The corrosion-inhibition mechanism of phosphates is not fully understood, it is believed that phosphate inhibitors react with the iron ions generated in the corrosion process
[41][16], or with ions present in the mortar, such as calcium, which forms calcium phosphate (Ca
3(PO
4)
2) precipitates, filling the pores and cracks of the mortar, thus impeding the diffusion of aggressive ions
[42,43,44,45,46][17][18][19][20][21]. It has been found that sodium phosphate (Na
3PO
4) can prevent pitting corrosion of steel in the simulated concrete pore solution if its concentration is equal to the chloride concentration
[47][22]. The presence of phosphates in the mortar increases the critical period of pitting initiation from 30 to 100 days and significantly reduces the chloride diffusion rate. Moreover, the apparent chloride diffusion coefficient calculated for mortar containing Na
3PO
4·12H
2O which is around 1.03 × 10
−12 m
2/s is lower than that obtained with the reference mortar (2.2 × 10
−12 m
2/s) for the same testing period
[48][23].
The concentration of phosphate species inside the pits was higher than the passive film zones without the pit. This indicates that phosphate ions could inhibit the corrosion process through a competitive adsorption mechanism with chloride ions where the chloride attack triggers the phosphate species to further adsorb at the pit locations on the metal surface
[49][24]. In addition, the presence of phosphate ions stabilizes ferrihydrite, a poorly crystallized FeOOH, which may be a protective layer for steel in Cl
−-contaminated concrete simulating solutions
[8]. As a counterpart, the inhibition efficiency of phosphate corrosion inhibitors is decreased in concrete because of the reaction of PO
43− ions with the concrete matrix
[50][25].
The surface analysis methods demonstrated that the inhibition mechanism of phosphate ions is attributed to the formation of a passive film with a duplex layer on the metal surface, including the inner layer of iron(hydro)oxides, formed by a solid-state mechanism and the outer layer of iron phosphate complexes mainly as FeHPO
4, Fe
3(PO
4)
2 and even Fe(PO
4), formed via a dissolution–precipitation mechanism
[49][24].
The impact of phosphate corrosion inhibitors on the stability of the passive film depends on the [Cl
−]/[OH
−] ratio. It can be explained by the fact that the high concentration of hydroxyl groups relative to chloride ions causes a predominant effect to form a robust passive film, which in turn stops the pitting corrosion without the aid of phosphate. This corrosion protection mechanism is associated with a continuous increase in the resistance of the passive film by adsorption of phosphate species in the weak points of the passive film, which block the anodic sites
[51][26].
Studies performed under applied mechanical stress revealed that in the presence of phosphate corrosion inhibitors, the critical [Cl
−]/[OH
−] ratio increased from 0.4 to 5 for strained electrodes under stress conditions (80% UTS)
[52][27].
The addition of phosphate corrosion inhibitors led to a decrease in chloride binding. This is mainly because phosphates hold a higher priority over chloride during ion exchange in Al
2O
3–Fe
2O
3-mono (AFm) and –tri (AFt) phases in the Ca–Al–S–O–H system of the concrete matrix. Moreover, phosphates exerted a significant influence on the chemical binding but a negligible effect on the physical binding
[53][28].
More recently, the use of alternative pentasodium triphosphate compounds Na
5P
3O
10, also known as sodium tripolyphospate, has shown an increased inhibition efficiency of around 80% for 480 days exposure in 3.5% NaCl, which is attributed to the development of a protective film barrier of PST on the steel rebar surface. Potentiodynamic polarization results revealed that PST affects the anodic and cathodic sites uniformly, thus presenting a mixed-type corrosion-inhibitor protection mechanism
[54][29]. The use of inorganic corrosion-inhibitor mixtures containing hexametaphosphate compounds (NaPO
3)
6, used in 3.5% NaCl contained SCPS, have been proven to reduce the corrosion rate by 8.60 and 25.52 times for 3 and 5% inhibitor addition, respectively
[55][30].
Disodium hydrogen phosphate (Na
2HPO
4, DHP) in simulated concrete pore solution (SCPS) and in mortar acts as an anodic corrosion inhibitor
[43,46][18][21]. Overall, phosphates require oxygen to be effective, as they are nonoxidizing anodic inhibitors
[38][31]. Trisodium phosphate (Na
3PO
4∙H
2O, TSP) in mortar acts as a mixed corrosion inhibitor
[56][32], as well as in chloride environments containing a [PO
43−]/[Cl
−] ratio higher than 0.6
[41,57][16][33] and even as a cathodic corrosion inhibitor, with a [PO
43−]/[Cl
−] ratio less than 0.6
[58][34]. Impedance techniques, such as electrochemical impedance spectroscopy (EIS), can be used to measure the corrosion improvement of inhibitors
[30,45,56,59,60][35][20][32][36][37]. Yohai et al. showed in their work with mortars great corrosion enhancement of the TSP (Mix C), even with the chloride addition, over the blank (Mix A) and the chloride contaminated specimen (Mix B) (see
Figure 42a), clearly seen in the Bode plot where Mix C is of one order of magnitude higher than the other two mixes (see
Figure 42b)
[56][32]. Similarly, Chaussadent et al. found an improvement in the corrosion protection by sodium monofluorophosphate as a corrosion inhibitor using the EIS results of mortar samples after carbonation (
Figure 53)
[61][38]. Another example of the use of EIS is the work from Etteyeb et al., in which the specimen containing inhibitors (see
Figure 64a) increased its impedance by two orders of magnitude compared to the blank (see
Figure 64b)
[45][20].
Figure 2. EIS spectra for mix designs:
Mix A:
Mix B:
Mix C:
registered after 720 days of exposure. Points represent the experimental EIS data and lines show the fitting results: (a) Nyquist plot, (b) and (c) Bode plots [32]. Reproduced with permission from Yohai, L. et al., Electrochim. Acta; published by Elsevier, 2016.
The corrosion inhibitor effect on the anodic and cathodic polarization curves can be seen in the Evans diagrams shown in
Figure 75. This is in agreement with the findings of Yohai et al., in which [PO
43−]/[Cl
−] = 1 and phosphate behaves as a mixed inhibitor
[60][37] and phosphate ions promote ferrous phosphate precipitation due to the higher solubility of ferric phosphate (p
Ksp = 26) than ferrous phosphate (p
Ksp = 32). The cyclic voltammograms in PSS, PSS + Cl
− and PSS + Cl
− + PO
43− showed the presence of a single negative and positive peak, attributed to the accumulation of magnetite on the steel surface, which is not fully reduced (see
Figure 86a)
[60][37]. However, for the PO
43− ions, which promote Fe
3(PO
4)
2 precipitation, the difference was not substantial and is due to lack of formation of Fe
3+ compound, hence not getting reduced in the following cycles (see
Figure 86b). By the impedance fitting, the protectiveness of the PO
43− ions was also seen, showing more ideal capacitors, which were related to the presence of a protective passive layer, however a small decrease in the impedance was associated with the change in the film composition, influencing the electronic properties (see
Figure 97).
Figure 5. Evans diagrams showing the effect of a corrosion inhibitor, (a) on the anodic branch (anodic inhibitor), (b) on the cathodic branch (cathodic inhibitor), and (c) on both the anodic and cathodic branches (mixed inhibitor).
Figure 6. Cyclic voltammograms for steel: (a) (tenth cycle) in PSS (—), PSS + Cl− (–□–), PSS + Cl− + PO43− (–○–), and (b) cycles 1–10th in PSS + Cl−. Scan rate: 10 mV s−1 [37]. Reproduced with permission from Yohai, L. et al., Electrochim. Acta; Published by Elsevier, 2013.
Figure 7. Impedance spectra recorded on steel electrodes aged over 24 h at Ecorr in SSP + Cl− with and without inhibitor. The symbols represent the data and the lines the fitting results. (a) Nyquist representation, (b) and (c) Bode representation, and (d) imaginary part of impedance as function of frequency, in logarithmic scale. PSS + Cl− (–□–), PSS + Cl− + PO43− (–○–) [37]. Reproduced with permission from Yohai, L. et al., Electrochim. Acta; Published by Elsevier, 2013.
Calcium monofluorophosphate (CaPO
3F)
[68][39], as well as zinc monofluorophosphate (ZnPO
3F) have been used as corrosion inhibitors for steel in 3 wt.% NaCl solution
[69][40], in both cases, the inhibitors are shown to be effective. Manganese monofluorophosphate (MnPO
3F) has been found to exhibit a mixed corrosion inhibition for steel in 3 wt.% NaCl solution
[70][41]. Aluminum tri-polyphosphate (AlH
2P
3O
10∙2H
2O) has been used as an ACI inhibitor with good results
[71][42].
Figure 108 shows Pourbaix diagrams with the stability fields for phosphoric ions as well as
EH−pH values of pore solution of two OPC pastes labelled CEM I and CEM III/B
[72][43]. The use of sodium monofluorophosphate (Na
2PO
3F, MFP) dual inhibitor and self-healing agent was found to recover 98.85% and 79.82% of the pH of the carbonated cement pastes compared to the untreated paste for CEM I and CEM III/B, respectively. The pH increased for higher concentrations of sodium in the treating agent.
Figure 8. Pourbaix diagram of two different OPC cement pastes (CEM I and CEM III/B) showing the stability fields for phosphoric ions in solution and the EH−pH values of the pore solutions. pH values of the pore solutions from untreated CEM I, carbonated paste (cc) and the self-healing agent sodium monofluorophosphate (MFA) for: (a) CEM I and (b) CEM III/B, respectively. pH values of the pore solutions from paste samples that were either only impregnated with MFA or were first carbonated and impregnated with MFA (cc + MFA) or were impregnated with MFA and carbonated (MFA + cc) for: (c) CEM I and (d) CEM III/B, respectively. The thick black lines frame the stability of HCO3− and CO32− [43]. Reproduced with permission from Kempl, J. et al., Cem. Conc. Comp.; Published by Elsevier, 2016.