Anion Capture and Exchange by Functional Coatings: New Routes to Mitigate Steel Corrosion in Concrete Infrastructure

Abstract

Chloride-induced corrosion is a major cause of degradation of reinforced concrete infrastructure. While the binding of chloride ions (Cl^-) by cementitious phases is known to delay corrosion, this approach has not been systematically exploited as a mechanism to increase structural service life. Recently, Falzone et al. [Cement and Concrete Research72, 54-68 (2015)] proposed calcium aluminate cement (CAC) formulations containing NO₃-AFm to serve as anion exchange coatings that are capable of binding large quantities of Cl^- ions, while simultaneously releasing corrosion-inhibiting NO₃^- species. To examine the viability of this concept, Cl^- binding isotherms and ion-diffusion coefficients of a series of hydrated CAC formulations containing admixed Ca(NO₃)₂ (CN) are quantified. This data is input into a multi-species Nernst-Planck (NP) formulation, which is solved for a typical bridge-deck geometry using the finite element method (FEM). For exposure conditions corresponding to seawater, the results indicate that Cl^- scavenging CAC coatings (i.e., top-layers) can significantly delay the time to corrosion (e.g., 5 ≤ d_f ≤ 10, where d_f is the steel corrosion initiation delay factor [unitless]) as compared to traditional OPC-based systems for the same cover thickness; as identified by thresholds of Cl^-/OH^- or Cl^-/NO₃^- (molar) ratios in solution. The roles of hindered ionic diffusion, and the passivation of the reinforcing steel rendered by NO₃^- are also discussed.

Keywords: (C) corrosion; (C) finite element analysis; (D) calcium aluminate cement; (D) chloride; (−) calcium nitrate.

Figure 1

The hydrated phase assemblages calculated using GEMS for CAC systems for w/c = 0.45 containing: (a) 0 mass % CN, (b) 10 mass % CN, and (c) 30 mass % CN [28]. These calculations consider 100 g of anhydrous CAC reacting with 45 g of water/CN solution. The dashed vertical line shows the maximum simulated degree of hydration for each system. At a degree of hydration greater than this value, chemical reactions will cease due to limitations on water availability.

Figure 2

The free Cl⁻ concentration (C_Cl,f) in solution as a function of time for CaCl₂·2H₂O solutions in contact with hydrated CAC pastes containing: (a) 0 mass % CN, (b) 10 mass % CN, and (c) 30 mass % CN over the concentration range: 0.01 mol/L ≤ $C_{Cl,f}^0$ ≤ 3 mol/L. The coefficient of variation of the C_Cl,f measurements was ≈5%.

Figure 3

The measured Cl⁻ binding isotherms for the hydrated CAC pastes following 21 d of immersion in CaCl₂·2H₂O solutions. The error bars indicate ± one standard deviation in the experimental measurements, which were performed in triplicate.

Figure 4

A representation of the different contributions to Cl⁻ binding which describe the uptake capacity of hydrated CAC pastes containing: (a) 0 mass % CN, (b) 10 mass % CN, and (c) 30 mass % CN.

Figure 5

(a) A Nyquist plot obtained from EIS measurements depicting the bulk resistance (R_b, kΩ) of the hydrated CAC + CN mixtures (ϕ_q =0.50). (b) The effective conductivity of hydrated CAC + CN mixtures as a function of the quartz dosage. The data points represent the average of two replicates, with error bars indicating the upper and lower values.

Figure 6

A schematic of the simulated concrete bridge-deck section subject to Cl⁻ ingress at (x = 0) that features a CAC-based top-layer of thickness x_c.

Figure 7

(a) The simulated Cl⁻ concentration profiles within the OPC concrete, and, (b) The Cl⁻/OH⁻ ratio within an OPC concrete as a function of time for different cover depths. The dashed line indicates the critical Cl⁻/OH⁻ ratio at which steel corrosion initiates. (c) The calculated time to corrosion initiation (t_init) as a function of the reinforcement cover depth x_r.

Figure 8

(a) The simulated Cl⁻ concentration profiles within an OPC concrete topped with a 0 mass % CN CAC top-layer (ϕ_q = 0.50, x_c=0.025 m) after 15 years of exposure to seawater. The dashed lines show scenarios wherein a Cl⁻ binding isotherm equal to that of the OPC paste was assumed in the CAC top-layer (Bind_top = Bind_bulk), or, wherein the CAC top-layer and the OPC concrete are assumed to have similar ionic diffusivities (D_top = D_bulk). (b) The Cl⁻/OH⁻ ratio as a function of time at a cover depth x_r = 0.050 m. The horizontal dashed line in (b) indicates the critical Cl⁻/OH⁻ ratio when steel corrosion initiates. (c) The corrosion delay factor (d_f) produced by replacement of the OPC concrete with a 0 mass % CN CAC top-layer, as a function of reinforcement cover depth x_r.

Figure 9

(a) The Cl⁻/OH⁻ ratio, and, (b) The Cl⁻/NO₃⁻ ratio within OPC concrete topped with a 30 mass % CN CAC top-layer (ϕ_q = 0.50, x_c=0.025 m) as a function of time for a cover depth x_r = 0.050 m. The horizontal dashed lines in (a) and (b) indicate the critical Cl⁻/OH⁻ and Cl⁻/NO₃⁻ ratios for the initiation of steel corrosion or inability for the re-passivation of steel by NO₃⁻. (c) The corrosion delay factor (d_f) produced by replacing OPC concrete with a 30 mass % CN CAC top-layer, as a function of reinforcement cover depth x_r.

Figure 10

The corrosion delay factor relative to OPC concrete as a function of: (a) The fractional thickness of the CAC top-layer (x_c/x_r), and (b) The surface Cl⁻ concentration (C^S_Cl), for a reinforcement cover depth x_r = 0.050 m.