Sol Gel Hybrid Coatings For Corrosion Inhibition Engineering Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Inorganic-organic hybrid sol-gel coatings have received increasing interest as potential replacements for chromates as corrosion inhibiting primers. The resulting polysiloxane layer formed by the sol-gel process provides excellent adhesion to the substrate through covalent bonding. The organic functional groups in the hybrid coatings allow covalent bonding to subsequent layers providing improved adhesion between the topcoat and primer. For hybrid sol-gel coatings, the siloxane precursors can be allowed to react beforehand forming colloidal particles with organic functional groups on the surface. These functional groups are then cross-linked with an organic cross-linking agent. For this study, colloidal particles with epoxide functional groups were formed from 3-glycidoxypropyltrimethoxysilane and tetramethoxysilane and cross-linked with aliphatic amines and the three isomers of phenylenediamine. The resulting cross-linking with phenylenediamine likely took the form of electroactive oligomers of poly(aniline) at various degrees of crosslink density. Subsequent measurements with electrochemical impedance spectroscopy (EIS) along with physical measurements and constant immersion testing showed improved corrosion performance over that of comparable coatings made with aliphatic amine cross-linking agents. We attribute this improvement to the in-situ oxidative polymerization of the phenylenediamine cross-linking agents acting as an oxygen scavenging mechanism, thus preventing the occurrence of the oxygen reduction reaction at the substrate surface, as well as increasing the crosslink density over time. Small but important levels of conductivity may also be present. Further studies of the unique properties of this coating are ongoing.

Keywords: sol-gel, phenylenediamine, electrochemical impedance spectroscopy, epoxy-amine


Organic-inorganic sol-gel hybrid coatings are seeing increased interest as potential replacements for chromium (VI) based pretreatments and primers.1-13 The coatings are covalently bonded to the underlying metal, providing excellent adhesion.3 Many can be prepared in water-only systems yielding low molecular weight alcohols as the only byproducts. Organic functional groups incorporated into the coatings provide covalent bonding to functional groups in subsequent layers yielding a complete coating system covalently bonded to the substrate.6-8 The hydrophobic nature of many of the organic cross-linking agents and tight cross-linking of the metal-oxide sol precursors yield excellent barrier properties to prevent water ingress. Corrosion inhibiting agents can be added directly to the coatings or encapsulated in responsive microcapsules and incorporated into the coatings.9-11 Inhibiting agents loaded into microcapsules are released as the microcapsules respond to localized increases in pH due to the earliest corrosion when water does finally reach the substrate, thereby passivating the metal and preventing further corrosion while those freely added travel with incoming water.

For this work, siloxane hybrid organic-inorganic coatings were formulated based on tetramethoxysilane (TMOS) and glycidoxypropyltrimethoxysilane (GPTMS). These silanes were first reacted to form inorganic silica colloidal particles with the epoxide moiety of GPTMS then cross-linked by the familiar epoxy-amine reaction. Previously, aliphatic amine cross-linking agents were utilized with analysis by electrochemical impedance spectroscopy (EIS).12-13 This work was extended with two additional aliphatic amine cross-linking agents, tris(2-aminoethyl)amine and triethylenetetramine. The current work also employed the isomers of the aromatic phenylenediamine as the amine cross-linking agents, namely, o-phenylenediamine, m-phenylenediamine, and p-phenylenediamine. The electrochemical properties were examined of these PDA cross-linked coatings which were quite different from those obtained when aliphatic cross-linking agents were used. These coatings also exhibited much better performance as anti-corrosion coatings than those cross-linked by aliphatic amines.


Materials. Aluminum alloy 2024T3 substrates were from McMaster-Carr and Q-Panel. Tris(2-aminoethyl)amine (TAEA), triethylenetetramine (TETA), o-phenylenediamine (o-PDA), and m-phenylenediamine (m-PDA) were from Aldrich, p-phenylenediamine (p-PDA) was from Fluka. Sol-gel precursors tetramethoxysilane (TMOS) and glycidoxypropyltrimethoxysilane (GPTMS) were from Aldrich and Alfa Aesar. Diethylenetriamine (DETA) was from Alfa Aesar. Novec™ Fluorosurfactant FC 4432 was from 3M. Acetic acid was from Fisher Scientific, sodium chloride (NaCl) from VWR, and ammonium sulfate ((NH4)2SO4) from ICN. All water was 18.2 MΩ (Millipore).

Substrate preparation. The aluminum slides were immersed in acetone or methanol for degreasing, rinsed with water, immersed in de-oxidizing solution, rinsed with water and allowed to dry in air. Substrates were surface treated by immersion in a dilute solution of GPTMS in 0.05 M acetic acid (1:150 v/v) for 20 minutes and allowed to dry in air as previously described.14

Coating preparation. TMOS (0.02 mol) and GPTMS (0.06 mol) were combined to form a 1:3 molar ratio mixture of silane monomers. The silane mixture was then added drop-wise to 22 mL of acidified water (0.05 M acetic acid) with stirring, yielding a solution of fifteen moles of water to one mole of silane monomer. This mixture was covered and aged with continuous stirring for 72 hours to form colloidal silica particles. Following aging, surfactant solution (9.4 mL, FC 4432, 0.1% w/w) and phenylenediamine cross-linking agents (0.015 mol in 10 mL of methanol/water 1:1 v/v), TAEA, TETA, or DETA (1:1 GPTMS:amino hydrogen plus enough water for equal volumes) were added to the silane mixture. Cleaned, surface treated substrates were dip-coated in the resulting mixture. Each substrate was immersed three times and allowed to dry in air. The final dry coating thickness is ca. 1.0µm.

Characterization techniques. Infrared attenuated total reflectance spectroscopy was performed on PDA cross-linked sol-gel coated aluminum alloy substrates. Measurements were obtained with a Nicolet 380 FT-IR spectrometer on a single bounce ATR accessory (Pike Technologies, MIRacle ATR) equipped with a ZnSe crystal and a standard DTGS room temperature detector. 150 scans were used to obtain desirable signal to noise ratios for the sol-gel samples, and 32 scans for precursor materials. GC-MS was used in conjunction with NMR spectroscopy to identify by-products associated with the PDA cross-linking agents. Samples of oxidized PDA were first separated using liquid column chromatography, with methanol as the solvent and silica gel as the stationary phase. Aliquots of each band isolated using these techniques were then analysed using an HP 5890 Series II Gas Chromatograph coupled with an HP 5971A Mass Selective-Detector. UV-Vis spectroscopy was used to measure the changing peak absorbance of the oxidizing PDA solutions over time. Experiments were performed on an Ocean Optics USB 2000 Spectrometer, in a 1.4 mL, 1 cm path length quartz cuvette.

Coating performance. To assess the corrosion inhibition abilities of the coatings, aluminum alloy substrates cross-linked with PDA isomers, as well as a fourth set with no cross-linking agent, were immersed in dilute Harrison's solution. Four taped substrates were coated for each type of gel: three were scribed down to bare metal over a 2.5 cm length to evaluate the coating performance at a site of intentional mechanical damage. The fourth was used as a control to measure overall corrosion protection. Samples were inspected visually; good corrosion inhibition was judged by a lack of aluminum oxide ("white rust"), as well as blistering, delamination, or spotting of the coating.

Electrochemical Impedance Spectroscopy (EIS). The impedance measurements were carried out in dilute Harrison's solution (0.05% NaCl and 0.35% (NH4)2SO4 in 18 MOhm water). The electrochemical cell that was used for the EIS experiments consisted of a poly(vinylchloride) pipe with an inner diameter of 2.6 cm attached to the substrate using Marine Goop® adhesive. Three cells, providing replicate results, were obtained for each cross-linking agent. A three-electrode configuration was used with the aluminum panel as the working electrode, platinum coated mesh as counter electrode and saturated calomel electrode (SCE) as the reference electrode. Gamry Instrument PC-4, FAS1, or Reference 600 potentiostats were used to conduct the EIS experiments. A perturbing voltage of 10 mV (rms) was applied to the samples. The frequency range studied was from 100,000 Hz to 0.01 Hz. The open-circuit potential (OCP) was used as the bias potential. The measurements started after 100s of OCP monitoring. A 30 minute delay was employed before the start of each consecutive test and continued for approximately 24 hours. The measurements were then taken once per day for several days and thereafter the number of days to the next measurement gradually increased.


The differences seen between the three aliphatic cross-linking agents in the previous work12 were compelling enough to extend that work to include cross-linkers with additional functional amines to further attempt to increase cross-link density. TETA was chosen so an extra functional amine (6 total) was introduced in a linear chain to see if a comparable increase in performance was obtained such as when DETA (5 total) was previously compared to ethylenediamine (EDA)(4 total). TAEA (also 6 functional amines) was chosen because it has a "built-in" crosslink point. Structures of all four cross-linkers are shown in Figure 1. Figure 2 compares EIS results obtained for DETA, TETA, and TAEA cross-linked samples. TETA showed the worst performance of the three. It appears the added flexibility of the longer chains more than offsets any gains imparted by the additional crosslink point. This also coincides with a difference seen in contact angles of 31.8 (±3.2) degrees for TETA and 52.1 (±2.7) for TAEA. The TAEA cross-linked samples performed the best of the three indicating the "built-in" crosslink does indeed improve the crosslink density and coating performance. These results, thus, indicate there is a law of diminishing returns attempting to increase crosslink density through the number of epoxy-amine reactions with this particular sol-gel coating system.

Figure 1. Structures of aliphatic cross-linking agents.

Figure 2. Average impedance values for three samples of each aliphatic cross-linked sol-gel coating on AA2024T3.

The three isomers of phenylenediamine also used as cross-linking agents are shown in Figure 3. Infrared spectroscopy (data not shown) of the cross-linked gels displayed peaks consistent with formation of the silica backbone and epoxy-amine cross-linking previously reported.13-14 In particular Si-O-Si stretches centered around 1000 cm-1 were visible, indicative of the siloxane network formation. Evidence of the epoxide ring opening cross-linking reaction can be found in the disappearance of peaks associated with the epoxide ring. For the GPTMS precursor an epoxide ring peak appears at 3045 cm-1 (epoxide CH2 antisymmetric stretch); this is the most useful band since it absorbs quite strongly and is not obscured by other peaks in the spectra. This peak completely disappeared after the gels were exposed to the PDA cross-linkers and the coatings solidified. The strong scissoring mode of primary amine attached to aromatic rings typically found at 1638-1602 cm-1 was of low intensity or missing altogether indicating conversion of the primary amines to more extensively substituted amines.

Figure 3. The three isomers of phenylenediamine used as cross-linking agents: A) p-PDA; B) m-PDA; C) o-PDA.

During exposure to DHS, all samples cross-linked with PDA isomers showed coloration that was not seen with aliphatic cross-linked samples. The p-PDA cross-linked coatings demonstrated the most significant coloration, changing from a colorless coating to a clear purple. m-PDA crosslinked samples also underwent substantial coloration changing to a green-brown color whereas o-PDA cross-linked samples changed somewhat less significantly to a pale blue color. The sample not cross-linked with PDA, like the aliphatics, was colorless, clearly implicating PDA. All three isomers of PDA are susceptible to oxidation, forming nitroaniline. These products were considered as possible candidates for the cause of coloration within the cross-linked gels. In order to investigate the formation of nitroaniline in these systems, UV-Vis spectra of solutions of PDA in methanol were acquired at the time of dissolution and again 24 hours later after standing in air. The resulting spectra are shown in Figure 4. The peaks at 433 nm and 450 nm for o- and p-PDA, respectively, are consistent with the formation of yellow and yellow-orange nitroaniline products.15 However, only very limited color change was observed for m-PDA. This is not consistent with the relative coloration of the PDA cross-linked sol-gels where p-PDA and m-PDA were the most highly colored with o-PDA showing only limited coloration. The oxidation product of p-PDA was further characterized by FTIR, GC-MS and 1H-NMR. Infrared spectroscopy revealed some new minor absorption bands in partially oxidized solid samples of p-PDA. These bands appeared at 1547 cm-1, 1326 cm-1, 1126 cm-1, and 857 cm-1, consistent with NO2 antisymmetric stretching, NO2 symmetric stretching, NO2 C-N stretching, and NO2 scissoring modes, respectively. A strong band at 1603 cm-1 in

Figure 4: UV-Visible spectra in aqueous solution of (a) o-PDA initially and after 24 hours, (b) m-PDA initially and after 24 hours, and (c) P-PDA initially and after 24 hours.

the spectrum of the oxidized solid indicated a di-substituted benzene ring with two different substituents. These results indicate that some of the amines are oxidized to nitro groups. This is confirmed by GC-MS and 1H-NMR (results not shown). Overall, these results reliably identify the monomeric oxidation product of p-PDA as p-nitroaniline. In summary, the oxidative formation of nitroaniline species by the isomers of PDA may occur to a limited degree within the cross-linked sol-gels under corrosive conditions. However, their limited production, their uniformly yellow color and the mismatch between the observed degree of coloration of the sol-gels and the relative reactivity of isomers of PDA towards nitroaniline formation clearly indicate that they are not the species responsible for the observed color change of the PDA cross-linked sol-gels. We attribute the coloration to the oxidative polymerization of the PDA isomers. All three isomers of PDA have been polymerized by mild oxidation; the colors seen in these samples are consistent with the colors reported for the products of oxidative polymerization of the isomers of PDA.16-18

Samples completely immersed in DHS were stable and protected the aluminum alloy substrates for extended periods of time. After twenty-two weeks of immersion, gels cross-linked with p- and o-PDA exhibited the best pinhole corrosion protection and scribe protection, respectively. This performance is better than that previously observed for aliphatic cross-linking agents.12 o-PDA cross-linked gels have the most significant pinhole corrosion; tiny flaws in the coating are attacked, however, little oxidized aluminum is present in these areas. Gels cross-linked with o-PDA do seem to protect the scribe quite well; very little oxidized aluminum or other signs of corrosion are present in the scribe after 8 months of continuous immersion. p-PDA cross-linked gels exhibit good scribe protection, and the fewest pinhole corrosion spots. However, where pinhole corrosion does occur, it is more severe than with the other cross-linking agents. m-PDA cross-linked gels exhibit reasonable pinhole protection, however, a considerable amount of oxidized aluminum is noted in the scribes, indicating poorer protection of exposed metal sections. An example is shown in Figure 5 comparing p-PDA cross-linked coating with a coating composed of uncross-linked coating, a sample in which any cross-linking present is provided by Si-O-Si bonds.


Figure 5. Aluminum alloy slides before and after 8 months of dilute Harrison's exposure (a) with p-PDA-cross-linked and (b) not PDA cross-linked.

EIS results for PDA cross-linked coatings yielded much different behavior than that seen for any of the aliphatic amine cross-linked samples; these results also indicate much better overall performance than those cross-linked with aliphatic amines as the duration of the immersion time was for a much longer period. All PDA cross-linked samples showed similar behavior regardless of isomer; a representative graph is shown in Figure 6. The samples show capacitive character at high frequencies, turning to resistive behavior at the middle frequencies, and then become capacitive again at low frequencies as demonstrated by the phase angle plots. The flat resistive region shifts to higher frequencies and lower impedance values with greater immersion time. Figure 7 shows the average impedance values obtained for the triplicate samples

Figure 6. EIS results for sol-gel sample cross-linked with o-PDA showing the decreasing resistive flat region; for this particular sample, even after 421 days, the low frequency region is only beginning to turn resistive as evidenced by the phase angle plot.

of each PDA isomer. It is seen they are rather comparable at earlier immersion times with p-PDA finally showing poorer performance after many months.

Figure 8 overlays plots from a coating cross-linked with ethylenediamine and the sample shown in Figure 6. It is seen that the aliphatic cross-linked sample begins to show resistive behavior at low frequency within a few days while that of the PDA sample is relatively unchanged after 42 days; all aliphatic cross-linked samples showed similar behavior. This change in EIS behavior and coating performance is quite dramatic considering the only difference between the samples is the cross-linking amine. Similar behavior to that seen in the PDA cross-linked EIS plots has been seen in previous investigations of sol gel coatings.8,19-23 It has been described as the effect of two distinct layers: a thin, dense Si-O-Al conversion layer formed at the metal interface, shown by the low frequency capacitive region, combined with a thicker, more porous sol-gel layer shown by the high frequency capacitance and resistive plateau.21 Few, if any, however, show this behavior for such extended periods of time as the current PDA cross-linked coatings.

To check for the presence and influence of an interfacial layer, an additional set of samples was prepared that did not have the GPTMS pre-treatment; that pre-treatment would be the most likely source

Figure 7. Average impedance values for three samples of aluminum alloy coated with sol-gels cross-linked with each PDA isomer.

Figure 8. Bode graph examples for 1,2-PDA and EDA cross-linked coatings.

for an interfacial layer with the present coating system. In the main sol, most of the Si-O-H moieties should be bound up to the inside of the colloidal particles to form the particles; there should still be enough to provide excellent adhesion to the substrate by itself, but with the absence of the pre-treatment, the density of the Si-O-Si bonds should be greatly reduced. Some evidence for this was seen in Figure 5; when an amine cross-linking agent wasn't used, the coating performed very poorly, indicating reduced amounts of Si-O-Si cross-linking. With an amine cross-linking agent, the space between the particles should be filled and cross-linked by either the aliphatic or PDA cross-linkers. Figure 9 shows a representative Bode plot of the samples without pre-treatment. While the resistive region is somewhat lower than when the GPTMS pre-treatment was present, neither the overall impedance nor the plot shape appear to change dramatically, at least not at earlier exposure time. It does begin to show resistive behavior at low frequencies much sooner than when the pre-treatment was used but this would most likely be due to a stabilizing effect from the PDA amines reacting to the epoxides of the pre-treatment. This suggests the behavior is due to something within the coating itself, for this behavior was not seen with any of the aliphatic cross-linking agents, thereby implicating the PDA isomers since that is the only difference between the coatings.

Figure 9. Bode graphs of 1,4-PDA cross-linked sol-gels, one with GPTMS pretreatment on the aluminum alloy substrate, one without.

There are several possible explanations for the enhanced durability of the PDA cross-linked coatings. One would be the presence of a "front" where the cross-linkers are more extensively polymerized, equivalent to the penetration of the oxygen-containing electrolyte; as noted earlier, we attributed the coloration seen to the oxidative polymerization of the PDA cross-linkers. Nearer the substrate, the PDA has not polymerized, or is less extensively polymerized, such that a capacitive-only response is measured for the lower portion of the coating. The ongoing polymerization of the PDA cross-linkers could serve as an oxygen scavenging mechanism, preventing it from penetrating through to the surface; conductive polymers have been shown to scavenge oxygen.24 Figure 10 depicts this scenario, along with a diagram of the circuit model used to fit the data at earlier exposure time; the model fits the data with all PDA isomers.

Figure 10. Schematic of coating with a capacitive region as modeled by CPE2 and a region that has localized conductive regions caused by the polymerization of PDA.

The second constant phase element (CPE) shown in Figure 10 can also be placed in series behind the second resistance, thus parallel to the first CPE, and that model fits the data equally well and yields almost identical values for the various elements. This suggests a transmission line model may also be appropriate; a transmission line model is commonly used with conductive polymers in which the charge transfer through the coating is through both ionic conduction in the pores and electronic conduction through the polymer.25 When polymerized, phenylenediamine yields poly(aniline)-like structures that are conductive, therefore, the PDA trapped in the interstitial spaces between the silica particles may be capable of passing current and providing a second conductive pathway with separate capacitance. Conductive polymers of poly(aniline), poly(pyrrole), and related species have been used to provide corrosion protection as surface treatments and in hybrid coating systems.26-27 Poly-PDA films have been examined for corrosion inhibition on steel.28-29 The PDA oligomers formed by oxidative polymerization of the cross-linkers used in this study could be expected to function in a similar manner, providing electrochemical corrosion inhibition. Attempts to find conductivity with 4-point probe and conductive AFM were unsuccessful but it should be noted that poly-PDA is normally only semi-conductive in nature; combined with the silica particles, the conductivity may simply be too low for measurement.

Finally, the EIS behavior seen could indeed be from an interfacial layer, just not aluminum oxide or a Si-O-Si layer. If the low frequency capacitance is from the aluminum oxide layer, it should be seen when aliphatic cross-linking agents are used and those samples should also last as long; neither was the case. If it is due to a Si-O-Si interfacial layer, it should also be visible with those samples because the only change was the cross-linking agent and, again, it should last as long. The density of epoxy-amine links between the pre-treatment epoxides and PDA amines may be far greater than with aliphatic cross-linkers, with less carbon chain at the immediate interface, acting to better stabilize the layer and increasing coating longevity; the flat aromatic rings could be expected to pack together more tightly. However, that does not explain the improved corrosion performance seen when a scratch is introduced through the coating; the bare aluminum in the scratch should corrode just as quickly as the metal in a scratch with any other coating.


Aluminum alloy 2024T3 has been coated with an organic-inorganic hybrid sol-gel coating with epoxide functionalized silica particles cross-linked by the isomers of phenylenediamine. These coatings demonstrated significant improvement in corrosion protection when compared to equivalent coatings in which the only difference was the amine cross-linking agents were aliphatic in nature. The coatings were studied with electrochemical impedance spectroscopy which showed vastly different behavior as compared to aliphatic amine cross-linked coatings. The in-situ oxidative oligomerization of the PDA isomers with potential conductivity has been identified as the likely source of the improved performance and impedance behavior.