Solid-state NMR study of aging of Colorbond polymer coating
Introduction
Polymer coatings, including nitrocellulose, polyester, latex and acrylic (a polymer of acrylates, methacrylates and similar monomers), are widely used as paints. The top coat is required for esthetic qualities (color and gloss) and as a protective layer since it interacts with the environment and must provide resistance to the conditions encountered including irradiation, oxidation, hydrolysis and the action of aggressive fluids. Coatings are normally exposed to the environment, leading to changes in density, void space, permeability, elasticity and resistivity [1]. These changes take place with time and are known as aging of coatings.
Chemical aging of coatings involves the process of oxidative degradation. Coating structures should be designed to reduce the absorption of short wavelength solar radiation so as to avoid the generation of radicals and photochemical reactions, which speed up the aging rate of polymer coatings. Physical aging is due to the slow changes that occur in coatings with time. When amorphous polymers are cooled through their Tg (glass transition temperature), they solidify and stiffen to form glasses. In this glassy state, the rate of structural relaxation becomes very slow and the polymer conformations are fixed in a non-equilibrium state. If the temperature is held at some value below Tg, the process of relaxation towards an equilibrium configuration continues and as a result many material properties change with time. The process of physical aging is highly temperature dependent, and is fastest at temperatures closest to Tg.
Chemical and physical changes due to aging may be detected by nuclear magnetic resonance (NMR) spectroscopy [2], [3], [4]. Although solution-state NMR provides the most detailed structural information, solid-state NMR provides important information about internal molecular motions and dynamics and is applicable to insoluble samples such as cured films and paints [5], [6]. In the present work, we report changes in NMR parameters of Colorbond polymer coatings with age. Colorbond samples were chosen because of their industrial relevance to current production techniques and their improved product durability. Colorbond is a polyester paint designed to withstand the harsh outdoor Australian climate and is subject to extremes in heat and humidity. Although previous studies have reported on the aging of polymers [7], to date no work has been reported in the open literature on the aging of this class of polymers, which are directly bound to the surface.
The mobility of groups within the polymer can be measured using solid-state 13C NMR techniques. The polymer basically consists of a carbon backbone surrounded by protons. These protons interact with each other via a dipole–dipole interaction resulting in broad resonance lines for solid samples and, since protons have a small chemical shift range, resonance lines overlap in proton NMR spectra. 13C–13C interactions, however, are rare since 13C abundance is only 1.108% and, since 13C nuclei have a large chemical shift range, better spectral resolution is obtained. 13C–1H interactions are particularly important since a 13C spectrum can be obtained by transferring magnetization from protons to the less abundant isotope, 13C. This enhances the 13C signals and is called cross polarization (CP). CP allows faster signal acquisition since the longitudinal relaxation times, T1, of 1H are shorter than 13C [7], [8]. Spectral resolution is enhanced by magic angle spinning (MAS), which involves placing a rotor containing the sample at an angle 54.7° to the magnetic field and spinning at several kHz. The line broadening caused by chemical shift anisotropy and dipolar interactions is reduced [9], [10].
Molecular motion can be measured using the NMR relaxation times T1, T1ρ and T2. T1 is the spin–lattice relaxation time and relates to fast molecular motion on the ω0 time scale or Larmor frequency (MHz). Spin–lattice relaxation involves a loss of energy by the excited nucleus to the surrounding environment or lattice [7], [8], [11]. The rate at which this relaxation process occurs is measured by the spin–lattice relaxation time (1/T1=rate). The closer the motional frequencies of the lattice are to the Larmor frequencies of the nuclei, the more efficient is the transfer of the magnetization (i.e. the shorter the T1). T1 measurements are usually limited to molecular motions in the 107 Hz frequency range.
Both proton and carbon T1 measurements are reported in this work. 1H T1 gives average information about dynamics from all the groups in the molecule since protons are ≈100% abundant [12]. The 13C T1 gives information about the dynamics of individual groups on molecules since the 13C nuclei are rare and there is little relaxation from other 13C nuclei [10], [13].
T1ρ is the spin–lattice relaxation time in the rotating frame [12]. It is similar to T1 in the static field but due to spin–lattice interactions in the rotating frame. In CP experiments, the magnetization initially builds up to a maximum due to the carbon–proton dipolar interactions in the spin-locking field, ω1, and decays exponentially to equilibrium with the time constant T1ρ. The values of T1ρ are used both to measure the molecular dynamics on the kHz time scale of ω1 and to report on the length scale of polymer mixing [14], [15].
T2 is the spin–spin relaxation time and is the time constant for the decay of the precessing X–Y component of the magnetization [7], [8], [11]. T2 gives information about the distribution of resonant frequencies and about the local fields experienced by the magnetic moments of the nuclei. In liquids, the local magnetic fields fluctuate very rapidly and, therefore, quickly average to ≈0 to yield long T2s and narrow resonance lines. The atoms in solids are in nearly fixed positions, and the internal fields are significant, which results in the rapid loss of coherence and short T2s (microseconds). T2 is related to the inverse of the resonance line width and thus T2 in solids is short and the resonance lines are broad.
Changes in the NMR relaxation times T2, T1ρ and T1 report on changes in molecular motion from slow to fast timescales. Colorbond polymer coatings were exposed for different periods in a weathering test situation and the effect on NMR relaxation times was studied.
Section snippets
Materials and methods
Two samples of polymer coatings, Merino and Off-White Colorbonds, were obtained from BHP Research Laboratory (Melbourne, Australia). Merino and Off-White polymer coatings are polyester paints whose composition includes, diols, diacids, triols, and cross-linkers. The monomer components (diol, diacid and triol) are heated from 60 to 120°C to polymerize. The liquid paint, mixture of polymer, solvent, cross-linkers, catalyst and blocker, are reacted together for a short time at 235°C to produce the
NMR chemical shift assignments
An example of a 13C CPMAS spectrum of a polymer sample, spinning at 8 kHz is shown in Fig. 2. The 13C NMR spectra of both Merino and Off White polymer coatings were similar with six main carbon chemical groups that were assigned as listed in Table 1.
Line width and T2
Since the value of T2 is related approximately to the inverse of line width, the line widths of individual groups from the 13C NMR spectrum were measured for fresh and aged Off-White coating. T2 estimated from the line width for all groups other than
Conclusions
In summary, there are only minor changes in 1H T1 and T1ρ between the fresh and aged Colorbond polymer coating. Similarly, no changes in line widths between fresh and aged coatings were seen. On the other hand, a change in 13C T1 values was observed between fresh and aged Colorbond polymers for carboxyl carbons adjacent to aromatic rings The carbon network was more sensitive to changes in intermolecular distances as a result of ‘shrinkage’ of the coating induced by weathering.
Acknowledgements
Special thanks to Kate Barbour (Dulux, Australia) for help with the QUV accelerated weathering tester.
References (18)
Polymer
(1999)- et al.
Polymer
(2000) J Magn Reson
(1978)- et al.
Polymer
(1998) Chemistry and physics of coatings
(1961)- et al.
Macromolecules
(1999) - et al.
Macromolecules
(1998) - et al.
Macromolecules
(1998) Spectroscopy of polymers
(1992)
Cited by (7)
Hypercrosslinked additives for ageless gas-separation membranes
2016, Angewandte Chemie - International EditionStabilization of a supersaturated solution of mefenamic acid from a solid dispersion with EUDRAGIT® EPO
2012, Pharmaceutical ResearchNMR structural elucidation of amino resins
2004, Journal of Applied Polymer ScienceChemistry of combustion products: Investigation of combustion chamber deposits formed in a gasoline direct injection stratified charge SI engine
2004, Doktorsavhandlingar vid Chalmers Tekniska Hogskola