Recycling of Plastic Materials Part 8 pot

RESULTS The post-consumer plastics waste, sample A, collected from households in Gothenburg contained two different paper components: milk packages lami-nated board and newsprint.. Sampl

industrial plastics waste containing high amounts of paper Such materials can-not be reprocessed without adding a polyolefin-containing fraction, in this case post-consumer waste from Kolding in Denmark The approximate compositions

of the samples A - D are given in Table 1

Hydrolysis

All samples were ground in a mill (Rapid GK 20, high-speed rotation knife mill, normally fitted with a 25 mm screen) When subjecting the waste material to the hydrolytic treatment, three methods can be employed:

• hydrolysis in suspension

• hydrolysis after impregnation with the acid solution and drying

• hydrolysis with acid in the gas-phase

Table 1

Approximate composition of samples A-D

Material Sample A (%) Sample B (%) Sample C (%) Sample D (%)

Other materials

All these methods were discussed previously.3Using the first method, the waste material was suspended in the acid solution (solid/liquid ratio 1/9) and the mix-ture boiled for about one hour Mineral acids with a concentration of 5-10% were used in this case With solutions of organic acids, a simple boiling was not suffi-cient, and the pretreatment had to be carried out at elevated temperatures in a pressure vessel A typical example is a treatment in 3 % oxalic acid solution at

121oC (pressure 2 bars) The hydrolyzed product was then washed with a tap water until neutral, and dried at 50oC for several days (moisture content below 4%) before milling The low temperature was chosen in order to preserve the original fibrous structure of the cellulose component However, the weight loss

of the cellulose component was typically 5-20%, depending on the extent of the hydrolytic breakdown

A technique circumventing the weight loss problem is based on impregnating the material with the acid solution and subsequently drying it at elevated tem-perature In this case the waste was impregnated with a dilute solution of an or-ganic acid such as citric or tartaric acids for an hour, then an excess acid was removed by vacuum filtering and the material was pre-dried at 105oC for 12 hours The actual hydrolytic process took place during homogenization of the material in the compounding extruder at 190oC These two methods of hydroly-sis proved feasible, producing comparable results when tested with LDPE waste containing 30% newsprint (waste from the Lövsta Plant in Stockholm and from the Idelux Plant, Arlon, Belgium).4In principle, one of the most positive results following hydrolysis is the improvement in processability The viscosity of the untreated melt was more than twice as high as that of the hydrolyzed sample, and it also showed significant fluctuations during the experiment These differ-ences appear to be associated with the size distribution of the paper fibers in the melt The average particle size was approximately 20µm and the sample con-tained less than 1% of particles larger than 100µm for samples subjected to hy-drolysis The mean fibre length of unhydrolyzed fibers after injection molding was 800µm; there was also a large number of agglomerates (parts of paper) in the melt

However, a more industrially applicable method of reprocessing paper-con-taminated plastics waste of various origins is the gas-phase method A semi-pi-lot plant reactor was constructed in order to handle larger quantities of waste (previous methods some kg) The reactor tank (length 1 m, inner diameter 0.64

m, volume 300 liters) was built of stainless steel The plastics waste and/or the

paper component was fed into the reactor by a single screw (75 mm diameter), and the material was circulated inside the reactor by a slow-moving PTFE-coated stirrer, which prevented the plastics parts from sticking to the wall The temperature was monitored and adjusted by a thermocouple inside the reactor The reactor was flushed with carbon dioxide and the selected test temperature was set above the boiling point of an acid, which was added to ob-tain a superheated gas phase Different gases were used, but the best results, from the point of view of environmental protection, were obtained with formic acid The 1.2 kg of formic acid (concentration 85%, technical quality) per m3 reac-tor volume was found to be an adequate dose for the hydrolytic treatment The time of hydrolysis in this reactor was quite long and varied from 1 to 3 hours The temperature varied between 150 and 240oC The mixture of paper and molten plastics flowed towards the bottom of the reactor by gravity, the melt tempera-ture was monitored, and the melt was finally extracted with a double screw (di-ameter 100 mm) into a water bath to prevent oxidation and achieve rapid cooling Typically, batch sizes of 25 kg could be handled in this reactor

After hydrolysis the material was ground in the knife mill fitted with an 8 mm screen and stored dry until processing (compounding) The reference material, which was not hydrolysed, was also ground once more with the 8 mm screen fit-ted to the mill

Processing

On a laboratory scale the material was melt-homogenized in a compounding machine (Buss-Kneader PR46) The residence time in the extruder was short, approximately 2 minutes At a melt temperature of 240oC, a good homogeniza-tion of the material was obtained For samples containing only polyolefines,

180oC was sufficient to obtain good homogenization of the material The com-pounded material was milled in the knife mill (8mm screen) and dried at 105oC for 24 hours prior to injection molding

Some samples (B, C, and D) were also melt-homogenized in a larger machine specially designed for mixed plastics waste, at the Cadauta plant in Italy The machine (“Revive” compounder from Cerrini, Busto Arsizio, Italy) has a screw diameter, D, of 120 mm, and a cylinder length of 30D It is a slow running ma-chine (33 rpm) The melt temperature was set at 240oC, but due to the special construction of the machine a high shear-induced heat gradient in the melt was

bly much higher than the set temperature The slow screw speed results in lon-ger residence time (approximately 15 minutes) than in the Buss-Kneader The compounded material was injection-molded into tensile test bars (cross section 3.5×10 mm2and an effective length of 75 mm) using a conventional ma-chine (Arburg Allrounder 211E/171R) The injection pressure was adjusted to obtain satisfactory moldings (upper limit 150 MPa) The melt temperature was

190oC In some cases the material was injection-molded into thin-walled prod-ucts, so called Raschig-rings, in Italy at the Cadauta plant using a normal injec-tion molding machine These rings are normally made of PP, and the holding pressure had to be increased by 50% in order to obtain satisfactory moldings us-ing the hydrolyzed samples

RESULTS

The post-consumer plastics waste, sample A, collected from households in Gothenburg contained two different paper components: milk packages (lami-nated board) and newsprint Lami(lami-nated board is normally very difficult both to grind and to reprocess Sample A contained 46 % paper and was found impossi-ble to injection-mould without hydrolysis The stiffness of sample A subjected to hydrolysis is very high (E-modulus 3.9 GPa, strength at break 15 MPa, elonga-tion 1.2%, impact strength 9 kJ/m2; Charpy test) This is a result of the large amounts of PS and PVC and a high cellulose content The presence of PS and PVC makes sample A very brittle In essence, the CUT-method was found to be the only way to make sample A moldable on a laboratory scale

Sample B, the plastics waste from Skive in Denmark mixed with edge trim-mings from the packaging industry (in total 30 % paper loading), was used in the tests to evaluate the influence of time and temperature using formic acid in the semi-pilot plant reactor In these experiments it was found that the thermal con-duction was low The mechanical properties such as modulus, strength, the cor-responding elongation at break, and impact strength are given in Figure 1 as a function of the melt temperature at the bottom of the reactor The values given

at room temperature (RT) in Figure 1 refer to the unhydrolyzed samples The time to reach the set temperature was approximately 1 hour It is clear from the E-value and IS-value data that the cellulose component, and probably the poly-mer as well, were subjected to severe thermal degradation for reaction tempera-tures above 200oC In general, it can be concluded that the optimum obtained Therefore, the actual melt temperature is not known, but it is

proba-lot reactor is approximately 2 hours (one hour to reach set temperature) Both unhydrolyzed and hydrolyzed samples (samples B, C and D) were com-pounded at the Cadauta plant in Italy using the “Revive” extruder.5After com-pounding, the granulated materials were injection-molded into test bars as well

as into molded cylindrical thin-walled parts (rings, see Figure 4 below)

The samples containing 30% paper could be compounded without difficulty us-ing the “Revive” machine The results obtained on the molded test bars are

Figure 1 Influence of the hydrolysis temperature (melt temperature in the ractor) on the mechani-cal properties of plastic waste containing 30% paper Sample B, formic acid in gas phase, treatment time 3 hours Values at room temperature (RT) refer to the unhydrolized sample.

temperature of hydrolysis is approximately 180 C, and the time in the

semi-pi-• the E-modulus decreased with hydrolysis, due to fibre length reduction

• the strength value was almost unaffected by hydrolysis

• there was an increase in elongation at break and in impact strength for the samples subjected to hydrolysis

In a series of experiments the cellulose content was varied between 18 and 45

% by changing the portion of paper component added to sample D The improve-ments in elongation at break following hydrolysis are illustrated in Figure 3 (un-filled symbols - unhydrolyzed samples; (un-filled symbols - hydrolyzed samples)

Figure 2 The mechanical properties of untreated and treated (formic acid in gas phase, 200oC, 3 h) plastic waste, sample D, containing 30% paper Sample D without paper: modulus - 1.1 GPa, strength − 17 MPa, elongation at break - exceeding 250%, no break indication at impact testing.

samples C and D:

shown in Figure 2 (sample D) The following behavior was observed for both

Figure 3 Elongation at break and E-modulus vs paper content for untreated (unfilled symbols) and treated sample D (filled symbols, formic acid in gas-phase, 200oC, 3 h).

The strength values are not influenced to a great extent for different paper con-tents (strength at break approximately 12 MPa), but the E-modulus values in-crease with increasing paper content for both unhydrolyzed and hydrolyzed samples as shown in Figure 3 Also, the flow behavior, as measured by spiral molding tests, was found to be improved 20% by the hydrolysis as a consequence

of the reduction of a fibre length of a cellulose component These short fibers do not act as reinforcing fibers and the adhesion between the fibers and the matrix

is poor; thus the strength is not improved by the incorporation of paper The stiffness, on the other hand, is naturally increased for increasing paper load be-cause of a high modulus of the fibers (approximately 20 GPa) In general, the re-sults are in an agreement with earlier findings for PP filled with hydrolyzed cellulose.6

Some of the samples compounded on the “Revive” machine were injec-tion-molded in Italy into thin-walled rings from material B, see Figure 4 The re-sults showed that thin-walled components containing 30% paper could be injection-molded (hold pressure has to be increased by 50%) When the paper content was reduced to 15% the components made out of sample B showed good elasticity and good surface finish for the hydrolyzed samples (E-modulus 1.1 GPa, strength 12 MPa and elongation at break 10%) Samples C and D could also be injection molded into rings (sample A not tested)

Figure 4 Photograph of injection-molded thin-walled parts (rings) Molded in Italy by Cadauta,

S Sebastiano da Po, Torino (sample B) Total width 310 mm, total mass, including runner 25 g, mass of each part 2 g (cylindrical, diameter 25 mm, height 25 mm, wall thickness 1 mm) These rings are normally made of PP.

The results presented here demonstrate that significant improvements in the homogeneity and processability of plastics waste of different compositions and origins contaminated with paper can be achieved using a relatively simple hydrolytic treatment Such treatment improves the mechanical properties of the material, and PFMW and similar plastics waste containing as much as 40% paper become processable into products with acceptable surface appearance The paper content at which the hydrolysis makes the most pronounced impact

on the properties of the material seems to be in the range 20 to 30 % However, the greater homogeneity of the paper-containing samples following hydrolysis

is in many cases a desired property, even at very low paper content, specially when molding thin-walled products

The gas-phase technique is an economical method of hydrolysis which seems to

be adaptable to the industrial practice The temperature of hydrolysis must, however, not exceed 180oC On an industrial scale the hydrolysis may be per-formed in a screw reactor in order to improve temperature control

In conclusion, the method of hydrolysis offers an efficient and economical way

of processing plastics waste, both post-consumer waste and industrial waste, contaminated with a cellulose component The addition of cellulose gives a de-sired stiffness to the final product, as reported in earlier publications.4,6Thus, the plastics waste containing a cellulose component can be used in several appli-cations, such as artificial wood

ACKNOWLEDGMENT

The support of the Commission of the European Communities and of the Na-tional Swedish Board for Technical Development in financing this work is grate-fully acknowledged

REFERENCES

1. S Gotoh, in Sorting of Household Waste and Thermal Treatment of Waste,

Eds M P Ferranti and G L Ferrero, Elsevier, London, pp 441-443, 1985.

2. C Klason and J Kubát, Sw Patent and Eur Pat Appl 8107444-5, US 4,559,376.

3. C Klason, J Kubát, A Mathiasson, M Qvist, and H R Skov, Cellulose Chem Tech,

23, 131 (1989).

4. A Mathiasson, C Klason, J Kubát, and H R Skov, Resources, Conservation and

Recycling, 2, 57 (1988).

Eds M P Ferranti and G L Ferrero, Elsevier, London, pp 441-443, 1985.

6. A Boldizar, C Klason, J Kubát, P Näslund, and P Saha, Int J Polym Mater., 11,

229 (1987).

5. R Fornasero, in Sorting of Household Waste and Thermal Treatment of Waste,

Processing of Mixed Plastic Waste

A Vezzoli, C A Beretta, and M Lamperti

CSI - Ricerca Applicata Montedison, Viale Lombardia 20,

20021 Bollate (MI), Italy

INTRODUCTION

Post-consumer plastic wastes can be divided into two different groups depend-ing on their source:

• mixed plastics from the household waste

• plastics from the industrial sectors

The first category involves the medium/short life articles that are used in food, pharmaceutical, and detergent packaging, shopping, and others The majority

of these articles are composed of thin protective films, sheeting for blisters, strapping, thermo-formed trays, as well as a variety of bottles for soft drinks, food and cosmetics

As shown in Figure 1, there are mainly five different polymers that contribute

to the total amount of plastic waste, and they are commodities (PE, PP, PS, PVC, PET) The plastic composition of the above mentioned mixture can change de-pending on the regional habits and the seasons of a year Also the mode of a waste collection can influence its final composition

The category of plastic waste from the industrial sectors concerns the me-dium/long life articles like the products for cars industry, furniture, appliances, etc The problem of these sectors is a wide variety of engineering materials and

a high number of components employed to build a final system

This paper reports two recycling approaches adapted in CSI to establish appro-priate technologies and/or design concepts for the above mentioned groups of plastic waste

MIXED PLASTICS FROM HOUSEHOLD WASTE

A study of this type of a plastic waste has been performed at CSI using samples that came from two different experimental collection systems, the first in plastic boxes (internal volume: 3 m3), and the second in arranged plastic areas inside