3.2. Chemical Structure Characterization
shows the full range and magnification of the 1400–1000 cm−1
range FTIR spectrum obtained for the PTFE target material which was used for deposition and examples of typical spectra of the films obtained by the PED and PLD techniques. The PTFE target spectrum typically has two characteristic peaks at 1201 cm−1
and 1150 cm−1
that are attributed to asymmetrical and symmetrical CF2
stretching. A third weaker peak corresponding to the CF2
wagging is observed at 642 cm−1
, which is consistent with literature data [2
The FTIR spectra of both types of deposited films indicate no major chemical differences in comparison to the target material, as all main bands are identified close to the maxima of the target. Nevertheless, the magnification of the 1300–1000 cm−1
wavenumber range (insert in Figure 5
) shows that a more accurate analysis exposes alteration of the chemical structure of the deposited macromolecules.
Considering the main absorption bands of PTFE (i.e. –CF2
–), it is observed that the ratio of the asymmetric and symmetric stretching bands intensity has been reversed and the absorption maxima have been shifted to higher wavenumbers (1230 and 1155 cm−1
, respectively). These two changes denote that the primary helix structure of the PTFE macromolecules undergo some transformation as a consequence of the UV and electron radiation [23
]. Additionally, the observed broadening of the whole absorption area may be attributed to the appearance of some new bands, for example from defluorinated group, that are difficult to recognize without additional data. Two regions with bands of very low intensity—marked by letter a and a′ in Figure 5
—remain unidentified. It is difficult to state unambiguously whether they are noise resulting from e.g., a small coating thickness, or whether they indicate the occurrence of absorption bands derived from oxygen groups (COOH and COF groups) or unsaturated groups, formed as a result of free radical reactions or just the moisture. To answer these questions, further structural analysis by the XPS method was carried out.
The elemental composition of the coating surfaces calculated basing on the results of X-ray photoelectron spectroscopy analysis are presented in Table 1
The surface of the samples obtained by the PED method consisted only of fluorine and carbon. The surface of the samples obtained by the PLD method also consisted predominantly of fluorine and carbon, however, some contamination with oxygen and nitrogen atoms was detected for these samples. In two out of the three examined samples prepared by the PLD method an XPS signal of silicon atoms was observed (insert in Figure 6
The presence of the silicon atoms on the surface observed by XPS may be explained by the low thickness of the PTFE layer deposited on the Si substrate as mentioned in the former chapter. It is possible that during the formation of a very rough PTFE surface some pits in the layer are also formed. As a result, the silicon substrate surface is exposed to the vacuum and detected by XPS analysis.
The fluorine to carbon ratio calculated as the quotient of the respective surface compositions (in atomic percent) for the target material is about 2.4. This value is higher than the theoretical F/C ratio for PTFE material, considered as a –(C
– polymer. However, a similar fluorine enrichment of the PTFE surface has been reported previously [25
]. The F/C ratio observed for the PED coating is very similar to the one observed for the target material. However, the F/C ratio measured for the PLD coating is less than 2. It is possible that a partial defluorination takes place on the surface of films prepared by PLD. A similar decrease of the F/C ratio has been previously observed for PTFE coatings formed by RF sputtering [31
The chemical composition of the thin films obtained by PED and PLD was analyzed based on high-resolution XPS spectra. The XPS F 1s spectra obtained for the target and the samples prepared by the PED method are presented in Figure 7
The position of the maximum of the XPS F 1s transition is identical for all samples and is located at a binding energy of 689.4 eV. This binding energy is characteristic for fluorine atoms present in the functional group with covalent C-F bonds such as: –CFR– or –CF2
]. The XPS F 1s spectra acquired for the sample produced by the PLD method are virtually identical to the ones shown in Figure 7
The XPS C 1s spectrum originating from the target is shown in Figure 8
. The maximum of the prominent peak observed for that sample is located at a binding energy of 292.3 eV. This position is characteristic for the –(C
– bonds which constitute the PTFE structure. A minor peak is observed at a binding energy of 285.0 eV, which is ascribed to adventitious carbon—a contamination of the target surface. These results indicate that the target surface consists of pure PTFE with negligible environmental contamination. Similar XPS results, both for the main peak and a contamination peak, are reported for other analyses of PTFE surfaces [25
In comparison to the target material the envelopes of the acquired XPS C 1s lines vary slightly for the samples prepared by PED and significantly for the samples prepared by PLD. To recognize the chemical transformations of the target material after exposure to the electron beam or laser pulses and its further deposition on the Si substrate, a detailed analysis of the high-resolution XPS spectra was carried out. The envelope of the XPS C 1s spectrum was deconvoluted into several components corresponding to the different chemical environments of carbon atoms. Since the majority of the XPS signal observed during the experiments originated from carbon and fluorine atoms only the presence of different C–F(H) functional groups was considered in the simulation model. Six basic C 1s components was assumed to be sufficient to properly deconvolute the envelopes of all XPS C 1s spectra recorded for the samples formed in both PED and PLD processes. The positions of the C 1s components were based on the binding energy shifts reported in [35
]. The model components are characterized in Table 2
and briefly described below.
Component C1 located at about 285 eV is ascribed to the non-functionalized sp3-hybridized carbon atoms observed for hydrocarbons. The component C2 located at about 287 eV represents carbon atoms that are not directly bound with fluorine atoms but which are in the vicinity of other fluorinated carbon atoms in the polymer chain. The component C3, with the maximum of its binding energy at about 288 eV, corresponds to carbon atoms bound to only one fluorine atom or the carbon atoms having no direct bonds with fluorine but located close to the chemical environment with several fluorine atoms. The component C4 is ascribed to carbon atoms located in a polymer chain with several –CFR– groups in a row. The most intense peak observed in all the recorded spectra is located at a binding energy of about 292 eV and is marked as C5. This corresponds to carbon atoms in –(CF2–CF2)n– functional groups in neat PTFE. At the high-energy side of the main maximum a peak at about 294 eV can be discerned. This environment is ascribed to –CF3 terminal groups of fluorinated polymer chains and denoted as C6.
The deconvolution procedure based on the above basic C 1s components was carried out with the application of several additional constraints. The position of the peak maximum was varied by no more than ±0.2 eV. The positions of all C1–C6 peaks given in Table 2
were used for all the considered samples: the target as well as the PED and PLD formed deposits. The full width in half maximum (FWHM) of the model lines was kept between 1.8 and 2.0 eV.
The positions of the C1–C6 peaks are indicated by arrows in Figure 8
. The results of the deconvolution procedure, indicating the fractions of the whole XPS C 1s signal corresponding to the considered chemical environments, are given in Table 3
and are shown as component peaks in Figure 8
The deconvolution of the target spectrum indicates that, apart from minor environmental contamination reflected as a small peak at 285 eV, the target material consisted only of –(C
– functional groups. On the surface of the samples prepared by the PED method (Table 3
) the component C1 was not observed at all. This indicates that the adventitious carbon contamination present on the surface of the target is not transferred onto the substrate during the deposition process. The surfaces prepared by the PED process are similar to the pure PTFE. The concentration of the C5 component exceeds 80% of the total XPS C 1s signal.
For PED, only a small number of CF3 groups along with a small amount of defluorinated carbon are recognized, the presence of which can explain the previously described absorption band changes in the 1300–1000 cm−1 region of the IR spectra. Nonetheless, the XPS data together with the FTIR results and micrographs of the samples obtained by PED show that the surface chemical composition is virtually identical to the target composition and remains almost pure PTFE.
The chemical composition of the surface of the coatings produced by the PLD method is very different from that observed for the target and the samples formed by the PED process (compare Figure 8
a,b). The envelope of the XPS C 1s transition is complex and extends over a wide range of binding energy from 285 to 294 eV with several prominent local maxima. The deconvolution of these spectra was carried out using an identical model as that described above. The results of this procedure are shown in Table 3
The presence of many chemical environments for the carbon atoms indicates that the PLD process results in a substantial defluorination and recombination of C–F functional groups of the PTFE target. Even at the lowest process pressure of 0.13 Pa the signal coming from –(C
– groups, considered as neat PTFE, is below 30% of the total XPS C 1s signal. The defluorination of the chemical environment of the carbon atoms may be partial as indicated by the presence of C3 and C4 components (–CFR–C
– groups). However, the abundance of C1 and C2 components indicates that in a substantial part of the material deposited during the PLD process, some carbon atoms are totally defluorinated and consequently crosslinked, since the aliphatic hydrocarbon groups are prone to such free radical reactions initiated by physical factors such as UV radiation [36
]. These defluorinated atoms constitute approximately 40% of all carbon atoms present in the final product of PLD deposition. Recombination of –(C
– groups into terminal –C
functional groups is also observed. The internal structure of the XPS spectrum observed for these samples is similar to the spectra observed for sputter-deposited fluorocarbon films [31
The XPS analysis of the surface of the samples obtained by the PLD process shows that a very prominent degradation of the initial PTFE material occurs. The final composition of these coatings is far from that of expected for a pure PTFE layer.
The XPS results clearly demonstrate a significant difference in chemical structure between the films obtained by the PED and PLD methods. Moreover, the findings regarding the films obtained by the PLD technique contrast with the results of the FTIR measurements, which did not show very significant changes in the chemical structure for these films. The chemical degradation of the samples formed by PED observed by XPS is negligible and in agreement with the results obtained from FTIR analysis.
XPS is a more surface-sensitive method, therefore some degree of the degradation observed by XPS is presumably located within the upper atomic layers of the coatings. Considering the results obtained for PLD coatings, the difference between the results obtained by XPS and FTIR analysis may be caused by the inhomogeneous surface topography of the PLD coatings, which affects each analysis method in a different way. There was large amount of particulates observed on the surface (Figure 3
). During the FTIR measurements the ATR crystal was in close contact with the sample surface, which results in flattening of the protruding particulates. The penetration depth of the IR radiation is of the order of a few micrometers which means that both the thin continuous film and the particulates are within the analyzed volume. Taking into account the size and number of the particulates their volume exceeded the volume of the continuous thin film by at least an order of magnitude. In this way the FTIR spectra are dominated by information originating from the structure of the particulates. In contrast, XPS is contactless and gives information about a very small volume of the material close to the surface, with a thickness of a few nanometers. In the case of morphologically complex surfaces such as those on the PLD coatings, the XPS spectrum contains information originating from the thin continuous film and a thin surface layer on the particulates.
In the light of the results obtained, it can be hypothesized that in the case of laser ablation, large non-degraded particles of material are extracted from the target. However, their transport to the substrate is probably possible due to an expansion of gaseous products of localized degradation, resulting e.g., from the heterogeneity of the PTFE structure [37
]. Such a mechanism would be in agreement with the theories regarding the mechanism of polymer transport during laser ablation presented in [38
]. Evaporated degraded polymer deposits from the vapor phase in the form of a thin film. This differs from the composition of the starting material and is characterized by a significant loss of fluorine and the presence of oxygen containing compounds, as confirmed by XPS analysis.
In the case of electron beam deposition, the results obtained would indicate a different transport mechanism. According to the model proposed in [39
], ablation can be the result of the evaporation of small pieces of the chain, formed as a result of macromolecule cutting, which then repolymerize on the substrate. While in the case of polymers the term “evaporation” may be controversial, for PTFE this mechanism is particularly plausible due to the presence of two fluorine atoms, thanks to which the formation of linear chains is thermodynamically favored [40
]. This would explain the very good preservation of the PTFE chemical structure in PED coatings.
A detailed analysis of the ablation mechanisms clearly requires further research, however, it should be emphasized that the proper characteristics of the coatings obtained are crucial for their understanding.
The issue with different results for different analysis methods does not arise in the subject-related literature. The majority of publications concerning PLD PTFE coatings present FTIR measurements only [12
]. In the work of Smausz at al. [15
], the spectra are not included in the XPS results presented and moreover, these results concerned coatings obtained at higher temperature than in the present work, or annealed, which could significantly change the coverage of the surfaces as a result of melting of the particulates. In our experiments owing to the combined use of both ATR-FTIR and XPS it was possible to thoroughly characterize the chemical structure of the PTFE films and the inhomogeneities in the coating morphology.