Analysis of polymer-based films is extremely central in the field of materials science since polymers are currently used for a multiplicity of manufactured goods and technological applications. Characterization of paints and pigment films, study of coating performance, development of polymer-based electronic devices such as poly-LED displays or organic photovoltaic systems, detailed analysis of food packaging, and investigation of ion-exchange membranes, are some important fields of interest that require detailed depth-resolved investigation of polymer-based materials. Such investigations often require not only knowledge of the elemental depth profile, but also detailed molecular information along the depth, i.e. ‘molecular depth profiling’. In general, several techniques can be used to obtain in depth information – mostly elemental – from polymer systems,1–5 but no existing technique meets the technological requirements for high-resolution molecular depth profiling in the sense defined above. Among mass spectrometry techniques, in the last few years the development of polyatomic primary ion sources and, more recently, the use of low-energy monoatomic beams have opened new perspectives in the secondary ion mass spectrometry (SIMS) of organic materials and polymers. In general, SIMS depth profiling is not universally applicable for molecular depth profiling of organic compounds because of the ion beam induced degradation phenomena active under high energetic (keV) bombardment intrinsic to the technique. The effect of such high energetic bombardment, in the so-called dynamic-SIMS mode, is a severe damage of the molecular structure of the organic compounds and the resulting detection of elemental information only. In the case of polyatomic ion beams, such as C60 beams, the capability to perform molecular depth profiling could be connected with the high sputter yield and the very low penetration depth of the polyatomic ions that ‘sputter away’ the damaged region leaving a more or less undamaged surface behind.6 Recently, some of us highlighted the fact that the feasibility of molecular depth profiling by means of polyatomic beams is strongly dependent on the radiation chemistry of the particular polymer under analysis,6 probably because fast erosion results from ion beam induced reactions occurring in the polymer, which of course depend on its chemical structure. Recently, it turned out that low-energy (∼0.1 keV) monoatomic beams (Cs+ in particular) can be used for depth profiling of selected polymeric systems, with retention of some molecular information along the profile.7 Although the underlying mechanisms are still unclear, it is certainly interesting to note that, as in the case of cluster beams, the penetration depth of very low-energy monoatomic primary ions is reduced with respect to usual keV monoatomic ion bombardment.
In this direction it could be interesting to make use of plasma erosion-based techniques, such as glow discharge mass spectrometry, since they could provide high currents of low-energy species hitting the surface for fast erosion and low penetration depth. Plasma-based techniques, such as glow discharge coupled with optical emission spectroscopy (GD-OES), are well established for the depth profiling analysis of very thin layers of inorganic materials. Although the application of GD-OES to the analysis of organic and polymeric materials is limited due to the almost purely elemental information provided, Shimizu et al. have reported sub-monolayer depth resolved profiles of self-assembled monolayers of thiourea adsorbed on copper using GD-OES.8 Radiofrequency glow discharge mass spectrometry (rfGD-MS) with a quadrupole mass analyzer has been used already for depth profiling a few perfluorinated synthetic polymers.9–11 However, the use of a quadrupole mass analyzer restricts the potential of the technique. A time-of-flight (TOF) spectrometer could greatly improve the performance, in particular in the analysis of polymer thin interfaces. Modern TOF analyzers are known for their high performance in terms of simultaneous detection of all masses of interest, good transmission, and high mass range.
In this paper, the feasibility of obtaining molecular depth profiles of polymer-based films by means of an innovative pulsed radiofrequency glow discharge time-of-flight mass spectrometry (pulsed-rfGD-TOFMS) technique is demonstrated. The pulsed-rf mode is used for thermally sensitive and non-conductive materials in GD-OES depth profiling as it permits lower average power deposition. This mode is therefore well adapted to polymer thin films. The orthogonal extraction time-of-flight mass spectrometer operating at repetition frequencies of 30 kHz is well suited for monitoring the transient signals generated by the pulsed-rfGD source. The orthogonal extraction geometry allows for a combination of good duty cycle and high mass resolving power.
The polymers selected for our experiments are common polymer materials used in a wide range of applications. They have been chosen on the basis of their known thermal and radiation degradation behavior. In particular, we obtained molecular depth profiles from films of poly(methylmethacrylate) (PMMA), poly(styrene) (PS), poly(alpha-methylstyrene) (PAMS) and polyethylenephthalates (PET, polyethyleneterephthalate and PETi, polyethyleneterephthalate-co-isophthalate). PMMA, a very common polymer material often used as an alternative to glass, is known to undergo main- and side-chain scission reactions caused by increasing temperature and by photon, electron, or ion-beam irradiation. PET, one the most important raw materials used in man-made fibers and bottle production, gives rise to extensive thermally induced depolymerization, making it easily recyclable. By contrast, the radiation behavior of PS is dominated by cross-linking reactions. PAMS has an intermediate behavior under keV cluster beam irradiation.6
It will be shown that a proper choice of the discharge conditions allows us to obtain molecular depth profiles of all polymers under investigation, including those (such as PS) that cannot be satisfactorily profiled by cluster SIMS. The feasibility of molecular depth profiling of polymer multilayers will also be demonstrated.
Poly(methylmethacrylate) (PMMA), poly(styrene) (PS) and poly(alpha-methylstyrene) (PAMS) were purchased from Scientific Polymer Products Inc. (Ontario, NY, USA). Poly-4-bromo(styrene) (PBrS), Cu(II) acetate, chloroform, acetone, ethanol and 1-chloropentane were purchased from Sigma-Aldrich Srl (Milano, Italy) and poly(ethyleneterephthalate-co-isophthalate) (PETi) was obtained from ICI (Diegem, Belgium). All the above chemicals were used without further purification.
Thin films of PMMA, PS, PAMS and PETi (approximate thickness 300–400 nm) were spin-cast at 2000 rpm from a 3% (w/w) solution in chloroform onto cleaned silicon wafers (Siegel Consulting, Aachen, Germany). Before polymer casting, the silicon substrates were first rinsed in CHCl3, then treated for 20 s with ‘piranha’ solution (a mixture of sulfuric acid and hydrogen peroxide), rinsed in water, etched in HF (6%) for 10 min, rinsed again in water, dried under nitrogen flux and finally rinsed in CHCl3. Multilayered films were prepared by first spinning a 5% (w/w) PMMA solution in acetone onto a PET substrate (GDC, Milano, Italy), and then spinning (onto the PMMA layer) a 3% (w/w) solution of PS in 1-chloropentane. In some instances, the three-layered structure was marked with inorganic tracers. For this purpose PBrS was used in place of PS and the PMMA solution was doped with Cu(II) acetate.
The rf-GD-TOFMS instrument (Fig. 1) has been developed at EMPA – TOFWERK (Thun, Switzerland) and consists of an adapted glow discharge (GD) source (Horiba Jobin Yvon, Longjumeau, France) and a fast orthogonal time-of-flight mass spectrometer (TOFWERK, Thun, Switzerland) with a micro-channel plate (MCP) detector (Burle Industries Inc., Lancaster, PA, USA).
Radiofrequency (rf) power is applied to the back side of a flat sample that seals the 4 mm diameter anode chamber. A flow of Ar of 0.5 L/min is directed towards the sample surface. Pure Ar (99.9995%, Air Liquide, Madrid, Spain) is used as plasma gas at pressures ranging from 150–300 Pa. During analysis, the sample temperature is kept at 4°C in order to minimize thermal damage.
The ions originating from the source are extracted through a sampler of 500 µm diameter and accelerated through the skimmer potential of about 600 to 900 V. The interface is designed to extract and focus the ions as well as to reduce the pressure between the GD source and the mass analyzer. A detailed description of the TOFMS instrument is given elsewhere.12, 13 The discharge was operated in pulsed mode, with a pulse width of 500 µs. The duty cycle, as defined by the ratio of rf pulse width over the GD period, was 12%.
The spectrometer was operated with an extraction frequency of 30 kHz to give a range of several hundred. The mass range is calibrated using the known times of flight of C and Ar2 peaks that are always present in the time-of-flight (TOF) mass spectra. One GD pulse period is monitored by 30 TOF spectra (each being the sum of three successive TOF extractions). A 30-point GD pulse profile is generated for any ion of interest identified in the mass spectra from the area under its TOF peak after 400 averaged GD pulse periods. Then integration in pulse profile produces a point in the depth profile. A schematic showing the timing scheme and illustrating the TOF extraction, pulse width, and GD pulse profile is shown in Fig. 2.
An important feature of this reflector-based spectrometer is the high mass resolving power (m/δm, δm being the half width at half maximum of the Gaussian-fitted peak). In the experimental conditions used in this work, the actual resolving power is ∼4000 at mass 208, although values as high as 6000 can be obtained.
TOF-SIMS measurements were performed with TOFSIMS IV (ION-TOF, Muenster, Germany) instrumentation. Depth profiles were obtained in dual beam mode, using 3 keV Ar+ ions for sputtering and 25 keV Bi+ ions for analysis. The sputter beam (30 nA) was rastered over a 100 µm × 100 µm area, and the analysis beam (1 pA) over a concentric area of 10 µm × 10 µm.
All mass spectrometric measurements reported in this paper were performed in positive ion mode.
RESULTS AND DISCUSSION
Operating the plasma in pulsed mode makes it possible to follow the temporal evolution of each ion present in the discharge during the GD pulse. One can select temporal intervals within the power pulse in which analyte ions are most intense and/or have highest signal-to-noise ratio for depth profile generation. Figure 3 shows the typical pulse profile of some ions collected during rf-GD-TOFMS analysis of PMMA. A different behavior is clearly observed for different peaks. In particular the Ar ion peak shows a profile characterized by a wide plateau, while the intensity of peaks for other ions, such as C+, C6H and C4H3O+, exhibit a considerable intensity enhancement when the rf pulse is turned off (i.e. in the so-called afterglow region). In the specific case of polymer-based films we found that the afterglow region is important as it mostly contains fragment ions that could be related to the polymer. In other words, pulsed-rfGD-TOFMS spectra of polymer-based films acquired in the afterglow region show the presence of ions characteristic of the chemical composition of the polymer. Therefore, the afterglow region seems to provide relatively soft ionization of the polymer-related species sputtered from the surface and that survived in the plasma environment.
Figure 4 shows the typical mass spectrum of PMMA, in the range up to m/z 85, taken from the afterglow region. Many fragment ions related to the polymer are observed in the spectrum, such as the series CxH, C2HxO+ and C4HxO+. It is difficult to know whether such ions originate directly from the sample or whether they result from chemical rearrangements of the sputtered species in the plasma environment. A detailed understanding of these phenomena is not a simple task and therefore further investigation is needed in this matter. Nevertheless, the spectra collected in the afterglow region are unquestionably related to the chemical structures of the polymers under investigation and the characteristic fragment ions allow for their identification. This can be seen, for example, by comparing the spectra of PMMA and PET, obtained in similar experimental conditions. Their spectra are distinct although the two polymers have similar elemental compositions (Fig. 5).
The presence of fragment ions related to the polymer structure can be exploited to obtain molecular depth profiles, provided that the discharge conditions are optimized to preserve molecular information as much as possible. Therefore, as a first step, we optimized the plasma conditions for profiling PMMA thin films deposited on silicon. In particular the effect of Ar pressure and rf power was investigated. The effect of the applied power (at a fixed pressure of 200 Pa) on the depth profile of the C4H3O+ fragment ion, acquired in the afterglow region, is shown in Fig. 6(a). The best profile, in terms of signal intensity and depth resolution, is obtained at 15 W. At a lower applied power the sputter rate is lower and the depth resolution is degraded, probably from a non-flat crater bottom. At higher applied power we observe a deterioration of depth resolution and a reduction of the intensity of polymer-related ions in the flat region of the PMMA depth profile. Both effects can be caused either by local thermal heating or by increased damage due to active species in the plasma (ions, fast neutrals, electrons, photons). Figure 6(b) shows that, at fixed applied power, there is an optimum pressure that gives the best profile. This is not surprising as it is well known from GD-OES profiling that the pressure strongly influences the quality of the sputter crater. Plasma condition optimization has been carried out for all the investigated polymer samples, by using characteristic ions such as C2O, C6H, C2H4O for PETi; C2H, C4H, C6H for PS; and C4H, C6H for PAMS.
Figure 7 shows the depth profiles of PMMA, PS, PETi and PAMS thin films on Si substrates. In all cases a plateau in the intensity of the structure-related ions is observed, confirming the feasibility of molecular in-depth analysis by rfGD-TOFMS. In some cases, as in Fig. 7(c), a transient intensity ion is observed at the interface with silicon, probably due to a change in the sputtering rate between polymer and substrate or to the presence of an interfacial silicon oxide layer. It is worth noting that in the case of PS, a polymer that is not successfully profiled by cluster SIMS, a relatively flat region of the polymer-related signals is observed, although after a severe reduction of their initial intensity. This is probably an indication of some damage accumulation, in accordance with the explanation given for the analogous behavior of other organic systems in cluster SIMS depth profiling.14 The presence of the plateau, however, indicates a depth profiling capability even in the case of PS, at least when dealing with thin films.
In view of the encouraging results reported above, the possibility of discriminating among polymer layers with similar elemental composition but different structure has been considered. For this purpose an additional experiment was devised, involving a three-layered structure composed of different polymers, two of them (namely PMMA and PET) having similar elemental compositions. The structure consists of a PET substrate on top of which we deposited sequentially a PMMA film and a poly(4-bromo-styrene) (PBrS) layer. The latter polymer was used instead of PS in order to provide a marker for elemental depth profiling. For the same reason the PMMA layer was doped with a copper salt.
Figure 8(a) shows the typical TOF-SIMS depth profile of the above described structure obtained by using a monoatomic primary sputter beam. As expected, detailed molecular characterization is impossible because of beam-induced damage, and the three layers can only be identified on the basis of the elemental tracers. No discrimination between the PMMA and PET layers is possible on the basis of polymer-related fragment ions. The elemental C+ intensity is constant in all layers and oxygen-containing ions (mostly CO+) – of course, present only in the oxygen-containing layers (PMMA and PET) – do not show any intensity variation along the two layers.
Figure 8(b) shows the typical rfGD-TOFMS depth profile obtained from the same multilayer sample. In this case, although the depth resolution is not optimized, one can easily recognize the presence of the three layers on the basis of polymer-related fragment ions. In particular, the intensity of the characteristic PBrS ions, such as C6H, is higher in the first layer, decreases at the first interface, reaches a lower steady state level because of the lower intensity of this ion in the spectrum of PMMA, and finally drops again at the PMMA/PET interface. The C2H4O fragment ion (as well as several additional ions not shown in Fig. 8(b)) assigned to the acrylate polymer is mostly present in PMMA and defines the second layer. Similarly, the second interface (PMMA/PET) is evidenced by a rise of the C2O signal identifying the PET slide used as substrate.
Pulsed radiofrequency glow discharge time-of-flight mass spectrometry (pulsed-rfGD-TOFMS) has been developed and employed for the depth profiling analysis of polymer-based materials. Organic fragment ions detected in the afterglow region of the discharge allow qualitative identification of the polymers as these ions are related to the polymer structure. Plasma conditions have been optimized in order to obtain molecular depth profiles of polymers with sharp interface. Molecular depth profiles of single- and multilayer polymer film structures (based on PMMA, PS, PET, and PAMS) have been successfully obtained by means of this innovative technique.
Financial support from the European Union (FP6 Contract STREP-NMP, N° 032202) is gratefully acknowledged.