Comparison of phase structures and surface free energy values for the coatings synthesised from linear polyurethanes and from waterborne polyurethane cationomers

The basic chain structure of the synthesised linear polyurethanes may be illustrated by the following formula:where:

The chain structures of the synthesised cationomers result from the molar ratios assumed for the reacting substances at successive stages of the polyaddition process, and they are more complex than above. They can be presented by the following molecular formula display in Scheme 1:

The chemical structures of the synthesised polyurethanes were confirmed by spectroscopic methods (an accurate structural analysis of cationomers at the similar structure by using ¹H, ¹³C and ¹⁹F NMR spectrometry was presented into our earlier works [22, 28]).

Table 1 provides same diffractive reflections which were noted for the analysed polyurethane coatings, and Fig. 1 shows an example of the WAXS diffraction pattern for linear polyurethane no. PU-6. The collected data indicate that polyurethanes PU-1, PU-2, PU-3, PU-6 and PU-9 exhibit a relatively high level of crystallinity, which relates to the coats obtained from the solutions both in 1,4-dioxane and in chlorobenzene. As opposed to that, the coats applied as aqueous dispersions of cationomers are generally amorphous. This is in line with the findings from our earlier research, as presented in [22]. That was also the case of cationomer coats KPU 5–8 which contained TFBD-derived fluorine segments. The amorphous phase makes the dominant phase in all those samples. The WAXS diffraction peak is composed of two sub-peaks: one of them is clearly amorphous, its full width at half maximum (FWHM) is higher than 10°, and the other is lower and narrower, which is a suggestion for some orderly arranged structures. There are also differences between samples within the size of their FWHM peaks. Those differences, however, are not sufficiently significant to support the conclusion that the presence of fluorine atoms controls the formation of crystalline phases in the synthesised cationomers.

Since the changes in specific heat were small and there were relaxation effects observed, it was hard to obtain the value of T _g directly from classic DSC curves—as was in the case of polyurethanes (Fig. 2a) and cationomers (Fig. 2b). The polyurethane coatings perform like elastomers, and they show two glass transition points: T _g1 < 0 °C for soft segments and T _g2 > 0 °C for hard segments. Hard segments are definitely superior in numbers in the studied polyurethanes, and they are formed by urethane groups (derived from a specific diisocyanate and BD, TFBD, N-MDA or N-MDA) and by urea groups which are present in relatively lower numbers (and which are derived from MDI and HMDA). Within the interactions between these segments, hydrogen bonds take the essential role in putting some order to the structure. Hence, the determined values of relaxation enthalpy may be incidental to a high degree. On the other hand, however, some clearly regular performance can be observed in the measured T _g values which would represent the glass transition of soft segments (there is no T _g1). That can be accounted for by a relatively low number of polyether-derived flexible segments in cationomers (the first equation on Scheme 1). Moreover, they must be separated from each other in fluorine-containing cationomers by rigid segments which are derived from MDI and TFBD.

The thermograms as shown in Fig. 2b reveal some increase in T _g2 for cationomers which contain 0% to 15% fluorine added with the polyether component. That increase speaks for some stiffening of the amorphous structure (Table 4). The stiffness can make it difficult to freely form the coatings from process solutions, which seems supported by the increased relaxation enthalpy value as observed for test samples KPU-5, 6 and 7.

The phase structure was additionally analysed by the atomic force microscopy (AFM) method. Figure 3 presents various AFM images for PU-3 with about 20% of crystalline phase that results from WAXS analysis. The crystalline phase agglomerations can be clearly observed against the amorphous polymer matrix. As opposed to that, AFM analyses of cationomer coatings revealed generally the presence of inhomogeneous amorphous phase, which resulted from limited miscibility of hard and soft segments. However, “islands” of hard segments with dimensions 50–100 nm (Fig. 3b, c) with sharply outlined shapes were visible even in those samples against the background of the pretty homogeneous phase of soft segments.

The physical state of the analysed area with AFM method, especially coarseness, is very important for getting reliable results of contact angle measurements. AFM analyses (Fig. 3a) demonstrate that irregularities (hollows and tops) are in some places big, but do not exceed the ±20 nm value, which allows to perform solid SEP measurements.

The question remains open whether the differences in supermolecular structures of linear polyurethanes and cationomers as presented above affect the surface structures of coats, and if they do—whether those effects can be observed in their SFE values. That problem is complex since the superficial structure is usually different from the internal structure of the polymer material. In this specific case, the analysed coats had the thickness of about 50 μm. One may hence assume that the internal phase and the strictly superficial layer, which is available for the SFE measurements with the use of model liquid drops applied to the materials under investigation, interpenetrate each other to a certain degree. The phase structure of a coating, i.e. also its stiffness, should thus affect the values of SFE parameters to be determined. When analysing the phase structure, one should pay attention to the nature and values of chemical interactions (hydrogen bonds, ionic interactions) and to physical interactions: polar and dispersion effects. In order to give consideration to both those aspects, the SFE value was calculated with the use of two complementary methods: by van Oss–Good and by Owens–Wendt, which utilise the same measured values of the wetting angle Θ to calculate the values of various components of the parameter γ _S: \( \gamma_{\text{S}}^{\text{LW}} \) and \( \gamma_{\text{S}}^{\text{AB}} \) (van Oss–Good method) or \( \gamma_{\text{S}}^{\text{d}} \) and \( \gamma_{\text{S}}^{\text{p}} \) (Owens–Wendt method). In order to authenticate the findings, different sets of model liquids were used. The characters of interactions in polyurethanes and in cationomers could be compared in this way. That turned out possible since the values of γ _S, as found with those two methods, are much similar, which is confirmed by the data presented in Table 5 for same polyurethane coatings and in Table 6 for exemplary cationomer coatings. A higher number of measurements made it possible to eliminate errors which result inter alia from interactions between the drop of a polar liquid, e.g. water and a tested coat. The SFE results calculated in the Owens–Wendt method for the water–diiodomethane system turned out the least reliable. The measurement results were presented for interpretation in the form of diagrams in Figs. 4, 5, 6 and 7. The SFE of cationomers was clearly higher than the SFE of polyurethanes which had been synthesised with the use of the same diisocyanates; samples with similar chain structures were compared.¹ As was shown in our earlier studies, the cationomer chain structure, which results from the employed components—diisocyanate, tertiary amines and compounds which add counterions, strongly affects the SFE parameters [28]. That was also confirmed by the results of our present research. The highest values of SFE (γ _S ≈ 50 mJ/m²) were noted for the cationomers which had been synthesised from MDI or IPDI, N-MDA and HCOOH. Much lower values were specific for the cationomers produced with the use of TDI and HDI (γ _S ≈ 40 mJ/m²), while the use of amines N-BDA and N-PhDA, and additional incorporation of counterions from C₄H₉Br or C₁₀H₂₁Br (γ _S ≈ 40 mJ/m²), yielded further decline of SFE (γ _S ≈ 34 mJ/m²; KPU-3). When fluorine is added, the SFE value may go as far down as to 30 mJ/m² (KPU-7 and -8). The presence of even small amounts of TFBD-derived segments considerably lowers all long-range interactions \( \gamma_{\text{S}}^{\text{LW}} \) which is apparent in the case of cationomers KPU-4–7. When analysing the values of components γ _S, one may arrive at the conclusion that the highest contribution to SFE comes from \( \gamma_{\text{S}}^{\text{LW}} \), where dispersion interactions \( \gamma_{\text{S}}^{\text{d}} \) are prevailing both in polyurethanes and in cationomers. The share of polar interactions, which by their nature are more dependent on the structures of functional groups and which control inter alia the presence of hydrogen bonds, is clearly inferior both in polyurethanes and in cationomers. In the case of cationomers, on the other hand, because of the presence of ionic groups, acid–base interactions \( \gamma_{\text{S}}^{\text{AB}} \) become apparent which can hardly be observed in nonionic polyurethanes. The effects as presented above, resulting from the nature of bonds and types of functional groups, must also be decisive for the phase structures of both types of polyurethane coatings. The presence of the crystalline phase in polyurethanes may thus be assumed to result from the improved structural arrangement and reduced extent of polar and dispersion interactions, which finally also gives lower SFE in polyurethane coats.