Research Progress on the Stability and Durability of Ag/AgCl Prepared by Anodic Chlorination Method for Chloride Ion Sensors in Cement-Based Materials

Abstract

A strong application potential for Ag/AgCl ion-selective electrodes (ISEs) used as chloride sensors in cement-based material is widely accepted, but their stability and durability have not been sufficiently addressed. This paper summarizes the research status of the stability and durability of Ag/AgCl ISEs used for the non-destructive detection of chloride in cement-based materials. Four topics including working principle and fabrication methods, the factors that influence stability, research status for stability and durability studies, and the reason for durability failure of Ag/AgCl ISEs in cement-based materials are reviewed. Meanwhile, the improving methods for Ag/AgCl ISEs are proposed based on discussions of various aspects of Ag/AgCl ISEs.

1. Introduction

Reinforced concrete as the primary structure material in the marine environment usually suffers from a serious risk of chloride-induced corrosion leading to premature durability deterioration. Once the chloride content surrounding steel achieves the so-called critical chloride concentration (639–6745 mg/L reported in [1,2]) for the corrosion initiation of steel, the localized depassivation of steel passive film initiates, followed by serious corrosion [3]. In general, the original chloride content in concrete is low, owing to the maximum limit of chloride content in raw materials, which benefits from restrictions of local relevant codes and standards. Menacing chloride regarding steel corrosion is from the environment [4,5,6]. The accumulation of chloride ions around the steel in concrete is the result of chloride penetration processes (including capillary absorption, diffusion, permeation, and convection [4,7]). Hence, to understand the deterioration process of concrete structures, monitoring the variation in chloride concentration is helpful.

Detecting chloride levels in concrete often involves a range of complicated methods (i.e., the Volhard method [8], potentiometric methods [9,10,11], ion chromatography method [12], chemical titration [13], and ion chromatography [14]), which are destructive, time-intensive, and labor-intensive, thereby rendering them unsuitable for sustained chloride monitoring. Non-destructive techniques include laser-induced breakdown spectroscopy (LIBS) [15], potentiometry [16], terahertz frequency electromagnetic wave spectroscopy (near infrared [17], microwave [18], and millimeter wave), and the optical fiber sensing method [19]. LIBS employs laser ablation to examine plasma spectral lines to assess the total chloride ion content, instead of just free chloride ions [20]. Potentiometry assesses potential differences but encounters challenges with electromagnetic interference and electrode stability [21]. Chloride ions are detected by terahertz spectroscopy through unique absorption, but carbonation and environmental influences can interfere [22]. The most reported non-destructive method for chloride detection in concrete is the electrochemical method using an Ag/AgCl ion-select electrode (ISE) [23,24] because of its excellent chloride selectivity, high precision, and low cost in fabrication [8]. With regard to the application of Ag/AgCl, it has been widely utilized as a reference electrode in analytical chemistry [25,26], and its role as a chloride sensor in the concrete industry has been under study in recent years [27,28,29,30,31]. Relative attempts to embed Ag/AgCl electrodes into cement-based materials have been gradually reported since the 1990s [32,33]. Since then, the use of Ag/AgCl as an ion-selective electrode for detecting chloride has been confirmed in both simulated concrete pore solutions and concrete with varying chloride levels [24,34]. This promising technology is still under development, but it has already attracted the attention of scientists as it presents good application prospects for use as a chloride sensor in cement-based materials.

Numerous studies have assessed the effectiveness of the Ag/AgCl electrode as a chloride sensor in cement-based materials [35,36,37,38], and an agreed-upon conclusion is obtained that it is possible to monitor the in situ concrete in real time and uninterruptedly via this method. The response potential of Ag/AgCl, depending on chloride concentration, conforms to Nernst’s Equation well in the right working environment [8,23]. Several attempts have also been made to correlate testing potential with chloride concentration. However, testing results obtained through this method always have errors due to the interference of many factors, such as interfering ions, temperature–humidity, and fabrication drawbacks [39]. The stability of Ag/AgCl ISEs used to monitor chloride in concrete has not been sufficiently addressed, and the origin of the errors needs to be further classified. In addition, long-term durability is necessary for Ag/AgCl ISEs to be used as in situ and non-destructive monitoring sensors. However, the dissolution and exfoliation of AgCl films caused by high alkalinity and multi-interfering ions may induce premature failure. To better use Ag/AgCl ISEs for the in situ monitoring of concrete components, especially for long-term monitoring, the long-term performance stability and durability of Ag/AgCl ISEs need to be clarified. Although there are relevant reviews on the research of Ag/AgCl ISEs at present, such as [8,23], which reviews the performance of Ag/AgCl ISEs for a chloride sensor in concrete, so far, the obtained results in different references of Ag/AgCl ISEs vary greatly. Currently, there are relatively few articles on the influencing factors of the test stability and durability of Ag/AgCl ISEs.

This paper reviews the progress of research on the application of Ag/AgCl ISEs as chloride sensors in cement-based materials regarding stability and durability. At first, the fabrication and principle of Ag/AgCl as a chloride sensor are illustrated. In addition, factors influencing Ag/AgCl ISE testing potential are analyzed, such as temperature, interfering ions, alkalinity, and AgCl film characterization. Finally, the sources of durability and measurement error are discussed. Herein, methods to improve the stability, durability, and accuracy of the Ag/AgCl ISE are recommended. The flow chart for this paper is shown in Figure 1.

2. Working Principle of Ag/AgCl as ISEs

Commonly used Ag/AgCl ISEs are solid-state membrane electrodes consisting of a fundamental silver core and a micron-thick layer of AgCl on the surface. When Ag/AgCl ISEs contact an aqueous solution, e.g., a simulated pore solution and a concrete/cement pore solution, a small amount of chloride from the AgCl film will diffuse into the solution. Meanwhile, some ions in the solution will absorb onto the Ag surface through the micro-channels over the AgCl film, as shown in Figure 2. As thus, the electrode surface is negatively charged, and the solution surrounding the electrode is positively charged, forming a double electron layer on the interface. Taking into account the redox reaction occurring on the electrode surface, there is a definite relationship between the electrical signal formed by the potential difference and the chloride activity, as described by the Nikolsky–Eisenman Equation (Equation (1)) [40,41,42]:

$$E_{Ag/AgCl}=E_{0,Ag/AgCl}+{\displaystyle \frac{\mathrm{RT}}{\mathrm{F}}}\ln K_{sp}-{\displaystyle \frac{\mathrm{RT}}{\mathrm{F}}}\ln \left[a_{\mathrm{C}{l}^{-}}+{\displaystyle \sum _{j\ne \mathrm{C}{l}^{-}}\left(K_{C{l}^{-}j}^{pot}·a_{\mathrm{j}}^{-1/z_{\mathrm{j}}}\right)}\right]$$

(1)

where E_0,Ag/AgCl represents the standard potential of the Ag/AgCl ISE, which has been determined experimentally until 325 °C in references [43,44]; K_sp is the solubility product of AgCl; R represents the ideal gas constant (8.314 JK⁻¹·mol⁻¹); T denotes the absolute temperature (K); F stands for the Faraday constant (96,485.3 C·mol⁻¹); α_Cl⁻ stands for the chloride activity; the effect of interfering ions is expressed as the $K_{C{l}^{-}j}^{pot}$ term for ion j; and the influence of common interfering ions in concrete on Ag/AgCl ISEs is evaluated in Section 4.2.

Since the Ag/AgCl ISE has a significantly stronger response to chloride ions compared to other ions, the test potential, namely the open circuit potential (OCP), can be expressed by the following Equation (2) [45]:

$$E_{\mathrm{measured}}=E_{0,Ag/Ag+}+{\displaystyle \frac{\mathrm{RT}}{\mathrm{F}}}\ln C_{C{l}^{-}}+E_{junction}$$

(2)

where E_measured is the measured potential of the Ag/Ag+ electrode, which is the mixed potential owned by a variety of silver salts formed on the electrode surface. E_junction is the liquid junction potential. C_Cl⁻ is the solubility product of AgCl. Equation (3) means that for determining the OCP of the Ag/AgCl ISEs at elevated temperatures, it is necessary to know the standard potential and the chloride activity in the electrolyte. Given that chloride activity is indicated by concentration, the electrode potential shows a proportional relationship with the logarithm of chloride concentration. The potential response of Ag/AgCl ISEs shows a Nernst behavior when the change in chloride activity in the medium affects the capacitance balance of the Ag/AgCl interface. The change in potential can therefore reflect the chloride content [46].

3. Fabrication Methods of Ag/AgCl ISEs

3.1. Anodic Chlorination

The fabrication process of Ag/AgCl ISEs refers to forming a uniform AgCl film adhered to the surface of the silver substrate. Anodic chlorination is commonly used, which performs with a constant voltage or constant current utilizing a three-electrode system in an aqueous chloride electrolyte. Under the action of an electric current, the formation of an AgCl film is a typical electroadsorption process, as shown in Figure 3. First, a single layer of AgCl grains is uniformly adsorbed on the surface of the silver substrate. Then, the multi-layer AgCl is formed through three-dimensional nucleation [47]. Micro-channels between AgCl grains provide the ion channel for chloride. Finally, a single layer of small AgCl grains is formed between the AgCl film and the silver substrate. The growth rate of the AgCl film during the anodic chlorination process is related to the ionic conductivity of micro-channels in the AgCl film [2]. The electrical conductivity of AgCl grains is minimal, and the overall conductivity of the AgCl film relies on the pores or micro-channels between these grains. The conductivity data for AgCl film range from 10⁻¹¹ to 10⁻⁷ Ω⁻¹·cm⁻¹, which is scattered due to the micro-channel conditions determined by the difference in fabrication regime and method [48].

Generally, Ag/AgCl ISEs prepared by a constant current are better compared to constant voltage in terms of stability [49]. The adopted current density determines the size of AgCl grains and the porosity of the AgCl thin layer, thus directly affecting the adsorption capacity of AgCl during chlorination [40,50]. A relatively lower current will induce a more compact AgCl film; moreover, the surface micro-channel distribution is more uniform. Inversely, the microstructure of the AgCl film prepared with a higher current density is much looser, and the bonding between the AgCl film and silver substrate is weak, as shown in Figure 3.

Additionally, with an increase in the applied current density, the thickness of the AgCl film increases linearly [51]. There is a recorded linear relationship with a slope of 2.67 between the thickness of the AgCl film and the current density [46]. The commonly used current density for preparing Ag/AgCl ISEs is presented in the literature [8].

The most common method for creating Ag/AgCl ISEs is electrochemical deposition, where the thickness of the AgCl film can be adjusted by altering the deposition current density and duration. The thickness (δ) of the AgCl film is calculated using Equation (3) [46,52]:

$$\delta ={\displaystyle \frac{I\cdot M\cdot t}{F\cdot \rho \cdot S}}={\displaystyle \frac{j\cdot M\cdot t}{F\cdot \rho }}$$

(3)

where S indicates the surface area of the Ag substrate in cm²; M is the molecular weight of AgCl, which is 143.32 g/mol; ρ is the density of AgCl, at 5.56 g/cm³; F is Faraday’s constant, which is 96,485.33 C/mol; I is the applied anodized current in milliamperes; j is the density of the anodized current applied; and t is anodization deposition time (s). The thickness of the AgCl film ranges from a few hundred nanometers to approximately 40 μm, with the most common thickness being between 2 and 12 µm [42,53].

3.2. Calibration

Ag/AgCl ISEs exhibit a well-defined linear response that correlates with the chloride activity in the solution. Thus, the chloride activity in the tested solution can be obtained by comparing the pre-calibrated potential diagram with the actual measured potential value [36]. The activity can be converted into concentration using the activity coefficient, and then the chloride concentration can be obtained, as given in Equation (4) [16,54]:

$${\alpha }_{C{l}^{-}}=C_{C{l}^{-}}\times {\gamma }_{C{l}^{-}}$$

(4)

where C_Cl⁻ is the chloride concentration and $\gamma$_Cl⁻$x=$ is the chloride activity coefficient. Note that it is not appropriate to assume that the concentration is equivalent to chloride activity, especially in a highly alkaline environment, as it would lead to a significant error [55]. However, it can be converted to concentration using the chloride activity coefficient [31]. Several attempts have been made to calculate chloride concentration based on the obtained potential. However, each electrode has a unique response curve despite being prepared in the same way in different papers [8,56,57]. Thus, to ensure the performance of Ag/AgCl ISEs or obtain the concentration of chloride ions, the calibration curve should be acquired. The Ag/AgCl ISE exhibits a Nernstian response to chloride ions when the concentration is 10⁻⁴ to 1 M. Notwithstanding these differences regarding the slopes of the response curves, the response potential and the concentration of the electrodes prepared by different researchers present a clear linear relationship in a specific range. The calibration curves’ slope of Ag/AgCl ISEs from 10⁻⁴ to 1 M of chloride ions [36,37,55,56,58,59,60,61,62,63,64,65,66] is shown in Figure 4.

It can be seen that the slope values are mainly concentrated between 55 and 60. A Quantile–Quantile plot illustrates that the slope has a normal distribution relationship, and the mode in the collected data is −57.5. Theoretically, this term is temperature dependent, and at a temperature of 25 °C, the Ag/AgCl ISE demonstrates a typical susceptibility of −59.16 [68]. The reasons for the different slopes of the check curve are related to the fabrication system, test temperature, interfering ions, etc., and are discussed in Section 5. Therefore, to further ensure the accuracy of the calculation results of the chloride concentration, the unique calibration curve of each electrode should be obtained by calibration before use.

4. The Stability of Ag/AgCl ISEs Used to Monitor Chloride in Concrete

4.1. Potential Interference

4.1.1. Liquid Junction Potential

Potential shifts in Ag/AgCl ISEs seem inevitable when used for in situ monitoring in cement-based material, which may cause considerable errors. For instance, Angst et al. [55] find that the concentrations obtained from the embedded sensors in mortar are ca. 25–70% higher than the external solution concentrations. Similarly, the results of the experiment and calculation according to Xu et al. [66] show that the error of the chloride diffusion coefficient measured by Ag/AgCl ISEs was ca. 30%. In terms of error sources for Ag/AgCl ISEs to monitor chloride concentration in concrete, the microstructure of AgCl films, depending on the fabrication methods, plays a crucial role in the test errors. Due to the incomplete coverage of the Ag substrate by deposited AgCl particles, which affects the mixed potential of Ag/AgCl ISEs, it is advised against using a thin AgCl film thickness below 1.34 µm, as reported in [53]. However, regarding electrodes having a thicker AgCl prepared with a higher current density, smaller-sized AgCl grains may be produced at the AgCl film, as well as the interface of the AgCl film and the silver substrate. The presence of inclusions increases ionic resistivity disproportionately to the AgCl film thickness and weakens the adhesion of the AgCl film to the Ag substrate [69]. Therefore, selecting and improving the fabrication process to improve the electrode performance is the basis of reducing the test error.

4.1.2. Diffusion Potential

In general, the calculated content based on Ag/AgCl ISEs is slightly lower than the actual chloride concentration, which is attributed to the diffusion potential, liquid junction potential, and membrane potential [31,70]. The potential readings of Ag/AgCl ISEs can be notably disrupted by diffusion potentials. To minimize errors related to diffusion potential, concentration gradients of any species, especially Cl⁻ and OH⁻, between the reference electrode and the chloride sensor should be avoided. To achieve this, reference electrodes should be placed at the same depth as the chloride sensors [69,71,72].

4.1.3. Membrane Potential

Differences in Cl⁻, OH⁻, or other ion concentrations between the chloride sensor and the reference electrode will cause membrane potentials. For concrete with a pH above 13.5, the membrane potentials resulting from practical chloride concentration variations are expected to be under 10 mV [42]. Ag/AgCl ISEs are in contact with the interfacial transition zone, inevitably leading to an error in the test value. The porosity in the interfacial transition zone is generally three times greater than the surrounding area [73], where ettringite and Ca(OH)₂ tend to form large crystals. Large amounts of Ca(OH)₂ are often observed around the electrodes. To reduce this error, the quality of the interface can be effectively controlled by applying a semi-permeable membrane that allows the free entry of ions and water.

4.2. Ion Interference

The activity of silver ions in a water solution determines the potential of Ag/AgCl ISEs, and this is inversely linked to chloride activity because of the solubility product. Any additional interfering species enabling the formation of an insoluble silver salt in sufficient concentration will influence the testing potential of Ag/AgCl ISEs. The effect of the interfering species on the Ag/AgCl ISE response can be schematically illustrated by dividing the diagram ISE potential versus activity of interfering species into three different regions, as shown in Figure 5. The Ag/AgCl ISE exhibits a stable potential in the ‘a’ region, where interference is minimal, and this potential is dictated by chloride ion activity, regardless of the interfering species [55,74]. In zone ‘b’, an increase in interfering species concentration modifies the potential response, which is now influenced by both chloride and the interfering species. This interference is due to the substitution of chloride on the ISE surface by the interfering species [41,75,76]. The surface distribution of precipitation between silver and interfering species can exhibit different phases due to varying concentrations of primary and interfering species [76]. The potential of the ISE in this zone depends on many factors, including the influence of time. At sufficiently high concentrations of interfering species (region “c” in Figure 5), the surface of the ISE is enveloped by the salt resulting from the reaction between silver and interfering species [76,77]. These influences are derived from the membrane potential and insoluble silver salt. Common cations (Li\Ca\Na\K\Mg) have little effect on electrode potential [78,79,80]. Most studies focus on anions such as sulfions, sulfate ions, halide ions, and hydroxide ions.

4.2.1. Sulfion and Sulfate Ions

The sulfate ion is a common ion in cement, but it also exists in soil, decaying organic matter, groundwater, seawater, and industrial waste water [81,82]. When Portland slag cement is used, the sulfion content stabilizes at 8–11 mM after hydration is basically completed, and the influence of the sulfion must be considered at that time [66,83]. As for laboratory tests, it is reported that sulfate ions (0.1 mol/L) have little influence on the chloride sensor in the simulated solution (0.2 mol/L NaOH, 0.6 mol/L KOH, saturated Ca(OH)₂) [80]. Jin et al. [24] studied the monitoring ability of the chloride sensor in a short period of time of about 100 days when mortar was subjected to a mass fraction of 7% sodium sulfate. The obtained results show that the potential increased slightly with the increase in exposure time, as shown in Figure 6. The attachment of Ag₂S to the electrode surface results in a significant deviation of the open-circuit potential [84,85]. It can be predicted that with the extension of the erosion age, erosion damage would worsen. Eventually, the composition and concentration of the pore solution will be changed, and the potential value monitored by the chloride sensor will eventually begin to decline.

4.2.2. Halide Ions

For the interference of iodide, selectivity coefficients ($K_{C{l}^{-I}}^{pot}$) varying from 2 to 360 have been reported [87]. It is the solution over a certain concentration that may lead to the formation of AgI crystals by consuming the Ag⁺ from Ag/AgCl [88]. The crystal layer on the Ag/AgCl ISE surface is partially changed into AgI over time. The conversion rate is influenced by both the thickness of the AgCl film and the iodide concentration in the solution. Moreover, time-dependent potentials are observed in solutions with iodide concentrations ranging from 10⁻⁴ to 10⁻³ M, and the potentials deviate from the linear approximation of the constant data over time. The presence of bromides results in a more negative potential for Ag/AgCl ISEs, which leads to an overestimation of chloride activity [37].

4.2.3. Hydroxide Ions

Cement-based materials usually display high alkalinity after hydration because of the generation of a large quantity of Ca(OH)₂ throughout the hydration process. Errors often appear in a low-concentration chloride environment when the pH ranges from 11.9 to 13.7 [36,60], leading to lower potentials than the theoretical values, which is related to easy OH⁻ adsorption on the Ag substrate [89]. Similar observations have been extensively reported in [90] that a pH value less than 12 has a negligible effect on Ag/AgCl ISEs. Still, the potential decreases distinctly by 50 mV for every pH increase of 1 once the pH value exceeds 13. The reasons explaining the above phenomenon are due to the formation of AgOH on the surface of Ag/AgCl ISEs, as shown in Figure 7. AgOH itself is not stable and can easily be converted into Ag₂O [40], as expressed by Equation (5):

$${2\mathrm{AgCl}+2\mathrm{OH}}^{-}\to {\mathrm{Ag}}_2{\mathrm{O}+2\mathrm{Cl}}^{-}{+\mathrm{H}}_2\mathrm{O}$$

(5)

When Ag₂O appears on the AgCl surface, the electrode potential reflected by the electrode includes the Ag-Ag₂O interface potential. The electrode potential with Ag/Ag₂O films also follows Nernst’s law, as shown in Equation (6):

$$E_{Ag/A{g}_{2}O}=E_{Ag/A{g}_{2}O}^0-{\displaystyle \frac{\mathrm{RT}}{\mathrm{F}}}\mathrm{In}{\alpha }_{O{H}^{-}}$$

(6)

where $E_{Ag/A{g}_{2}O}^0$ is estimated at ca. 345 mV [31] and ${\alpha }_j$ stands for the activity of OH⁻. According to Zhang et al. [42], the negative OCP shift in Ag/AgCl ISEs in cement extract with low chloride levels should not be attributed to the transformation of AgCl into Ag₂O. This is because an amorphous Ag-Cl-O mixture, rather than an Ag₂O crystal, formed on the Ag/AgCl ISEs surface.

The appearance of Ag₂O could influence the thickness and heterogeneity of the AgCl film. Ag₂O-related impurities emerge on the sensor surface, corresponding to the thickness and non-uniformity of the AgCl film [41]. Both Ag₂O and AgOH forming on the Ag/AgCl ISE surface will destroy the AgCl thin layer, resulting in a potential decline. A more negative potential is associated with an Ag/AgCl ISE that has a thinner AgCl layer, which facilitates the adsorption of OH⁻ on the Ag substrate [39]. Moreover, under the condition of high alkalinity, when the chloride concentration is low, the electrode potential decreases with time [90]. However, when the chloride concentration reaches a certain level, alkalinity no longer affects the Ag/AgCl ISE. More specifically, when the chloride concentration is less than 0.01 M, alkalinity does not affect the Ag/AgCl ISE [55], as depicted in Figure 8. According to Jin et al. [31], the minimum concentration of unaffected chloride depends on the pH. When the pH is 13.5 and 12.5, this value of engagement is 0.003 M and 0.0003 M, respectively. The available research shows that Ag/AgCl continues to be affected by fluctuations in chloride concentration in highly alkaline concrete environments [24]. But the long-term durability of the chloride probe is in doubt because the sensing surface layer is oxidized by hydroxyls in the concrete pore solution.

The original concentration of chloride in cement-based materials is usually low, ranging from 5 × 10⁻³ to 0.7 M [59,91]. Although the Ag/AgCl ISE surface generates Ag₂O or AgOH in the absence of chlorides, the potential can be quickly restored once the sensors contact the chloride solution. This process is reversible, as AgCl will gradually reform when chloride ions are added or subsequently present. This might be attributed to the reason that chlorides can repair the AgCl film [24,55].

4.2.4. Ion Selectivity Coefficient

The interference of various ions on the Ag/AgCl ISE response can be expressed as the γ term in Equation (2). The chloride selectivity coefficient ($K_{C{l}^{-}j}^{pot}$) represents the interference of other ions. For chemical species in the solid phase, the activity coefficient equals one, and the chloride activity coefficient is nearly one at low-to-moderate chloride concentrations [92,93]. For the interference of j ions, the selectivity coefficient is represented by Equation (7):

$$K_{C{l}^{-}, j}^{pot}={\displaystyle \frac{eF\left(E_{Ag/AgCl}^0-E\right)}{RT·{\alpha }_j}}-{\alpha }_{C{l}^{-}}$$

(7)

The value of $K_{C{l}^{-}j}^{pot}$ can be obtained by experiments or theoretical models, and corresponding references provide several values listed in

Table 1. The chloride selectivity coefficient for the Ag/AgCl ISE reported in the literature (25 °C).

Interfering Ions	${\mathit{K}}_{\mathit{C}{\mathit{l}}^{-}\mathit{j}}^{\mathit{p}\mathit{o}\mathit{t}}$	Exposure Time to the Solution	References
I−	86.5	Not specified	[41]
	2.9 × 106	Theoretical model	[94]
	1~2.1 × 106	Theoretical model	[95]
	3~14	<1 day	[87]
	1.8 × 102~2.2 × 106	<1 day	[96,97]
	86.5~1.8 × 106	<1 day	[98]
Br−	3.63 × 102	Theoretical model	[94]
	1~3.5 × 102	Theoretical model	[95]
	1.2	Not specified	[41]
	1.1 × 102~3.5 × 102	<1 day	[97,98]
	2.1~3.3 × 102	<1 day	[98]
$\mathrm{S}$, ${\mathrm{SO}}_4^{2-}$	2.04 × 1015	Not specified	[94]
	4.73 × 10−8	Not specified	[94]
OH−	2.4 × 10−2	Not specified	[41]
	9.33 × 10−3	Theoretical model	[94,95]
	10−2	Theoretical model	[95]
	≈10−2	Theoretical model	[95]
	4 × 10−3	>6 months	[65]
	≈10−2	Not specified	[16]
	2 × 10−3~9.1 × 10−3	<1 day	[97,98]

.

As can be seen in Table 1, selectivity coefficients vary greatly in different academic sources. Other parameters, such as surface coverage, diffusion processes, and membrane morphology, are believed to influence the response of the AgCl electrode [76,77,87,97,99,100].

4.3. Temperature and Humidity

Ag/AgCl ISEs are sensitive to temperature, responding quickly to temperature variation [8]. Temperature will affect the RT/F term in Equation (2) [37]. It is reported by [55,101] that for every 10 °C decrease in temperature, the potentials drop by approximately 6 mV in E_0,Ag/AgCl and around 3 mV in the RT/F term. Considering the influence of temperature, the measured potential (E_m) under a certain concentration can be written as follows based on Equation (8):

$$E_{\mathrm{m}}=a+b\mathrm{T}+c\mathrm{T}\cdot \mathrm{InT}$$

(8)

According to thermodynamic principles, temperature significantly influences the potential of Ag/AgCl ISEs, leading to noticeable measurement variations [102]. Additionally, the potentials of the Ag/AgCl ISEs and the used reference electrodes also depend on temperature [55,101]. Proper values for E_0,Ag/AgCl and reference electrodes should be used. Even minor temperature differences between the Ag/AgCl ISE and the reference electrode can cause considerable errors [55]. The impact of temperature can be minimized by employing Ag/AgCl-based reference electrodes, such as Ag/AgCl/sat. KCl, since both the reference electrode and the Ag/AgCl ISE have the same standard potential [37]. Although temperature plays a critical role in the potential following the laws of thermodynamics, resulting in non-negligible measurement shifts, these offsets are the overall movement of the calibration curve. It seems these errors can be corrected using a coefficient for temperature correction.

Besides, the humidity is inversely proportional to the chloride ion concentration. Lower humidity will induce a higher electrode potential value of the chloride sensor [63]. Tian et al. [63] conducted experiments with dry–wet cycles on test blocks that included Ag/AgCl ISEs, but did not perform a quantitative analysis of how temperature and humidity specifically affected the Ag/AgCl ISE. Cui et al. [103] conducted quantitative research to investigate the influence of temperature and humidity coupling on the potential of Ag/AgCl ISEs, which was examined throughout the temperature cycling test. A basic model was established to correlate the Ag/AgCl ISE potential with temperature, humidity, and chloride ion concentration. It is reported that at the critical value of 40 °C, temperature, humidity, and response potential are correlated. In an environment below 40 °C, the influence of temperature on the response potential is not obvious. Zhang et al. [61] investigated the working performance of Ag/AgCl ISEs in mortar with different chloride content (1% and 3%) and moisture, and the results illustrate that the potential was changed by nearly double from 75% to 95% RH, as shown in Figure 9 It can be concluded that the change in humidity may induce a distinct test error. This is because saturation directly affects the concentration of the pore solution. However, most current studies have not considered the influence of saturation. If the electrode is used for ordinary concrete that is not in direct contact with water, it is necessary to consider saturation.

4.4. Effect of AgCl Film Properties on Stability of Ag/AgCl ISE

The potential of Ag/AgCl ISEs in solution is not only determined by the redox equilibrium at the metal surface, but influenced by the solubility of AgCl [104]. Since AgCl films as a porous crystalline layer have numerous micro-channels for ionic reactions, surface chemistry alterations are related to their thickness, morphology, and microstructure, and all of them highly depend on the fabrication method [69]. Most of the research regarding AgCl film characteristics has focused on the Ag/AgCl ISE prepared by anodic chlorination [8,42]. By this means, an AgCl monolayer is created on a silver substrate via an adsorption process [105], followed by the gradual formation of a thin film through three-dimensional nucleation and growth [106]. According to Jin et al. [24], the SEM observations indicated that layer-by-layer formation of AgCl grains may occur during electrochemical deposition. Zhang et al. [107] investigated the size of the deposited AgCl grains, and the results indicated that the deposition current density was a key factor influencing the composition and microstructure of the AgCl film on the Ag substrate. The AgCl grains deposited were primarily between 800 and 1300 nm in size, with anodized current densities varying from 0.5 to 8.0 mA/cm², as shown in Figure 10. A low current density during AgCl film deposition resulted in a more uniform, homogeneous, and compact film, contributing to a stable potential response of the Ag/AgCl ISE in an alkaline medium [69,85].

The AgCl film created through anodic chlorination is porous, allowing the Ag/AgCl ISE to have a large effective surface area and quick ion diffusion, as shown in Figure 3. Therefore, the Ag/AgCl ISE is very sensitive to changes in the concentration of chloride ions. When the electrode is immersed in different concentrations, the response time required for potential stabilization is just ca. 30 s [39]. Nonetheless, the swift migration of interfering ions such as OH⁻ and Br⁻ on the porous Ag/AgCl ISE might hasten the conversion of the AgCl film into other insoluble compounds like Ag₂O and AgBr, leading to mixed potential and errors during chloride detection [9,16,71].

Theoretically, the AgCl thickness in response to chloride concentration shows a linear relationship based on Equation (4). However, the experimental results according to the research of Pargar [69] suggested the measured thickness was significantly greater than the theoretical thickness, indicating that the surface layers were less dense. Due to the lack of microscopic testing methods, the distribution of electrode surface density and the effect of density on electrode performance remain unclear.

Figure 11 depicts the AgCl film thickness used in Ag/AgCl ISEs for chloride detection in the referenced studies. The thickness of the AgCl film varies from a few hundred nanometers to approximately 40 µm, with the most common thickness being between 2 and 12 µm [42,53]. The increase in AgCl film thickness until a certain level obviously influences the OCP stabilization time. Potential fluctuations are relevant for AgCl film thickness > 32.04 µm [53,107]. However, the influence of AgCl film thickness on electrode stability is not evident in a specific range, and the maximum value of this range increases with the current density used for electrode fabrication [69].

Although a more significant current density may lead to a thicker AgCl film having better stability and service life, the performance of the prepared electrode decreases when the current density exceeds a certain value. As shown in Figure 12, a relatively higher current density does not produce the expected increase in thickness but results in an additional inner layer of smaller AgCl grains forming near the Ag substrate. The formation of a layer of small AgCl grains with open pores between them may suggest that the top AgCl film (or the entire AgCl film) is not strongly adhered to the Ag substrate [53].

5. The Durability of Ag/AgCl ISEs Used to Monitor Chloride in Concrete

5.1. Applications of Ag/AgCl ISEs for Long-Term Monitoring in Cement-Based Materials

Ag/AgCl ISEs used in cement-based materials as chloride sensors often act as the anode by half-cell potentiometric measurements. Calomel electrodes having a temperature coefficient close to Ag/AgCl ISEs (ca. −0.6 to −0.65 mV/C) are often used as the reference electrode [31,108,109]. Solid-state electrodes such as MnO₂ and Pt electrodes are also reported as reference electrodes due to a better mechanical performance [28,45,63,110,111,112,113,114].

Many existing investigations regarding Ag/AgCl embedded in cement-based materials as chloride sensors show that Ag/AgCl can be used efficiently to monitor chlorides [24,115]. One of the common research methods is to immerse embedded Ag/AgCl ISE samples in a chloride salt solution to simulate the chloride penetration process [24]. For instance, Dong et al. [28] buried Ag/AgCl ISEs in cylinder concrete samples immersed in NaCl solutions (0, 0.1, 0.5, and 1 M). Similarly, Elsener et al. [9] embedded Ag/AgCl ISEs in the mortar. For time intervals between 200 and 550 days, the sensor data was analyzed statistically, and the findings confirm that chloride sensors enable the continuous in situ monitoring of chloride levels in concrete, as shown in Figure 13. To assess the penetration depth and chloride diffusion rate across different concrete configurations, Montemor et al. [36] installed Ag/AgCl ISEs at varying depths. In addition, Zhang et al. [113] buried Ag/AgCl ISEs in the mortar and compared the relationship between the test concentration of Ag/AgCl ISEs and the concentration of free chloride ions. The electrode exhibited excellent stability and a Nernst reaction in the mortar. The above studies show that Ag/AgCl ISEs can respond sensitively to the erosion of chloride, and the obtained potentials are converted to the concentration, but the specific error is not given. For cement-based materials, the double electric layer effect and adsorption phenomenon in the pores of cement stone may affect the activity of dissolved chloride ions [116]. The double electric layer in the pores of cement slurry refers to the arrangement of charged particles and oriented dipoles existing at the solid–liquid interface between the pore wall and the pore solution. This bilayer is composed of an ion fixation layer (the Stern layer) adsorbed on the pore surface and an ion diffusion layer extending into the pore solution. In cement slurry, the cement slurry has a complex pore structure and a highly alkaline pore solution. This double layer can significantly affect ion transport and reactivity. Further research is indeed needed to fully describe these effects and their influence on chloride ion sensing in concrete.

In some experiments, the electrodes are directly embedded in cement-based materials doped with chloride ions. In this case, chloride ions react first with the hydration products to form calcium trichloroaluminate and calcium monochloroaluminate, as shown below [84]:

$$\begin{gathered} \mathrm{Ca}{(\mathrm{OH})}_2+2\mathrm{C}{\mathrm{l}}^{-}\to \mathrm{CaC}{\mathrm{l}}_2+2\mathrm{O}{\mathrm{H}}^{-} \\ 3\mathrm{CaO}·{\mathrm{Al}}_2\mathrm{O}·6{\mathrm{H}}_2\mathrm{O}+3{\mathrm{CaCl}}_2+25{\mathrm{H}}_2\mathrm{O}\to 3\mathrm{CaO}·3{\mathrm{CaCl}}_2·31{\mathrm{H}}_2\mathrm{O} \end{gathered}$$

(9)

The above reaction reduces the content of free chloride ions in the pore solution, thus increasing the potential value monitored by the chloride sensor [118].

5.2. Reason for Durability Failure

The durability failure of Ag/AgCl ISEs is deemed to be the result of AgCl film corrosion. Generally, as immersion time progresses, the AgCl film on the Ag/AgCl ISE surface experiences corrosion due to simultaneous dissolution and exfoliation, creating a porous and loose appearance. The ratio of Cl to Ag on the electrode surface diminishes as immersion time increases. When immersed in cement extract solution, an amorphous mixture of Ag-Cl-O forms on the electrode surface with a Cl⁻/OH⁻ ratio ranging from 0.0127 to 0.0139. [46]. Exfoliation, caused by the weak bond between the Ag rod and AgCl film, is suggested as the main factor behind the poor performance of Ag/AgCl ISEs in an alkaline medium. Due to these weak adhesions, OH⁻ from the alkaline medium may rapidly oxidize AgCl at the Ag/AgCl interface [69], and it is thought that the dissolution of the AgCl film is essentially the inverse of AgCl growth. The micro-channels are quickly widened, allowing the electrolyte solution within them to primarily facilitate ionic conduction [46]. Therefore, the life of the AgCl electrode is also determined by the AgCl thickness, which can be expressed as resistance. As depicted in Figure 14, the correlation between AgCl film resistance and the service life of Ag/AgCl ISEs is established, allowing for the precise evaluation of its performance in concrete [53].

As mentioned above, the failure of Ag/AgCl ISEs is attributed to the dissolution and exfoliation of AgCl film, so interfering ions in this process, such as sulfate ions, OH⁻, and Cl⁻, will influence the life of the AgCl electrode. Despite the potential interference from sulfides, sulfion and sulfate ions in cement-based materials do not significantly affect the stability and durability of AgCl electrodes, likely because of their low initial concentration [37]. In the case of a sulfate attack in the external environment, no relevant results have been found.

When applied to concrete, the effect of OH⁻ in the pore solution of concrete on the AgCl film is considered the main reason for the deterioration of Ag/AgCl ISEs in the absence of chloride. Many scientists have attempted to clarify why Ag/AgCl sensors produce unexpected results in alkaline environments such as concrete. The solubility product constant (Ksp) of AgCl, which is 1.77 × 10⁻¹⁰, is much lower than that of AgOH, which is 1.94 × 10⁻⁸ [82]. In the early stages of service, the concentration of OH⁻ in the concrete pore solution near the Ag/AgCl ISE is much greater than chloride due to the raw materials having a very low chloride content, and the penetration of chlorides through concrete is limited at this stage [8]. As a consequence, in alkaline concrete pore solutions with minimal chloride concentration, AgCl is swiftly converted to Ag₂O on the Ag/AgCl ISE surface. The above transformation is believed to be possible for causing potential fluctuations and even mixed potential, and subsequently, a significant negative potential shift in Ag/AgCl ISEs. For example, the study by Angst et al. [38] revealed that the stable OCP of Ag/AgCl ISEs was generally lower with increased alkalinity in solutions lacking chlorides, and defects were observed on the electrode surface following immersion in an alkaline solution [30]. According to Jin et al. [119], the rate at which AgCl converts to Ag₂O on the electrode surface depends on the OH⁻ concentration in the immersion solution.

The instability of the AgCl electrode in completely chloride-free alkaline solutions over time is probably caused by transformation reactions with the environment. Once chloride is introduced, the AgCl electrode again behaves according to Nernst’s law [37]. Additionally, a significant presence of chlorides aids in keeping the AgCl film intact, thereby ensuring the stable potential of the Ag/AgCl ISE [23]. The potentials for Ag/AgCl in concrete with lower chloride concentrations show a progressive decrease [59]. A significant negative potential shift (50–150 mV) was observed for Ag/AgCl ISEs placed in mortars containing minimal chlorides [113,116] or submerged in a simulated concrete pore solution with low chloride concentrations (0.05 and 0.1 mol L⁻¹) during the initial service stage [24].

Some discrepancies can be noted between relevant studies [120,121] regarding the long-term stability of the Ag/AgCl ISEs. These might be ascribed to methods and techniques of sensor fabrication [116]. Enhancing the contact area between Ag and AgCl or adopting a more elaborate Ag/AgCl probe design is accepted as a method to improve the durability of the probe in highly alkaline environments by increasing the time required for the hydroxyl ions to oxidize the silver at the Ag/AgCl interface [122]. The electrode manufactured by an electrolytic process with the proper fabrication regime shows high precision, while it has worse reproducibility and poor long stability [59,90]. It is possible to increase the reproducibility and long stability of Ag/AgCl ISEs by improving the bonding of the Ag substrate and the Ag/AgCl film. Previous works showed [55,116] that an Ag substrate coated with AgCl deposited by anodizing and additionally dipped in a melt of AgCl contributes to a more stable membrane. The Ag/AgCl ISE created using this technique shows a precise potential response for specific chloride concentrations, with an accuracy of about ±0.05 mol·L⁻¹. In addition, the working performance of the electrode can also be significantly improved by forming a protective film with a better performance on the electrode surface; for example, Xu et al. [123] proposed a highly sensitive and stable Ag/AgCl ISE by employing a protective coating layer based on polypyrrole. As can be seen from the above, improving the preparation process, increasing the bonding strength of the surface layer of the silver chloride electrode, and providing a protective film for the silver chloride layer are important measures to enhance the durability and stability of the silver chloride electrode.

6. Conclusions and Perspectives

The application of Ag/AgCl as ISEs to monitor chloride erosion in cement-based materials is reviewed. Common fabrication methods, interference factors and error sources, and durability are summarized and discussed. In addition, methods aiming to improve the stability and durability of Ag/AgCl ISEs are proposed. The conclusions can be summarized as follows:

-

The influence of environmental factors, including temperature, interfering ions, and alkalinity, on the performance of Ag/AgCl ISEs is systematically analyzed in this work. Temperature may lead to potential shifts, and this disadvantage can be overcome with temperature correction. Influences of interfering ions on the stability of the Ag/AgCl ISEs used to monitor chloride in concrete are derived from the membrane potential and insoluble silver salt, which should be considered only as interfering ions over a certain high concentration. Errors often appear in such high-alkalinity materials, such as concrete, with low concentrations of chloride due to the formation of AgOH on the surface of Ag/AgCl ISEs, leading to a lower potential than the theoretical values. However, when chloride concentration reaches a certain level, the electrode is no longer affected by alkalinity.

-

The microstructure of AgCl films, depending on the fabrication methods, plays a crucial role in the test errors; thus, selecting and improving the fabrication process to improve the electrode performance is the basis of reducing the test error. Besides, the calculated content based on Ag/AgCl ISEs is slightly lower than the actual chloride concentration, which is attributed to the diffusion potential, liquid junction potential, and membrane potential. It is noted that the change in concrete humidity may induce a distinct test error. Due to the differences in testing methods, the test results of electrodes vary greatly. Therefore, it is necessary to promote the standardization process of electrode preparation and use.

-

The Ag/AgCl ISEs can be used in cement-based materials by half-cell potentiometric measurement. The durability studies of Ag/AgCl ISEs in cement-based materials vary significantly, ranging from 20 to 550 days. The long-term durability of Ag/AgCl ISEs is related to the AgCl film corrosion caused by dissolution and exfoliation. The AgCl thickness, interfering ion, especially OH⁻, and the bonding between the AgCl film and Ag substrate will influence the durability life of Ag/AgCl ISEs.

-

Although long-term durability is necessary for Ag/AgCl ISEs to be used as in situ monitoring sensors, the relevant research enhancing long-term durability is still limited. To enhance the stability and durability of the electrode, its chemical stability in complex environments can be improved by modifying the microstructure and coating materials of the electrode. Given that the electrode test results are influenced by many factors, but the influence of multiple factors on the electrode results is still unclear, research can be carried out through methods such as machine learning.