Surface and interface characterization of chlorinated polyolefin coated thermoplastic polyolefin.

Laser scanning confocal fluorescence microscopy measurements were performed on thermoplastic polyolefin (TPO) substrates that were coated with chlorinated polyolefin (CPO). The TPO investigated was a blend of high modulus polypropylene with an ethylene-butene copolymer (EBR9) containing 9 wt% butene. The CPO was a maleated chlorinated polypropylene containing 20 wt% Cl. The purpose of these experiments was to obtain detailed mechanistic information about the CPO-TPO interaction. To achieve acceptable contrast in these measurements, a fluorescent dye was covalently attached to a small portion of the CPO. Solvent wiping of the TPO substrates with isopropyl alcohol followed by xylenes prior to coating with CPO increased the mean roughness of the TPO surface by more than 100 nm; but it had a much larger effect on the roughness of the (several micrometers) CPO-TPO interface after coating. The EBR component of the TPO was shown to be exclusively responsible for the roughness increase. We also found evidence of a diffuse interface between the CPO and ERB components that was localized to sites in which the EBR was present at the TPO surface.

Keywords: Fluorescence spectroscopy, surface analysis, confocal microscopy, chlorinated polyolefin, adhesion promotion, solvents, solvent wiping, thermoplastic olefins 

**********

Many plastic automotive parts such as bumpers and fascia are fabricated from thermoplastic polyolefin (TPO). The term TPO refers to a family of polypropylene (PP) blends in which a second polyolefin, such as ethylene-propylene rubber (EP), ethylene-propylene-diene rubber (EPDM), or ethylene-butene rubber (EBR) serves as the impact modifier. Coating failure on thermoplastic polyolefin substrates has been a problem confronting the automotive industry for many years. The poor paintability of TPO substrates results from its low surface energy, the lack of active groups, and the chemical inertness of the main TPO components. Although the introduction of the impact modifier into the blend tends to improve the paintability of TPO, it is still difficult to obtain a strong coating directly on a TPO substrate. For proper paint adhesion to TPO, additional surface treatments are required. These include plasma or flame treatment or corona discharge to oxidize the surface in order to introduce polar groups. (1) Another major strategy for surface preparation is the application of a thin layer of adhesion promoter (AP) to the TPO surface to enhance the coating adhesion to TPO. Most adhesion promoters are formulations containing chlorinated polyolefin (CPO). The adhesion promoter is regarded as a coupling agent in the painted TPO system, and TPOs of different composition normally require different CPOs for optimum performance. Although CPO has been used as an adhesion promoter in the automotive industry for many years, knowledge about how CPO promotes coating adhesion to TPO is limitroducible results is challenging because of the complexity and variability of the underlying TPO substrate. (2)

To appreciate some of the complexities, consider first that while paint will adhere to CPO-primed injection-molded TPO, the adhesion is substantially poorer if the TPO of the same composition is compression-molded. (3) One concludes that the processing conditions must affect the surface morphology of the TPO blend in a way that is essential for obtaining good paint adhesion. For polymer blends like TPO, injection-molded components are formed under nonequilibrium conditions in which shear forces associated with flow fields and temperature gradients associated with the mold surface affect the overall morphology in ways that are not well understood. These processing conditions can affect the fraction of the impact modifier component at the TPO surface as well as the nature of the crystalline PP domains near or at the surface. In one study, Ryntz and Ramamurthy (4) found that a thin transcrystalline PP region developed at the TPO surface of an injection-molded PP/EPR blend. Beneath this layer was a thick region containing large PP spherulite crystals followed by a layer rich in EPR. EPR domains near the surface tended to be elliptical, with the long axis of the domains following the flow direction of the TPO melt. Beyond these global aspects of the morphology of this particular blend, the authors commented that TPO has a high spatial heterogeneity in all dimensions that is difficult to characterize. (5)

A number of investigations have been focused on developing an understanding of the interaction between the CPO and the TPO substrate. (6-12) These experiments have been carried out on TPO samples in which the primarily isotactic PP had a significant amorphous content, and on CPO samples chosen for their adhesion to this type of TPO. Ryntz et al. (3) examined the role played by the choice of solvent used as the CPO vehicle. They found that different solvents and different baking protocols appeared to change the morphology of the TPO substrate, and they concluded that solvent-induced swelling of the substrate plays a vital role in TPO substrate paintability. This study also demonstrated that the presence and nature of crystalline polypropylene at the surface of the TPO substrate affects the subsequent paintability. (3) A reasonable hypothesis about CPO promoting adhesion on TPO substrate is that the adhesion of CPO to TPO comes from the penetration of CPO chains into the rubber component of the TPO matrix. The adhesive strength is due either to entanglements or to resistance to chain pull out. (8,13) While many of these conclusions are likely to apply to any adhesion promoter-TPO combination, the way in which these interactions operate, and the relative importance of the different factors involved, may well depend upon the specific choice of CPO and TPO.

[FIGURE 1 OMITTED]

In this article, we examine the interaction of CPO with a high-modulus TPO consisting of a blend of a high melting isotactic polypropylene with a highly crystalline metallocene EBR as the impact modifier. This TPO can be thought of as a model for the next generation of plastic automotive parts which are thinner and lighter in weight, and which provide comparable mechanical strength to current TPOs formulated with EPR and EPDM. The CPO we employed (20 wt% chlorine) is itself a semicrystalline polymer with approximately 30% degree of crystallinity, as indicated by the manufacturer. (14) Product literature suggests that PP-based CPO with low chlorine content (18 or 20 wt%) provides better adhesion to high modulus PP-based TPO than CPO materials with a higher chlorine content. A sample of this CPO was covalently labeled with a fluorescent dye, and this labeled polymer was reintroduced into the CPO as a tracer for fluorescence microscopy experiments. One of the most interesting variables that we investigated is the influence of solvent wiping, with xylenes and isopropyl alcohol as the wiping solvents, on the CPO/TPO interface in the coated substrate. On the other hand, we found very little influence of the baking temperature used for freshly coated TPO substrate.

[FIGURE 2 OMITTED]

EXPERIMENTAL

Materials

Injection-molded talc-free TPO plaques were provided by Visteon Automotive Systems. These were fabricated from a blend of 75 wt% isotactic polypropylene (PP1042) and 25 wt% ethylene-butene rubber (EBR9, 9 wt% butene) (Exact 3125) from Exxon. Some of our experiments were carried out on these individual components, which were obtained from Exxon. Solvents such as xylenes, cyclohexane, and 2-propanol were purchased from Aldrich and used without further purification. The CPO sample was a chlorinated, maleated polypropylene (Superchlon 872S, Nippon Paper Industries Ltd.) with 20 wt% Cl content. By gel permeation chromatography (GPC, polystyrene standards), we determined the molecular weight of this CPO sample to be [M.sub.w] = 92,000, [M.sub.n] = 41,000. The anhydride content was determined by titration and found to be 0.19 mmol/g polymer. (15) A fluorescent-dye-labeled sample of this polymer CPO (CPO-HY) was added as a tracer to the CPO samples employed here. Its synthesis and characterization are described elsewhere. (15) A stock solution of this polymer at 20 wt% solids, containing 97.5/2.5 by weight, CPO/CPO-HY, was prepared in a solvent consisting of xylene/cyclohexane, 3/1 by weight.

Sample Preparation

SAMPLES FOR SURFACE ROUGHNESS MEASUREMENTS: In addition to the TPO plaques described earlier, polypropylene (PP) and ethylene-butene rubber (EBR9) substrates for coating experiments were prepared by compression-molding using a Carver press. To prepare PP substrates, two to three grams of the PP 1042 sample in powder form were sandwiched between two sheets of poly(ethylene terephthalate) (Melinex[TM] PET) film and placed between two preheated (170[degrees]-180[degrees]C) stainless steel plates separated by a 0.5 mm metal spacer. Then the entire assembly was set into the Carver press and subjected to two metric tons of pressure for 90 sec at 170[degrees]-180[degrees]C. To minimize the influence of post-pressing thermal history, the pressed sample was quenched into ice water immediately after pressing. The EBR substrate was prepared in the same manner, but the preheating temperature for stainless steel plates and the compression-molding temperature were kept in the range of 120[degrees]-140[degrees]C. For solvent wiping experiments, pieces of polymer substrate (2.5 X 7.5 cm) were cut from the as-molded substrate. To eliminate the possible influence of impurities on the sample surface in surface-roughness measurements, all of the polymer substrates were subjected to a gentle cleaning protocol consisting of immersion into an Ovrus soap water solution, followed by rinsing with deionized distilled water, and finally with acetone.

SOLVENT-WIPING EXPERIMENTS: Experiments to test the influence of solvent wiping on polymer substrates were carried out using a special device originally designed for coating drawdown. After the test substrate was mounted on the device, solvent wiping was performed by moving the drawdown bar back and forth across the surface four times at a speed of ca. 10 cm/sec. A folded Kimwipe[TM] paper, saturated with solvent, was attached to the drawdown bar. The wiping direction on the TPO substrate was along the injection flow direction of the TPO plaque, while the wiping direction on the compression-molded PP and EBR substrates was along the length of the substrate. Two springs were used to maintain a constant force on the substrate for different pieces during the solvent-wiping procedure. The solvent-wiped substrates were dried at room temperature for at least 24 hr before surface roughness measurements were carried out using the WYKO surface profiler in the VSI mode.

[FIGURE 3 OMITTED]

COATING CPO ONTO TPO SUBSTRATES: Prior to CPO coating, each TPO substrate, cut to 2.5 X 7.5 cm, was cleaned by the gentle cleaning protocol previously described, followed by drying at room temperature. Some samples were further subjected to solvent wiping. Each TPO sample was mounted on the drawdown device, and a small amount (0.2-0.5 mL) of CPO solution (20% solids of which 2.5 wt% was CPO-HY, in xylenes/cyclohexane 3/1 w/w) was placed on the surface. A wet thin CPO layer in the range of 30-50 [micro]m thick was coated onto the TPO substrate by moving the drawdown bar at a speed of ca. 10 cm/sec. Two different types of TPO substrates were coated with CPO solution: One was cleaned only with soap water and then rinsed, and the second was washed and then subjected to solvent wiping.

CPO-coated TPO samples were subjected to three different sets of baking conditions. One set of samples was simply dried at room temperature (RT, 23[degrees]C) for 24 hr. A second set of samples was dried at room temperature for 10 min and then baked at 82[degrees]C for 30 min. A third set of samples was dried at room temperature for 10 min and then baked at 121[degrees]C for 30 min. The samples baked at 82[degrees]C or 121[degrees]C were quenched into cold water immediately after baking to minimize possible CPO crystallization. The thickness of the dry CPO layer coated onto TPO was in the range of 5-10 [micro]m.

Surface Roughness Measurements

Surface roughness measurements were carried out using the WYKO surface profiler in the vertical-scanning interferometery (VSI) mode. The back-scan and scanning distance were set at 5 and 15 [micro]m, respectively. A 120 X 90 [micro]m area was scanned by choosing a magnification of 50 X 1. The saturation value was chosen in a range of 0.01-0.05%. For the VSI mode, the vertical resolution can be 3 nm for a single measurement and 1 nm for multiple measurements. The largest vertical distance that can be measured in VSI mode is 500 [micro]m. Two parameters describing the surface roughness (or smoothness) are the average roughness ([R.sub.a]) and the root mean square (rms) roughness ([R.sub.q]). [R.sub.a] is the mean height variation calculated over the entire measured array; while [R.sub.q] is the root mean square average of the measured height deviations taken within the evaluation length or area and measured from the mean linear surface. Higher values of [R.sub.a] or higher [R.sub.q] indicate a rougher surface.

[GRAPHIC OMITTED]

Laser Confocal Fluorescence Microscopy Measurements

The morphology of CPO-coated TPO samples was visualized using a Zeiss LSM 510 laser scanning confocal fluorescence microscope (LCFM) system. Confocal images were obtained by choosing the 488 nm argon laser line as the excitation light and collecting the light emitted in the range 510-550 nm by using a band-pass filter. Distilled water was used as medium between the sample and the water-immersion objective of the microscope. An optical slice ranging from 0.6-0.8 [micro]m thick was obtained by setting the pinhole in the range 65-100 [micro]m. The instrument was operated in the frame scan mode to obtain top-view images of CPO-coated TPO; while the Z-direction multiline scan mode was used to get side-view images of CPO-coated TPO.

[FIGURE 4 OMITTED]

For each side-view image of CPO-coated TPO, the fluorescence intensity across a CPO layer coated onto TPO substrate was analyzed using Analytical Image Station[TM] image analysis software (AIS, Imaging Research Inc., St. Catharines, Ont., Canada). To carry out this analysis, we converted each confocal image from a 24-bit RGB tif file to an 8-bit grayscale tif file. From this analysis, we obtained fluorescence intensity data as a function of depth along the profile line across the coated CPO layer and into the TPO. The data were then imported into Origin[TM] software to obtain plots of the fluorescence intensity profile across the CPO layer. In these plots, we used a moving five-point averaging protocol to smooth the profile line.

RESULTS AND DISCUSSION

The experiments described in this article were carried out on a high-modulus TPO in which a Ziegler-Natta polypropylene (PP1042 from Exxon with a low amorphous content) was blended with 25 wt% of a crystalline metallocene ethylene-butene (Exact 3125) copolymer containing 9 wt% butene as the impact modifier. We refer to this EBR composition as EBR9. Talc-free plaques were injection-molded at Visteon. Experiments were also carried out on plaques of PP1042 and plaques of EBR9 compression-molded in Toronto.

Solvent Wiping and Surface Roughness

It is common practice in industrial laboratories, in the preparation of coated TPO plaques, to wipe the TPO surfaces before they are coated. In this procedure, the TPO substrate was wiped with a fabric soaked with 2-propanol (isopropyl alcohol, IPA) and with mixed isomers of xylene ("xy-lenes"), either in discrete steps or as a mixture. One function of the solvent-wiping procedure is to remove impurities like dust due to transportation and storage, and to remove any mold-release agent left over from the TPO part fabrication. As shown below, another function of the solvent-wiping step, which has not previously been documented, is to create a rougher TPO surface. This rougher surface not only can promote mechanical adhesion between CPO and TPO substrate, but it can also increase the contact area between CPO and TPO.

In our experiments, we first wiped the TPO surface with 2-propanol and then immediately wiped the surface again with xylenes. Different TPO samples were wiped under constant force and with similar wiping rates. These samples were then examined with a WYKO surface profilometer to measure the surface roughness. These measurements were carried out on wiped and unwiped samples. For each sample, images were obtained from at least eight to 10 different places. Values of the average roughness, [R.sub.a], and the mean-squared roughness, [R.sub.q], were calculated by combining data from all of these measurements.

The WYKO surface profiling system is a noncontact optical profiler for measuring a wide range of surface heights. With this system, experiments in the phase-shift interferometry (PSI) mode are optimum for the measurement of smooth surfaces, whereas the vertical scanning-interferometry (VSI) mode is more useful for the measurement of rough surfaces and surfaces with steep steps. In the VSI mode, a white-light beam passes through a microscope objective to the sample surface. A beam splitter reflects half of the incident beam to the reference surface. The beams reflected from the sample and from the reference surface recombine at the beam splitter to form interference fringes, where the best contrast fringe occurs at best focus.

Figure 1 shows the 2D average surface roughness profile and a 3D surface image of the original TPO substrate without solvent wiping. In both the 2D profiles and the 3D image, one can clearly see peaks and valleys on the TPO surface. Although this roughness looks severe in the image, the z-axis scale is highly magnified. The values of [R.sub.a] and [R.sub.q] indicate that the roughness is on a rather fine scale. The images themselves emphasize the high resolution available with this technique. Since the injection flow direction is the y direction of Figure 1A, the direction of the valleys suggests that removing the TPO substrate from the mold during manufacture may tear off some material and cause the formation of valleys and peaks on the TPO surface.

[FIGURE 5 OMITTED]

Figure 2 provides similar information for a solvent-wiped TPO substrate. Although the surface looks smoother, the z-axis scale bars and roughness data in Figure 2 indicate that the solvent-wiped TPO has a much rougher surface than the original TPO substrate. By comparing the scale bar of Figure 2B (x-direction profile) with that of Figure 2C (y-direction), one can see the roughness along the x direction is greater than that along the y direction. We note that the y direction is the solvent-wiping direction. It appears that the influence of solvent wiping on the sample along the y direction is less than that on the x direction. Although the surface of the TPO substrate was changed significantly, the average surface roughness was still less than 500 nm.

The results shown in Figures 1 and 2 provide evidence that the surface smoothness of TPO substrate was changed significantly upon solvent wiping. To understand how this surface roughening occurred, we carried out similar solvent wiping experiments on the TPO components, PP1042 and EBR9. The wiping procedure for the PP substrate and EBR9 substrate was the same as that for the TPO samples described earlier. The results are presented in the form of a histogram in Figure 3. Here, one can see that the average surface roughness ([R.sub.a]) of solvent-wiped EBR9 substrate is about 500 nm, while that of the unwiped substrate is about 170 nm. In contrast, for the PP sample, the values of surface average roughness for both solvent-wiped and unwiped PP substrates are similar. Average surface roughness for the unwiped PP was about 145 nm, while that of the wiped samples is about 150 nm. In Figure 3, one also can see that the influence of solvent wiping on the surface roughness of TPO substrates is quite similar to that of EBR9 itself. The resistance to solvent wiping of the PP surface points to the absence of a thin amorphous PP layer, (8) which would likely be removed by wiping with xylene. Thus, it is unlikely that an amorphous PP component contributes to the increase in the [R.sub.a] of the injection-molded TPO substrate when wiped with organic solvent. It is more likely that the solvent-wiping step involves only EBR domains at the TPO surface.

Characterization of CPO-Coated TPO Substrates

It is widely believed that, in CPO-coated TPO systems, the interfacial properties between CPO and TPO play a vital role in CPO when promoting the adhesion of coatings to TPO substrates. Operational variables are the composition, molar mass, and molar mass distribution of the CPO; the composition and processing conditions of the TPO; the cleaning conditions for the TPO surface; the choice of solvent for the coating operation; and the bake temperature. In addition, most automotive parts are coated wet-on-wet, which means that the CPO coat is not dry before the basecoat and clearcoat are applied to the part. From a mechanistic point of view, one has to consider factors such as the morphology of the TPO substrate, the roughness of the substrate surface before coating, the nature of the interaction (thermodynamic and kinetic) of the CPO with the TPO components, and the nature of the interface between the CPO and the TPO substrate. These factors are largely unknown. It is our hope in undertaking these experiments to provide some information about these factors for the specific system that we examine. Our major tool for studying morphology and the CPO/TPO interface is laser scanning confocal fluorescence microscopy (LCFM). The operational variables to explore are the influence of solvent wiping and the bake temperature of the CPO-coated substrate.

[FIGURE 6 OMITTED]

For contrast, LCFM experiments require a fluorescent component. In a previous publication, (15) we described the synthesis of a dye-labeled CPO sample, prepared by reacting the anhydride groups of CPO with a reactive benzothioxanthene dye related to the commercial fluorescent dye Hostasol Yellow 3G[TM]. This synthesis is summarized in Scheme 1. We employed this dye-labeled CPO (CPO-HY) as a tracer, and prepared a blend containing 2.5 wt% CPO-HY in CPO. This blend was used in all of the experiments to be described. The interpretation of these experiments is based on the assumption that the dye-labeled CPO acts as a tracer for the CPO itself and that the blend interacts with the TPO substrate in a manner very similar to that of the unmodified CPO.

In Figure 4 we present a series of top-view images of a CPO-coated unwiped TPO substrate baked at 121[degrees]C for 30 min. These images were taken as part of a series of scans at different depths (multi-z-section scanning) by LCFM. The CPO layer is fluorescent and is represented by the bright areas in these images. To save space, we show here only a selection of images at various depths near the CPO/TPO interface. The upper left image, somewhere within the CPO layer, is arbitrarily denoted as z = 0. Subsequent images were taken at lower depths, indicated in [micro]m from the first image. In the images obtained by moving the focus from the CPO layer into the TPO substrate, one can see that the transition region between the CPO and TPO layer is very narrow (less than 0.5 [micro]m thick). In this unwiped TPO sample, the CPO/TPO interphase is very sharp, at least on the scale of the optical resolution available by this technique. There is no indication of interdiffusion at the CPO/TPO interface from this series of top-view images.

In the images at 1.1 and 1.45 [micro]m depth, one can see bright patches surrounded by a darker background. We assume that these patches are due to the valleys in the surface of the TPO substrate that we observed in Figure 1A. The bright patch suggests higher local CPO concentration at this depth into the TPO substrate. The orientation of these bright patches is the same as the injection flow direction for fabrication of the TPO substrate. For other samples of CPO-coated unwiped TPO substrates subjected to different baking conditions, the top-view images are very similar. Thus, we are unable to detect any influence of the baking conditions on the CPO/TPO interface for these samples.

Rather different results were obtained for the solvent-wiped samples. In Figure 5 we present a similar series of top-view images of a CPO-coated solvent-wiped TPO substrate. This sample was baked at 121[degrees]C for 30 min. As described in Figure 4, the CPO is fluorescent and the bright color represents the CPO layer. The first image was taken from within the CPO layer near the CPO/TPO interface and is arbitrarily designated as z = 0. The numbers in the image windows for subsequent images indicate their depth in [micro]m relative to the first image. Unlike the unwiped substrate, in which the transition region is very narrow and in which the CPO/TPO interface appears to be sharp, the CPO-coated solvent-wiped TPO has a much broader transition region. From this series of images, the transition region is about 2.6 [micro]m thick. The rougher and broader transition region is oriented along the solvent-wiping direction. In a previous section, we showed that the average surface roughness of the solvent-wiped TPO substrate was less than 500 nm. In this CPO-coated solvent-wiped TPO sample, one can observe the increase of roughness between CPO and TPO upon CPO solution-coating and subsequent baking. Similar observations were made for the CPO-coated solvent-wiped TPO substrate under different baking conditions.

To obtain a better understanding of the CPO/TPO interface in the CPO-coated TPO sample, we needed to examine the CPO layer in cross-section. We initially prepared microtomed samples by cutting samples in a range of 20-30 [micro]m thick from the CPO-coated TPO. LCFM depth profiling indicated that the microtoming process distorted the sample. For example, the overall thickness of the CPO layer observed varied, depending upon where the image was taken. In some regions, the CPO layer was thinner than that determined by the optical sectioning method described in the following paragraph. We conclude that the process of room temperature microtoming distorts the sample and, as such, the results may be misleading.

A very useful way to obtain cross-section images in a nondestructive way is through z-section line scans. In this approach, one measures a series of intensities along a single line, through multiple scans, in which the depth of focus is advanced by 100 nm for each scan. These intensities are processed through software to construct a two-dimensional cross-sectional image. We used this approach to construct cross-section (side-view) images of CPO-coated TPO substrates. The side-view images obtained in this way for two CPO-coated TPO substrates are shown in Figure 6. In both images, the bright layers represent the CPO layer since CPO contains the fluorescent dye. Figure 6A shows a cross-section image of a CPO-coated unwiped TPO sample, as described in Figure 4; while Figure 6B presents a sideview image of a CPO-coated solvent-wiped TPO, as described in Figure 5. Although there is a rougher CPO/TPO interface in the CPO-coated solvent-wiped TPO than in the unwiped sample, it is difficult to see any evidence of CPO penetration into the TPO component. It is clear, however, that the characteristics of the transition region from the CPO coating to the TPO substrate, as discerned from the optical cross-sections, are completely consistent with those inferred from the multiple images seen in Figures 4 and 5.

[FIGURE 7 OMITTED]

In order to see if more information is available about the CPO/TPO interface, we constructed intensity profiles across this interface for many series of measurements. These fluorescence intensity profiles across the CPO layer traverse the distance from the water surrounding the water-immersion objective to the TPO substrate, as shown by the two white vertical arrows in Figures 6A and 6B. Since the resolution of the confocal microscope in the z direction is on the order of 0.5 [micro]m, we would not expect to resolve the features of an interface on a smaller length scale. Thus, we are looking for features of these profiles that indicate a transition zone not due to surface roughness larger than this distance. We present two such intensity profiles across the CPO layer, for a solventwiped sample, in Figure 7.

In Figure 7, we use dashed lines to delimit the z-axis distances corresponding to zero fluorescence intensity and 90% of the maximum fluorescence intensity. The left side of the image refers to the top of the sample, where we expect a sharp interface between the CPO layer and the water adjacent to the microscope objective. The finite width of this interface (denoted [D.sub.1]) is a measure of the resolution limits of the microscope. On the right side of the figure, there appears to be greater curvature in the transition from the region of high intensity into the dark TPO substrate. We define the distance between the lines at the CPO/TPO interface corresponding to zero and 90% of the maximum fluorescence intensity as [D.sub.2]. In Figure 7A, [D.sub.2] is significantly larger than [D.sub.1]. The increased width of the interface may indicate that some CPO diffused into the TPO substrate, accompanied perhaps by diffusion of some of the TPO into the CPO layer.

[FIGURE 8 OMITTED]

[FIGURE 9 OMITTED]

Since the inferences that we will attempt to draw in the following paragraphs are based upon data close to the resolution limit of the microscope, we must carefully consider the precision of each measurement. We begin with a discussion of [D.sub.1] values. For more than 30 measurements on CPO-coated unwiped substrates, we obtained [D.sub.1] = 0.68 [+ or -] 0.09 [micro]m. For the corresponding solvent-wiped samples, we found [D.sub.1] = 0.69 [+ or -] 0.08 [micro]m. These values establish the precision of the measurement and the resolution limits of the instrument in the z direction.

[D.sub.2] values showed larger dimensions and much larger scatter. For the unwiped substrates, we found [D.sub.2] = 1.01 [+ or -] 0.18 [micro]m ([D.sub.2]-[D.sub.1] = 0.33 [+ or -] 0.20 [micro]m), (16) and for the solventwiped substrates, [D.sub.2] = 0.96 [+ or -] 0.14 [micro]m ([D.sub.2]-[D.sub.1] = 0.27 [+ or -] 0.14 [micro]m). Inspection of individual images showed differences as well in the nature of the curvature seen in the intensity profile plots near the [D.sub.2] edges. In profiles in which [D.sub.2] is significantly larger than [D.sub.1] (Figure 7A), there is a soft downward curvature to the intensity profile to the [D.sub.2] side of the intensity maximum. In contrast, in profiles in which [D.sub.2] and [D.sub.1] are similar in magnitude, the profile between [D.sub.1] and [D.sub.2] is symmetrical (Figure 7B). We began to suspect that there was a significant site-to-site variability in the width of the CPO/TPO interface as measured by this technique. To follow up on this idea, we constructed histograms of the ([D.sub.2]-[D.sub.1]) differences for each sample. These results are plotted in Figure 8A for the CPO-coated unwiped substrates and in Figure 8B for the solvent-wiped substrates. In both figures, one sees that there is a range of ([D.sub.2]-[D.sub.1]) differences centered at 0.2 to 0.3 [micro]m. The similarity in the shapes of the histograms for the solvent-wiped and unwiped samples lends support to the idea that the fluorescence intensity profiles are not very sensitive to the interface roughness characteristic of the former set of samples.

We would like to interpret these differences in the ([D.sub.2]-[D.sub.1]) values as a measure of the breadth of the interface between the CPO and TPO polymers. In making this inference, we assume that analyses along vertical lines drawn perpendicular to the CPO/TPO interface, as shown in Figure 7, are insensitive to the roughness of the interface. Since polymer/polymer interfaces are normally much thinner (a few nm) than the values of ([D.sub.2]-[D.sub.1]) seen in Figure 8, the diffuse boundary region where CPO and TPO meet is more properly referred to as an interphase.

In a previous publication, (15) we described LCFM experiments on blends of this CPO, PP1042 and EBR9. We found that in ternary blends containing 2.5 wt% CPO, this component completely wetted EBR droplets dispersed in the PP matrix. In addition, the CPO showed partial miscibility with EBR. Analysis of the image droplets, in a manner similar to that employed in Figure 7, showed that the EBR/CPO interphase was substantially more diffuse than the PP/CPO interface. In the blend experiments, there was much less variation in the width of the boundary region between CPO and EBR than one finds here for CPO and TPO. Based on these findings, we can explain the results obtained here if we assume that the surface of the TPO has regions in which the PP component is at the surface and other regions where the EBR component is at the surface. In other words, we envision the surface of the TPO to be patchy. In terms of this hypothesis, cross-sectional images analyzed in regions in which the CPO layer is adjacent to EBR exhibit interphases with dimensions on the order of 0.3-0.5 [micro]m, and other sections analyzed in regions in which the CPO layer is adjacent to PP exhibit sharp interfaces. These ([D.sub.2]-[D.sub.1]) values are small and close to the resolution limit of the LCFM instrument. Nevertheless, they appear to be significantly different. Rather than drawing a firm conclusion about our sample, we prefer to put forward this result as a working hypothesis for further investigation. A drawing depicting this type of morphology for the CPO-coated TPO is shown in Figure 9.

To put our results in proper context, we note that there have been a number of studies published previously about the interaction of CPO with TPO. The thickness of the CPO/TPO interface in a CPO-coated TPO system was investigated by different groups. Mirabella et al. (11) attempted to place upper and lower bounds on the thickness of the CPO/TPO interface in a TPO/CPO/paint system by combining a theoretical model and experiments that included scanning transmission X-ray microscopy (STXM). In their system, they predicted a lower bound of 11 nm for the interface thickness, and an upper bound of about 400 nm. From the STXM measurements, they estimated the thickness to be 340 [+ or -] 80 nm. Ellis (10) calculated an interface thickness for polypropylene and chlorinated polypropylene from the estimated Flory-Huggins chi-parameter for the two polymers. In this way he calculated a value of about 8 nm. This type of interface would be considered to be broad from the perspective of strongly segregated polymer blends, but very narrow from the perspective of the broader interphases inferred for CPO and TPO.

Ryntz, in collaboration with Treado, (5, 17) carried out several experiments intended to examine the penetration of CPO into the TPO matrix. These experiments employed a lower modulus TPO with ethylene-propylene rubber (EPR) as the impact modifier and a CPO with a higher chlorine content than the polymers described here. Confocal Raman experiments provided support for [micro]m penetration of the CPO into the TPO itself. LCFM experiments using Nile Red as a free dye provided evidence that the solvent used in the CPO coating step, coupled with the post-application bake step, led to a change in the impact modifier morphology near the TPO/CPO interface. Nile Red is more soluble in EPR than in PP, and, as a result of CPO coating, droplets of EPR, with diameters ranging from 3 to 6 [micro]m, appeared to bloom to the TPO surface. The authors also commented that they found some evidence for penetration of the EPR up to 2 [micro]m into the CPO layer and some droplets of EPR at the CPO surface.

In an earlier study, (3) there were more ambiguous results with respect to the depth of CPO penetration into the TPO. Time-of-flight secondary-ion mass spectrometry (TOF-SIMS) experiments, intended to measure the profile of chlorine atoms as a function of depth, suggested that there may be a diffuse nature to the CPO/TPO interface, but the information was at the limits of spatial resolution of the experiment. They also carried out LCFM experiments using a CPO labeled with a fluorescent dye (ANS). These experiments gave some evidence for a weak diffuse fluorescence beneath the otherwise (visually) sharp interface between the CPO and the TPO layers. The images examined were obtained by microtoming, and the authors of this article commented that there was some sample distortion caused by microtoming, consistent with the difficulties we have experienced here. Perhaps the most striking result from this paper, which deserves further investigation, is the patchiness of the CPO layer on the TPO as seen in top-view images of the coated substrate (Figures 16 and 17 of reference 3). While more detail would be welcome, these images suggest that the as-applied CPO (which was sprayed on) did not fully wet the TPO substrate, and that baking the sample induced spreading of the CPO over a greater fraction of the TPO substrate.

In this article, we describe results for TPO coated by a drawdown process with a CPO solution. In every instance we obtained films in which the TPO was completely coated with CPO. With no special precautions beyond the use of the drawdown bar, the coating thickness was uniform to within ca. 1 [micro]m. In the experiments of Ryntz and Treado, (5,17) which also employed confocal microscopy, there was also no indication of incomplete coverage or dewetting of the CPO coating on TPO. For these model experiments, one uses a CPO solution normally without the additives present in an industrial formulation. For example, commercial CPO contains carbon black to impart electrical conductivity to the part prior to painting. Under these circumstances, it might be easier to detect if a CPO coating gave imperfect wetting of a TPO substrate than if a model study created a clear CPO overcoat.

SUMMARY

We have described experiments that examine the nature of a CPO coating on a high-modulus TPO fabricated as a blend of a highly crystalline Ziegler-Natta polypropylene (PP) and a crystalline poly(ethylene-co-butene) (EBR9) impact modifier. The CPO itself had a relatively low chlorine content (20 wt%). The CPO was doped with a small amount (2.5%) of a similar CPO to which a fluorescent dye was covalently attached.

One variable that we investigated was the influence of solvent wiping on the TPO substrate surface and on the CPO/TPO interface. Surface profilometry measurements indicate that surface wiping increased the rms roughness of the TPO itself to a value of 500 nm. Model experiments suggest that the PP is unaffected by solvent wiping, whereas the EBR9 is affected in much the same way as the TPO itself. More impressive is the influence of solvent wiping on the after-coating, after-bake roughness of the CPO/TPO interface. CPO coated onto unwiped TPO had a smooth, straight profile on the scale of 1 [micro]m, as seen in LCFM images. In contrast, the CPO/TPO interface, as seen in side-view, for the solvent-wiped substrate showed considerable roughness.

Detailed investigation of the fluorescence intensity profiles across the CPO layer showed that the interface between the CPO and TPO was more diffuse than the sharp interface we found at the surface of the CPO layer. It was difficult to establish precisely the thickness of the CPO/TPO interface, and this value appeared to vary with the site at which the measurement was made. These results could best be explained in terms of a model for the surface morphology of the TPO in which both PP and EBR domains existed. In regions of the sample where CPO was in contact with PP, the interface was sharp, and in regions of the sample where CPO was in contact with EBR9, the interface (interphase) was more diffuse, with dimensions on the order of 0.5 [micro]m. This result was consistent with findings described in reference 15, that this particular CPO exhibited partial miscibility with EBR9.

ACKNOWLEDGMENTS

The authors would like to thank Visteon Company and Nippon Paper Ind. Ltd. for kindly providing samples for this study. We also thank NESRC and Material and Manufacturing Ontario (MMO) for financial support of this research.

References

(1) Polymer Surfaces: From Physics to Technology, Garbassi, F., Morra, M., and Occhiello, E. (Eds.), Wiley, New York, Chap. 10, 1994.

(2) Prater, T.J., Kaberline, S.L., Holubka, J.W., and Ryntz, R.A., "Examination of the Distribution of TPO Adhesion Promoter Material in a Painted TPO System," JOURNAL OF COATINGS TECHNOLOGY, 68, No. 857, 83-91 (1996).

(3) Ryntz, R.A., Xie, Q., and Ramamurthy, A.C., "Effects of Coating Solvents on the Morphology of Thermoplastic Polyolefins (TPO)," JOURNAL OF COATINGS TECHNOLOGY, 67, No. 843, 45 (1995).

(4) Ryntz, R.A., Ramamurthy, A.C., and Mihora, D.J., "Thermal and Impact Induced Stress Failure in Painted TPO: The Role of Surface Morphology," JOURNAL OF COATINGS TECHNOLOGY, 67, No. 840, 35 (1995).

(5) Morris, H.R., Munroe, B., Ryntz, R.A., and Treado, P.J., "Fluorescence and Raman Chemical Imaging of Thermoplastic Olefin (TPO) Adhesion Promotion," Langmuir, 14, 2426 (1998).

(6) Schmitz, P.J. and Hoiubka, J.W.J., "Investigation of the 'Surface' and 'Interface' Composition of Adhesion Promoter/Thermoplastic Olefin Systems: The Effect of Adhesion Promoter Bake Temperature," J. Adhes., 48, 137-148 (1995).

(7) Ryntz, R.A. and Buzdon, B., "The Relationship between Clearcoat Permeability and Adhesion Promoter Penetration to Gasoline Resistance of Painted TPO Substrates," Prog. Org. Coat., 32, 167-172 (1997).

(8) Ryntz, R.A., "The Effect of Thermoplastic Poly(olefin) (TPO) Morphology on Subsequent Paintability and Thermal Shock Performance," Prog. Org. Coat., 27, 241-254 (1996).

(9) Clemens, R.J., Batts, G.N., Lawniczak, J.E., Middleton, K.P., and Sass, C., "How Do Chlorinated Poly(olefins) Promote Adhesion of Coatings to Poly(propylene)," Prog. Org. Coat., 24, 43-54 (1994).

(10) Ellis, T.S., "Interaction of Chlorinated Polyolefins with Ethylene-propylene Copolymers and Their Relevance to Painted TPOs," Polym. Eng. Sci., 41, 2065-2072 (2001).

(11) Mirabella, F.M., Dioh, N., and Zimba, C.G., "Theoretical Analysis and Experimental Characterization of the TPO/Adhesion Promoter/Paint Interface of Painted Thermoplastic Polyolefins (TPO)," Polym, Eng. Sci., 40, 2000-2006 (2000).

(12) Tomasetti, E., Vandorpe, S., Daoust, D., Boxus, T., Marchand-Brynaert, J., Poleunis, C., Bertrand, P., Legras, R., and Rouxhet, P.G., "Diffusion of an Adhesion Promoter (Chlorinated Polypropylene) into Polypropylene/Ethylene-propylene Copolymer (PP/EP) Blends: Methods of Quantification," J. Adhes. Sci. Technol., 14, 779-789 (2000).

(13) Raphael, E. and de Gennes, P.G., "Rubber-Rubber Adhesion with Connector Molecules," J. Phys. Chem., 96, 4002-4007 (1992).

(14) Information supplied by Nippon Paper Industries Cl. Ltd., 2001. (www.npaper.co.jp/html/chem/english)

(15) Ma, Y., Farinha, J.P.S., Winnik, M.A., Yaneff, P.V., and Ryntz, R.A., "Compatibility of Chlorinated Polyolefin (CPO) with the Components of Thermoplastic Polyolefin (TPO): a Study by Laser Scanning Confocal Fluorescence Microscopy," Macromolecules, 37, 6544-6552 (2004).

(16) Calculated from the expression [DELTA] ([D.sub.2]-[D.sub.1]) = {([DELTA][D.sub.1])[.sup.2] + ([DELTA][D.sub.2])[.sup.2]}

(17) Morris, H.R., Turner, J.F. II, Munro, B., Ryntz, R.A., and Treado, P.J., "Chemical Imaging of Thermoplastic Olefin (TPO) Surface Architecture," Langmuir, 15, 2961-2972 (1999).

Yuechun Ma and Mitchell A. Winnik -- University of Toronto*

Phillip V. Yaneff -- E.I. Dupont Canada([dagger])

Rose A. Ryntz -- Visteon Automotive Systems**

* Department of Chemistry, 80 St. George St., Toronto, Ontario, M5S 3H6 Canada.

[dagger] 377 Fairall St., Ajax, Ontario, Canada, L1S 1R7.

** 401 Southfield Rd., P.O. Box 6231, Dearborn, MI 48121.