United States Environmental Protection Agency Atmospheric Research and Exposure Assessment Laboratory Research Triangle Park NC 2771 1 Research and Development EPA/600/S3-88/044 Feb. 1989 &EPA Project Summary Analytical Techniques for Measuring the Effects of Acid Deposition on Coatings on Wood C. M. Balik, R. E. Fornes, and R. D. Gilbert Preliminary experiments have been carried out to characterize the potential deleterious effects of acidic deposition on three representative paints: an oil alkyd paint and two acrylic latex formulations. The base polymer latex common to both latex paints was also studied individually. Free films of paint have been exposed to relatively high levels of gaseous SO2 and ultraviolet light, and have been immersed in aqueous SC>2 at pH 2.0. Several analytical techniques have been used to assess the resulting chemical and physical changes in the paint films, including sorption and diffusion measurements, attenuated total reflectance infrared spectroscopy, dynamic mechanical analysis, sol- gel analysis, contact angle meas- urements, differential scanning calorimetry, and electron spin resonance. All techniques show promise for characterizing the early stages of damage to paint films caused by acidic deposition. The major effects noted in this study include leaching of acid-soluble extender components upon immers- ion in aqueous SOz, and enhanced degradation of the base polymer upon exposure to gaseous SOa and ultraviolet light This Project Summary was devel- oped by EPA's Atmospheric Research and Exposure Assessment Laboratory, Research Triangle Park, NC, to announce key findings of the research project that is fully docu- mented in a separate report of the same title (see Project Report ordering information at back). Introduction The objectives of this project are: (1) to study the effects of acid deposition products in combination with near-uv radiation of the micro- and macro- structures of acrylic latex paint coatings, (2) to assess microdamages experienced by acrylic copolymers upon exposure to acid deposition; (3) to determine mechanistic paths that lead to polymer structural changes and microdamage; and, (4) to relate molecular structural changes to macrodamages that affect the service life of exterior coatings used on wood substrates. The following analytical techniques will be employed: Intrinsic viscosity; Sol/Gel Analyses; Gel Permeation Chroma- tography; Surface Contact Angles; Dif- ferential Scanning Calorimetry (DSC); Dynamic Mechanical Analysis (DMA); Stress-Strain Analysis; Fatigue Tests; Infrared Spectroscopy; Electron Spin Resonance Spectroscopy; ESCA and SIMS; SEM - Fracture Surfaces studies. Changes will be followed in molecular weight, molecular weight distribution, glass transition temperature, changes in bulk properties, the amount of polymer chain scission, crosslinking and oxidation upon exposure of the paint coatings to ultra-violet radiation, S02, NOX and mixtures of the SC>2 and NOX with air and/or water. The resulting data will permit deter- mination of the mechanistic paths that result in the degradation of the base polymer in the paint coating and allow ------- the prediction of useful lifetimes of coatings exposed to acid rain pollutants. Experimental Procedures Films of the base and compounded latexes were prepared by the drawdown procedure, using standard draw-down bars, on release paper. They were dried at R.T. for 4 hr, stored in a desiccator oven drierite for 48 hr and then heated at 40°C for 48 hr under vacuum. The films prepared in this manner absorbed negligible amounts of water when stored at R.T. and 64% R.H. A sample of each film before and after exposure to various environments was obtained and placed in tetrahydrofuran (THF) for 3 days. The soluble portion was separated from the insoluble portion by filtration, the insoluble portion dried and weighed to determine the gel content. The amount of sol was determined by evaporation of the THF. Intrinsic viscosities were determined at 25°C using modified Cannon-Fenske viscometers. Thermal analyses were performed with either a Perkin-Elmer DSC-2 or -7, each equipped with a data station. Samples 1 cm by 6 cm were cut from each film using care to obtain uniform widths. The thicknesses were measured with a micrometer. Dynamic mechanical analyses (DMA) tests were made with an Autovibron DDV-II-C at a frequency of 11 hz and a scanning rate of 2.5°C/min. The films were exposed to uv light (254 or 350nm) in a Rayonnet U.V. reactor. The samples were mounted in quartz tubes equipped with inlet tubes for SOa, NOX, air or various combinations. Contact angles were measured with an NRL goniometer using distilled water. The solubility/diffusivity data for gases in polymers is obtained using conven- tional weight-gain techniques. A sensitive electrobalance system en- closed in a vacuum chamber was constructed for this purpose. Provisions are made for maintaining constant diffusant pressure and temperature. Mass changes (as small as a few micrograms) are monitored as a function of time, and continuously displayed on a chart recorder. A second electrobalance system that can be used in corrosive environments, to be built in the second year, will greatly speed up the data collection process. An Analect FX-6260 FTIR spectrometer equipped with an MCT detector and a flat-plate ATR sampling accessory was used to obtain the infrared spectra. A ZnSe paral- lelogram crystal with an angle of incidence of 45° was used. All ATR spectra were collected at 4cm-1 resolution, and 128 scans were accum- ulated for each sample. The paint films for FTIR Analysis were cast on a clean glass plate surface, dried overnight in air, and transferred to a desiccator containing anhydrous CaSC>4 for a minimum of three days. Film thickness was measured using a sensitive micrometer having a precision of ± 1.2 nm. The thickness of the films used in this study was 127 urn (about 5 mils). A coating device for putting a thin, uniform film of paint on a cylindrical ATR crystal was constructed for FTIR analysis of diffusion kinetics and in-situ analysis of chemical reactions. The associated gas delivery system for these experi- ments has also been completed. The system will be tested, and should be operational by July 1988. Experimental Results Preliminary Mechanical/ Structural Effects of SO? The effect of uv light (350 nm) in the absence and presence of SOa on the intrinsic viscosity, gel content, contact angle, Tg, and dynamic modulus of the contained polymer in the base latex is shown in Table 1. All the films were from a single casting. Preliminary Solubility and Diffusivity Data for SOa Sorption isotherms at 28°C for SC>2 in the Latex without CaCOs (LO), Latex with CaCC>3 (LC) and polyer base samples have been obtained. The solubility of SOa in the samples varies linearly with pressure up to 650 torr (Henry's law behavior) for the two paint samples and the polymer base. The paint sample containing CaC03 superimposes on the base polymer isotherm, while the sample without CaCC>3 has a slightly higher slope. This difference could be due to an inconsistency in the reported amount of polymer in this paint; a value of 35% was used to normalize the data for both paint samples to grams polymer. Reproduci- bility tests are in progress to check this. Diffusion coefficients (D) were obtained for S02 In each sample at each experimental pressure. D was found to increase with increasing pressure. This trend has been observed before for SOa In a polyimide.This is attributed to plasticization of the polymer by the penetrant. The increase in D is most pronounced for the paint sample without CaC03. | The total flux of SO? through these samples is proportional to the permeability, P, which is equal to the product of the diffusivity D and the solubility ko (Henry's law constant). The pressure-dependent diffusivity trans- lates to a pressure-dependent perme- ability. The permeabilities of the polymer base and the CaCOs-containing paint are again very comparable, and are in the neighborhood of 10'7 cc(STP) cm2/cm3 polymer-torr-sec. The permeabilities of the CaCC>3-free paint are an order of magnitude higher. Simple Immersion Tests in Aqueous SO2 Free films of LC and LO latex paints, as well as the polymer base, were exposed to aqueous SOa (pH = 2.0) for periods ranging from 1 minture to 14 days. The polymer base showed no weight loss after 14 days. However, significant weight loss occurred upon exposure to sulfurous acid for the two latex paint samples. For the latex paint without the CaCOa extender, only 7.2% of the sample weight is lost after 14 days of exposure. For the latex paint with CaCOs extender, the weight loss levels off after 4 hours; the maximum weight 4 loss being 27% after 14 days. The weight loss of the LC samples in deionized distilled water (pH = 5.4, due to the presence of HgCOs) is on'y 8.5% after 14 days. This clearly indicates that the presence of SOa In water accelerated the rate at which materials are leached out of the films. The initial linearity of the weight loss vs. t°-5 plot (curve A, Figure 1) suggests a Fickian diffusion mechanism for removal of material from the sample. The diffusion coefficient is 1.84 x 10'9 cm2/sec. The ATR spectra of the polymer binder and the two latex paints are shown in Figure 2. Due to the multicomponent nature of the paints and the presence of infrared-absorbing inorganics, assign- ment of all the bands is difficult, especially in the region between 400 cnrr1 and 1600 cm'1 where over- lapping of bands is evident. The polymer binder is a terpolymer of vinyl acetate, vinyl chloride, and butyl acrylate, but the relative composition is not known. Despite these complications, differences between the spectra of samples with and without CaCOs extender can be observed, as seen by comparing spectra A and B in Figure 2. The two peaks at ^ 1416 cm'1 and 875 cm'1 are not found ------- 040 - 0.32 - 0.24 - "0 • 0.16 - 008 - 40 60 80 tos(mmf5 100 120 140 Figure 1. Weight loss of latex paint film with exposure time fa) Films with CaCO3 at pH = 2.0, (b) films without CaCOa at pH = 2.0: (c) films with CaCOa in deionized distilled water. 1728 1416 4000 3600 3200 2800 2400 2000 1600 1200 800 400 Wave Number Figure 2. FTIR-A TR spectra of fa) latex with CaCOs, (b) latex without CaCOi (c) base polymer latex. ------- in the spectrum of the LO sample, and are identified as the stretching and wagging vibrations of the C0$2~ group of 003, respectively. The broad band between 850 and 450 cnr1 is absent in the spectrum of the polymer base and therefore has been attributed to the pigment (TiC>2) or the china clay (aluminum silicate) added to the paints. The vibrational frequencies of these additives are located in this region. The band at 1728 cnrr1, due to the C = 0 stretching mode of the polymer, is common to all three spectra. This peak is well separated from others and is suitable for use as a reference for spectral subtraction. In Figure 3, ATR spectra of the LC samples exposed to sulfurous acid for 1, 6, and 30 minutes are shown. The in- tensities of the bands at 1416 cnv1 and 875 cnr1 decrease markedly with increasing exposure time, and are essentially gone after 30 minutes of exposure, indicating a loss of CaCO^. The broad absorbance centered around 700 cnv1 becomes sharper with increasing exposure. This result is also found for the spectra of exposed LO samples. In contrast, the spectra of the polymer base show essentially no change after 14 days of exposure. Subtraction of the spectrum of the polymer base exposed for 14 days from the spectrum of an unexposed sample resulted in a flat line across the entirJ spectral range. Determination of D from the FTIP-ATR Data Difference spectra between the unexposed LC and LO samples revealed the two CaCOa peaks at 875 and 1416 cm~1. The integrated intensities of these two peaks were followed as a function ol exposure time to aqueous SOa- The penetration depth in an ATR experimenl is only a few pm, therefore the CaCOa will be removed from this thin surface layer in a much shorter time than is required to remove all of the CaCO$ from 10 -e 8 •Q 2000 7700 1400 1100 Wave Number 800 450 Figure 3. FTIR-ATR spectra of latex paint containing CaCO3 exposed to aqueous SOz fpH = 2.0) for la) 0 mm; (b) 1 min.; (c) 6 mm., (d) 30 mm ------- the bulk film. The penetration depths are 2.3 and 1.4 urn for the 875 and 1416 cnr1 peaks respectively. The very rapid loss of CaCOa from the surface layer is evident in Figures 4-(a) and (b). The diffusion coefficient cannot be obtained from Figures 4-(a) and (b) in a straightforward manner. Exposure times below 1 minute could not be reproducibly controlled, thus the (linear) short-time regions of the curves could not be obtained. Secondly, there is considerable uncertainty in the assign- ment of the effective film thickness ;, given that the amplitude of the evanescent wave in an ATR experiment decays exponentially with depth into the sample, and that the concentration profile of the diffusant remaining in the surface layer are nonlinear. The results of this analysis for the data in Figure 4 are D = 1.77 x 10'9 cm2/sec for the 1416 cm'1 peak, and 2.25 x 10-9 cm2/sec for the 875 cm-1 peak. Both are in very good agreement with the value of 1.84 x 10-9 cm2/sec obtained from the bulk weight loss data. The solid lines in Figure 4 were obtained by fixing these values for D and back- calculating the expected values of It/loo. Discussion Mechanism of UV/SO2 Effects The data in Table 1 demonstrate that exposure to uv causes both chain 12 t°s(min)°* (b) Figure 4. Data points: removal of CaC03 from LC paint films exposed to sulfurous acid as measured by the integrated intensity ratio of la) the CaCOa band at 875 cm'\- (b) the CaCOs band at 1416 cm'\ Solid lines: calculated variation of the integrated intensity ratio with f°'5. ------- Table 1. Effect of U.V. and/or SOg on Base Latex Properties Sample Reference No. 1-75-BL 1-77-BL1 1-78-BL2 1-79-BL3 1-80-BL4 1-81-BL5 1-81-BL6 Exposure Time (Hr) SC>2 UV (350nm) (ml/min) 0 48 96 48 96 0 0 0 0 0 2 2 2 (48 hr) 2 (96 hr) In] (dl/gm) 1.00 0.82 0.50 0.17 0.11 1.00 1.00 Gel Content (%) 22 32 60 66 68 22 22 Contact Angle (6) 78 71 65 52 * 63 67 Tg E1X1010 dynes/cm2 (°C) (at -30"C) 22 22 24 35 33 34 34 — 83 97 100 80 35 25 24 24 2.50 2.50 2.16 2.72 2.40 1.90 1.90 1.24 1.27 4.54 2.29 1.49 1.64 2.44 1.25 2.07 "Water drop spread immediately scission and crosslinking of the polymer. The intrinsic viscosity decreases and the gel content increases with uv exposure time. The higher molecular weight species will be preferentially converted to gel (crosslinked polymer) increasing the fraction of low molecular weight species in the soluble portion and resulting in a decrease in intrinsic viscosity. However, the decrease in contact angle indicates an increase in the carbonyl group content which must result from chain scission. The small increase in Tg confirms crosslinking has occurred. Exposure to SC>2 alone has relatively little effect. However, the combination of uv and SOa results in rapid polymer microstructural changes. The intrinsic viscosity decreases by nearly an order of magnitude, the gel content triples and the Tg increases markedly. For example, the Tg increases from approximately 22 °C for the unexposed base latex to about 75-80 °C for samples exposed for 48 hrs to both uv and SC>2 The polarity of the surface also significantly increases. This indicates that uv and SC>2 in combination cause rapid microstruc- tural polymer changes, namely chain scission and crosslinking. Mechanism of CaCOa Removal The bulk weight loss data and FTIR spectra clearly suggest that the CaCOa extender added to the latex paint is removed when the film is exposed to solutions containing acidic ions. Exposure to deionized distilled water, in which the acidic species come from HaCOa, rsults in loss of less than 10% of the sample weight in 14 days. However, upon exposure to pH = 2.0 sulfurous acid, essentially all of the CaCOj is leached out in about 4 hours. The approximate composition (in weight %) of the dry LC latex samples is 37% polymer base, 21% CaC03, 35% TiOa, and 7% china clay. The maximum weight loss is 27%, which is nearly equal to the sum of the weights of CaCOa and china clay. The changes in the infrared spectra also support the loss of these two components. Although the sharpening of the broad band between 450 and 850 cm~1 could be attributed to loss of TiOg rather than china clay, the extremely low solubility of TiC>2 in acids, relative to china clay makes this rather unlikely. No detectable evidence for sulfite or bisulfite ion was found in the FTIR spectra of exposed films. The mechanism of removal of CaCOa from the latex film must involve three steps: (1) diffusion of the components of the SOa aqueous solution into the film, (2) reaction and/or dissolution of CaCOa, and (3) diffusion of CaCO^ (or its ions) out of the water-swollen film. Given the Fickian kinetics for the overall removal of material from the film, the rate limiting step must be diffusion controlled. Therefore, if any of the above chemical reactions do occur, they must be very rapid in comparison. It is not possible to distinguish between steps (1) and (3) with the present data, although one of these processes should be much fastei than the other, otherwise, Fickian kineticj would not be observed. Thus, the diffusion coefficient measured in these experiments represents either step (1) oi step (3) in the mechanism proposec above, but not a combination of the two It would seem that diffusion of CaCOa (oj its ions) out of a water-swollen filnr could occur much faster than the initia diffusion of water, SOa, and ions into £ dry, unswollen film. Summary During the first year of this project preliminary studies were conducted tc determine the best techniques to identify chemical and physical effects or polymeric coatings used on wood due tc wet and dry acid deposition. The following techniques were employed Fourier transform IR, gravimetric measurements to obtain diffusivities anc solubilities, dynamic mechanical analysis (DMA), sol-gel analyses, intrinsic viscosities (IV), differential scanninc calorimetry (DSC), contact angle anc electron spin resonance (ESR). Emphasis during the initial phase of ou investigations has been on free films o latex paint. Some preliminary work hai begun on alkyd paint films and coating; on wood substrates. Preliminary exposure studies show tha the techniques employed are well suitei to follow changes in the microstructure c latex films as a function of exposure ti acid deposition. A ------- C. M. Batik, R. E. Fornes, and R. D. Gilbert are with North Caromia oia.c University, Raleigh, NC 27695. J. W. Spence is the EPA Project Officer (see below). The complete report, entitled "Analytical Techniques for Measuring the Effects of Acid Deposition on Coatings on Wood," (Order No. PB 89-127 518/AS; Cost: $15.95, subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Atmospheric Research and Exposure Assessment Laboratory U.S. Environmental Protection Agency Research Triangle Park, NC 27711 United States Environmental Protection Agency Center for Environmental Research Information Cincinnati OH 45268 Official Business Penalty for Private Use $300 EPA/600/S3-88/044 000032V PS U S ENVIR PROTECTION AGENCY REGION 5 LIBRARY „ 230 S D£A*80K*i STREET CHICAGO II- 60604 ------- |