Influence of Zirconia and Organic Additives on Mechanical and Electrochemical Properties of Silica Sol-Gel Coatings

This paper presents the result of the investigation of organically modified silica (ORMOSIL)-zirconia coatings used to enhance their protective properties, namely corrosion and scratch resistance. Two different materials, i.e., SiO2/ZrO2 and SiO2/GPTMS/ZrO2, were synthesized, measured, and analyzed to find the difference in the used organosilane precursor (dimethyldiethoxysilane and (3-glycidoxypropyl)trimethoxysilane, respectively). SiO2/ZrO2 coatings showed higher hardness than SiO2/GPTMS/ZrO2. Moreover, the value of polarization resistance (Rp) for SiO2/GPTMS/ZrO2 coated 316L steel relative to the uncoated one was obtained. It was nearly 84 times higher. The coating delamination was observed with load 16N. Additionally, the corrosion mitigation for 316L coated by SiO2/GPTMS/ZrO2 was observed even after extended exposure to corrosion agents.


Introduction
Many types of coatings, including inorganic or organic ones (paints, enamels, and lacquers) produced to protect against corrosion were developed as a result of modern technological advances. It should be added that the research that contributed to these latest scientific developments was conducted on corrosive degradation. The sol-gel method allows good corrosion protection capabilities to be attained thanks to the possibility of using the numerous types of sol-gel synthesis precursors, other reagents (e.g., inhibitors), and chemical or physical treatments (e.g., etching or polishing). It also makes it possible to obtain coatings with wear resistance properties in a corrosive environment [1][2][3][4][5][6][7]. More and more emphasis is being placed on developing fluoride-free materials for corrosion mitigation [5]. Taking into consideration the elimination of fluorinated compounds used in corrosion protection, the application of various sol-gel oxide matrices modified by organic groups (ORMOSIL), inhibitors [8][9][10], or nanoparticles [10][11][12] may be a successful alternative.
Inorganic sol-gel oxides are the simplest way to achieve corrosion mitigation [13]. This group of materials is characterized by good adhesion to metal substrates, thus they are often applied as interlayers [14]. During the research on such types of materials, it was discovered that they have the ability to create many Van der Waals interactions, which can be transformed into stable covalent bonds [15]. Additionally, inorganic coatings show better wear resistance than pure organic ones, e.g., those containing polymer compounds. Typical inorganic sol-gel coatings for corrosion mitigation are obtained as simple oxides, The silica and two different silica-zirconia hybrid materials were obtained by the sol-gel method with the reagents set as presented in Table 1. The organically modified SiO 2 based matrices were prepared with tetraethoxysilane 98% (TEOS) and dimethyldiethoxysilane 98% (dMdEOS) as oxide precursors, both from Sigma Aldrich (Darmstadt, Germany). The base SiO 2 sol was used as an interlayer and for further modification. The zirconia modification of organically modified silica matrices (SiO 2 /ZrO 2 ) was obtained by mixing the base SiO 2 (A) sol with ZrO 2 (B) sol (after 1 h following reaction start, sol A was mixed with B, Table 1) based on zirconium(IV) butoxide 80% in n-butanol (ZrOBu, Sigma Aldrich). Both sols, SiO 2 and ZrO 2 , were homogenized together for 0.5 h. Another kind of organically modified silica-zirconia hybrid (SiO 2 /GPTMS/ZrO 2 ) was obtained by mixing an organically modified SiO 2 (C) sol with ZrO 2 (B) sol (1 h after reaction start, C was mixed with B, Table 1), as mentioned above. The organically modified silica materials SiO 2 /GPTMS/ZrO 2 were conducted by using (3-glycidoxypropyl)trimethoxysilane 97% (GPTMS, Alfa Aesar, Kandel, Germany) in a mixture with TEOS as a precursor. All materials were synthesized in acid condition with the use of chydrochloric acid 31-38% (HCl, Stanlab, Lublin, Poland) and/or acetic acid 99.5% (CH 3 COOH, Chempur, PiekaryŚląskie, Poland) in amounts presented in Table 1. Additionally, to stabilize zirconia sol, by complexing a zirconium cation in the zirconium terakis-acetylacetone configuration [34] (it slows down hydrolysis and condensation), the acetyloacetone 99.6% (AcAc, Alfa Aesar) was added. The solvents for obtaining materials were ethanol 96% (EtOH, POCH, Gliwice, Poland) and butanol 99.9% (BuOH, POCH). All reagents were used as received without further purification. The used volume ratio and hydrolysates configuration are presented in Table 1 below. A block diagram of the obtained sol-gel network is presented in Figure 1. further purification. The used volume ratio and hydrolysates configuration are presented in Table 1 below. A block diagram of the obtained sol-gel network is presented in Figure  1.

Preparation of Steel Substrates
Structural, low carbon steel (P265GH) and austenitic, stainless steel (316L) were used as substrates. The chemical compositions of steels are shown in Table 2 (according to  PN/EN 10027-2 and EN 10273). Before coating application, the surface of P265GH steel samples was sanded with sandpaper (grade 800) (Ra = 137 ± 14 nm after sanding, according to the PN-ISO 4288 standard with 20 measurement points per an individual sample, 0.8 mm length of measurement section, Olympus, LEXT OLS4000, Tokyo, Japan), cleaned in detergent and distilled water, and then etched in sulfuric acid (12%). After that, P265GH substrates were cleaned in distilled water and ethanol and left to dry at room temperature (20.5-21.5 °C). After sanding, the 316L steel samples (Ra = 194 ± 9 nm, according to the PN-ISO 4288 standard with 20 measurement points per an individual sample, 0.8 mm length of measurement section, Olympus, LEXT OLS4000) were rinsed in distilled water and etched in hydrochloric acid (9%), then rinsed with distilled water and ethanol and left to

Preparation of Steel Substrates
Structural, low carbon steel (P265GH) and austenitic, stainless steel (316L) were used as substrates. The chemical compositions of steels are shown in Table 2 (according to PN/EN 10027-2 and EN 10273). Before coating application, the surface of P265GH steel samples was sanded with sandpaper (grade 800) (R a = 137 ± 14 nm after sanding, according to the PN-ISO 4288 standard with 20 measurement points per an individual sample, 0.8 mm length of measurement section, Olympus, LEXT OLS4000, Tokyo, Japan), cleaned in detergent and distilled water, and then etched in sulfuric acid (12%). After that, P265GH substrates were cleaned in distilled water and ethanol and left to dry at room temperature (20.5-21.5 • C). After sanding, the 316L steel samples (R a = 194 ± 9 nm, according to the PN-ISO 4288 standard with 20 measurement points per an individual sample, 0.8 mm length of measurement section, Olympus, LEXT OLS4000) were rinsed in distilled water and etched in hydrochloric acid (9%), then rinsed with distilled water and ethanol and left to dry in room temperature (20.5-21.5 • C). The 316L steel surface was not sanded (assuming that coating would be deposited on an existing installation system in which an ideally treated surface was not possible).

Obtained Coatings and Stabilization
The ten-layer coatings were obtained by the dip-coating methods with controlled process parameters, such as dipping rate (34.12 mm/min), immersion time (60 s for the first layer, 30 s for the second layer, and 15 s for the other layers), and pulling out rate (34.12 mm/min). Each layer was deposited after a 24 h break at room temperature (20.5-21.5 • C) in open air. The sequence of layers in the obtained coatings is presented in Figure 2 ( Figure 2 only represents how thick a given coating was during subsequent layers deposition). In both cases, the organically modified SiO 2 interlayer (the layer nearest to the substrate) coatings was stabilized at 250 • C in open air (laboratory drying oven, POL-EOKO-APARATURA SP.J., SL 53 STD, WodzisławŚląski, Poland). After the sixth and the tenth layers, the samples were dried as follows: SiO 2 /ZrO 2 set was stabilized at 250 • C for 3 h, and SiO 2 /GPTMS/ZrO 2 set was stabilized at 120 • C for 1 h in open air (POL-EOKO-APARATURA SP.J., SL 53 STD). dry in room temperature (20.5-21.5 °C). The 316L steel surface was not sanded (assuming that coating would be deposited on an existing installation system in which an ideally treated surface was not possible).

Obtained Coatings and Stabilization
The ten-layer coatings were obtained by the dip-coating methods with controlled process parameters, such as dipping rate (34.12 mm/min), immersion time (60 s for the first layer, 30 s for the second layer, and 15 s for the other layers), and pulling out rate (34.12 mm/min). Each layer was deposited after a 24 h break at room temperature (20.5-21.5 °C) in open air. The sequence of layers in the obtained coatings is presented in Figure  2 ( Figure 2 only represents how thick a given coating was during subsequent layers deposition). In both cases, the organically modified SiO2 interlayer (the layer nearest to the substrate) coatings was stabilized at 250 °C in open air (laboratory drying oven,POL-EOKO-APARATURA SP.J., SL 53 STD, Wodzisław Śląski, Poland ). After the sixth and the tenth layers, the samples were dried as follows: SiO2/ZrO2 set was stabilized at 250 °C for 3 h, and SiO2/GPTMS/ZrO2 set was stabilized at 120 °C for 1 h in open air (POL-EOKO-APARATURA SP.J., SL 53 STD).

SEM and EDX
The surface morphology, the elemental composition, and the distribution of chemical elements of the obtained samples were examined by scanning electron microscopy (SEM)

SEM and EDX
The surface morphology, the elemental composition, and the distribution of chemical elements of the obtained samples were examined by scanning electron microscopy (SEM) with an EDX detector (HITACHI S-3400N, Tokyo, Japan). For each sample, 3 measurement points were selected. The most representative results are presented in Section 3.1.

Raman Spectroscopy
The chemical composition and the types of bonding in the obtained materials were examined by Raman spectroscopy (Raman spectrometer LabRAM HR800 HORIBA JOBIN YVON, Lonqjumeau, France). The Raman spectra were measured in the range 100-4000 cm −1 using an argon laser with a 514.5 nm wavelength. For the individual samples, 6 measured points were performed. The most representative spectra are presented in the manuscript.

Scratch Test
The adhesion of the obtained coatings was determined with Micro Combi Tester (MCT), CSM Instruments, Graz, Austria according to the PN-EN 1071-3:2007 standard. A Rockwell indenter, Graz, Austria with a diameter of 100 µm was used to perform 3 scratches on individual samples, and each of them was 3 mm long. A load in the range from 30 mN to 20 N with linearly increasing speed of 1.5 mm/min was used. During the measurement, normal force (F n ), friction coefficient (µ), and penetration depth (Pd) were recorded. Based on these parameters (FN, µ, Pd), the adhesion of the coatings to the substrate was specified, with characteristic loads (LC): LC 1 -cracking of coating, LC 2 -characteristic chipping of coatings, LC 3 -penetration of the coating into the substrate in the middle of the scratch.

Nanointendation Test
Hardness and the Young modulus were specified by the nanoindentation test, using Micro Combi Tester (MCT), CSM Instrument, with a pyramidal Berkovich tip. ISO 14577-1 standard was used to perform nanoindentation tests. The tests were taken by the load of the normal force of 0.1 mN and the rate of loading of 0.6 mN/min. For an individual sample, 12 measure points with a 15 s pause between loading and unloading were performed. The Young modulus was specified using the Oliver-Pharr method.

Electrochemical Test
Electrochemical tests were carried out in 3% NaCl solution after 0 h and 24 h of sample exposure in the solution with an SI 1286 potentiostat, Schlumberger, Houston, TX, USA. A three-electrode system with a calomel electrode as the reference electrode and a platinum electrode as the auxiliary electrode was used. The stationary potential and subsequently the polarization resistance, which used the Stern-Geary law (defined by R p = ∆E/∆i equation) in the range of −0.015 V to +0.015 V, were measured.
The open circuit potential (OCP) was recorded until changes were smaller than 5 mV/5 min. The polarization curves were determined in the potential range from −0.75 V to +0.5 V relative to the saturated calomel electrode (SCE). The investigated samples were polarized at a potential change rate of 1 mV/s. Figure 3 shows the morphology of uncoated P265GH (S1) and 316L(S2) steels. The surface of P265GH steel was characterized by the significant growth of surface area and porosity. Characteristic features were observed on the uncoated 316L surface. The possible reason for the occurrence of these features is described in Section 2.2. Figure 3 shows the morphology of the obtained coatings on both steel substrates. The sample with a SiO 2 /ZrO 2 coating ( Figure 3A1) was characterized by a significant growth of structure. This meant that the specific surface area increased, which may have increased the number of potential points of interaction between the corrosive medium and the sample surface [35]. The sample with a SiO 2 /GPTMS/ZrO 2 coating ( Figure 3C1) was smoother with some inhomogeneous inclusions (marked by white circles). These inclusions were not located on the coating surface (SiO 2 /ZrO 2 ) but in the volume of the coating (SiO 2 /GPTMS/ZrO 2 ). Both coatings, SiO 2 /ZrO 2 ( Figure 3A2) and SiO 2 /GPTMS/ZrO 2 ( Figure 3C2), on P265GH substrates were characterized by a regular distribution of Si and Zr, as can be seen in the result of EDX mapping. For both coatings on 316L substrates ( Figure 3B1,D1), the inclusions were not observed. Furthermore, the result of EDX mapping ( Figure 3B2,D2) also showed the regular distribution of Si and Zr in SiO 2 /GPTMS/ZrO 2 , similarly as in the case of SiO 2 /ZrO 2 ( Figure 3A2,C2). on the coating surface (SiO2/ZrO2) but in the volume of the coating (SiO2/GPTMS/ZrO2). Both coatings, SiO2/ZrO2 ( Figure 3A2) and SiO2/GPTMS/ZrO2 ( Figure 3C2), on P265GH substrates were characterized by a regular distribution of Si and Zr, as can be seen in the result of EDX mapping. For both coatings on 316L substrates ( Figure 3B1,D1), the inclusions were not observed. Furthermore, the result of EDX mapping ( Figure 3B2,D2) also showed the regular distribution of Si and Zr in SiO2/GPTMS/ZrO2, similarly as in the case of SiO2/ZrO2 ( Figure 3A2,C2).  Figure 4 shows the Raman spectra for uncoated carbon steel P265GH and the steel with coatings. For the uncoated steel samples, the characteristic bands for oxide iron were detected.

Scratch Test
The scratch resistance and the damages resulting from the tests are shown in Figures  6 and 7. The coatings with organic compounds (SiO2/GPTMS/ZrO2) showed elastic deformations along the cracks, which was consistent with the literature data [17]. Additionally, for this type of coating, acoustic emission was not recorded during the test, which indi-

Scratch Test
The scratch resistance and the damages resulting from the tests are shown in Figures 6 and 7. The coatings with organic compounds (SiO 2 /GPTMS/ZrO 2 ) showed elastic deformations along the cracks, which was consistent with the literature data [17]. Additionally, for this type of coating, acoustic emission was not recorded during the test, which indicated the lack of brittle cracking of the coating materials [17]. For inorganic coating materials (SiO 2 /ZrO 2 ), the deformations characteristic for the fragile type of destruction were observed. In particular, the acoustic emission and the deformed cracks of the conformal type were detected. Both of the obtained materials showed similar adhesion to both types of steel, i.e., P265GH and 316L. Each of the analyzed deformation characteristics for the coatings on P265GH steel were also observed for the coatings on the 316L steel. This is a valuable result, as it is extremely rare in the literature and indicates the wide applicability of the used coatings. The value of critical forces (L c1 , L c2 , L c3 ) for the specific damages of coatings on P265GH corresponded with the results obtained for the 316L coated samples. The first cracks (L c1 ) on both types of coatings were recorded under the load of 1.4 N ( Table 3, Figures 6 and 7). The highest value of the loading force at which coating delamination was observed was recorded for SiO 2 /GPTMS/ZrO 2 ( Figure 6B3, Figure 7B3). The delamination of this coating occurred with the load force value twice as high as SiO 2 /ZrO 2 . These results demonstrate that the first cracks for the analyzed coatings were detected at the similar load force value, however, their specifications were different: of sol-gel coatings and laser-oxidized coatings were completely different. The laser-oxidized coatings resulted from the oxidation of the substrate surface, which meant strong physicochemical interaction between the obtained coating and the substrate. In addition, this mechanism made the obtained materials very adhesive [53]. The oxide coatings originating from the sol-gel synthesis were produced by different mechanisms, which resulted in the predominance of electrostatic and weak hydrogen interactions in the first step (and covalent ones after heat treatment) at the coating-substrate level [15]. Thus, the sol-gel method allows one to obtain coatings characterized by high adhesiveness to metal substrates.    of sol-gel coatings and laser-oxidized coatings were completely different. The laser-oxidized coatings resulted from the oxidation of the substrate surface, which meant strong physicochemical interaction between the obtained coating and the substrate. In addition, this mechanism made the obtained materials very adhesive [53]. The oxide coatings originating from the sol-gel synthesis were produced by different mechanisms, which resulted in the predominance of electrostatic and weak hydrogen interactions in the first step (and covalent ones after heat treatment) at the coating-substrate level [15]. Thus, the sol-gel method allows one to obtain coatings characterized by high adhesiveness to metal substrates.     Interestingly, the obtained sol-gel SiO 2 /GPTMS/ZrO 2 coatings showed adhesion similar to gradient-oxide coatings obtained by laser radiation-the layer in the focal plane ("yellow_d" and "orange_d") [53]. It should be noted that the mechanisms of the synthesis of sol-gel coatings and laser-oxidized coatings were completely different. The laser-oxidized coatings resulted from the oxidation of the substrate surface, which meant strong physicochemical interaction between the obtained coating and the substrate. In addition, this mechanism made the obtained materials very adhesive [53]. The oxide coatings originating from the sol-gel synthesis were produced by different mechanisms, which resulted in the predominance of electrostatic and weak hydrogen interactions in the first step (and covalent ones after heat treatment) at the coating-substrate level [15]. Thus, the sol-gel method allows one to obtain coatings characterized by high adhesiveness to metal substrates.

Nanointendation Test
The Young modulus (E IT ) and hardness (H IT ) were calculated for the maximum load of 0.1 mN with the Oliver-Pharr protocol; the results are presented in Table 4. The loading and the unloading curves for the 316L steel with coatings are shown in Figure 8. The analysis of the obtained results indicates that the mechanical properties of SiO 2 /ZrO 2 coating are similar to the surface of the uncoated 316L steel. The data obtained for the coating based on zirconia showed that its penetration depth was about six times lower (with the same force), which meant that the hardness of this coating was also significantly higher than in the case of SiO 2 /GPTMS/ZrO 2 . For the coating with zirconia, the mean Vickers hardness was about 236 ± 14 HV for 316L and about 236 ± 22 HV and 41 ± 12 HVSiO 2 /GPTMS/ZrO 2 (according to the Oliver-Pharr procedure [54]). It was expected that the obtained E and H would result from the structure of sol-gel materials as, in general, the SiO 2 /ZrO 2 consist of an inorganic network with CH 3 moieties, and here, the inorganic part is dominant. Elasticity depends more on the organic part of the network. In the case of hardness, the organic part gains the advantage of being more influential. Both statements are supported by the results obtained for the SiO 2 /GPTMS/ZrO 2 coating, where, during the growth of the organic part of networks, it was observed that the value of hardness decreased. In other words, the behavior observed for coating materials in the plastic and elastic region differed significantly from the behavior of the substrate without a coating. To sum up, in the case of oxide thin films on metallic substrates, H and E were influenced by the amount of organic part moieties in the network structure [55]. It may also be concluded that the crystal phase increased in the inorganic part of the sol-gel, observed in particular for SiO 2 /ZrO 2 , which had impact on the H value. The results showed that the addition of zirconia led to an increase in hardness [56]. The course of the nanoindentation curves ( Figure 8) indicated that all samples had elastic-plastic features.

Nanointendation Test
The Young modulus (EIT) and hardness (HIT) were calculated for the maximum load of 0.1 mN with the Oliver-Pharr protocol; the results are presented in Table 4. The loading and the unloading curves for the 316L steel with coatings are shown in Figure 8.  The analysis of the obtained results indicates that the mechanical properties of SiO2/ZrO2 coating are similar to the surface of the uncoated 316L steel. The data obtained for the coating based on zirconia showed that its penetration depth was about six times lower (with the same force), which meant that the hardness of this coating was also significantly higher than in the case of SiO2/GPTMS/ZrO2. For the coating with zirconia, the mean Vickers hardness was about 236 ± 14 HV for 316L and about 236 ± 22 HV and 41 ± 12 HVSiO2/GPTMS/ZrO2 (according to the Oliver-Pharr procedure [54]). It was expected that the obtained E and H would result from the structure of sol-gel materials as, in general, the SiO2/ZrO2 consist of an inorganic network with CH3 moieties, and here, the inorganic part is dominant. Elasticity depends more on the organic part of the network. In the case of hardness, the organic part gains the advantage of being more influential. Both statements are supported by the results obtained for the SiO2/GPTMS/ZrO2 coating, where, during the growth of the organic part of networks, it was observed that the value of hardness decreased. In other words, the behavior observed for coating materials in the plastic and elastic region differed significantly from the behavior of the substrate without a coating. To sum up, in the case of oxide thin films on metallic substrates, H and E were influenced by the amount of organic part moieties in the network structure [55]. It may

Electrochemical Test
The following results were obtained from the measurements in 3% NaCl: open circuit potential, polarization resistance (Rp), and polarization curves (Tafel plots). The density corrosion current was calculated based on the polarization resistance measurement using the Equation (1): where: R p -resistance polarization, I corr -corrosion current densities, b c /b a -cathodic and anodic Tafel's parameters (established that b c /b a = 0.120 V/dec). The polarization resistance measurements were also used to E corr . Figures 11B, 12B, 15B and 16B show polarization curves-Tafel plots. However, due to the fact that Tafel's method was used to calculate the I corr for pure metals and that the 316L steel was characterized by complex composition (additions such as Mo, Ni, Cr), which generated errors during the calculation of the characteristic R p , I corr, and the determination of E corr , these values were calculated using polarization resistance measurements. In addition, the Tafel plots could be distorted by concentration polarization, resistance drop effects, passivation phenomena, and complex redox reactions observed on the alloy steel surface. However, the Tafel plots also presented results such as polarization resistance in the widest range of potential.
The tests were performed for the uncovered P265GH steel and 316L steel and for both steels with deposited coatings: SiO 2 /ZrO 2 and SiO 2 /GPTMS/ZrO 2 . Figures 9 and 10 present the open circuit potential for coated and uncoated P265GH after the first contact and after 24 h contact with a 3% NaCl solution, respectively. Figures 11 and 12 show: (A) the obtained polarization resistance and (B) the polarization curves-Tafel plots-for the uncovered P265GH samples and the samples with oxide layers. After the first contact with a 3% NaCl solution, the shapes of the polarization curves for both the uncoated P256GH steel and the samples coated with oxide layers were very similar ( Figure 11). However, it can be clearly seen that the lower densities of anode currents were obtained for the carbon steel coated with SiO 2 /GPTMS/ZrO 2 layers. The analysis of the obtained electrochemical results, presented in Table 5, indicated that the modification of the P256GH steel substrate with SiO 2 /GPTMS/ZrO 2 mixed coatings treated at 120 • C increased resistance to corrosion (R p ) from 0.889 kΩ·cm 2 to 7.273 kΩ·cm 2 and decreased the densities of corrosive currents (i corr ), which were in the range of 29 µA/cm 2 to 3.6 µA/cm 2 . To confirm the barrier properties of the obtained layers, Figure 12 presents polarization curves after 24 h exposure in a 3% NaCl solution. This longer exposure confirmed the increased corrosion resistance of samples with SiO 2 /GPTMS/ZrO 2 layers. Unfortunately, for SiO 2 /ZrO 2 , the densities of corrosive currents increased after 24 h of duration in the solution. Table 5 presents the numerical results after 24 h. The results obtained for the samples with SiO 2 /ZrO 2 layers demonstrated that this type of coating material did not show long term corrosion protection. The results for both types of coatings after 24 h in a corrosive medium showed values similar to the R p and i corr of the P265GH steel (approximately 0.6 kΩ·cm 2 ). The highest R p (1878 kΩ·cm 2 ) after 24 h of exposure in 3% NaCl was obtained for the sample with SiO 2 /GPTMS/ZrO 2 coating. The observations of the uncoated steel before stabilization and after stabilization at 250 • C confirmed that the heat treatment affected the reduction of corrosive processes after long term exposure in 3% NaCl. The presented studies showed that P265GH steel is not a corrosion-resistant material in 3% NaCl. The highest R p parameters were obtained for samples with SiO 2 /GPTMS/ZrO 2 layers. However, the application of thin oxide films increased the properties of its anticorrosion behavior about fourfold (based on R p, where anticorrosion properties for uncoated P265GH steel treated at 250 • C was 1) after 24 h exposure to 3% NaCl solutions. that P265GH steel is not a corrosion-resistant material in 3% NaCl. The highest Rp parameters were obtained for samples with SiO2/GPTMS/ZrO2 layers. However, the application of thin oxide films increased the properties of its anticorrosion behavior about fourfold (based on Rp, where anticorrosion properties for uncoated P265GH steel treated at 250 0 C was 1) after 24 h exposure to 3% NaCl solutions. Figure 9. Open circuit potential for P265GH steel without coating and with oxide layers after duration in 3% NaCl 0 h. Figure 10. Open circuit potential for P265GH steel without coating and with oxide layers after duration in 3% NaCl 24 h. Figure 9. Open circuit potential for P265GH steel without coating and with oxide layers after duration in 3% NaCl 0 h. that P265GH steel is not a corrosion-resistant material in 3% NaCl. The highest Rp parameters were obtained for samples with SiO2/GPTMS/ZrO2 layers. However, the application of thin oxide films increased the properties of its anticorrosion behavior about fourfold (based on Rp, where anticorrosion properties for uncoated P265GH steel treated at 250 0 C was 1) after 24 h exposure to 3% NaCl solutions.      In the second part of the experiment, SiO 2 /ZrO 2 and SiO 2 /GPTMS/ZrO 2 coatings applied to 316L steel were tested. Figures 13 and 14 showed the open circuit potential for 316L uncoated and with coatings after the first contact and after 24 h contact with a 3% NaCl solution, respectively. In the polarization resistance ( Figure 15A) and the obtained polarization curves ( Figure 15B), immediately after immersion in 3% NaCl, one could observe changes in their course. The heat treatment of the 316L steel at 250 • C caused the reduction of corrosion currents in the anode part on the Tafel plot. The lower densities of corrosion currents in both the cathodic and the anodic part were observed for SiO 2 /ZrO 2 and mixed SiO 2 /GPTMS/ZrO 2 layers in relation to the uncoated 316L steel. The analysis of the obtained values, presented in Table 6, allowed us to observe that the highest R p was obtained for the steel with SiO 2 /ZrO 2 layers (2.0 MΩ·cm 2 ) immediately after insertion in the solution. Additionally, for the remaining coating, the value was increased in relation to the non-heated 316L steel. The densities of the corrosion currents of the uncoated substrate heated at 250 • C were about six times lower than those of 316L steel. This confirmed the significant influence of heat treatment on the increase in the corrosion properties of austenitic steel and materials with self-passivation features. The presented polarization resistance ( Figure 16A) and polarization curves ( Figure 16B) after 24 h of exposure to a 3% NaCl solution indicated a decrease in corrosion current density for SiO 2 /ZrO 2 and SiO 2 /GPTMS/ZrO 2 samples (in comparison to uncoated 316L). The SiO 2 /GPTMS/ZrO 2 sample was also characterized by a shift towards the positive values of the cathodicanode transition potential (E C-A ) ( Figure 16B), which was confirmed by better protective properties [57]. Table 6 presents the results obtained after 24 h of exposure to a 3% NaCl solution. The results indicated that SiO 2 /ZrO 2 coatings did not show long term corrosion protection. Their polarization resistance decreased after 24 h of exposure in an aggressive environment, while R p for the steel pre-heated at 250 • C after 24 h of exposure decreased about 20 times. The results obtained after a 24 h exposure confirmed the high protection of the SiO 2 /GPTMS/ZrO 2 coatings. The highest polarization resistance (5.9 MΩ·cm 2 ) and the lowest densities of corrosive currents (4.5 nA/cm 2 ) were obtained for these layers. With reference to the uncoated and non-heated steel substrate, the polarizing resistance was about 85 times higher. The obtained results showed the high corrosion resistance of the 316L steel with SiO 2 /GPTMS/ZrO 2 layers. The high polarization resistance and the low densities of the corrosion currents determined indicated the possible use of these layers as long term protection against aggressive environments [57][58][59][60]. A clear improvement in the corrosion resistance for both steels (P256GH and 316L) was observed after applying the proposed sol-gel coatings on them. For both steels, SiO 2 /GPTMS/ZrO 2 layers resulted in increasing barrier properties after long term exposure to 3% NaCl. Further investigation may explain if the long term protection effect is caused by the "self-healing" process described in the literature [15,61] as the SiO 2 /ZrO 2 coating did not show that. The obtained result seemed to be correlated with results from SEM suggesting that the deposition of SiO 2 /ZrO 2 ( Figure 3A1), especially on the P265GH steel, increased the specific surface area, which increased the number of potential interactions between corrosion agents (Cl − ) and the surface, accelerating the damage of the material [35]. In result of long term interaction of metallic substrate with Cl − , the corrosion products may have grown under the coatings. According to results obtained by Croll [62], corrosion products are less dense than the metal and push the coatings away from the surface metal, which can account for the behavior of SiO 2 /ZrO 2 coatings in NaCl solution. may explain if the long term protection effect is caused by the "self-healing" process described in the literature [15,61] as the SiO2/ZrO2 coating did not show that. The obtained result seemed to be correlated with results from SEM suggesting that the deposition of SiO2/ZrO2 ( Figure 3A1), especially on the P265GH steel, increased the specific surface area, which increased the number of potential interactions between corrosion agents (Cl -) and the surface, accelerating the damage of the material [35]. In result of long term interaction of metallic substrate with Cl -, the corrosion products may have grown under the coatings. According to results obtained by Croll [62], corrosion products are less dense than the metal and push the coatings away from the surface metal, which can account for the behavior of SiO2/ZrO2 coatings in NaCl solution. Figure 13. Open circuit potential for 316L steel without coating and with oxide layers after duration in 3% NaCl 0 h.

Figure 14.
Open circuit potential for 316L steel without coating and with oxide layers after duration in 3% NaCl 24 h. may explain if the long term protection effect is caused by the "self-healing" process described in the literature [15,61] as the SiO2/ZrO2 coating did not show that. The obtained result seemed to be correlated with results from SEM suggesting that the deposition of SiO2/ZrO2 ( Figure 3A1), especially on the P265GH steel, increased the specific surface area, which increased the number of potential interactions between corrosion agents (Cl -) and the surface, accelerating the damage of the material [35]. In result of long term interaction of metallic substrate with Cl -, the corrosion products may have grown under the coatings. According to results obtained by Croll [62], corrosion products are less dense than the metal and push the coatings away from the surface metal, which can account for the behavior of SiO2/ZrO2 coatings in NaCl solution. Figure 13. Open circuit potential for 316L steel without coating and with oxide layers after duration in 3% NaCl 0 h.

Figure 14.
Open circuit potential for 316L steel without coating and with oxide layers after duration in 3% NaCl 24 h.

Conclusions
The protective, organically modified oxide coatings were obtained by the sol-gel method. The influence of zirconia and organic additives on mechanical and electrochemical properties of silica coatings was studied.
The microscopy studies showed differences in the structure morphology between coated substrates as well as between different coatings. However, the differences were more due to the substrate than to the coating material. The EDX and the Raman spectroscopy showed rather uniform structures of both coatings. The analysis of Raman spectra showed that the coatings were built of a SiO 2 -ZrO 2 hybrid material with organic groups corresponding to the precursors involved in the synthesis (non-hydrolyzing moieties).
The highest Vickers hardness was observed in the case of the silica-zirconia coatings with methyl moieties (SiO 2 /ZrO 2 )-it was similar to the uncoated 316L steel sampleswhile the SiO 2 /GPTMS/ZrO 2 coating had five times lower hardness compared to the substrate. The cause seemed to be the organic moieties in this coating compared to the -CH 3 moieties from SiO 2 /ZrO 2.
The SiO 2 /GPTMS/ZrO 2 coatings showed the highest adhesion to the metallic substrate at the level of gradient layers (performed using laser oxidation).
Additionally, the hybrid SiO 2 /GPTMS/ZrO 2 coatings exhibited the best properties in terms of corrosion mitigation in seawater. After 24 h in a 3% NaCl solution, the values of polarization resistance were nearly 84 times higher (in comparison to the uncoated 316L steel). This confirmed the positive influence of the organic part on the barrier properties of thin films using corrosion agents (such as Cl -). The results obtained for P265GH may have resulted from the low corrosion resistance of this kind of steel. This, in turn, was connected with a substantial rise in oxide iron scale, increasing during the corrosion process, and was also due to high volume that damaged the coatings.
For the P265GH samples, the nanoindentation results were not presented due to significant scattering of results, which probably correlated with an oxide iron scale arising on the substrate surface under the coating, influencing the inhomogeneities on the surface and enhancing difficulties in contact measurements.
However, it seems that only a combination of zirconia (ZrO 2 ) with silica (SiO 2 ) modified by large organic moieties (GPTMS) gives a larger chance to obtain coating materials with high adhesion to a metallic substrate and with corrosion mitigation ability. However, the influence of the substrate on the degree of degradation in corrosive environments should also be taken into account. The corrosion mitigation was observed even after extending exposure to corrosion agents, which could have resulted from the reinforcement of an organically modified silica network with zirconia.