Recycling of Glass Fibers from Fiberglass Polyester Waste Composite for the Manufacture of Glass-Ceramic Materials ()
1. Introduction
The sustainable elimination of composites still remains as a challenge nowadays. On one hand, global production of composites materials increases every year, and it is expected to reach 10.3 Mt in 2015. Of all these composites, about 90% corresponds to thermostable composites with glass fibers (fiberglass reinforced plastics (FGRP)). The recycling of these composites is not, at present, profitable in economic terms, because obtained fibers present lower mechanical properties than the original ones, and cannot be employed in the manufacture of structural materials. Therefore, most of the waste glass fiber composites are stored in landfills or buried. This arises serious environmental problems, due to this kind of wastes are usually non-biodegradable and very bulky.
In absence of specific legislation in the EU, these wastes are often apply the European Directive on End-of Life Vehicles (1999/31/EC, 2000/53/EC) [1,2] and the European Directive on Waste Electrical and Electronic Electronic Equipment, (WEEE), (2002/96/EC) [3]. These legislations limit the amount of waste that may be set aside for landfills. Besides, UK and Germany has implemented a total ban for the dumping of this waste.
Recycling composites is difficult since they typically contain two or more components (filler, fiber, resin, etc.). Recycling thermoset composites is a particular challenge since, once the thermoset matrix molecules are crosslinked, the resulting material can no longer be remelted or remoulded [4]. Further, the most common thermosetting resins, such as polyester and epoxy resin, cannot be depolymerised to their original constituents [5]. Thermoset composites therefore commonly end up in landfills, and since the components are nonbiodegradable, the economic costs of such disposal can be very high.
A number of technologies have been proposed for recycling thermoset composites:
1) Mechanical comminution-based processes to obtain a specific grain size that renders the material useful as reinforcement in new formulations [6-8].
2) Thermal processing such as:
a) Combustion and the use of the heat generated [9, 10].
b) Pyrolysis [11] and thermolysis [12].
c) Fluidised bed thermal processes to recover the carbon fiber reinforcement from composites [13,14].
3) Composite depolymerisation based on chemical processes such as hydrolysis, glycolysis and solvolysis to recover organic materials [15].
The glass fibers obtained as a solid residue in pyrolysis suffer the degradation of their physical properties, limiting their reuse [5,14]. Nonetheless, the residual glass fiber was used in the manufacture of new composites with no important decline in the mechanical properties of the final product [16,17]. None of these techniques achieves more than the partial recovery of glass fiber and packing.
This work proposes the reuse of glass fiber in the production of glass-ceramic materials. Glass-ceramics are polycrystalline materials of fine microstructure that are produced by the controlled crystallisation (devitrification) of a glass. The first glass-ceramics, developed in the 1950s, were produced via a conventional glass-making process, followed by crystallisation involving heating and later cooling (allowing nucleation and crystallisation respectively). In recent years, however, the sintering method has become a technologically viable route of glass-ceramics manufacture. Sintered glass-ceramics are usually made by milling a glass frit into particles of adequate size, heating to provide sintering, and then allowing crystallisation of glass particles.
Glass-ceramics have found a wide variety of applications in different technological fields [18]. The most important glass-ceramic for architectural applications is Neoparies®. This wollastonite material is produced on a large scale for building interior and exterior walls. Large flat or curved sheets of this material can also be produced for facing buildings. The main properties of Neoparies® include great resistance to weathering, zero water absorption rate, hardness (which is greater than that of natural stone), light weight (30% lighter than natural stone building materials), and the ease with which curved panels can be made [19].
This paper evaluates the valorisation of waste composites by a thermolysis and gasification process and the suitability of the glass fiber as alternative raw materials in the manufacture of wollastonite-plagioclase glassceramic material for architectural applications. The thermolysis process basically involves pyrolysis the waste composite at high temperature (typically 550˚C) in the very little oxygen present (<3 vol.%) this prevents combustion reactions taking place to decompose the polymer matrix into oil, char and gas. After the polymer has been removed, the fibers are heated in air to oxidize residual char and remove surface contamination. The fibers are then recovered for reuse via glass-ceramic material.
2. Experimental
The polyester fiberglass (PFG) waste (Polifibra S.A., Guadalajara, Spain) used in this work was composed of E-glass fiber (SiO2: 54.3 wt%, Al2O3: 15.2 wt%, CaO: 17.2 wt%, MgO: 4.7 wt%, B2O3: 8.0 wt%) plus unsaturated polyester resin made from orthophtalic acid and styrene. Table 1 shows the composition of PFG waste.
The total amount of organic matter contained in such PFG waste, deduced from Table 1, is 35.5 wt%.
The elemental composition of the PFG waste was determined using an automated LECO CHNS 923 analyser (Table 2) [20].
The thermal behavior of the PFG samples (particle size = 100 - 200 µm) was studied using a Setaram Sensys Evolution 1500 thermal analysis system equipped with a differential thermal analyser (DTA) and a thermogravimetric analyser (TGA). The samples were heated at 800˚C at heating rates of 10˚C·min−1 in pure air (20 ml·min−1).
PFG was treated at 550˚C for 3 h in a 9.6 dm3 thermolytic reactor, which consists of a heating system and a gas condensation device. Temperature of 550˚C was selected as the working temperature based on preliminary studies. The experiment was performed in triplicate. This process of thermolysis yielded a solid residue, oil and a non-condensed gas [20]. The amount of gas generated was estimated by the difference between the initial weight of PFG and the amount of liquids and solids obtained.
The solid residue obtained in thermolysis was oxidized in air atmosphere. After thermolysis, pressurized air was injected into the reactor (20 l/h) maintaining the temperature at 550˚C. The final result is a glass fiber without organic matter. The recovering of weight in glass fiber during gasification stage was calculated by using Equation (1):
(1)
Table 1. Composition of the PGF waste used in the experiments.
Table 2. Elemental composition and gross calorific value (GCV) of the PGF waste and the glass fiber residue obtained by thermolysis.
where 
is the yield (mass) of the solid residue obtained during thermolysis, 
is the mass loss during gasification and 
is the glass fiber content of the initial PGF waste (see Table 1).
Elemental analyses of the solid residue obtained in the thermolysis process were undertaken using a LECO TGA 701 and LECO CHNS 923 analyser respectively. Morphological studies of the post-thermolysis solid residue (char-covered glass fibers) were performed using a Hitachi model S-2100 scanning electron microscope (SEM). Samples were coated in graphite for observation. The clean glass fibers recovered after gasification were gold plated and examined using a Jeol JSM 6500 F field emission microscope (FEM). Fiber diameter was determined according to British Standard ISO 11567 [22].
Tensile tests were performed on individual fiber glass filaments. The fibers were extracted by carefully pulling them with standard tweezers and then glued onto a cardboard frame with an epoxy adhesive (Araldit), the glue was allowed to cure for a day. The frame consist of a small thin rectangle about 1 mm by 8 mm, with a small rectangle cut from the middle about 0.5 mm by 4 mm. The cardboard frame was fixed to the grips of a microelectro-mechanical testing machine for the tensile tests (Kammrath & Weiss’ Tensile/Compression Stage). After cutting the cardboard sides, the load was immediately exerted on the fiber and the tensile test carried out under stroke control at a cross head speed of 2 µm per minute. The load carried by the fibers was measured with the 500 mN load cell of the testing machine while the fiber elongation was determined directly with the cross-head displacement and, therefore, including the elastic compliance of the load system. This method, as the opposite to measure the actual cross-section of each individual fiber, is specially suited for fibers whose diameter distribution is fairly constant as in this case [23]. The Young’s modulus of the fiber glass filament was also calculated.
The clean glass fiber recovered after gasification was ground using a BIOMETAL RETSCH PM 100 ball mill at 500 rpm for 15 min. A glass was formulated with incorporation 5% of Na2O to facilitate the melting process.
The components (204.25 g of the resulting powder (particle size < 250 µm) and 18.4 g of Na2O (as Na2CO3) were mixed for 30 minutes in a blender (TURBULA) to get a homogeneous mixture. The batch was placed in an aluminosilicate crucible and heated at 10˚C·min−1 in an electric furnace up to 1450˚C. After a holding time of 120 min at the melting temperature, the melt was quenched by pouring into water producing a glass frit. This frit was then ground using a BIOMETAL RETSCH PM 100 ball mill at 400 rpm, and several fractions of different sized particles separated (see Table 3) with the aim of determining the effect of particle size on glass crystallisation.
The thermal stability of these different glass fractions and their preferential crystallisation mechanisms (surface or bulk) were studied by DTA employing a SETARAM LABSYS TG apparatus. DTA analyses were performed between 25˚C and 1400˚C in air, using calcined Al2O3 as a reference material. All analyses were performed at a heating rate of 50˚C·min−1. The DTA curves were normalised to sample weight. The evaluation of the amorphous nature of glass after melting and the mineralogical study of the crystalline phases devitrified after thermal treatment was performed by X-ray diffraction (XRD) (Philips model X’PERT MPD) with Ni-filtered Cu Kα radiation operating at 30 mA and 50 kV.
The feasibility of the sintering + crystallization process to produce glass-ceramic tiles was evaluated on a mixture of different particle size glass powders: 1600 - 2000 µm, 160 - 250 µm and 80 - 100 µm and on glass powder with particle size of 80 - 100 µm. The percentages of the different fractions (1600 - 2000 µm (42 wt%), 160 - 250 µm (34 wt%) and 80 - 100 µm (24 wt%)) and 80 - 100 µm (100 wt%) respectively, were randomly chosen. The samples were compacted by vibration in a plaster mould and afterward fired at 1013˚C for 60 minutes with heating and cooling rate of 50˚C·min−1.
3. Results and Discussion
Figure 1 shows the TG/DTG curves for the thermal degradation of PFG waste in air. The first phase (dehydration) temperature peak occurs at around 210˚C (mass loss = 1.8 wt%). The second and main phase of degradation temperature peak occurs between 259˚C and 392˚C (mass loss = 26.4 wt%). In the isophthalic acid-based polyester resin the second step involves scission at the cross-link and formation of styrene and the linear poly