It has been emphasized above that a surface molecule on a rough surface will often have a different environment—for example, fewer nearest neighbors, more ‘‘broken bonds’’— than a similar molecule on a smoother surface. In addition to this, it must be remembered that most, if not all, of the chemical or physical treatments used to produce a rough surface will also alter the chemical nature of the surface molecules. There are many reports in the literature of treatments which produce both mechanical and chemical effects. Sometimes these are seen as supplementing, sometimes as opposing each other.
Zhuang and Wightman’s work on carbon fiber-epoxy adhesion provides a recent example [74]. They studied both the surface topography and the surface chemistry of carbon fibers modified by treatment with an oxygen plasma prior to incorporation into a epoxy matrix. Two types of fibers, differing in surface roughness, were studied. An increase in surface oxygen content was observed on treatment, mirrored by increases in the polar component of surface energy and in interfacial shear strength (IFSS). Here the rougher fibers had somewhat lower IFSS. The lower adhesion was associated with incomplete filling by the resin of valleys on the fiber surface striations. However, there is evidence that the rougher surface imparts better durability in a humid environment.
PTFE is a notoriously difficult substrate to bond, but severe treatment producing both roughening and surface chemical changes have been found to ease the difficulty. Recently, Koh et al. have used argon ion irradiation as a pretreatment both in the presence and absence of oxygen [54]. The treatment produced increasing roughness, eventually giving a fibrous forestlike texture (Fig. 4(b)). These treated surfaces were bonded with a thermoplastic adhesive cement, and generally considerably enhanced adhesion was found. The level of adhesion appeared to rise to a peak, which occurred at a treatment level of 1016ions/cm2.
High-resolution X-ray photoelectron spectra showed chemical changes also occurring. In the absence of oxygen, a 285 eV (C-C and C-H) peak developed with maximum intensity at a dose of 1016ions/cm2. In the presence of oxygen a strong O 1s signal developed which was attributed to the reaction of oxygen atoms with the free radicals created by argon ion bombardment. Here again, the enhanced adhesion is attributed to a combination of improved wettability and chemical reactivity of the surface, combined with mechanical keying to the increasingly rough surfaces. There is
no convincing explanation of the fall in adhesion at the highest treatment time. It would be interesting to know at what level the difference between adhesion at 1016 land 1017ions/ cm2 was statistically significant.