The aim of the study described in this thesis is to obtain depth – resolved insight into the film formation of alkyd-based coatings. In this project, confocal Raman microspectroscopy (CRM) measurements and the mechanical characterization techniques Atomic Force Microscopy (AFM) nanoindentation, tensile testing and dynamic mechanical thermal analysis (DMTA) were combined.
There is an increasing demand for using water-borne coatings due to environmental and health concerns. For this reason water-borne alkyd emulsion coatings are becoming ever more important in the coatings market. Knowledge about film formation after application and the build-up of mechanical and chemical properties during cure of water-borne alkyds is important for ensuring a competitive performance of these coatings and for instance allows the development of improved cross-linking binders. Thus, there is a need for sensitive, quantitative techniques to measure the extent of cross-linking as a function of time and also as a function of the position within the drying film. There is also a need to understand and quantitatively describe the profile and temporal spatial development of the drying front within the coating.
Chapter 2 serves as a general introduction to organic coatings. The chemical drying process of alkyd coatings is described in this chapter. Subsequently, the temporal development of mechanical properties related to cross-link density is discussed. Finally, the main measurement methods used to investigate the film formation of alkyds, confocal Raman microspectroscopy and AFM nanoindentation, are introduced.
In Chapter 3, the results of the successful application of CRM for the investigation of the oxidative cross-linking as a function of film depth in drying alkyd-based coatings are described. Raman microscopy offers a powerful combination of spatial resolution and chemical characterization. This technique provides new insights into the network formation of alkyds. In addition, results shown were obtained in a non-invasive fashion. During the chemical drying, double bonds in the fatty acid chains react with oxygen, resulting in their disappearance. The consumption of double bonds in the alkyd films is monitored in depth by CRM. The results show that the chemical cross-linking is not homogenous in the depth of the coating. Different drying profiles were obtained for the long and short oil alkyd films, respectively. Adding thickener to the coating and changing the water into organic solvent has a similar influence on the drying process: both plasticize the coating film, thus accelerate drying. Based on the measured concentration profiles and the spatial laser beam intensity distribution along the film thickness, the real concentration profiles were calculated. The shape of the drying profiles suggests that there is oxygen penetration control over the cross-linking process.
Chapter 4 describes an attempt to quantitatively describe the measured depth profiles of the C=C concentration during oxidative drying of alkyd coating films. The depth profiles are described using a quantitative model borrowed from chemical process technology which predicts concentration changes in liquid films during chemical reactions with gasses, accompanied by mass transfer of gas molecules. The experimental data presented in Chapter 3 are complemented by model calculations using parameters which are either measured or taken from literature. The model encompasses partial differential equations. First the influence of the model parameters (oxygen solubility, diffusion coefficient of oxygen, and the rate constant of double bond disappearance) on the consumption of double bonds is assessed numerically. In order to obtain a better match between the calculated and measured C=C consumption curves, the oxygen solubility was allowed to vary. The simulation model shows that the drying profiles are the result of a competition between oxygen penetration and chemical oxidation. Effects of the solvent, thickener and oil length on the calculated profiles are assessed. The numerical model can also be used to predict drying profiles in alkyd films.
In Chapter 5, tensile testing was used to determine the elastic modulus of free coating films. The elastic modulus of an elastic network is proportional to the cross-link density. Thus, based on the CRM results, a model was developed to give information on modulus variation in the depth of the film during the drying process, on the completely dried film thickness at a given time and on the maximum possible thickness of the completely cross-linked coating film. The outer layers of the films become completely cross-linked after a very short time, compared to the time of complete drying. A threshold thickness was experimentally determined, beyond which complete network formation becomes practically impossible (on the time scales of this study). Two assumptions were made: the first takes the elastic response as proportional to the accomplished degree of drying, the second assumes that the modulus shows a parabolic distribution with the distance from the midplane for the partially dried coating at room temperature. The second assumption was supported by confocal Raman microscopy results. From these assumptions a set of polynomial relationships were derived whose parameters were evaluated from the fitting of a reduced set of experimental data. This allowed a reasonable reproduction of the drying-stiffening behavior of the coatings observed experimentally.
In Chapter 6, AFM nanoindentation experiments enabled us to measure Young’s modulus on the nanometer scale on the film surface. During nanoindentation, a given load is applied to an indenter (in this case the AFM probe) in contact with the sample, making an indentation into the surface. During the indentation a force - distance curve is recorded which contains information on the applied load and the depth of the penetration. With a known contact area and based on the Sneddon and Hertz models, the value of the Young’s modulus can be calculated from the slope of the unloading curve. This technique relies on the facts that film deformation recovered during unloading is largely elastic, and during the initial withdrawal of the AFM probe the contact area remains constant.
Most commercial alkyd coatings are in the glassy state at room temperature. In the glass transition region the modulus drastically decreases and above Tg the coatings are in a rubbery phase. There is a significant difference between the viscoelastic behaviour of long and short oil alkyds. The long oil alkyds attain a higher final cross-link density both on the surface and in the bulk. Adding thickener to the coating does not cause a significant change in the mechanical behaviour of the alkyd coatings. The solvent does influence the mechanical properties of the final films. Solvent-borne alkyds have a higher cross-link density than water-borne ones.
AFM nanoindentation provides a unique possibility to measure time dependent mechanical behavior of alkyd coatings on the surface of the films. Using simple time – temperature superposition, it was possible to create modulus master curves from AFM nanoindentation data. The master curves obtained allow the viscoelastic behaviour of alkyd coatings to be determined at frequencies outside the practical measurement ranges at constant temperatures.
Measured values of elastic moduli using DMTA and AFM are within comparable range. The absolute values of the measured E moduli differ but are in the same order of magnitude. The AFM nanoindentation method at present is primarily suitable to establish trends, but there are promising results opening the possibility to use it as a quantative measurement after further optimizing the experiment.
The aim of Chapter 7 is to establish a correlation between the consumption of C=C bonds and changes in mechanical properties of the alkyd films studied. During the drying process, simultaneous information is obtained on chemistry, using CRM (described in Chapter 3), and on mechanical properties using AFM nanoindentation (described in Chapter 6), DMTA (described in Chapter 6) and tensile testing (described in Chapter 5). The CRM results are correlated with the mechanical properties of alkyd resins of long and short oil lengths. A combination of the results allows us to quantitatively monitor the advancement of the drying process within the coating films. The relationship between the C=C bond consumption and the cross-link density was found to depend on the oil length of the alkyd sample. However, for all samples it was shown that free radical addition to double bonds was not the only contribution to cross-link formation. In addition to this reaction, cross-links resulting from the combination of radicals, as well as mechanical cross-links (entanglements) contribute to the overall cross-link density.
By measuring films of different thickness both on the surface and on the lower side (in direct contact with the substrate) of the film, we could estimate the gradient in the elastic modulus in the depth of the film during the cross-linking process.