Stainless Steel Voltammetric Sensor to Monitor Variations in Oxygen and Humidity Availability in Reinforcement Concrete Structures

The present work presents the results obtained with a stainless steel (SS) voltammetric sensor to detect variations in humidity (H2O) and oxygen (O2) availability in concretes. First, studies in solution were run by preparing several solutions to represent the different conditions that can be found in concrete pores. Second, the sensor’s response was studied by varying O2 availability by argon or synthetic air bubbling. Then concrete conditions with different degrees of carbonation were simulated using solutions with a pH between 13 and 8.45. After characterization in solution, a study by means of concrete samples with several water/cement ratios (0.6, 0.5 and 0.4) was performed, in which sensors were embedded and studied under different O2 and H2O saturation conditions. The obtained results revealed that with the voltagram, it is possible to evaluate O2 availability variation from the slopes of the lines identified logarithmically in the voltagram for the obtained cathodic sweeping. All the results obtained with the sensor were correlated/validated by standard assays to characterize porosity in hardened concretes.


Introduction
The follow up and forecasting of reinforced concrete structures' (RCS) service lives are some of the main recent objectives for reducing economic and environmental costs that are a direct result of inefficient maintenance actions, repairs that could have been minimized, and even demolishing buildings which could have been avoided if suitable preventive measures had been taken. Moreover, control and efficacious estimations would have helped to ensure people's safety.
In order to develop models that allow optimum follow up and forecast structures' service lives, it is necessary to know which phenomena threaten their durability. With RCS, one of the main problems is rebar corrosion [1][2][3][4]. The basic most extensive service life model related to this phenomenon was developed by Tutti [1,5]. This basic model distinguishes two periods: the first is the period of initiation, which includes the time from when aggressive agents penetrate the concrete matrix to the time when they reach rebars and bring about their depassivation; the second is the period of propagation, which covers the time from when rebars depassivate to when corrosion starts developing. The structure's service life ends when an unacceptable degree of corrosion is reached.
These durability models applied to a given structure can be built by taking measurements of the structure in real time with monitoring systems. One of the most promising systems is that formed by embedded sensors, which is why recent studies have focused on developing them [3,4,[6][7][8][9][10][11][12][13].
When following up corrosion processes in structures, sensors can be made that center on the period of initiation by detecting the presence of agents that trigger rebar depassivation by, for example, concrete carbonation or the presence of chlorides [2,4,14]; on the other hand, sensors can be used that assess the period of propagation either by quantifying the substances directly involved in corrosion processes like O 2 and H 2 O, or stating the rebar corrosion status [15]. Developing voltammetric multi-sensor systems that assess both the aforementioned periods is another possibility and would, therefore, allow complete service life models to be developed.
The voltammetric sensors operation is based on the application of potential signals continuously modified as a linear function of time. Each potential generates an electric current response related to a specific reaction, which involves the interest agents (oxygen, water, chlorides, etc.). By means of the current analysis as a function of the potential, we can quantify these agents [16]. This is clearly an advantage over other type of sensors, such as potentiometric or resistance sensors, because it allows for developing more accurate models in order to quantify different analytes contained in the electrolyte.
Apart from seeking sensor systems' effectiveness, it is fundamental to create autonomous and very reliable low-cost and high-durability systems. This is why it is so important that the material used to make sensors meets these requirements.
Stainless steel (SS) use is widespread in industry thanks to its stability in many aggressive environments, and also its low cost compared to other metals [17]. This alloy presents excellent mechanical and corrosion resistance and is a good conductor of both heat and electricity [18]. Thus, its use has grown in recent years in different sectors, such as energy, pharmaceuticals, shipping, and construction [18][19][20].
In short, we understand SS to be an iron carbon alloy that contains a minimum of 10.5% chrome. Although this steel is called stainless, it does not mean that it does not corrode, rather a protective layer of chrome oxide forms on its surface instead of common rust. If under normal atmospheric conditions this chrome oxide layer deteriorates, it is immediately formed again by chrome combining with oxygen (O 2 ) in environmental atmospheres to protect the metal from poorly stable and highly soluble iron oxide formation [21].
Today this metal has been employed to manufacture sensors used in the food industry [22] and those embedded in concrete [8,23]. With concrete, Correia, et al. [6] demonstrate the effectiveness of using an SS electrode to detect O 2 variations inside concrete. In the same article, Correia, et al. [6] also mention how the degree of concrete saturation is observed by Rs measurement (resistance to the ions circulating through concrete [23]) by the impedance spectroscopy technique. The article by McCarter and Vennesland [8] refers to SS being used as a reference electrode and counter electrode in sensor systems to control RCS corrosion. The article by Xu, et al. [24], presents a new kind of stainless steel (SS) galvanic sensor system for the study of rebar corrosion in different pore solution conditions. Tabatabai and Aljuboori [25] developed a resistance sensor made with SS in order to detect ice and water on roads and bridges.
On the other hand, many authors propose that O 2 reduction in SS is associated with the zones in which Fe(II) compounds act as absorption centers, promoting the oxygen reduction reaction [6,21,[26][27][28].
Bearing this in mind, developing a voltammetric sensor would be interesting to monitor RCS' service life with this metal because the effectiveness of sensors to detect variations in O 2 availability inside the RC matrix has been proven. Therefore, being able to develop O 2 availability quantification models by sensor measurements would be very interesting indeed. O 2 is one of the main agents that induces rebar corrosion in RCS [4], [6,29,30]. Furthermore, unexpected increase in O 2 availability can mean a physical defect of the concrete cover in RCS, such as the formation of cracks. It could identify indirect reactions as a result of the concrete carbonation, because this phenomenon produces the pH change of the concrete pore solution. The change of pH significantly affects the oxygen reduction in the SS surface [21].
Moreover, this metal's low cost and its good services would allow sensor systems to be produced that cover large surfaces and offer excellent durability at a much lower cost than if noble metals were employed.
Our objective was, therefore, to assess the possibilities of using SS voltammetric sensors embedded in concrete to develop a sensor system with SS. To do so, we assessed their reactivity to the different conditions that could come into play in concrete (carbonation, chloride penetration, variations in O 2 availability, etc.) by employing solutions that simulate concrete pore solution. Then we assessed the behavior of sensors when they were embedded in concrete by varying O 2 availability conditions and humidity before checking if cracks in concrete could be detected, and even different cracking patterns, by these sensors [31].
The objective of the analysis herein performed was to understand how the SS sensor works to know its limitations, to define simple correlations, and to identify its potential to discriminate changes in O 2 availability and humidity by bearing in mind that they are most important in the corrosion reaction of RCS.

Materials and Methods
In order to fulfil the proposed objectives, we defined the following study phases: • Phase (1) Study in solution. In this preliminary phase, the processes that took place on the SS electrode surface when immersed in solution were studied, which represent the different conditions that could come into play in concrete. The obtained results were compared with those found in the literature. • Phase (2) Studies performed with the SS sensor embedded in conventional concrete. SS voltammetric sensors were embedded in conventional concrete samples at different water/cement ratios (w/c). Samples were studied under distinct O 2 availability conditions.
The sections below explain in detail the experiments that were run in each phase and how SS sensors were produced. The results were compared to those found in the literature to understand how the O 2 reduction reaction takes place on the SS electrode and how these conditions can affect it, as we focused on this reaction in the present study.

The Applied Electro-Analytical Techniques
The following electro-analytical techniques of cyclic voltammetry and impedance spectroscopy were applied by Autolab PGSTAT10 equipment and the Nova 1.11 software.
The test configuration applied with both these techniques was the 3-electrode configuration. The same SS 304 wire to manufacture sensors was employed as the working electrode (0.8 mm diameter, 12 cm long) was used. An INOX plate was utilized as the auxiliary electrode (105 cm 2 surface). The reference electrode was a saturated calomel electrode (SCE), while the potential values herein presented referred to the SCE.
The cyclic voltammetry technique was applied to the range within which O 2 reduction took place and was a range of sufficiently negative potentials to induce the process [23].
Here the intention was only sweeping in the area of the negative potentials. We made sure to not exceed the "critical pitting potential" during anodic sweeping where the passivity of SS steel was lost and became active [26,33].
Potential sweeps commenced from the OCP (open circuit potential or the system's equilibrium) cathodically until −1.2 V, before returning to 0 V (anodic sweeping) and ending back at the OCP to make the system as least polarizing as possible. The sweep rate was 20 mV/s. The aim of such sweeping was to disturb the real O 2 availability inside the concrete matrix as little as possible.
The ohm drop in the voltagrams was corrected with the Rs value obtained by the impedance spectroscopy technique with the same 3-electrode configuration. The range of sweeping frequencies went from 100,000 Hz to 1000 Hz. Sweeping at high frequencies was justified by simplifying the system similarly to Randles simple equivalent electrical circuit (Rs-Rp/Cdl). The effective value of the applied tension was 10 mV.

Studies Performed with the Sensor Embedded in Concrete
In order to manufacture SS sensors, a 0.8 mm-diameter SS 304 wire was used. The sensor's working surface was limited by covering the ends with waterproof polyurethane paint. Sensors' average unsupported length was 12.3 ± 0.5 cm. They were made U-shaped to easily insert them into samples ( Figure 1). • Solution 0.1 m NaOH (pH = 13) to simulate non-carbonated concrete by adding H2O2 at 0.02 m, and to assess the possible appearance or stabilization of this intermediate product in reducing O2, as suggested in the literature The results were compared to those found in the literature to understand how the O2 reduction reaction takes place on the SS electrode and how these conditions can affect it, as we focused on this reaction in the present study.

The Applied Electro-Analytical Techniques
The following electro-analytical techniques of cyclic voltammetry and impedance spectroscopy were applied by Autolab PGSTAT10 equipment and the Nova 1.11 software.
The test configuration applied with both these techniques was the 3-electrode configuration. The same SS 304 wire to manufacture sensors was employed as the working electrode (0.8 mm diameter, 12 cm long) was used. An INOX plate was utilized as the auxiliary electrode (105 cm 2 surface). The reference electrode was a saturated calomel electrode (SCE), while the potential values herein presented referred to the SCE.
The cyclic voltammetry technique was applied to the range within which O2 reduction took place and was a range of sufficiently negative potentials to induce the process [23]. Here the intention was only sweeping in the area of the negative potentials. We made sure to not exceed the "critical pitting potential" during anodic sweeping where the passivity of SS steel was lost and became active [26,33].
Potential sweeps commenced from the OCP (open circuit potential or the system's equilibrium) cathodically until −1.2 V, before returning to 0 V (anodic sweeping) and ending back at the OCP to make the system as least polarizing as possible. The sweep rate was 20 mV/s. The aim of such sweeping was to disturb the real O2 availability inside the concrete matrix as little as possible.
The ohm drop in the voltagrams was corrected with the Rs value obtained by the impedance spectroscopy technique with the same 3-electrode configuration. The range of sweeping frequencies went from 100,000 Hz to 1000 Hz. Sweeping at high frequencies was justified by simplifying the system similarly to Randles simple equivalent electrical circuit (Rs-Rp/Cdl). The effective value of the applied tension was 10 mV.

Studies Performed with the Sensor Embedded in Concrete
In order to manufacture SS sensors, a 0.8 mm-diameter SS 304 wire was used. The sensor's working surface was limited by covering the ends with waterproof polyurethane paint. Sensors' average unsupported length was 12.3 ± 0.5 cm. They were made U-shaped to easily insert them into samples ( Figure 1). L-shaped 3 cm-wide SS plates were used as counter electrodes. The average surface of these employed counter electrodes was 113.34 ± 2 cm 2 , which was about 37-fold bigger than the surface of SS sensors to ensure that those reactions taking place on the counter L-shaped 3 cm-wide SS plates were used as counter electrodes. The average surface of these employed counter electrodes was 113.34 ± 2 cm 2 , which was about 37-fold bigger than the surface of SS sensors to ensure that those reactions taking place on the counter electrodes would not interfere with the response obtained in the electro-analytical tests [34].
These sensors were embedded in prismatic concrete samples (4 × 4 × 16 cm 3 ) made with three different types of concrete, which are set out in Table 1.  Figure 1 illustrates how sensors were placed inside concrete samples. Sensors were placed at a depth of more or less 1 cm at the top of samples.
To confer the study statistical reliability, three mixes were produced with each concrete type and two samples were made per mix. Table 2 includes the number of samples per dosage. In accordance with the O 2 availability conditions, the experimental part was divided into two subparts (see Table 3). The system described in Figure 2 was set up in the non-saturated state to submit samples to different O 2 availability conditions. In this state, partial O 2 pressure was regulated inside the desiccator by a vacuum pump. To keep temperature constant, the desiccator was left in a thermostatic bath and the temperature inside the desiccator was left at 23 • C ± 2 • C. To achieve constant relative humidity (RH) conditions, saturated salt solution was used to keep RH at around 84.5% ± 0.3 (saturated KCl solution) [35]. With an approximately constant temperature and RH during tests, constant environmental absolute humidity conditions were achieved, which enabled us to ensure that the humidity which samples could absorb did not differ due to the environmental conditions, rather owing to their own porosity.

RATED STATE
sample was tested 3 times in each state. All the samples were tested under SAT conditions. Each sample was tested 3 times in this state 2.2.1. Study in the Non-Saturated State The system described in Figure 2 was set up in the non-saturated state to submit samples to different O2 availability conditions. In this state, partial O2 pressure was regulated inside the desiccator by a vacuum pump. To keep temperature constant, the desiccator was left in a thermostatic bath and the temperature inside the desiccator was left at 23 °C ± 2 °C. To achieve constant relative humidity (RH) conditions, saturated salt solution was used to keep RH at around 84.5% ± 0.3 (saturated KCl solution) [35]. With an approximately constant temperature and RH during tests, constant environmental absolute humidity conditions were achieved, which enabled us to ensure that the humidity which samples could absorb did not differ due to the environmental conditions, rather owing to their own porosity. The following test sequence was repeated three times per dosage: 1. Vacuum conditions: To homogenize the initial behavior of the sensors embedded in the different samples, the studied samples were subjected to a vacuum for 12 h. After finishing the adaptation period, the aforementioned electro-chemical studies were carried out. 2. Atmospheric conditions: Once the tests under the vacuum conditions ended, synthetic air was allowed to enter the desiccator. When a 1-bar pressure was achieved inside, 60 min were allowed for the system to stabilize before testing began. 3. Partial 500-mbar air pressure: when the tests under atmospheric conditions ended, the pressure inside the desiccator was lowered until vacuum conditions were The following test sequence was repeated three times per dosage: 1.
Vacuum conditions: To homogenize the initial behavior of the sensors embedded in the different samples, the studied samples were subjected to a vacuum for 12 h. After finishing the adaptation period, the aforementioned electro-chemical studies were carried out.

2.
Atmospheric conditions: Once the tests under the vacuum conditions ended, synthetic air was allowed to enter the desiccator. When a 1-bar pressure was achieved inside, 60 min were allowed for the system to stabilize before testing began.

3.
Partial 500-mbar air pressure: when the tests under atmospheric conditions ended, the pressure inside the desiccator was lowered until vacuum conditions were achieved. At this point, 30 min were allowed before letting air in until a sTable 500-mbar pressure was accomplished. After a 15-min wait, argon was injected until the 1-bar pressure was recovered. In this way, the O 2 molar fraction inside the desiccator diminished to half in relation to the working atmospheric conditions. Measurements were taken on sensors after 1 h under 500 mbar conditions.
A preliminary study was performed which demonstrated that the 1-h adaptation time before starting each test series sufficed to ensure the reproducibility of the obtained results.
While testing, the temperature was controlled inside the desiccator by the thermocouple placed inside it. After repeating the test sequence for 3 days, samples were left in a closed container under 100% humidity conditions to prevent concrete carbonation.

Study in the Saturated State
In order to achieve the saturated state, all the samples were placed inside a vacuum chamber ( Figure 3). Vacuum conditions were set up in this chamber and lasted 1 h, after which time H 2 O was allowed to enter and completely cover samples. The system was left under these conditions without breaking the vacuum for another 1-h period. After this time, the vacuum was broken, and samples were left immersed for 48 h before the test series commenced under these conditions. before starting each test series sufficed to ensure the reproducibility of the o While testing, the temperature was controlled inside the desiccator couple placed inside it.
After repeating the test sequence for 3 days, samples were left in a cl under 100% humidity conditions to prevent concrete carbonation.

Study in the Saturated State
In order to achieve the saturated state, all the samples were placed in chamber ( Figure 3). Vacuum conditions were set up in this chamber and l which time H2O was allowed to enter and completely cover samples. The under these conditions without breaking the vacuum for another 1-h pe time, the vacuum was broken, and samples were left immersed for 48 h series commenced under these conditions. Tests were carried out under these conditions on three consecutive d the dosages at the same time.

Applied Electro-Analytical Techniques
The same techniques as in Phase I were applied. The only difference o was that work was done in a 2-electrode configuration because the setup in and the test sequence did not allow for placing a reference electrode inside came into contact with each studied sample's surface. This configuration because the counter electrode had a much bigger surface than the working [34].

Concrete Characterization Tests
The results of these standardized tests were used to support the result the sensors.
The following tests were done: Tests were carried out under these conditions on three consecutive days and for all the dosages at the same time.

Applied Electro-Analytical Techniques
The same techniques as in Phase I were applied. The only difference on this occasion was that work was done in a 2-electrode configuration because the setup in the desiccator and the test sequence did not allow for placing a reference electrode inside the system that came into contact with each studied sample's surface. This configuration was acceptable because the counter electrode had a much bigger surface than the working electrode (WE) [34].

Concrete Characterization Tests
The results of these standardized tests were used to support the results obtained with the sensors.
The following tests were done: ). This test allows information to be acquired about the volume of interconnected pores and their size distribution. These tests were performed by the ITC of the Universitat Jaume I (Spain). The MIP method consists of injecting mercury at different pressures. The volume of absorbed mercury at each pressure is recorded so that the volume of absorbed mercury within a certain range of pressures is associated with a given access size range.
The equipment used was Mercury Posorimeter Micromeritics, AutoPore IV (9500). In order to perform tests, a sample was taken from specimen nuclei. Before running tests, samples were dried at 100 • C. Table 4 shows the mean values obtained during the characterization tests, as well as the coefficient of variation (CV), defined as the quotient between the standard deviation and the average value. In short, the lower the w/c ratio, the higher the fck for 28 days, and the less the H 2 O accessibility inside the matrix for both absorption and pressure; at a higher strength, concretes present higher electrical resistivity in the saturated state and less air permeability.

Results
In order to show the study results to clarify the following mechanism, Figure 4 is presented.

Results
In order to show the study results to clarify the following mechanism, Figure 4 is presented.

Phase 1: Studies Performed in Solution
We observed that the SS sensor in solution behaved slightly differently to that embedded in concrete. This agrees with what Correira et al. [6] reported. This difference might be due to the diffusion phenomena in concrete being very different to that in solution, which could influence the passive oxide layer that forms on the SS surface.
In line with previously published works [23], the nomenclature used in voltammetry graphs of electrical potentials includes ΔEWE for those cases in which the ohm drop is corrected and work is done with a 3-electrode configuration, and ΔVWE for the case when the ohm drop is corrected but work is done with a 2-electrode configuration.

Reducing O2 on the SS Electrode's Surface
According to Babic and Metikos-Hukovic [26] and Bösing et al. [36], the peaks observed in Figure 5a,c during anodic sweeping (P.Ox.) are associated with the electroformation of Fe(OH)2 layers and their oxidation to form FeOOH layers, while the peaks noted during cathodic sweeping (P.Re.) are associated with a reduction in these products. The Fe(OH)2 and FeOOH layers form on a pre-existing protective Cr2O3 layer [26,36]. Figure 5a,c depicts the voltagrams obtained using the SS sensor when moving from an airless atmosphere (the black line) to an aired atmosphere (the lilac line). Figure 5a shows the potentials vs. SCE and Figure 5c shows the potentials vs. OCP.
These figures show a large difference in current density in the zone where a reduction peak appeared (P.Re.). Bearing in mind both test conditions, this increase could only be due to the reduction in O2 (because N2 is an inert gas). This agrees with many authors in the literature who propose that O2 reduction in SS is associated with the zones in which Fe(II) compounds act as absorption centers [6,21 ,26-28].

Phase 1: Studies Performed in Solution
We observed that the SS sensor in solution behaved slightly differently to that embedded in concrete. This agrees with what Correira et al. [6] reported. This difference might be due to the diffusion phenomena in concrete being very different to that in solution, which could influence the passive oxide layer that forms on the SS surface.
In line with previously published works [23], the nomenclature used in voltammetry graphs of electrical potentials includes ∆E WE for those cases in which the ohm drop is corrected and work is done with a 3-electrode configuration, and ∆V WE for the case when the ohm drop is corrected but work is done with a 2-electrode configuration.

Reducing O 2 on the SS Electrode's Surface
According to Babic and Metikos-Hukovic [26] and Bösing et al. [36], the peaks observed in Figure 5a,c during anodic sweeping (P.Ox.) are associated with the electroformation of Fe(OH) 2 layers and their oxidation to form FeOOH layers, while the peaks noted during cathodic sweeping (P.Re.) are associated with a reduction in these products. The Fe(OH) 2 and FeOOH layers form on a pre-existing protective Cr 2 O 3 layer [26,36]. Figure 5a,c depicts the voltagrams obtained using the SS sensor when moving from an airless atmosphere (the black line) to an aired atmosphere (the lilac line). Figure 5a shows the potentials vs. SCE and Figure 5c shows the potentials vs. OCP.
These figures show a large difference in current density in the zone where a reduction peak appeared (P.Re.). Bearing in mind both test conditions, this increase could only be due to the reduction in O 2 (because N 2 is an inert gas). This agrees with many authors in the literature who propose that O 2 reduction in SS is associated with the zones in which Fe(II) compounds act as absorption centers [6,21,[26][27][28].
The reaction of the reduction in the O 2 dissolved on the SS surface can take place with four (1) or two (2) electrons [17,18,26,27,37,38].
Which reaction predominates apparently depends on the electrode's surface composition and also on the conditions it is under [6,37,38].
Which reaction predominates apparently depends on the electrode's surface composition and also on the conditions it is under [6,37,38].   Figure 5a,c. The j/j0 logarithmic value is represented on the Yaxis, where j0 is the reference current density of 1 A/cm 2 and j is the density obtained at each potential applied in A/cm 2 .    Figure 5a,c. The j/j 0 logarithmic value is represented on the Y-axis, where j 0 is the reference current density of 1 A/cm 2 and j is the density obtained at each potential applied in A/cm 2 .
With these representations, the following can be stated: • The open circuit potential (OCP) shifts to the left if the atmosphere is aired.

•
In this representation, two straight-lined sections appear. These lines are associated with O 2 , H 2 O 2 , and H 2 O reduction processes. The following can be deduced from the study: • The first straight line is that which varies the most with changes in O 2 availability, and the value of its slope in the airless atmosphere goes from −2.630 V −1 to −4.272 V −1 ( Table 5). The variation in the intercept is slight. This variation in the representation of potentials versus the SCE is 6.8% and 4.5% versus the OCP (Table 5). These minor variations do not enable us to establish significant differences between both states.

•
The second line is practically horizontal. The value of this line's slope seems to be independent of the solution's airing (Table 5). This plateau aspect indicates that it is due to a limitation in the speed of the material's transport process.  Figure 5b,c for the airless and aired atmospheres.

Line 1 Line 2
Line Range m1 n1 Line Range m2 n2 Airless atmosphere Parameter m (slopes of straight lines) is an interesting parameter because it does not depend on either the reference being used in the technique or the electrode's surface. This means that it is very efficient as a comparative parameter among several tests run with different sensors (demonstrated in Appendix A).

Influence of pH on the SS Electrode
Poorly soluble carbonates form when concrete is carbonated, mainly by the reaction with calcium hydroxide, which lowers the hydroxide concentration in the concrete pore solution and, therefore, also its pH, with pH values coming close to neutrality (pH ≈ 8) [34].
According to the study performed by Babic and Metikos-Hukovic [26], pH influences the surface conditions of SS, hence the importance of studying this influence because the electrode's surface conditions impact O 2 reduction owing to passive oxide layer formation [6,37,38].
In order to assess how variations in pH influence the sensor's response, Figure 6 compares the results obtained with the solutions that simulated the three different pH conditions of the concrete pore solution: • Normal conditions (pH 13 and pH 12.72).

•
The initial carbonation state (pH of 11.45). • Carbonated concrete (pH 8.35). With these representations, the following can be stated: • The open circuit potential (OCP) shifts to the left if the atmosphere is aired.

•
In this representation, two straight-lined sections appear. These lines are associated with O2, H2O2, and H2O reduction processes. The following can be deduced from the study: o The first straight line is that which varies the most with changes in O2 availability, and the value of its slope in the airless atmosphere goes from −2.630 V −1 to −4.272 V −1 ( Table 5). The variation in the intercept is slight. This variation in the representation of potentials versus the SCE is 6.8% and 4.5% versus the OCP (Table 5). These minor variations do not enable us to establish significant differences between both states. o The second line is practically horizontal. The value of this line's slope seems to be independent of the solution's airing (Table 5). This plateau aspect indicates that it is due to a limitation in the speed of the material's transport process.
Parameter m (slopes of straight lines) is an interesting parameter because it does not depend on either the reference being used in the technique or the electrode's surface. This means that it is very efficient as a comparative parameter among several tests run with different sensors (demonstrated in Appendix A).

Influence of pH on the SS Electrode
Poorly soluble carbonates form when concrete is carbonated, mainly by the reaction with calcium hydroxide, which lowers the hydroxide concentration in the concrete pore solution and, therefore, also its pH, with pH values coming close to neutrality (pH ≈ 8) [34].
According to the study performed by Babic and Metikos-Hukovic [26], pH influences the surface conditions of SS, hence the importance of studying this influence because the electrode's surface conditions impact O2 reduction owing to passive oxide layer formation [6,37,38].
In order to assess how variations in pH influence the sensor's response, Figure 6 compares the results obtained with the solutions that simulated the three different pH conditions of the concrete pore solution: • Normal conditions (pH 13 and pH 12.72).

•
The initial carbonation state (pH of 11.45).  Figure 6a shows how the sensor's response changes with pH. As pH lowers (with pH close to neutrality), the shown peaks in both the anodic and cathodic sweepings shift to  Figure 6a shows how the sensor's response changes with pH. As pH lowers (with pH close to neutrality), the shown peaks in both the anodic and cathodic sweepings shift to the right and their current density (j) lowers. This agrees with the study on SS performed at different pHs by Babic and Metikos-Hukovic [26].
As medium basicity diminishes, the OHconcentration drops, Fe(OH) 2 formation is inhibited and, therefore, the current density of the associated peak lowers.
Bearing in mind that some authors believe the O 2 reduction occurs on Fe(II) compound layers, the alteration to these layers that results from varying pH may influence the measurement taken by the sensor to characterize O 2 availability. Figure 6b and Table 6 show the difference in the four pH values studied by means of the logarithmic representation. In all cases, line 2 is seen as a practically constant horizontal line with a similar intercept value for this line. We can see that the slope values for line 1 are similar. Therefore, we can state in a first instance that varying pH would not significantly affect identifying the availability of O 2 and humidity. We also observe how at the start of the curve associated with H 2 O reduction, less negative potentials appear as pH lowers, which is consistent with the H 2 O pourbaix.

Influence on the Electro-Chemical Behavior of Other Chemical Species
In concrete pore solution, we find both hydroxides and carbonates, and occasionally chlorides in aggressive environments. Another interesting reaction from the mechanistic point of view is that H 2 O 2 as the O 2 reduction reaction on the SS surface can occur with two or four electrons to generate OH − or H 2 O 2 [17,18,26,27,37,38]. Depending on the electrolyte conditions and the metal surface, one product or another being produced is favored. With H 2 O 2 , its presence can affect O 2 reduction on the steel surface [17,27,37,38].
This section aims to identify if the presence of carbonates, chlorides, and H 2 O 2 influence in the O 2 reduction on the steel surface , and the solutions were: From the results in Figure 7, we note how the presence of carbonates and chlorides barely affected the results obtained with the SS electrode, provided that pH is not affected.
According to the study by Correia, et al. [6], the surface conditions of steel embedded in concrete can be affected by the presence of Cl anions. Figure 8a shows the voltagrams obtained with the stainless steel (SS) electrode at pH 12.72 (solution 0.1 m NaOH) to reduce dissolved oxygen (O 2 ) before (voltagram in black) and after adding (in green) 0.02 m of H 2 O 2 to the system. The voltammetry obtained when H 2 O 2 is lacking, which shows that reduced O 2 is associated with peak B, appears at this pH at the potential of −0.78 V vs. SCE. This peak is followed by a shoulder of slight current density (peak C) at a maximum potential of −1 V vs. SCE. During anodic sweeping, D peaks appear (−0.74 V vs. SCE), as does a subsequent E shoulder at approximately −0.6 V vs. SCE. Peak D may be related to the oxidation process of the species obtained after reducing molecular O 2 (peak B). According to the study by Correia, et al. [6], the surface conditions of steel embedded in concrete can be affected by the presence of Cl anions. Figure 8a shows the voltagrams obtained with the stainless steel (SS) electrode at pH 12.72 (solution 0.1 m NaOH) to reduce dissolved oxygen (O2) before (voltagram in black) and after adding (in green) 0.02 m of H2O2 to the system. The voltammetry obtained when H2O2 is lacking, which shows that reduced O2 is associated with peak B, appears at this pH at the potential of -0.78 V vs. SCE. This peak is followed by a shoulder of slight current density (peak C) at a maximum potential of −1 V vs. SCE. During anodic sweeping, D peaks appear (-0.74 V vs. SCE), as does a subsequent E shoulder at approximately -0.6 V vs. SCE. Peak D may be related to the oxidation process of the species obtained after reducing molecular O2 (peak B).
The addition of H2O2 to the studied system brought about a morphological variation in the voltagram with a new peak appearing (A), as well as a marked modification in the currents recorded for the other peaks.
It is well known that H2O2 can act as a very weak acid (pKa = 11.6) to generate the perhydroxyl anion (HO 2 − ) and to release protons to the medium. Under our working conditions (pH 12.72), the non-dissociated H2O2 concentration was only 9.5% of the 0.02 m added to the system, while the remaining 90.5% came in the form of HO 2 − . It is also known that on surfaces like that of iron, this species' decomposition is catalyzed in molecular O2 and H2O. The voltagram obtained by adding H2O2 to the system (Figure 8a, green line) seems to fall in line with the above reasoning. We associate shoulder A, which came before the B peak that we related to the reduction in molecular O2, with the reduction in non-dissociated H2O2. Peak B current density significantly increased due to the O2 generated by the HO 2 − dismutation catalyzed by the sensor's surface. Under these conditions, the current density of the third peak (C) also significantly increased. This was associated with the bioelectronic HO 2 − reduction in the presence of one H2O molecule, which led to the formation of species OH -. When the anodic cycle was run, peak E disappeared in the presence of H2O2 and peak D intensified, possibly because the processes overlapped, which would mask the slower process associated with peak E. The way the system behaved in the CV study is reflected in the logarithmic representations of the current's intensity vs. the potential. The system's OCP value significantly shifted, new slopes appeared (Figure 8b), and the slope m1′s absolute value increased when H2O2 was added (Table 7). This higher value indicates the increased presence of O2, a finding that falls in line with what we discuss. O2 availability increased due to the greater HO 2 − dismutation caused by this species increasing when H2O2 was added to the solution, as discussed.

Phase 2: Studies Performed with the Sensor Embedded in Concrete
The objective of the analysis carried out in this study phase was to understand how the sensor embedded in concrete works, to know its limitations, to define simple correlations, and to identify the sensor's discrimination potential.
The article by Correia, et al. [6] mentions how the passive oxide layer that forms in the SS embedded in concrete differs from that produced in solution. Hence the obtained voltagrams differ even under similar pH conditions, as seen in the voltagram shown in Figure 9a (black solid line). In concrete in the non-saturated H2O state, we cannot distinguish the peaks associated with the formation of Fe(OH)2 and FOOH layers.
The cited article [6] explains how the passive oxide layer that forms on the surface of the sensor embedded in concrete favors O2 reduction. The cited authors state that a bigger metal surface covered by Fe(II), and not by solution, probably exists.
In the voltagrams in Figure 9a, obtained in our study, we see how current density diminishes on the cathodic sweeping curve under the 500 mbar and VAC conditions, i.e., when O2 availability is reduced. The range of potentials within which this decrease took place coincided with the affected zone in the voltagrams in solution by varying O2 availability ( Figure 5). This would support what the article by Correia, et al. sets out [6]. As the humidity conditions also remain practically constant between both states as temperature does, this diminished current density would be related to reduced O2 availability. The The addition of H 2 O 2 to the studied system brought about a morphological variation in the voltagram with a new peak appearing (A), as well as a marked modification in the currents recorded for the other peaks.
It is well known that H 2 O 2 can act as a very weak acid (pKa = 11.6) to generate the perhydroxyl anion (HO 2 − ) and to release protons to the medium. Under our working conditions (pH 12.72), the non-dissociated H 2 O 2 concentration was only 9.5% of the 0.02 m added to the system, while the remaining 90.5% came in the form of HO 2 − . It is also known that on surfaces like that of iron, this species' decomposition is catalyzed in molecular O 2 and H 2 O.
The voltagram obtained by adding H 2 O 2 to the system (Figure 8a, green line) seems to fall in line with the above reasoning. We associate shoulder A, which came before the B peak that we related to the reduction in molecular O 2 , with the reduction in non-dissociated H 2 O 2 . Peak B current density significantly increased due to the O 2 generated by the HO 2 − dismutation catalyzed by the sensor's surface. Under these conditions, the current density of the third peak (C) also significantly increased. This was associated with the bioelectronic HO 2 − reduction in the presence of one H 2 O molecule, which led to the formation of species OH − . When the anodic cycle was run, peak E disappeared in the presence of H 2 O 2 and peak D intensified, possibly because the processes overlapped, which would mask the slower process associated with peak E. The way the system behaved in the CV study is reflected in the logarithmic representations of the current's intensity vs. the potential. The system's OCP value significantly shifted, new slopes appeared (Figure 8b), and the slope m1's absolute value increased when H 2 O 2 was added (

Phase 2: Studies Performed with the Sensor Embedded in Concrete
The objective of the analysis carried out in this study phase was to understand how the sensor embedded in concrete works, to know its limitations, to define simple correlations, and to identify the sensor's discrimination potential.
The article by Correia, et al. [6] mentions how the passive oxide layer that forms in the SS embedded in concrete differs from that produced in solution. Hence the obtained voltagrams differ even under similar pH conditions, as seen in the voltagram shown in Figure 9a (black solid line). In concrete in the non-saturated H 2 O state, we cannot distinguish the peaks associated with the formation of Fe(OH) 2 and FOOH layers. Moreover, the logarithmic representation of the voltagrams (Figure 9b) helps to distinguish the two lines observed in solution which, as shown by a former study, have been associated with a reduction in the O2, H2O2, and H2O systems.
In the study in solution, line 1 is related to dissolved O2 availability, whereas line 2 is practically horizontal during the H2O reduction process. In the voltagrams presented in Figure 9, line 2 presents no horizontal trend section because we do not contemplate a current saturation case caused by the transport of H2O molecules towards the electrode. Figure 10a,c,e shows the voltagrams obtained in both the saturated and non-saturated state ATM of the concrete pore system for one concrete sample w/c = 0.6, one concrete sample w/c = 0.5, and another sample w/c = 0.4. As shown, when the capillary system is under H2O saturation conditions, the peaks associated with the reduction in Fe(OH)2 and FeOOH layers (around -0.7 V) can be seen, especially in less porous concrete. Under the H2O-saturated concrete sample conditions, the conditions of the sensor's environment are similar to the study in solution conditions, which is why the same peaks are noted.
The logarithmic representation (Figure 10b,d,f) of samples in the H2O saturation state shows that line 2 is not horizontal, but its slope tends to diminish. The same can be stated of line 1. This indicates less O2 availability according to that studied in solution.
In order to assess the sensor's effectiveness in distinguishing the different O2 availability conditions by simple data processing techniques, the average data obtained from the slopes and intercepts of the lines associated with O2 availability and humidity per The cited article [6] explains how the passive oxide layer that forms on the surface of the sensor embedded in concrete favors O 2 reduction. The cited authors state that a bigger metal surface covered by Fe(II), and not by solution, probably exists.
In the voltagrams in Figure 9a, obtained in our study, we see how current density diminishes on the cathodic sweeping curve under the 500 mbar and VAC conditions, i.e., when O 2 availability is reduced. The range of potentials within which this decrease took place coincided with the affected zone in the voltagrams in solution by varying O 2 availability ( Figure 5). This would support what the article by Correia, et al. sets out [6]. As the humidity conditions also remain practically constant between both states as temperature does, this diminished current density would be related to reduced O 2 availability. The voltagrams and cathodic Tafel representations in Figure 9 present the same morphology and effect with the O 2 availability reductions on the three studied concretes due to the figure showing the results obtained in one of the w/c = 0.6 samples as an example.
Moreover, the logarithmic representation of the voltagrams (Figure 9b) helps to distin-guish the two lines observed in solution which, as shown by a former study, have been associated with a reduction in the O 2 , H 2 O 2 , and H 2 O systems. In the study in solution, line 1 is related to dissolved O 2 availability, whereas line 2 is practically horizontal during the H 2 O reduction process. In the voltagrams presented in Figure 9, line 2 presents no horizontal trend section because we do not contemplate a current saturation case caused by the transport of H 2 O molecules towards the electrode. Figure 10a,c,e shows the voltagrams obtained in both the saturated and non-saturated state ATM of the concrete pore system for one concrete sample w/c = 0.6, one concrete sample w/c = 0.5, and another sample w/c = 0.4. As shown, when the capillary system is under H 2 O saturation conditions, the peaks associated with the reduction in Fe(OH) 2 and FeOOH layers (around −0.7 V) can be seen, especially in less porous concrete. Under the H 2 O-saturated concrete sample conditions, the conditions of the sensor's environment are similar to the study in solution conditions, which is why the same peaks are noted.
tinguish the two lines observed in solution which, as shown by a former study, have been associated with a reduction in the O2, H2O2, and H2O systems.
In the study in solution, line 1 is related to dissolved O2 availability, whereas line 2 is practically horizontal during the H2O reduction process. In the voltagrams presented in Figure 9, line 2 presents no horizontal trend section because we do not contemplate a current saturation case caused by the transport of H2O molecules towards the electrode. Figure 10a,c,e shows the voltagrams obtained in both the saturated and non-saturated state ATM of the concrete pore system for one concrete sample w/c = 0.6, one concrete sample w/c = 0.5, and another sample w/c = 0.4. As shown, when the capillary system is under H2O saturation conditions, the peaks associated with the reduction in Fe(OH)2 and FeOOH layers (around -0.7 V) can be seen, especially in less porous concrete. Under the H2O-saturated concrete sample conditions, the conditions of the sensor's environment are similar to the study in solution conditions, which is why the same peaks are noted.
The logarithmic representation (Figure 10b,d,f) of samples in the H2O saturation state shows that line 2 is not horizontal, but its slope tends to diminish. The same can be stated of line 1. This indicates less O2 availability according to that studied in solution.
In order to assess the sensor's effectiveness in distinguishing the different O2 availability conditions by simple data processing techniques, the average data obtained from the slopes and intercepts of the lines associated with O2 availability and humidity per dosage and for each studied condition were compared (Figure 11,13).  Figure 11 depicts the value of the slope (m1) and the intercept (n1) for each studied dosage. The m1 values show a clear correlation between O2 availability and the slope obtained with the cathodic Tafel line. When examining the standard deviations, the results are not conclusive due to overlapping, but when we contemplate the minimum and maximum of the clearance range marked by them, we find that the correlation between O2 The logarithmic representation (Figure 10b,d,f) of samples in the H 2 O saturation state shows that line 2 is not horizontal, but its slope tends to diminish. The same can be stated of line 1. This indicates less O 2 availability according to that studied in solution.
In order to assess the sensor's effectiveness in distinguishing the different O 2 availability conditions by simple data processing techniques, the average data obtained from the slopes and intercepts of the lines associated with O 2 availability and humidity per dosage and for each studied condition were compared (Figures 11 and 13).
(e) (f)  Figure 11 depicts the value of the slope (m1) and the intercept (n1) for each studied dosage. The m1 values show a clear correlation between O2 availability and the slope obtained with the cathodic Tafel line. When examining the standard deviations, the results are not conclusive due to overlapping, but when we contemplate the minimum and maximum of the clearance range marked by them, we find that the correlation between O2 availability and the slope remains. The data obtained for the non-saturated H2O samples (VACUUM, 500 mbar, and ATM) reveal that the parameter m1 value rises as O2 availability increases for all concretes. The highest absolute parameter m1 value is obtained in all concretes for higher O2 availa-  Figure 11 depicts the value of the slope (m1) and the intercept (n1) for each studied dosage. The m1 values show a clear correlation between O 2 availability and the slope obtained with the cathodic Tafel line. When examining the standard deviations, the results are not conclusive due to overlapping, but when we contemplate the minimum and maximum of the clearance range marked by them, we find that the correlation between O 2 availability and the slope remains.
The data obtained for the non-saturated H 2 O samples (VACUUM, 500 mbar, and ATM) reveal that the parameter m1 value rises as O 2 availability increases for all concretes. The highest absolute parameter m1 value is obtained in all concretes for higher O 2 availability (ATM conditions), while the lowest absolute value is obtained under the VAC conditions (VAC, [O 2 ] ≈ 0). This is consistent with that observed in solution. We also note that, from the results obtained in water-accessible porosity and mercury porosity (Table 4), the absolute parameter value for all the O 2 levels in the less porous concrete (w/c = 0.4) is lower than that obtained in more porous concretes (w/c = 0.6 and w/c = 0.5) (Figure 11a).
When studying the values of the slopes of lines m1 and m2 with the open porosity results according to MIP (Figure 12), a direct correlation appears. The greater slopes correspond to higher air porosities. This makes sense in light of the parameter m1 analysis, in which we identify a higher absolute value for this parameter with greater O 2 availability. . This is consistent with that observed in solution. We also note that, from the results obtained in water-accessible porosity and mercury porosity (Table 4), the absolute parameter value for all the O2 levels in the less porous concrete (w/c = 0.4) is lower than that obtained in more porous concretes (w/c = 0.6 and w/c = 0.5) (Figure 11a). When studying the values of the slopes of lines m1 and m2 with the open porosity results according to MIP (Figure 12), a direct correlation appears. The greater slopes correspond to higher air porosities. This makes sense in light of the parameter m1 analysis, in which we identify a higher absolute value for this parameter with greater O2 availability. For the saturated state, we observe that the value obtained for slope m1 (Figure 11a) is similar for all concretes (m1 = 2.53 ± 0.14 V -1 ), which might be due to the O2 concentration in H2O having a certain equilibrium value determined by the temperature of the system which, having lowered, must be transported from the outside by a winding network of saturated solvent capillaries. For the saturated state, we observe that the value obtained for slope m1 (Figure 11a) is similar for all concretes (m1 = 2.53 ± 0.14 V −1 ), which might be due to the O 2 concentration in H 2 O having a certain equilibrium value determined by the temperature of the system which, having lowered, must be transported from the outside by a winding network of saturated solvent capillaries. Figure 11a depicts how for saturated concrete, the slope is practically the same for the three concretes. This agrees with the O 2 concentration being the same regardless of the porosity-independent one (because gas solubility in H 2 O only depends on temperature). Although porosity affects the H 2 O/metal contact surface, parameter m1 does not depend on the electrode's surface (Appendix A).
The results of the intercept of line 1 (n1), which are provided in Figure 11b, have similar values for all the studied cases, which allows us to consider that this value could be related to the exchange kinetics parameters of the metal/O 2 system corrosion process.
The interpretation of the results obtained from the line 2 analysis is complex. Figure 13a shows a similar behavior to that noted for line 1. The slope of the second section depends on O 2 availability, while the intercept can practically be considered to take a constant value. Nevertheless, the tests in solution show that line 2 is practically horizontal and does not vary with variations in O 2 availability. A proposal to justify the horizontal section observed when O 2 in solution was reduced was previously mentioned. As H 2 O is now no longer the majority agent of the capillary system, we cannot expect it to impose its kinetic behavior on the system. The argument to justify the observed facts means having to revise the reaction mechanisms set out in the bibliography. It is accepted that the O2 reduction reaction occurs by means of a sequence of different stages. For neutral or acidic media, the most frequently proposed process is that described in Figure 14a. As we are dealing with a basic medium (the pore solution pH of a recent concrete is higher than or equals 12.5), the reaction we contemplate is that illustrated in Figure 14b, instead of H2O2 appearing and, bearing in mind that proton availability would be low, the perhydroxyl species appearing. We propose that the slope of line 2 seen in the Tafel representation depends on the second reaction process, during which the perhydroxyl radical reacts with two electrons and one H2O molecule to give three hydroxide anions.  It is accepted that the O 2 reduction reaction occurs by means of a sequence of different stages. For neutral or acidic media, the most frequently proposed process is that described in Figure 14a. As we are dealing with a basic medium (the pore solution pH of a recent concrete is higher than or equals 12.5), the reaction we contemplate is that illustrated in Figure 14b, instead of H 2 O 2 appearing and, bearing in mind that proton availability would be low, the perhydroxyl species appearing. We propose that the slope of line 2 seen in the Tafel representation depends on the second reaction process, during which the perhydroxyl radical reacts with two electrons and one H 2 O molecule to give three hydroxide anions.
In accordance with the outline shown in Figure 14b, the slope of the second section depends on O 2 availability, as noted from the data in Table 5. When the porous concrete system is saturated, O 2 availability is low because of scarce O 2 solubility in H 2 O and the slow speed of transporting O 2 from the outside to the surface of the metal embedded in the material. Thus, the second process occurs to a lesser extent than when the porous system contains only residual humidity. stages. For neutral or acidic media, the most frequently proposed process is that described in Figure 14a. As we are dealing with a basic medium (the pore solution pH of a recent concrete is higher than or equals 12.5), the reaction we contemplate is that illustrated in Figure 14b, instead of H2O2 appearing and, bearing in mind that proton availability would be low, the perhydroxyl species appearing. We propose that the slope of line 2 seen in the Tafel representation depends on the second reaction process, during which the perhydroxyl radical reacts with two electrons and one H2O molecule to give three hydroxide anions. In accordance with the outline shown in Figure 14b, the slope of the second section depends on O2 availability, as noted from the data in Table 5. When the porous concrete system is saturated, O2 availability is low because of scarce O2 solubility in H2O and the slow speed of transporting O2 from the outside to the surface of the metal embedded in the material. Thus, the second process occurs to a lesser extent than when the porous system contains only residual humidity.
Therefore, the slope of line m2 diminishes, which also occurs with the slope of line 1, but without the horizontality that takes place when testing in solution. This last point is logical because, in the potentials region, H2O and O2 reduction processes overlap and pore saturation does not suffice to impose H2O reduction kinetics on HO 2 − species reduction kinetics. The slope m2 value is higher with increasing porosity (Figure 13a). The higher porosity is, the more favorable diffusion phenomena are. This fact allows the kinetics of reactions, reagents, and products to move more easily. Therefore, the stages 2, 3, and 4 of the O2 reduction reaction are more favored the higher the porosity is.
The interpretation of the intercept of line 2 (n2) (Figure 13b) is more difficult because it must correspond to a commitment dynamics state situation between the speed with Therefore, the slope of line m2 diminishes, which also occurs with the slope of line 1, but without the horizontality that takes place when testing in solution. This last point is logical because, in the potentials region, H 2 O and O 2 reduction processes overlap and pore saturation does not suffice to impose H 2 O reduction kinetics on HO 2 − species reduction kinetics.
The slope m2 value is higher with increasing porosity (Figure 13a). The higher porosity is, the more favorable diffusion phenomena are. This fact allows the kinetics of reactions, reagents, and products to move more easily. Therefore, the stages 2, 3, and 4 of the O 2 reduction reaction are more favored the higher the porosity is.
The interpretation of the intercept of line 2 (n2) (Figure 13b) is more difficult because it must correspond to a commitment dynamics state situation between the speed with which the HO 2 − species is generated and the dismutation process that competes with the intermediate reduction to be transformed into OH − . As with n1, its value is similar for all states, but is slightly lower when concrete samples are H 2 O-saturated.

Discussion
Although it is true that the standard deviation of the m1 and m2 parameters among the various O 2 availabilities is high, the little discrimination power can be solved by data analysis techniques following work protocols (RNA, PCA, and PLS). We can establish more accurate models to quantify O 2 and humidity availability inside a hardened concrete matrix. Moreover, with these methods, we can include as many study variables as we wish. Furthermore, if we consider that the sensor responds to the presence of chlorides and carbonates, we could study their feasibility for also detecting the presence of Cl − and concrete carbonation, which are the main precursor agents of steel rebar depassivation.
Furthermore, voltammetric SS sensors will allow us to define more economic sensor networks than one of noble metals. They can be made with a bigger surface which would provide a more robust response. We could install more control points which would allow more information to be collected because the cost of the material would lower, allowing us to intensify the statistical control which is a key factor to achieving success in strategies such as the SHM (Structural Health Monitoring). Moreover, being able to present resistance to different aggressive environments and respond to changes in the concrete matrix which compromise structures' durability (variation in O 2 availability, presence of chlorides, and concrete carbonation) offer a very high potential for monitoring structures.
According to this, the voltammetric SS sensors could be suitable to develop sensor systems to evaluate the initiation and propagation period of the rebars corrosion processes in the reinforcement of concrete structures.

Conclusions
Firstly, in accordance with the obtained results, parameters m1 and m2, the slopes of the lines identified logarithmically in the voltagram for the cathodic sweeping obtained with the SS sensor, are related to the different O 2 reduction stages on the sensor's surface. This allows us to assess the availability of both O 2 and humidity in hardened concrete matrices.
Secondly, the parameters m1 and m2 present a direct correlation with the results of the standardized porosity tests and, with them, these concrete characterization parameters can be assessed in situ in RCS.
Thirdly, the voltametric behavior observed in dissolution agreed with the proposed O 2 mechanism in Figure 14a.
Fourthly, the behavior of the sensor embedded in concrete is similar to the behavior in alkaline dissolution, but considering the differences due to the diffusion phenomena, limitations or transport limitations occurred in hardened concrete matrix.
Finally, as was demonstrated in the dissolution study, the SS sensor is sensitive to the presence of chlorides and carbonates. In Equation (A1) we can see how the parameter does not depend on the electrode's surface.
In order to see how this parameter does not depend on the reference, we considered expressing our results in accordance with a second reference (Ref2). To do so, we had to express the potentials referring to our Ref1 in relation to Ref 2, which was done by subtracting the difference of the existing potential measured between Ref1 and Ref2, by connecting Ref2 to the multimeter negative pole (or COM) from the potential expressed in Ref1. As Equation (A3) remains the same as Equation (A1), it shows that the parameter did not depend on either the reference with which we expressed the potentials or the electrode's surface. However, as the parameter depended directly on the intensities ratio, we were unable to compare the parameters obtained with the different electrode configurations because field intensities differed. In other words, we cannot compare the measurements taken with the 2-electrode configuration to those obtained with the 3-electrode ones.