The Corrosion Performance of Hybrid Polyurea Coatings Modified with TiO₂ Nanoparticles in a CO₂ Environment

Abstract

To enhance the corrosion resistance of underground pipelines made of low carbon steel, nano-TiO₂-modified polyurea was applied to their surface. The anti-corrosion performance of these nano-TiO₂-modified coatings was tested by immersing them in a NACE (5 wt.% NaCl + 0.5 wt.% CH3COOH) solution under high temperatures and high CO₂ pressures. The mass variation, SEM morphology, and open-circuit potential were determined. EIS tests, neutral salt spray tests, and contact angle measurements were carried out to analyze the effect of nanoparticles on corrosion resistance. Within the same pressure range, the polyurea coating shows the highest corrosion resistance when 5% TiO₂ nanoparticles were added compared to that of polyurea coatings with 0%, 10%, and 15% TiO₂ added. Coatings with 5% TiO₂ nanoparticles showed rapid diffusion after being immersed for 96 h, indicating that the anti-corrosion performance of the coating weakened.

1. Introduction

Salt caverns have the potential to store significant amounts of gas and oil underground due to the large sealing capacity of rock salt, the inert nature of the salt structure, and the operational flexibility of injection–extraction processes [1]. These caverns are formed by the dissolution of salt in deep salt beds and domes by hot water circulating through wells drilled into the salt formation. Typically, they are located at depths of around 500 to 2000 m, with pipelines running vertically to them [2,3]. The circumferential stresses of pipelines buried at this depth were assessed to be in the range of 0 to 2.5 MPa by Lin Li et al. [4]. Those pipelines are exposed to severe corrosion environments, including flowing sediments, and Cl⁻ and CO₂ dissolved in underground water, gradually increasing geostatic stress with depth. Chloride is a well-known factor that can cause corrosion issues in steel [5,6]. It is therefore important to note that CO₂, as a component of oil field-associated gas or natural gas, exists in oil and gas fields and is difficult to eliminate. Although CO₂ itself is not corrosive, it is highly soluble in water, forming carbonic acid, which increases the concentration of hydrogen ions in water. The solubility of CO₂ in underground water increases with geostatic pressure [7].

Carbon steel is the material most commonly used in these pipelines due to its good mechanical properties and cost-effectiveness. However, it is susceptible to corrosion [8,9]. To protect the underlying pipelines against corrosion, various polymeric coatings are applied to the carbon steel. These coatings include polyurea (PU) [10], polyethyleneimine (PEI) [11], poly-dimethylamino siloxane (PDMAS) [12], epoxy coatings [13], and so on. These protective layers offer help by introducing a barrier to prevent ionic transport and electrical conduction [14]. Among them, polyurea (PU) is a synthetic high-strength elastomeric material that can be sprayed as a coating over existing structures for protection against weathering, such as by UV radiation and corrosion.

Nanoparticles with diameters less than 100 nm exhibit unique thermal, magnetic, and optical properties, as well as enhanced surface stability due to their small size, surface effects, quantum size effects, and macroscopic quantum tunneling effects [15]. Nanocomposite technology can improve mechanical properties such as the tensile strength, modulus, ultraviolet stability, corrosion resistance, and reduced permeability to gasses. TiO₂ [16,17], SiO₂ [18], ZnO [19,20], ZrO₂ [21], nano-silver [22,23], and Fe₂O₃ [24] nanoparticles have been reported to improve the performance of coatings in various environments. M. Behzadnasab et al. [21] reported that clear epoxy coatings with 2~3 wt.% ZrO₂ nanoparticles possessed the best corrosion performance in a 3.5% NaCl solution due to the higher-barrier properties and ionic resistance of the coatings, caused by possible chemical interactions between the polyurea and ZrO₂ particles. S. Bordbar et al. [22] demonstrated that adding Ag-NPs to the polyurea coating improved the anti-corrosion behavior of polyurea by 1.7 times. S.K. Dhoke et al. [24] showed that adding nano-Fe₂O₃ into an alkyd-based waterborne coating could improve its corrosion resistance, UV resistance, scratch resistance, and abrasion resistance. The coating system with a concentration of 0.3 wt.% nano-Fe₂O₃ particles provided better performance than the other systems. The addition of 4 wt.% nanometer-sized TiO₂ into unsaturated polyester (UP) resin resulted in significant improvements in mechanical properties: the tensile strength increased by 47%, bending strength by 173%, tensile elasticity modulus by 22%, bending elasticity modulus by 12%, impact strength by 60%, and elongation at break by 48% [25]. Small-sized nanofillers can occupy the void between the polyurea polymer chains, reduce the defects at the nanoscale, and improve the barrier performance of the coatings [21,23,25]. TiO₂, ZnO, SiO₂, and carbon nanotubes are the most promising nanoparticles employed in corrosion protection coatings. Due to their availability, low cost, non-toxicity, chemical stability, and UV resistance, TiO₂ nanoparticles were thus incorporated into polyurea to improve the anti-corrosion performance in this work.

Previous research confirmed that the protective performance of each coating in the salt spray test was ranked as follows: coating with 1 wt.% nano-TiO₂ > coating with 1 wt.% nano-SiO₂ > coating without nano-TiO₂ [26]. S. Radhakrishnan et al. [27] showed that coatings containing 4.18% TiO₂ exhibit the highest OCP and the lowest corrosion rate among coatings with 0%, 1.3%, 2.8%, and 4.18% TiO₂ added. It is reasonable to say that the additions of more nanosized TiO₂ would show better anti-corrosion performance. Therefore, coatings with 0%, 5%, 10%, and 15% TiO₂ added were designed in this work to find the TiO₂ addition with the best corrosion resistance.

Most research on the corrosion resistance of coatings with nanoparticles has been carried out with carbon dioxide gas dissolved in solutions at atmospheric pressure and room temperature. There are limited studies on TiO₂ nanoparticle corrosion resistance under high temperatures and high CO₂ pressures. The present work aims to demonstrate the anti-corrosion performance of polyurea containing TiO₂ nanoparticles at high temperatures and high CO₂ pressures, specifically with higher concentrations of CO₂ dissolved in solutions. Coatings with different contents of TiO₂ nanoparticles were immersed in a NACE solution (5 wt.% NaCl + 0.5 wt.% CH₃COOH) at different CO₂ pressures for various durations. After the immersion test, the mass variation, scanning electron microscopy, open-circuit potential, and electrochemical impedance spectroscopy (EIS) results were used to evaluate and analyze the corrosion performance of the coating.

2. Materials and Methods

The substrate used in this work is a P110 steel. The chemical composition is listed in

Table 1. The chemical composition of the steel (all numbers are in wt.%).

C	Si	Mn	P	S	Cr	Mo	Ni	V	Ti	Cu	Fe
0.270	0.260	1.410	0.014	0.003	0.089	0.080	0.049	0.007	0.036	0.030	Bal.

.

The steel was machined to a sheet with dimensions of 40 mm × 13 mm × 2 mm. The surface finish of the specimen was achieved by grinding with 400-, 1200-, and 2500-grit SiC abrasive paper, followed by polishing with an MD-Chem polishing cloth (Struers Ltd., Marne, France) that was lubricated with 0.04 μm OP-S Colloidal Silica suspension (Struers Ltd., Rotherham, UK) to achieve a mirror surface. Samples were ultrasonically cleaned with deionized water (15 MΩ cm) for 20 min.

The polyurea coating comprises components A and B in a volume fraction ratio of 1:1. Component A is a semi-prepolymer synthesized from 40% polyisocyanates, 30% polyol, and 30% organosilane (in volume fraction). Component B is a mixture of 30% nanopowders, 55% amino-terminated polyether, and 15% amine chain extender (in weight fraction). An epoxy polyurethane primer with a thickness of 30 ± 0.5 µm was applied to the polished and degreased steel substrates using a film applicator (Model 352, Erichsen Co., Iserbach, Germany), and dried in air for 24 h. Subsequently, the polyurea containing 0 wt.%, 5 wt.%, 10 wt.%, and 15 wt.% nano-TiO₂ was coated to the surface using the same method with an interval of 1 h. The thickness of the coating in each step was measured with an ultrasonic thickness gauge (model: 45 MG, Olympus, Toyko, Japan). The thickness of the prime and the top coating was 30 ± 0.5 µm and 150 ± 1 µm, respectively. After the completion of the coating process, specimens were left in the air for 7 days before further experiments.

Samples were immersed in solutions within a high-temperature and high-pressure reactor (3L double-body magnetic drive reactor, Dalian, China) for varying durations. Three samples were measured for each point in each test. The solution was 5 wt.% NaCl and 0.5 wt.% CH₃COOH, and it should not exceed 2/3 of the volume of the reactor. Samples were suspended in the reactor and immersed in the solution. Then, CO₂ with a purity of 99.99% was introduced for more than 1 h to remove the air in the reactor. Valves of the reactor were closed. The temperature in the reactor was increased to 80 °C. CO₂ with a purity of 99.99% continued to be pumped for a specified time until the required CO₂ pressure (0.5 MPa, 1.5 MPa, 2.5 MPa) was achieved. Samples were taken out from the reactor after specific durations (24 h, 48 h, 96 h, 192 h, and 240 h), rinsed with deionized water, and weighed in an analytical balance (ME104T/00, Mettle Toledo, Zurich, Switzerland) with 0.1 mg readability. Sample morphologies before and after corrosion tests were characterized by a scanning electron microscope (SEM, S-4800, Hitachi, Toyko, Japan). The particle size analysis was carried out by a Zetasizer Nano ZES (BeNano 180 Pro, Bettersize, Dandong, China). Contact angle measurements were carried out by a drop shape analyzer (DSA30, KRŰSS, Hamburg, Germany). Measurements were performed by the needle-in-drop sessile drop.

Corrosion tests were conducted in a magnetic-driven reaction kettle (Kemao Company, Beijing, China). The solution consisted of 5 wt.% NaCl and 0.5 wt.% CH₃COOH in deionized water (≥15 MΩcm, Merck Millipore, Darmstadt, Germany), commonly referred to as the NACE solution. The test was carried out by a conventional three-electrode system, which consists of a sample as the working electrode (working area: 4.76 cm²), a graphite electrode as an auxiliary electrode (working area: 20 cm²), and a saturated calomel electrode (SCE) as the reference electrode. Electrochemical impedance measurements (EIS) were performed at an AC amplitude of 5 mV using an electrochemical workstation (CHI660D, Chenhua Instruments, Shanghai, China). Before the impedance test, the open-circuit potential (OCP) was recorded. The frequency range was 10⁻² Hz up to 10⁵ Hz, with an amplitude of 10 mV at OCP, using five points/decade. All experiments were executed in a Faraday cage to minimize the external electronic interference. The fitting of EIS data was performed by Echem Analyst^TM (Gamry Instruments, Warminster, PA, USA) with the simplex algorithm method.

Neutral salt spray (NSS) tests lasting up to 30 days were performed for each type of sample. The testing conditions were (i) a NaCl concentration of 5% (±1%); (ii) a temperature of 35 °C (±2); (iii) demineralized water, and (iv) a pH range between 6.5 and 7.2. After NSS testing, samples were cleaned and dried. The appearance was used to assess the protection effect.

3. Results and Discussion

3.1. Mass Variation

Figure 1 illustrates the mass variation in samples with different contents of TiO₂ nanoparticles roughly possessing the same tendency. Samples with polyurea coatings containing no nanoparticles lose weight from the beginning due to the weak protection of coating under 0.5 MPa, 1.5 MPa, and 2.5 MPa CO₂ pressures. Samples with polyurea coatings containing TiO₂ particles gain weight quickly at the beginning (around 40 h) under different CO₂ pressures. The initial gain weight in the nanoparticle coatings is attributed to the penetration and adhesion of corrosion media. The point of 10% TiO₂ nanoparticle coatings after 100 h of testing in Figure 1a is a turning point where the rate of mass variation changes. This rate of weight gain slows down. The rate at which the coating absorbs the electrolyte solution becomes slower until the coating is saturated. It slows down the mass variation in the sample. And some points, such as the mass variation in 10% TiO₂ and 15% TiO₂ coatings after 100 h of testing under 1.5 MPa in Figure 1b, exhibit negative values, indicating a decrease in mass. As the electrolyte solution penetrates to the coating/matrix metal interface, corrosive microcells form, and the matrix metal corrodes. There is a small corrosion product at this stage, and it can be transported outside the coating through the microporous channel in the coating. As the corrosion products continue to diffuse into the solution, the overall mass decreases. As corrosion progresses, the corrosion rate accelerates, and the volume of corrosion products increases. Thus, the corrosion products are difficult to transport through the microporous channels, and gradually accumulate under the coating, resulting in the gradual mass increase. The final stage of the mass increase can be explained by the accumulation of corrosion products, suggesting that the rate of the production of these products is greater than their rate of diffusion away from the coating.

3.2. Microstructures

Figure 2 shows the initial SEM morphologies of polyurea coatings with different TiO₂ contents. Coatings with 0 %, 5%, and 10% TiO₂ nanoparticles show good surface condition without any cracks or defects. In contrast, the coating with 15% TiO₂ appears slightly rougher. Higher resolution images in the upper right corner of Figure 2 show the distribution of nanoparticles in polyurea coatings. Figure 3 shows the size distribution of nanoparticles by Dynamic Light Scattering (DLS). It shows that the dispersion of 5, 10, and 15 wt.% TiO₂ particles in polyurea coatings was in the range of 6.3~15.6, 9.0~27.9, and 23.6~45.7 nm, respectively, which is also consistent with SEM images in Figure 2. This is probably due to the poor dispersion of small nanoparticles. The higher the content of nanoparticles, the more likely they will agglomerate. Nanoparticles of smaller sizes have higher specific surface energy and specific surface area, which puts them in the thermodynamic unstable state. They are likely to aggregate into clusters and lose the inherent characteristics of nanoparticles. The clusters can be regarded as defects in the coating, and they show weak protection. Figure 4 presents the morphologies of the aforementioned samples tested under 2.5 MPa CO₂ pressure for 240 h. All the samples’ surface conditions became worse after the immersion test, especially the one without TiO₂ nanoparticles. For the coating without TiO₂ nanoparticles, micro-cavities and micro-cracks show up on the sample surface (Figure 4a). Samples coated with 5%, 10%, and 15% nanoparticles show some micro-cavities, but no obvious micro-cracks. The additions of nanoparticles can enhance the crosslinking structure of coatings and improve the interface between nanoparticles and resin, which are valuable for enhancing the barrier performance of coatings. The coatings with 0 wt.% and 15 wt.% TiO₂ show small cavities. It may be speculated that the insulation of the polyurea coating with 0 wt.% and 15 wt.% TiO₂ was compromised after 240 h of immersion. The corrosive media may reach the substrate through holes and cavities.

3.3. Open-Circuit Potential

Figure 5 shows the coatings with different concentrations of TiO₂ nanoparticles under three CO₂ pressures: 0.5 MPa, 1.5 MPa, and 2.5 MPa. The corrosion potential decreases significantly within the first 24 h of testing. It fluctuates slightly with the extension of immersion time, but it overall shows a downward trend. Notably, regardless of the CO₂ pressure, the coating with 5% TiO₂ exhibits higher OCP than that with 10% and 15% TiO₂ throughout the entire testing periods. OCP reflects the thermodynamic stability of the interface. The initial large drop of the OCP is attributed to the solution penetrating into the coating and establishing a liquid channel to the metal surface. Subsequently, dissolved ions (H⁺, HCO₃⁻, CO₃²⁻, etc.) reach the substrate through the established liquid channel and change the charge distribution. This process accelerates the electrochemical reaction in which the cathodic reaction produces hydrogen. The coating begins to wrinkle and peel off at a high corrosion rate. The large negative shift in the OCP is consistent with the penetration of the solution, which accelerates the anodic reaction. With the extension of time, approximately after 50 h, the OCP only varies within a small range, indicating that the equilibrium is almost established at the interface.

3.4. EIS Analysis

3.4.1. Effect of CO₂ Pressure

EIS is an effective analytical technique for measuring the electrical response of an electrochemical system by applying a sinusoidal perturbation. EIS data can be correlated with some key physical properties, such as diffusion coefficients, chemical reaction rates, and microstructural characteristics of the electrochemical system. However, an EIS analysis is prone to potential ambiguities in interpretation. Rigorous modeling and its combination with statistics can be used to overcome these challenges and enhance the insight one can gain from EIS. EIS plays an important role in monitoring and predicting the degradation of organic coatings [28]. To investigate the effect of CO₂ pressure, immersion time, and TiO₂ content on the electrochemical stability of polyurea coating, EIS was carried out on coatings with different contents of TiO₂ nanoparticles under different CO₂ pressures. Figure 6 and Figure 7 present the Nyquist and Bode plots of polyurea coatings with 0, 5, 10, and 15% TiO₂ nanoparticles under 0.5, 1.5, and 2.5 MPa CO₂ pressures after 24 h and 240 h immersion tests.

When the sample was immersed for 24 h, the electrolyte slowly permeated the coating, leading to the change in the Bode impedance plot at low frequencies. The anti-corrosion ability of the coating was reflected in changes in its impedance and impedance modulus. Thus, coatings with a higher low-frequency impedance modulus and impedance exhibit better anti-corrosion performance. At 24 h, the impedance spectra under 0.5 MPa CO₂ pressure all have double capacitive reactance arcs, and the coating impedance modulus (|Z|) reaches more than 10⁷ Ω·cm², indicating that the coating has a good protective effect on the substrate [29]. As the CO₂ pressure increases from 0.5 MPa to 1.5 MPa, |Z| is diminished by half an order of magnitude. When the CO₂ pressure increases to 2.5 MPa, the impedance modulus continues to be lowered to 10⁶ Ω·cm² (Figure 6c). Previous research shows that the organic coating lost the protective effect on the substrate when its resistance value was lower than 10⁶ Ω·cm² [30]. Otherwise stated, the polyurea coating containing 5% TiO₂ nanoparticles provides little protection to the substrate once exposed in a 2.5 MPa CO₂ environment. At 240 h, |Z| remains highest under 0.5 MPa CO₂ pressure, exceeding 10⁶ Ω·cm², which has a good protective effect on the substrate. As the CO₂ pressure increases from 0.5 MPa to 1.5 MPa, |Z| continuously decreases to 10⁶ Ω·cm². When the CO₂ pressure increases to 2.5 MPa, |Z| decreases to 10⁵ Ω·cm². Figure 7 shows that no matter how much TiO₂ nanoparticles are contained, |Z| decreases with the increase in CO₂ pressure, clearly indicating that higher pressures accelerate the corrosion of steel with the polyurea coating. Most organic coatings exhibit semi-permeable membrane characteristics, and the increase in pressure is favorable to the penetration of a solution, oxygen, and other corrosive media into the coating [31].

The equivalent circuits of polyurea coatings with different TiO₂ contents under different CO₂ pressures immersed in the NACE solution for 24 h and 240 h are shown in Figure 8a and Figure 9a, respectively. The data for equivalent circuits in Figure 8a and Figure 9a are referred to as Appendix A 

Table A1. List of values of each component in equivalent circuit in Figure 7a.

TiO2 Content	CO2 (MPa)	Rs (ohm)	Qc (F)	Rc (ohm)	Qdl (F)	Rct (ohm)
0%	0.5	2376	1.728 × 10−10	3.083 × 105	1.399 × 10−8	3.084 × 106
	1.5	803.2	2.100 × 10−10	2.555 × 104	1.274 × 10−7	2.545 × 105
	2.5	691	2.224 × 10−10	1.325 × 104	4.1924 × 10−8	1.325 × 105
5%	0.5	1.105 × 104	9.061 × 10−15	3.353 × 105	5.599 × 10−9	7.035 × 106
	1.5	1.335 × 104	2.371 × 10−12	3.074 × 105	8.342 × 10−9	7.754 × 105
	2.5	1.126 × 104	2.752 × 10−10	2.046 × 105	3.245 × 10−5	625
10%	0.5	1.181 × 104	9.039 × 10−12	3.066 × 105	1.435 × 10−9	1046
	1.5	1.205 × 104	9.715 × 10−11	2.386 × 105	1.35 × 10−8	786
	2.5	1.102 × 104	1.098 × 10−10	2.150 × 105	1.662 × 10−7	1053
15%	0.5	1.011 × 104	2.095 × 10−11	2.846 × 105	4.004 × 10−9	4.166 × 105
	1.5	1.015 × 104	4.656 × 10−11	2.277 × 105	1.898 × 10−7	1.88 × 105
	2.5	1.131 × 104	3.148 × 10−10	2.109 × 104	2.104 × 10−8	4.886 × 104

and

Table A2. List of values of each component in equivalent circuit in Figure 8a.

TiO2 Content	CO2 (MPa)	Rs (ohm)	Qc (F)	Rc (ohm)	Qdl (F)	Rct (ohm)	Warburg
0%	0.5	1195	1.586 × 10−10	3.047 × 105	5.468 × 10−11	3.243 × 105	9.342 × 105
	1.5	2012	1.630 × 10−10	2.820 × 105	1.260 × 10−10	6494	4.152 × 105
	2.5	1953	2.871 × 10−10	1292	3.921 × 10−9	2181	4.470 × 105
5%	0.5	5420	5.439 × 10−11	1.583 × 105	1.587 × 10−7	6.966 × 105	1.142 × 104
	1.5	8266	4.087 × 10−10	3.982 × 104	1.094 × 10−6	1.15 × 105	3.002 × 105
	2.5	2746	6.014 × 10−10	1.089 × 104	2.655 × 10−6	5.663 × 104	1.733 × 105
10%	0.5	9005	3.236 × 10−11	7.922 × 104	0.004749	17.9	1.97 × 107
	1.5	8695	5.054 × 10−11	3.145 × 104	1.58 × 10−8	2.11 × 105	5.998 × 106
	2.5	3393	1.258 × 10−10	875.4	8.103 × 10−6	0.5669	1.872 × 106
15%	0.5	7635	6.189 × 10−11	4.57 × 104	4.243 × 10−7	3.275 × 105	6.354 × 106
	1.5	7632	2.623 × 10−10	4590	1.941 × 10−6	8.314 × 104	1.072 × 106
	2.5	2651	2.921 × 10−10	655.2	4.86 × 10−6	4.137 × 104	1.208 × 107

. Both equivalent circuits have two sets of R-C circuits, due to the two-time constant shown in the Nyquist plots. The first-time constant is related to the process of the organic layer, and the second-time constant (the second loop with or without the Warburg tail) accounts for the corrosion process at the coating/substrate interface. The equivalent circuits consist of solution resistance (R_s), coating capacitance (Q_c), coating resistance (R_c), double-layer capacitance (Q_dl), and charge transfer resistance (R_ct). Moreover, Figure 8a includes one more component, Warburg impedance (Z_w) connecting in series with R_ct. The main basis for adding it is the straight line with a 45° angle to the abscissa (Figure 6f). The time when it becomes visible depends on the double-layer capacitance; hence, not all Nyquist plots at 240 h have the straight line. The process is controlled by the Faraday process and diffusion. After 240 h, corrosion products accumulated at the interface and the polyurea coating acted as a barrier that hindered and inhibited the diffusion of these corrosion products into the solution. In this case, Z_w is introduced to the equivalent circuit to simulate the diffusion process. Figure 8b and Figure 9b show the value of Q_c and R_c and how they change with CO₂ pressure. R_c decreases with a CO₂ pressure increase and Q_c shows the opposite tendency. This further confirms that higher pressures weaken the protection of polyurea coatings.

Figure 8b and Figure 9b show the coating resistance and the coating capacitance after having the samples immersed for 24 h and 240 h. With the increase in CO₂ pressure, the coating resistance decreases and the coating capacitance increases.

3.4.2. Effect of Content of TiO₂ Nanoparticles

The polyurea coating containing 5% nanoparticles possesses slightly higher OCP than those with 0%, 10%, and 15% nanoparticles, as indicated in Figure 5. For a clear comparison, |Z| of polyurea coatings with different TiO₂ nanoparticles under 1.5 MPa CO₂ pressure tested after varying immersion times is summarized in Figure 10. |Z| of polyurea coatings increases with TiO₂ added from 0% to 5%, and then decreases with more TiO₂ nanoparticle additions. The coating with 5% TiO₂ nanoparticles consistently demonstrates the highest |Z| from 24 h to 240 h.

In TiO₂ nanoparticles, titanium atoms are connected by bridging oxygen, and this structure has a certain hydrophobicity [32]. The condensation reaction between hydrophobic groups and terminal amine or hydroxyl groups in polyurea augments the number of chemical crosslinking points and fills the micropores. Hence, the compactness and hydrophobicity of the coatings are improved. This occurred when 5% TiO₂ nanoparticles were added. However, as more nanoparticles were added to the coating, such as 10% or 15%, they accrete to larger particles, which is unfavorable for the reaction with amine or hydroxyl groups in polyurea [32], as shown in Figure 2, Figure 3 and Figure 4. The compactness and hydrophobicity of the coating with 10% or 15% nanoparticles are compromised compared with that with 5% nanoparticles. Furthermore, the larger particles cause more bumps and holes on the polyurea surface instead of filling the micropores, leading to more inhomogeneities.

H.W. Shi et al. [26] reported the improved protective performance of coatings reinforced with 1 wt.% nanometer-sized TiO₂ in 3.5% NaCl solution immersion testing. The coating impedance modulus (|Z|) in a Bode plot was less than 10⁷ Ω·cm². The effect of TiO₂ nanoparticles in the present work was studied under high CO₂ pressure at 80 °C, more harsh environments than the environment in reference [26]. The degradation of coating was accelerated, and |Z| in a Bode plot for coating with 5% TiO₂ maintained more than 10⁷ Ω·cm², higher than the |Z| in reference [26]. The nanosized copper was added to improve the polyurea corrosion resistance [33]. The coating resistance with nanosized Cu was 2.474 × 10⁵ Ω cm² after being immersed for 0 h. The coating resistance of polyurea coating with 5% TiO₂ in the present work was measured to be 3.353 × 10⁵ Ω cm² after being immersed for 24 h, indicating that the polyurea coating with TiO₂ has better corrosion resistance.

3.4.3. Degradation of Hybrid Polyurea Coating (Effect of Immersion Time)

To further investigate the degradation of polyurea coatings, EIS was carried out on samples after different immersion durations. The Bode plots and Nyquist diagrams are presented in Figure 11 and Figure 12.

At the initial stage of immersion (0 h), only a one-time constant is evident, showing that the corrosion protection performance of coating is mainly under the control of barrier effects. This can be proved by impedance modulus |Z| in the low-frequency region (around 10⁸ Ω·cm²), indicating the excellent protective effect of coating at this time. Also, as revealed by SEM in Figure 2, the coating is rarely defective. The Nyquist diagram shows a single capacitive reactance arc, which is simulated by a simple electric circuit consisting of solution resistance (R_s), coating capacitance (Q_c), and coating resistance (R_c), as shown in Figure 13.

As the immersion time increases to 24 h, |Z| in the low-frequency region decreases rapidly. The test solution gradually penetrates the coating and fills up the defects (pores, cracks, etc.), which leads to the decrease in coating resistance and the increase in coating capacitance. |Z| in the low-frequency region soon decreases to about 10⁷ Ω·cm² (Figure 11). The solution reaches the interface between the coating and the substrate, accompanied by the accumulation of corrosion products. The corrosion products severely affect the combination of the coating and the substrate, resulting in blisters and swelling of the coating. This explains the decrease in the impedance compared with that at 0 h. The Nyquist plot shows that another capacitive reactance arc appears in the low-frequency region. Although it is not a complete arc, it should still be a two-time constant circuit. The fitted equivalent circuit is shown in Figure 8a.

After 24 h, Nyquist plots exhibit the characteristic of a two-time constant. A comparison of the Nyquist plot of 96 h and 192 h reveals some similarities in the high-frequency regions, while the main difference appears in the low-frequency region. It can be speculated that the coating capacitance no longer increases significantly after 96 h. Instead, the time constant in the low-frequency region representing the double-layer capacitance (Q_dl) and polarization resistance (R_t) has a dominant effect on the electrochemical system. The fitted equivalent circuit is illustrated in Figure 9a including a Warburg impedance to simulate this diffusion process.

After 192 h, corrosion products continued to accumulate between the metal surface and the polyurea coating. At this stage, they can be seen as a single layer instead of just adhering to polyurea coating. Hence, the equivalent circuit was adjusted, as shown in Figure 14. It also corresponds to the mass variation, which indicates that the samples gain weight at the final stage (Figure 1).

The corrosion protection mechanism of coatings with nanoparticles during the immersion in the electrolyte and the electrochemical corrosion reaction on the interface between the steel and coatings is depicted schematically in Figure 15. The starting state is shown in Figure 15a, corresponding to the Nyquist plot in Figure 12a. As immersion time increased, a liquid channel to the metal surface was established. Subsequently, dissolved ions (H⁺, HCO₃⁻, CO₃²⁻, etc.) were transported to the interface between the substrate and the coating and they changed the charge distribution corresponding to the Nyquist plot in Figure 12b. Then, the steel started to corrode as an anode; the anodic reaction was as follows:

$$\mathrm{F}\mathrm{e}\to {\mathrm{F}\mathrm{e}}^{2+}+2\mathrm{e}$$

(1)

$$\mathrm{F}\mathrm{e}+{\mathrm{H}\mathrm{C}\mathrm{O}}_3^{-}\to {\mathrm{F}\mathrm{e}\mathrm{C}\mathrm{O}}_3+{\mathrm{H}}^{+}+2\mathrm{e}$$

(2)

The substrate steel at adjacent defects can serve as a cathode. The cathodic reaction may include reduction reactions of H⁺, H₂O, H₂CO₃, and HCO₃⁻, as shown in Figure 15e.

$$2{\mathrm{H}}_3{\mathrm{O}}^{+}+2\mathrm{e}\to {\mathrm{H}}_2+2{\mathrm{H}}_2\mathrm{O}$$

(3)

$${\mathrm{H}}_2\mathrm{O}+2\mathrm{e}\to 2{\mathrm{O}\mathrm{H}}^{-}+{\mathrm{H}}_2$$

(4)

$$2{\mathrm{H}}_2\mathrm{C}{\mathrm{O}}_3+2\mathrm{e}\to 2\mathrm{H}\mathrm{C}{\mathrm{O}}_3^{-}+{\mathrm{H}}_2$$

(5)

$$2\mathrm{H}\mathrm{C}{\mathrm{O}}_3^{-}+2\mathrm{e}\to 2\mathrm{C}{\mathrm{O}}_3^{2-}+{\mathrm{H}}_2$$

(6)

The overall corrosion reaction is shown as follows:

$$\mathrm{F}\mathrm{e}+\mathrm{C}{\mathrm{O}}_2+{\mathrm{H}}_2\mathrm{O}\to \mathrm{F}\mathrm{e}\mathrm{C}{\mathrm{O}}_3+{\mathrm{H}}_2$$

(7)

As the immersion time increased, the liquid channel or the micro-cavities interconnected with each other and accelerated the corrosion process. The volume of corrosion products increased and accumulated under the coating. When it went to 192 h or 240 h, more serious corrosion happened until failure.

For the coatings with 15% nanoparticles, there are more micro-cavities and defects than coatings with 5% and 10% nanoparticle additions, as shown in Figure 3, which mainly resulted from the aggregation of nanoparticles. These micro-cavities and defects helped to transport the corrosive media and accelerated the corrosion.

3.5. NSS Testing Results

The durability of coatings with different contents of nanoparticles was evaluated by the blister. Figure 16 shows photographs of coating samples with 30 days of exposure in a neutral salt spray chamber. Blisters are visible to the naked eye after 30 days of testing. Most blisters are located near the edge of the samples. The areas of defects are 0.15%, 0.02%, 0.08%, and 0.14% for polyurea coatings with 0%, 5%, 10%, and 15% TiO₂. The best corrosion performance in NSS testing is the coating with 5% TiO₂ nanoparticles, which coincides with the electrochemical testing results in Section 3.4.2, implying that the proper additions of nanoparticles can effectively improve corrosion resistance.

3.6. Contact Angle

Contact angle measurements were carried out to understand the surface hydrophobicity behavior. Figure 17 shows the contact angles for coatings with 5% TiO₂ nanoparticles after being immersed for different durations under 1.5 MPa CO₂ pressure. The contact angle value gradually decreases with the increase in immersion time. When the immersion time is 0 h, 24 h, 96 h, 192 h, and 240 h, the contact angles are 88.96°, 86.86°, 83.66°, 82.15°, and 77.89°, respectively. The contact angle was reduced by 12%, and the wettability changed significantly. As the surface roughness increases, the contact angle decreases [34]. It is consistent with the SEM morphology change (Figure 2 and Figure 4). The smaller contact angle means better hydrophilicity. The corrosive medium can stay on the coating surface for a longer time, and it helps the corrosion process. The surface energy of the coating shows an upward trend with the increase in immersion duration. It increased by about 34% after 240 h of immersion. The significant increase in surface energy is probably associated with the increase in surface polar functional groups caused by the cleavage of chemical bonds.

4. Conclusions

The present work investigated the corrosion effect of polyurea modified with TiO₂ nanoparticles coated on P110 steel surfaces. Coatings with different contents of TiO₂ nanoparticles were immersed in a NACE solution under different CO₂ pressures for various durations. It was found that the polyurea coating with 5% TiO₂ nanoparticles exhibited the best corrosion resistance in the NACE solution under CO₂ pressures of 0.5~2.5 MPa among polyurea coatings with 0, 5, 10, and 15% TiO₂ added. This coating maintained relatively good corrosion resistance after the immersion test in 0.5 MPa and 1.5 MPa CO₂ pressure for 120 h. The degradation process indicated that rapid diffusion occurred in the polyurea coating with 5% TiO₂ nanoparticles added after 96 h of immersion, suggesting that its anti-corrosion performance was weakened.