Raman spectroscopy in coatings research and analysis: Part II. Practical applications

JCT CoatingsTech, Sept, 2005 by Neil J. Everall

In this two-part tutorial, the use of Raman spectroscopy for characterizing and testing coatings is examined. Part I covered the type of information that can be obtained using Raman spectroscopy (August 2005 JCT CoatingsTech, page 38), and described how a coating is studied in a typical Raman experiment. Part I also discussed the thickness of the coating to be studied, and the influence of the substrate. In this section, we focus on examples that illustrate the capabilities and limitations of Raman spectroscopy.

APPLICATIONS AND EXAMPLES

Compositional Mapping

Perhaps the most obvious application of Raman spectroscopy to coating technology is compositional mapping, whereby a map or image of the composition of a coating can be built up in one, two, or three dimensions. Raman mapping involves sequential acquisition of Raman spectra as a laser beam is incrementally rastered over a sample. An image can then be constructed based upon the intensity (or some other parameter, such as position or width) of a Raman band as a function of position on the sample. (Global) Raman imaging involves defocusing the laser to illuminate a large, 2D area of a sample, and then sequential acquisition of images at distinct Raman frequencies (selected using a tuneable filter). In principle, this is an attractive approach, and for simple, stable systems, where it is only necessary to record images at a few (well-separated) Raman frequencies, it can be very fast. However, if many Raman bands have to be analyzed to understand the coating chemistry, then it can take a long time to acquire images at all of the necessary frequencies. In addition, Raman imaging requires a high power laser beam (to generate a reasonable power density at the sample when the beam is defocused), and it turns out that samples are more prone to thermal decomposition under global illumination than point illumination, due to the difficulty in conducting heat away from a "sheet" of hot material at the sample surface. There are also hybrid technologies available, where the laser is focused to a line on the sample; spectra are acquired simultaneously at each point along the line, using a spectrograph/CCD (ChargeCoupled Device) camera combination, then the line is moved to a new point on the sample. Approaches to Raman mapping and imaging have been summarized in detail elsewhere. (18-20,33)

Many of the examples discussed in this tutorial involve measuring spatial variations in a property, so only a few examples of compositional mapping and imaging will be given here. Roberts and Evans studied the influence of manufacturing variables on the surface quality of paper laminates impregnated with melamine-formaldehyde (MF) and urea-formaldehyde (UF) resins. (34) They used Raman microscopy to map the MF distribution as a function of the amount of UF that was added to the system. They showed that if the UF level was too low, then the MF tended to fill voids in the center of the laminate, leaving the surface MF-deficient and causing visible defects. Increasing the amount of UF prevented ingress of the MF into the body of the laminate, thereby increasing the MF concentration at the surface, giving fewer defects. Larsson et al. used confocal Raman axial profiling with a water-immersion objective to map ligand distributions in surface-treated chromatographic adsorbent particles. (35) They showed how accurate axial profiles of allyl, sulphopropyl, and dextran concentrations could be obtained from particles immersed in water. This proved that the sulphopropyl groups were confined to thin (~20 [micro]m) shells on the surface of 100 [micro]m diameter bead, and allowed coating thickness to be assessed for different particle types. The power of global Raman imaging was shown in a detailed study of the near-surface composition of a thermoplastic olefin (TPO) that was sprayed with a chlorinated polyolefin (CPO) to form a primer layer for subsequent painting. (36) The TPO was a blend of ethylene-propylene rubber (EPR) and poly(propylene) (PP). The Raman image showed separation of the TPO in distinct EPR and PP phases, with the EPR and CPO being co-localized near the surface. The authors suggested that the solvent carrier for the CPO spray was responsible for inducing separation of the TPO into its EPR and PP components. This work also highlighted the need for multivariate data processing to analyze the large volume of data that is generated in spectroscopic images (Figure 9).

Crystallinity in Polymeric Coatings

On occasion, one needs to characterize crystallinity gradients in polymer films and coatings. For example, in order to achieve good adhesion between the polyester coatings and metal substrates used in can production, it is important to control the polymer crystallinity so as to retain an amorphous layer near the metal surface, while maintaining a reasonable level of crystallinity elsewhere to avoid compromising the mechanical performance of the coat. The critical changes can occur within just one or two microns of the metal surface, so excellent spatial resolution is required. Figure 10 shows a Raman line map, taken at 1 [micro]m increments through a cross-section of a steel-polyester laminate. As one approaches the metal, the crystallinity falls dramatically (decrease in intensity of the 1096 [cm.sup.-1] band over a distance of just 2 [micro]m to yield amorphous polymer at the metal interface). This result confirms that both the coating and the can-forming processes have been controlled to produce an acceptable film structure both in the bulk and at the interface. Similarly, using polarized confocal Raman microscopy, it is possible to map changes in polymer molecular orientation rather than crystallinity, because the intensity of a Raman band depends on the polarization of the laser and Raman fields and the orientation of the scattering unit. (37) We have used this approach to map gradients through the thickness of uniaxially-oriented PET films. (38)