Introduction

To protect metallic devices and infrastructures from corrosive environments and sudden mechanical damage, such as external impact, epoxy resins, due to their superior chemical, mechanical, and thermal resistance, are extensively used as binders in heavy-duty coatings.1,2 However, thermoset epoxy coatings suffer from inherent brittleness with an insufficient resistance to crack initiation, and as early as during curing, premature cracks can be observed.3,4

Following this, once the driving energy for crack propagation, influenced by local stress and film thickness, surpasses the critical fracture energy (defined as the energy concentrated at the crack tip needed for crack propagation), brittle crack propagation ensues, and the energy needed for crack propagation per unit area is recognized as the fracture toughness.5,6,7,8 During service, the crack formation can severely deteriorate the anticorrosive coating performance and result in failures, such as delamination and de-bonding.9,10,11,12

To improve the fracture toughness of epoxy coatings, a second phase material, of which the most commonly used ones are reactive liquid rubbers, such as hydroxyl-terminated polybutadiene (HTPB), epoxy-terminated polybutadiene (ETPB), carboxyl-terminated butadiene-acrylonitrile (CTBN), amino-terminated butadiene-acrylonitrile (ATBN), or vinyl-terminated butadiene-acrylonitrile (VTBN), can be introduced.13,14,15,16,17 Due to poor compatibility, an insufficient interfacial adhesion between the HTPB rubbers and the resin matrix has been observed. To improve on this, the functionalization of HTPB rubbers with epoxide and carboxyl groups was performed.18 As an example, Yahyaie et al.,19 using synthesized ETPB rubbers, improved the adhesion and mechanical properties for a diglycidyl ether of bisphenol-A (DGEBA) epoxy cured with a cycloaliphatic amine. They found that the toughening mechanism essentially consists of an expanding rubber phase combined with adequate shear deformation, under loading, of the bulk epoxy. In thin epoxy films, due to the energy dissipation induced by the presence of the rubber phase, the coating fracture toughness gets improved. Barcia et al.,18 on the other hand, compared the properties of epoxy resins modified with either isocyanate-functionalized HTPB or a copolymer of polybutadiene and diglycidyl ether of bisphenol-A (DGEBA) epoxy. Due to phase separation, the use of isocyanate-functionalized HTPB resulted in a coating with a milky appearance. However, the simultaneous presence of the copolymer modifier provided an outstanding toughening effect with enhanced impact resistance and resulted in a transparent cured material. Earlier works, reviewed in reference (13), also investigated the effects on epoxy toughening of matrix ductility and rubber particle size distribution and concentration.

Due to high thermal stability and chemical resistance, as well as low surface tension and water repellency, silicones can be combined with epoxies to improve the flexibility.20,21 However, as a result of phase separation, the application of silicones is limited by their deterioration of mechanical and anticorrosive coating properties.22

To achieve better dispersion of the modifiers in epoxy coatings and composites, nanoscale rigid particles with a well-defined size, were also used. Yuan et al.23 modified epoxy coatings for Q235 carbon steel with a sepiolite powder and saw that the toughened coatings exhibited superhydrophobicity and superior mechanical and anticorrosive properties.

Using functionalized graphene oxide, Li et al.24 optimized the long-term anticorrosive properties of epoxy coatings. Their results demonstrate that to improve mechanical properties and UV and thermal stabilities, as well as anticorrosion properties and flexibility, only 0.5 wt% functionalized graphene oxide is needed. Compared to rubber modifiers, however, the toughening effect of rigid nanoparticles is quite small.22

Recently, core-shell rubber (CSR) particles, consisting of a soft rubber core and a rigid shell, have been used to toughen epoxy coatings without compromising mechanical properties, such as modulus and glass transition temperature (Tg), of the coating.22,25 The rubber core, using materials such as styrene-butadiene, polybutadiene, acrylate polyurethane, or silicone rubbers, is expected to provide impact resistance, while the hard shell, for instance poly-methyl methacrylate (PMMA), ensures compatibility with the epoxy resin and enables CSR dispersion. Le et al.,26 to toughen their epoxies, used 5–10 vol% of 55 nm core-shell particles and found that the fracture toughness increased around 200% with only a small lowering of the coating modulus, strength, and Tg.

Since 1980, extensive research has been carried out to toughen the epoxies with improved crack propagation resistance. Various approaches have been explored, including the incorporation of carboxyl-terminated butadiene nitrile liquid rubber,27,28,29,30,31,32,33,34 core-shell particles,35,36 glass beads,37,38,39 hollow latex particles,40 nylon particles,41 as well as the utilization of thermoplastic modifiers, such as polyether sulphone29,36 poly(phenylene oxide),42 and pre-formed polyamide−12 particles.43 Furthermore, these investigations have delved into assessing the influence of temperature, loading rate,44 modifier types, content, size and distribution on the fracture toughness of epoxy. These studies utilized single-/double-edge-notched specimens containing pre-existing cracks and employed testing methods like 3-/4-point bending or symmetrical double-grooved strip tensile tests to evaluate these effects.

It was found that for rubber-modified epoxy, the primary toughening mechanism was shear yielding through cavitation of the rubber particles. In addition, an enhancement in crack propagation resistance was observed accompanied by an increase in shear plasticity via shear band formation at the crack tip.28,44

Despite decades of efforts to improve the crack propagation resistance of epoxy coatings, the underlying mechanisms governing premature crack initiation during the curing process remain inadequately understood. Particularly, a quantification of internal stress and premature crack initiation susceptibility is lacking. In the present work, with the aim of quantifying the premature crack initiation susceptibility and the anticorrosion properties, we investigate the toughening of a solvent-based novolac epoxy/4, 4′-diaminodicyclohexylmethane (PACM) coating. Modifiers used are SrTiO3 nanoparticles, epoxypropoxypropyl-terminated polydimethylsiloxane (PDMS), and two types of CSR particles. In addition, the influence of the CSR modifier concentrations on coating property transients, premature crack initiation susceptibility, and barrier performance, is studied.

Experimental

Materials and coating preparation

A protective D.E.N. 438-X80 epoxy/PACM coating with an amine/epoxy functional group stoichiometric ratio of 1.0 was formulated. The binder was a novolac epoxy resin in xylene, supplied by Dow Chemical Company with an epoxy equivalent weight of 223 g/mol and a solids weight of 80%, used as received. A 4, 4′-diaminodicyclohexylmethane (PACM) supplied by Evonik Corporation with a hydrogen equivalent weight of 54 g/mol was used as hardener. DMS-E11, an epoxypropoxypropyl-terminated polydimethylsiloxane (PDMS) supplied by Gelest, USA, with an epoxy equivalent weight of 500–600 g/mol, was used as siloxane-containing toughening modifier and the weight content of the modifier was controlled to 5 wt%.

Molecular structures of novolac epoxy, epoxypropoxypropyl-terminated polydimethylsiloxane (PDMS), and PACM are shown in Fig. 1.

Fig. 1
figure 1

Molecular structures of novolac epoxy, epoxypropoxypropyl terminated polydimethylsiloxane (PDMS), and 4, 4′-diaminodicyclohexylmethane (PACM)

Full size image

Strontium titanate (SrTiO3, average particle size of 100 nm), supplied by IoLiTec Ionic Liquid Technologies GmbH, Germany, was used as a rigid modifier in a concentration of 5 wt%. The CPVC for SrTiO3 is 0.498 and the reduced PVC is 0.011 (λ = 0.022).

Two types of pre-dispersed core-shell rubber (CSR) particles in epoxy resins, namely MX 217 and MX 267, supplied by Kaneka, Belgium, were used as CSR modifiers (details in Table 1). To vary the CSR concentration in the coatings, MX 217 and MX 267 were mixed into the novolac epoxy.

Table 1 Chemical composition and properties of the pre-dispersed core-shell rubber (CSR) particles
Full size table

Preparation of the coating samples were initiated by mixing the binder, modifier, and hardener. Using an Elcometer 3580 casting film applicator, the wet film thickness (WFT) was controlled to 300 µm. After application, the samples were stored in an Aralab climate chamber for 120 h with a controlled temperature and relative humidity of 23 ± 0.5°C and 35 ± 1 %, respectively.

Characterizations

ATR-FTIR spectroscopy

To track epoxy conversion, the coatings were applied on a 0.1 mm thick plastic substrate (supplied by Erhvervsskolernes Forlag, Denmark). The epoxy conversion as a function of curing time was monitored with a Thermo Scientific Nicolet iS5 ATR-FTIR spectrometer. To avoid the influence of a carbonation reaction at the coating surface, the ATR-FTIR spectra for each sample was measured from the coating-substrate interface on free films, and for all experiments three replicates were used. Details on how to calculate the epoxy group conversion from ATR-FTIR spectra can be found in a previous work.45

Dynamic mechanical thermal analysis (DMTA) measurements

The elastic modulus of coatings at different curing times was tracked using an advanced rheometer DHR-2 (TA Instruments) with rectangular free films. To record the modulus for 5 min at 23°C with a frequency of 1 Hz, an oscillation-time ramp procedure was performed. For all experiments, three replicates were used.

Curing shrinkage measurement

The volumetric shrinkage was calculated from the current value of the coating thickness monitored using an optical 3D profilometer (KEYENCE VR-3000) throughout the curing. For all experiments, three replicates were used and details of the calculation method can be found in reference (45).

Internal stress measurements

The coatings were applied on 316L stainless steel shims (200 mm × 15 mm × 0.1 mm), cleaned beforehand with ethanol, and the curing-induced internal stress monitored with a deflection method. The coated samples were placed horizontally in a climate chamber until gelation and then mounted vertically. Laser position sensors were used to record the deflection at the free end and details on how to calculate the internal stress are described elsewhere.45 Three replicates were used for each test.

Tensile test

Using an advanced rheometer DHR-2 (TA Instruments), the stress-strain curves of the coatings were measured in tensile mode at 23°C and a linear tension rate of 5 µm/s. For all experiments, three replicates were used. Details on how to apply the rectangular free film to rigid substrates are described elsewhere.4

Crack initiation susceptibility quantification and cracks morphology characterizations

The crack initiation susceptibility was calculated with the stress-strain curves of samples obtained using tensile testing (details in previous work4). A digital microscope (KEYENCE VHX) was used to observe the premature crack morphology for embedded and polished samples.

Electrochemical corrosion tests

For evaluation of anticorrosive coating properties, electrochemical impedance spectroscopy (EIS) tests are widely used.46,47 To study the influence of modifiers, coatings with a wet film thickness (WFT) of 300 µm were applied on 0.8 mm thick aluminum panels with dimensions of 150 mm × 15 mm. After curing in a climate chamber for 120 h, the EIS measurements were performed using a Gamry Potentiostat Reference 600+ in a three-electrode cell at room temperature. A specimen with a circular exposure area of 10 cm2, a saturated calomel electrode (SCE), and a graphite rod were used as the working (WE), reference (RE), and counter electrode (CE), respectively. The EIS tests were conducted in 3.5 wt% NaCl electrolyte, keeping the working electrode at the open circuit potential (OCP) with a signal amplitude of 10 mV in the frequency range of 105 to 10−1 Hz. Three replicates were used for each test and the acquired data was analyzed with the software ZSimpWin.

Results and discussion

Effect of modifier type

The effects of four modifiers on the coating property transients during curing, including the epoxy conversion, the elastic modulus, the relative volumetric shrinkage, the internal stress, and the crack initiation susceptibility, are shown in Fig. 2.

Fig. 2
figure 2

Experimental data of coating properties during curing at 23 ± 0.5°C for a novolac epoxy/PACM coating with 5 wt% of different modifiers. The relative volumetric shrinkage and the crack initiation susceptibility is defined as the ratio of the current coating volume to the initial coating volume minus one (Vt/V0 − 1) and the reduction-in-strain factor, respectively. The higher the crack initiation susceptibility of the coating, the higher the possibility that cracks can form in the coating. In all the plots, one standard deviation is provided with error bars

Full size image

During the first 10 h, only a small difference in the epoxy group conversion of the modified and non-modified coatings can be seen. After 24 h, as a result of vitrification, the crosslinking reactions become diffusion-controlled48,49 and only a slight increase in the epoxy group conversion takes place. After 120 h of curing, the highest final conversion was detected in the coating modified with solvent-containing CSR 217 particles, followed by PDMS, SrTiO3, and CSR 267 particles. This can be attributed to the plasticizing effect from the residual solvent.

At the 24 h mark, as a result of the slightly higher epoxy conversion, the elastic modulus of the samples modified with PDMS and CSR 217 is higher than for CSR 267, SrTiO3, and non-modified coating. Unlike the samples modified with solvent-free SrTiO3 and CSR 267, which reached the maximum modulus at the 48 h mark and then exhibited a decrease, a steady increase is seen for PDMS and CSR 217 modified coatings after 48 h, which reach the maximum at the 72 and 96 h mark, respectively.

For all coatings, at the 120 h mark, the final relative volumetric shrinkage decreases with a decreasing final epoxy conversion (compare Fig. 2a, c). Although the final conversion is the highest for the CSR 217 coating, the rubber core in the modifier can release strain and shrinkage, and a relatively small volumetric shrinkage is therefore seen for this particular coating.22

The internal stress in the PDMS modified sample, as a result of an enhanced final conversion, is slightly higher than for the non-modified coating. Also, rigid SrTiO3 and CSR particles, due to the improved elastic modulus, lead to an increased internal stress. As a result of the extra solvent plasticizing effect, the measured internal stress in solvent-containing CSR 217 modified coating is much smaller than that of solvent-free CSR 267 modifiers.

During the first 96 h, there is little difference in the crack initiation susceptibility of non-modified, SrTiO3, and PDMS modified coatings, and a reduced final crack initiation susceptibility is seen in SrTiO3 and PDMS modified samples. However, as a result of a higher internal stress, a higher crack initiation susceptibility is seen for CSR 217 and CSR 267 modified coatings. Introduction of nanoscale SrTiO3 and PDMS is a potential way to reduce the crack initiation susceptibility. However, in PDMS modified epoxy protective coatings, it should be noted that the anticorrosion performance can be severely influenced by the phase separation (Figs. 3c, 4c, and S2).22,50

Fig. 3
figure 3

Morphology of defects and cracks (observed with a digital microscope on a rigid, flat substrate) for a novolac epoxy/PACM coating (a), and with 5 wt% of different modifiers (b–e). The red markings emphasize defects and cracks

Full size image
Fig. 4
figure 4

Morphology of defects and cracks (observed with a digital microscope) for a novolac epoxy/PACM coating (a) with 5 wt% of different modifiers (b–e) on a U-channel substrate (the red markings indicate defects and cracks). (f) Schematic diagram of the morphology characterized with embedding and polishing treatment

Full size image

In Fig. 3, the morphology of defects and cracks in the modified coating applied on a rigid glass substrate and observed using a digital microscope at the coating surface is shown. Compared to the non-modified coating (Fig. 3a), a tiny crack and patterns that can develop into cracks (indicated with red circles) can be found in the SrTiO3 modified coating (Fig. 3b) and for the PDMS modified sample, no visible defects are seen (Fig. 3c). Moreover, compared to morphology at the interface (Fig. S2b in the Supporting Information), because the PDMS copolymer can aggregate into spherical domains and more domains are formed at the interface than at the film surface, less phase separation is detected at the surface area.50 In CSR 217 (Fig. 3d) and CSR 267 (Fig. 3e) modified samples, larger-size cracks and patterns that can develop into cracks (indicated with red circles) are observed. This is consistent with Figs. 2d and 2e.

Compared to coatings applied to flat substrates, cracks and defects are more frequently observed at inner angles. Figure 4, using a digital microscope, shows the crack and defect morphology observed in modified coatings applied at an inner 90-degree angle (Fig. 4f). Many defects and several long cracks can be seen at the coating substrate interface for the non-modified coating (Fig. 4a), while the SrTiO3 modified coating only shows tiny cracks, indicating that the SrTiO3 modified coating is less susceptible to premature crack initiation, which agrees well with Fig. 2e. As expected, phase separation occurs in the PDMS modified coatings (Fig. 4c) and few defects can be observed. In addition, larger-size cracks and defects are seen in CSR 217 (Fig. 4d) and CSR 267 (Fig. 4e) modified coatings, proving the efficiency of the method for crack initiation susceptibility evaluation (Fig. 2e).4 The morphology of cracks and defects observed after a further polishing treatment, to a deeper position, exhibited similar trends and validated the results for the non-modified (Fig. S3a), CSR 217 (Fig. S3b), and CSR 267 (Fig. S3d) modified coatings (details in the Supporting Information).

Effects of CSR concentrations

The effects of CSR 217 concentrations on the transients of coating properties, including the epoxy conversion, the elastic modulus, the relative volumetric shrinkage, the internal stress, and the crack initiation susceptibility, are shown in Fig. 5.

Fig. 5
figure 5

Experimental data of coating properties during curing at 23 ± 0.5°C for a novolac epoxy/PACM coating modified with different concentrations of CSR 217. In all the plots, one standard deviation is provided with error bars

Full size image

During the first 8 h, because the reactants are more diluted in coatings with lower modifier concentrations, the epoxy conversion in 0, 0.5, and 5 wt% CSR 217 modified samples is much smaller than for the 25 wt% sample. After 24 h, because the particles reduce chain mobility,51 a lower final conversion is observed in the 25 wt% CSR 217 modified coating.

After the 24 h mark, the elastic modulus increases with increasing modifier concentrations in the range from 0 to 5 wt%. For the 25 wt% CSR 217 modified coating, it was impossible to prepare the rectangular DMTA samples applied to rigid substrates at 120 h curing and the elastic modulus was taken as the results measured after one-month curing (Fig. S1a), corresponding to the first point in time where the elastic modulus allowed for testing. Therefore, the measured modulus after one month, which is much smaller than for the non-modified coating, is used to calculate the internal stress (Fig. 5d), and the modulus and crack initiation susceptibility results for the 25 wt% CSR 217 modified samples are not shown in Figs. 5b and 5e. Unlike the non-modified and 5 wt% modified samples, which reach the maximum modulus at 72 and 96 h, respectively, the elastic modulus of the 0.5 wt% CSR 217 modified coating keeps increasing with time.

Due to a reduced final epoxy conversion and a higher content of CSR particles, the relative volumetric shrinkage decreases with increasing CSR 217 concentrations. In addition, the measured internal stress increases with increasing modifier concentrations from 0 to 5 wt%. For the 25 wt% CSR 217 modified sample, as a result of a relatively small final conversion, elastic modulus, and volumetric shrinkage, the measured internal stress is much smaller than the non-modified coating.

The crack initiation susceptibility of the non-modified and the 0.5 wt% CSR 217 modified sample does not show much difference and a slightly higher crack initiation susceptibility is seen in the 5 wt% CSR 217 modified coating. In addition, the strain value corresponding to crack initiation is slightly improved (Fig. S1b, c), indicating an improved flexibility in the 25 wt% CSR 217 modified coating.

The morphology of defects and cracks in the coatings containing different CSR 217 modifier concentrations applied to rigid substrates at the coating surface is shown in Fig. 6. Unlike the non-modified coating (Fig. 6a) and the 5 wt% CSR 217 modified sample (Fig. 6c), in which many defects and larger-size cracks can be found, no visible defects and cracks can be observed in the 0.5 wt% (Fig. 6b) and 25 wt% CSR 217 modified sample (Fig. 6d). However, patterns that can develop into defects and cracks, indicated with red circles, are observed.

Fig. 6
figure 6

Morphology of defects and cracks for a novolac epoxy/PACM coating (a) with different CSR 217 modifier concentrations (b–d) on a rigid flat substrate. Red arrows indicate defects and cracks

Full size image

Figure 7 shows the crack and defect morphology observed in coatings containing different concentrations of CSR 217 modifiers applied inside an inner 90-degree angle (Fig. 4f), where cracks are more frequently seen. Compared to the non-modified coating (Fig. 7a) and the 5 wt% CSR 217 modified sample (Fig. 7c), less defects and smaller-size cracks are seen in the 0.5 wt% CSR 217 modified sample (Fig. 7b). When the CSR 217 concentration increases to 25 wt% (Fig. 7d), no visible defects and cracks can be observed and this is consistent with the measured internal stress in Figs. 5d and 5e. The observed morphology of cracks and defects after a further polishing, to a deeper position, showed similar trends and validated the results for the non-modified (Fig. S3a), 5 (Fig. S3b), and 25 wt% CSR 217 (Fig. S3c) modified coatings (details in the Supporting Information).

Fig. 7
figure 7

Morphology of defects and cracks (observed with a digital microscope) for a novolac epoxy/PACM coating (a) with different CSR 217 modifier concentrations (b–d) on a U-channel substrate. Red arrows indicate defects and cracks

Full size image

The effects of CSR 267 concentrations on the transients of coating properties are shown in Fig. 8.

Fig. 8
figure 8

Experimental data of coating properties during curing at 23 ± 0.5°C for a novolac epoxy/PACM coating modified with different concentrations of CSR 267. In all the plots, one standard deviation is provided with error bars

Full size image

The epoxy conversion increases with increasing CSR 267 concentrations during the first 8 h. After 24 h, due to a combined effect of vitrification and reactant mobility reduction induced by CSR particles,51 a minimum conversion is seen in the 15 wt% CSR 267 modified coating.

Due to the same problem as described for the 25 wt% CSR 217 modified coating, the measured modulus of the 37 wt% CSR 267 modified coating after one-month curing (Fig. S1a) is used to calculate the internal stress (Fig. 8d). In addition, the modulus and crack initiation susceptibility results for the 37 wt% CSR 267 modified samples are not shown in Figs. 8b and 8e. Unlike the non-modified and 0.5 wt% modified samples, which reach the maximum modulus at the 72 h mark, the elastic modulus of the 2.5, 5, and 10 wt% CSR 267 modified coating reaches the maximum earlier, at 48 h, and the 15 wt% CSR 267 modified coating keeps increasing.

The relative volumetric shrinkage decreases with increasing CSR 267 concentrations and the measured internal stress increases with increasing modifier concentrations and reaches the peak at 5 wt% and then decreases with increasing modifier concentrations. Compared to the non-modified coating, a slightly higher crack initiation susceptibility can be seen in CSR 267 modified samples with the concentration ranging from 0.5 to 10 wt% and a much lower crack initiation susceptibility is observed in the 15 wt% CSR 267 modified coating. In addition, the strain values corresponding to crack initiation of 15 and 37 wt% CSR 267 modified coating are much larger than the non-modified coating (Fig. S1b, c), suggesting an enhanced premature crack initiation resistance.

The morphology of defects and cracks in the coatings containing different CSR 267 modifier concentrations applied to rigid substrate at the coating surface is shown in Fig. 9. Unlike the non-modified coating (Fig. 9a) and the 5 wt% CSR 267 modified sample (Fig. 9d), in which many defects and larger-size cracks can be found, only tiny cracks can be observed in the 0.5 (Fig. 9b), 2.5 (Fig. 9c), and the 10 wt% CSR 267 modified sample (Fig. 9e). The purple arrows here indicate the scratches produced during the polishing treatment. As CSR 267 concentration increases to 15 (Fig. 9f) and 37 wt% (Fig. 9g), no visible defects and cracks can be observed and only patterns (indicated with red circles) that can develop into defects and cracks are seen.

Fig. 9
figure 9

Morphology of defects and cracks (observed with a digital microscope) for a novolac epoxy/PACM coating (a) with different CSR 267 modifier concentrations (b–g) on a rigid substrate. The red arrows indicate defects and cracks, and the purple markings indicate scratches originating from the polishing process

Full size image

Figure 10 shows the crack and defect morphology observed in coatings containing different concentrations of CSR 267 modifiers applied inside an inner 90-degree angle (Fig. 4f). Compared to the non-modified (Fig. 10a) and the 5 wt% CSR 267 modified sample (Fig. 10d), less defects and smaller-size cracks are seen in the 0.5, 2.5, and 10 wt% CSR 267 modified samples (Figs. 10b–10e). When the CSR 267 concentration increases to 15 and 37 wt% (Figs. 10f and 10g), no visible defects and cracks can be observed and this is consistent with the measured internal stress in Figs. 8d and 8e. The observed morphology of cracks and defects after a further polishing treatment, to a deeper position, showed similar trends and validated the results for the non-modified (Fig. S3a), 5 (Fig. S3d), and 37 wt% CSR 267 (Fig. S3e) modified coatings (details in the Supporting Information).

Fig. 10
figure 10

Morphology of defects and cracks (observed with a digital microscope) for a novolac epoxy/PACM coating (a) with different CSR 267 modifier concentrations (b–g) on a U-channel substrate and the red markings indicate the defects and cracks

Full size image

Effects of modifiers on coating barrier properties

Figure 11 shows the EIS results and fitting curves of coatings with different modifiers in 3.5 wt% NaCl solution at room temperature, including the Nyquist plot (a–c), the Bode plot of impedance (d–f), the phase angle (g–i), and the corresponding equivalent circuit illustrated in Fig. 11j.

Fig. 11
figure 11figure 11

The Nyquist plots (a–c), the Bode plots of impedance (d–f), and the phase angle (g–i) for a novolac epoxy/PACM coating with different modifiers on an aluminum substrate, as well as the corresponding equivalent circuit (j)

Full size image

The Bode plot of impedance (Figs. 11d–11f) and the phase angle shown in Figs. 11g–11i suggest a single time constant, implying that no electrochemical reactions occur at the coating and the aluminum substrate interface.52,53 Therefore, the equivalent circuit shown in Fig. 11j was used to fit the EIS results, in which RS represents the electrolyte resistance and RC denotes the coating resistance. Our previous work4 shows that premature cracks are formed in curing coatings. Consequently, the electrolyte can penetrate the coating through defects and cracks. As a result of the non-uniform electrolyte uptake in the sample, a constant-phase element (CPE) is used instead of a pure capacitance. A good fitting of results can be found in the measured experimental results and the fitting curves (Figs. 11a–11i), indicating a sufficient efficiency of the equivalent circuit and the fitted CPE capacitance and coating resistance results are summarized in Fig. 12. When evaluating barrier performance of coatings, the CPE capacitance and the coating resistance are considered the key parameters. A reduced CPE capacitance signifies an enhanced corrosion resistance because it is more difficult for the electrolyte to penetrate the coating. Conversely, an increased coating resistance indicates an improved barrier performance.

Fig. 12
figure 12

The fitted constant-phase element (CPE) capacitance (a) and the coating resistance RC (b) for EIS results of a novolac epoxy/PACM coating with different modifiers on an aluminum substrate

Full size image

Compared to a non-modified coating, a reduced CPE capacitance can be observed in 0.5, 25 wt% CSR 217, 0.5, 10, 15, and 37 wt% CSR 267 modified coatings (indicated by the blue dashed line in Fig. 12a), suggesting an enhanced anticorrosion performance due to smaller internal stress and reduced premature crack initiation susceptibility. Though few defects and premature cracks are observed in the PDMS-modified coating, the coating barrier performance, due to phase separation induced by copolymer aggregation, is severely deteriorated and a higher CPE capacitance is seen. The increase in CPE capacitance in the coatings modified with 5 wt% SrTiO3, 5 wt% CSR 217, and 2.5 and 5 wt% CSR 267 can be attributed to the increased curing-induced internal stress and elevated premature crack initiation susceptibility. In addition, an increase in coating resistance can be observed in 25 wt% CSR 217 and 37 wt% CSR 267 modified coatings (indicated by the blue dashed line in Fig. 12b), providing additional substantiation for the improvement in the barrier performance.

On the other hand, the anticorrosion properties of the coating increase with increasing concentrations of CSR 217 from 0.5 to 25 wt% (indicated with the red arrows in Figs. 12a and 12b). For CSR 267 modified samples, the barrier properties exhibit a decrease from 0.5 to 5 wt% (indicated with the green arrows in Figs. 12a and 12b), followed by an increase from 5 to 37 wt% (indicated with the blue arrows in Figs. 12a and 12b).

Conclusions

A novolac epoxy coating was modified with the addition of epoxide-terminated PDMS, SrTiO3 nanoparticles, or core-shell rubber (CSR) nano-particles. The coating properties, internal stress, and premature crack initiation susceptibility were investigated. Using an electrochemical impedance spectroscopy (EIS) measurement, the anticorrosion properties of the modified coatings were also quantified.

Digital microscopy images of the defect and crack morphology show that although the epoxide-terminated PDMS and the SrTiO3 nanoparticle modifiers can provide a slightly lower premature crack initiation susceptibility, the EIS results suggest that the coating barrier performance is severely deteriorated.

Coatings modified with 200 nm diameter MX 217 and MX 267 CSR particles, on the other hand, have a great influence on the curing-induced internal stress, the premature crack initiation susceptibility, and anticorrosion properties. The commercial 25 wt% MX 217 and 37 wt% MX 267 CSR particles, despite a reduced elastic modulus, can reduce the curing-induced internal stress and provide strong corrosion resistance.

Future research into epoxide-terminated PDMS modifiers to optimize protective epoxy coatings, particularly for enhancement of the anticorrosive barrier performance, should focus on establishing a dispersing procedure that can avoid and eliminate phase separation. In addition, to minimize the detrimental effects on mechanical properties, further work is needed to improve the modulus of core-shell rubber (CSR) modified coatings.