Confocal Raman microscopy study of the melamine distribution in polyester-melamine coil coating.

Abstract Hexamethoxymethylmelamine, a common crosslinker used in combination with polyester resins to prepare thermosetting coatings, can self-condense during curing and in theory build up regions with high crosslink density (melamine enrichment). This distribution may affect such properties as the "local" glass transition temperature and structural heterogeneity and thereby modify formability. stain resistance, and weatherability. The self-condensation regions have not previously been observed in detail so their distribution in the film is unclear. Confocal Raman microscopy (CRM) has been used to characterize the melamine distribution in three polyester-melamine paint systems: one clear and two pigmented coil coatings. Both the surface mapping and depth profiling features of CRM were used to perform a 3D analysis. The depth profiling of the clear coating was performed in a nondestructive way, but due to the opaque nature of the pigments, cross sections of the pigmented coatings were prepared in order to facilitate the analysis of the melamine distribution through the film thickness. Melamine-enriched zones were found on the surfaces of all three samples. They appeared to be approximately spherical with diameters of around 5 [micro]m. It was also observed from the cross sections that the melamine-enriched zones appeared as "particles" distributed randomly through the coating. The Raman spectra collected at these regions show significantly higher triazine ring band intensity. All of these findings confirm the hypotheses constructed over the past 20-30 years.

Keywords Coil coating, Melamine enrichment. Confocal Raman microscopy

Introduction

Coil coating is a highly efficient process that applies polymer coatings onto metal substrates, normally aluminum and steel, before fabrication; it is also known as "prepaint." Coil-coated metal is used in a wide range of industrial and commercial applications, including building cladding for walls and roofs, automotive parts, and household electrical appliances. The coating films must have sufficient formability in order to withstand crack formation during the profiling process. In addition, they must possess enough surface hardness for scratch and stain resistance. Moreover, when coil coatings are employed in outdoor applications, factors such as strong sunlight, atmospheric pollution, and humidity will affect the coating's appearance over time. Therefore, the durability of coil coatings is another important property to be considered. The mechanical and chemical properties are usually attributed to the structure of the resin in the paint formulation and the crosslinking structures that are established through curing. For example, in the polyester-melamine coating system, the use of linear high [T.sub.g] polyesters results in good formability, whereas the use of a branched polyester leads to an increase in the number of crosslinks which will enhance hardness at the expense of formability.

Polyester-melamine coil coatings still dominate the market due to their high performance in formability, scuff resistance, adequate durability, and cost efficiency. There is a large variety of polyester resins available to meet different requirements. The most commonly used melamine crosslinking resin is hexamethoxymethylmelamine (HMMM), others with lower degrees of methylation can be used together with HMMM. The curing of coil coatings involves the trans-etherification of the melamine methoxy groups with the polyester hydroxyl groups, resulting in the loss of methanol (co-condensation). Melamine can also self-condense during this process. It is of great importance to study the melamine self-condensation because the presence of mainly self-condensation domains may affect coating properties such as glass transition temperature and effective crosslink density, thereby modifying the formability, hardness, stain resistance, and chemical resistance of the paint. For example, it was previously found that improved film formability at equivalent hardness can be related to the low tendency of HMMM to undergo self-condensation during curing. (1)

Much work has been carried out to determine the reaction mechanism of the trans-etherification and to build up the general model of melamine self-condensation so as to control the coating's internal crosslink networks and to improve coating performance. As reported in the literature, melamine self-condensation is a complex process, and it depends on factors such as the type and amount of catalysts, curing conditions, polyester hydroxyl values, etc. Calbo (2) found that di-nonylnaphthalene di-sulfonic acid (DNNDSA) produced less melamine self-condensation than the normal para-toluene sulfonic acid (p-TSA); Blank's investigations showed that increased moisture in the air significantly increased self-condensation. (3) Gan et al. (4) proposed the reaction mechanism and the general model of melamine clustering. Jones et al. (5) found the self-condensation rate also depended on the types of melamine resins and proposed the possible products.

Melamine self-condensation has been studied for decades; it is a very important phenomenon that will affect the crosslinking density of the coatings. Control of these reactions may lead to the development of specialized coatings. Therefore, the great importance of this phenomenon has also stimulated the application of new techniques on both qualitative and quantitative studies. These studies have tended to concentrate on the surface chemistry of coatings. For example, Urban et al. (6) used attenuated total reflection (ATR) Fourier transform infrared spectroscopy (FTIR) to study polyester-melamine coatings and found melamine enrichment on the surface when the hydroxyl value of polyester is low. Another ATR-FTIR study was reported by Hamada et al. (7) who quantified the degree of melamine enrichment in the surface region of polyester-coatings as a function of infrared penetration depth. More recently, Gamage et al. (8) used X-ray photoelectron spectroscopy (XPS) to analyze the distribution of melamine at the surface of polyester melamine based coatings cured under nonisothermal conditions and also found melamine enrichment at the coating/air interface.

Although melamine enrichment near the coating surface has been previously studied in details, melamine self-condensation in the bulk has not been particularly well elucidated as ATR-FTIR and XPS are both surface-sensitive techniques (a few nanometers/micrometers). Very recently, confocal Raman microscopy (CRM) has been demonstrated to be very suitable to determine the component distribution in polymer films or multilayer polymeric systems and is nondestructive. For example, Fleming et al. (9) used CRM to analyze dye diffusion in PET fibers and found its distribution to be inhomogeneous; Schrof et al. (10) studied the distribution of photostabilizers in UV-cured coatings using CRM: Dupuie et al. (11) used CRM to study the multilayer paint systems. Both Mura et al. (12) and Schmidt el al. (13) found heterogeneous distribution of certain functional groups in polymer films using CRM. The resolution and accuracy of CRM has also been previously investigated by Everall. (14), (15) The compatibility of this technique with commercial coil coatings is the subject of a forthcoming paper. (16)

The current investigation is focused on utilizing the CRM surface mapping and depth profiling techniques together with its optical imaging function to investigate the distribution of melamine in polyester-melamine coil coatings. A 3D view of the melamine distribution in coatings will be shown.

Experimental

Materials and methods

The coil coating samples investigated in this work were all laboratory prepared aliphatic polyester-melamine coatings. Two pigmented coatings and one clear coating were assembled. The samples and their pigmentation are listed in Table 1. Due to reasons of commercial confidentiality, only a general description of the components included in the coil coating formulations employed is provided here.

Table 1: Sample information

Sample  Base resin     OH    HMMM  High NH  PE  Pigmentation
                      value        MMM (a)

Clear   Aliphatic PE   35     >5     >5     82  N/A
White   Aliphatic PE   35     >5     >5     82  Titanium dioxide (52)
Red     Aliphatic PE   35     >5     >5     82  Iron oxide (19)

Note: PE, polyester

(a) The ratio of HMMM to high NH MMM was the same in every formulation

All samples were prepared using a standard coil coating laboratory simulation method. The coatings were cast as liquids on to the primer-coated hot dip galvanized (HDG) steel substrates using a wire wound draw down bar of suitable diameter to achieve a dry film thickness of approximately 20 [micro]m. They were cured for 30 s in an electric oven with efficient air flows to reach a peak metal temperature (PMT) of 232[degrees]C. Panels were quenched with tap water directly after curing and air-dried. Discs with diameter of 8 mm were punched out and used for the CRM analysis.

The cross-sectional samples were prepared and polished using a standard method. A strip was cut from the panel and mounted vertically in a mold. The position was maintained so as to keep the strip standing vertically. The mould was then filled with curing resin specially chosen not to swell the coating. After curing overnight, the samples were demolded, finely polished using a Struer Labopol 5 grinding machine with laboforce 3 head. P120, P180, and P320 sand papers were used to grind samples; they were then polished with abracloth 9 [micro]m and durasilk 3 [micro]m. Finally, the samples were washed with distilled water.

Confocal Raman microscopy instrumentation

All Raman spectra were recorded using a Nicolet Almega visible dispersive Raman spectrometer (Madison, USA) with a 785-nm laser excitation. A l00x dry-objective (NA = 0.90, Olympus) was used for both surface mapping and depth profiling experiments. This theoretically yields a depth resolution of 1.02 [micro]m; a lateral resolution of 0.44 [micro]m and a laser spot size of 0.96 [micro]m. A laser power of 300 mW was used for spectra collection of the clear and white coil coatings: while for the red coil coating a lower laser power of 120 mW was set. The spectra collection at different positions was precisely controlled by a motorized stage. The stage movement was controlled automatically by the Raman instrument during the data collection. All of the data processing were performed using Omnic V7.3 software.

For the CRM surface mapping experiments, areas with the sizes of 10 x 10 [micro]m to 20 x 20 [micro]m were investigated. The step size was set at 1 [micro]m to achieve a high spatial resolution. Therefore, each map contains 100-400 sampling points with one spectrum collected at each position.

The depth profiling of the clear coating was carried out by collecting Raman spectra throughout the coating thickness. For the pigmented coatings, the melamine distribution analysis was performed using CRM to scan through the cross-sectional samples (lateral scanning). A step size of 1 [micro]m was set for both depth profiling and lateral scanning to enhance the resolution.

Results and discussion

Melamine estimation method

Figure 1 shows the Raman spectrum of a clear polyester-melamine coil coating under investigation in this work. The band assignments associated with this current study are summarized in Table 2. In order to estimate the quantity of melamine from the Raman spectra, a peak area ratio method was adopted to overcome the problems associated with fluorescence in the samples, especially the pigmented coatings. The peaks chosen for this ratio calculation are the melamine ring band at 980 [cm.sup.-1] (the stronger of the three melamine bands) and the polyester band at 950 [cm.sup.-1]. Thus, the melamine-to-polyester peak area ratio ([Ratio.sub.MEL]) is used to represent the relative quantity of melamine in the coil coatings.

[FIGURE 1 OMITTED]

Table 2: Selected band assignments of polyester-melamine coil coatings

Band position ([cm.sup.-1])   Assignment

1557                         Melamine (a) side chain C-N
1395                         Melamine side chain C-O
984                          Melamine ring
914                          Residual methoxy group (b)
953                          Polyester
810                          Polyester

(a) The melamine mentioned in this work refers to HMMM

(b) The residual methoxy groups mentioned here are the nonreacted
methoxy groups on the HMMM side chains

CRM surface mapping analyses

Figure 2a shows an optical image of the surface of the clear coaling; an indistinct feature is observed in the marked region. Figure 2b shows a CRM [Ratio.sub.MEL] map (the intensity bar indicates the values of [Ratio.sub.MEL] with corresponding colors in the maps) that shows the concentration of melamine in the scanned area marked in Fig. 2a, while Fig. 2c shows a 3D plot of the melamine distribution that provides a more instant view. It should be noted that this has no topographical significance: it is purely a device to show the [Ratio.sub.MEL], intensity at different regions on the surface. It is clearly observed that the melamine concentration is much higher in the region marked with a white circle in Fig. 2a. The value of [Ratio.sub.MEL] (Fig. 2a) in the red zone (~3.4) is approximately 2.3 times higher than that in the blue regions (~1.5) and the size of this melamine-enriched zone is approximately 5 [micro]m (the green zone) in diameter. These observations can be possibly explained by the general melamine self-condensation model proposed by Gan et al., (5) who suggested that phase separation might occur in regions that they referred to as melamine clusters due to a higher crosslink density. Therefore, the observed feature of the melamine-enriched zone may arise from such high crosslink density.

[FIGURE 2 OMITTED]

The Raman spectra collected in the red. green, and blue regions in Fig. 2b are illustrated in Fig. 2d; the intensity scale is the same for all spectra shown. It can be seen that the melamine bands at 1557, 1395, and 984 [cm.sup.-1] all increase in height, whereas the residual methoxy band at 914 [cm.sup.-1] is fairly constant. It is well known that the methoxy groups are eliminated by either the trans-etherification of the polyester with melamine and/or the melamine self-condensation reactions. Therefore, the relatively constant distribution of residual methoxy groups suggests the observed melamine-enrichment zones are not simply developed by inhomogeneous mixing of the polyester and melamine resins. The melamine molecules must be crosslinked via self-condensation that results in localized melamine enrichment.

Figure 3 shows the surface mapping of a white pigmented coil coating. Unlike the optical image of the clear coating, the corresponding image shown in Fig. 3a is much clearer in its definition of a possible melamine-enriched zone (marked with the red circle).

[FIGURE 3 OMITTED]

The CRM [Ratio.sub.MEL] map of this area confirms that this spherical-like particle with a diameter of ~5 [micro]m is a melamine-enriched region. The value of [Ratio.sub.MEL] in the red core of the melamine-enrichment zone (~2.95) is about 2.3 times higher than the surrounding blue areas (~1.3). The spectra shown in Fig. 3c suggest the melamine bands at 1556, 1397, and 980 [cm.sup.-1] all increase in height while the residual methoxy band at 914 [cm.sup.-1] is relatively constant. These findings are all in good agreement with the CRM surface mapping of the clear coil coating. Therefore, the inclusion of the titanium dioxide pigment in the coil coating has not interfered with the localized melamine enrichment.

The distribution of the titanium dioxide pigment in the same scanned area (Figs. 3a and 3b) on the white coating surface is also analyzed by CRM surface mapping, as shown in Fig. 4a; this analysis is based on the Raman intensity of the titanium dioxide band rather than a peak area ratio (the intensity bar indicates the relative Raman intensities with corresponding colors in the maps). The titanium dioxide characteristic peaks are shown in Fig. 4b. It is observed from Fig. 4a that the pigment concentration is much lower in the middle region of the scanned area. Therefore, the region with higher concentration of melamine contains less pigment.

[FIGURE 4 OMITTED]

The CRM surface mapping of the red coil coating is exhibited in Fig. 5. A possible melamine-enriched region approximately 5 [micro]m in diameter is observed and shown within the white circle (Fig. 5a). As shown in Fig. 5b, the [Ratio.sub.MEL] map of the scanned area (the blue square in Fig. 5a) suggests a significantly higher concentration of melamine within the white circle.

[FIGURE 5 OMITTED]

As illustrated in Fig. 5c, the Raman spectra collected at three different positions in the [Ratio.sub.MEL] map show that the intensities of all three melamine bands increase from the blue to the red regions, while the residual methoxy and polyester corresponding band intensities are fairly constant. The value of [Ratio.sub.MEL] of the melamine-enriched zone (~5) is roughly 3.3 times higher than the surrounding area (~1.5). These findings are generally in good correlation to those of the clear and white coatings, however, it is also observed that the maximum [Ratio.sub.MEL] value is higher and the spectra contain more noise. This is mainly due to the lower laser power (120 mW) used here since a higher power will destroy the red sample. Consequently, the signal intensity and data accuracy are reduced.

The analyses of the iron oxide pigment distribution are shown in Fig. 6; the iron oxide characteristic bands are shown in Fig. 6b. Once again, less pigment is found in the region with more melamine, as shown in Fig. 6a. The normalized Raman spectra shown in Fig. 6b also support the above finding. They are all in good agreement with the pigment distribution analysis of the white coil coating.

[FIGURE 6 OMITTED]

Therefore, it can be summarized that the melamine-enriched regions are formed in both clear and pigmented coil coatings, possibly via melamine self-condensation reaction. In addition, the pigment distribution in the regions is significantly lower.

Depth profiling and lateral scanning analyses

A nondestructive depth profiling has only been applied to the clear coil coating. For the pigmented coil coatings, a lateral profile was obtained through the cross sections.

Figure 7 describes several different ways in which the depth profiles of the clear coat can be illustrated for a "normal" (no melamine enrichment) region. The variation with depth into the coating of the melamine ring (980 [cm.sup.-1]), polyester (950 [cm.sup.-1]), and residual methoxy groups (914 [cm.sup.-1]) are highlighted in the Raman intensity map shown in Fig. 7a. It is a very different map from Fig. 2b, which is a topographical map of the surface. This map looks at the variation in the chemistry with depth into the film under an area no bigger than 1 [micro][m.sup.2]. Color from red to blue indicates the corresponding Raman intensities; the X axis has no spatial significance and indicates the band position while the relative laser penetration depth is plotted in the Y axis; the surface and bottom of the clear coating are as indicated in Fig. 7a. Although higher intensity values for the melamine ring band are observed at depths of 5-13 [micro]m; the other bands also perform similar trends. Therefore, the decrease in Raman intensities at the depths of 0-5 urn and 20-25 urn is actually due to either the out of focus signals and/or the reduction of the signal magnitude in the coating. This does not represent an inhomogeneous distribution of melamine.

The [Ratio.sub.MEL] against relative depth is plotted in Fig. 7b, i.e., the depth is plotted on the X axis starting at the bottom of the film approximately 25 urn from the surface. A variation ([+ or -]0.1) in the [Ratio.sub.MEL] value is observed due to the Raman signal fluctuation; however, this small variation will not affect the melamine distribution analysis in this work. A dotted line is fitted to the graph to indicate where the homogenous distribution of melamine in the clear coating would sit. The mean value of the [Ratio.sub.MEL] is around 1.4. As observed in Fig. 7c, the Raman spectra collected at three different depths in the clear coating also suggest a homogenous distribution of melamine in the" normal" regions.

The depth profiling of the melamine-enriched zone in the clear coating is shown in Fig. 8. The figures are in the same order as Fig. 7. It is observed from Fig. 8a that the melamine band (984 [cm.sup.-1]) is very strong at relative depths from 5 to 10 [micro]m. while the other bands all show similar intensities (e.g., compare the distribution of the 984 [cm.sup.-1] band to the distribution of the 1050 [cm.sup.-1] band in Figs. 7a and 8a). The [Ratio.sub.MEL] depth profile shown in Fig. 8b clearly exhibits a very different behavior compared with the CRM depth profiling in the "normal" region. The [Ratio.sub.MEL] increases to a maximum value of [~3.4] (the core of the melamine-enriched zone) and then decreases to an approximately constant value of [~1.5] that is quite similar to the mean value in Fig. 7b. The vertical size of the melamine-enriched zone is approximately 8 urn as indicated in Fig. 8b.

As shown in Fig. 8c, the Raman spectra collected at three different depths (in Fig. 8b) reinforce the above finding. The melamine band (984 [cm.sup.-1]) is much stronger in the middle of the melamine-enriched zone (spectrum at depth 7 [micro]m) and lower at both the surface and bottom regions. Therefore, the observed melamine-enriched region is found approximately 3 [micro]m below the coating surface and is not seen throughout the coating thickness.

Unlike the depth profiling of the clear coating, a cross section of the white coating is scanned by the laser to investigate the distribution of melamine in the coating. An optical image of the white cross section is shown in Fig. 9a with the different layers clearly observed. A distinct feature that appears in the white pigmented coaling is observed as a "gray particulate." CRM surface mapping within the blue square area shown in Fig. 9a is plotted in Fig. 9b. This map provides an instant view of the melamine distribution near the "particle"; the 980 [cm.sup.-1] band intensity is used as the indication of melamine rather than the use of [Ratio.sub.MEL]. No melamine resin is incorporated into either the curing resin or the steel substrate; therefore, their corresponding layers are shown as blue. The primer contains less melamine resin in the formulation and the layer is relatively thin, so the light blue color in the primer region is observed. However, the "particle" region in the white coating layer (Fig. 9a) appears as red in Fig. 9b, which indicates significantly higher melamine concentration in this region. In other words, the "particle" with diameter around 8 [micro]m is actually a melamine-enriched region in the white coating. Moreover, the spectra (Fig. 9c) collected at the high and normal melamine concentration regions as seen in Fig. 9b show clearly less titanium dioxide band intensity at the high melamine concentration region. This observation correlates to the previous results quite well and suggests that the pigment distribution at the high melamine concentration area is quite low.

The lateral scanning pathway through the "particle" is shown in Fig. 9a with a dark blue dotted line. The Raman intensity map is shown in Fig. 9d; this map is different again from Figs. 2a, 3a, 7a, and 8b because here the Raman intensity of the various bands is plotted from several spectra taken along the dotted line. It can be observed that the "particle" region contains significantly higher melamine concentration. The [Ratio.sub.MEL] lateral profile is plotted in Fig. 9e and it can be seen that there is no melamine in the curing resin layer (as expected), moderate melamine concentration ([Ratio.sub.MEL] [approximately equal to] 1.7) in the white coating layer and significantly higher melamine concentrations ([Ratio.sub.MEL] [approximately equal to] 3.5) in the "particle" region. Moreover, different layers can also be distinguished using the [Ratio.sub.MEL] values as indicated in Fig. 9e. (Note: The high value of [Ratio.sub.MEL] in the primer is due to signal fluctuation.)

Similarly, a "gray particle" with a diameter of ~8 [micro]m is also observed in the cross section of the red coating, as shown in Fig. 10a. It is observed from Fig. 10b that the [Ratio.sub.MEL] profile obtained by lateral scanning through the "particle" in the red cross section starts increasing from the curing resin/red coaling boundary (i.e., the surface of the coil coating) and reaches a maximum value of ~5. Similar to the observation from the CRM surface mapping of the red coil coating (Fig. 6), the pigment Raman intensity in the high melamine concentration region as observed in Fig. 10a is found to be lower than that in the normal regions.

[FIGURE 10 OMITTED]

Therefore, it can be generally concluded from the above findings that the melamine-enriched zones give unique optical features; this is primarily due to the lower pigment concentration in these areas. Moreover, as demonstrated in a previous paper, (16) the pigment band can be observed even when the Raman laser is focused on the surface of a multilayer coil coating sample (a polyester/melamine clear coating on the titanium dioxide pigmented polyester/melamine coating). Thus it is also quite possible that the melamine-enriched regions contain no pigment: the observation of pigment bands with lower Raman intensity is due to the laser refraction at the deeper layer throughout the clear particles (melamine-enriched zones). The detailed mechanisms for the observation of melamine-enriched zones are not clear; a possible explanation is that phase separation occurred during the curing process. After the mixing of the resin and pigment, there will be molecules of polyester resin that are associated with pigment and molecules that are not. Those that are not are free to react with melamine; but once reacted, the melamine could be available to react with the next most likely candidate. If this happens to be another melamine molecule because the stoichiometric concentration of this material is higher, then this will reduce the local concentration of melamine and so more will migrate into this zone, giving even more opportunities to increase the melamine concentration and further self-condensation. Thus the exclusion of the pigment particles and phase separation as observed is more of an indirect consequence of the processes that are on going during cure.

In summary, the melamine-enriched regions with less pigment concentration are observed from both the surface and deeper parts of the coatings. A clearer image is obtained from the sample cross sections. Generally, the melamine-enriched zones appear like "particles" with sizes of around 5-8 [micro]m in diameter distributed randomly throughout the coating. The [Ratio.sub.MEL] values obtained by different methods are summarized in Table 3; they are generally consistent (the data from the red coating show larger errors due to the low laser power used, as discussed above). Moreover, it should be emphasized that as the melamine/polyester ratio ([Ratio.sub.MEL]) is based on the relative Raman spectral intensity ratio of the two bands (melamine band and polyester band), they cannot be correlated to the actual weight ratio of the melamine and polyester resins used according to the coating formulation. However, the [Ratio.sub.MEL] obtained at the normal regions can be considered as the reference and then compared with the [Ratio.sub.MEL] obtained at the high melamine concentration regions. According to Table 3. the [Ratio.sub.MEL] in normal regions has a value of approximately 1.5 while the high melamine concentration region has a value of 3-5. This indicates the high [Ratio.sub.MEL] regions have more melamine molecules (more precisely, higher ratio of melamine molecules to polyester molecules) than that in the normal regions.

Table 3: Summary of [Ratio.sub.MEL]

Sample    Analytical  [Ratio.sub.MEL]  [Ratio.sub.MEL]  [Ratio.sub.MEL]
            method       (high) (a)                        (high)/
                                                        [Ratio.sub.MEL]

Clear     Surface          3.4              1.5              2.3
          mapping

White     Surface          2.95             1.3              2.3
          mapping

Red       Surface          5.0              1.5              3.3
          mapping

Clear     Depth            3.4              1.5              2.3
          profiling

White-CS  Lateral          3.5              1.7              2.1
          scanning

Red-CS    Lateral          5.0              2.0              2.5
          scanning

Note: CS, cross-sectional sample

(a) The high [Ratio.sub.MEL] value obtained from the melamine enriched
regions

Conclusion

The melamine distribution in three polyester-mela-mine coil coating systems has been successfully analyzed by CRM. This technique has been demonstrated to be very suitable to study the distribution of functional groups in coil coatings. CRM mapping at the coating surface together with depth/lateral profiling throughout the coating thickness provide a general 3D view of the melamine enrichment in the polyester-melamine coil coatings. The localized melamine-enriched regions are found in all of the coatings at both surface and deeper regions; a clearer view is obtained only from the cross sections. Their distribution is found to be generally random throughout the coating thickness. The pigments incorporated in the paint formulation do not appear to affect the melamine enrichment. The sizes of the melamine-enriched zones range from approximately 5 to 8 [micro]m in diameter and the [Ratio.sub.MEL] value in the melamine-enriched zone is ~2 times higher than that of the surrounding regions.

We believe this is the first time that the formation of discrete zones of melamine self-condensate have been characterized in such detail throughout the thickness of a polyester-melamine coil coating. The findings in this work are in good correlation with the theory and models of the melamine self-condensation that have been previously developed. The different optical features of the melamine-enriched zone especially in the pigmented coatings are possibly due to the compact crosslinks formed during curing process restricting the pigment particles incorporation. All of the above findings suggest a different phenomenon from the melamine segregation on the coating surface that has been previously investigated using surface-sensitive techniques such as ATR-FTIR or XPS.

The formation of these self-condensation zones has only just been observed and so the effect of the components on the melamine self-condensation is not clear. However, they do seem to be related to the presence of strong bases. Perhaps, self-condensation is encouraged by a lack of catalyst. A future paper will discuss the evidence for a greater crosslink density and the effects that the zones have on mechanical and durability properties.

Acknowledgments The authors thank Mr. James Maxted and Mr. James Smith (Becker Industrial Coatings. UK) for useful discussions and help, and acknowledge Becker Industrial Coatings for financial support.

References

(1.) Koral, JN, Petropulos, JC, "Hexakis (MethoxyMethyl) Melamine: Its Chemistry and Utilization in Surface Coatings," J. Paint Technol., 38 600-609 (1996)

(2.) Calbo, LJ, "Effect of Catalyst Structure on the Properties of Coatings Cross-Linked with Hexa(MethoxyMethyl) Melamine," J. Coat. Technol., 52 75-83 (1980)

(3.) Blank, WJ, "Reaction Mechanism of Melamine Resins." J. Coat. Technol., 51 61-70 (1979)

(4.) Gan, S, Solimeno, RD, Jones, FN, Hill, LW, "Recent Studies of the Curing of Polyester-Melamine Enamels. Possible Causes of Overbake Softening." Water-Borne and Higher-Solids Coatings Symposium (1989)

(5.) Jones, FN, Chu, G. Samaraweera, U, "Recent Studies of Self-Condensation and Co-Condensation of MelamineFormaldehyde Resins; Cure at Low Temperatures." Prog. Org. Coat., 24 189-208 (1994). doi:l0.1016/0033-0655(94) 85014-3

(6.) Hirayama, T, Urban, MW, "Distribution of Melamine in Melamine/Polyester Coatings; FT-IR Spectroscopic Studies." Prog. Org. Coat., 20 81-96 (1992). Doi: 10.1016/0033-0655(92)85006-H

(7.) Hamada, T, Kanai, H, Koike, T, Fuda, M, "FTIR Study of Melamine Enrichment in the Surface Region of Polyester/Melamine Film." Prog. Org. Coat., 30 271-278 (1997). doi:10.1016/S0300-9440(97)00003-9

(8.) Gammage, NJW, Hill, DJT, Lukey, CA. Pomer, PJ, "Distribution of Melamine in Polyester-Melamine Surface Coatings Cured Under Nonisothermal Conditions." J. Polym. Sci. A: Polym. Chem., 42 83-91 (2004). doi: 10. 1002/pola.11006

(9.) Fleming, OS, Kazarian, SG, Bach, E, Schollmeyer, E, "Confocal Raman Study of Poly (Ethylene Terephthalate) Fibres Dyed in Supercritical Carbon Dioxide: Dye Diffusion and Polymer Morphology." Polymer, 46 2943-2949 (2005). doi:10.1016/j.polymer.2005.02.067

(10.) Schrof, W, Beck, E, Koniger, R, Reich, W, Schwalm, R, "Depth Profiling of UV Cured Coatings Containing Photo-stabilizers by Confocal Raman Microscopy." Prog. Org. Coat., 35 197-204 (1999). doi:10.1016/S0300-9440(99)00026-0

(11.) Dupuie, JL, Weber, WH, Scholl, DJ, Gerlock, JL, "Clear-coat Analysis in Isolated and Multi-Layer Paint Systems by Confocal Raman Microscopy." Polym. Degrad. Stab., 57 339-348 (1997). Doi:10.10l6/S0141-39l0(97)00029-3

(12.) Mura, C, Yarwood, J. Swart, R, Hodge, D, "Raman Microscopy Studies of the Distribution of the Fungicide Fluorfolpet in Plasticized PVC Films." Polymer, 41 8659-8671 (2000). doi:10.1016/S0032-3861(99)00446-2

(13.) Schmidt, P, Fernandez, MR, Pastor, JM, Roda, J, "Study of the Conformations of Poly ([epsilon]-Caprolactam) and Poly ([epsilon]-Caprolactam)-Polybytadiene Block Copolymers by FTIR Spectroscopy with Photoacoustic Detection and by MicroRaman Confocal Spectroscopy." Polymer. 38 2067-2075 (1997). doi:10.1016/S0032-3861(96)00766-5

(14.) Everall, NJ, "Modeling and Measuring the Effect of Refraction on the Depth Resolution of Confocasl Raman Microscopy." Appl. Spectrosc., 54 773-782 (2000). doi:10.1366/0003702001950382

(15.) Everall, NJ, "Confocal Raman Microscopy: Why the Depth Resolution and Spatial Accuracy Can be Much Worse than You Think." Appl. Spactrose., 54 1515-1520 (2000). doi: 10.1366/0003702001948439

(16.) Zhang, WR, Lowe, C, Smith, R., "Depth Profiling of Coil Coating by Confocal Raman Microscopy." Prog. Org. Coat. (2008), submitted

This paper was awarded Second Place in the 2008 FSCT Roon Awards competition held as part of the FutureCoat! conference sponsored by the Federation of Societies for Coatings Technology, in Chicago, IL, on October 14-16, 2008.

W. Zhang ([??]), R. Smith

Queen Mary, University of London, London, UK

e-mail: [e-mail]

C. Lowe

Becker Industrial Coatings, Liverpool. UK