Inhibition Effect of Natural Pozzolan and Zinc Phosphate Baths on Reinforcing Steel Corrosion

Zinc phosphate (ZnP) baths are widely used for increasing corrosion resistance and surface preparation for painting. Studies on exploiting these baths in the reinforced concrete (RC) are still in the early stages. This is probably due to the shortcomings, such as the alkaline instability and high porosity of the obtained coatings. Use of natural pozzolan (NP) as cement replacement is growing rapidly due to its economic, ecological, and technical benefits. The combined effect of using ZnP baths and NP-based cement on the resistance of concrete against damage caused by corrosion has been investigated. Four phosphating baths were prepared: ZnP, ZnP-Ni, ZnP-Cu, and ZnP-Mn. Steel specimens were phosphated at 55-60C for 15 min. Concrete specimens were produced with four different levels of NP: 0% (control), 10 %, 20%, and 30%. The investigation was carried out using RC specimens where a constant anodic potential was impressed after 28 and 90 days of concrete curing. The electrochemical behavior of the coated steel has further been evaluated in chloride contaminated Ca(OH)2 saturated solution (CH-Cl) using the open circuit potential (OCP), the potentiodynamic polarization, and the polarization resistance with time. The bond strength between the coated steel and concrete has been evaluated by the pull-out test. Test results showed that concrete containing NP at higher replacement levels and steel specimens treated in bication baths exhibited corrosion initiation times several times longer than the control concrete with uncoated steel. In addition, the best corrosion performance was noted in the steel specimen treated in the ZnP-Cu bath. Its corrosion density was about twentyfold lower with respect to the bare steel, and its inhibition efficiency exceeded 95% in (CH-Cl) solution. In addition, its polarization resistance was about fifteenfold lower with respect to the bare steel. SEM, EDX, and XRD techniques have been employed, as well.


Introduction
Reinforcing steel embedded in fresh concrete develops a protective passive layer on its surface.This layer, which is formed as a result of the high alkalinity of the concrete pore solution (pH ∼ 13), consists of -Fe 2 O 3 adhering tightly to the steel.As long as that oxide film is present, the steel remains intact [1].However, chloride ions attack and concrete carbonation can destroy the film and, in the presence of H 2 O and O 2 , reinforcement corrosion takes place.Chloride ions, which were described by Verbeck [2] as the specific and unique destroyer, can be present in concrete through the use of either contaminated aggregate, sea water, brackish water, or admixtures containing chlorides [1].
In the literature, there are state-of-the-art review studies on numerous methods of protection against reinforcement corrosion [3][4][5][6][7][8].A variety of protective coatings such as zinc and epoxy coatings have widely been used to improve the corrosion resistance of reinforcing bars.Treatments of reinforcing steel in zinc phosphate baths could probably be a vital approach in the future [9][10][11][12].
The insulation properties make an important contribution to the prevention of reinforcement corrosion [19].Further, the zinc phosphate baths are currently considered eco-friendly [20].However, the obtained phosphate coatings usually contain pores which are pathways for the corrosive International Journal of Corrosion electrolyte diffusion into the metal substrate.As a result, the corrosion resistance of the conventional phosphate coating is not high enough.Additives such as Ni 2+ , Cu 2+ , Mn 2+ , Ca 2+ , and Mo 2+ and nanoparticles have been extensively used to produce phosphate coatings with uniform structure, lower porosity, higher corrosion resistance, and improved adhesion properties [11,15,[21][22][23][24][25].A brief state-of-the-art review of these additives and their effects on the characteristics of zinc phosphate baths has recently been published by the author [26].
Zinc phosphating is essentially an electrochemical process [27].When the steel comes into contact with phosphate solution which is basically a phosphoric acid-based solution, an electrochemical reaction takes place.Fe 2+ ions start to dissolve and release electrons at the anodic site, and the reduction of H + occurs at the cathodic site.The evolution of hydrogen and consumption of H + ions result in a local increase of pH at the surface leading to precipitation of zinc phosphate crystals [28].According to Zimmermann et al. [23], Donofrio [29], and Kunst et al. [30], the possible reactions taking place in the zinc phosphating bath are as follows: The alkaline stability of the obtained zinc phosphate coating on steel is of great importance, as it will be embedded in a highly alkaline environment (pH∼13).Literature shows that the alkaline stability of zinc phosphate coatings depends on their chemical composition and crystal structure [31].Hopeite (Zn 3 (PO 4 ) 2 .4H 2 O) has a higher alkaline solubility than phosphophyllite (Zn 2 Fe(PO 4 ) 2 .H 2 O) [29,32,33].Thus, alkaline solutions first dissolve the coating composed of hopeite.Then the dissolution continues for layers rich in iron and those mainly composed of phosphophyllite [34].In general, the presence of phosphophyllite is associated with a better corrosion resistance than hopeite, probably due to its enhanced chemical stability relative to alkaline electrolytes [35].
Natural pozzolan (NP) is being widely used as cement replacement due to its ecological, economic., and performance-related advantageous properties [36][37][38][39][40][41].However, its use caused a lower alkalinity compared with plain Portland cement [42].This could be due to the reduction in cement content (i.e., the dilution effect) when adding natural pozzolan and the consumption of portlandite (CH) through the pozzolanic reaction.
Syria is rich in natural pozzolan with estimated reserves of about one billion tonnes [43].There are a significant number of studies that deal separately with the effect of using natural pozzolan or that of zinc phosphating process on the corrosion of reinforcing steel.However, very little or even no previous works, as the author thinks, have been carried out in the past to investigate the combined effect of adding NP as cement replacement and zinc phosphate coated steel on the anticorrosion properties of reinforcing steel exposed to highly chloride ion concentrations.In addition, literature did not cover the benefit, in terms of alkaline stability, which can be achieved when the zinc phosphate coated steel is embedded in concrete of relatively lower alkalinity.Further, the reduced concrete permeability offered by NPbased cements could be considered an additional physical barrier between steel and its environment.
The objective of this paper is to investigate the influence of adding NP as cement replacement on the corrosion performance of reinforcing steel treated in different zinc phosphate baths.Four binders with different replacement levels of NP (0%, 10%, 20%, and 30%) have been produced for this investigation.In addition, four zinc phosphating baths were employed: the first one is a monocation bath (i.e., it contains only Zn-cation) and three bication baths, Zn-Ni, Zn-Cu, and Zn-Mn, respectively.Two test environments were created for this investigation: concrete specimens and a contaminated Ca(OH) 2 saturated solution in which the coated steel was embedded or immersed.
The study is of particular importance for the following points: (i) Replacement of OPC by natural pozzolan could significantly minimize CO 2 released into atmosphere and save energy.In addition, zinc phosphating process is considered eco-friendly.
(ii) This study is the first of its kind in Syria.However, it is not limited to the country.It can be applied to other countries of similar geology, e.g., Harrat Al-Shamah, a volcanic field which covers a total area of some 45,000 km 2 , and about 15,000 km 2 is located in the country.
The rest covers parts of Jordan and KSA.
(iii) As our country begins preparations for the huge reconstruction after the war comes to its end, the encouraging results can be considered a motivation of other studies, for instance, use of these local supplementary cementing materials in enhancement of the anticorrosion properties of reinforced concrete made from recycled concrete aggregates (RCA).RCA are expected to be an inevitable building material during the postwar reconstruction in Syria.

Experimental Procedure
2.1.NP-Based Concrete.Natural pozzolan (NP) used in the experiments was collected from a Tal Shihan quarry, 70 km southeast of Damascus.More detailed information related to its characteristics can be found in the latest research work of the author and colleagues [44].Four binder specimens have been prepared: one plain Portland cement CEM I (control) and three binary binders with three replacement levels of 10%, 20%, and 30% NP .5% of gypsum was added to all the binder specimens.This ratio is frequently used by all Syrian cement plants, providing about 2.35% SO 3 content.The amount of gypsum added to the cement clinker is expressed usually as the mass of SO 3 present; this is limited by European Standard BS EN 197-1: 2000 to a maximum of 3.5 percent [45].All replacements were made by mass of cement.All binders were designated according to the replacement level.For instance, NP10 and NP30 refer to the binders containing 10% and 30% of NP, respectively.More data on the characteristics of the binder components, namely, particle size distribution and the grinding method, can also be found in the scientific work of the author and colleagues [44].Four concrete mixes have been prepared using the same constituents and procedure followed by the author in his latest work [44].Concrete cubes (150 mm) were cast for the determination of compressive strength and water permeability.Concrete cylinders of 75 mm × 150 mm and 100 mm × 200 mm were also cast for testing the concrete porosity and the concrete chloride ion penetrability, respectively.The RC specimen for the accelerated corrosion test was 100 mm × 200 mm concrete cylinder in which 14 mm diameter steel bar was centrally embedded.The steel bar was embedded into the concrete cylinder such that its end was at least 45 mm from the bottom of the cylinder, and it was coated with epoxy at the exit from the concrete cylinder in order to eliminate crevice corrosion.

Zinc Phosphate Baths.
Four zinc phosphate baths were created, conventional bath or monocation bath without modifying elements and three bication baths modified either by nickel (Ni 2+ ), copper (Cu 2+ ), or manganese (Mn 2+ ) cations.Their chemical compositions and designations are illustrated in Table 1.All the chemical reagents applied in the present work were of analytic grade.Deionized water was used in all zinc phosphating baths.Phosphating process was carried out by the conventional immersion of reinforcing steel specimens at 55-60 ∘ C, for 15 min in the phosphating baths at a pH value of 2±0.1.The phosphating process primarily included precleaning, activation, phosphating, rinsing, and drying.The full steps of the phosphating process are schematically illustrated in Figure 1.The chemical composition of steel rebar with some mechanical properties is presented in Table 2.A macrograph of treated steel specimens with an optical microscopic photo is shown in Figure 2. A ZnP-Ni coating of about 8 m thickness can clearly be seen in Figure 2(b).

Morphology of Zinc Phosphate
Coatings.The morphology and elemental composition of the obtained zinc phosphate coatings were studied by scanning electron microscope (SEM, VEGA II TESCAN) and energy disperse X-ray (EDX) spectrometer.reported represent the average of six readings for compressive strength test and the average of three readings for all other tests.

Accelerated Corrosion Test.
A rapid corrosion test was used to compare the corrosion performance of NP-based cement concrete in which either straight phosphate rebars or bent phosphate rebars are embedded, as shown in Figure 3.
The objective of evaluating the bent phosphated rebars is to verify the coating soundness and to investigate whether or not the bending action might affect its corrosion performance.Similar techniques with little differences were reported by other researchers [36,[46][47][48][49][50].The cylindrical specimen shape was adopted to provide uniform cover and easier fabrication.
In the study, RC specimens were immersed in a 15 % NaCl solution leveling half of the concrete cylinder and the steel bar (working electrode) was connected to the positive terminal of a DC power source while the negative terminal was connected to a steel plate (counter electrode) placed near the concrete specimen in the solution.The corrosion process was initiated by impressing a relatively high anodic potential of 12 V to accelerate the corrosion process.The instrument used in the test is a Laboratory DC power supply, Model GPC-60-300 equipped with a Digital Multimeter Sanwa, CD 721.
Figure 3 shows a schematic representation of the experimental setup for the accelerated corrosion test.The specimens were monitored periodically to see how long it takes for corrosion cracks to appear on the specimen surface.The current readings with time were recorded at 3 h intervals.Three specimens from each concrete mix and phosphating bath were tested after 28 and 90 days of curing.

Bond Strength Test.
Pull-out tests and beam tests are the common experimental methods used for assessment of bond performance [51].In the study, the concrete cube specimens were tested to determine the bond strength between steel rebars and concrete using the concentric pullout test.This concentric pull-out test was similar to that outlined in ASTM C 234.Although this type of test does not represent the actual situation in a structure, the authors used this method since the study is basically a comparative study.The pull-out test was carried out using 150 mm cubic specimens with the bar centrally embedded.Deformed steel bars with 16 mm in diameter were used instead of no.6 (19 mm) bars specified in ASTM C 234 which is used as the basis of casting and testing procedures.Each mold was designed to cast one 150 mm cubic specimen as shown in Figure 4(a).The mold was made of steel plates.The holed cap was designed to support the bar in a vertical position.
The embedment length of reinforcing bars was 72 mm (∼4.5 times the rebar diameter).The required embedment length has been obtained by breaking the bond between steel and concrete using polyvinyl chloride (PVC) sleeves to cover the unembedded length.The gap between the reinforcing steel and sleeves was sealed with a silicon sealant.
Sketch map of the bond test setup and a photograph of the bond test frame with the specimen in position are shown in Figures 4(b) and 4(c).The specimen was mounted in the frame with the bar passing through the slot in the frame.A digital dial indicator has been installed on the free end to measure the rebar's displacements.
A 2000 kN capacity universal testing machine has been used for the pull-out testing.The bond stress () has been calculated by dividing the applied load () by the surface area (Ø) of reinforcing steel in contact with concrete, as shown in the following formula: where  is the length of embedment (72 mm) and Ø is the diameter of the reinforcing bar (16 mm).The free-end slip values were plotted against the bond stresses, and the bond strengths at failure have been determined for each concrete mixture and curing age.solution to simulate reinforced concrete exposed to chloride attack.The obtained chloride contaminated solution is denoted (CH-Cl).Its pH value was about 12.3.In numerous studies of reinforcement corrosion, saturated Ca(OH) 2 has been used as a substitute for concrete pore solution [52].
The electrochemical tests in the (CH-Cl) solution were performed with a three-electrode system.The working electrode was either the coated steel or the bare steel, the counter electrode was a platinum sheet, and the reference electrode was a saturated Calomel electrode (SCE).The surface area of the working electrode exposed to the solution was 6 cm 2 .Different electrochemical methods were employed to evaluate the corrosion behavior of the coated steel specimens.First, the coated and bare steel specimens were immersed in the test solution for one hour to reach stationary open circuit potential (OCP).Second, polarization curves have been measured with a scan rate of 0.5 mV/s and a scan range from -0.25 V for OCP to + 0.25 V for pitting potential.Third, periodic measurements of linear polarization resistance (LPR) over time were performed during 96 h.The specimens were polarized at ±20 mV with respect to the open circuit potential (E ocp ) at a scan rate of 0.5 mV/s and the LRP measurements were taken every 30 minutes.For a small perturbation about the open circuit potential, there is a linear relationship between the change in applied current per unit area of electrode (Δi) and the change of the measured voltage (ΔE).The ratio (ΔE/Δi) is called polarization resistance (R p ).The testing temperature was 25 ∘ C.
After the test, the specimens were rinsed at room temperature and the surface was observed and analysed with SEM and EDX.SATOE STADI X-ray diffractometer (XRD) using CuK radiation, operated at 40 KV and 30 mA with a scan mode ranging from 5 to 85 ∘ and a scan speed of 2 ∘ /min, was also employed in order to investigate the phases formed on the coated steel surface after immersion in (CH-Cl) solution.
All the electrochemical tests were performed using a potentiostat (TACUSSEL, PGZ 301) at ambient temperature.The tests were repeated three times with the same conditions for confirming the reproducibility of the obtained zinc phosphate coatings.All electrochemical experiments were remarkably reproducible.

Morphology and Chemical Composition of Zinc Phosphate Coating.
Scanning electron micrographs of the zinc phosphate coatings obtained during the phosphating process are shown in Figures 5-8.The SEM observations give more detail about the modifications of coating morphology with the cation type in the phosphating solutions.
For the monocation bath, ZnP bath, the coating seems to be porous.It consists of platelet shaped crystallites of 30-40 m length and 5-10 m width, which uniformly covered the surface of the specimen, Figure 5.However, the presence of Ni 2+ , Cu 2+ , and Mn 2+ cations in the phosphating solutions resulted in a significant modification of the crystalline form and size.
The ZnP-Ni coating is smoother than the monocation ZnP coating.Its structure is characterized by platelets with sizes up to 20 m in length and up to 3 m in width, as shown in Figure 6.The grain refinement was also reported in literature [9,35,53,54].The relatively high weight of Ni in the coating as confirmed by the EDX analysis may indicate the substitution of Zn 2+ by Ni 2+ in hopeite or the deposition of Ni in the coating.This result was in agreement with the results obtained by Zimmermann et al. [23] and Su and Lin The ZnP-Cu coating consists of smaller grains, the dimensions of which varied between 2 and 6 m.Copper, as clearly seen in Figure 7, modified and refined the crystal structure and enhanced the coverage of the surface significantly.This, which was in agreement with the results obtained by Abdalla et al. [57], can be attributed to the effect of CU-Fe galvanic couple on dissolution of iron and acceleration of crystal deposition [57].From the EDX analysis, Figure 7, it can clearly be distinguished that the dominant element was copper.Therefore, the metal surface exposed between the formed crystals might be a surface of copper rather than the base metal.This result is in well agreement with the result of Ogle and Buchheit [35] who concluded that Cu 2+ may deposit as copper metal on the surface.
From Figure 8, it is noted that the ZnP-Mn coating is well crystallized and it completely covers the steel surface.Its structure is characterized by small platelets whose width is about 3 m and length is 15 m.Manganese (Mn 2+ ) behavior, to some extent, in the phosphating bath is similar to that of Ni 2+ , as clearly seen in Figure 8.In addition, Mn 2+ may replace Zn 2+ in the hopeite crystal lattice.The presence of Mn 2+ in the coating may support this assumption, as clearly seen from the EDX analysis in Figure 8.This was in agreement with the results obtained by Su and Lin [55], who proposed the formation of Mn 2 Zn(PO 4 ) 2 .4H 2 O. Rezaee et al. [21] also observed this formation.In addition, adding Mn 2+ reduced the phosphate coating porosity and improved surface coverage which led to an enhanced corrosion resistance [21].3 and 4 show the test results of compressive strength, water penetration depth, chloride ion penetrability, and porosity of concrete.The test results will not be discussed here as the same trend of results using the same or a similar natural pozzolan was studied in depth in recent works carried out by the author [44,58].In addition, the microstructural investigation of the hydration products can also be found in these works.

Corrosion Resistance of Coated Steel Specimens Embedded in NP-Based Concrete. Typical curves of corrosion current
versus time for the reinforced concrete specimens made with NP-based binders and phosphate steel rebars are illustrated in Figures 9-12, respectively.As shown in Figures 9-12, currenttime curve initially descended till a time value after which a steady low rate of increase in current was observed, and after a specific time period a rapid increase in current was detected until failure.The decreasing tendency of current at the very early time could be explained by the filling of the pores with salt and other deposits in the salt water [36].Almost a similar variation of the corrosion current with time has also been observed by other researchers [46,50].The first visual evidence of corrosion was the appearance of brown stains on the surface of the specimens.Cracking was observed shortly thereafter, and it was associated with a sudden rise in the current.
Figure 13 presents the average corrosion times required to crack the specimens made with NP-based binders and phosphated steel rebars.Time to cracking in NP0-based concrete specimens was in the range of 87-134 h (3.6-5.6 days), whereas that in NP30-based concrete was in the range  of 197-367 h (8.2-15.3days).In addition, it was observed in Figure 13 that the corrosion resistance of NP-based cement concrete specimens increased significantly with age while that of the plain cement concrete had a slight increase which has also been indicated by other researchers [36,50].
The combined effect of using NP-based cement concrete and zinc phosphated steel rebars can clearly be seen in Figure 13.From the graphs, it is very clear to note that the time taken for initiation of 28-day cured concrete cracking was found to be 197, 215, 246, 263, and 232 h for bare steel rebar, ZnP coated steel rebar, ZnP-Mn coated steel rebar, ZnP-Ni coated steel rebar, and ZnP-Cu coated steel rebar which were embedded in NP30-based concrete, respectively.The best  This delay in corrosion time can be attributed to the following: (i) The pozzolanic reaction between glassy phase in NP and CH released during cement hydration contributes to filling the voids and pores in concrete with an additional C-S-H.This leads to decrease of pore size and to a smaller effective diffusivity for chloride [36].This, which was confirmed by the water penetration depth, chloride penetrability, and concrete porosity tests, can also improve the long-term corrosion resistance of RC structures and make concrete denser and less permeable [36,50].(ii) The role of Cu 2+ cation in refinement of the coating was confirmed by SEM and EDX analysis, Figure 7. (iii) The deposition of Cu 2+ -based layer on the phosphated steel rebars contributes to forming a physical barrier between the base metal (steel) and its environment.(iv) Bication baths enhance the alkaline stability of the zinc phosphate coatings.According to Simescu and Idrissi [11] study, the dissolution of ZnP-Ni coating in the alkaline medium (pH ∼12.5) is accompanied by the formation of hydroxyapatite Ca 10 (PO 4 ) 6 (OH) 2 .Hydroxyapatite formation occurs even in the presence of aggressive chloride ions [11].This chemical compound provides an effective protection against reinforcement corrosion and contributes to the reduction in chloride aggressiveness [10].(v) Adding NP as cement replacement reduced pH values of concrete as seen in Table 4.This reduction can be due to the consumption of CH through the pozzolanic reaction and the lower cement content (i.e., the dilution effect).

International Journal of Corrosion
From this accelerated test, it is also confirmed that the bication phosphating baths have higher corrosion resistance when compared to the monocation bath at all replacement levels of NP.
Further, it is worth mentioning that the bending action did not negatively affect the corrosion resistance of the phosphated steel rebars.The maximum reduction which was noted in ZnP treated rebar did not exceed 6%.

Bond Strengths.
The bond stresses at 0.25 mm slip and at failure were presented in Table 5.It can be clearly seen that phosphate coatings applied to the embedded steel did not significantly reduce the bond strength of steel with concrete.After 28 days of curing, some of the phosphating baths increased the bond strength (i.e., ZnP-Cu, ZnP-Ni) while the other slightly decreased the bond strength with concrete (i.e., ZnP-Mn and ZnP).However, after 90 days of curing, all phosphated steel rebars achieved higher bond strengths when compared with the bare steel embedded in the control concrete.

Corrosion Behavior of Coated and Bare Steel in (CH-Cl)
Solution.The open circuit potential values which were recorded after stabilization for the coated and bare steel specimens are illustrated in Figure 14.The results show that the potentials are becoming more noble for bication zinc phosphate coatings, and the ZnP-Cu coating is the most noble one.However, an inversion appears for the monocation zinc phosphate coating (i.e., ZnP bath).This, according to the author, can be explained by the following: (i) The presence of Cu or Ni in the bication zinc phosphate coatings may increase the potential values, as copper and nickel are more noble than steel.
(ii) The formation of passive phases may make the potential more noble over time.This was confirmed by SEM, EDX, and XRD analysis, as shown in Figures 17-20.
(iii) The additives contribute to refinement of zinc phosphate crystals [26].
The polarization curves are shown in Figure 15.Upon increasing the potential above E corr which corresponds to the minimum current density, a passive region was observed where the current density was of the order of A/cm 2 .Pitting  occurred in the bare steel at a potential value of less than zero.On the other hand, at potential values higher than 0.6 V/SCE an increase in the current density was observed in ZnP-Cu and ZnP-Ni specimens.This sharp increase in the current density corresponds to the pitting potential E pit .Due to the effective barrier offered by the bication zinc phosphate coatings, the average polarization current declines considerably as compared to the bare steel.Table 6 presents E corr , I corr , E pit , and passivity range of specimens after the polarization test.Among all the investigated coatings, the ZnP-Cu and ZnP-Ni coated specimens exhibited the lowest values of I corr , about twentyfold lower with respect to the bare steel.In addition, these coatings revealed a passive region of about 1000 mV.The inhibition efficiency (IE) has further been estimated using the following equation [59]: Here (I corr ) 0 and I corr denote corrosion current density of reinforcing steel in the absence and presence of zinc  phosphate coating, respectively.As illustrated in Table 6, inhibition efficiency values of more than 90% were observed in the zinc phosphate coatings modified by either Cu 2+ or Ni 2+ cations.This result suggests that ZnP-Cu and ZnP-Ni coatings provide an effective corrosion resistance.
The periodic measurements of linear polarization resistance are displayed in Figure 16.For the bication zinc phosphate coatings, the differences in the behavior are not very significant.The polarization resistance decreases slightly at the beginning of the immersion, meaning that the dissolution rate of the bication coating slows down; then the polarization resistance increases, probably due to the formation of a passive phase by the resulting products.This result was confirmed by SEM, EDX, and XRD analysis, as shown in Figures 17-20.The formation of more than one passive phase, even in the presence of chloride ions can clearly be seen in Figure 19.These passive phases led to a better protection of steel as evidenced by the polarization tests; I = f(E) and R p = f(t).After 96 h of immersion in the (CH-Cl) solution, ZnP-Cu and ZnP-Ni coatings recorded polarization resistance of about 40 kOhm.cm 2 .On the other hand, the polarization resistance of the bare steel and the monocation coating recorded values of less than 10 kOhm.cm 2 .
Based on the XRD analysis, the presence of calcium hydroxyzincate, Ca(Zn(OH) 3 ) 2 .2H 2 O, is clearly seen in all zinc phosphate coatings after 96 h of immersion in the (CH-Cl) solution.This compound is the main corrosion product of galvanised steel in chloride contaminated concrete [60,61].
New peaks related to the formation of other products, such as hydroxyapatite, Ca 10 (PO 4 ) 6 (OH) 2 , and copper oxide were detected in ZnP-Cu coating.The formation of such passive products in a highly alkaline solution contaminated with chloride ions was reported by other researchers [9][10][11]62].These products contribute to the decrease of the chloride aggressiveness and provide an effective protection against the reinforcement corrosion, although further investigation is needed to study their electrochemical behavior in a wider range of pH and at different chloride ion concentrations.It is worth mentioning that the copper particles deposited in the zinc phosphating remained intact as evidenced by the strong peaks of copper shown in either EDX analysis or XRD diffraction test.Compared with that of bication baths, the monocation bath has the worst corrosion performance where uncovered steel zones can easily be distinguished in Figure 17.These zones are attributed to the rapid dissolution of the phosphating coating and the complete breakdown of the local passive layer.

Conclusions
This paper was an attempt to investigate the effects of some additives such as alkali metal ions on zinc phosphate bath properties.In addition, the effect of using NP-based binders on corrosion performance was mainly studied.
(i) Bication baths led to improvements in the zinc phosphate coating performance in terms of reinforcement  corrosion resistance and bond strength between concrete and steel.The best corrosion resistance was observed in the ZnP-Cu-NP30-based binder system.It took corrosion initiation periods 4 times longer than the bare steel-NP0based cement system.The reduced chloride penetrability, water permeability, and porosity of NP-based concrete specimens made a further significant improvement in terms of reinforcement corrosion resistance.A similar excellent corrosion performance was also noted in the ZnP-Ni system, as it performed very well in the electrochemical tests carried out using either (CH-Cl) solution or concrete specimens.Although ZnP-Mn coating performed, to some extent, well, but its corrosion performance was not as high as the ZnP-Cu and ZnP-Ni coatings.
(ii) Bent phosphated steel rebars did not significantly affect the corrosion performance.(iii) Based on the results of bond strength, it should be noted that the hydrogen embrittlement which can be encountered in acidic baths may be considered negligible in the studied phosphating baths.
(iv) The stability of the phosphate layer in a highly alkaline solution was significantly increased by modifying the zinc phosphating bath with Cu 2+ , Ni 2+ , and Mn 2+ .(v) Electrochemical test results revealed that the bication coatings performed very well in highly alkaline solutions contaminated with chloride ions.The inhibition efficiency values ranged from about 85% to 95%.The SEM, EDX, and XRD analysis supported this result, where passive resulting products were detected.
(vi) The microscopical examination of the monocation bath specimens, after the polarization resistance test, revealed uncovered steel zones, indicating that corrosion performance of this bath may not be relied upon.
(vii) Further studies on the chemical stability of the modified zinc phosphate coatings are highly recommended.Long-term tests when these coatings are exposed to chloride ions are also recommended.Testing the efficiency of the formed phosphate coatings when immersed in concrete exposed to carbonation is recommended, as well.
(viii) Finally, there is a need to develop an optimum chemical composition of the phosphating bath and treatment conditions that provide maximum protection against reinforcement corrosion.

Figure 1 :
Figure 1: A schematic representation of the full steps of phosphating process.

Figure 2 :
Figure 2: (a) Photograph of the some phosphated steel rebars; (b) an optical microscopy photo of cross-sectional ZnP-Ni treated steel.

Figure 3 :
Figure 3: ((a) and (b)) Schematic representation of the concrete specimen with bent steel rebar.(c) Schematic representation of experimental setup for the accelerated corrosion test.

2. 5 . 2 InternationalFigure 4 :
Figure 4: (a) Photograph of the mold used in the bond test with the bar in place; (b) sketch map of the bond set up; (c) photograph of the bond test setup with the specimen in position.

Figure 9 :
Figure 9: Typical curves of corrosion current versus time of concrete specimens tested after 28 days of curing.

Figure 10 :
Figure 10: Time to cracking of RC specimens made with NP10based cement and phosphated steel rebars after 28 days of curing.

Figure 11 :
Figure 11: Time to cracking of RC specimens made with NP20based cement and phosphated steel rebars after 28 days of curing.

Figure 12 :Figure 13 :
Figure 12: Time to cracking of RC specimens made with NP30based cement and phosphated steel rebars after 28 days of curing.

Figure 14 :Figure 15 :
Figure 14: OCP values of the obtained coatings after stabilization.

2 )Figure 16 :
Figure 16: Polarization resistance over time of coated steel specimens and bare steel after 96 h of immersion in (CH-Cl) solution.

Figure 17 :
Figure 17: SEMs and EDX (a) and XRD analysis (b) of ZnP coating after immersion in (CH-Cl) solution for 96 h.

Figure 18 :
Figure 18: SEMs and EDX (a) and XRD analysis (b) of ZnP-Ni coating after immersion in (CH-Cl) solution for 96 h.

Figure 19 :
Figure 19: SEMs and EDX (a) and XRD analysis (b) of ZnP-Cu coating after immersion in (CH-Cl) solution for 96 h.

Figure 20 :
Figure 20: SEMs and EDX (a) and XRD analysis (b) of ZnP-Mn coating after immersion in (CH-Cl) solution for 96 h.

Table 1 :
Chemical composition of the prepared zinc phosphating baths.

Table 2 :
Chemical composition with some physical characteristics of the steel bars used in the experiments.

Table 3 :
Compressive strength development of concrete.

Table 4 :
Water penetration depths, porosity, chloride penetrability, and pH of the investigated concrete.

Table 5 :
Bond stresses (MPa) of the investigated phosphate steel rebars.

Table 6 :
I corr , E corr , E pit , the passivity range, and the inhibition efficiency for the coated steel specimens and the bare steel.