All-polymer composites from recycled woven polypropylene fabrics and polyethylene film.

INTRODUCTION

The use of woven fabrics of oriented polypropylene tape has increased considerably in recent years, for applications such as sacks for the bulk transport of agricultural and constructional materials, packaging tape, and ground stabilization mesh. As a consequence, more of this material is now finding its way into the waste stream, where it poses certain challenges for economic and environmentally beneficial disposal. It has been estimated that there are about 10,000 tonnes of woven polypropylene sacks disposed of from the agricultural stream in the UK annually [1], and probably double this from the construction sector, now that safety legislation has precluded the use of reuse through a deposit system [2]. In the UK, the majority of this material is currently disposed off via landfill.

Although it is possible to include waste polypropylene from these sources into low-grade applications, such as wood replacement products [3-5], there are some problems in this, associated with the material's higher melting point compared with polyethylene, which makes up the bulk of the recyclate, and the difficulties in shredding the woven material. In addition, conventional recycling involving melting of the polypropylene will remove the orientation of the polymer tapes, thereby losing the potentially higher stiffness and strength of these tapes. One possible alternative to this, which is investigated in this paper, is to explore the potential for producing an all-polymer composite, where the oriented polypropylene tapes act as the reinforcing fibers in a composite material.

The use of all-polymer composites has also been growing considerably in recent years. Much of the early work in this area was performed by Ward and coworkers, in the development of the hot-compaction method [6-9]. This method involves heating woven polymer fabrics to just below their melting point while under pressure. The outer surface of each fiber or tape melts and fuses together, giving a well-compacted composite, with retention of considerable orientation and very good mechanical properties. Such materials are now commercialized and used in a growing range of applications. One potential drawback for this method is that it has a narrow processing window and so requires very good temperature control. Another approach that overcomes this to some extent is to use a tape that has an outer coating of a lower-melting point copolymer [10]. Compaction can then occur at any temperature between the melting point of the copolymer and the melting point of the polypropylene.

If all-polymer composites were to be made from waste woven polypropylene, the second of these methods would not be possible as it requires the copolymer coating to be applied during production of the tapes. The method of hot-compaction will also be less than ideal in this case, for two reasons. First, the requirement for a very well-controlled process with a narrow temperature window is likely to be difficult with a recycling operation, where low cost and fast throughput are normally required. Second, while there is a significant amount of woven polypropylene entering the waste stream, there is even more low density polyethylene (LDPE) that needs a viable recycling route. These reasons led to the approach adopted here, which was to use the LDPE as a matrix material to bind the polypropylene tapes together [11].

The aims of this work, therefore, were to investigate the potential for production of simple flat sheet composites from a combination of waste woven polypropylene fabrics and LDPE film, using a simple production method with wide process windows. The potential applications for such material are expected to be for low-cost board products, especially for outdoor applications, where wood-based materials tend to be limited.

MATERIALS AND PROCESSING

The raw materials for this study were all provided by Second Life Plastics, Llandeilo, UK, who collect waste agricultural plastics from farms across the West of the UK. Waste fertilizer sacks were used as the source of woven polypropylene fabric, which had a plain weave of 3 mm tapes, with a compressed thickness of 0.21 mm and a weight of 115 g/[m.sup.2]. Two types of LDPE film were also supplied: one with a thickness of 0.15 mm and a weight of 103 g/[m.sup.2] and one with a thickness of 0.12 mm and weight of 73 g/[m.sup.2]. The two types of LDPE were identical apart from different thickness, and these were both used to enable a wider range of compositions to be laid up with alternating sheets. DSC tests on the raw materials had shown that the PP had a melting point (midpoint) of 169[degrees]C, indicating that it is a copolymer grade, while the LDPE had a melting point of 109[degrees]C.

For most of the study, the materials were cut into 200 mm squares, for compression molding. In most cases, the materials were used as-received, with a small amount of surface dirt and water, and unless otherwise stated all results are for material in such condition. For some samples, the squares were water cleaned and/or air-dried. For the final set of samples, shredded polypropylene was used. This was produced either by a controlled manual method of cutting that produced individual tapes of approximately 2 cm length (as shown in Fig. 1), or by a commercial method that produced much finer pieces (as shown in Fig. 2).

The sheets of PP and LDPE were then laid up with alternate layers of PP and LDPE and then compression molded between heated platens. After a certain time period, the compression molder was water-cooled and the compacted composite removed. The variables studied included:

* Amount of polyethylene. This was varied by using different combinations of the polypropylene fabric and polyethylene film, to give samples with weight percentage of polyethylene between 30 and 100% (just LDPE).

[FIGURE 1 OMITTED]

* Cleaning. In most cases, the materials were used as-received, with a moderate amount of dirt on the surface and a very small amount of surface water. These were designated as dirty/dry. Water cleaning and air drying was used to produce clean/dry or clean/wet (with no drying). Dirty/wet samples were also produced by adding some water to the as-received sheets.

* Compression conditions. The compression temperature (platen temperature) was varied from 150 to 200[degrees]C and the applied pressure was varied from 0.5 to 5 MPa. The time the material was held at the compression temperature before compression was held constant at 10 min.

* Lay-up. In most cases, the 200 mm squares of PE and PP were laid up in an alternate arrangement, with samples cut from a 0/90 direction. Some samples were cut in a [+ or -]45 direction. One plaque was produced with a random arrangement of smaller (~20 mm) pieces. Plaques were also produced with a random arrangement of shredded PP and small pieces of LDPE.

The compression molded plaques were cut into parallel-sided strips of size 200 X 15 mm (with a thickness of approximately 6 mm). These samples were used for the flexural tests. For impact tests, the samples were cut into two 100 mm lengths, while for tensile tests, dog-bone samples were cut from the strips.

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

MECHANICAL TESTING AND MICROSCOPY

Flexural, tensile, and impact tests were conducted on the samples produced. Flexural testing was conducted using an Instron electromechanical test machine with a three-point bend geometry. A span length of 100 mm was used, with a deflection rate of 5 mm/min. From the force/deflection curves, the flexural modulus and flexural strength were calculated.

Tensile testing was conducted using a Hounsfield electromechanical test machine, with an extension rate of 5 mm/min. Impact testing was performed using a Ray-Ran impact test machine, using an unnotched Izod configuration. For the flexural and tensile tests, an average of 5 samples was used for each condition; for the impact tests, an average of 10 was used.

Optical microscopy was performed on polished sections through the samples using a Reichart-Jung Microscope. Scanning electron microscopy of the fracture surface of some impact samples was performed.

MECHANICAL PROPERTY RESULTS

Figure 3 shows the effects of material composition on the flexural strength and stiffness (for woven PP, compressed at 160[degrees]C and 5 MPa, with a 0/90 orientation). It can be seen from this that both the strength and stiffness increase as the proportion of PP increases, reaching a maximum at about 50% PP. This suggests that a certain amount of LDPE is needed to give good bonding of the PP layers, but as the LDPE content goes too high, the benefits of the stiffer and stronger PP are diminished. For further tests, a composition of 61% PP was used, as this is most conveniently achieved with alternate lay-ups of the most common thickness materials found in the agricultural waste stream.

Figure 4 shows the effect of compression pressure (for woven PP, compressed at 160[degrees]C, with a composition of 61% PP and a 0/90 orientation). From this it can be seen that there is a continued slight increase in stiffness as the pressure increases, but that the strength only shows slight changes above 2 MPa.

[FIGURE 4 OMITTED]

Figure 5 shows the effect of compression temperature (for woven PP, compressed at 5 MPa, with a composition of 61% PP and a 0/90 orientation). This shows that the flexural strength (and to a lesser extent the stiffness) increase to a maximum at about 170[degrees]C and then decrease as the temperature is increased further to 200[degrees]C. This effect is likely to be due to a balance between improved bonding between PP and LDPE layers as the temperature increases and loss of PP orientation at higher temperatures. Both of these effects are considered further below. The improved contact between materials can be seen in Figs. 6-8, which show optical micrographs of these samples. Figure 6 has been compressed at 150[degrees]C, and shows the PP tapes (darkest regions), the LDPE (mid-grey regions), and the mounting resin filling air-gaps (lightest regions). It can be seen that there is a relatively poor contact between the PP and LDPE, which will presumably lead to poor bonding. As the compression temperature is increased to 170[degrees]C, as shown in Fig. 7, there appears to be much better contact between the PP tapes and LDPE film. The PP tapes are still well defined. With compression molding at 200[degrees]C, above the melting point of the PP, it can be seen from Fig. 8 that although all the material has consolidated, the PP has flowed into the LDPE and presumably all the orientation has been lost.

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

The effects of cleaning and drying are shown in Fig. 9 (for woven PP, compressed at 160[degrees]C and 5 MPa, with a composition of 61% PP and a 0/90 orientation). This shows that there are quite small differences between the samples, although the clean samples have slightly higher strength and stiffness. Surprisingly, the presence of water seems to slightly improve the properties in most cases, possibly owing to the additional internal pressure being generated because of water evaporation during the high temperatures of compression molding. Certainly, the presence of a small amount of water does not lead to significant loss of properties.

The effects of varying the lay-up arrangement are shown in Figs. 10 and 11. Figure 10 shows the flexural strength, the tensile strength, and the impact strength, while Fig. 11 shows the flexural modulus. For all these samples, the proportion of PP was 61%, compressed at 160[degrees]C and 5 MPa. The woven 0/90 had an alternate lay-up of woven pieces, with the testing direction along the PP tapes. The woven [+ or -]45 was similar apart from samples being cut at 45[degrees] to the tape direction. A comparison of these two shows that there is little significant difference in the flexural strength, while the tensile strength is significantly higher for the 0/90 orientation. This shows that for the flexural test, the bonding between layers is the dominant factor, and so the orientation of the tapes is less important. For the tensile tests, the bonding between layers is much less important, and so the 0/90 orientation is considerably stronger. The flexural stiffness is slightly higher for the 0/90 samples, showing that in the initial part of a flexural test, tape orientation does have some importance. Interestingly, the impact strength of the [+ or -]45 samples is higher, possibly because of the reduced stiffness.

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

The careful alternate lay-up of LDPE and PP does tend to give the best properties, but it must be recognized that in a potential commercial operation, the task of cutting and laying up the creased and folded fabrics may be prohibitive. In this case, a more automated process may be required, with the use of a random orientation of smaller pieces. The remaining data sets in Figs. 10 and 11 show this, for a random arrangement of small pieces (about 20 mm square) and the two types of shredded material. These show that not surprisingly, as the degree of automation increases, all the mechanical properties decrease. It seems that if a random arrangement is going to be used, it is better to have smaller pieces, as seen by the better properties of the hand-shredded samples compared with the small pieces. This is presumably because with a random arrangement, there is less chance of PP to LDPE contact, which is required for good consolidation. If the pieces are smaller, then there is likely to be a more uniform distribution of these contacts than with larger pieces. The properties of the hand-shredded samples, while not as good as the woven samples, are still moderate, with the exception of the tensile strength, and this may prove a viable recycling route.

[FIGURE 9 OMITTED]

[FIGURE 10 OMITTED]

The commercially shredded samples gave the poorest mechanical properties, partly because of the very small scale of shredding, and also because of the inclusion of the thread used for stitching of the sacks. Figure 12 shows a fracture surface of an impact sample of the commercially shredded material, which clearly shows the presence of some of the stitching thread that has not bonded at all to the PP or LDPE. Information from sack manufacturers indicated that the stitching thread used is polyester, and so poor bonding to the polyolefins is not surprising.

[FIGURE 11 OMITTED]

[FIGURE 12 OMITTED]

ORIENTATION RECOVERY

It would seem from the results described above that as the temperature increases, the PP tapes lose some of their strength and stiffness because of the recovery of orientation. This is a well-known phenomenon with oriented polymers [11-14]. To study this in more detail, controlled recovery tests were performed on individual PP tapes. These were performed using a Rheometrics Dynamic Mechanical Thermal Analyser (DMTA), operating in thermomechanical analysis (TMA) mode, where a single PP tape was held in a tensile orientation, under a controlled zero load, at an elevated temperature (of 140, 150, or 160[degrees]C) for a period of 30 min. The length of the tape was continually monitored and this was then used to calculate the overall shrinkage (%) caused by the heat treatment. The results of these tests are shown in Fig. 13. From this, it can be seen that the tapes shrink by about 25% at 140[degrees]C, rising to almost 60% at 160[degrees]C. Tests were attempted at 170[degrees]C, but at this temperature, the tapes were not able to support their own weight.

As the orientation recovers and the tapes shrink, it is expected that the modulus of the tapes will decrease. The tensile modulus of individual tapes was also measured on the DMTA before and after heat treatment using a small strain tensile test mode. The results for heat treatment times of both 10 and 30 min are shown in Fig. 14. From this, it can be seen that the modulus of the PP tapes before any heat treatment is about 4.1 GPa, but this drops rapidly with any heat treatment above 140[degrees]C. The loss of modulus occurs primarily within the first 10 min of any heat treatment. With heat treatment at 160[degrees]C, the modulus has dropped to about 700 MPa. This value is at the low end of what would be expected for PP, confirming the DSC findings that it is a copolymer, of fairly low-stiffness.

[FIGURE 13 OMITTED]

BOND STRENGTH TESTS

It is also clear from the results above that the bonding between LDPE and the PP tapes is important, and this was measured for individual PP tapes. Lap-shear samples were made up using individual PP tapes overlapping single pieces of LDPE film, with an overlap length of 10 mm. These were heated in the compression molder (between glass slides) to various temperatures, and then compressed at a pressure of between 5 and 10 MPa. After cooling and careful removal, it was possible to test the samples in the DMTA in a tensile test mode. The shear strength results of these tests are shown in Fig. 15.

As the temperature is increased from 140 to 170[degrees]C, the bond strength between the LDPE and PP increases, reaching 0.45 MPa. Interestingly, performing the same tests between two PP tapes gave reasonable bond strength at 160 and 170[degrees]C. again indicating that the material is a copolymer with a lower melting range.

CONCLUSIONS

Overall, it would seem that the incorporation of waste woven PP into an all-polymer composite with waste LDPE is a potentially viable option for recycling of these materials. Optimum properties were obtained with approximately equal amounts of each material; the LDPE required to provide good bonding and the oriented PP to provide increased strength and stiffness.

[FIGURE 14 OMITTED]

[FIGURE 15 OMITTED]

Processing of such a recycled composite is relatively straightforward via a compression molding technique, with quite wide processing windows in terms of temperature and pressure. The method is also relatively insensitive to the presence of dirt and water; indeed a small amount of water may be beneficial because of additional internal pressure as the water evaporates.

The use of shredded PP does give poorer performance, but may allow much greater amounts of automation on a commercial scale. If shredding is used, it seems important to remove the polyester stitching threads used for the PP sacks.

Considerable recovery of orientation of the PP tapes does occur above 140[degrees]C, accompanied by a dramatic reduction of the PP modulus. Conversely, as the temperature increases, the bond strength between PP and LDPE, and between PP tapes increases. It, therefore, seems that there is a balance between these two effects, with considerable loss of orientation inevitable if reasonable consolidation is to be achieved.

Given that the original modulus of the PP tapes is 4.4 GPa, it is somewhat disappointing that the best composite samples only achieve a flexural modulus of about 1.5 GPa. This is due to the considerable drop in modulus of the PP tapes with heat treatment, coupled with the inclusion of the much lower modulus LDPE and the importance of interfacial bonding. Despite this, the properties achieved are better than for a completely melt-processed mixture (as shown in Fig. 5).

The type of material achievable via this route is never going to compete in terms of properties with virgin all-polymer composites, but it may have a place in cheaper flat-board products, as the property profile (stiffness of about 1.5 GPa, strength up to 40 MPa, and good impact performance) compares well with existing wood-based materials and other recycled polymer products [15]. It will have good outdoor performance compared with wood-based materials and can claim to have reduced environmental impact, being a totally recycled (and recyclable) material.

REFERENCES

1. Department of Environment Transport and Regions, Option for Tackling the Problem of Waste Non-packaging Farm Plastics, DETR, London (1998).

2. Health and Safety at Work Act, HMSO, London (2002).

3. J. Cooper, Mater. Recycl. Weekly, 170, 9 (1997).

4. S. Mason, Mater. Recycl. Weekly, 178, 8 (2001).

5. V.T. Breslin, U. Senturk, and C.C. Berndt, Resour. Conservat. Recycl., 23, 243 (1998).

6. M.I.A. ElMaaty, D.C. Bassett, R.H. Olley, P.J. Hine, and I.M. Ward, J. Mater. Sci., 31, 1157 (1996).

7. P.J. Hine, I.M. Ward, and J. Teckoe, J. Mater. Sci., 33, 2725 (1998).

8. P.J. Hine, I.M. Ward, N.D. Jordan, R. Olley, and D.C. Bassett, J. Macromol. Sci. Phys., B40, 959 (2001).

9. P.J. Hine, I.M. Ward, N.D. Jordan, R. Olley, and D.C. Bassett, Polymer, 44, 1117 (2003).

10. N. Cabrera, B. Alcock, T. Peijs, T. Schimanski, E. Klompen, and J. Loos, "All-Polypropylene Composites for Ultimate Recyclability," in EcoComposites Conference, ESCM, London (2001).

11. J.C. Arnold, W.P. Sinnott, and B.C. Suddell, U.K. Patent WO2004048072 (2004).

12. K. Yamada, M. Kamezawa, and M. Takayanagi, J. Appl. Polym. Sci., 26, 49 (1981).

13. Y.S. Yadav and P.C. Jain, Thermochim. Acta, 117, 97 (1987).

14. E. Andreassen, K. Grostad, O.J. Myhre, M.D. Braathen, E.L. Hinrichsen, A.M.V. Syre, and T.B. Lovgren, J. Appl. Polym. Sci., 57, 1075 (1995).

15. J.C. Arnold and F. O'Brien, "Recycled Composites Produced from Woven Polypropylene Fabrics and Polyethylene Film," in Proceedings of Ecocomposites, T. Peijs, editor, ESCM, London, (2003).

J.C. Arnold, F. O'Brien, M. Moody

Materials Research Centre, School of Engineering, University of Wales Swansea, Singleton Park, Swansea SA2 8PP, United Kingdom

Correspondence to: J.C. Arnold; e-mail: [e-mail]