Crosslinking of Polyolefin Foam II. Applicability of Parameters to Assess Crosslinking/Foam Density Relationships

Transcription

1 Crosslinking of Polyolefi n Foam II. Applicability of Parameters to Assess Crosslinking/Foam Density Crosslinking of Polyolefin Foam II. Applicability of Parameters to Assess Crosslinking/Foam Density C.S. Sipaut a, G.L.A. Sims b and Z.M. Ariff c a School of Chemical Sciences, Universiti Sains Malaysia, 11800, Penang, Malaysia b Victoria University of Manchester-legacy, The University of Manchester, Grosvenor Street, Manchester M1 7HS, UK c School of Material and Mineral Resources, Engineering Campus Universiti Sains Malaysia, Nibong Tebal, Penang, Malaysia Received: 21 May 2007, Accepted: 7 January 2008 ABSTRACT The effect of introducing triallyl cyanurate (TAC) monomer, into dicumyl peroxide (DCP) crosslinking systems for low-density polyethylene (LDPE) is considered. In the foam formation, chemical blowing used in a fi xed amount of 8.0 phr. Effects are characterised as a function of relative concentrations in solid, melt and foamed states. It was observed that gel content could only be used as a reasonable indicator to predict foaming behaviour only for traditional crosslinking with DCP alone. The results also showed that melt modulus seemed a better indicator when TAC was incorporated in the system but appeared only to be particularly relevant at a specifi c TAC concentration. It was found that swell ratio better controls expansion prediction whereas foam density determines physical and mechanical properties independent of formulation. Moreover, swell ratio appeared to be able to define expansion characteristics not only of traditional crosslinking systems but also those containing triallyl cyanurate monomer. INTRODUCTION In the manufacturing of crosslinked polyolefin foam, the foam expansion and cell stabilization are highly dependent on the degree of crosslinking. It has been reported that at higher crosslinking level the foams produced have higher density (1). Foams properties are usually depend on their density but other factors such as base polymer (2), cell size (3), cell shape (3-4), processing techniques, amount and types of crosslinking agent and blowing agent (3,4,6) also influence Smithers Rapra, 2008 Cellular Polymers, Vol. 27, No. 1,

2 C.S. Sipaut, G.L.A. Sims and Z.M. Ariff foam properties. However, the design of molecular interaction to generate crosslinking in the polymer is the most important structural modification. These consequently provide different degrees of crosslinking and if used in foam production will give different properties (7). Therefore the degree of crosslinking formed in the polymeric system must be predicted. The methods of assessing the crosslinking level in the polymer depend on the type of polymer. Gel content determination is the popular method used in measuring crosslinking levels of polyethylene (8). In this method, the insoluble fraction in crosslinked polyethylene was measured after refluxing with a specific solvent. Mahapatro et al (9) correlation between melt modulus and foam expansion is based on the measurement of melt modulus at a temperature similar to the final foam expansion temperature. The melt modulus was calculated from slope of stress-extension ratio curve then used to predict the foam density. The limitation of this method is that during the test, the tensile testing temperature must be similar to the final foam expansion temperature in the foaming process. The sample softens and expands during the placement in a hot chamber causing the distance between the grips of the tensile machine to be unstable. An alternative approach for characterising the degree of crosslinking is infrared (IR) spectral analysis. This was used for a crosslinked copolymer ethylene and vinyltrimethoxy silane (10), ethylene vinylsilane (11), polypropylene (12) and ethyleneoctene in polyethylene (13). Changes in the peak area or intensity of a specific functional group during the reaction were measured. This IR method is unsuitable for determining crosslinking in low density polyethylene (LDPE) which does not contain specific peaks indicating crosslinking of the LDPE molecule. Part I was the paper by Sims and Sipaut (14). The effect of introducing triallyl cyanurate TAC monomer in a low density polyethylene-dicumyl peroxide LDPE-DCP system for foam formation was investigated. It is reported that comparing results of solid and foamed samples of identical gel content, formulation containing of TAC produce higher foam density compared to formulation with DCP alone. It also found that the gel content measurement to determine foam density is not an appropriated parameter or method if TAC is introduced in the formulation. Therefore the study was extended to investigate appropriate parameters or method that suitable to predict and correlate foam behaviour. In this study the effect of introducing TAC on crosslinking LDPE-DCP will be investigated as a function of gel content, melt modulus and swell ratio. 12 Cellular Polymers, Vol. 27, No. 1, 2008

3 Crosslinking of Polyolefi n Foam II. Applicability of Parameters to Assess Crosslinking/Foam Density EXPERIMENTAL 1. Raw Materials Base polymer Base polymer used was Low density polyethylene (LDPE) Stamylan 1808 supplied by DSM with a density of 918 kg m -3 and melt flow index of 8 dg min -1. Crosslinking agent Dicumyl peroxide (DCP) was selected as crosslinking agent. It has a half-life of 10 hours at 115 C and 1 minute at 171 C (14), containing 40% DCP on 60% inert clay carrier manufactured by Akzo Nobel Chemicals. Formulations were investigated using a range of DCP concentrations up to 1.0 phr (parts per hundred resins). Polyfunctional monomer Polyfunctional monomers used was triallyl cyanurate (TAC) supplied by Degussa. Concentrations in the range of to 5.0 phr were used. Blowing agent The chemical blowing agent used was DP45/1, manufactured by Bayer AG. It is based on activated azodicarbonamide (ADC) with a decomposition temperature of 165 C. All foam formulations used a fixed concentration of 8.0 phr. 2. Sample Preparation Compounding Mixing was conducted on a thermostatically controlled, electrically heated tworoll mill. The roll temperatures were initially set at 115 C then adjusted to 105 C after the LDPE full melted. The mixing and folding of the melted LDPE on the two-roll mills was timed at approximately 5 minutes. Immediately following the banding step, the polyfunctional monomer was added (where applicable) cutting and folding over a period of 5 minutes to obtain a homogeneous mix, followed by the blowing agent (where applicable) which was added progressively over a period of 10 minutes. The final addition was the peroxide crosslinking agent over a period of 3 minutes with continuous cutting and folding. Peroxide was added last to minimise the chance of premature dissociation and crosslinking. The compounded sample was allowed to cool to room temperature and stored Cellular Polymers, Vol. 27, No. 1,

4 C.S. Sipaut, G.L.A. Sims and Z.M. Ariff in a controlled environment at 23 ± 2 C and 50 ± 5% relative humidity for a period of 24 hours before further processing. Solid crosslinked matrix Crosslinked matrix samples were prepared by compression moulded of 45 g of compounded samples in a square plate with a thickness of 1.5 mm for 20 minutes. The mould was then water-cooled to 30 C under pressure. The crosslinked samples then stored at 23 ± 2 C and 50 ± 5% relative humidity for a period of 24 hours before testing. Foam production Foaming was carried out by a two-stage heat and chill process using pre-heated 10 mm deep mould. After compounding, the foamable samples were compress moulded at 14 MPa for 20 minutes at 165 C and cooled under pressure to 30 C. The partially expanded moulding was then immediately transferred to a circulating hot air oven at 130 C for 20 minutes to gradually complete the expansion. 3. CHARACTERISATION 3.1 Assessment of crosslinking Gel content Gel content was determined by refluxing in a stainless-steel mesh cage in boiling xylene for 24 hours and expressing the weight of the vacuum-dried insoluble fraction as a percentage of sample weight before extraction (3-6,8). Melt tensile measurements To conveniently gain an assessment of melt strength of the polymer during foam expansion, tensile testing of the crosslinked matrix was carried out at 130 C (the expansion temperature in the final stage of the heat and chill process). Dumbbell tensile specimens were placed in the jaws of an Instron tensile testing machine fitted with an environmental chamber set at 130 C. The samples were conditioned at the test temperature for 10 minutes before begins the test. The gauge length was set at 30 mm and the crosshead speed at 20 mm min -1. Results allowed assessment of melt modulus and are quoted as the average of five tests. The melt modulus in this work and Mahapatro work is calculated from the tangent of the initial slope of stress-strain curve and stress-extension ratio curve respectively. 14 Cellular Polymers, Vol. 27, No. 1, 2008

5 Crosslinking of Polyolefi n Foam II. Applicability of Parameters to Assess Crosslinking/Foam Density Swell ratio Crosslink density can be estimate from equilibrium swelling measurements (18,19). The swell ratio decreases as the crosslink density increases (8,20). The swell ratio is the ratio of the gel volume in the swollen state to its volume in the unswollen state. The swell ratio of crosslinked samples was determined in p-xylene at 130 C in accordance with ASTM D (8). The average of three determinations had an experimental error of ± 2%. 3.2 Foam Characterisation Foam density Foam density was determined from the weight and volume of regular parallelepiped samples (50 x 50 x 20 mm) free of skin, voids or other irregularities (3,5,14). Cell imaging and cell size determination Cellular structures of foam specimens were assessed from Scanning Electron Microscope (SEM) images. The mean apparent cell size was obtained from SEM images by a modified cell count per unit length method described elsewhere (3,5). Foam compression properties Parallelepiped specimens (free of defects and skins) 50 mm x 50 mm with an average thickness of 20 mm were cut from bulk foamed samples. Compression measurements were performed on an Instron Universal Testing Machine fitted with a compression cage at a crosshead speed of 20 mm min -1 and compressing the samples to 80% of their original thickness. Elastic compression modulus was calculated from the initial linear portion of the stress-strain curve and the compressive stress at 50% strain was recorded. Results were reported as the mean of ten measurements. RESULTS AND DISCUSSION Part I of this paper (14), reported that the gel content parameter could only be used to predict the foam density for crosslinking with DCP alone, and not for with the addition of triallylcyanurate monomer. The effect of melt modulus parameter on foam density however has not been discussed in detail. From the results, it only postulated that at similar gel content, the introduction of TAC in the crosslinking system gave different melt modulus. Cellular Polymers, Vol. 27, No. 1,

6 C.S. Sipaut, G.L.A. Sims and Z.M. Ariff The Effect of Traditional Crosslinking on Melt Modulus and Gel Content The dependence of foam density on gel content and melt modulus of typical commercial foaming systems (i.e. LDPE-DCP system with varying DCP concentration) is shown in Figure 1. The results showed that whilst foam density appeared to be more directly related to gel content than melt modulus, the trends were very similar. The theory propounded by Mahapatro et al. (9) suggested that foam density is directly related to melt modulus at equilibrium expansion. Whilst the melt modulus in this work is calculated from the initial slope of the melt stress-strain curves, Mahapatro et al. (9) theory actually defines an alternative more complex melt modulus based on extension ratio (λ). They calculated melt modulus from the slope of the melt stress-strain curve above a strain of 2 to 3 where the strain is a function of λ 2 - λ -1. It was found however, that the two alternative melt moduli gave identical trends in foam density as shown in Figure 2. This suggested that the theory propounded by Mahapatro et al could be used to predict foam behaviour for this particular system. As both moduli produced identical trends, the melt modulus quoted in all further results is the more familiar Young s modulus of the melt. Other workers (15) suggest that, above the critical gel content for stable foam formation, cell size reduces with increasing gel content. The cell size reduction with increasing gel content was considered to be directly related to the higher Figure 1. Effect of gel content and melt modulus on foam density 16 Cellular Polymers, Vol. 27, No. 1, 2008

7 Crosslinking of Polyolefi n Foam II. Applicability of Parameters to Assess Crosslinking/Foam Density Figure 2. Effect of foam density on melt modulus (determined from initial slope and Mahapatro theory) of LDPE crosslinked by DCP alone Figure 3. Mean cell size dependence on gel content and melt modulus for traditional crosslinking system degree of crosslinking and consequent increased melt modulus, which then restricted expansion as show in Figure 3. These results suggested that cell size is also directly related to gel content and melt modulus which then gives Cellular Polymers, Vol. 27, No. 1,

8 C.S. Sipaut, G.L.A. Sims and Z.M. Ariff further credence to the use of the theory of Mahapatro et al. (9). Thus cell size appears to be restricted as melt modulus increases. This could be the result either of resistance to expansion restricting cell growth or an increase in cell nucleation suggested by Colton and Suh (16). The Effect of Introducing TAC in the System on Gel Content and Melt Modulus Figure 4 shows that the increase in foam density for systems containing 0.5 phr TAC was related to the increase in gel content and melt modulus. These results are not surprising and similar to those observed with DCP alone (Figure 1). Comparing the general trends shown in Figure 1 (DCP alone) and Figure 4 (DCP/0.5 phr TAC), a more linear dependence of foam density on melt modulus was observed with the DCP/TAC system but a hint of a step change in melt modulus appeared to be indicated when using DCP alone. This appears to occur over the gel content range 60 to 70% and melt Young s moduli from to MPa (Figure 2). However, the results tend to follow the general trend predicted by the theory of Mahapatro et al., for samples crosslinked by varying DCP concentration in the presence of 0.5 phr TAC or by DCP alone. To assess the phenomenon further, the foam density range of formulations with gel contents of approximately 62%, 70% and 77.5% is shown in Figure 5. Figure 4. Dependence of foam density on gel content and melt modulus for samples containing varying concentrations of DCP and a TAC concentration fixed at 0.5 phr 18 Cellular Polymers, Vol. 27, No. 1, 2008

9 Crosslinking of Polyolefi n Foam II. Applicability of Parameters to Assess Crosslinking/Foam Density Figure 5. The effect of formulation on foam density at similar gel content Generally, within the limits of experimental error, at any specific gel content, the higher the TAC concentration, the higher the foam density. The trend line slopes in Figure 5 increase with the TAC concentration. Thus gel content is not the only parameter that determines foam density when TAC is introduced into the system. The foam density increases with increasing TAC which tends to be supported by general findings of Part I of this paper where increasing TAC at any specific DCP concentration seemed to give higher crosslink density. This in turn would explain the increases in melt modulus (restricting cell expansion and limiting foam density reduction) with increasing TAC concentration as indicated in Figure 6. Figure 7 gives a wealth of information. It is important to understand that each series of three separate results at each nominal gel content represents an adjustment in DCP concentration to give the nominal gel content in the presence of 0, 0.5 and 2.0 phr TAC (increasing TAC concentration produces progressive increase in melt modulus and foam density). Figure 7 also positively indicates that gel content is not a suitable indicator to predict foam density (as previously suggested). However, if all data points could be considered as part of a mastercurve then the Mahapatro theory appears to hold some validity. However, when experimental errors are taken into account, three separate curves are strongly suggested. It appears that the Mahapatro theory may be valid in Cellular Polymers, Vol. 27, No. 1,

10 C.S. Sipaut, G.L.A. Sims and Z.M. Ariff Figure 6. The effect of formulation on melt modulus at similar gel contents Figure 7. The dependence of foam density on melt modulus at specific gel contents (DCP,TAC ratios [phr] shown in parentheses) the presence of TAC but only at specific gel content. Overall, however, the incorporation of TAC in the crosslinking system seems to invalidate general application of the Mahapatro theory. It is considered that crosslink density may contribute to this phenomenon. 20 Cellular Polymers, Vol. 27, No. 1, 2008

11 Crosslinking of Polyolefi n Foam II. Applicability of Parameters to Assess Crosslinking/Foam Density This behaviour again supported earlier findings that as the TAC concentration increased in the formulation, crosslink density appeared to increase, resistance to expansion (modulus) increased, melt extensibility decreased and this was the prime reason for the observed foam density increase. This also suggested that the increase in melt modulus (expansion resistance) may promote cell nucleation (16,18). Generally, within the limits of experimental error, the results in Table 1 show that an increase in foam density was accompanied by a significant increase in compression modulus and reduced mean cell size. Table 1. Effect of formulation at similar foam densities on melt modulus, gel content and foam properties DCP, phr TAC, phr Gel content, % Density, kg m -3 Compression modulus, MPa Cell size, mm Melt modulus, MPa None ± ± ± ± ± ± ± ± None ± 1.52 ± ± ± ± ± ± ± ± ± ± ± None ± ± ± ± ± ± ± ± ± ± ± ± The Effect of Traditional Crosslinking on Swell Ratio Figure 8 shows that the effect of increasing DCP concentration was as expected, to decrease the swell ratio. A higher crosslink density in the network, reduces the ability of the sample to swell. This applied for any specific TAC concentration. The large drop in swell ratio caused by the addition of TAC at a particular DCP concentration, confirmed the crosslinking promotion effect of TAC. A 0.5 phr TAC concentration requires a minimum DCP concentration of 0.15 phr to become particularly effective; above this concentration, the swell ratios are of similar order, confirming previous gel content findings (14). Cellular Polymers, Vol. 27, No. 1,

12 C.S. Sipaut, G.L.A. Sims and Z.M. Ariff Figure 8. Effect of crosslinking systems on swell ratio Figure 9. The effect of TAC concentration on swell ratio at a fixed DCP concentration of 0.25 phr Moreover, the curves of 2.0 and 5.0 phr TAC, have a more linear relationship compared with those crosslinked with DCP alone or in the presence of 0.5 phr TAC. This strongly suggested that at higher TAC concentrations, a more rigid network was present which restricted the ability of samples to swell. Figure 9 appears to indicate that there are only two regimes in swell ratio measurement 22 Cellular Polymers, Vol. 27, No. 1, 2008

13 Crosslinking of Polyolefi n Foam II. Applicability of Parameters to Assess Crosslinking/Foam Density (whereas gel content results had suggested three general regimes) as the concentration of TAC increases. The first regime, similar to the gel content results, occurs up to approximately 0.5 phr TAC, where swell ratio decreased sharply suggesting efficient use of the crosslinking promoter. On increasing the TAC concentration further, swell ratio tended to decrease at a much reduced rate. This behaviour suggested that gel content (network fraction) tends to an asymptotic value with increasing TAC concentration whereas a progressive decrease in swell ratio indicates an increasingly more crosslinked structure. This explains why gel content is not a suitable parameter upon which to base foam density predictions and implies that swell ratio could be a more suitable parameter when the TAC is introduced into the system. Effect of Swell Ratio on Foaming Behaviour The relationship between swell ratio and foam properties was initially established for LDPE crosslinked by DCP alone. Figure 10 shows that an increase in swell ratio is accompanied by a decrease in foam density and an increase in mean cell size. A higher swell ratio is associated with a more loosely crosslinked network (i.e. lower crosslink density) that has a greater ability to expand and hence allows a lower foam density. When the foam density dependence on swell ratio of various crosslinking systems is compared (Figure 11). The data points fall closer to a mastercurve Figure 10. The effect of swell ratio on foam density and mean cell size for DCP crosslinked LDPE foam Cellular Polymers, Vol. 27, No. 1,

14 C.S. Sipaut, G.L.A. Sims and Z.M. Ariff Figure 11. The dependence of foam density on swell ratio as a function of crosslinking system formulation than the melt modulus results (Figure 7). Hence the swell ratio might be used to predict foam density. Therefore an attempt was made to investigate similar swell ratios of approximately 39, 22 and 17. Table 2 indicates that formulations of similar swell ratio give approximately similar foam density, compression modulus and mean cell size. This suggested that all selected formulations of similar swell ratio had similar expansion characteristics and presumably, similar matrix properties. A further consideration is that swell ratio is based on volume whereas melt modulus measurements consider only uniaxial stress which may explain why swell ratio has a more universal applicability. CONCLUSIONS When foaming traditional DCP/LDPE systems, the foam density appeared to be controlled by gel content and/or the melt modulus (above 50% gel fraction). It was concluded that either gel content or melt modulus was inadequate to define foaming behaviour when TAC is included in the crosslinking system. The swell ratio (crosslink density related) was a more universal parameter for predicting foam expansion than either gel content or melt modulus. 24 Cellular Polymers, Vol. 27, No. 1, 2008

15 Crosslinking of Polyolefi n Foam II. Applicability of Parameters to Assess Crosslinking/Foam Density Table 2. Effect of formulation on foam properties at similar swell ratios DCP, phr TAC, phr Swell ratio Foam density, kg m -3 Compression modulus, MPa Average Cell Size, mm ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ACKNOWLEDGEMENTS The authors wish to thank to Universiti Sains Malaysia, Penang, Malaysia for the financial support and also The University of Manchester. The authors wish to thank to Akzo Nobel Chemicals for the supply of DCP, Bayer for DP45/1, Degussa for TAC and DSM for polymer supply. REFERENCES 1. Klempner D. and Frisch K.C., Handbook of Polymeric Foams and Foam Technology, Hanser, Puri R.R. and Collington K.T., Cell. Polym., 7, (1988), Sims G.L.A. and Khunniteekool C., Cell. Polym., 13, (1994), Sims G.L.A. and Khunniteekool C., Cell. Polym., 15, (1996), Lewis C., Rodlum Y., Misaen B., Changchum S. and Sims G.L.A., Cell. Polym., 22, (2003), Sims G.L.A. and Khunniteekool C., Cell. Polym., 15, (1996), Beveridge C. and Sabston A., Material and Design, 8, (1987), ASTM D , Standard Test Methods for Determination of Gel content and Swell Ratio of Crosslinked Ethylene Plastics, ASTM, Philadelphia, Cellular Polymers, Vol. 27, No. 1,

16 C.S. Sipaut, G.L.A. Sims and Z.M. Ariff 9. Mahapatro A., Mills N.J. and Sims G.L.A., Cell. Polym., 19, (1998), Hjertberg T., Palmlof M. and Sultan B.A., J. Appl. Polym. Sci., 42, (1991), ,. 11. Palmlof M., Hjertberg T., and Sultan B.A., J. Appl. Polym. Sci., 42, (1991), Sirisinha K. and Kawko K., Macromol. Mater., 290, (2005), Sirisinha K. and Chimdist. S., Polym. Testing, 25, (2006), ,. 14. Sims G.L.A. and Sipaut C.S., Cell. Polym., 20, (2001), Sims G.L.A. and Khunniteekool C., Cell. Polym. III, RAPRA Technology, Paper 13, Coventry, U.K., Colton J.S. and Suh N.P., Polym. Eng. Sci., 27, (1987), Colton J.S. and Suh N.P., Polym. Eng. Sci., 27, (1987), Tai H-J., Polymer, 42, (2001), Gunton C.I.S., The Preparation and Characterisation of Halobutyl Rubber Networks, PhD Thesis, University of Manchester Baba M., Gardette J-L. and Lacoste J., Polym. Degrad. Stab., 63, (1999), Cellular Polymers, Vol. 27, No. 1, 2008