*2.6. Chloride Migration*

The NT Build 492 standard was used to determine the non-steady state chloride migration coefficient in concrete or cement-based materials [18]. The standard requires hardened samples to determine the material's resistance to chloride penetration. The method requires samples with a diameter of 100 mm and a thickness of 50 mm cut from the casting cylinders. The samples are immersed in a saturated solution of calcium hydroxide— Ca(OH)2—and prepared with distilled water, using a desiccator and a vacuum pump, for

25 h. Once the saturation cycle is complete, the anode contains a 0.3 N NaOH solution (12 g NaOH in 1000 mL distilled water), and the cathode containsa2N NaCl solution (100 g NaCl in 900 mL water).

**Figure 2.** Experimental setup for the characterization of reinforced concrete samples.

#### **3. Results and Discussion**

## *3.1. Microstructural Characterization*

The 29Si NMR spectra of the blast-furnace slag are shown in Figure 3, and the deconvolutions of the starting materials are listed in Table 1. The 29Si spectrum of the anhydrous cement showed 5 peaks; the peaks at −57, −71, −80, and −91 are assigned to the glassy component of the slag, and these signals confirmed the presence of natural pozzolans. The last peak in the spectrum is a very wide signal, which is indicative of low crystallinity. The gehlenite −71 ppm signals were still observed. The two signals of the natural pozzolan (−80.3 and −90.8 ppm) were also detected; this implies that the pozzolan is a non-reactive material, as its percentage was maintained after the hydration of the cement, and the peak at −89 ppm corresponded to the crystalline mullite [14]. The signal assigned to the Q4(0Al) units remained very wide, but had shifted slightly downward (greater chemical shift) and was centered at −106 ppm.

**Figure 3.** 29Si NMR spectra of the raw materials.


**Table 1.** Deconvolution of the 29Si MAS-NMR spectra.

#### *3.2. Electrochemical Impedance Spectroscopy*

Figure 4 shows the equivalent circuit characterized by the resistors and CPEs (constant phase elements) used in the impedance data simulation, indicating the distribution of the ions in the sample-solution system. The first interface corresponds to cementitious material in contact with the saline solution (Rs combined with Yp1-R1). The second corresponds to the cementitious material containing the reinforcing steel (R1-Yp1) combined with Yp2-R2.

**Figure 4.** Equivalent circuit corresponding to the different interfaces.

Figure 5 shows the Bode plots as a function of time for the alkaline-activated concrete samples immersed in a 3.5% sodium solution. Table 2 includes the parameters used in the simulation. The values of these parameters have been obtained using a complex least-square nonlinear program (CNLS) [19]. In addition, the EIS test allows the diffusion coefficient to be calculated and the degree of porosity of the material to be related [20]. Finally, the Bode plots have been simulated using the equivalent electrical circuit in Figure 3.

**Figure 5.** Bode plots for steel samples embedded in concrete and evaluated over a nine-year period.


**Table 2.** Values of the characteristic parameters used in evaluating the behavior of steel embedded in alkaline-activated concrete, along with its response over time.

In the Bode plots, as a function of the evaluation time (Figure 4), a capacitive behavior is observed by two time-constants determined by two flattened mid- and low-frequency inflections in which the center is located below the real axis [21]. The inflection defined at high frequencies is associated with the porosity of the samples [22]. This phenomenon of lower inflection attenuation is associated with a dispersion process in the frequency, since the surface of the steel rod is corrugated. The phase vs. frequency Bode plots define an inflection at high frequencies after 3 years of evaluation. This type of behavior can be interpreted by the barrier effect generated in the system, being more significant at 6 years and stabilizing at 7.5 years [23,24].

The Bode plots showing magnitude vs. frequency (Figure 5) show the evolution of the concrete's protective effect over time when immersed in a solution of 3.5% by weight of NaCl. For 1.5 years, the value of the magnitude of impedance doubled regarding the material after the curing process and without the presence of a chloride ion. This condition was called zero time. Subsequently, the impedance values increased dramatically up to the 6-years point [25]. After 7.5 years of testing, the magnitude of the impedance decreased, stabilizing, and the values were like those obtained at 9 years [26]. The results showed that the impedance values increased in the first years of the test due to the alkaline activation of the aluminosilicates, which played an essential role in the stability of the steel. After the sixth year of testing, the decomposition of the steel's protective layer was accelerated, similar to the behavior noted without protection; the loss of stability was minimal after 7.5 years, and there was no evidence of a decrease in impedance values [27].

The evolution of the impedance occurs in the first 28 days of concrete curing, during which time the formation of the hydration products of the pozzolanic reaction favors the initial values of protection against the corrosion phenomena [28]. Likewise, the samples in contact with the NaCl solution evaluated between 18 months and 9 years showed that the impedance values were related to the decrease in the degree of porosity of the mixtures. This process is attributed to a higher generation of hydration products as a result of the protection of steel. Using materials such as iron and steel slag in concrete mixtures contributes to the optimization of the mechanisms that control the entry of chloride ions into the samples [29]. Using materials such as iron and steel slag in concrete mixes helps optimize the mechanisms that control the entry of chloride ions into the samples [29]. Blast furnace slag, with its pozzolanic, chemical, and physical properties, can reduce the permeability of mixtures during their hydration processes. These characteristics prevent the migration of chloride ions into the samples, improving their durability [30]. Alkaline activation generates dense and uniform zones at the steel–concrete interface owing to the slag activated with Na2SiO3. This behavior is attributed to factors such as the water content that reduces the function of Na2SiO3. The low porosity as a function of time is explained by the high initial concentrations of SiO4 in the pore solution and the increasing dissolution of quartz [31]. Another factor is the production of calcium hydroxide in the cementitious hydration processes, resulting in the formation of hydrated calcium silicates that fill the existing pores in the concrete mixtures, reducing porosity [32,33].

The impedances shown in Figure 5 illustrate the different assessment times, indicating that when compared to the average migration of the chloride ion in the concrete, the first years of evaluation show that the reaction was slow, and the migration values were high. However, after 7.5 years of evaluation, it was established that the pozzolanic activity generated a reduction in the migration of the chlorides. Therefore, time was determined to be a factor in reducing the mobility of the chloride ions across the porous matrix.

#### *3.3. Resistance to Chloride Ion Penetration*

Figure 6 shows the results obtained by applying the ASTM C1202 standard to determine the permeability of concrete to chloride ions. This methodology measures the passage of an electrical current through concrete samples [31,32]. The behavior of the analyzed concretes as a function of the transferred load is observed (Figure 6). The results show that the alkali-activated concrete evaluated for up to 9 years presented a lower permeability to chlorides than did the samples at 28 days, which was called zero value. Slag concrete activated with sodium silicate is classified as having a low permeability to chlorides, and the behavior is observed as a function of time, up to nine years [33]. This behavior is associated with the plugging of the pores. The permeability of concrete is explained by the relationship or factor that most influences the fixation of chloride ions to the sample. The high concentration of chlorides in the studied systems is due to the exposure time (nine years), which can increase the capacity of the sample to fix chlorides [34]. It was also determined that the concretes evaluated at nine years could still fix chlorides; therefore, the existence of free chloride ions (within the concrete samples) looking for a way to fix themselves to the specimens was noted. These free chloride ions can affect the durability of concrete because they are responsible for initiating the corrosive process in the steel reinforcement within the structures [35]. By correlating this phenomenon with the EIS technique, it was established that the high impedance values are due to the presence of alumina in activated concrete, which is associated with the ability of concrete mixtures to fix and immobilize free chloride ions within the pore solution [36].

**Figure 6.** Determination of resistance to chloride ion penetration.

#### *3.4. Volume of Permeable Pores*

Figure 7 shows the results obtained for the concrete used in this study, based on the volume of permeable pores. In the graph, the concretes with more exposure time to the saline solution had fewer permeable pores [37]. That is, they are good quality concretes. In the case of alkali-activated iron and steel slag concrete, this property is attributed to the more significant densification of the paste—due to the presence of the hydration products— along the evaluation times, generating a lower percentage of pores [38]. This behavior is compared with the results obtained by chloride ion permeability and electrochemical

impedance spectroscopy, since chloride ions can be absorbed or dissolved in the water concentrated in the pores. As previously mentioned, free chlorides are responsible for the corrosion of the steel used to reinforce concrete structures. However, the low volume is associated with the hydration of the cement, so there is no corrosion promoter [39]. This behavior is attributed to the reactivity of the granulated blast furnace slag in the presence of the alkaline activator, causing hydration products that block the pores, resulting in slower chloride ion ingress rates when evaluating the systems as a function of time [40].

**Figure 7.** Variation of permeability in concrete with alkaline activation.

#### *3.5. Chloride Ion Migration*

Figure 8 shows the average evaluation times of chloride ion migration in concrete. The graph shows high values in the first years of evaluation. However, after the third year of evaluation, it was determined that the pozzolanic activity reduces the migration of chlorides. From the analysis, it can be concluded that the mobility of the chloride ions through the porous matrix is reduced over time.

**Figure 8.** Non-steady-state migration coefficient as a function of time.
