Corrosion of steel rebars embedded in One-part Alkali activated concrete mixes

. To reduce CO 2 emissions and turn a variety of industrial/agricultural wastes into valuable cementitious products, alkali-activated materials (AAM) are recognized as suitable substitutes for regular Portland cement (OPC). However, the concentrated aqueous alkali solutions used in conventional two-part alkali activated materials are highly corrosive, viscous, and are difficult to handle in direct field applications. As a result, the potential for developing so-called "just add water" type one-part AAMs, as compared to traditional two-part AAM, is being explored, particularly in cast-in-situ applications. In the present study on corrosion of reinforcing steel bars in fly ash-slag (FA-GGBS) based one-part AAC mixtures, three parameters — the total binder content, the relative proportions of GGBS and Fly-ash and the percentage of sodium oxide (Na 2 O) - are recognized as the key factors in determining the strength and durability performance (including corrosion of rebars embedded in it) of a given AAC mix. Accordingly, experiments were conducted on AAC mixes with three binder contents (440, 460, and 480 kg/m 3 ), three Slag/FA ratios (80/20, 70/30 and 60/40, by volume) and three alternate Na 2 O percentages (5, 6, and 7%, by weight of total binder content). Prismatic cylindrical test specimens of reinforced geopolymer concrete were prepared and half-cell potential and corrosion rate measurements were made after 28, 56, and 90 days of continuous exposure to 3% of NaCl solution, to accelerate the corrosion process. Measured corrosion current density and corrosion rates using a Electro-chemical Corrosion Analyser have indicated that the AAC mixture having a total binder content 440 kg/m 3 , GGBS/FS ratio of 70/30 and 6% Na 2 O content, exhibits best corrosion resistance amongst the various mixes tested herein, as measured up to the end of 90-days.


Introduction
Due to the significant amount of CO2 emitted during the calcination of limestone at temperatures between 1400 and 1450°C and the associated high energy requirements, the traditional manufacture of ordinary Portland cement (OPC) is a significant contributor to the global carbon dioxide emissions.For instance, in 2016, the total CO2 global emissions brought on by the manufacturing of OPC reached nearly 1.45 Gt, or around 8% of all ICSTCE 2023 https://doi.org/10.1051/e3sconf/202340503024 E3S Web of Conferences 405, 03024 (2023) anthropogenic CO2 emissions [1].One strategy to reduce net CO2 emissions from construction industry is to develop alternate low-carbon binders [2,3].Promising amongst such alternate materials are geopolymers, considered as a sub-group of alkali-activated materials (AAMs) [4], which are regarded as an environmentally acceptable green cementitious materials with a minimal carbon footprint [5].Geopolymers can be produced by activating aluminosilicate precursors, such as kaolin clays found in nature or industrial by-products such as fly ash and slag [6].
AAMs can be produced in one of two ways, using either a " two-part" or a "one-part" approach.A broad range of alkali-activated materials are typically manufactured using the traditional "two-part" methodology, which involves activating the raw aluminosilicate precursors with liquid alkali activators in addition to water when necessary [6].Since these alkali solutions are, most often, viscous, highly corrosive, and unfriendly, working with traditional "two-part" AAMs in the field is quite challenging.Provis 2018, illustrated that the traditional "two-part" AAMs would, most likely, continue to be used in precast production in controlled environments in the factories.However, a new class of AAMs were to be developed which could be used comfortably in in-situ applications.This new class of "onepart" or "just add water" type alkali activated materials use solid activators [7] and are supposed to be more cost-effective and environment friendly than their two-part counter studies on one-part AAMs [7,8,9].
Chloride-induced corrosion of embedded reinforcement is a significant issue affecting durability of reinforced concrete structures contribute to the premature failure of reinforced concrete structures.On RC structures', a significant amount of money is spent yearly on repairs and maintenance due to damage caused by chloride induced corrosion [10].To determine whether AAMs are suitable replacements for Portland cement-based reinforced concrete structures, a detailed examination of resistance of AAMs to chloride-induced corrosion is required.The published research on chloride-induced corrosion of rebars in AAM concretes is, however, still scarce.
When chlorides enter the concrete, the pseudo-passive layer that prevents corrosion on the steel is harmed [11].According to specific reports [12], enhancing Geo-polymer concrete (GPC) quality may refrain the reinforcing bar from corroding.Tennakoon performed longterm studies on the corrosion of steel rebars in fly-ash-based and slag-based GPCs [13].Their research demonstrated that the embedded rebars in fly ash and slag-based GPC mixes had greater corrosion resistance than in OPC-based mixes and their chloride diffusion coefficients were lower than those of OPC concrete.Reddy [14] experimented on the durability of reinforced GPC in marine environment.They used centrally reinforced beams manufactured with 8M and 14M concentrations of sodium-based solutions to test the corrosion-based durability of low calcium fly ash-based, two-part GPC mixes.The results of these experiment have demonstrated that two-part GPCs are superior to OPC-based concrete mixes in terms of corrosion resistance.Using three different types of OPC-fly ash mortars, [15] assessed the corrosion stability of the passive state with additions of 0%, 0.20%, 0.40%, and 2.0% chlorides (by total weight of binder material in the mix).According to them, 2% of pre-mixed chlorides multiplies the rate of corrosion of steel reinforcement by hundred times.Gunasekara et al. [10] examined the impact of chlorides on GPC and OPC-based samples over 540 days.In their studies, chloride-admixed fly ash geopolymer samples showed higher levels of corrosion resistance as compared to similar OPC-based samples while reduced chloride diffusion was seen in GPC samples ponded in 3% sodium chloride solution.
However, there needs to be more knowledge regarding the durability of steel reinforcement in GPCs in various corrosive environments, which then promotes a more extensive range of construction applications with them [16].To determine if AAMs are suitable replacements for OPC-based reinforced concrete structures, detailed studies of resistance of AAMs to chloride-induced corrosion is required.The published research on studies on chloride-induced corrosion in alkali-activated concretes is still scarce.
Herein results of a detailed study on corrosion of reinforcing steel bars in fly ash-slag (FA-GGBS) based One-part AAC mixtures additionally premixed with 3% of sodium chloride is presented, with three parameters-the total binder content, the relative proportions of GGBS and Fly-ash and the percentage of sodium oxide (Na2O) -recognized as being the key factors in determining the strength and durability performance (including corrosion of rebars embedded in it) of a given AAC mix, Accordingly, experiments were conducted on AAC mixes with three total binder contents (440, 460, and 480 kg/m 3 ), three Slag-to-FA ratios (80/20, 70/30 and 60/40, by volume) and three alternate Na2O percentages (5,6, and 7%, by weight of total binder content).Prismatic cylindrical test specimens of reinforced geopolymer concrete were prepared and corrosion rate measurements were made after 28-, 56-, and 90 days of their casting by using electrochemical method.Herein the test specimens were kept in 3% of NaCl solution up to 90 days to accelerate the corrosion process.The corrosion of the rebar was analysed using cyclic sweep method on a state-of-art, Electrochemical Corrosion Analyser (Make: M/s Gill AC, ACM instruments, United Kingdom).The results obtained during the tests are then analysed and discussed.

Aluminosilicate precursors
Low calcium class F fly ash is normally produced by burning anthracite or bituminous coal.This type of fly ash usually has a CaO content in the range of 5-10%.It has pozzolanic properties only.For the present work, class F type of fly ash from M/s Udupi Power Corporation Ltd, Padubidri, Udupi, Karnataka State has been used.Again, ground granulated blast-furnace slag (GGBS), a by-product of Iron and Steel Industry, is probably the most effective cement replacement material used in concrete constructions.In the present investigation, GGBS from M/s JSW Iron and Steel Company, Bellary, Karnataka, India, supplied by a local supplier, is also used as a source material.

Alkaline Activator
Anhydrous sodium meta-silicate (Na2SiO3-Anhydrous) powder is the solid powder activator chosen for this work based on the recommendations of Nematollahi [17].The chemical composition of the anhydrous sodium metasilicate powder has been determined as 50.46%Na2O and 47.24% SiO2 by weight, with a modulus ratio (Ms) of 0.9-1.0(where Ms = SiO2/Na2O).This material has been procured from a local dealer, in bulk.Specific Gravity of the activator is 2.614

Aggregates in the mix
In the present investigation, locally available manufactured sand conforming to IS : 383-2016 has been used as fine aggregate (Specific Gravity 2.65).Combined crushed granite aggregates of MAS 20 mm and 12.5 mm (Specific Gravity 2.7), conforming to IS : 383-2016 have been used to produce the one-part alkali-activated concrete mixes.

Water
The quantity and quality of water is required for mixing to be looked into carefully.Potable water is generally considered satisfactory.In the present investigation, however, tap water which is available in the laboratory has been used for mixing.

Details of One-part AAC mixes used
Experiments were conducted herein on AAC mixes with three alternate binder contents (440, 460, and 480 kg/m 3 ), three Slag-to-FA ratios (80/20, 70/30 and 60/40, by volume) and three alternate Na2O percentages (5, 6, and 7%, by weight of total binder content).Coarse aggregates of 20 mm and 12.5 mm, mixed in 3:2 proportions by weight, were used in each of the AAC mixes.Water to binder ratio is fixed at a moderate value of 0.42 in all the mixes.
It has been well recognized that alkali activated Concretes/Geopolymer concretes are quite impermeable and hence longer durations of time are required for the aggressive ions to permeate through them.Hence, in order to accelerate the corrosion of embedded rebars in the test specimens, sodium chloride powder is pre-mixed each test mi at 3% (by weight) of their total binder content at the mixing stage.
The choice of the three primary design variables considered in each of the nine design mixes in the calibration phase, based on a 3x3 orthogonal array in a Taguchi's Design of Experiments framework are shown in Table 1.The mix-quantities, per cubic meter, of all the ingredients used in each of the nine calibration mixes are detailed in Table 2.

Mixing, specimen preparation and curing methods
For the preparation of AAC mixes, specified quantities of fine aggregate, coarse aggregate, source materials (fly ash and GGBS) and solid activator, as per mix-design, are initially dry mixed in a Ribbon-Type concrete mixer.In addition, powdered NaCl, 3 % by weight of total binder, has been added to the dry materials in the concrete mixer.To initiate the reactions, water is added and mixed for 4-5 min to obtain a homogenous AAC mixture.After that slump test was performed.
In order to test the quality of the prepared concrete mixes, by testing their compressive strengths at 28-days, 100 mm cube specimens were cast with the various one-part alkali activated concrete mixes.For testing the corrosion characteristics of steel rebars embedded in the various one-part alkali activated concrete mixes, prismatic cylindrical specimens of 75 mm diameter and 300 mm length, as shown in Fig. 1. were cast.Three test specimens were cast from each of the mixes.The prismatic cylindrical specimens have a 12 mm diameter TMT steel bar embedded centrally inside them.A 30 mm AAC cover was maintained for the Steel reinforcing bar.Before being centrally put into the prismatic specimen, the reinforcing steel bar was wire-brushed to remove the surface scales.The TMT reinforcing bar was wrapped in the insulating tape and then coated with epoxy at either of the ends avoid crevice corrosion.The interior of all the moulds were initially cleaned.Grease is applied on the inside surfaces of the moulds.The concrete mix was poured into the moulds, in layers, and initially hand-compacted using tamping rod.Further, the moulds were placed on a table vibrator and were machine vibrated.

Fig. 1. AAC test specimen for corrosion measurement
The moulds were kept in the ambient laboratory conditions for 24 hours after which all the test specimens were demould.Further, the cube specimens were left in the ambient laboratory environment i.e., room temperature (generally 23 ± 2 0 C) and the corrosion test specimens were kept in a 3% NaCl solution for specified number of days to accelerate the corrosion process.

Compressive strength
A 2000 kN-capacity Compression Testing Machine (CTM) has been used for testing the 100 mm-sized cube specimens.Compressive strength tests were performed on cube specimens of various one-part alkali activated concrete mixes, after 28 days of curing under ambient laboratory conditions.The average of compressive strengths of three replicate AAC cube specimens made of each mix has been considered as the compressive strength of that mix.

Corrosion monitoring methods
The corrosion characteristics of the rebar embedded inside the test specimens made of different AAC mixes were evaluated based on electrochemical techniques.Tests were based half-cell potential method, cyclic sweep tests were performed to evaluate the corrosion characteristics at 28, 56, and 90-days of continuous exposure to aggressive 3% sodium chloride solution.A state-of-art, corrosion monitoring device [Make: M/s Gill AC, ACM instruments, United Kingdom] was used to track the corrosion performance of the cylindrical prismatic specimens Fig. 2. For corrosion monitoring, the performance of each AAC mix was determined by averaging the results of three replicate reinforced AAC cylindrical prismatic specimens.

Fig. 2. Corrosion Testing using Electrochemical Corrosion Analyzer [Make: GILL AC, USA]
In the corrosion measurements in the present study, three electrodes were employed and were connected to the corrosion monitoring device Fig. 3.The TMT steel bar in the prismatic AAC specimen was considered as the working electrode (WE), the saturated calomel electrode (SCE) as the reference electrode (RE), and the graphite electrode as the auxiliary electrode (AE).The reinforced prismatic specimen was immersed in a solution with a NaCl concentration that was the same (3%) as that used in the production of AAC during the corrosion study.Also same as the concentration of NaCl solution in which the specimen is kept cured.

Fig. 3. Arrangements of electrodes during the measurements
The half-cell potential of the reinforcing steel in the AAC specimens was calculated using the SCE as a reference, ASTM-C876 was followed to monitor the possibility of corrosion of the steel bar embedded within the geopolymer concrete mix.By using the Gill AC corrosion analyser, the reinforced steel bar in the AAC specimen was polarized to ± 270mV at a repetition rate of 30mV/minute during the cyclic sweep testing.
The Icorr and corrosion rate of rebars embedded in one part alkali activated concrete specimens are calculated based on the Tafel plots.By using Tafel's Rulers, the anodic and Corrosion rate, (mm / year) = (Metal Factor × Icorr)/1000 where Icorr = A/m 2 and Metal factor = (t × K)/ρ In the above, t = total number of seconds in year = 31556930 seconds ρ is the metal density (steel) in g/cm 3 K is the electrochemical equivalent in g / coulomb.K = ∑ (atomic % of element ×atomic weight of element) / (96487×Valency of element)

Results and discussion
In the following section, the results obtained herein for compressive strengths, and corrosion measurements made on a set of one-part AAC mixtures are discussed.

Compressive strength
The compressive strength of a material is a crucial property that determines its ability to withstand applied loads or pressure.In this case, we are specifically discussing the compressive strength of a set of one-part alkali activated concrete mixes.The average 28days compressive strength values for all the calibration mixes are shown in Fig. 4.

Fig. 4. 28th-day Compressive strength of AAC Mixes -Calibration Phase
Analysing the reported compressive strength values, it can be observed that the compression strengths of these mixes range from a minimum of 48.02 MPa (Mix-M3) to a maximum of 68.84 MPa (Mix-M9).
The compressive strength values of the nine mixes allow for a comparison of their respective strengths, providing insights into the effectiveness of different compositions and proportions.Further analysis, such as identifying the specific ingredients and proportions used in each mix, can help determine the factors influencing the compressive strength and guide the optimization of Alkali activated concrete formulations for desired applications.

Corrosion current densities (Icorr) and corrosion rates
Table 3. presents the average corrosion current densities (Icorr) and the corrosion rates calculated for each of the nine calibration mixes at 28-, 56-, and 90-days.The graphical representation of the calculated corrosion rates is in Fig. 5. From the results in Table 1, at an early age of 28-Days, the corrosion currents density (Icorr) ranged from 0.00351 A/m 2 (Mix-M2) to 0.00677 A/m 2 (Mix-M8).The corresponding corrosion rates at this age ranged between 0.0036 mm/year (Mix-M2) to 0.0068 mm/year (Mix-M8).As the age of continuous immersion of all the test specimens in 3% sodium chloride solution was continued, in general, the values of both the corrosion current density and corrosion rate, as measures of corrosion of the embedded rebar in the various one-part alkali-activated concrete mixes, also have got increased.While corrosion currents in the range of 0.00416 A/m 2 (Mix-M1) to 0.00648 A/m 2 (Mix-M8) and corrosion rates ranging 0.0042 mm/year (Mix-M1) to 0.0066 mm/year (Mix-M8) are recorded at 56-days.increased corrosion currents in the range of 0.00492 A/m 2 (Mix-M2) to 0.00709 A/m 2 (Mix-M8) and enhanced corrosion rates varying between 0.0050 mm/year (Mix-M2) to 0.0072 mm/year (Mix-M8) were recorded at 90-Days.
Overall, Mix-M8 consistently shows higher corrosion currents and corrosion rates compared to all the other mixes.On the other hand, Mix-M2 consistently exhibited lower corrosion measures.In their corrosion experiments on steel bars embedded a class of Two-part GPC mixes, admixed with 3.5% NaCl, [18], have obtained Icorr values ranging from 0.00276-0.00428A/m 2 at the age of 60 days.(Were their samples Therein, the chloride admixed GPC mixes had shown a higher probability of steel corrosion and also exhibited higher Icorr value as compared to control GPC mixes (with no chloride admixed initially in the mixes).
The results discussed above suggest that an increase in percentage of fly-ash in the onepart AAMs can lead to lower corrosion resistivity.Lower resistivity of fly ash, can cause increases in the corrosion rate of steel rebars embedded in GPC [12].
Concrete mixes can be categorized to be under passive, low, moderate, high risk high, and very high risks, as detailed in Table 4. [19].Now, amongst the present set of one-part AAC mixes, Mix-M2 may be considered to be still at low corrosion risk, while the rest of mixes have reached the moderate level of corrosion risk, because of continued exposure to aggressive chloride environment for 90-days.[12] Corrosion rate (µA/cm 2 ) Probable Corrosion level < 0.1 0.1-0.5 0.5-1 >1

Negligible -Passive Low Moderate High
According to Kupwade-Patil and Allouche [20], well designed GPC concrete mixes might be attractive replacement for OPC-based concrete in reinforced concrete structures located in marine environments or those exposed to deicing salts or salty water for extended periods in cold countries.

Conclusions
• The results obtained showed that the higher value of compressive strength was detected in AAC mix M9 prepared with 480 kg/m 3 of total binder content, GGBS/FS ratio of 60/40 and 7% Na2O.
• Measurement of corrosion current density and corrosion rates have indicated that addition of fly ash has significant effect on corrosion behaviour of GPC mixes.Again, the corrosion current and corrosion rates measured at 28-, 56-and 90-days of age reveal variations in the corrosion behaviour among the nine trial mixes with different mixproportions • While Mix M8 (480 kg/m 3 of total binder content, GGBS/FS ratio of 70/30 and 5% Na2O) exhibits the highest corrosion current density and corrosion rate, Mix M2 recorded (440 kg/m 3 of total binder content, GGBS/FS ratio of 70/30 and 6% Na2O) the lowest values, indicating the best of corrosion resistance amongst the various mixes, as measured up to the end of 90-days of continuous exposure to chloride environment.
• Rebar in Mix M-2 can be considered to be at a low corrosion risk and those in rest of the trial one-part AAC mixes can be considered as at moderate-risk, after continuous exposure to chloride environment up to 90 days.
• The results obtained for corrosion currents densities and corrosion rates provide insights into the durability of the concrete mixes.Lower corrosion rates indicate better corrosion resistance.

Table 1 .
Factors and Levels of primary variables considered -Calibration Phase -Taguchi's Design of Experiments Methodology

Table 3 .
Results of corrosion studies on one-part AAC mixes