An investigation of the effect of pigment on the degradation of a naturally weathered polyester coating.
Keywords Polyester-melamine coating, Pigment, Weathering, Degradation
Organic coatings have been used and developed for many centuries to improve the appearance of and/or to protect the substrate materials. Coil coating, also known as "pre-paint," is a highly automated and continuous process for coating long strips of thin metal prior to fabrication. The polymeric paints are applied to unwound coils of steel or aluminum in a range of colors, cured in seconds, and re-coiled for distribution and use. (1)
Coil coatings are employed in many outdoor applications, including industrial and domestic cladding and automotive parts. Such applications often mean the coil coating is exposed to aggressive environmental conditions, including large temperature variations, strong sunlight, humidity, and atmospheric pollution. Such factors may affect both the appearance and performance of a coating over time. Through a process of progressive improvement, today's coil coatings have much improved exterior durability; however, this does not always ensure that these coating systems achieve the required performance, particularly in regions with high humidity, intense UV, or frequent acid rain.
When exposed to the environment, organic coatings are subject to photo-oxidation, (2), (3) and in some instances moisture enhanced photo-oxidation mechanisms are believed to account for the increased degradation rate observed when coatings are used in high humidity environments. (4), (5) More recently, acid rain has become a prominent problem as it often results in etching effects on coil coatings. Mori et al. found that crosslinkers in acrylic-melamine systems degrade very rapidly in the acidic solution present in acid rain. (6) Many coil coating formulations contain high concentrations of pigments; typically around 10-40% by weight of the cured coating is pigment, and some pastel shades contain even more. Therefore, it is reasonable that the photo-oxidation and environmentally induced degradation, such as acid rain attacks, can also cause the decomposition of pigments. This usually results in the surface morphology and color changes. It is generally accepted within the paint industry that pigments such as titanium dioxide are able to catalyze the degradation of binder resins due to their photocatalytic properties. (7-9) Even when stabilized by encapsulation, the average particle is never 100% coated, which means even the best grade of titanium dioxide pigment has the capacity for photodegradation.
It is important commercially to investigate the durability of coil coating formulations, both to improve the coating performance and to predict the coating's long-term service lifetime before the coating is used in an aggressive external environment. Many analytical techniques have been used to characterize coating durability. Color and gloss measurements have long been used as industrial quality control methods. Scanning electron microscopy (SEM) and atomic force microscopy (AFM) are widely used to analyze surface morphologies, (10), (11) while X-ray photo-electron spectroscopy (XPS) (12), (13) and time-of-flight secondary iron mass spectrometry (ToF-SIMS) (14) have been used to investigate the surface chemistry of coatings, before and after exposure to the environment. Photo-acoustic (PA) Fourier transform infrared spectroscopy (FTIR) with step-scan capacity, (15), (16) on the other hand, provides excellent bulk chemistry analysis with a sampling depth ranging from 6 to 20 [micro]m. Utilizing these and other techniques, much research has been carried out with the aim of developing more fully an understanding of organic coating degradation mechanisms, when such coatings are exposed to different environmental conditions.
Although accelerated weathering procedures and instrumentation such as QUV-A, QUV-B and EMM-AQUA have been developed as aids with which to study coating durability, results obtained from such methods are difficult to correlate with the performance data obtained when coil coatings are exposed to the natural environment. The aim of this study was to investigate why the durability of a polyester-melamine-based coil coating formulations had performed so badly at one exposure site when previous experience and other tests had indicated otherwise. Different pigments (copper phthalocyanine blue, lead chromatc yellow, titanium dioxide white, and iron oxide red) had been added to the formulations to achieve specific colors. The resulting coatings were exposed to the environment in Hainan, PRC, for a duration of 2 years. XPS and PAS-FTIR were used to analyze the surface and bulk chemistry of the coatings, respectively. Coating surface morphology changes were investigated by gloss and color measurements, as well as SEM and AFM. Along with the investigation of the general degradation of the coatings, the effects of pigmentation on coating durability, surface chemistry, and morphology were also investigated.
Materials and sample preparation
The coil coating samples investigated in this work were all laboratory-prepared. The coil coating formulation employed was based on a cyclo-aliphatic polyester and crosslinked with melamine in a ratio of 85:15. A mixture of hexamethoxymethylmelamine (HMMM) and a higher amino ratio methoxymethylmelamine (MMM) crosslinking agents was used in a ratio of 5:3. Lead chromate, iron oxide, and copper phthalocyanine blue pigments were included in the formulations to prepare yellow, red, and blue coatings, respectively. Titanium dioxide pigment was used in all of the coating formulations to provide opacity. The size of the pigment particles was checked using a Hegman Gauge after beadmill grinding, so as to ensure the pigment particle size was smaller than 5 [micro]m in the mill base. The coating sample names and associated pigmentations are listed in Table 1; all formulations were prepared with a pigment volume concentration (PVC) well below the critical PVC (CPVC).
Table 1: Coil coating sample formulations Sample Color Pigment weight (%) Overall PVC CPVC CY Yellow Lead chromate (44%) 15.8 37.2 Titanium dioxide (3%) CR Red Iron oxide (23%) 7.4 40.3 Titanium dioxide (1%) CB Blue Copper phthalocyanine (10%) 10.5 47.3 Titanium dioxide (12%)
All coatings were cast as liquid films, on a polyester-melamine-primed aluminum substrate, using a wire-wound draw-down bar of a diameter suitable to obtain a film thickness of ~20 [micro]m. The coatings were cured for 30 s, reaching a peak metal temperature of 232[degrees]C in an electric oven with efficient air flow. Discs ~8 mm in diameter were punched from the coated panel for analysis.
Weathering of coil coatings
The samples were exposed to the environment at a 45[degrees] angle to the horizontal and facing south towards the sun in Hainan province, southeastern China, for 2 years, and an average taken of the results from each panel. Subsequent exposure of panels of similar color in this site has shown similar durability. The site chosen for sample exposure is subject to high temperatures, high humidity, and a high solar UV dosage. A summary of the general weather conditions experienced in Hainan are provided in Table 2. For analysis, one-half of the coated panel after weathering was cleaned by rinsing the coated panel with distilled water, followed by swabbing the coated panel with cotton wool to remove any loosely adhered particles (washed coil coatings). The untouched half of the weathered coated panel was analyzed as received (unwashed coil coatings). Panels were inspected after both 1- and 2-year exposures; because 1-year exposure samples showed similar trends (but less degradation) compared to 2-year samples, 2-year exposure results were selected and are shown in this paper.
Table 2: Climatic data for Hainan, PRC (30) Average annual high temperature ([degrees]C) 37 Average annual low temperature ([degrees]C) 11 Average annual relative humidity (%) 81 Total UV radiation (MJ [m.sup.-2]) at 26[degrees]S tilt 306 Annual rainfall (mm [year.sup.-1]) 2013 Distance from ocean (m) 300
A set of yellow and white coatings with the same PVC of CY and CB coatings, respectively, were also exposed in the QUV-A chamber (The Q-Panel Company, USA) equipped with QUV-A-340 nm fluorescent tubes, each with peak irradiance at 340 nm. The sample panel was irradiated for 8 h in each 12-h cycle at 60[degrees]C. During the dark period (4 h) the panel was subjected to a moist environment at a temperature of 50[degrees]C so that water condensation appeared on the panels.
X-ray photoelectron spectroscopy
XPS analyses were performed on a modified Thermo VG Scientific (East Grinstead, UK) ESCALAB Mk II spectrometer. The instrument was equipped with an XR4 twin anode X-ray source (Al K[alpha]/Mg K[alpha]) and an Alpha 110 electron energy analyzer. In this work the Al K[alpha] X-ray source (hv = 1486.6 eV) was used at 300 W (15 kV x 20 mA). The pass energy employed for all survey spectra acquisition was 100 eV. The pass energy was set at 20 eV for the acquisition of the C 1s, O 1s, and N 1s high resolution spectra. For other elements (invariably at low concentration in the surface analysis), a pass energy of 50 eV was used to acquire high-resolution spectra. Sample mounting for XPS analysis was achieved by fixing a specimen to a VG sample stub using double-sided adhesive tape. Quantitative surface chemical analyses were calculated from the high resolution core level spectra following the removal of a nonlinear (Shirley) background. The manufacturer's Avantage software was used, which incorporates the appropriate sensitivity factors and corrects for the electron energy analyzer transmission function.
Photoacoustic (PA) FTIR
PA-FTIR spectra were recorded on a Nicolet 8700 spectrometer (Madison, USA) with an MTEC model 200 PA accessory (Ames, USA), in the mid-IR region (400-4000 [cm.sup.-1]) with a resolution of 8 [cm.sup.-1]. The PA cell was purged with dry helium at 20 [cm.sup.3] [s.sup.-1] for 5 min before acquiring a spectrum.
The step-scan PA-FTIR (SSPA-FT1R) experiments were performed by varying the modulation frequency of the IR beam at a constant amplitude of 3.5[[lambda].sub.HeNe]. Spectra were collected at the mid-IR range (400-4000 [cm.sup.-1]) with a resolution of 8 cm"1. The correlation between the IR modulation frequency and sampling depth is provided in Table 3; this is based on taking the thermal diffusivity of the coatings as 0.01 x [10.sup.-5] [m.sup.2] [s.sup.-1] (a general value for most polymeric materials; a more detailed sampling depth calculation can be found elsewhere (16)). A carbon black reference material was used as a strong surface absorber to calibrate the instrument. The in-phase (I) spectra were used as a direct indication of the chemical composition within the corresponding sampling depths. All SSPA-FTIR spectral data processing was accomplished using Omnic 7.3 software.
Table 3: IR sampling depth at different modulation frequencies Modulation frequency (Hz) IR sampling depth ([mu]m) 50 25 100 18 200 13 400 9 600 7 800 6 1000 5.6
Atomic force microscopy
Surface topography data was obtained using an NT-MDT NTEGRA-Spectra AFM (Moscow, Russia). The AFM was operated in the semi-contact mode using NT-MDT NSG03 cantilevers (Moscow, Russia). AFM images of 100 x 100 [micro][m.sup.2], 50 x 50 [micro][m.sup.2], 25 x 25 [micro][m.sup.2], and 15 x 15 [micro][m.sup.2] were acquired from each sample. The surface size distribution and image processing was performed using the supplier's Nova software.
Scanning electron microscopy
Coating surface morphology was also investigated using the Hitachi S-3200 SEM (Tokyo, Japan) with an accelerating voltage of 25 keV. Samples were sputter coated with 6 nm thick gold to prevent electrostatic charging.
Gloss retention and color measurements
Gloss measurements were investigated with a sheen 60[degrees] gloss meter calibrated with a glossy black panel with a reflection of 93.4. Color change data was obtained using the Datacolour Spectraflash 300 fitted with an integrating sphere (Cheshire, UK).
Results and discussion
Surface morphology characterization
Figure 1 shows the morphology of the unweathered (CB), weathered/unwashed (CB-UW), and weathered/washed (CB-W) CB coating surfaces. It can be clearly observed that the unweathered CB surface is relatively smooth but not entirely flat; several discrete lumps can also be clearly seen. After 2 years' weathering, the CB coating surface becomes rough. Some large particles were deposited on the coating surface. Holes observed show erosion of the coating surface during the weathering process. In order to clean loosely bonded dirt from the coating surface, half of the weather panels were cleaned using distilled water. It is clearly observed from the washed CB coating that its surface is even rougher, with hollow regions all over. Small titanium dioxide particles are clearly observed in the hollows.
[FIGURE 1 OMITTED]
The unweathered (CR/CY), weathered/unwashed (CR-UW/CY-UW), and weathered/washed (CR-W/CY-W) CR and CY coating surface morphologies are shown in Figs. 2 and 3, respectively. Generally, the CR coating surface is rougher than the CB coating surface. The CY coating surface is even rougher, with small protrusions all over the surface. Moreover, the CR and CY surfaces both show the raised features (lumps) that are similar to those observed on the CB coating surface. As the same binder resin was incorporated in the CB, CR, and CY coatings, and as they are all formulated well below their CPVC (Table 1), the rougher surface of CR/CY would appear to be due to the pigment particle size. The copper phthalocyanine blue pigment has an average particle size of less than 0.1 [micro]m, (17) which is below the resolution of the SEM images shown in Fig. 1; thus, it is considered to have no effect on the coating surface roughness. The CR coating shows a rougher surface than the CB coating, because the size of a few pigment particles is sufficiently large to disrupt the surface. The CY coating surface is even rougher, as it has an even higher concentration of large pigment particles.
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
It is known that different pigments with different hydrophilic/hydrophobic properties will disperse/stabilize differently in the binder system. As CY, CR, and the majority of CB coatings are based on inorganic pigments, their dispersing ability is quite similar (hydrophilic nature). Regarding the blue pigment (organic, hydrophobic), the surfactants and polyester binder are all actually very good at dispersing and stabilizing organic pigments. As all three coatings are formulated well below CPVC, the different surface features observed are primarily due to the pigment particle size (e.g., lead chromate > iron oxide > titanium dioxide > phthalocyanine blue), as well as their concentration.
It is observed from Fig. 2 (CR-UW) and Fig. 3 (CY-UW) that the surface features for the naturally weathered CR and CY coatings are quite similar to those observed from the CB-UW surface.
The holes observed on the CR-UW and CY-UW surfaces suggest that environmental erosion has occurred on their surfaces during the weathering. Moreover, their surfaces are both much rougher than the unweathered CR and CY coating surfaces. This may be due to the pigment exposure, dirt pick-up, and binder resin degradation during the weathering (this will be confirmed in a later discussion).
The surface features of the CR-W and CY-W coatings are quite similar but very different from the CB-W surface; the latter becomes much "cleaner" compared with the CB-UW surface. Hence, the washing process seems to be the most effective at removing loose dirt from the CB-UW surface. This may be due to the titanium dioxide chalking effect, which makes the bonding between dirt and coating weaker. This is not unusual, because titanium dioxide particles are widely used in the self-cleaning coating system.18 It is known from Table 1 that the CB coating contains the highest titanium dioxide content. This may explain why only the CB coating shows significant changes after the washing process (this will be confirmed in a later discussion). Therefore, the pigment seems to play an important role in the type of degradation taking place in exposed polyester-melamine coil coatings.
Furthermore, the large holes observed only on the CB-W coating surface may be due to the removal of melamine-enriched zones, which can be even larger than 5 [micro]m in diameter. As observed previously, the melamine-enriched regions in the white polyester-melamine coil coating seem to contain no titanium dioxide pigment. (19) Therefore, the resinous features could also be the melamine-enriched regions that were not as enriched as zones that phase separated, and are thus a more integral part of the coating. Further work will be undertaken to investigate this.
The AFM analyses of the unweathered coatings are shown in Fig. 4; an AFM tapping mode image and a surface size distribution chart (i.e., the size and distribution of particles on the coating surfaces) are both included. It is generally found that the AFM images correlate with the previous SEM results quite well. The main particle size on the CB coating surface as determined by AFM is ~1.2 [micro]m, which is mainly due to the raised features observed on the CB coating surface. Smaller particles with size ~0.4 [micro]m may be mainly attributed to agglomerated titanium dioxide pigment particles. Similarly, the CR and CY surfaces both show significant numbers of particles with sizes larger than ~1 [micro]m; this is again mainly due to the raised features. However, both the CR and CY surfaces show significant increments of particles with size smaller than 0.5 [micro]m; this is mainly due to the pigment particles. The fact that CR and CY coatings both exhibit this effect indicates that pigments with relatively large particle sizes are close to the surface, causing some degree of roughness. Furthermore, the higher pigmentation of CR and CY coatings may also contribute to the surface roughness even when the PVC is well below the CPVC. Therefore, the AFM results also confirm that the coating surface morphology is highly affected by the pigment particle sizes. The same raised features (as observed from SEM images) with size larger than 1 [micro]m are certainly not due to the pigmentation in the formulation. Again, as the coil coating formulation contains several different types of resins, additives, and pigments, the observation of raised features on the coating surface may be associated with matting agents or even melamine enrichment. (19)
[FIGURE 4 OMITTED]
The AFM results for the weathered and washed coatings are shown in Fig. 5. The CB-W surface again demonstrates rougher surface features (more small particles) when compared to the CR-W and CY-W coatings. Small particles can also be clearly seen from the AFM image of the CB-W coating. The surface size distribution chart suggests that the size of particles on the CB-W coating surface is ~0.3-l [micro]m. Therefore, it is quite possible that the small particles seen previously on CB-W surfaces (Fig. 1) are titanium dioxide pigment particles. The CR-W coating surface is smoother than the CB-W coating surface; the main particles on the CR-W coating surface are larger than 1 [micro]m. The surface roughness of the CY-W coating is between those of the CB-W and CR-W coatings. The main particles on the CY-W coating surface have sizes between ~0.5 and 1.5 [micro]m. The size of pigment particles is usually less than 1 [micro]m. They may possess a slightly larger size after weathering exposure, as dirt may cover the pigment particles. Therefore, the CB-W surface shows a higher quantity of smaller particles, as titanium dioxide particles are highly exposed. The CY-W coating also shows a considerable number of smaller particles, as it has the highest pigment volume concentration of the three coatings investigated. Thus, the AFM results are found to be quite consistent with the previous SEM results. With the particle size distribution feature of AFM (quantitative analysis), the effect of pigment on coating surface morphology before and after weathering is clearly revealed.
[FIGURE 5 OMITTED]
The color and gloss measurements of the three coatings before and after weathering are shown in Figs. 6a and 6b, respectively. It can be observed that the CY coating shows the most significant change in color after weathering, while the CR coating shows the least change in color. All washed surfaces show some recovery in color; this is mainly due to the removal of loose dirt attached to the coating surface. It is quite interesting to observe that the CB-W coating shows the most recovery in color after washing (over 80%). This observation actually correlates well with the SEM/ AFM results, suggesting that the washing process has removed significantly more dirt from the weathered CB coating surface than from the other two. This is again quite possibly due to titanium dioxide photo-activity that weakens the bond between dirt and resin (i.e., the chalking effect).
[FIGURE 6 OMITTED]
The gloss measurements are summarized in Fig. 6b. It can be generally found that CB coating retains the least gloss after weathering. All three coatings show recovery of gloss after the washing process; this is again due to removal of dirt on the surface. This is in line with the observation from SEM/AFM suggesting that the CB coating surface is generally rougher than the other two after weathering. It also indicates that the titanium dioxide chalking effect may be responsible for the significant changes observed in the CB coating surface after weathering.
Surface chemistry characterization
The elemental surface compositions for the CB, CR, and CY unweathered coatings as determined by XPS are summarized in Table 4. They are all quite similar and well correlated with previous findings. (20) All of the coating surfaces are primarily composed of C (78 [+ or -] 2 at%) and O (18 [+ or -] 2 at%), which are assigned to the presence of the polyester resins and flow aids. N (1.3 [+ or -] 0.1 at%) arises due to the melamine crosslinking resin (the only source of N in the formulation). The N level is relatively low, as the majority of the true coating surface is usually covered by a thin layer of flow aid (20) that has very low surface energy and is usually included in the formulation to help the wet paint flow on the substrate. The observation of small quantities of Na, Ca, and Si are due to the inclusion of minor additives in the coating formulation and/or the adventitious surface contamination during the sample handling process. No elements indicative of the pigments are observed, except in the case of the CY coating (Pb, 0.1 at%). The absence of chrome can be explained by the difference in the XPS cross-section of the two elements. As lead has a much greater cross-section than chrome, at such low concentration levels chrome may be present at a concentration below the detection limit. Because the SEM and AFM analyses shown in Figs. 3 and 4 suggest that the CY yellow coating surface is much rougher than the rest, the presence of Pb originating from the lead chromate yellow pigment is very likely due to the exposure of the pigment particles. Because the CR and CB coatings have lower PVC than the CY coating, and the particle sizes are smaller, their surfaces are smoother, with no pigment signal obtained by the XPS surface chemistry analysis.
Table 4: Elemental surface composition (at%) of unweathered coil coatings Sample C O N Na Ca Si Pb CB 79.0 16.9 1.2 0.2 0.8 1.9 - CR 76.9 19.2 1.4 0.3 0.4 1.9 - CY 77.2 19.9 1.3 - 0.3 1.2 0.1
The elemental surface compositions for the weathered and unwashed (UW) CB, CR, and CY coatings are summarized in Table 5. Each of the coating surfaces is also primarily composed of C (35 [+ or -] 2 at%) and O (48 [+ or -] 2 at%). However, compared to the unweathered coating surfaces (Table 4), the C content drops and the O content increases dramatically. This is due to the degradation of coating during the weathering process, because polyester resin generally undergoes both photo-oxidation and hydrolysis when exposed to the environment with strong sunlight and high humidity. The C content drops because most of the polyester on the coating surface is oxidized and hydrolyzed; this also results in significant increases of the O content (note that the O concentration
is actually contributed by a combination of dirt, pigment, resin, and surface oxidation; although O is not directly related to surface oxidation, it can still indicate surface chemical changes when comparing only the three coating surfaces). Moreover, the dirt accumulated on the coating surface (as also observed from SEM images) suggests that the decreasing C content could well be partially due to the dirt that masks the true coating surface. The observation of the increasing Si and Al contents also confirms that the coating surface is covered by dirt, usually clay (aluminum silicates) originating from the natural environment. The N content in the three coatings all increases because the thin flow aid layers were removed during the weathering process; the melamine crosslinking resin was exposed to the atmosphere.
Table 5: Elemental surface composition of weathered and unwashed coil coatings Sample C O N Na Ca Si Al CB-UW 37.5 46.0 3.7 1.1 0.1 5.8 2.6 CR-UW 32.6 49.7 2.9 1.1 0.1 7.1 3.7 CY-UW 32.9 50.3 1.8 0.4 0.2 7.7 3.9 Sample Mg Ti Fe Pb Cr Cu CB-UW 0.3 2.4 0.4 - - 0.2 CR-UW 0.7 0.4 1.7 - - - CY-UW 0.4 0.6 0.8 0.1 0.9 -
Traces of different pigments on the coating surfaces are also found. Ti is found on all three coating surfaces because titanium dioxide pigment is incorporated in all formulations. The CB-UW coating surface shows the highest Ti concentration because of its high titanium dioxide pigmentation in the formulation. Fe is also found on all three coating surfaces because dirt contains a low concentration of Fe. The CR-UW shows more Fe than the other two coatings because only the CR coating has iron oxide pigment in the formulation. Pb and Cr are found only on the CY-UW coating surface; this is because of the lead chromate pigment in the CY coating formulation. For the same reason, Cu is found only on the CB-UW coating surface. The observation of small amounts of Na, Ca, and Mg is because of the dirt pick-up during the weathering process. Therefore, the XPS results suggest that the coating surface was degraded during the natural weathering process; the pigment particles were then exposed to the atmosphere. A certain amount of dirt was also accumulated on the coating surface. These are the facts that led to the loss of color and gloss of the coating during weathering exposure over time.
Another observation from Table 5 is that the CB-UW coating surface has the least Si, Al, Mg, and Fe. This indicates that the least dirt was picked up by the CB-UW coating compared to the other two coatings. Because the CB-UW coating surface has more titanium dioxide particles, the loss of surface layers to which the dirt particles are attached through chalking may lead to less dirt being found on the CB-UW coating surface.
The surface chemistry of the weathered and washed coatings is shown in Table 6. It is found that the C and N contents increase by roughly 10-20 and ~1 at%; the O content decreases by ~10 at% after the washing process. The Si, Al, Mg, and Fe contents all decrease because loosely bonded dirt was removed from the coating surface. There are still traces of pigment (e.g., Pb, Cr, Ti, Fe, and Cu) on the coating surface; their surface compositions either increase (removal of dirt from the surface exposing more pigment) or decrease (removal of pigment from the surface). However, it is also observed that the Ti, O, and dirt (Si, Al) contents in the CB-W coating decrease the most (by 1, 14, and 3.9 at%, respectively); while the C and N contents increase the most (by 19 and 1.3 at%, respectively) among the three coatings. This indicates that the washing process is the most effective for removing dirt from the weathered CB coating surface. This can be again explained by the chalking effect of the titanium dioxide pigment.
Table 6: Elemental surface composition of weathered and washed coil coatings Sample C O N Na Ca Si Al CB-W 56.2 32.0 5.0 0.5 0.3 1.9 2.0 CR-W 42.4 42.1 3.7 0.5 0.3 6.2 2.7 CY-W 43.5 44.4 2.0 0.3 0.2 4.9 2.3 Sample Mg Ti Fe Pb Cr Cu CB-W 0.3 1.4 0.1 - - 0.2 CR-W 0.3 0.3 1.0 - - - CY-W 0.3 0.4 0.6 0.1 1.1 -
It is known that the titanium dioxide particles can cause photodegradation even when stabilized21 because the inert coating is never 100% complete; free radicals can be promoted when the titanium dioxide particles are in contact with UV light and moisture. Although the titanium dioxide used here is a rutile type with [SiO.sub.2] and [Al.sub.2][O.sub.3] coatings, the high photo-activity of the exposed titanium dioxide core will still cause severe degradation. The exposure site (Hainan, PRC) has a subtropical environment and is very close to the sea; hence the UV and moisture levels are both considered to be very high. In other words, this exposure site is especially aggressive. Therefore, it is reasonable that the radicals promoted by titanium dioxide cause more degradation in the CB coating (it contains the highest titanium dioxide concentration), especially in the regions that are in direct contact with the titanium dioxide pigment particles. The highly degraded binder can be more easily removed from the coating surface, which results in the observation of titanium dioxide pigment particles.
Because all three coatings contain more than one type of pigment, two panels of yellow and white coatings were made with only lead chromate and titanium dioxide pigment, respectively, and exposed in a QUV-A cabinet for 3000 h. The gloss and PA-FTIR data (note: the melamine degradation measured by PA-FTIR was calculated using equation (1) below) are shown in Fig. 7. It can be clearly seen that the white coating generally shows more surface (lower gloss) and bulk (higher MELDeg%) degradation after QUV-A exposure. Thus, the previous hypothesis about titanium dioxide pigment photo-activity is proved.
[FIGURE 7 OMITTED]
MELDeg% = [[Ratio.sub.MEL] (STD) - [Radio.sub.MEL] (EXP)/[Radio.sub.MEL]] (1)
where MELDeg% is the percentage of melamine loss in the coating after exposure; [Ratio.sub.MEI]. (STD) is the melamine (1550 [cm.sup.-1]) to hydrocarbon band (1450 [cm.sup.-1]; as this band is quite stable upon weathering, it is selected as the spectral internal reference) height ratio of the unexposed coating; and [Ratio.sub.MEL] (EXP) is the melamine to hydrocarbon band height ratio of exposed coating. A typical infrared spectrum can be found in Fig. 8.
[FIGURE 8 OMITTED]
Bulk chemistry characterization
The 800 Hz SSPA-FTIR spectra (sampling depth = 6 [micro]m) of CY, CR, and CB coatings before and after exposure are shown in Fig. 8; the general band assignment can be found in Table 7. It is generally observed that the coatings undergo oxidation and hydrolysis upon weathering. The increase in the hydroxyl group band region (3700-2500 [cm.sup.-1]) indicates the hydrolysis and/or oxidation of polyester-melamine linkages, as well as polyester molecules. The oxidation is often attributed solely to photo-oxidation. The broadening of the carbonyl band (1730 [cm.sup.-1]) indicates the photo-oxidation of the polyester backbone. The drop of the melamine band (1550 [cm.sup.-1]) suggests that the mela-mine crosslinker degraded during the weathering process. These findings are also well correlated with the previous findings. (3-5) However, the CY-W sample shows the highest hydroxyl and carbonyl band intensity and the lowest melamine band intensity among the three coatings. This indicates that the CY coating degraded more than the CB and CR coatings during the weathering.
Table 7: General PA-FTIR band assignment Band position ([cm.sup.-1]) Assignment 3700-2500 Hydroxyl group (-OH) vibration 1730 Carbonyl group (-C = O) vibration 1550 Melamine ring and side group vibration
The degradation depth profiling of CY, CR, and CB coatings were then quantified using equations (1) and (2); the results are shown in Fig. 9. It needs to be pointed out that the signal collected at each depth comes from the entire thermal diffusion depth (i.e., sampling depth).
[FIGURE 9 OMITTED]
OH% = [[Radio.sub.OH] (EXP) - [Radio.sub.OH] (STD)/[Radio.sub.OH] (STD)] (2)
where OH% is the increase of the hydroxyl group in the coating after weathering; [Ratio.sub.OH] (EXP) is the ratio of hydroxyl group (3700-2500 [cm.sup.-1]) to hydrocarbon band (3000-2800 [cm.sup.-1]--because this band is quite stable upon weathering, it is selected as the spectral internal reference) area in the exposed coatings; [Ratio.sub.OH](STD) is the ratio of the hydroxyl and hydrocarbon band areas in the unexposed coating. Similar coating degradation quantification methods can also be found elsewhere. (22), (23)
It is generally observed that the overall degradation trend follows that observed in Fig. 8, which is that the CY coating degrades more than the other two. This is believed to be because of the lead chromate pigment that accelerates coating degradation during weathering. However, all coatings show decreasing degradation trends in deeper layers. This is believed to be because of both the UV and moisture penetration in the coating. As the coating surface tends to have the higher concentration of moisture and directly interacts with UV radiation, degradation is more significant on the surface. It is also observed from Fig. 9b that the CB coating shows more difference between surface (~4%) and deeper layer (~2%) melamine degradation than the other two; this is consistent with previous findings indicating that CB coating undergoes more surface degradation due to titanium dioxide photo-activity that accelerates binder degradation. However, as titanium dioxide also has excellent opacity, deeper layers are actually well protected against weathering.
The result observed in Fig. 9 (i.e., lead chromate causes more degradation than titanium dioxide) is also quite different from that observed in Fig. 7 (titanium dioxide causes more degradation than lead chromate). This inconsistency may arise from the nature of the two weathering methods. The QUV-A cabinet condition only simulates UV and moisture that are in the atmosphere, thus ignoring other aspects such as acid rain, long-term oxidation, etc. Hence, the catalytic property of lead chromate may not be well activated by the QUV-A weathering condition.
It is known that the chrome containing catalyst Cr-MCM-41 is usually used to oxidize cyclohexane because of the strong oxidizing property of Cr(VI). (24), (25) Moreover, the Cr(VI) ion is capable of performing a variety of oxidizing reactions in the natural environment. (26) The Cr(VI) may be reduced to Cr(V), Cr(IV), and Cr(III) after the reaction; and it has been proved that Cr(III) is also an effective catalyst in the oxidizing reaction of polymers. (27) It is then highly possible that the high pigmentation of lead chromate (~40% by weight) in the CY coating formulation leads to the accelerated oxidation of the binder resin in the coating during the long-term natural weathering process. According to Table 2, the sample exposure site (Hainan, PRC) has a condition of high humidity and intense sunlight; rainfall is also quite frequent. With the increasing global air pollution, the rainfall in Hainan is also acidic. It was previously found that the annual average acid rain frequency in Hainan is around 24.5% with pH of approximately 5. (28) It is also known that the solubility of chromate ions increases with the solution acidity (29); hence the acid rain may also lead to more chromate ions being dissolved. Thus the degradation of lead chromate pigment leads to significant change in color (Fig. 6a) as well as degradation of binder in the coating. Moreover, the degradation of lead chromate pigment may also enhance moisture permeation during weathering, which leads to more degradation than with the CB/CR coatings, even in deeper layers. Although the CB-W coating shows the most superficial chemical and physical changes, the SSPA-FT1R bulk chemical analysis shows that the CY-W coating actually undergoes the most severe degradation among the three coatings. The CR-W coating shows similar surface features with the CY-W coating, but the least degradation among the three coatings. This is because iron oxide pigment is more stable during weathering than lead chromate and titanium dioxide.
Both surface and bulk changes of polyester-melamine coil coatings that contain three different pigmentations exposed in Hainan, PRC, for 2 years were successfully analyzed using various techniques. The combination of color/gloss measurements; AFM/SEM, XPS, and PA-FTIR analyses provides a clearer view on how polymeric coatings degrade in natural environments.
It was primarily found from the surface morphological (SEM and AFM) analyses, as well as the color/ gloss measurements, that the pigmentation has a very significant effect on the coating surface appearance. Larger pigment particles and/or higher pigment concentration tends to roughen the initial coating surface. The pigment also has a large impact on the coating surface degradation after weathering exposure. Most of the degradation occurs evenly at the surfaces of CY and CR coatings. However, in the case of a CB coating with a higher concentration of titanium dioxide in the formulation, the action does not occur evenly but tends to form hollows and holes. Moreover, significantly less dirt is found on both the unwashed and washed CB coating surfaces. This is believed to be because of the high photo-activity of the titanium dioxide particles, despite their being coated and stabilized. The above observations were then confirmed chemically by the surface chemistry analysis (XPS) and PA-FTIR analysis of yellow and white only panels exposed in QUV-A cabinets.
The bulk chemistry investigation (SSPA-FTIR) shows a very different picture: that although the CY coating retains higher gloss, it shows the most bulk degradation and color change among the three coatings. This is believed to be mainly because of the lead chromate, which is a highly oxidizing material that over the duration of the exposure degrades and oxidizes much of the organic binder. This is a very different mechanism than the radically driven photo-oxidation caused by high levels of titanium dioxide pigmentation, which leads to significant surface changes in CB coating after weathering. Moreover, the causes of large holes on the CB-W coating surface could be related to the exclusion of pigment, especially from the melamine enrichment zones (19); otherwise, the highly concentrated titanium dioxide regions may undergo more rapid degradation during the weathering. These will all lead to the uneven coating of surface features as observed from SEM/AFM images (future work will be undertaken to investigate these hypotheses further). Because of the low concentration of titanium dioxide, as well as the better stability of iron oxide, the CR coating shows the least surface change and bulk degradation after weathering.
In summary, this study not only provides a comprehensive investigation into coating degradation upon natural weathering, but also shows the effect of pigment on coating durability, which suggests different pigment catalyzes coating degradation under different mechanisms. Moreover, the coating surface changes may tell a completely different story from the actual bulk degradation. The coating degradation may even undergo different mechanisms when exposed under different weathering conditions (e.g., QUV-A, natural environment). Therefore, it is necessary to analyze coating performance using both surface and bulk sensitive techniques to fully understand the mechanisms behind the organic coating degradation in the natural environment. It is highly recommended that the design of the coil coating formulation should take both the coating's service environment and its pigment properties into account in order to prevent unexpected failures.
Acknowledgments The authors thank Mr. James Maxted, Mr. James Smith (Becker Industrial Coatings, UK), Mr. Paolo Marino (University of Surrey, UK), Mr. Colin Lovell and Dr. Lesley Wears (Exeter College, UK) for useful advice, and acknowledge Becker Industrial Coatings for funding this work.
(1.) Sanders, D, Waterborne and Solvent Based Saturated Polyesters and Their End User Applications, Wiley. London, 1999
(2.) Ranby, B, "Photodegradation and Photo-Oxidation of Synthetic Polymers." 7. Anal. Appl. Pyrol., 15 237-247 (1989)
(3.) Bauer, DR, Mielewski, DF, Gerlock, JL, "Photooxidation Kinetics in Crosslinked Polymer Coatings." Polym. Degrad. Stab., 38 57-67 (1992)
(4.) Bauer, DR, Mielewski, DF, "The Role of Humidity in the Photooxidation of Acrylic Melamine Coatings." Polyin. Degrad. Stab., 40 349-355 (1993)
(5.) Nguyen, T, Martin, J, Byrd, E, Embree, N, "Relating Laboratory and Outdoor Exposure of Coatings. III. Effect of Relative Humidity on Moisture-Enhanced Photolysis of Acrylic-Melamine Coatings." Polym. Degrad. Stab., 77 1-16 (2002)
(6.) Mori, K, Tachi, K, Muramatsu, M, Torita, K, "Mechanism of Acid Rain Etching of Acrylic-Melamine Coatings." Prog. Org. Coat., 36 34-38 (1999)
(7.) Gesenhues, U, "Al-Doped [TiO.sub.2] Pigments: Influence of Doping on the Photocatalytic Degradation of Alkyd Resins." J. Photochem. Photobiol. A: Chem., 139 243-251 (2001)
(8.) Almeida, AR, Moulijn, JA, Mul, G, "In Situ ATR-FTIR Study on the Selective Photo-Oxidation of Cyclohexane Over Anatase [TiO.sub.2]." J. Phys. Chem., 112 1552-1561 (2008)
(9.) Scierka, S, Forster, A, Kosik, W, "An XPS Study on the Effect of Pigment on the UV Degradation of an Epoxy System." Proc. 81st Annual Meeting Technical Program of the FSCT, 2003
(10.) Biggs, S, Lukey, CA, Spinks, GM, Yau, S, "An Atomic Force Microscopy Study of Weathering of Polyester/Melamine Paint Surfaces." Prog. Org. Coat., 42 49-58 (2001)
(11.) Faucheu, J, Wood, KA, Sung, LP, Martin, JW, "Relating Gloss Loss to Topographical Features of a PVDF Coating." J. Coat. Technol. Res., 3 29-39 (2006)
(12.) Lavaste, W, Watts, JF, Chehimi, MM, Lowe, C, "Surface Characterisation of Components Used in Coil Coating Primers." Int. J. Adhes. Adhes., 20 1-10 (2000)
(13.) Haverkamp, RG, Siew, DC, Barton, TF, "XPS Study of the Changes During the Service Life of Polyester Powder Coatings." Surf. Interface Anal., 33 330-334 (2002)
(14.) Hinder, SJ, Lowe, C, Maxted, JT, Watts, JF, "Migration and Segregation Phenomena of a Silicon Additive in a Multilayer Organic Coating." Prog. Org. Coat., 54 104-112 (2005)
(15.) Urban, MW, "Recent Advances in Coatings Characterisation by Photoacoustic FTIR Spectroscopy." Prog. Org. Coat., 16 321-353 (1989)
(16.) Zhang, WR, Lowe, C, Smith, R, "Depth Profiling of Coil Coating Using Step-Scan Photoacoustic FTIR." Prog. Org. Coat., 65 469-476 (2009)
(17.) Heliogen[R] Blue L 7072 D Pigment Data Sheet. The BASF Company
(18.) Euvananont, C, Junin, C, Inpor, K, Limthongkul, P, Thanachayanont, C, "[TiO.sub.2] Optical Coating Layers for Self-Cleaning Applications." Ceram. Int., 34 1067-1071 (2008)
(19.) Zhang, WR, Smith, R, Lowe, C, "Confocal Raman Microscopy Study of the Melamine Distribution in Polyester-Melamine Coil Coating." J. Coat. Technol. Res., 6 315-328 (2009)
(20.) Perruchot, C, Abel, M, Watts, JF, Lowe, C, Maxted, JT, White, RG, "High-Resolution XPS Study of Crosslinking and Segregation Phenomena in Hexamethoxymethyl Mela-mine-Polyester Resins." Surf. Interface Anal., 34 570-574 (2002)
(21.) Gesenhues, U, "Contribution of [TiO.sub.2] to the Durability and the Degradation of Organic Coatings." Proc. Forum de la Connaissance Symposium, Paris, November 1996
(22.) Gerlock, JL, Smith, CA, Cooper, VA, Dusbiber, TG, Weber, WH, "On the Use of Fourier Transform Infrared Spectroscopy and Ultraviolet Spectroscopy to Assess the Weathering Performance of Isolated Clearcoats from Different Chemical Families." Polym. Degrad. Stab., 62 225-234 (1998)
(23.) Nichols, NE, Gerlock, JL, "Rates of Photooxidation Induced Crosslinking and Chain Scission in Thermoset Polymer Coatings." Polym. Degrad. Stab., 69 197-207 (2000)
(24.) Laha, SC, Glaser, R, "Characterisation and Catalytic Performance of [Cr]MCM-41 and [Cr]MCM-48 Prepared by Either Classical or Microwave Heating." Microporous Mes-oporous Mater., 99 159-166 (2007)
(25.) Samanta, S, Mal, NK, Bhaumik, A, "Mesoporous Cr-MCM-41: An Efficient Catalyst for Selective Oxidation of Cyclo-alkanes." J. Mol. Catal. A Chem., 236 7-11 (2005)
(26.) Weckhuysen, BW, Wachs, IE, Schoonheydt, RA, "Surface Chemistry and Spectroscopy of Chromium in Inorganic Oxides." Chem. Rev., 96 3327-3349 (1996)
(27.) Muzart, J, "Chromium-Catalysed Oxidations in Organic Synthesis." Chem. Rev., 92 113-140 (1992)
(28.) Huang, Q, "Hainan Acid Rain and Analysis of Synoptic System in 2007." Chin. J. Trop. Crops, 30 238-243 (2009)
(29.) Willard, HH, Kassner, JL, "Solubility of Lead Sulfate, Chromate and Molybdate in Nitric and in Perchloric Acid." J. Am. Chem. Soc, 52 2402-2408 (1930)
(30.) Becker Industrial Coatings internal research report
W. R. Zhang, R. Smith
School of Engineering and Materials Science, Queen Mary, University of London, Mile End Road, London E1 4NS, UK
S. J. Hinder, J. F. Watts
The Surface Analysis Laboratory, Faculty of Engineering and Physical Sciences, University of Surrey, Guildford, Surrey GU2 7XH, UK
The Long Term Development Group, Becker Industrial Coatings, Goodlass Road, Speke. Liverpool L24 9HJ, UK
W. R. Zhang (*)
AkzoNobel Decorative Paints, Room 306, Building 183, Wexham Road, Slough, Berkshire SL2 5DS, UK
[C]ACA and OCCA 2010
|Printer friendly Cite/link Email Feedback|
|Comment:||An investigation of the effect of pigment on the degradation of a naturally weathered polyester coating.|
|Author:||Zhang, W.R.; Hinder, S.J.; Smith, R.; Lowe, C.; Watts, J.F.|
|Date:||May 1, 2011|
|Previous Article:||The role of organic modification of layered nanosilicates on mechanical and surface properties of organic-inorganic coatings.|
|Next Article:||Influence of some vegetable oils on the photocrosslinking of coatings based on an o-cresol novolac epoxy resin and a bis-cycloaliphatic diepoxide.|