- Email: [email protected]
Phosphonate-functionalized ORMOSIL coatings for magnesium alloys
Progress in Organic Coatings 65 (2009) 381–385
Contents lists available at ScienceDirect
Progress in Organic Coatings journal homepage: www.elsevier.com/locate/porgcoat
Phosphonate-functionalized ORMOSIL coatings for magnesium alloys A.N. Khramov a,∗ , J.A. Johnson b a b
Universal Technology Corporation, 1270 N. Fairfield Rd., Dayton, OH 45432-2600, USA Air Force Research Laboratory, Materials and Manufacturing Directorate, Wright-Patterson AFB, OH 45433-7750, USA
a r t i c l e
i n f o
Article history: Received 17 April 2008 Received in revised form 13 February 2009 Accepted 4 March 2009 Keywords: Sol–gel ORMOSIL Interpenetrating network Coating Corrosion Phosphonate Magnesium
a b s t r a c t Hybrid organically modified silicates, or ORMOSIL materials, with covalently attached pendant phosphonate groups were processed by a sol–gel method. A hydrolysis and condensation route was used involving a phosphonato-silane and resultant films were evaluated as prospective coatings for magnesium AZ31B alloy substrates. The inclusion of phosphonate functionalities into the coating matrix was achieved by a cocondensation of the phosphonato-silane with other organo-functional silanes followed by blending the resultant phosphonate-functionalized silicate oligomers with another sol–gel matrix material to form an interpenetrating network structure. The effect of synthesis composition on the hydrophobic/hydrophilic balance and film barrier properties was examined by several methods, including constant immersion and electrochemical tests. The observed enhancement in corrosion protection properties was attributed to a combination of the barrier properties of the organo-silicate matrix along with strengthening of the coating/substrate interface due to chemical bonding of the phosphonate groups to the metal surface. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Sol–gel-derived organo-silicate hybrids, also known as organically modified silicates (ORMOSILs) have attracted significant interest as versatile, easily processed coating materials with potential to replace chromate-based corrosion protection surface treatments for low density metals and their alloys [1–4]. The performance of such surface treatments is dependant upon a combination of barrier properties, adhesive bond strength to the substrate, and compatibility with any subsequent organic protective paints/coatings to be applied. Optimal adhesion and barrier properties can be achieved through introduction of various organofunctional groups into the ORMOSIL material, thus tailoring its chemical composition. In a recent publication on surface treatments for magnesium alloy substrates, we investigated phosphonate-modified sol–gel based surface coatings processed by a co-reaction of the phosphonato-silane with triethoxysilane [5]. The approach represented a promising way to incorporate phosphonate groups into ORMOSIL coating materials through sol–gel processing with commercially available phosphonato-silanes. The presence of pendant phosphonate groups covalently linked to the ORMOSIL network can substantially increase the ability to chemically react with certain metal substrates, leading to improved metal/coating adhesion. In particular, the phosphonate functionalities are capable
∗ Corresponding author. E-mail address: [email protected] (A.N. Khramov). 0300-9440/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.porgcoat.2009.03.001
of strong chemical bonding with the surface metal oxide layer on magnesium, aluminum and titanium substrates [6–8], making phosphonate-modified ORMOSILs very attractive materials for developing surface treatments and protective coatings for these low density metals and their alloys. The development of protective surface treatments for magnesium alloys is especially important due to the growing industrial demand for structural applications in the automotive, aerospace, and electronics industries. Because of the high chemical reactivity of magnesium, industrial applications of such materials almost always require the use of surface treatments to provide the necessary corrosion protection and paint adhesion [9,10]. In this paper, we report further developments on these phosphonate-containing ORMOSILs. In the first part, the effect of various silane co-reagents that impart different degrees of hydrophobicity on resultant films was studied. In the second part, various phosphonate-modified sol–gel oligomers were blended with another crosslinkable ORMOSIL matrix to form films with an interpenetrating network (IPN) structure. The corrosion protection properties of the films obtained by both these approaches and a comparison of the coatings with interpenetrating network formation versus coatings made via layer-by-layer deposition are discussed. 2. Experimental Tetraethoxysilane (TEOS), methyltriethoxysilane (MTEOS), dimethyldiethoxysilane (DMDES), diethylphosphonatoethyl-triethoxysilane (PHS), and 3-glycidoxypropyl-trimethoxysilane (GPTMS)
382
A.N. Khramov, J.A. Johnson / Progress in Organic Coatings 65 (2009) 381–385
were purchased from Gelest (Tullytown, PA, USA). Diethylenetriamine (DETA) was purchased from Aldrich. All chemicals were used as-received without additional purification. Reagent grade Type I water and HPLC grade ethanol (Aldrich) were used for preparation of sols. The organo-silicate sols were prepared via acid-catalyzed hydrolysis and condensation of a mixture of PHS and a chosen silane (typically at 1:4 molar ratios) in ethanol/water solution. For example, the sol of PHS:TEOS with 1:4 molar ratio was prepared by mixing of 0.02 mol of PHS and 0.08 mol of TEOS with 6.8 ml of 0.05 M acetic acid, followed by the addition of ethanol to make a total volume of 70 ml. The solution was vigorously stirred and aged for 3 days in a closed vial. The prepared sols were applied on cleaned AZ31B magnesium alloy coupons by a dip-coating procedure using a Model 201 dip-coater (Chemat Technology, Inc.) with a withdrawal speed of 10 cm/min. In the IPN formulations, the abovementioned phosphonatemodified sols were mixed with sol–gel-processed nano-sized particles with peripheral epoxy-groups [4]. The residual mixture was crosslinked through the epoxy-amine chemistry using DETA as a low molecular weight amine crosslinker. For example, 10 ml of a 3-day-old sol made with PHS and TEOS (1:4 molar ratio) was mixed with 20 ml of the epoxy-functionalized GPTM-TMOS sol prepared as previously described [4]. After addition of non-ionogenic surfactants [4], 0.5 ml of DETA, and the final dilution with water up to 35 ml of total volume, the solution was vigorously stirred for about 0.5 min, and applied by a dip-coating procedure to a pre-cleaned magnesium substrate. Coupons of sheet magnesium alloy AZ31B (2 × 3 ) were initially degreased in hexane and then cleaned in an acetic-nitrate pickle solution according to ASTM D 1732-03 standard procedure [11], and subsequently rinsed with distilled water and wiped with dry lintless cloth immediately after the bath. The coated samples were dried under ambient conditions for at least 24 h prior to testing. The thickness of the coatings was estimated from the scanning electron microscopy (SEM) micrographs of the cross-section view of the samples. The SEM was performed with a Leica Cambridge 360 FE electron microscope using a secondary electron image detector at 15 kV and 100 pA probe current. The water contact angle data were obtained using a FTA200 contact angle analyzer (First Ten Angstroms Inc.). Constant immersion tests and potentiodynamic scan (PDS) measurements were performed in dilute Harrison’s solution (0.35 wt% (NH4 )2 SO4 and 0.05 wt% NaCl). Electrochemical measurements were performed using a Gamry PC4/300 potentiostat coupled with Gamry Corrosion Measurement System CMS100. A one-chamber, three-electrode cell set-up was used, with the working electrode consisting of an exposed area of 6 cm2 on the surface of the sample. The reference and auxiliary electrodes were a saturated calomel electrode (SCE) and a platinum mesh electrode, respectively. The PDS scans were acquired in the range from −0.1 V up to 0.3 V versus open circuit potential at the rate of 1 mV/s. Prior to testing the samples were immersed in electrolyte solution either for 2 min (thinner films, Fig. 4) or for 10 min (IPN samples, Fig. 5).
3. Results and discussion Chemically, magnesium is the most reactive among structural metals and highly susceptible to corrosion. When magnesium is protected by a barrier coating, localized corrosion in aqueous solution can be initiated from any pin–hole defects and/or water uptake through the full coating system. The onset of corrosion is accompanied by an increase in pH due to the formation of magnesium hydroxide as the corrosion by-product [9]. Therefore, the ability of the coating system to withstand exposure to an alkaline envi-
Fig. 1. Chemical structures of the organo-functional silanes used for the sol–gel processing of hybrid coatings with phosphonate functionalities.
ronment will most likely result in improved corrosion protection properties. In the case of ORMOSIL hybrid coating materials, it is expected that the stability against alkaline hydrolysis can be enhanced by increasing the amount of non-hydrolizable silicon–carbon bonds, suggesting a higher ratio of organic content for better performance. Since silica itself is slightly soluble at a basic pH, it is reasonable to assume that the presence of hydrolytically stable silicon–carbon bonds will improve the stability of hybrid organo-silicates against alkaline hydrolysis as compared to pure silicate sol–gel materials. Under this assumption, we have tried to choose matching silanes with a systematically increasing number of silicon–carbon bonds (as shown in Fig. 1) for co-reaction with PHS. Since PHS itself has only one silicon–carbon bond, the PHS–TEOS co-condensation products have the least amount of such bonds and have the highest inorganic (silica) content, whereas the PHS–DMDES oligomers have the highest number of silica–carbon bonds and the lowest inorganic-to-organic ratio. Accordingly, as the number of such non-hydrolizable methyl groups is increasing in the following order, PHS–TEOS < PHS–MTEOS < PHS–DMDES, the sol–gel material becomes increasingly more hydrophobic, as indicated by the water contact angle data shown in Table 1. However, contrary to our expectations, as we examined the performance of resultant coatings subjected to the constant immersion test, we found the corrosion protection decreasing in the order of the increased hydrophobicity of the materials: PHS–TEOS > PHS–MTEOS > PHS–DMDES (Fig. 3). The PHS–TEOS coating demonstrates the highest corrosion protection, whereas PHS–DMDES coating shows very limited corrosion protection despite initial thoughts that such a material with a high amount of non-hydrolizable methyl groups will be the most stable against alkaline hydrolysis. This behavior might be explained by the possibly lower crosslinking density of the PHS–DMDES films (due to lower amount of the silanol groups) and the possible nanoscale phase separation due to increasing differences between the hydrophilic silica and hydrophobic methyl-siloxane clusters, leading to increased porosity (i.e., lower film barrier properties). The electrochemical (PDS) measurements data for resultant coatings are shown in Fig. 4 and are in generally good agreement with constant immersion test results. In the PDS experiments however, both PHS–TEOS and PHS–MTEOS demonstrate similar corTable 1 Coatings characteristics. Composition
Molar ratio
Thickness, m
Contact angle (water), ◦
PHS–TEOS PHS–MTEOS PHS–DMDES IPN (matrix only) PHS–TEOS–IPN PHS–MTEOS–IPN PHS–DMDES–IPN
1:4 1:4 1:4 – – – –
0.65–0.70 0.65–0.70 0.65–0.70 1.5–2 1.5–2 1.5–2 N/Aa
62 75 81 55 57 55 78
a
Visible phase separation resulting in a non-uniform coating.
A.N. Khramov, J.A. Johnson / Progress in Organic Coatings 65 (2009) 381–385
383
Fig. 2. Schematic illustration of the IPN ORMOSIL coating obtained by blending of the epoxy/amine organo-silicate matrix material with the phosphonate-functionalized oligomers.
rosion protection due to a shorter (∼60 min) overall immersion in electrolyte solution as compared to longer (24 h) constant immersion test exposure. They both show a significant improvement in corrosion protection as compared with control TEOS-processed, silica-only coating as it is indicated by a substantial reduction in corrosion current. At the same time, the PHS–DMDS coating demonstrates poor performance despite its more hydrophobic properties and expected higher hydrolytic stability. In another approach to increase the ability of the ORMOSIL material to withstand the alkaline environment caused by the corrosion of a magnesium substrate, we have tried blending the abovementioned phosphonate-modified silica oligomers with the previously described sol–gel hybrids obtained by the epoxyamine chemistry [4,12]. These hybrid sol–gel materials can be best described as nano-structured hybrids, where the nano-sized sil-
ica clusters with peripheral epoxy-groups are first synthesized by sol–gel processing and then crosslinked with amino-functionalized cross-linkers. We have found that these epoxy/amine sol–gels can be easily compatible with the abovementioned phosphonatefunctionalized silicates; though it is worth mentioning that these two types of sol–gel materials are processed in different conditions of hydrolysis and condensation of the organo-silanes. The sol–gel processing of phosphonate-functionalized silicates occurs at low water-to-silane ratio and leads to the mostly linear or weakly branched phosphonate-modified silicate oligomers [5]. The synthesis of epoxy-functionalized silicates was conducted at high water-to-silane ratio that favors the formation of the highly branched or particulate structures of the silicate aggregates [4,12,13]. We believe that when the mostly linear/weakly branched phosphonate-modified silicate oligomers are blended with the
Fig. 3. Visual inspection records of the constant immersion test for the two groups of the coating samples: upper row—phosphonate-functionalized coatings formulated with various silane co-reagents (exposed for 24 h); lower row—IPN ORMOSIL coatings formulated with blended phosphonate-functionalized silicates (exposed for 48 h).
384
A.N. Khramov, J.A. Johnson / Progress in Organic Coatings 65 (2009) 381–385
nano-sized epoxy-functionalized silica clusters and subsequently crosslinked through epoxy-amine chemistry, an interpenetrating network is formed as depicted in the schematic diagram (Fig. 2). The resulting IPN nano-structured ORMOSILs have relatively large organic content and the essential amount of phosphonate and silanol functionalities to be able to provide effective barrier protection with the improved coating/substrate interface. As indicated by the water contact angle data (Table 1), the matrix material dominates within the whole IPN structure and the increasing number of non-hydrolizable methyl groups in the phosphonate-modified silicates has no effect on hydrophobichydrophilic properties of the IPN coatings for all oligomers except for those obtained by co-condensation of PHS with DMDES. In the latter case, the addition of PHS–DMDES oligomers causes visible phase separation, resulting in non-uniform “spotted” coatings that demonstrated inconsistent corrosion protection and were excluded from further comparative analysis. As shown in Fig. 3, the incorporation of phosphonate-modified silicates into the IPN structure substantially improves the corrosion protection properties of the coating. An improvement over the IPN matrix material was observed in both cases when either PHS–TEOS or PHS–MTEOS was blended into the IPN structure, which could then withstand more than double the exposure time to the corrosive electrolyte and the alkaline magnesium hydroxide by-product. When we further visually examined the exposed area for all tested coatings, a different characteristic of corrosion became noticeable. The corrosion products appeared on the tested samples as following: severe corrosion on bare AZ31B sample as a dense dark layer; almost no visible corrosion on the PHS–TEOS–IPN coating sample; minor corrosion on the PHS–MTEOS–IPN coating as random dark spots; moderate corrosion on samples coated with the matrix material without addition of any phosphonate-modified silicates (Fig. 4). The enhancement of corrosion performance for these coating samples was also confirmed by the PDS analysis shown in Fig. 5. The coating of the AZ31B substrate with the matrix material alone reduces the corrosion current by more than 2 orders of magnitude. Then the addition of phosphonate-modified silicates into the IPN structure further decreases the corrosion current by approximately another order of magnitude. In agreement with the results of the constant immersion test, both PHS–TEOS–IPN and PHS–MTEOS–IPN coatings demonstrate comparable improvement in corrosion protection performance. However, the PDS analysis reveals additional information that gives some insight into possi-
Fig. 5. Potentiodynamic scans of the phosphonate-functionalized IPN ORMOSIL coatings (substrate—AZ31B; in dilute Harrison’s solution, PHS–TEOS–LL coating is obtained by layer-by-layer deposition and shown for comparison).
ble mechanisms of corrosion protection provided by the presence of the phosphonate-ester groups within the bulk of these coatings. When the phosphonate group is present, the upper half of the polarization curve exhibits a distinctive and reproducible noise pattern. This region of the PDS curve is associated with the anodic dissolution of the magnesium substrate and formation of magnesium hydroxide. When the phosphonate-ester groups are hydrolyzed by the magnesium hydroxide they can further form insoluble magnesium phosphonate thus “plugging” the pores or corrosion defects in the coating with precipitation product and increasing corrosion resistance. We hypothesize that the interruption in the flow of magnesium ions “scavenged” by the phosphonate groups is causing disruptions in the electrochemical reaction circuit and the resultant noisy spikes on polarization curves. Therefore, such noisy patterns on the polarization curves are the “finger-prints” of a pore-plugging mechanism that may play an important role in the corrosion protection provided by the incorporation of phosphonate-modified silicates into the IPN coating. It is interesting to note the performance of coatings obtained by means of a “layer-by-layer” deposition process as shown in Fig. 5 and designated with the LL abbreviation. This coating is obtained by placing a first layer of PHS–TEOS sol–gel material followed by the deposition of the ORMOSIL matrix material on top of it without blending them together into an IPN structure. This PHS–TEOS–LL coating also shows the noisy pattern on the anodic branch of the polarization curve although with much reduced magnitude of the spikes. The corrosion current for this coating is slightly less than that for the IPN coatings, but the anodic peak height is also lower indicating somewhat diminished barrier properties as compared to either IPN coatings with added phosphonate-modified silicates or comprised of the matrix material alone. 4. Conclusions
Fig. 4. Potentiodynamic scans of the phosphonate-functionalized coatings obtained by sol–gel processing of PHS and a variable silane co-reagent at constant (1:4) molar ratio (substrate—AZ31B; in dilute Harrison’s solution).
In summary, we have performed a comparative study on the corrosion protection properties of phosphonate-modified hybrid organo-silicate coatings based on the condensation products of a phosphonate-functionalized silane. We have investigated the effectiveness of several approaches to utilize these compounds as coating materials to protect AZ31B magnesium alloy substrates. It appears that the major challenge still to overcome is related to the generally insufficient hydrolytic stability of sol–gel coatings against alkaline environments associated with corrosion of magnesium materials. The straightforward approach to increase
A.N. Khramov, J.A. Johnson / Progress in Organic Coatings 65 (2009) 381–385
effectiveness through increasing the amount of non-hydrolyzable methyl groups within sol–gel films failed to provide adequate corrosion protection due to unfavorable nano-scale phase separation and resultant porous film defects. However, the performance can be substantially improved by the implementation of an interpenetrating network coating morphology, which was achieved by blending phosphonate-functionalized silicates with a nanostructured matrix material. Such phosphonate-modified ORMOSIL materials can provide effective corrosion protection, combining the ability of phosphonate groups to chemically react on the magnesium surface, the enhanced barrier properties and tolerance of the ORMOSIL material to withstand pH changes associated with magnesium corrosion, and the “self-healing” capacity of the material to block the corrosion defects with magnesium/phosphonate precipitation products. Acknowledgment The authors are grateful to Maj. J. Gresham for financial support on this project through the Air Force Office of Scientific
385
Research (AFOSR) Surface and Interfacial Science “Smart Surfaces” program. References [1] Y.J. Du, M. Damron, G. Tang, H. Zheng, C.-J. Chu, J.H. Osborne, Prog. Org. Coat. 41 (2001) 226. [2] T.L. Metroke, O. Kachurina, E.T. Knobbe, Prog. Org. Coat. 44 (2002) 185. [3] T.L. Metroke, J.S. Gandhi, A. Apblett, Prog. Org. Coat. 50 (2004) 231. [4] A.N. Khramov, V.N. Balbyshev, N.N. Voevodin, M.S. Donley, Prog. Org. Coat. 47 (2003) 207. [5] A.N. Khramov, V.N. Balbyshev, L.S. Kasten, R.A. Mantz, Thin Solid Films 514 (2006) 174. [6] M.B. Mitchell, V.N. Sheinker, E.A. Mintz, J. Phys. Chem. B 101 (1997) 11192. [7] W. Gao, L. Dickinson, C. Grozinger, F.G. Morin, L. Reven, Langmuir 12 (1996) 6429. [8] G. Guerrero, P.H. Mutin, A. Vioux, Chem. Mater. 12 (2000) 1268. [9] G.L. Song, A. Atrens, Eng. Mater. 1 (1999) 11. [10] J.E. Gray, B. Luan, J. Alloys Compd. 336 (2002) 88. [11] D 1732-03, Standard Practices for Preparation of Magnesium Alloy Surfaces for Painting, ISTM International Standards, West Conshohocken, PA, USA, 2003. [12] M.S. Donley, R.A. Mantz, A.N. Khramov, V.N. Balbyshev, L.S. Kasten, D.J. Gaspar, Prog. Org. Coat. 47 (2003) 401. [13] A.J. Vreugdenhil, V.N. Balbyshev, M.S. Donley, J. Coat. Technol. 73 (915) (2001) 35.
Contents lists available at ScienceDirect
Progress in Organic Coatings journal homepage: www.elsevier.com/locate/porgcoat
Phosphonate-functionalized ORMOSIL coatings for magnesium alloys A.N. Khramov a,∗ , J.A. Johnson b a b
Universal Technology Corporation, 1270 N. Fairfield Rd., Dayton, OH 45432-2600, USA Air Force Research Laboratory, Materials and Manufacturing Directorate, Wright-Patterson AFB, OH 45433-7750, USA
a r t i c l e
i n f o
Article history: Received 17 April 2008 Received in revised form 13 February 2009 Accepted 4 March 2009 Keywords: Sol–gel ORMOSIL Interpenetrating network Coating Corrosion Phosphonate Magnesium
a b s t r a c t Hybrid organically modified silicates, or ORMOSIL materials, with covalently attached pendant phosphonate groups were processed by a sol–gel method. A hydrolysis and condensation route was used involving a phosphonato-silane and resultant films were evaluated as prospective coatings for magnesium AZ31B alloy substrates. The inclusion of phosphonate functionalities into the coating matrix was achieved by a cocondensation of the phosphonato-silane with other organo-functional silanes followed by blending the resultant phosphonate-functionalized silicate oligomers with another sol–gel matrix material to form an interpenetrating network structure. The effect of synthesis composition on the hydrophobic/hydrophilic balance and film barrier properties was examined by several methods, including constant immersion and electrochemical tests. The observed enhancement in corrosion protection properties was attributed to a combination of the barrier properties of the organo-silicate matrix along with strengthening of the coating/substrate interface due to chemical bonding of the phosphonate groups to the metal surface. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Sol–gel-derived organo-silicate hybrids, also known as organically modified silicates (ORMOSILs) have attracted significant interest as versatile, easily processed coating materials with potential to replace chromate-based corrosion protection surface treatments for low density metals and their alloys [1–4]. The performance of such surface treatments is dependant upon a combination of barrier properties, adhesive bond strength to the substrate, and compatibility with any subsequent organic protective paints/coatings to be applied. Optimal adhesion and barrier properties can be achieved through introduction of various organofunctional groups into the ORMOSIL material, thus tailoring its chemical composition. In a recent publication on surface treatments for magnesium alloy substrates, we investigated phosphonate-modified sol–gel based surface coatings processed by a co-reaction of the phosphonato-silane with triethoxysilane [5]. The approach represented a promising way to incorporate phosphonate groups into ORMOSIL coating materials through sol–gel processing with commercially available phosphonato-silanes. The presence of pendant phosphonate groups covalently linked to the ORMOSIL network can substantially increase the ability to chemically react with certain metal substrates, leading to improved metal/coating adhesion. In particular, the phosphonate functionalities are capable
∗ Corresponding author. E-mail address: [email protected] (A.N. Khramov). 0300-9440/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.porgcoat.2009.03.001
of strong chemical bonding with the surface metal oxide layer on magnesium, aluminum and titanium substrates [6–8], making phosphonate-modified ORMOSILs very attractive materials for developing surface treatments and protective coatings for these low density metals and their alloys. The development of protective surface treatments for magnesium alloys is especially important due to the growing industrial demand for structural applications in the automotive, aerospace, and electronics industries. Because of the high chemical reactivity of magnesium, industrial applications of such materials almost always require the use of surface treatments to provide the necessary corrosion protection and paint adhesion [9,10]. In this paper, we report further developments on these phosphonate-containing ORMOSILs. In the first part, the effect of various silane co-reagents that impart different degrees of hydrophobicity on resultant films was studied. In the second part, various phosphonate-modified sol–gel oligomers were blended with another crosslinkable ORMOSIL matrix to form films with an interpenetrating network (IPN) structure. The corrosion protection properties of the films obtained by both these approaches and a comparison of the coatings with interpenetrating network formation versus coatings made via layer-by-layer deposition are discussed. 2. Experimental Tetraethoxysilane (TEOS), methyltriethoxysilane (MTEOS), dimethyldiethoxysilane (DMDES), diethylphosphonatoethyl-triethoxysilane (PHS), and 3-glycidoxypropyl-trimethoxysilane (GPTMS)
382
A.N. Khramov, J.A. Johnson / Progress in Organic Coatings 65 (2009) 381–385
were purchased from Gelest (Tullytown, PA, USA). Diethylenetriamine (DETA) was purchased from Aldrich. All chemicals were used as-received without additional purification. Reagent grade Type I water and HPLC grade ethanol (Aldrich) were used for preparation of sols. The organo-silicate sols were prepared via acid-catalyzed hydrolysis and condensation of a mixture of PHS and a chosen silane (typically at 1:4 molar ratios) in ethanol/water solution. For example, the sol of PHS:TEOS with 1:4 molar ratio was prepared by mixing of 0.02 mol of PHS and 0.08 mol of TEOS with 6.8 ml of 0.05 M acetic acid, followed by the addition of ethanol to make a total volume of 70 ml. The solution was vigorously stirred and aged for 3 days in a closed vial. The prepared sols were applied on cleaned AZ31B magnesium alloy coupons by a dip-coating procedure using a Model 201 dip-coater (Chemat Technology, Inc.) with a withdrawal speed of 10 cm/min. In the IPN formulations, the abovementioned phosphonatemodified sols were mixed with sol–gel-processed nano-sized particles with peripheral epoxy-groups [4]. The residual mixture was crosslinked through the epoxy-amine chemistry using DETA as a low molecular weight amine crosslinker. For example, 10 ml of a 3-day-old sol made with PHS and TEOS (1:4 molar ratio) was mixed with 20 ml of the epoxy-functionalized GPTM-TMOS sol prepared as previously described [4]. After addition of non-ionogenic surfactants [4], 0.5 ml of DETA, and the final dilution with water up to 35 ml of total volume, the solution was vigorously stirred for about 0.5 min, and applied by a dip-coating procedure to a pre-cleaned magnesium substrate. Coupons of sheet magnesium alloy AZ31B (2 × 3 ) were initially degreased in hexane and then cleaned in an acetic-nitrate pickle solution according to ASTM D 1732-03 standard procedure [11], and subsequently rinsed with distilled water and wiped with dry lintless cloth immediately after the bath. The coated samples were dried under ambient conditions for at least 24 h prior to testing. The thickness of the coatings was estimated from the scanning electron microscopy (SEM) micrographs of the cross-section view of the samples. The SEM was performed with a Leica Cambridge 360 FE electron microscope using a secondary electron image detector at 15 kV and 100 pA probe current. The water contact angle data were obtained using a FTA200 contact angle analyzer (First Ten Angstroms Inc.). Constant immersion tests and potentiodynamic scan (PDS) measurements were performed in dilute Harrison’s solution (0.35 wt% (NH4 )2 SO4 and 0.05 wt% NaCl). Electrochemical measurements were performed using a Gamry PC4/300 potentiostat coupled with Gamry Corrosion Measurement System CMS100. A one-chamber, three-electrode cell set-up was used, with the working electrode consisting of an exposed area of 6 cm2 on the surface of the sample. The reference and auxiliary electrodes were a saturated calomel electrode (SCE) and a platinum mesh electrode, respectively. The PDS scans were acquired in the range from −0.1 V up to 0.3 V versus open circuit potential at the rate of 1 mV/s. Prior to testing the samples were immersed in electrolyte solution either for 2 min (thinner films, Fig. 4) or for 10 min (IPN samples, Fig. 5).
3. Results and discussion Chemically, magnesium is the most reactive among structural metals and highly susceptible to corrosion. When magnesium is protected by a barrier coating, localized corrosion in aqueous solution can be initiated from any pin–hole defects and/or water uptake through the full coating system. The onset of corrosion is accompanied by an increase in pH due to the formation of magnesium hydroxide as the corrosion by-product [9]. Therefore, the ability of the coating system to withstand exposure to an alkaline envi-
Fig. 1. Chemical structures of the organo-functional silanes used for the sol–gel processing of hybrid coatings with phosphonate functionalities.
ronment will most likely result in improved corrosion protection properties. In the case of ORMOSIL hybrid coating materials, it is expected that the stability against alkaline hydrolysis can be enhanced by increasing the amount of non-hydrolizable silicon–carbon bonds, suggesting a higher ratio of organic content for better performance. Since silica itself is slightly soluble at a basic pH, it is reasonable to assume that the presence of hydrolytically stable silicon–carbon bonds will improve the stability of hybrid organo-silicates against alkaline hydrolysis as compared to pure silicate sol–gel materials. Under this assumption, we have tried to choose matching silanes with a systematically increasing number of silicon–carbon bonds (as shown in Fig. 1) for co-reaction with PHS. Since PHS itself has only one silicon–carbon bond, the PHS–TEOS co-condensation products have the least amount of such bonds and have the highest inorganic (silica) content, whereas the PHS–DMDES oligomers have the highest number of silica–carbon bonds and the lowest inorganic-to-organic ratio. Accordingly, as the number of such non-hydrolizable methyl groups is increasing in the following order, PHS–TEOS < PHS–MTEOS < PHS–DMDES, the sol–gel material becomes increasingly more hydrophobic, as indicated by the water contact angle data shown in Table 1. However, contrary to our expectations, as we examined the performance of resultant coatings subjected to the constant immersion test, we found the corrosion protection decreasing in the order of the increased hydrophobicity of the materials: PHS–TEOS > PHS–MTEOS > PHS–DMDES (Fig. 3). The PHS–TEOS coating demonstrates the highest corrosion protection, whereas PHS–DMDES coating shows very limited corrosion protection despite initial thoughts that such a material with a high amount of non-hydrolizable methyl groups will be the most stable against alkaline hydrolysis. This behavior might be explained by the possibly lower crosslinking density of the PHS–DMDES films (due to lower amount of the silanol groups) and the possible nanoscale phase separation due to increasing differences between the hydrophilic silica and hydrophobic methyl-siloxane clusters, leading to increased porosity (i.e., lower film barrier properties). The electrochemical (PDS) measurements data for resultant coatings are shown in Fig. 4 and are in generally good agreement with constant immersion test results. In the PDS experiments however, both PHS–TEOS and PHS–MTEOS demonstrate similar corTable 1 Coatings characteristics. Composition
Molar ratio
Thickness, m
Contact angle (water), ◦
PHS–TEOS PHS–MTEOS PHS–DMDES IPN (matrix only) PHS–TEOS–IPN PHS–MTEOS–IPN PHS–DMDES–IPN
1:4 1:4 1:4 – – – –
0.65–0.70 0.65–0.70 0.65–0.70 1.5–2 1.5–2 1.5–2 N/Aa
62 75 81 55 57 55 78
a
Visible phase separation resulting in a non-uniform coating.
A.N. Khramov, J.A. Johnson / Progress in Organic Coatings 65 (2009) 381–385
383
Fig. 2. Schematic illustration of the IPN ORMOSIL coating obtained by blending of the epoxy/amine organo-silicate matrix material with the phosphonate-functionalized oligomers.
rosion protection due to a shorter (∼60 min) overall immersion in electrolyte solution as compared to longer (24 h) constant immersion test exposure. They both show a significant improvement in corrosion protection as compared with control TEOS-processed, silica-only coating as it is indicated by a substantial reduction in corrosion current. At the same time, the PHS–DMDS coating demonstrates poor performance despite its more hydrophobic properties and expected higher hydrolytic stability. In another approach to increase the ability of the ORMOSIL material to withstand the alkaline environment caused by the corrosion of a magnesium substrate, we have tried blending the abovementioned phosphonate-modified silica oligomers with the previously described sol–gel hybrids obtained by the epoxyamine chemistry [4,12]. These hybrid sol–gel materials can be best described as nano-structured hybrids, where the nano-sized sil-
ica clusters with peripheral epoxy-groups are first synthesized by sol–gel processing and then crosslinked with amino-functionalized cross-linkers. We have found that these epoxy/amine sol–gels can be easily compatible with the abovementioned phosphonatefunctionalized silicates; though it is worth mentioning that these two types of sol–gel materials are processed in different conditions of hydrolysis and condensation of the organo-silanes. The sol–gel processing of phosphonate-functionalized silicates occurs at low water-to-silane ratio and leads to the mostly linear or weakly branched phosphonate-modified silicate oligomers [5]. The synthesis of epoxy-functionalized silicates was conducted at high water-to-silane ratio that favors the formation of the highly branched or particulate structures of the silicate aggregates [4,12,13]. We believe that when the mostly linear/weakly branched phosphonate-modified silicate oligomers are blended with the
Fig. 3. Visual inspection records of the constant immersion test for the two groups of the coating samples: upper row—phosphonate-functionalized coatings formulated with various silane co-reagents (exposed for 24 h); lower row—IPN ORMOSIL coatings formulated with blended phosphonate-functionalized silicates (exposed for 48 h).
384
A.N. Khramov, J.A. Johnson / Progress in Organic Coatings 65 (2009) 381–385
nano-sized epoxy-functionalized silica clusters and subsequently crosslinked through epoxy-amine chemistry, an interpenetrating network is formed as depicted in the schematic diagram (Fig. 2). The resulting IPN nano-structured ORMOSILs have relatively large organic content and the essential amount of phosphonate and silanol functionalities to be able to provide effective barrier protection with the improved coating/substrate interface. As indicated by the water contact angle data (Table 1), the matrix material dominates within the whole IPN structure and the increasing number of non-hydrolizable methyl groups in the phosphonate-modified silicates has no effect on hydrophobichydrophilic properties of the IPN coatings for all oligomers except for those obtained by co-condensation of PHS with DMDES. In the latter case, the addition of PHS–DMDES oligomers causes visible phase separation, resulting in non-uniform “spotted” coatings that demonstrated inconsistent corrosion protection and were excluded from further comparative analysis. As shown in Fig. 3, the incorporation of phosphonate-modified silicates into the IPN structure substantially improves the corrosion protection properties of the coating. An improvement over the IPN matrix material was observed in both cases when either PHS–TEOS or PHS–MTEOS was blended into the IPN structure, which could then withstand more than double the exposure time to the corrosive electrolyte and the alkaline magnesium hydroxide by-product. When we further visually examined the exposed area for all tested coatings, a different characteristic of corrosion became noticeable. The corrosion products appeared on the tested samples as following: severe corrosion on bare AZ31B sample as a dense dark layer; almost no visible corrosion on the PHS–TEOS–IPN coating sample; minor corrosion on the PHS–MTEOS–IPN coating as random dark spots; moderate corrosion on samples coated with the matrix material without addition of any phosphonate-modified silicates (Fig. 4). The enhancement of corrosion performance for these coating samples was also confirmed by the PDS analysis shown in Fig. 5. The coating of the AZ31B substrate with the matrix material alone reduces the corrosion current by more than 2 orders of magnitude. Then the addition of phosphonate-modified silicates into the IPN structure further decreases the corrosion current by approximately another order of magnitude. In agreement with the results of the constant immersion test, both PHS–TEOS–IPN and PHS–MTEOS–IPN coatings demonstrate comparable improvement in corrosion protection performance. However, the PDS analysis reveals additional information that gives some insight into possi-
Fig. 5. Potentiodynamic scans of the phosphonate-functionalized IPN ORMOSIL coatings (substrate—AZ31B; in dilute Harrison’s solution, PHS–TEOS–LL coating is obtained by layer-by-layer deposition and shown for comparison).
ble mechanisms of corrosion protection provided by the presence of the phosphonate-ester groups within the bulk of these coatings. When the phosphonate group is present, the upper half of the polarization curve exhibits a distinctive and reproducible noise pattern. This region of the PDS curve is associated with the anodic dissolution of the magnesium substrate and formation of magnesium hydroxide. When the phosphonate-ester groups are hydrolyzed by the magnesium hydroxide they can further form insoluble magnesium phosphonate thus “plugging” the pores or corrosion defects in the coating with precipitation product and increasing corrosion resistance. We hypothesize that the interruption in the flow of magnesium ions “scavenged” by the phosphonate groups is causing disruptions in the electrochemical reaction circuit and the resultant noisy spikes on polarization curves. Therefore, such noisy patterns on the polarization curves are the “finger-prints” of a pore-plugging mechanism that may play an important role in the corrosion protection provided by the incorporation of phosphonate-modified silicates into the IPN coating. It is interesting to note the performance of coatings obtained by means of a “layer-by-layer” deposition process as shown in Fig. 5 and designated with the LL abbreviation. This coating is obtained by placing a first layer of PHS–TEOS sol–gel material followed by the deposition of the ORMOSIL matrix material on top of it without blending them together into an IPN structure. This PHS–TEOS–LL coating also shows the noisy pattern on the anodic branch of the polarization curve although with much reduced magnitude of the spikes. The corrosion current for this coating is slightly less than that for the IPN coatings, but the anodic peak height is also lower indicating somewhat diminished barrier properties as compared to either IPN coatings with added phosphonate-modified silicates or comprised of the matrix material alone. 4. Conclusions
Fig. 4. Potentiodynamic scans of the phosphonate-functionalized coatings obtained by sol–gel processing of PHS and a variable silane co-reagent at constant (1:4) molar ratio (substrate—AZ31B; in dilute Harrison’s solution).
In summary, we have performed a comparative study on the corrosion protection properties of phosphonate-modified hybrid organo-silicate coatings based on the condensation products of a phosphonate-functionalized silane. We have investigated the effectiveness of several approaches to utilize these compounds as coating materials to protect AZ31B magnesium alloy substrates. It appears that the major challenge still to overcome is related to the generally insufficient hydrolytic stability of sol–gel coatings against alkaline environments associated with corrosion of magnesium materials. The straightforward approach to increase
A.N. Khramov, J.A. Johnson / Progress in Organic Coatings 65 (2009) 381–385
effectiveness through increasing the amount of non-hydrolyzable methyl groups within sol–gel films failed to provide adequate corrosion protection due to unfavorable nano-scale phase separation and resultant porous film defects. However, the performance can be substantially improved by the implementation of an interpenetrating network coating morphology, which was achieved by blending phosphonate-functionalized silicates with a nanostructured matrix material. Such phosphonate-modified ORMOSIL materials can provide effective corrosion protection, combining the ability of phosphonate groups to chemically react on the magnesium surface, the enhanced barrier properties and tolerance of the ORMOSIL material to withstand pH changes associated with magnesium corrosion, and the “self-healing” capacity of the material to block the corrosion defects with magnesium/phosphonate precipitation products. Acknowledgment The authors are grateful to Maj. J. Gresham for financial support on this project through the Air Force Office of Scientific
385
Research (AFOSR) Surface and Interfacial Science “Smart Surfaces” program. References [1] Y.J. Du, M. Damron, G. Tang, H. Zheng, C.-J. Chu, J.H. Osborne, Prog. Org. Coat. 41 (2001) 226. [2] T.L. Metroke, O. Kachurina, E.T. Knobbe, Prog. Org. Coat. 44 (2002) 185. [3] T.L. Metroke, J.S. Gandhi, A. Apblett, Prog. Org. Coat. 50 (2004) 231. [4] A.N. Khramov, V.N. Balbyshev, N.N. Voevodin, M.S. Donley, Prog. Org. Coat. 47 (2003) 207. [5] A.N. Khramov, V.N. Balbyshev, L.S. Kasten, R.A. Mantz, Thin Solid Films 514 (2006) 174. [6] M.B. Mitchell, V.N. Sheinker, E.A. Mintz, J. Phys. Chem. B 101 (1997) 11192. [7] W. Gao, L. Dickinson, C. Grozinger, F.G. Morin, L. Reven, Langmuir 12 (1996) 6429. [8] G. Guerrero, P.H. Mutin, A. Vioux, Chem. Mater. 12 (2000) 1268. [9] G.L. Song, A. Atrens, Eng. Mater. 1 (1999) 11. [10] J.E. Gray, B. Luan, J. Alloys Compd. 336 (2002) 88. [11] D 1732-03, Standard Practices for Preparation of Magnesium Alloy Surfaces for Painting, ISTM International Standards, West Conshohocken, PA, USA, 2003. [12] M.S. Donley, R.A. Mantz, A.N. Khramov, V.N. Balbyshev, L.S. Kasten, D.J. Gaspar, Prog. Org. Coat. 47 (2003) 401. [13] A.J. Vreugdenhil, V.N. Balbyshev, M.S. Donley, J. Coat. Technol. 73 (915) (2001) 35.