2000) 267–273. [35] D.A. Silva, H.R. Roman, P.J.P. Gleize, Evidences of chemical interaction between EVA and hydrating Portland cement, Cem. Concr. Res. 32 (2002) 1383–1390. [36] A.H.M. Abdalkader, C.J. Lynsdale, J.C. Cripps, The effect of chloride on cement mortar subjected to sulfate exposure at low temperature, Constr. Build. Mater. 78 (2015) 102–111. [37] A.M. Ramezanianpour, R.D. Hooton, Thaumasite sulfate attack in Portland and Portland-limestone cement mortars exposed to sulfate solution, Constr. Build. Mater. 40 (2013) 162–173. [38] O.S.B. Al-Amoudi, M. Maslehuddin, M.M. Saadi, Effect of magnesium sulfate and sodium sulfate on the durability performance of plain and blended cements, ACI Mater. J. 92 (1995) 15–23.