Printer Friendly

Multilayer Method as a Tool for Depth Dependent Polymer Film Photodegradation Studies.

A new polymer film destructive depth profiling protocol is presented for the analysis of photo- and thermally degraded thin films on the depth scale of less than 100 microns. The method, demonstrated here on thin films of poly(vinyl chloride) (PVC), provides a means of preparation of thin laminates of high optical quality comprised of many ([greater than]20) thin layers of individual thickness less than 15 microns. The constituent layers are fused together under appropriate pressure, temperature and time treatment to yield a film assembly of high optical quality that behaves like a uniform single layer during photodegradation exposure, but which may still be separated after treatment. Compared to previous techniques, this new method is relatively simple and non-labor intensive. Film adhesive properties are controlled to within [+ or -] 5% Concentration depth profiles of polymer photolysis products were reconstructed by analyzing each of the separated layers using UV-visible spectrophotometty. The continuity of these film assemblies with respect to mechanical properties, adhesive properties and the depth distribution of key photolysis reagents and products was confirmed using photothermal and reference microscopy techniques. Optical absorption depth profiles examined in UV- photodegraded poly(vinyl chloride) (PVC) films exhibited the classic dependencies expected in the presence of nitrogen and oxygen atmospheres.


In studies of the degradation and stabilization mechanisms in thin polymer films and coatings, the characterization of depth dependent changes in composition on the micron length scale is of essential importance [1-10]. Depth profile analyses furthermore may give insight into the influence of many external variables on the degradation process. These variables include the intensity and spectral distribution of visible and ultraviolet radiation, temperature, humidity, oxygen and ozone concentrations and other factors influencing material reactivity. A classical example of photo-induced depth non-uniformity, is the formation of chromophore centers in many materials, which progressively limit the propagation of photolysis radiation deeper into a sample, as irradiation proceeds. Material disintegration may then preferentially occur in shallow subsurface regions.

A number of previous works have been devoted to investigations of depth variations of composition in polymer films as the result of photodegradation [2-8]. A marked obstacle to these investigations is the absence of a convenient and reliable method for depth sampling of material composition on length scales in the range 1-100 [micro]m. Previous depth sampling methods typically involve cutting or scraping thin layers from the surface of a sample, with subsequent analysis of the removed (or sometimes the remaining) material. Analysis methods have included infrared spectroscopy, in which the removed material is pressed into pellets with potassium bromide [5], and microtitration performed on the sampled dust [8]. Problems with labor intensiveness, potential low reproducibility and complexities in calibration are serious limitations of such destructive methods. The need for a convenient, reliable and reproducible method of depth profile analysis, therefore persists.

In the present work, we introduce a new strategy for the depth dependent investigation of degradation mechanisms in thin polymer films. This method consists in principle of preparing the material under test as an assembly of many very thin layers, which are fused together under known treatment conditions. The adhesion between layers is maintained under a compromise condition: the adhesive force must be maintained high enough to ensure that the properties of the assembly remain as close as possible to those of a continuous single layer medium, but sufficiently low to permit separation, by mechanical peeling after degradation.

This work will present a new method for preparing multilayer film assemblies of high optical quality and of repeatable composition under known conditions of adhesion. This method is convenient, reliable and much less labor intensive than those reported previously. Film assemblies prepared by this method are subjected to examination by a number of optical and thermal techniques to confirm the continuity of the material with respect to its transport properties. Our study will be implemented using unstabilized poly (vinyichioride) (PVC) as a well characterized test material. Its' photodegradation mechanism has accompanying depth patterns of photo-oxidation and photodehydrochiorination which have been extensively studied [2, 5-8, 10].


I. Materials

Unstabilized poly(vinyl chloride) (PVC) was received from Aldrich Chem. Co. (No 9002-86-2, Relative viscosity 2.23, average molecular weight [M.sub.n] = 55,000; average molecular weight [M.sub.w] = 97,000). Traces of catalyst and other impurities were removed by a threefold precipitation with methanol (HPLC grade) from a solution of tetrahydrofurane (THF, HPLC grade), with subsequent vacuum desiccation.

II. Film Preparation Methods

Thin PVC films of 5-15 [micro]m thickness were used as the individual lamina in the multilayer assemblies. These films were prepared by spray casting onto glass surfaces, by the technique described below.

Spray casting was performed by feeding a 4% w/w solution of PVC in THF through an airbrush hooked up to a supply of compressed nitrogen. The spraying operation was carried out in a tent enclosure, constructed in house, under an atmosphere of nitrogen saturated with THF. This procedure was used to prevent bubble formation in the films by rapid outgassing of solvent vapors. Drying of the films was conducted in the enclosure under conditions of a slow decrease of the THF vapor concentration in the nitrogen carrier, starting from THF saturation, with reduction eventually to zero concentration. The precise time period required for this operation depended on the film thickness, but was in the range of 3-5 hours.

Spray deposition of the films was made onto glass with the airbrush maintained at a distance of 500 mm from the glass surface. Films were prepared by deposition onto microscope slides, which had been precleaned by mixture of chromic and sulfuric acids. This was followed by treatment in an ultrasonic bath using deionized water, and finally, HPLC grade methanol, as the cleaning medium. Film drying was conducted in the tent enclosure as described above, by progressive removal of solvent from the gas flow. The final drying operation was carried out in an atmosphere of dry nitrogen at 70[degrees]C for 30 minutes and after that in vacuum (0.001 mm Hg) at 100[degrees]C, 2 hr. After drying, the prepared PVC films could be easily released from the glass surfaces.

III. Multilayer Assemblies

The procedure outlined below was used to prepare multilayer film assemblies for photodegradation tests. Multilayer assemblies used for these purposes must be highly free of optical defects, bubbles, and folds. It is normally very difficult to avoid the formation of these defects in preparing multilayer films by conventional methods, which use a combination of overlay and pressure. However, the method described below introduces new procedural elements which enable the fabrication of multilayer assemblies composed of many thin layers (15-20 layers and +), with maintenance of high overall optical quality.

The individual films in a multilayer assembly must be adhered together in fabrication and yet must be separable after processing. To ensure eventual separability of the layers, circular cardboard spacer rings were prepared (70-100 [micro]m thickness and 30 mm diameter) and attached to the individual sheets of (5-15 [micro]m thickness) film before the films were released from the glass backings (see preparation stage, above). The rings were attached to the film using cyanoacrylate adhesive. After curing of the adhesive, each individual specimen of the film was trimmed to a circular shape (30 mm diameter just at the outside edge of the cardboard spacer ring) and then released from the surface of the glass.

About 15-20 films are assembled in a typical multilayer. The number of films as well as their thickness are not limited and depend on desired depth resolution. The lower limit of thickness of individual films has not yet been determined but a thickness of 1 [micro]m should be feasible. We have used 5-15 [micro]m films for convenience of handling, since they meet acceptable depth resolution criteria for the present application. The selected quantity of films backed with cardboard spacer rings was overlaid in a vacuum die assembly of a proprietary design (O. Nepotchatykh, US Patent Pending). The die chamber was immersed in a thermostatic oil bath and then evacuated to a pressure of [10.sup.-4] mm Hg over a period of 20 minutes, to release interlayer gases. The die element was externally clamped near the top position of travel to prevent the application of mechanical pressure to the films during the initial evacuation phase. After a complete removal of interlayer gases from the chamber, the die was released and a mech anical pressure equivalent to 1 kg/[cm.sup.2] was applied to the films.

The degree of adhesion between individual layers in the assembly was characterized by means of peel force testing. A dynamometer with a full-scale resolution of 1 kg was used to make measurements of the peel strength to within a [+ or -] 5% repeatability.

During drying and fabrication of a multilayer assembly, the PVC polymer must be exposed to temperatures in the range of 100-120[degrees]C, in nitrogen atmosphere and under vacuum. The possible contribution of thermal degradation to the PVC assembly during preparation was considered. Based on past work on the low temperature thermal degradation of PVC [11], we have estimated that the degree of thermal degradation of unstabilized PVC under these conditions to be characterized as 2 X [10.sup.-6] to 5 X [10.sup.-5] mole HC1/mole PVC corresponding to a % wt. degradation of less than 0.002% of the prepared materials. This degree of degradation is known to contribute negligibly to the experimental error in photodegradation testing.

IV. Film Characterization Methods

Optical Microscopy

Analyses using light microscopy were performed on prepared cross sections of multilayer assemblies, with the objective of detecting the presence (or absence) of interfaces between the adhering layers.

Thin cross sections of multilayer films were prepared by setting the sample in an Epon 812 resin matrix (J. B. EM Montreal, Canada) at 60[degrees]C, followed by a cure at room temperature. 5 [micro]m thin cross sections were tomed using a Reichart Ultracut (Austria) microtome apparatus. Thin sections were examined using a Zeiss Ultraphot Light Microscope, under Nomarski differential interference contrast (quartz first order red retardation).

Photothermol Analysis

Another method for possible testing of material continuity is provided by photothermal characterization. Techniques of photothermal depth profilometry have been used in a number of recent works [12-17] to characterize variations of both optical and thermal properties of thin film materials with depth. The optical depth profiling methods are based on the laser induced generation of heat sources in materials [17], with the observation of temperature changes occurring in the time or frequency domain in response to the heat source generation. An analysis of the thermal transient or modulated signals permits the detection of discontinuities in the thermal properties at buried subsurface interfaces [13].

In previous works in this laboratory, photothermal depth profilometry techniques were developed for the detection of depth resolved optical absorption in thin film materials, and also for the detection of interfacial properties of layer assemblies containing a small number of discrete interfaces [12-16]. These works included a demonstration of the photothermal depth profiling of optical absorption in photo-degraded PVC films via the laser photopyroelectric effect [14].

More recently, a more quantitative method of photothermal depth profiling has been developed based on the laser mirage effect, a method of photothermal beam deflection which measures heat conduction from the surface of an optically heated sample into an adjacent fluid medium. This method, which is described in detail in references [16] and [18-19], is included in the present work as an optical and a thermal continuity confirmation method for the photodegraded films. It has the special advantage of having good immunity to background signals, when optical absorption in the sample is relatively weak, as is true of the measurements encountered in the present work.

Film preparation techniques for mirage effect analysis are simple and involved roll pressing cut sections of the photodegraded material onto plexiglas backings. using a thin (1-5 [micro]m) layer of adhesive. After the adhesive was cured, the mounted samples were assembled into a mirage effect cell for photothermal analysis.

V. Photodegradation Procedures

The photodegradation of PVC film samples was carried out in a photodegradation chamber, which was designed and constructed in house. The degradation apparatus contained a light source which may be used to simulate the solar spectrum, and which consisted of a 1000 W Xenon arc lamp (Hanovia 982C0011) installed in a housing equipped with a system of quartz lenses and spherical concave mirrors (Oriel Corp. Model 66023). A quartz water cell, acting as a broadband infrared filter, was placed between the lamp housing and the photolysis cell. An adjustable aperture was placed between the water cell and the photolysis cell, and used to select a highly spatially uniform region of the light beam for the photolysis. Under typical operating conditions, this setup delivered an irradiation beam of 45 mm diameter to the sample, with an irradiance which was spatially uniform to within [+ or -] 3.5%. The photolysis cell consisted of a PMMA hermetic cell body. The cell was equipped with quartz entrance and exit windows, fitting s for gas supplies (nitrogen or oxygen) and had an internal microfan intended for cooling the sample during photolysis.

Measurements of the total irradiance of the photolysis beam were made by means of a bolometer designed in-house on the basis of a poly(vinylidene fluoride) PVDF pyroelectric thin film detector (28 [micro]m thickness, material supplied by Atochem North America inc., part No: TO28NA). The detector film was overcoated with a blackbody absorber layer, and installed into a housing of previously reported design [20]. Because of known frequency response limitations of the detector circuit at low frequencies, the light incident on the pyroelectric was modulated using a mechanical chopper at a frequency of 15 Hz. The modulated detector signal was recorded using a lock-in amplifier (Princeton Applied Research, Model 5101), the chopper frequency was monitored using a frequency counter, and a digital voltmeter was used to read the output channel of the lock-in amplifier.

The pyroelectric bolometer was calibrated against a Newport Model 835 optical power meter in the wavelength range of 400-1000 nm. This meter gave no coverage of the ultraviolet, however, so that the pyroelectric device, which is spectrally flat over the entire UV-visible range, was used for actual calibration measurements on the lamp system. The total irradiance of the photolysis beam incident on the sample under an electrical power of 900 W supplied to the lamp, was 8400 W/[m.sup.2] (1580 W/[m.sup.2] for [lambda] [less than]350 nm).

The spectral irradiance of the photolysis beam was monitored periodically using an Oriel Model 77250 monochromator equipped with a set of quartz lenses to increase throughput of the input beam. The pyroelectric bolometer was aligned immediately past the exit slit plane of the monochromator.

The spectral irradiance distribution in the photolysis beam gave a good approximation to the extra-terrestrial solar spectrum [21], when the beam was directed through near UV absorbing filters. However, for the purposes of the present study, the intent was not to simulate solar exposure, but rather to repeatably induce depth dependent changes in the film composition by known photolysis mechanisms on an accelerated time scale. To achieve this, the lamp output was not filtered by any mechanism, beyond the use of a water cell. The spectral radiance distribution of our lamp system had a greatly enhanced contribution in the UV. Our measurements indicate that our photolysis beam had an integrated UV irradiance (in the range [lambda] = 200 nm- 350 nm) which is 27.8 times that of the estimated solar extraterrestrial value.

The UV-visible absorption spectra of all films was made on an HP Model 8452A Photodiode Array Spectrophotometer.


The successful use of a multilayer thin film assembly to replace a monolithic, single layer in photodegradation testing rests on two premises. First, the model film should be mechanically resolvable into separate layers (albeit with sufficient adhesion between layers), which is necessary for a convenient depth resolved sampling of the material. Second, in depth dependent studies of photodegradation, the replacement of a monolithic single layer test sample by a multi-layer assembly is valid only in cases where the multilayer material shows a good approximation to depth continuity in the following properties. These consist of (i) optical properties; (ii) thermal transport properties; and (iii) mass transport properties, particularly molecular diffusion coefficients for atmospheric gases and species such as HC1, which is an essential degradation product [2, 5, 9, 22].

In light of the need to determine the validity of these assumptions, a series of characterization tests was performed to establish the degree of material continuity for multilayer assemblies prepared under known conditions.

I. Film Preparation Variables

In order to avoid changes in the orientation of polymer chains and therefore changes in material microstructure with preparation conditions, adhesion between individual films was induced at relatively low mechanical pressures and temperatures. In order to ensure that film assemblies were prepared under known and repeatable conditions, a study was made of the adhesive force present between layers as a function of the preparation temperature and time duration, for application of mechanical pressure. The results of this study are summarized in Table 1. Measurements of adhesive force were repeatable to within a [+ or -]5% error over replicate preparations of the assemblies under constant conditions, for all entries in this table. The adhesive force measured between layers was furthermore independent of the number of layers making up the assembly.

We determined that for PVC, there is an accessible working range of temperatures and time duration of mechanical pressure, which may be used to produce films with useful mechanical properties. In particular, an adhesive force within the limits of 30-150 g/[cm.sup.2] between layers in a model film, is sufficient for the convenient mechanical separation into individual layers. Above 400 g/[cm.sup.2] the adhesive force becomes comparable to the tensile strength of the individual material layers and the film layers tear before separating. Below 30 g/[cm.sup.2] adhesion is too weak to withstand minimal handling without separation of the individual layers. A preparation temperature above the glass transition temperature ([T.sub.g] [sim] 85[degrees]C for the present material) promotes good adhesion, because of the enhanced mobility (therefore increased probability of entanglement) of the polymer chains. Temperatures which are too low fail to promote adhesion, while temperatures substantially above [T.sub.g] and close to the polymer melting point ([T.sub.m] [sim] 170[degrees]C) promote fusion of the two phases with no possible separation of the individual layers. As seen in Table 1, successful preparations of multi-layer assemblies of our test material with properties in the desired working range of adhesive force, are obtained at temperatures in the range of 110-120[degrees]C and time duration for application of mechanical pressure in the range of 10-30 minutes. The properties of these preparations are repeatable to within an error of [+ or -]5%.

II. Continuity of Optical Properties

As noted earlier, the optical quality of films prepared by our assembly procedure is high, and comparable to the quality of the individually cast layers. Figure I shows a photograph of a typical multilayer film assembly before photodegradation. It should be noted that adhesion (and ultimately, photodegradation) occurs efficiently in a 1 [cm.sup.2] circular center region of the assembly (corresponding to the dimensions of the die element used to press the films together). Prior to photolysis, we measured the transparency (optical transmission) of single layer films and of the central adhering region of the multi-layer assemblies in the wavelength range of 200-820 nm. These measurements have shown an absence of any significant difference in the ability of single or assembled films to transmit light.

In an attempt to observe the structure of possible interfaces in the undegraded PVC assemblies, we executed microscopy studies on multilayer films along the depth cross section of these samples, with preparations at different levels of interlayer adhesive force. The method of differential interference contrast (Nomarski-quartz first order red retardation) was chosen because of its high sensitivity to refractive index and thickness related inhomogeneities in transparent media [23]. Microscopy examinations were carried out on multi-layer cross sections having interlayer adhesive forces lying within the limits of 5-1000 g/[cm.sup.2]. Here, however, we encountered a number of difficulties with the sample preparation procedure. As a result of the process of cutting microtomes from an embedded sample matrix (see experimental section), many of the multilayer materials fell apart in sectioning. Commonly, we observed ribbon-like separations of one or more layers from these cross sectional cuts, with folding of the di splaced material over intact sections of the cut. This was observed even at relatively large values of the interlayer adhesive force ([great than] 50 g/[cm.sup.2]), and was an increasingly common problem as the number of interfaces in the sample increased. Because of this problem, our microscopic examinations were restricted to assemblies having only 1-3 constituent layers. Another feature of these preparations is the presence of a regular pattern of score marks in regions of the cross section and the sample matrix images. These artifacts are induced by the toming process and may further complicate image interpretation.

Notwithstanding, one point can be clearly made about the micrographs obtained on intact thin sections of these multiply layered samples, of which Fig. 2a shows a typical example. No detectable interfaces are observable by this method. Interlayer interfaces are only seen on corrupted thin sections in which the material is either clearly separated or was on the verge of rupture (Fig. 2b). Based on these images, there is thus no evidence of an observable departure of the multilayer assembly from the properties of an optically and structurally continuous medium.

A more recent study of these assemblies was made using the newly developed technique of light profile microscopy [24]. This is a method of crossed beam microscopy, which employs contrast based on elastic scattering (among other mechanisms). Light profile micrographs were capable of detecting interfaces in intact specimens of these materials, with very high sensitivity. However, because of the unusual contrast of this new method, which is still under investigation, the significance of this observation, in terms of the transport continuity model of the multilayered material is still unclear. Further investigations are ongoing and will be reported on in future work.

III. Continuity of Interfacial Thermophysical Properties

The absence of detectable thermal interfaces in the multilayer assemblies prepared by our experimental method was confirmed using a recently developed photothermal depth profiling technique based on laser mirage effect spectrometry [18, 19]. A bilayer assembly, prepared at a temperature of 110[degrees]C and subjected to mechanical pressure for a period of 20 minutes was used as a material to test adhesion between interfaces. A diagram of the test sample is shown in Fig. 3.

The film assembly was surface deposited on front (F) and rear (R) surfaces with a thin blackbody index marker, which was used to absorb excitation light from a modulated laser beam. The sample was adhered to a flat poly(methylmethacrylate) (PMMA) backing block and then sealed into an optical cell containing a water medium for photothermal beam deflection analysis. In this measurement, the sample was heated by light absorption from a wideband modulated laser beam, by which a thermo-optical impulse response is extracted from the measurement [15]. The measured impulse response was used to reconstruct a depth profile of light absorption induced heat flux in the material at t = 0, past excitation by an (albeit mathematical) impulse.

When a thermally continuous sample is excited at the front surface by an impulse, all light absorption (and generated heat source density) is confined to the front surface of the sample. The depth profile of heat source density generated in the sample by the pulse, and reconstructed by this method, should approximate a spatial Dirac delta function corresponding to heat source density confined to the front surface. In practice, even for a theoretically continuous material, the reconstruction algorithm contributes a well-characterized broadening [12, 13, 16, 17], which is typical of that seen in the observed profile. Similarly, when the back surface of the sample is irradiated by the laser impulse, in a thermally continuous material, the heat source reconstruction should contain a peak precisely at the depth of the rear surface (as measured from the front surface position of the sample). With rear interface irradiation (R), the reconstruction of a broadened peak results, positioned at a depth of 140 [micro]m from the front surface. Again, the broadening with depth has been found to be characteristic of the reconstruction algorithm in the presence of a thermally continuous material.

In a material with detectable thermal discontinuities at the interior interface, reconstruction of heat source density in the presence of light absorption at the front surface would clearly contain, in addition to a large surface heat source peak, significant satellite peaks at the depth positions of value [2n1.sub.1] (n = 1, 2,...) (where [l.sub.1] = 70 [micro]m). These contributions arise from the reflection of thermal energy at an interface where interfacial adhesion is poor [13]. In the case of optical excitation at the rear interface, the effect of poor interfacial adhesion would produce in a delay in the arrival of thermal energy at the front surface. The reconstructed depth of the rear surface absorber would then be larger than the physical distance of the rear surface from the front of the sample.

As both of these effects are visibly absent, the test confirms a good first order approximation to thermal homogeneity, for the bilayer material.

IV. Continuity of Pphotoproduct Distributions in Photodegraded Thin Films

The polyene sequences form in PVC from the well studied process of dehydrochlorination [2, 9, 10]. The polyene photoproducts have a significant visible wavelength absorption owing to multiply conjugated olefinic structures which form over many repeat steps of the dehydrochlorination reaction [2, 5, 9, 10]. This process occurs efficiently, and without competition from photooxidation only in the absence of oxygen. When oxygen is present, the well known bleaching effect occurs, due to olefinic bounds oxidation, which disturbs conjugation of polyenes and forms structures which are transparent in the visible wavelength range (2), and excluded from analysis using the mirage effect spectroscopy.

The results shown in Fig. 4 demonstrate the level of accuracy available for absorption coefficient depth profiling using our mirage effect methodology. The sample in this case consisted of a multilayer assembly photodegraded in an oxygen atmosphere. After photolysis the central region of the multilayer assembly (see Fig. 1) was cut into two halves: one half was analyzed by the mirage effect while the other was mechanically separated, and subjected to peel force testing followed by absorption spectrophotometry on the separated layers. The absorption coefficient depth profile, [beta](x) at 476 nm, recovered by mirage effect spectrometry as described in References [18, 19], is integrated over the average thickness of the layers used in the multilayer assembly. This block integrated profile was superimposed with that recovered from absorption spectrophotometry, as measured directly on the individual layers. The agreement between the depth profiles, is measured as a root mean square (rms) difference between the p rofiles normalized to the peak absorption coefficient value of the profile(s). This agreement is consistently better than 12% rms of full scale for the two techniques. The inherent uncertainty (random error) in the destructive multilayer depth profile is limited at present by photometric errors.

These profiles follow a well-known depth pattern of optical absorption in photodegraded PVC films in the presence of oxygen. In the photodegraded film photooxidation occurs efficiently within one oxygen diffusion depth from the surface (10-40 [micro]m), giving a distribution of photoproducts which are nearly transparent in the visible wavelength range. Deeper into the film ([great than] 30-40 [micro]m) the oxygen concentration is effectively zero, dehydrochiorination is efficient and long chain length polyene sequences build up rapidly in the course of the photolysis. As the sequence lengths increase, however, the photoproduct UV absorption also increases dramatically, attenuating photolysis light by absorption before it reaches greater depths. At depths significantly greater than [sim] 100 [micro]m, the film becomes again, optically transparent.

A test was devised to evaluate the continuity of the transport properties of these multilayer assemblies, involving a comparison of the depth profile of the optical absorption coefficient, [beta](x), obtained in a separable multilayer assembly, to that obtained in a fused (non-separable) assembly. To make this test, two identical assemblies of the material were prepared but with differing heat treatment as outlined in Table 1. The separable assembly had adhesion in the range 100 [+ or -] 20 gm/[cm.sup.2] while the fused material was heat treated at 150[degrees]C for 15 minutes, giving non-separable layers with adhesive force greater than 600 gm/[cm.sup.2]. Laser mirage effect depth profiling was used to recover [beta](x) for these assemblies, because of its special sensitivity to interfacial discontinuities in heat flow. The profiles, [beta](x), obtained for both assemblies, are compared in Fig. 5, and show no significant difference in the depth profile features, to within an experimental error better than 10 % rms of full scale. This finding is consistent with an absence of barriers to the transport of significant photolysis products in the separable assembly.

A further comparison was made between the mirage effect based absorption depth profiles obtained in a separable multilayer assembly, and in a single layer (200 [micro]m) spray cast film of PVC, photolysed under parallel conditions. Such a comparison was more difficult here because the single layer casting contained a different loading of solvent (THF) from the multilayer assembly.

Several effects of this solvent on the photolysis kinetics, and therefore the depth distribution of photo-products in PVC have been identified in past work. Stabilizers and peroxides are common impurities in reagent grade THF. Significant concentrations of peroxides, particularly the [alpha]-peroxide of THF, have a direct influence on the photooxidation rate, through their tendency to form free radicals. The effect of stabilizers, through light screening or other mechanisms is also problematic. In the absence of these impurities, THF itself has been furthermore shown to have a slightly inhibiting effect on the degree of photo-dehydrochlorination in PVC [25].

In the present work, all film preparations were conducted using HPLC grade solvent, in which such impurities were not present above the trace level. We have experimentally determined that the presence of THF in PVC film reduces the rate of polyene formation during the photodegradation of PVC in the presence of [O.sub.2] or [N.sub.2], consistent with past work [25]. This was in agreement with the very good repeatability of the depth profiles determined in the photolysis of our multilayer assemblies.

In the case of the multilayer assemblies, the individual layers ([less than] 15 [micro]m thickness) were dried thoroughly in a vacuum oven prior to assembly, whereas thick (200 [micro]m) single castings required excessively long periods of drying under vacuum to remove the solvent. Furthermore, we have found evidence of differences in particle morphologies based on the differences in spray casting procedures [24], so an additional variable arises in comparing photodegradation in very thin assembled layers ([less than] 15 [micro]m) and thick ([greater than] 200 [micro]m) layers. Notwithstanding, when corrected for peak absorption in the [beta](x) depth profiles, as recovered by the mirage effect, a close agreement was found between the polyene absorbance distribution found in the separable multilayer assembly and that of the single cast layer.

Another test carried out on photodegraded samples of this material was a measurement of the adhesive force between layers as a function of depth. Prior to photolysis, there was no depth dependence of interlayer adhesion observed in samples of the material. After photolysis, samples showed a depth pattern of adhesive force which varied generally according to the profile shown in Fig. 6. The sample examined in this case had an adhesion prior to photolysis of 30 gm/[cm.sup.2], but in regions where photodegradation was most extensive, the adhesive force increased by [sim] 100 g/[cm.sup.2]. A similar adhesion enhancement was also observed for samples photolysed under nitrogen, with a peak in adhesive force occurring near the sample surface, where the photoproduct concentration was a maximum. The most direct explanation for this effect is an increase in the sample's internal temperature, which may arise through the combination of the high photolysis source irradiance with the strongly enhanced UV and visible absor ption of the photoproducts, relative to the starting material. A relatively small total extent of chemical conversion of the starting material is required to produce very large changes in the optical absorption accounting for this effect. However, adhesion may also be enhanced by other photoinitiated processes such as polyene crosslinking [2] (which would be most efficient in the absence of oxygen), or the interaction of polar structures arising from photo-oxidation (in the near surface regions of the oxygen photolysed material). Further study would be required, however, to discriminate the relative contributions of these processes. This enhanced adhesion would appear to promote interlayer continuity, and indeed, there is no evidence, in terms of polyene absorption profiles, that the increased adhesive force disrupts the continuity of transport properties.

By way of further illustration we may also extend the results, recovered by the multilayer and nondestructive depth profiling methods, to a comparison of the photodegradation of PVC films, conducted under an atmosphere of nitrogen (Fig. 7). From these results we observe another typical pattern in the absorption coefficient depth dependence in PVC at visible wavelengths. After photodegradation under an oxygen atmosphere, the maximum concentration of polyenes occurs at depths of 30-60 [micro]m, with a monotone decrease deeper into the polymer. Under an inert atmosphere such as nitrogen, we observe the maximal concentration of polyenes in a 5-10 [micro]m superficial layer of the sample, with an enhanced degree of peak optical absorption. This pattern is known in the literature [26, 27].

It should be noted that an important advantage of the multilayer method is that it is not restricted in spectral range, or for that matter, even to optical analysis of the separated layers. While the mirage effect depth profile method has many attractive characteristics, especially its non-destructive character, it can operate only at available laser wavelengths. The multilayer method in principle may be extended to the analysis of many more chemical species than are accessible by optical methods.

Finally, while the results of these studies show great promise for applications of the multilayer method in film testing, we must also keep in mind the limitations it poses. Under extreme conditions, for example at high temperature, in the presence of solvents or under extreme degrees of degradation, the layered model can fail in several directions. The assembly may be predisposed for example, to delamination, in which case the boundary effects become enhanced. Alternatively, at high temperature, the individual layers may become fused together and therefore inseparable after photolysis.

Nevertheless under controlled experimental conditions, and conditions of weak to moderate photodegradation, we have established that this method has a number of essential advantages in comparison to previous techniques for obtaining depth dependent information in thin film PVC. The extension of these results to other polymeric materials will require obvious adaptations of our experimental procedures. However, the present work lays a groundwork of methodological principles and procedures to accomplish that extension.


Mechanically separable PVC film assemblies can be prepared with high optical quality, and with predictable adhesive properties under specified conditions. Our preparation method is reliable, unambiguous and non-labor intensive. The results of thermal and conventional optical microscopic testing failed to detect any evidence of significant interfacial discontinuity between the assembled layers. Agreement between depth profiles of dehydrochlorination in photodegraded films, obtained between the multilayer destructive analysis and non-destructive (mirage effect) optical absorption depth profiling is observed to a nearly quantitative extent. Optical absorption depth profile test results give a picture consistent with a continuity in mass transport properties for oxygen and hydrogen chloride. This suggests that the multilayer assembly gives a good approximation to a continuous single layer under photolysis, and yet remains separably analyzable afterwards.


The authors would like to thank Dr. S. W. Fu for contributing mirage effect depth profile data in support of this method. The authors would also like to thank the Natural Sciences and Engineering Research Council of Canada (NSERC) for their ongoing support of this work.


(1.) G. C. Marks and G. Butters, J. MacromoL Sci-Chern. A12, 569 (1978).

(2.) K. S. Minsker, S. V. Kolesov, and G. E. Zaikov, Aging and Stabilization of Polymers Based on Vinylchloride, Nauka, Moscow (1982).

(3.) G. E. Schoolenberg and P. Vink, Polymer, 32, 432 (1991).

(4.) G. E. Schoolenberg and H. D. F. Meijer. Polymer, 32, 438 (1991).

(5.) A. P. Alexandrov, V. N. Genkin, and V. V. Sokolov, Polymer Science USSR, 27, 1188(1985).

(6.) J. Gardette and J. Lemaire, Polym. Degr. Stab., 25, 293 (1989).

(7.) J. Gardette and J. Lemaire, Polym. Degr. Stab., 33, 77 (1991).

(8.) N. M. Biturin, V. N. Genkin, V. P. Lebedev, A. V. Nikitin, V. V. Sokolov, L. D. Strelkova, and T. G. Fedoseyeva, Polymer Science USSR, 25, 1, 91(1983).

(9.) K. Maruyama, Y. Kuramoto, M. Yagi, and Y. Tanizaki, Polymer. 29, 24 (1988).

(10.) D. Braun and E. Bezdadea, Encyclopaedia of PVC, Second Edition. ch. 8, pp. 397-433, L. I. Nass and C. A. Heiberger, eds., Marcel Dekker, New York (1986).

(11.) O. V. Nepotchakyh, M. I. Adullin, K. S. Minsker, G. E. Zaikov, and C. R. H. I. de Jonge, Polym. Degrad. Stab., 18, 349 (1987).

(12.) J. F. Power and M. C. Prystay, Appl. Spec., 49, 709 (1995).

(13.) J. F. Power and M. C. Prystay, Appl. Spec., 49, 725 (1995).

(14.) M. C. Prystay and J. F. Power, Polym. Eng. Sci., 36, 685 (1996).

(15.) M. A. Schweitzer and J. F. Power, Appl. Spec., 48, 1076-1087 (1993).

(16.) J. F. Power and M. A. Schweitzer. in Photacoustic and Photothermal Phenomena: Tenth International Conference, M. Bertolotti and F. Scudieri, eds. AIP Conference proceedings 463, pp. 190-94, Woodbury, N.Y. (1999).

(17.) J. F. Power, in Volume 3 III International Workshop--Advances in Signal Processing for NDE of Materials, pp. 57-63, X. P. Maldague, technical editor, Topics in Nondestructive Evaluation Series, American Society for Non-destructive Testing (1998).

(18.) J. F. Power, S. W. Fu, and M. A. Schweitzer, Appl. Spec., 54, 110 (2000).

(19.) S. W. Fu and J. F. Power, Appl. Spec., 54, 127 (2000).

(20.) J. F. Power, App. Spec., 45, 1252 (1991).

(21.) M. Iqbal. An Introduction to Solar Radiation, Academic Press Inc., Toronto (1983).

(22.) K. Patel, A. Velazquez, H. S. Calderon, and G. R. Brown, J. Appl. Polym, Sci., 46, 179 (1992).

(23.) L. C. Sawyer and D. C. Grubb, Polymer Microscopy, p. 21, Chapman-Hall London (1987).

(24.) J. F. Power and S. W. Fu. Appl. Spec, 52, 1507(1999).

(25.) K. S. Minsker, M. I. Abdullin. and V. A. Kraikin, Plasticheskie Massy, 3, 31(1980).

(26.) L. D. Strelkova, G. T. Fedoseyeva, and K. S. Minsker, Vysokomolecular. Soed., 18. No. 9, 2064 (1976).

(27.) N. M. Biturin, V. N. Genkin, and V. V. Sokolov, Vysokomolecular. Soed, short reports, 23. No. 3, 221 (1981).
COPYRIGHT 2000 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2000 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Publication:Polymer Engineering and Science
Geographic Code:1CANA
Date:Aug 1, 2000
Previous Article:Coupled Thermo-Mechanical Analysis for Plastic Thermoforming.
Next Article:Rotational Molding of Plastics: Comparison of Simulation and Experimental Results for an Axisymmetric Mold.

Related Articles
Photothermal imaging of dehydrochlorination in poly(vinyl chloride) films.
The effect of interfacial instabilities on the strength of the interface in two-layer plastic structures.
Effect of interdiffusion regions on penetrant flux through multilayer films.
Layers break apart in controlled way. (Making Polymers That Self-Destruct).
Production of barrier films by chaotic mixing of plastics.
Characterization of combinatorial polymer blend composition gradients by FTIR microspectroscopy.
LLDPE-based mono- and multilayer blown films: property enhancement through coextrusion.
Multilayer Film Comprising An Amorphous Polymer.
Influence of UV absorber on photodegradation processes of poly(vinyl chloride) with different average degrees of polymerization.
Role of the interphase in the flow stability of reactive coextruded multilayer polymers.

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters