- Email: [email protected]
Property changes in polyoxymethylene (POM) resulting from processing, ageing and recycling
Available online at www.sciencedirect.com
Polymer Degradation and Stability 92 (2007) 2181e2189 www.elsevier.com/locate/polydegstab
Property changes in polyoxymethylene (POM) resulting from processing, ageing and recycling V.-M. Archodoulaki*, S. Lu¨ftl, T. Koch, S. Seidler Institute of Materials Science and Technology, Vienna University of Technology, Favoritenstrasse 9-11, A-1040 Vienna, Austria Received 29 November 2006; received in revised form 15 January 2007; accepted 5 February 2007 Available online 11 August 2007
Abstract The degradation processes initiated by thermo-mechanical and thermo-oxidative loading as well as exposure to ultraviolet irradiation weathering were examined in commercially available semi-crystalline polyoxymethylene (POM), using predominantly thermo-analytical methods. With reference to different injection moulding conditions and moulding geometries (such as loudspeaker grilles and safety-belt components) it is demonstrated that POM-copolymer chain is not affected, even under higher shear stresses and complicated moulding geometries, if an appropriate additive is used. Thermogravimetric Analysis (TGA) was used to observe stabiliser consumption and further degradation, whereas investigations of melt-flow index and molar mass show effects correlated to late-term changes of the molar mass distribution. Mass-spectrometry investigations performed in parallel to the thermogravimetric analysis identified formaldehyde and carbon dioxide as the main degradation products. Weathering by ultraviolet irradiation results in damage similar to the thermo-oxidative impact. Generally, the copolymers are less sensitive to thermo-mechanical and thermo-oxidative degradations than the homopolymers due to the chain modification. Furthermore, reprocessing results in a decrease of the elongation at break and thermo-oxidative induced degradation proceeds faster for the additional UV-stabilised materials. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Polyoxymethylene; Reprocessing; Degradation; Thermal ageing; UV-irradiation; Thermal analysis (TGA)
1. Introduction A specific characteristic of polymeric materials is the alteration of their properties by processing, use and recycling. Because of their polymeric structure, chain damage mechanisms initiated by temperature, shear or UV-light lead to dissociation and in further consequence to failure of the component. The degradation is of thermo-mechanical, thermo-oxidative and photo-oxidative nature. Kern and Cherdron investigated the degradation behaviour of POM in detail in the early 1960s [1,2]. Copolymers and their specific reduction of the degradation sensitivity were investigated by several authors [3e6]. A good overview about the stabilisers used in POM is given in Refs. [7,8]. * Corresponding author. Tel.: þ43 1 58801 30850; fax: þ43 1 58801 30895. E-mail address: [email protected] (V.-M. Archodoulaki). 0141-3910/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.polymdegradstab.2007.02.024
It is well known that a large number of reactions take place during processing and small changes in the molecular structure result in considerable modification of properties. In the case of POM, a scheme is presented in Fig. 1 showing the interrelation of reactions. Many examples can be found in the literature in which a chemical characteristic is used to monitor the degradation progress. In some cases a good correlation with the deterioration of the mechanical properties is found, but knowledge about the correlation between chemical changes and the consequent mechanical changes is very poor. Methods for assessing the changes that are promoted by molecular degradation and under service lifetime are of great importance. The most sensitive methods may indicate that the degradation has occurred long before it is apparent in the engineering properties. For POM a good approach for the quantification of degradation induced under service lifetime is to determine
2182
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189 Table 1 Materials of interest Material A Copolymer
Copolymer
B Copolymer
Copolymer
Fig. 1. Scheme of the resulting effects caused by the interactions of thermomechanical, thermo-oxidative and photo-oxidative loadings in a polymer.
stabiliser consumption and beginning degradation by means of TGA. Basically the decrease of the onset temperature in the thermogravimetric Analysis (TGA) is the first indication for thermo-oxidative damage, whereas a decrease of the onset temperature without a change of the first derivative of the TG signal (DTG) refers explicitly to stabiliser consumption. Degradation can be determined by a change of the DTG signal [9].
2. Experimental 2.1. Materials Two differently stabilised homopolymers DELRINÒ 900P (solely heat stabiliser) and DELRINÒ 927 P and 127 P (heat & UV-stabiliser) and four copolymer types UltraformÒ W2320 003 (solely heat stabiliser), UltraformÒ W2320 U03 (heat & UV-stabiliser) and HostaformÒ 27021(solely heat stabiliser) and HostaformÒ 27021 LS (heat & UV-stabiliser) were used for the investigations. The resins are commercially available and designated for injection moulding. The samples for the experiments were injection moulded specimens (sample geometry according to ISO 527-2, Type 1A), injection moulded loudspeaker grilles and injection moulded safety-belt components (press button). All samples were produced in accordance with processing conditions recommended by the producer (homopolymers: temperature of the nozzle 190 C, mould temperature 90 C; copolymers: temperature of the nozzle 215 C, mould temperature 90 C). In order to exclude hydrolytic degradation all materials were pre-dried for 4 h at 80 C before moulding. An overview of the sample geometries and assignments is given in Table 1. All investigated materials had similar Melt Volume Rates (20e25 cm3/10 min) except the heat & UV-stabilised DELRINÒ 127 P with a MVR of 5 cm3/10 min and are available for comparable applications.
Stabiliser
Acronym
Heat
A1 pellet A11 first processing step A17 seventh processing step Heat þ UV A2 pellet A21 first processing step A27 seventh processing step
Investigated geometries Pellet Specimens Pellet Specimens
Heat
B1 pellet Pellet B11 first processing step Specimens B17 seventh processing step Loudspeaker grilles Heat þ UV B2 pellet Pellet, specimens B21 first processing step Loudspeaker grilles B27 seventh processing step Press button
H Homopolymer Heat
H1 Pellet H11 first processing step H17 seventh processing step Homopolymer Heat þ UV H2 pellet H21 first processing step H27 seventh processing step
Pellet Specimens Pellet Specimens Press button
Thermal ageing was performed at 140 C (maximum temperature deviation: 1.5 C) in an oven (Heraeus) for 21, 35 and 56 d. Samples will be labelled as 21 d/140 C, 35 d/ 140 C and 56 d/140 C in the following. For UV-irradiation an artificial weathering test using a Xenontester (Suntest XLSþ, Atlas) was carried out. This test simulates the standard climatic conditions of car indoor irradiation for Central Europe (irradiation dose equivalent for 1 year 1709 MJ/m2, black panel temperature w50 C). Samples exposed to UV-irradiations will be denoted as 1a UV (¼1 year) and 2a UV (¼2 years). 2.2. Methods Non-isothermal TGA: Samples of about 10 mg were heated in air and nitrogen atmosphere with a purge gas stream of 100 ml/min in ceramic pans with a heating rate of 10 K/min. The thermogravimetric measurements were performed using a TGA 2050 (TA-Instruments). The onset temperature (i.e. the degradation start temperature) was defined as the temperature at which a mass loss of 3% of the initial sample mass was reached [10]. Coupling with a mass-spectrometer (MS) provided a simultaneous analysis of specific degradation products that evolved during thermal degradation. The mass spectrum reflects the mass-to-charge ratio and the relative abundance of the ion fragments, thus providing a fingerprint of each compound. Gases evolving during the decomposition of the sample were registered after every 10 C step of temperature increase. The MS used was a Thermostar (Balzers). The coupling consisted of a heated quartz capillary tube (120 C) connecting the TGA furnace outlet with the MS gas inlet through a pinhole diaphragm.
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189
For isothermal TGA investigations samples of about 10 mg were heated in air atmosphere with a purge gas stream of 100 ml/min in ceramic pans until 200 C and then kept isothermal for hours. Melt Volume Rate (MVR): The MVR is obtained by measuring the amount of molten material flowing through a defined orifice under a defined mass. Comparing the same polymer, melts of low-molar mass polymers flow more rapid than melts of high-molar mass polymers. The MVR experiments were carried out with 2.16 kg load and at 190 C in a ZWICK 4105.01/03 device. Tensile tests: The monotonic loading tests conducted under constant crosshead speed conditions were performed in accordance to ISO 527-2. The crosshead speed for assessing the modulus of elasticity was 1 and 50 mm/min for the determination of the tensile strength and the elongation at break. Nanoidendation: For the determination of mechanical properties of the complex shaped press buttons small pieces were cut and a typical cross-sectional preparation including embedding in epoxy resin, grinding and polishing was carried out. Instrumented microhardness tests were done using a Nanoindenter XP (MTS Systems Inc.). The indentation depth was 1 mm and the indentation rate 0.05 mm/s. After a holding time of 30 s at maximum load the specimens were unloaded. The indentation hardness HIT was calculated at the maximum load by means of the contact depth, according to Ref. [11]. The measurements were done at two different positions on the cross-section, at a distance of 50 mm from the surface and in the core region. Average values were calculated from 10 indents at every position. Molar mass investigations: The method used was gel permeation chromatography (GPC). The investigations were performed at elevated temperature (140 C) with a PL-GPC 220 instrument (Polymer Laboratories), equipped with two separation columns and a differential refractometer. N,N-Dimethylformamide (DMF) was used as eluent at a flow rate of 0.5 ml/min. The calibration was made with polystyrene standards. 3. Results 3.1. Influence of the processing conditions and comparison of different geometries Several processing conditions (e.g. injection velocity and nozzle temperature) were investigated in order to verify the influence of the processing conditions and to find out to what extent degradation takes place during processing. Furthermore, different injection moulding geometries and their influence on the degradation behaviour were compared. Fig. 2a and b compare the results of the copolymers and homopolymers after variation of the processing conditions. It can be shown that even when the temperature of the nozzle is increased the copolymers remain thermally stable (Fig. 2a), whereas the homopolymer (H1) in Fig. 2b shows a shoulder indicating some chains that are affected due to the higher temperature. By increasing the injection moulding velocity the formed flash
2183
Fig. 2. TGA of POM copolymer (a) and POM homopolymer (b), after different processing conditions.
(obviously degraded material) shows a decreased onset temperature and an altered DTG signal corresponding to a total degraded POM (Fig. 2b). TGA investigations in nitrogen atmosphere reveal that there is no influence of the used mould geometry on the degradation behaviour of POM copolymer (Fig. 3). On the other hand, isothermal TGA investigations on pellets and specimens clearly show that the first processing step leads to stabiliser consumption and reduces the thermal stability of all investigated materials extracted from specimens or structural components (Fig. 4). 3.2. Influence of multi-processing and ageing TGA investigations reported in an earlier publication [9] reveal that thermo-oxidative processes decrease the degradation onset of the homo- and copolymer pellets, since the secondary products of the auto-oxidation enhance the auto-oxidation and/ or auto-catalytic cleavage of the CeOeC-bonds. Random scission of the polymer chains and the formation of macro-alkyl
2184
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189
changes in the crystallinity for either material could be found by means of DSC with respect to ageing [13]. The TGA investigations of the heat-stabilised homopolymer (H1) in comparison to additive-extracted material indicate continuous stabiliser consumption and chain damage after thermo-oxidative loading (Fig. 5a). The TG and DTG curves of the oven-stored material are shifted towards the curve of the additive-extracted material; the onset temperature is lowered close to the temperature range in which the thermal degradation of the pure POM-chain starts (according to Kern [1], thermal degradation onset is above 270 C in nitrogen atmosphere). The onset temperature of the H1156 d/140 C is 218 C, the onset temperature of the additivepurified sample is 206 C. Because of the change in the shape of the DTG signal we can conclude that degradation took place. Fig. 3. TG and DTG curves of POM copolymer: influence of the used mould geometry on the degradation behaviour.
radicals, which form alkyl peroxides and hydroperoxides in the presence of even a small amount of oxygen, lead to a subsequent degradation. While the solely heat-stabilised homopolymer (H1) shows some stabiliser consumption after multiple processing, the solely heat-stabilised copolymers (B1 and A1) are more stable against thermo-mechanical degradation. According to the TGA/MS results for the homopolymer, stabiliser consumption occurs during reprocessing and further oven-storage. In the copolymers, the stabiliser is less affected during accelerated ageing due to the presence of co-monomer units; the degradation takes place only in the amorphous phase [12]. The homopolymers and the copolymers differ in crystallinity because of their molecular structure, but no significant
Fig. 4. Isothermal TGA on pellets and specimens in air atmosphere, performed at 200 C.
Fig. 5. TGA of the homopolymer after thermo-oxidative loading in comparison to the pellet and the additive-extracted pellet (a); MS-spectra of the homopolymer after thermo-oxidative loading in comparison to the initial state (b) (m/z: 30eformaldehyde; 31emean peak trioxane, methoxy fragment; 44ecarbon dioxide, ethylene oxide, urea fragment; 57eadditive fragment; 60eacetic acid, methyl formate, urea; 77ephenyl fragment).
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189
Fig. 6. TG curves of two homopolymers with different stabiliser packages in initial state and after thermo-oxidative loading.
The MS-spectra show that the ion current response for antioxidant fragments (m/z ¼ 57 (t-butyl group) and 77 (phenyl group)) is relatively poor after thermo-oxidative loading in comparison to the initial state (Fig. 5b). Total stabiliser
2185
consumption had happened during an oven-storage of 56 d at 140 C. In Fig. 6 the TG curves of two homopolymers (H1 and H2) with different stabiliser packages are illustrated, in initial state and after oven-storage. It is clearly shown that already in the initial state the additional UV-stabilised homopolymer H2 is thermally superior to the heat-stabilised H1. After thermo-oxidative loading, although homopolymer H2 is also subjected to auto-oxidation and degradation due to the secondary products of the auto-oxidation, the thermal stability of H2 remains superior. One explanation for the higher thermal stability of the homopolymer H2 is that it may contain reagents such as urea, polyamides or polyurethane, which prevent splitting by secondary products of the auto-oxidation as is referred to in Ref. [1]. However, these reagents could not be verified with the help of mass spectrometry. A comparison of two different heat- and UV-stabilised copolymers (left/right) can be seen in Fig. 7. The DTG curves (top) show a different degradation performance due to thermo-oxidative loading. The molar mass distributions in initial state and after thermo-oxidative loading (bottom) show that copolymer 2 is clearly more affected than copolymer 1 (due to business rivalry no further identity of copolymers 1
Fig. 7. TGA of two copolymers after thermo-oxidative loading in comparison to the specimen, pellet and the additive-extracted pellet (top); copolymer 1 (a) and copolymer 2 (b); molar mass distributions in initial state and after 56 d/140 C (bottom); copolymer 1 (c) and copolymer 2 (d).
2186
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189
and 2 is available). This implies that different commercial products with comparable stabiliser packages for similar applications show different thermal stability after thermo-oxidative loading. 3.3. UV-irradiation The irradiation exposed surfaces of homo- and copolymers (H1, A1 and B1) without any UV-stabiliser after 1 year artificial exposure (even indoor exposure dose) are totally chalked and cracks have developed. According to Gardette et al. [14] the following degradation mechanism for photo-oxidation takes place in POM: the first step leads to the oxidation of the carbon atoms with the formation of secondary hydroperoxides that decomposes into two radicals: an alkoxy-macroradical and a hydroxy radical. The alkoxy-macroradical can react in two possible ways: a cage reaction may occur leading to the formation of carbonate and water, and a b-scission which leads to a formate and an alkoxy-macroradical. Hydrogen abstraction from the polymeric chain by macroradicals leads to the formation of alcohols. Thereby a new macroradical is created. In contrast, the formation of hydroperoxides leads to chain scission via b-scission reaction and formaldehyde liberation [1]. The results of the TGA and GPC investigations indicate that similar degradation mechanisms occur in the case of thermo-oxidative loading and UV-irradiation. Fig. 8 illustrates the TGA and GPC traces of the heat-stabilised homopolymer, similar results were obtained for the heat-stabilised copolymers. Fig. 8a presents the DTG results in the initial state compared with the conditions after thermo-oxidative loading and UV-irradiation, the shift of the curves towards lower temperature indicates degradation. Fig. 8b demonstrates the corresponding GPC results. Because of the very narrow molar mass distribution (MMD) in the initial condition, only a shift of the MMD and no change in the MMD shape is expected. The UV-irradiated samples of all heat- and UV-stabilised polymers show no visible changes on the surface, TGA investigations of the irradiated samples indicated no degradation. All materials were stabilised properly against photo-oxidation. 3.4. MVR and mechanical properties Oven-storage leads to an increase in the MVR of homopolymers and copolymers, all UV-stabilised materials show a higher MVR increase than the solely heat-stabilised materials, i.e. the melt viscosity is reduced, although the oxidative stability of both systems is comparable in the initial state. This reduction can be explained as a change in the MMD. The used UV-stabiliser evidently decreases the thermo-oxidative stability of the material during ageing in the oven at 140 C. Similar results were reported in Ref. [15], where in contrast to the solely heat-stabilised materials all heat- and UV-stabilised grilles show a more intense decrease in the deflection at break after ageing of 35 d/140 C. Fig. 9 illustrates the changes in the MVR as a funtion of the storage duration. Higher values of the MVR after storage
Fig. 8. Degradation behaviour of heat-stabilised homopolymer after thermooxidative loading and UV-irradiation. DTG curves of the homopolymer (seventh processing step) in initial condition, after thermo-oxidative loading (21 d/ 140 C) and 1a UV-irradiation (car indoor irradiation, Central Europe) (a); GPC curves of the homopolymer (seventh processing step) in initial condition, after thermo-oxidative loading (21 d/140 C) and 1a UV-irradiation (car indoor irradiation, Central Europe) (b).
duration of 56 d at 140 C are measured for the copolymer 2, which also shows the major MMD changes. Whereas for all investigated materials the elongation at break decreases due to reprocessing [13], the modulus of elasticity is less sensitive to changes in the chemical and/or physical structure. The modulus of elasticity of the homopolymers and copolymers remains stable even though the number of processing steps is increased. It seems, that the interaction between amorphous and crystalline phase has changed, for example a variation of the nature and number of the tie-molecules. Another explanation for the decrease of the elongation at break is an increase of voids due to shrinkage effects of the polymer during cooling from the mould. Especially in thick parts like tensile test specimens and because of suboptimal
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189
2187
Fig. 9. Changes in the MVR and elongation at break in dependence of the thermo-oxidative loading duration, results illustrated for the copolymer B2. 1st PS: first processing step, 4th PS: fourth processing step, 7th PS: seventh processing step.
conditions used for processing such shrinkage effects are not unexpected. Polarised optical microscopy of thin section cuts through the samples reveals that the crystalline super molecular structure is almost not altered; i.e. the MMD should be unchanged after reprocessing. TGA investigations have already indicated stabiliser consumption, which cannot be monitored by GPC. However, the tensile test is not a sensitive method to detect embrittlement since the embrittled layer makes up only a small proportion of the overall sample cross-section. The modulus of elasticity from the tensile test is an integral property for injection moulded specimens, which have a heterogeneous structure from edge to core depending on the cooling conditions. Measurements of the modulus of elasticity and hardness over the specimen cross-section can be made with instrumented nanoindentation tests and the dependence of the modulus of elasticity and hardness with the crystalline structure can be confirmed. At the edge, the area that has been cooled with the fastest cooling rate and less crystalline structure is visible; the modulus of elasticity is low. In the shear zone, an area with a lower cooling rate and a fine spherulitic structure is present; the modulus of elasticity is higher than at the edge. In the core of the specimen, in which ‘‘constant’’ crystallisation conditions dominate, cooling and heat of crystallisation are interactive; the modulus of elasticity is high [13,16]. The influence of thermo-oxidative ageing at elevated temperature and UV-irradiation on the hardness of the press button is illustrated in Fig. 10. An increase in hardness is provoked by degradation, leading to decreased inter-crystalline interactions. Annealing effects could not be proved. Results from the nanoindentation testing showed increased hardness and modulus in thermo-oxidative loaded regions for the homopolymer (Fig. 10a). While oven-storage of 56 d leads
Fig. 10. Hardness determined by instrumented nanoindentation measurements in initial state and after thermo-oxidative loading and UV-irradiation for the homopolymer (a) and copolymer (b).
to an increase of the modulus of elasticity and hardness caused by embrittlement of the homopolymers (Fig. 10a), the hardness of the copolymers shows less change (Fig. 10b). The data also indicate that the mechanical properties and level of thermo-oxidation are strongly correlated. Polarised optical microscopy of thin sections cut through the press buttons reveals that the crystalline structure is different for the investigated materials. While for the homopolymers, due to the higher melt viscosity and the resulting higher shear rates, shear induced crystallisation effects cannot be excluded, for the copolymers a typical skin-score morphology e as described above e can be deduced. This explains the higher values of hardness for the homopolymers and the unusual higher values of the hardness (and indentation modulus) for the skin region compared to the core for these materials (Fig. 10a). In Fig. 10 the influence of UV-irradiation on the hardness measured in the initial state and after 1a UV and 2a UV is plotted.
2188
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189
As discussed in the previous section, for the solely heatstabilised homopolymers repeated processing leads to stabiliser consumption after six reprocessing steps, while for the copolymers (in any stabilisation state) no stabiliser consumption after repeated processing could be found.
4. Conclusions
Fig. 11. Influence of recycling on the degradation behaviour of POM copolymer determined by TGA; reprocessing of copolymer grills, 1st PS: first processing step, 3rd PS: third processing step (a) and addition of 20% recycling material (b).
UV-irradiation causes no changes in the mechanical properties in comparison to the initial state in both of the heat- and UV-stabilised materials. 3.5. Influence of addition of recycled material The most common form of recycling is done directly in the factory, where waste such as sprues and runners is regranulated and used to produce new products of the same kind. Normally the fraction of reground material mixed with virgin material is small to avoid sensitivity to molecular degradation. The influence of recycling on the degradation behaviour of POM-copolymer components can be seen in Fig. 11. Injection moulded grilles after three processing steps (Fig. 11a) and ones with an addition of a mass fraction of 20% recycled material show no altered degradation behaviour (Fig. 11b).
Differently stabilised POM homopolymers and copolymers were investigated to monitor the degradation initiated by processing, artificial ageing and recycling. The TGA method is well known as a good approach to prove stabiliser consumption and induced degradation. The TGA results indicate random scission of the main chain as the initiation mechanism of degradation of POM. Depression of the onset temperature of the TG curves is attributable to continuous stabiliser consumption. Changes in the DTG signal compared to the initial state are directly related to chain damages. In dependence of the applied stabiliser system homopolymers show stabiliser consumption and chain damages due to multiple processing. Thermo-oxidative ageing leads to further stabiliser consumption followed by chain damage and degradation. In spite of multiple processing all investigated copolymers are thermally stable. After 21 d/140 C, degradation can be detected for the homopolymers. Due to the presence of the co-monomer units copolymers are more stable, and therefore chain damage could be proved only after a duration of 56 d/140 C, obviously oven-storage affects the homopolymers more than the copolymers. After 1 year irradiation (car indoor, Central Europe), the solely heat-stabilised materials present a chalked surface and cracks. The results of the TGA and GPC investigations indicate that similar degradation mechanisms occur in case of thermo-oxidative loading and UV-irradiation. In the both heat- and UV-stabilised materials (homo- and copolymers) UV-irradiation does not affect the mechanical properties. Thermo- and photo-oxidation of POM proceed according to a two-fold mechanism: (a) decomposition of the primary peroxides formed by mechanical-oxidative degradation induced by processing and (b) oxidation of radicals as discussed in the Bolland mechanism of auto-oxidation of polymers. In thermo-oxidation b-scission of the alkoxy radicals is the dominating reaction whereas in photo-oxidation hydroperoxide decomposes through a cage reaction or to an alkoxy radical and a hydroxy radical. Alternatively an alkoxy radical can form an aldehyde and a new radical by b-scission. The investigations proved that the progress of the degradation of POM is independent of the used mould geometries, in a similar manner. Stabiliser consumption takes place, primarily during the first processing step from pellet to part, no differences in the extent could be proved by differently investigated mould geometries. It is demonstrated that POM copolymers tolerate higher fluctuations of the processing parameters before degradation occurs than POM homopolymers. On the other hand
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189
homopolymers show enhanced mechanical properties compared to the copolymers, even after artificial ageing. The impact of chain damage on the mechanical properties is different for the analysed materials: thermo-oxidative loading of the heat- and UV-stabilised materials leads to higher embrittlement and higher-MVR values. In contrast to the solely heat-stabilised material the thermo-oxidative induced degradation progresses faster, due to the added stabiliser. This effect has been found in the homopolymers as well as copolymers. Reprocessing leads to a decrease of the elongation at break in all materials. The explanation for this behaviour is that the nature as well as the number of tie-molecules are altered.
References [1] Kern W. Chemische Elementarvorga¨nge bei der Alterung von Kunststoffen. Chemiker-Ztg/Chem Apparatur 1967;91:255e62. [2] Kern W, Cherdron H. Der Abbau von Polyoxymethylenen. Makromol Chem 1960;40:101e17. [3] Hermann HD, Weissermel K. Die Copolymerisation von Formaldehyd mit cyclischen Formalen. Makromol Chem 1966;94:91e102. [4] Weißermel K, Fischer E, Gutweiler K, Hermann HD. Zur Copolymerisation des Trioxans. Kunststoffe 1964;54:410e5. [5] Inoue M. Crystallization and melting of copolymers of polyoxymethylene. J Appl Polym Sci 1964;8:2225e38.
2189
[6] Ivanova LL, Grebennikova NA, Sokolova GP, Dudina LA, Irzhak VI, Yenikolopyan NS. Heat stability of a formaldehyde-1,3-dioxolane copolymer. Vysokomol Soyed 1976;A18(4):920e5. [7] Zweifel H, editor. Plastic additives handbook. 5th ed. Mu¨nchen: Hanser Verlag; 2000. [8] Sabel HD, Schlaf H, Unger P, Ziegler U. Polyacetale. In: Bottebruch L, editor. Technische Thermoplaste: Polycarbonate, Polyacetale, Polyester, Cellulosester. Mu¨nchen: Carl Hanser Verlag; 1992. p. 300e95. [9] Archodoulaki VM, Lu¨ftl S, Seidler S. Thermal degradation behaviour of poly-(oxymethylene): 1. Degradation and stabiliser consumption. Polym Degrad Stab 2004;86:75e83. [10] Lu¨ftl S, Archodoulaki VM, Seidler S. Thermal-oxidative induced degradation behaviour of poly(oxymethylene) copolymer detected by TGA/MS. Polym Degrad Stab 2006;91:464e71. [11] Oliver WC, Pharr GM. An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. J Mater Res 1992;7/6:1564e83. [12] Mucha M. Thermooxidation rate of diacetate terminated polyoxymethylene with various morphological structures. Colloid Polym Sci 1984;262: 841e50. [13] Archodoulaki VM. Eigenschaftsa¨nderungen von Polyoxymethylenen induziert durch Verarbeitung, Alterung und Recycling. Fortschritt-Berichte VDI, Reihe 5, Nr. 705. Du¨sseldorf: VDI-Verlag; 2005. [14] Gardette JL, Sabel HD, Lemaire J. Photooxidation of polyacetal copolymer I. Angew Makromol Chem 1991;188:113e28. [15] Archodoulaki VM, Lu¨ftl S, Seidler S. Effects of coloration on the physical and mechanical properties and on the thermal stability of polyoxymethylene. Adv Polym Tech 2006;25(1):63e72. [16] Wolters E, Buchberger HL, Kohlhepp KG. Zusammenhang zwischen Gefu¨geaufbau und mechanischem Verhalten von Spritzgussteilen aus Hostaform. Plastverarbeiter 1973;9(24):549e55.
Polymer Degradation and Stability 92 (2007) 2181e2189 www.elsevier.com/locate/polydegstab
Property changes in polyoxymethylene (POM) resulting from processing, ageing and recycling V.-M. Archodoulaki*, S. Lu¨ftl, T. Koch, S. Seidler Institute of Materials Science and Technology, Vienna University of Technology, Favoritenstrasse 9-11, A-1040 Vienna, Austria Received 29 November 2006; received in revised form 15 January 2007; accepted 5 February 2007 Available online 11 August 2007
Abstract The degradation processes initiated by thermo-mechanical and thermo-oxidative loading as well as exposure to ultraviolet irradiation weathering were examined in commercially available semi-crystalline polyoxymethylene (POM), using predominantly thermo-analytical methods. With reference to different injection moulding conditions and moulding geometries (such as loudspeaker grilles and safety-belt components) it is demonstrated that POM-copolymer chain is not affected, even under higher shear stresses and complicated moulding geometries, if an appropriate additive is used. Thermogravimetric Analysis (TGA) was used to observe stabiliser consumption and further degradation, whereas investigations of melt-flow index and molar mass show effects correlated to late-term changes of the molar mass distribution. Mass-spectrometry investigations performed in parallel to the thermogravimetric analysis identified formaldehyde and carbon dioxide as the main degradation products. Weathering by ultraviolet irradiation results in damage similar to the thermo-oxidative impact. Generally, the copolymers are less sensitive to thermo-mechanical and thermo-oxidative degradations than the homopolymers due to the chain modification. Furthermore, reprocessing results in a decrease of the elongation at break and thermo-oxidative induced degradation proceeds faster for the additional UV-stabilised materials. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Polyoxymethylene; Reprocessing; Degradation; Thermal ageing; UV-irradiation; Thermal analysis (TGA)
1. Introduction A specific characteristic of polymeric materials is the alteration of their properties by processing, use and recycling. Because of their polymeric structure, chain damage mechanisms initiated by temperature, shear or UV-light lead to dissociation and in further consequence to failure of the component. The degradation is of thermo-mechanical, thermo-oxidative and photo-oxidative nature. Kern and Cherdron investigated the degradation behaviour of POM in detail in the early 1960s [1,2]. Copolymers and their specific reduction of the degradation sensitivity were investigated by several authors [3e6]. A good overview about the stabilisers used in POM is given in Refs. [7,8]. * Corresponding author. Tel.: þ43 1 58801 30850; fax: þ43 1 58801 30895. E-mail address: [email protected] (V.-M. Archodoulaki). 0141-3910/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.polymdegradstab.2007.02.024
It is well known that a large number of reactions take place during processing and small changes in the molecular structure result in considerable modification of properties. In the case of POM, a scheme is presented in Fig. 1 showing the interrelation of reactions. Many examples can be found in the literature in which a chemical characteristic is used to monitor the degradation progress. In some cases a good correlation with the deterioration of the mechanical properties is found, but knowledge about the correlation between chemical changes and the consequent mechanical changes is very poor. Methods for assessing the changes that are promoted by molecular degradation and under service lifetime are of great importance. The most sensitive methods may indicate that the degradation has occurred long before it is apparent in the engineering properties. For POM a good approach for the quantification of degradation induced under service lifetime is to determine
2182
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189 Table 1 Materials of interest Material A Copolymer
Copolymer
B Copolymer
Copolymer
Fig. 1. Scheme of the resulting effects caused by the interactions of thermomechanical, thermo-oxidative and photo-oxidative loadings in a polymer.
stabiliser consumption and beginning degradation by means of TGA. Basically the decrease of the onset temperature in the thermogravimetric Analysis (TGA) is the first indication for thermo-oxidative damage, whereas a decrease of the onset temperature without a change of the first derivative of the TG signal (DTG) refers explicitly to stabiliser consumption. Degradation can be determined by a change of the DTG signal [9].
2. Experimental 2.1. Materials Two differently stabilised homopolymers DELRINÒ 900P (solely heat stabiliser) and DELRINÒ 927 P and 127 P (heat & UV-stabiliser) and four copolymer types UltraformÒ W2320 003 (solely heat stabiliser), UltraformÒ W2320 U03 (heat & UV-stabiliser) and HostaformÒ 27021(solely heat stabiliser) and HostaformÒ 27021 LS (heat & UV-stabiliser) were used for the investigations. The resins are commercially available and designated for injection moulding. The samples for the experiments were injection moulded specimens (sample geometry according to ISO 527-2, Type 1A), injection moulded loudspeaker grilles and injection moulded safety-belt components (press button). All samples were produced in accordance with processing conditions recommended by the producer (homopolymers: temperature of the nozzle 190 C, mould temperature 90 C; copolymers: temperature of the nozzle 215 C, mould temperature 90 C). In order to exclude hydrolytic degradation all materials were pre-dried for 4 h at 80 C before moulding. An overview of the sample geometries and assignments is given in Table 1. All investigated materials had similar Melt Volume Rates (20e25 cm3/10 min) except the heat & UV-stabilised DELRINÒ 127 P with a MVR of 5 cm3/10 min and are available for comparable applications.
Stabiliser
Acronym
Heat
A1 pellet A11 first processing step A17 seventh processing step Heat þ UV A2 pellet A21 first processing step A27 seventh processing step
Investigated geometries Pellet Specimens Pellet Specimens
Heat
B1 pellet Pellet B11 first processing step Specimens B17 seventh processing step Loudspeaker grilles Heat þ UV B2 pellet Pellet, specimens B21 first processing step Loudspeaker grilles B27 seventh processing step Press button
H Homopolymer Heat
H1 Pellet H11 first processing step H17 seventh processing step Homopolymer Heat þ UV H2 pellet H21 first processing step H27 seventh processing step
Pellet Specimens Pellet Specimens Press button
Thermal ageing was performed at 140 C (maximum temperature deviation: 1.5 C) in an oven (Heraeus) for 21, 35 and 56 d. Samples will be labelled as 21 d/140 C, 35 d/ 140 C and 56 d/140 C in the following. For UV-irradiation an artificial weathering test using a Xenontester (Suntest XLSþ, Atlas) was carried out. This test simulates the standard climatic conditions of car indoor irradiation for Central Europe (irradiation dose equivalent for 1 year 1709 MJ/m2, black panel temperature w50 C). Samples exposed to UV-irradiations will be denoted as 1a UV (¼1 year) and 2a UV (¼2 years). 2.2. Methods Non-isothermal TGA: Samples of about 10 mg were heated in air and nitrogen atmosphere with a purge gas stream of 100 ml/min in ceramic pans with a heating rate of 10 K/min. The thermogravimetric measurements were performed using a TGA 2050 (TA-Instruments). The onset temperature (i.e. the degradation start temperature) was defined as the temperature at which a mass loss of 3% of the initial sample mass was reached [10]. Coupling with a mass-spectrometer (MS) provided a simultaneous analysis of specific degradation products that evolved during thermal degradation. The mass spectrum reflects the mass-to-charge ratio and the relative abundance of the ion fragments, thus providing a fingerprint of each compound. Gases evolving during the decomposition of the sample were registered after every 10 C step of temperature increase. The MS used was a Thermostar (Balzers). The coupling consisted of a heated quartz capillary tube (120 C) connecting the TGA furnace outlet with the MS gas inlet through a pinhole diaphragm.
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189
For isothermal TGA investigations samples of about 10 mg were heated in air atmosphere with a purge gas stream of 100 ml/min in ceramic pans until 200 C and then kept isothermal for hours. Melt Volume Rate (MVR): The MVR is obtained by measuring the amount of molten material flowing through a defined orifice under a defined mass. Comparing the same polymer, melts of low-molar mass polymers flow more rapid than melts of high-molar mass polymers. The MVR experiments were carried out with 2.16 kg load and at 190 C in a ZWICK 4105.01/03 device. Tensile tests: The monotonic loading tests conducted under constant crosshead speed conditions were performed in accordance to ISO 527-2. The crosshead speed for assessing the modulus of elasticity was 1 and 50 mm/min for the determination of the tensile strength and the elongation at break. Nanoidendation: For the determination of mechanical properties of the complex shaped press buttons small pieces were cut and a typical cross-sectional preparation including embedding in epoxy resin, grinding and polishing was carried out. Instrumented microhardness tests were done using a Nanoindenter XP (MTS Systems Inc.). The indentation depth was 1 mm and the indentation rate 0.05 mm/s. After a holding time of 30 s at maximum load the specimens were unloaded. The indentation hardness HIT was calculated at the maximum load by means of the contact depth, according to Ref. [11]. The measurements were done at two different positions on the cross-section, at a distance of 50 mm from the surface and in the core region. Average values were calculated from 10 indents at every position. Molar mass investigations: The method used was gel permeation chromatography (GPC). The investigations were performed at elevated temperature (140 C) with a PL-GPC 220 instrument (Polymer Laboratories), equipped with two separation columns and a differential refractometer. N,N-Dimethylformamide (DMF) was used as eluent at a flow rate of 0.5 ml/min. The calibration was made with polystyrene standards. 3. Results 3.1. Influence of the processing conditions and comparison of different geometries Several processing conditions (e.g. injection velocity and nozzle temperature) were investigated in order to verify the influence of the processing conditions and to find out to what extent degradation takes place during processing. Furthermore, different injection moulding geometries and their influence on the degradation behaviour were compared. Fig. 2a and b compare the results of the copolymers and homopolymers after variation of the processing conditions. It can be shown that even when the temperature of the nozzle is increased the copolymers remain thermally stable (Fig. 2a), whereas the homopolymer (H1) in Fig. 2b shows a shoulder indicating some chains that are affected due to the higher temperature. By increasing the injection moulding velocity the formed flash
2183
Fig. 2. TGA of POM copolymer (a) and POM homopolymer (b), after different processing conditions.
(obviously degraded material) shows a decreased onset temperature and an altered DTG signal corresponding to a total degraded POM (Fig. 2b). TGA investigations in nitrogen atmosphere reveal that there is no influence of the used mould geometry on the degradation behaviour of POM copolymer (Fig. 3). On the other hand, isothermal TGA investigations on pellets and specimens clearly show that the first processing step leads to stabiliser consumption and reduces the thermal stability of all investigated materials extracted from specimens or structural components (Fig. 4). 3.2. Influence of multi-processing and ageing TGA investigations reported in an earlier publication [9] reveal that thermo-oxidative processes decrease the degradation onset of the homo- and copolymer pellets, since the secondary products of the auto-oxidation enhance the auto-oxidation and/ or auto-catalytic cleavage of the CeOeC-bonds. Random scission of the polymer chains and the formation of macro-alkyl
2184
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189
changes in the crystallinity for either material could be found by means of DSC with respect to ageing [13]. The TGA investigations of the heat-stabilised homopolymer (H1) in comparison to additive-extracted material indicate continuous stabiliser consumption and chain damage after thermo-oxidative loading (Fig. 5a). The TG and DTG curves of the oven-stored material are shifted towards the curve of the additive-extracted material; the onset temperature is lowered close to the temperature range in which the thermal degradation of the pure POM-chain starts (according to Kern [1], thermal degradation onset is above 270 C in nitrogen atmosphere). The onset temperature of the H1156 d/140 C is 218 C, the onset temperature of the additivepurified sample is 206 C. Because of the change in the shape of the DTG signal we can conclude that degradation took place. Fig. 3. TG and DTG curves of POM copolymer: influence of the used mould geometry on the degradation behaviour.
radicals, which form alkyl peroxides and hydroperoxides in the presence of even a small amount of oxygen, lead to a subsequent degradation. While the solely heat-stabilised homopolymer (H1) shows some stabiliser consumption after multiple processing, the solely heat-stabilised copolymers (B1 and A1) are more stable against thermo-mechanical degradation. According to the TGA/MS results for the homopolymer, stabiliser consumption occurs during reprocessing and further oven-storage. In the copolymers, the stabiliser is less affected during accelerated ageing due to the presence of co-monomer units; the degradation takes place only in the amorphous phase [12]. The homopolymers and the copolymers differ in crystallinity because of their molecular structure, but no significant
Fig. 4. Isothermal TGA on pellets and specimens in air atmosphere, performed at 200 C.
Fig. 5. TGA of the homopolymer after thermo-oxidative loading in comparison to the pellet and the additive-extracted pellet (a); MS-spectra of the homopolymer after thermo-oxidative loading in comparison to the initial state (b) (m/z: 30eformaldehyde; 31emean peak trioxane, methoxy fragment; 44ecarbon dioxide, ethylene oxide, urea fragment; 57eadditive fragment; 60eacetic acid, methyl formate, urea; 77ephenyl fragment).
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189
Fig. 6. TG curves of two homopolymers with different stabiliser packages in initial state and after thermo-oxidative loading.
The MS-spectra show that the ion current response for antioxidant fragments (m/z ¼ 57 (t-butyl group) and 77 (phenyl group)) is relatively poor after thermo-oxidative loading in comparison to the initial state (Fig. 5b). Total stabiliser
2185
consumption had happened during an oven-storage of 56 d at 140 C. In Fig. 6 the TG curves of two homopolymers (H1 and H2) with different stabiliser packages are illustrated, in initial state and after oven-storage. It is clearly shown that already in the initial state the additional UV-stabilised homopolymer H2 is thermally superior to the heat-stabilised H1. After thermo-oxidative loading, although homopolymer H2 is also subjected to auto-oxidation and degradation due to the secondary products of the auto-oxidation, the thermal stability of H2 remains superior. One explanation for the higher thermal stability of the homopolymer H2 is that it may contain reagents such as urea, polyamides or polyurethane, which prevent splitting by secondary products of the auto-oxidation as is referred to in Ref. [1]. However, these reagents could not be verified with the help of mass spectrometry. A comparison of two different heat- and UV-stabilised copolymers (left/right) can be seen in Fig. 7. The DTG curves (top) show a different degradation performance due to thermo-oxidative loading. The molar mass distributions in initial state and after thermo-oxidative loading (bottom) show that copolymer 2 is clearly more affected than copolymer 1 (due to business rivalry no further identity of copolymers 1
Fig. 7. TGA of two copolymers after thermo-oxidative loading in comparison to the specimen, pellet and the additive-extracted pellet (top); copolymer 1 (a) and copolymer 2 (b); molar mass distributions in initial state and after 56 d/140 C (bottom); copolymer 1 (c) and copolymer 2 (d).
2186
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189
and 2 is available). This implies that different commercial products with comparable stabiliser packages for similar applications show different thermal stability after thermo-oxidative loading. 3.3. UV-irradiation The irradiation exposed surfaces of homo- and copolymers (H1, A1 and B1) without any UV-stabiliser after 1 year artificial exposure (even indoor exposure dose) are totally chalked and cracks have developed. According to Gardette et al. [14] the following degradation mechanism for photo-oxidation takes place in POM: the first step leads to the oxidation of the carbon atoms with the formation of secondary hydroperoxides that decomposes into two radicals: an alkoxy-macroradical and a hydroxy radical. The alkoxy-macroradical can react in two possible ways: a cage reaction may occur leading to the formation of carbonate and water, and a b-scission which leads to a formate and an alkoxy-macroradical. Hydrogen abstraction from the polymeric chain by macroradicals leads to the formation of alcohols. Thereby a new macroradical is created. In contrast, the formation of hydroperoxides leads to chain scission via b-scission reaction and formaldehyde liberation [1]. The results of the TGA and GPC investigations indicate that similar degradation mechanisms occur in the case of thermo-oxidative loading and UV-irradiation. Fig. 8 illustrates the TGA and GPC traces of the heat-stabilised homopolymer, similar results were obtained for the heat-stabilised copolymers. Fig. 8a presents the DTG results in the initial state compared with the conditions after thermo-oxidative loading and UV-irradiation, the shift of the curves towards lower temperature indicates degradation. Fig. 8b demonstrates the corresponding GPC results. Because of the very narrow molar mass distribution (MMD) in the initial condition, only a shift of the MMD and no change in the MMD shape is expected. The UV-irradiated samples of all heat- and UV-stabilised polymers show no visible changes on the surface, TGA investigations of the irradiated samples indicated no degradation. All materials were stabilised properly against photo-oxidation. 3.4. MVR and mechanical properties Oven-storage leads to an increase in the MVR of homopolymers and copolymers, all UV-stabilised materials show a higher MVR increase than the solely heat-stabilised materials, i.e. the melt viscosity is reduced, although the oxidative stability of both systems is comparable in the initial state. This reduction can be explained as a change in the MMD. The used UV-stabiliser evidently decreases the thermo-oxidative stability of the material during ageing in the oven at 140 C. Similar results were reported in Ref. [15], where in contrast to the solely heat-stabilised materials all heat- and UV-stabilised grilles show a more intense decrease in the deflection at break after ageing of 35 d/140 C. Fig. 9 illustrates the changes in the MVR as a funtion of the storage duration. Higher values of the MVR after storage
Fig. 8. Degradation behaviour of heat-stabilised homopolymer after thermooxidative loading and UV-irradiation. DTG curves of the homopolymer (seventh processing step) in initial condition, after thermo-oxidative loading (21 d/ 140 C) and 1a UV-irradiation (car indoor irradiation, Central Europe) (a); GPC curves of the homopolymer (seventh processing step) in initial condition, after thermo-oxidative loading (21 d/140 C) and 1a UV-irradiation (car indoor irradiation, Central Europe) (b).
duration of 56 d at 140 C are measured for the copolymer 2, which also shows the major MMD changes. Whereas for all investigated materials the elongation at break decreases due to reprocessing [13], the modulus of elasticity is less sensitive to changes in the chemical and/or physical structure. The modulus of elasticity of the homopolymers and copolymers remains stable even though the number of processing steps is increased. It seems, that the interaction between amorphous and crystalline phase has changed, for example a variation of the nature and number of the tie-molecules. Another explanation for the decrease of the elongation at break is an increase of voids due to shrinkage effects of the polymer during cooling from the mould. Especially in thick parts like tensile test specimens and because of suboptimal
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189
2187
Fig. 9. Changes in the MVR and elongation at break in dependence of the thermo-oxidative loading duration, results illustrated for the copolymer B2. 1st PS: first processing step, 4th PS: fourth processing step, 7th PS: seventh processing step.
conditions used for processing such shrinkage effects are not unexpected. Polarised optical microscopy of thin section cuts through the samples reveals that the crystalline super molecular structure is almost not altered; i.e. the MMD should be unchanged after reprocessing. TGA investigations have already indicated stabiliser consumption, which cannot be monitored by GPC. However, the tensile test is not a sensitive method to detect embrittlement since the embrittled layer makes up only a small proportion of the overall sample cross-section. The modulus of elasticity from the tensile test is an integral property for injection moulded specimens, which have a heterogeneous structure from edge to core depending on the cooling conditions. Measurements of the modulus of elasticity and hardness over the specimen cross-section can be made with instrumented nanoindentation tests and the dependence of the modulus of elasticity and hardness with the crystalline structure can be confirmed. At the edge, the area that has been cooled with the fastest cooling rate and less crystalline structure is visible; the modulus of elasticity is low. In the shear zone, an area with a lower cooling rate and a fine spherulitic structure is present; the modulus of elasticity is higher than at the edge. In the core of the specimen, in which ‘‘constant’’ crystallisation conditions dominate, cooling and heat of crystallisation are interactive; the modulus of elasticity is high [13,16]. The influence of thermo-oxidative ageing at elevated temperature and UV-irradiation on the hardness of the press button is illustrated in Fig. 10. An increase in hardness is provoked by degradation, leading to decreased inter-crystalline interactions. Annealing effects could not be proved. Results from the nanoindentation testing showed increased hardness and modulus in thermo-oxidative loaded regions for the homopolymer (Fig. 10a). While oven-storage of 56 d leads
Fig. 10. Hardness determined by instrumented nanoindentation measurements in initial state and after thermo-oxidative loading and UV-irradiation for the homopolymer (a) and copolymer (b).
to an increase of the modulus of elasticity and hardness caused by embrittlement of the homopolymers (Fig. 10a), the hardness of the copolymers shows less change (Fig. 10b). The data also indicate that the mechanical properties and level of thermo-oxidation are strongly correlated. Polarised optical microscopy of thin sections cut through the press buttons reveals that the crystalline structure is different for the investigated materials. While for the homopolymers, due to the higher melt viscosity and the resulting higher shear rates, shear induced crystallisation effects cannot be excluded, for the copolymers a typical skin-score morphology e as described above e can be deduced. This explains the higher values of hardness for the homopolymers and the unusual higher values of the hardness (and indentation modulus) for the skin region compared to the core for these materials (Fig. 10a). In Fig. 10 the influence of UV-irradiation on the hardness measured in the initial state and after 1a UV and 2a UV is plotted.
2188
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189
As discussed in the previous section, for the solely heatstabilised homopolymers repeated processing leads to stabiliser consumption after six reprocessing steps, while for the copolymers (in any stabilisation state) no stabiliser consumption after repeated processing could be found.
4. Conclusions
Fig. 11. Influence of recycling on the degradation behaviour of POM copolymer determined by TGA; reprocessing of copolymer grills, 1st PS: first processing step, 3rd PS: third processing step (a) and addition of 20% recycling material (b).
UV-irradiation causes no changes in the mechanical properties in comparison to the initial state in both of the heat- and UV-stabilised materials. 3.5. Influence of addition of recycled material The most common form of recycling is done directly in the factory, where waste such as sprues and runners is regranulated and used to produce new products of the same kind. Normally the fraction of reground material mixed with virgin material is small to avoid sensitivity to molecular degradation. The influence of recycling on the degradation behaviour of POM-copolymer components can be seen in Fig. 11. Injection moulded grilles after three processing steps (Fig. 11a) and ones with an addition of a mass fraction of 20% recycled material show no altered degradation behaviour (Fig. 11b).
Differently stabilised POM homopolymers and copolymers were investigated to monitor the degradation initiated by processing, artificial ageing and recycling. The TGA method is well known as a good approach to prove stabiliser consumption and induced degradation. The TGA results indicate random scission of the main chain as the initiation mechanism of degradation of POM. Depression of the onset temperature of the TG curves is attributable to continuous stabiliser consumption. Changes in the DTG signal compared to the initial state are directly related to chain damages. In dependence of the applied stabiliser system homopolymers show stabiliser consumption and chain damages due to multiple processing. Thermo-oxidative ageing leads to further stabiliser consumption followed by chain damage and degradation. In spite of multiple processing all investigated copolymers are thermally stable. After 21 d/140 C, degradation can be detected for the homopolymers. Due to the presence of the co-monomer units copolymers are more stable, and therefore chain damage could be proved only after a duration of 56 d/140 C, obviously oven-storage affects the homopolymers more than the copolymers. After 1 year irradiation (car indoor, Central Europe), the solely heat-stabilised materials present a chalked surface and cracks. The results of the TGA and GPC investigations indicate that similar degradation mechanisms occur in case of thermo-oxidative loading and UV-irradiation. In the both heat- and UV-stabilised materials (homo- and copolymers) UV-irradiation does not affect the mechanical properties. Thermo- and photo-oxidation of POM proceed according to a two-fold mechanism: (a) decomposition of the primary peroxides formed by mechanical-oxidative degradation induced by processing and (b) oxidation of radicals as discussed in the Bolland mechanism of auto-oxidation of polymers. In thermo-oxidation b-scission of the alkoxy radicals is the dominating reaction whereas in photo-oxidation hydroperoxide decomposes through a cage reaction or to an alkoxy radical and a hydroxy radical. Alternatively an alkoxy radical can form an aldehyde and a new radical by b-scission. The investigations proved that the progress of the degradation of POM is independent of the used mould geometries, in a similar manner. Stabiliser consumption takes place, primarily during the first processing step from pellet to part, no differences in the extent could be proved by differently investigated mould geometries. It is demonstrated that POM copolymers tolerate higher fluctuations of the processing parameters before degradation occurs than POM homopolymers. On the other hand
V.-M. Archodoulaki et al. / Polymer Degradation and Stability 92 (2007) 2181e2189
homopolymers show enhanced mechanical properties compared to the copolymers, even after artificial ageing. The impact of chain damage on the mechanical properties is different for the analysed materials: thermo-oxidative loading of the heat- and UV-stabilised materials leads to higher embrittlement and higher-MVR values. In contrast to the solely heat-stabilised material the thermo-oxidative induced degradation progresses faster, due to the added stabiliser. This effect has been found in the homopolymers as well as copolymers. Reprocessing leads to a decrease of the elongation at break in all materials. The explanation for this behaviour is that the nature as well as the number of tie-molecules are altered.
References [1] Kern W. Chemische Elementarvorga¨nge bei der Alterung von Kunststoffen. Chemiker-Ztg/Chem Apparatur 1967;91:255e62. [2] Kern W, Cherdron H. Der Abbau von Polyoxymethylenen. Makromol Chem 1960;40:101e17. [3] Hermann HD, Weissermel K. Die Copolymerisation von Formaldehyd mit cyclischen Formalen. Makromol Chem 1966;94:91e102. [4] Weißermel K, Fischer E, Gutweiler K, Hermann HD. Zur Copolymerisation des Trioxans. Kunststoffe 1964;54:410e5. [5] Inoue M. Crystallization and melting of copolymers of polyoxymethylene. J Appl Polym Sci 1964;8:2225e38.
2189
[6] Ivanova LL, Grebennikova NA, Sokolova GP, Dudina LA, Irzhak VI, Yenikolopyan NS. Heat stability of a formaldehyde-1,3-dioxolane copolymer. Vysokomol Soyed 1976;A18(4):920e5. [7] Zweifel H, editor. Plastic additives handbook. 5th ed. Mu¨nchen: Hanser Verlag; 2000. [8] Sabel HD, Schlaf H, Unger P, Ziegler U. Polyacetale. In: Bottebruch L, editor. Technische Thermoplaste: Polycarbonate, Polyacetale, Polyester, Cellulosester. Mu¨nchen: Carl Hanser Verlag; 1992. p. 300e95. [9] Archodoulaki VM, Lu¨ftl S, Seidler S. Thermal degradation behaviour of poly-(oxymethylene): 1. Degradation and stabiliser consumption. Polym Degrad Stab 2004;86:75e83. [10] Lu¨ftl S, Archodoulaki VM, Seidler S. Thermal-oxidative induced degradation behaviour of poly(oxymethylene) copolymer detected by TGA/MS. Polym Degrad Stab 2006;91:464e71. [11] Oliver WC, Pharr GM. An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. J Mater Res 1992;7/6:1564e83. [12] Mucha M. Thermooxidation rate of diacetate terminated polyoxymethylene with various morphological structures. Colloid Polym Sci 1984;262: 841e50. [13] Archodoulaki VM. Eigenschaftsa¨nderungen von Polyoxymethylenen induziert durch Verarbeitung, Alterung und Recycling. Fortschritt-Berichte VDI, Reihe 5, Nr. 705. Du¨sseldorf: VDI-Verlag; 2005. [14] Gardette JL, Sabel HD, Lemaire J. Photooxidation of polyacetal copolymer I. Angew Makromol Chem 1991;188:113e28. [15] Archodoulaki VM, Lu¨ftl S, Seidler S. Effects of coloration on the physical and mechanical properties and on the thermal stability of polyoxymethylene. Adv Polym Tech 2006;25(1):63e72. [16] Wolters E, Buchberger HL, Kohlhepp KG. Zusammenhang zwischen Gefu¨geaufbau und mechanischem Verhalten von Spritzgussteilen aus Hostaform. Plastverarbeiter 1973;9(24):549e55.