The Effect of Cure Temperature and Humidity on the Properties of Solvent-Borne Zinc Silicate Coatings

Transcription

1 The Effect of Cure Temperature and Humidity on the Properties of Solvent-Borne Zinc Silicate Coatings A method developed to measure cure rate showed that temperature and humidity during initial film-forming dictate the properties of solvent-borne zinc silicates. by Gerald Eccleston, Scientific Services Laboratory Fig. 1 - (above) Steel sections used for the construction of the Omega Telecommunications Tower Victoria Photos courtesy of Scientific Services Laboratory Forty years after their invention by Nightingall, 1 an Australian engineer, inorganic zinc silicate coatings are widely used as primers to protect structural steelwork (Fig. 1). Today the most important class of self-curing zinc silicate is that based on alkyl silicates (solvent-borne). An important virtue of these coatings it that they typically develop resistance to rain within 30 minutes of application in contrast to the alkali silicate (water-borne) coatings, which remain sensitive to water for up to 24 hours and thus run the risk of being washed away. Despite their advantages, there is ample qualitative and anecdotal information indicating that alkyl silicates can form soft, friable coatings or be prone to delamination (Fig. 2) when overcoated after application and cured under conditions of low humidity. For example, it is a common practice in the coating industry to spray the 36 JANUARY 1998 / JPCL PMC

2 coating with water if the humidity at the time of application is low. It is also reported that these alkyl silicates do not cure at humidities of 40 percent relative humidity (RH) and below. 2 Fig. 2 - Delamination of overcoated zinc silicate coating The mechanism of cure of these coatings is fairly involved and includes hydrolysis (Fig. 3a), condensation (Fig. 3b), transesterification (Fig. 3c), and solvent evaporation. In the initial production of the alkyl silicate binder, the material is prehydrolyzed. Prehydrolysis is achieved by the addition of a small, precise quantity of water, and the solution is made slightly acidic. The resulting hydrolysis reaction (Fig. 3a) involves the removal of some of the alkoxy groups to produce silanol groups and alcohol. In the resulting prehydrolyzed binder, the silanol groups are stable in the acidic binder 3 (Fig. 3a). Immediately before application, when the zinc is mixed with the prehydrolyzed binder, the acid is neutralized, and the silanol groups become reactive. 4 In the resulting condensation reactions (Fig. 3b), the silanol groups can react either with each other to form the silicate matrix and water, or with an adjacent alkoxy group (Et-O-)[Et=C 2 H 5 ] to produce the corresponding alcohol (Et-O-H). The additional water produced can then promote further hydrolysis of unreacted alkoxy groups. Tracking these complex reactions is in consequence somewhat involved. In recent years, several techniques have been used to monitor the curing of these binders. These techniques include infrared spectroscopy, 5 silicon nuclear magnetic resonance (NMR) 6, and gas chromatography. 7 Of these techniques, gas chromatography allows for quantitative determination of the reactions involved in the curing of these coatings. The present article describes a study in which a method was developed using gas chromatography to determine the rate of cure of alkyl silicate coatings at different temperatures and humidities. The method involved hydrolyzing unreacted alkoxy groups in the alkyl silicate to their corresponding alcohol and measuring the amount produced using gas chromatography. By determining the silicon content, researchers were also able to determine the alkoxy/silicon ratio of the alkyl silicate and hence the degree of cure. Initially, the prehydrolyzed binder has a high alkoxy/silicon ratio, reflecting the unreacted alkoxy groups. However, when the curing reaction is complete, and all of the alkoxy groups have reacted, the alkoxy/ silicon ratio is much lower. The results were then correlated with the physical properties of the coating as measured by Taber abrasion resistance and pencil hardness. The aim of the work was to determine the effects of the initial temperature and humidity on the degree of cure and several physical properties of the coating JPCL PMC / JANUARY

3 a.) Hydrolysis OEt OEt OEt Si OEt + H 2 O = OEt Si OH + EtOH OEt OEt b.) Condensation OEt OEt OEt OEt OEt Si OH + OEt Si OEt = OEt Si O Si OEt + EtOH OEt OEt OEt OEt OEt OEt OEt OEt OEt Si OH + HO Si OEt = OEt Si O Si OEt + H 2 O OEt OEt OEt OEt c.) Transesterification OR OR OR Si OR + R'OH = R'O Si OR + ROH OR OR and, consequently, whether curing is significantly retarded or impaired at low humidities. The studies showed that the temperature and humidity during the initial film forming process dictate the physical properties of these coatings. Material and Methods The alkyl silicate was produced from polyethoxysilane. The material was partially hydrolyzed to 80 percent, i.e., 80 percent prehydrolysis, based on the reaction: (EtO) 12 Si 5 O 4 + 6H 2 O = 5SiO EtOH In the acidic solution, however, only the hydrolysis component of the reaction (Fig. 3a) proceeds while the condensation reaction, which generates the H 2 O required for the reaction to proceed to completion, is very slow. The alkoxy to silicon ratio of the polyethoxysilane 40 is 2.4 (i.e., 12:5); after prehydrolysis, the reaction was found to be 1.48 rather than the 0.48 that would have occurred if the reaction had proceeded to completion. (1.48 is based on the author s analysis of the 80 percent prehydrolyzed product; 0.48 is the theoretical ratio based on the reaction going to completion.) The silicon content was measured using a gravimetric method. 8 After curing, the coating was dried overnight in a vacuum desiccator to remove residual solvent g of the coating was weighted into a 30 ml culture bottle with screw cap. 5.0 ml of concentrated hydrochloric acid was added to the bottle to dissolve the zinc powder and then sealed. After 5 minutes, 20 ml of 20 percent sodium hydroxide was added to dissolve the silica, and the bottle was again sealed and allowed to clear for 30 minutes. An internal standard of 50 ml of methyl ethyl ketone was added to the solution. The solution was then injected into the gas chromatograph, and the areas of the response peaks for the alcohols and the standard were measured and compared against those of prepared ethanol standards. The silicon content was determined by weighing 0.15 g of the material into a Fig. 3 - The reaction involved in the curing of alkyl silicates JPCL PMC / JANUARY

4 platinum crucible and dried overnight in a vacuum desiccator. Four to five drops of 1:1 sulfuric acid were added, the crucible was ignited in a furnace for 1 hour and weighted. The residue was then moistened with 1:1 sulfuric acid (mixed with water) and 10 ml of hydrofluoric acid added. The residue was evaporated to dryness and weighed. For researchers to determine the physical properties, the formulation was also applied by conventional spray to steel panels blast cleaned to an AS class 3 finish (SSPC-SP 5). 9 To ensure that the zinc dust remained in suspension, the solution was constantly agitated in the spray gun during application. Immediately after application, the zinc silicate-coated panels were placed in the humidity cabinet. Periodically, panels were removed, and the Taber abrasion resistance was determined in accordance with AS , Inorganic Zinc Silicate Paints (ASTM D 4060). 10 Results and Discussion The Effect of Relative Humidity on Cure Rate Relative humidity is defined as the ratio of the actual vapor pressure to the saturation vapor pressure at the same temperature. It is a measure of the relative saturation with water vapor of the atmosphere at certain temperatures. Thus, at a constant temperature, increasing the relative humidity increases the absolute humidity of water Alkoxy/Silicon Ratio content of the atmosphere. Both relative or absolute humidity and temperature must be known to adequately specify curing conditions. Once each coating had been applied, 80% RH & 25 C 60% RH & 25 C 40% RH & 25 C Time (hours) Fig. 4 - Degree of cure vs time at constant temperature it was cured for various periods, respectively at 80 percent RH, 60 percent RH, or 40 percent RH at a constant temperature of 25 C (77 F). By keeping the temperature constant and varying the humidity, researchers could assess the effect of atmospheric moisture on the cure rate. Air flow was maintained at constant velocity across the curing film to preclude convection effects. The results for the cure rate at different humidities (Fig. 4) show that the coating cures much more rapidly at 80 percent RH than at 60 percent RH and 40 percent RH. At 60 percent and 80 percent RH, both coatings approach the same steady state cure conditions, although at 60 percent RH, cure takes nearly 3 times as long. However, at 40 percent RH, the degree of cure does not approach the steady state achieved at 60 percent RH and above, even after a much longer period. It is also interesting to note that while the alkyl silicate binder was prehydrolyzed 40 JANUARY 1998 / JPCL PMC

5 to have 2 alkoxy groups per silicon atom or alkoxy/silicon ratio of 2.4, when zinc dust is added the number of alkoxy groups immediately decreases to just under 0.8. This decrease indicates that further hydrolysis Fig. 5 - Taber abrasion resistance vs time at constant temperature. The data do not readily fit to a smooth curve. The data trends are discussed in the text. Fig. 6 - Degree of cure vs time at constant absolute humidity has occurred during the mixing of the zinc dust with the alkyl silicate and before atmospheric exposure. The author has investigated the mechanism of this reaction further in an unpublished paper. The physical properties as measured by Taber abrasion (Fig. 5) show a strong correlation to that obtained for the alkoxy/ silicon at the same humidity. At 80 percent RH and 25 C (77 F), the coating reaches its steady state maximum abrasion resistance after 30 hours while at 60 percent RH and 25 C (77 F), the maximum abrasion resistance, which is slightly less than that cured at 80 percent RH and 25 C (77 F), is reached after 50 hours. A coating cured at 40 percent RH and 25 C (77 F), however, does not approach the abrasion resistance achieved at the higher humidities, even after 170 hours. The results show that the extent and rate of cure are directly related to the relative humidity of the atmosphere to which the coating is exposed. More important, they show that a coating cured at 40 percent RH (or less) and 25 C (77 F) is unlikely to achieve a satisfactory cure and will remain soft and friable even after prolonged cure. The Effect of Temperature on Cure Rate To determine the effect of temperature on cure rate of these coatings, the absolute humidity was kept constant at g JPCL PMC / JANUARY

6 moisture/g dry air while the temperature, and therefore the relative humidity, varied as follows: 20 C (68 F) at g moisture/g dry air (80 percent RH) 25 C (77 F) at g moisture/ g dry air (60 percent RH) 32 C (90 F) at g moisture/g dry air (40 percent RH) A graph of alkoxy/silicon ratio versus time for the coatings cured under these conditions is given in Fig. 6. The rate of decrease of alkoxy groups is most rapid at 20 C (68 F) and reaches a steady state maximum after 30 hours, while at 25 C (77 F) the rate of decrease is slower. But after 75 hours, the alkoxy/silicon is similar to that at 20 C (68 F). At 32 C (90 F), the rate of decrease is very slow, and the alkoxy/silicon ratio does not drop below 0.4, even after 100 hours, indicating that at this temperature, the coating is unlikely to achieve satisfactory cure. Again, there is a strong correlation between the Taber abrasion resistance of the coating (Fig. 7) and the alkoxy/silicon ratio. Thus, as the Taber abrasion resistance increases (i.e., the coating becomes harder), the alkoxy/silicon ratio decreases. The results indicate that at constant moisture content, as the temperature increases, the rate or extent of cure and the Taber abrasion resistance decrease. These results might seem surprising, since an increase in temperature would be expected to lead to an increase in the rate of hydrolysis and condensation reactions; hence, the loss of alkoxy groups should be greatest at the higher temperature. Since the hydrolysis and condensation rates do not increase, other factors must be involved. An explanation of the results would Fig. 7 - Taber abrasion resistance vs time at constant absolute humidity. The data do not readily fit to a smooth curve. The data trends are discussed in the text. appear to lie in the fact that during the curing process, the film changes from a liquid to a solid, and the reactivity of the components within the film must therefore change. In the liquid phase, before evaporation of the solvent is complete, the mobility of the reacting species allows the condensation and hydrolysis reactions to occur at a much faster rate than in the drier (solvent-depleted) film. In the drier film, the mobility of the reacting species is severely retarded, and the reaction proceeds more slowly, if at all. Thus, as the temperature increases, even though the hydrolysis and condensation reaction rates increase, the rate of evaporation of the solvent also increases, and hence the film will dry out more quickly and be reactive for a shorter period. As indicated by NMR, these decreases in drying time and reactivity lead to isolated pockets of polymeric SiO 2 (friable film) rather than an extended polymeric 42 JANUARY 1998 / JPCL PMC

7 network of silica (hard film), which is the desired result. 6 The Effects of Water Addition on Zinc Silicate Coatings It is frequently recommended that a fine mist of water be applied to alkyl zinc silicate coatings that have been applied under conditions of low humidity. To quantify the effectiveness of this practice, researchers decided to monitor the effects of adding water to the zinc silicate coating some time after the start of cure. They chose to add water 5 hours after the start of cure, at both 80 percent RH and 40 percent RH with the temperature kept constant at 25 C (77 F). After 5 hours, the coatings were immersed in water for 10 minutes. The purpose was to determine if a coating cured initially under poor conditions could be redeemed by water immersion (i.e., improve the degree of cure and hardness of the coating). The effect of water immersion on the coating cured at 80 percent RH and 40 percent RH were compared, both in terms of alkoxy/silicon ratio and Taber abrasion resistance of the coatings. After the coatings cure at 80 percent RH for 5 hours, immersing them in water results in a significant decrease in the early alkoxy/silicon ratio and a corresponding early increase in the Taber abrasion resistance. However, after the coatings cured at 40 percent RH, immersion did not result in Atmospheric conditions before solvent evaporation is complete are critical to coating performance. any appreciable loss in the alkoxy groups or any change in the Taber abrasion resistance. This result suggests that the film cured at 40 percent RH is dry and unreactive after 5 hours and has dried out much faster than a similar film cured at 80 percent RH. If water from the condensation reaction is also considered as a significant solvent, then the solvent loss would be expected to be greater at low humidities and the film would be wet, and hence reactive, for a shorter period than for a film cured at a higher humidity. Conclusions This work indicates that the atmospheric conditions prevailing during the initial reactive period of the coating, before solvent evaporation is complete, are critical to the performance of the coating. During this time, the film formation is influenced by temperature and humidity. Initially, the film is in solution, and the addition of zinc dust allows the silanol groups to condense to produce water; the presence of zinc dust accelerates the hydrolysis reaction. In the early stages, the reacting species are mobile in the liquid phase, and the curing reaction occurs rapidly. As the solvent starts to evaporate, the reacting species are brought together, and the equilibrium of both the hydrolysis 44 JANUARY 1998 / JPCL PMC

8 (Fig. 3a) and condensation reactions (Fig. 3a) is shifted to the right. With most of the solvent removed, the mobility of the reacting species decreases, and the reaction rate drops dramatically. Water plays an important role as one of the main reactants in alkyl silicate coatings. Although initially not present in the prehydrolyzed alkyl silicate solution, the condensation of silanol groups is capable of providing enough water to hydrolyze a further 40 percent of the alkoxy groups (to give a total of 80 percent hydrolysis). Further water is released in the binder by the addition of zinc dust. At low humidities, this water evaporates from the coating rapidly; hence, water is not available for the hydrolysis reaction. In addition, because the water is also a solvent, the film dries out much more quickly, and the coating is reactive for a shorter period. In contrast, at high humidities, the rate of evaporation is slower; more water is retained in the film for a longer period; and the hydrolysis and condensation reactions occur to a much greater extent. As the temperature increases at constant atmospheric moisture content, both the water and the organic solvent evaporation rates increase. The film, therefore, has less time to react with either atmospheric moisture or the water released from the condensation of silanol groups before the cure reaction ceases. Immersion of a coating in water after it has dried out does not measurably reduce the number of alkoxy groups or increase either the hardness or the abrasion resistance. After solvent evaporation, the physical and chemical condition of the coating appears to be irreversible. Indeed, attempts to improve the properties of a coating that has been cured under low humidity conditions by the spray application of water will only succeed if carried out before the coating dries out. In practical terms, hot or dry conditions need to be avoided during application and drying if an inorganic alkyl zinc silicate is to achieve its performance potential. If these conditions cannot be avoided, the coating would need to be sprayed with water before the coating completely dries out. Hot conditions are only detrimental if accompanied by low humidity. References 1. V. Nightingall, U.S. Patent 2,440,969, May W. McKeown, Fourth Meeting of Zinc- Rich Technologists Group, Melbourne, Australia, July H. Baumann, Kolloidzchr, Vol. 162, No. 28 (1959). 4. R. Richardson, J.A. Waddam, Research, Vol. 7 (1954), D.G. Weldon et al., The Cure and Topcoating of Inorganic Zinc-Rich Primers, Journal of Protective Coatings & Linings (April 1989), 21-23, N.J. Clayden, C.J. Rix, G. Eccleston, Reactions in Polyethoxysilane Coatings using Solution and Solid-state Si NMR Spectroscopy, Journal of Material Science, Vol. 22, No. 8 (1987), 2,913-2, G.B. Byrnes and L.D. Vincent, Measuring the Rate of Cure of Solvent-Based Inorganic Zinc Primers, Materials Performance (February 1994), R. K. Iler, Industrial & Engineering News, Vol. 39 (1942), p Australian Standard A.S. 1627, Abrasive Blast Cleaning of Steel Surfaces, Part 4 (Sydney, Australia: Standards Association of Australia, 1974). 10. Australian Standard A.S. 2105, Inorganic Zinc Silicate Paints, (Sydney, Australia: Standards Association of Australia, 1974). Gerald Eccleston is the Manager of the Scientific Services Laboratory (SSL) Painting Contractor Certification Program. He is also a Senior Materials Scientist in SSL s Coating and Building Materials Section. Eccleston has worked as a consultant for a range of private and public sector clients. He has set up and operated a third party appraisal and certification program for active fire protection products, as well as a program for protective coating applicators. He has performed selection, appraisal, testing, quality assurance, and failure investigation for a variety of coating and building products. Eccleston has more than 20 years experience in testing, investigation, and consulting for the building and construction industry. He previously headed a team within SSL s Building Material section responsible for providing testing facilities and technical advice for paints, corrosion protection, metallurgy, and building materials such as sealants, flooring membranes, and cladding. He then established and began operation of the Painting Contractor Certification Program to provide quality assurance for painting of steel structures. He can be reached at Scientific Services Laboratory, 177 Salmon Street, Port Melbourne, Victoria 3207, Australia; ; fax: JPCL PMC / JANUARY