Processing and Characterization of PET Composites Reinforced With Geopolymer Concrete Waste

In the present study, poly (ethylene terephthalate)-based composites were produced and characterized. These composites were composed by poly (ethylene terephthalate) (PET) reinforced with geopolymer concrete waste (GCW). Both untreated (U-GCW) and treated with oleic acid (OA) geopolymer concrete waste (T-GCW) were used in the production of the composites. The PET/GCW ratios used for either treated or untreated GCW bodies were 80/20 (wt%), 60/40 (wt%) and 50/50 (wt%). Chemical compositions were assessed by X-ray fluorescence spectroscopy (XRF), crystallinity by differential scanning calorimetry (DSC), thermal stability by thermogravimetry (TGA), microstructure by field emission gun scanning electron microscopy (FEG-SEM) with energy dispersive X-ray spectroscopy (EDS), and mechanical properties were assessed by compression tests. Fourier transform infrared spectroscopy (FT-IR) was used to check the efficiency of the treatment with OA, as well as the interaction between PET and GCW. The T-GCW PET composites showed better thermal, physical, and mechanical properties, for non-structural applications, when compared to U-GCW.


Introduction
2][3][4][5] Gorrasi et al. 2 prepared and characterized the physical properties of a PET-based composite reinforced by dispersed halloysite nanotubes for packaging applications.According to these authors, the behavior of the composites was similar to that observed in the literature.Kilinc et al. 3 investigated the flame retardancy and tensile properties of PET-based composites containing phosphorus and boron-based additives, which were processed by extrusion and subsequently molded in an injection machine.These authors reported a high flame retardancy and strength improvement after the additions.However, other authors have investigated the properties derived by the addition of the Fe 2 O 3, CaCO 3 , TiO 2 , and ZnO nanofillers 4 .Chae et al. 5 studied the thermal and rheological properties of highly concentrated PET composites with ferrite nanoparticles.The results showed that ferrite nanoparticles played an important role in PET crystallization, due to heterogeneous nucleation.
Consequently, other characteristics were influenced, such as the thermal stability of the matrix by obstruction of the heat permeability and low diffusion of volatile degradation products.The tensile strength and rupture energy of hybrid poly (methyl vinyl siloxane) composites reinforced with short PET fibers and wollastonite whiskers were studied by Fu et al. 6 The authors observed an increase in strength with increasing wollastonite content, from 5 to 45 wt%.The authors expected an opposite effect, due to change on the interface properties between PET fibers and matrix, caused by the incorporation of wollastonite whiskers.However, interfacial frictional stress caused by matrix shrinkage during the curing process may have accounted for the strengthening of the composite.In another study, Scaffaro et al. 7 reported the effect of nature and content of organically modified clay on the properties of PET nanocomposites.However, the filler content used were low (3, 5 and 10 wt%) for each kind of Cloisite clay (15A and 30B), when compared to the contents used in the present study.These authors indicated intercalated structure for the lowest clay content (3 wt%), while increasing filler amount induced the coalescence of clay particles, and a coarse morphology.These authors observed the trend of strong degradation of the PET during processing due the presence of organic modifier, a slight increase in crystallinity and a slight change on Young's modulus for all clay contents below 10 wt%.
The class of aluminosilicates is based on [SiO 4 -] and [AlO 4 -] tetrahedral units, that are reactive inorganic oxides in solution.Cement-derived materials, such as geopolymers, are composed by several kinds of aluminosilicates, ferroaluminates and other compounds, which can be extracted from fly ash, mud, ground granulated blast furnace slags, coal ash, rice husk ash, kaolinite, metakaolinite, palm oils and others.This utilization decreases the amount of urban solid waste 8,9 .
The purpose of this study was to develop composites containing high amounts of pulverized geopolymer concrete waste, treated and untreated with oleic acid.The results highlight the potential of the composites in the substitution of conventional materials by novel composites with optimized thermal stability, thermal-acoustic insulation, density, mechanical resistance and processability.

Materials and Methods
The poly (ethylene terephthalate) (PET) was provided by the Instituto de Macromoléculas Professora Eloisa Mano (IMA); the geopolymer concrete waste (GCW) was supplied by Lafarge Concrete Ltda and the oleic acid (OA, 99% purity) was purchased from Sigma Aldrich.
The GCW fillers were modified by a treatment with oleic acid (OA), in which 100 g of the geopolymer concrete particles were added to 450 ml of a 0.33% (v/v) OA solution in ethanol.The OA treatment was kept under stirring during 24 hours, and the treated geopolymer materials were filtered and washed with ethanol to remove the unreacted oleic acid.After that, treated and untreated materials were dried.Consequently, the materials were referred as untreated geopolymer concrete waste (U-GCW) and treated geopolymer concrete waste (T-GCW).The modification was performed in an attempt to compatibilize organic and inorganic phases, dispersing the hybrid fillers and increasing fluidity of the melt polymer 13 .
The composite specimens were processed by hot pressing the mixed pulverized PET and T-GCW or U-GCW at 270ºC, in proportions 80/20 (wt%), 60/40 (wt%) and 50/50 (wt%).Cylinder PET/U-GCW and PET/T-GCW specimens were produced for the compression tests, according to ASTM D695.Therefore, 06 groups of composites were produced and characterized.
Chemical composition triplicates of the geopolymer concrete waste were evaluated by X-ray fluorescence analysis (XRF) in a Rigaku RIX 3100 spectrometer.The amount of each oxide was determined and expressed in terms of mass percentage (wt%).
Fourier transform infrared spectroscopy (FT-IR) was performed with a Shimadzu, Model IR-21 PRESTIGE spectrometer, using a KBr disk.The spectra were taken from 400 to 4000 cm -1 with 128 scans and 2 cm -1 of resolution.Characteristic absorptions bands of the processed composites were registered.
Thermal degradations of the composites were assessed by means of a TA Instrument Q500 thermogravimetry (TGA) equipment.The analyses were carried out from 30 to 700°C, at 10°C/min, under nitrogen atmosphere.10 mg samples were analyzed.The onset, maximum degradation and end temperatures (T onset , T max and T end ) were registered and compared.
Differential Scanning Calorimetric (DSC) analyses were carried out using a TA Instruments Q1000 calorimeter.The polymer composites were heated from 30 to 300°C at 10°C/min.All composites were heated and kept at the maximum temperature during 2 min, and cooled to the minimum temperature (30°C) at 10°C/min.This first step aims to bring the composite to the as-received condition.After this, the bodies were heated up to 300°C at 10°C/min.The crystallinity grades of the composites were calculated for each material according to the methodology described elsewhere 14 .The overlay of the curves is shown between the temperatures at 30 and 300°C.
The morphology, dispersion of particles and interface structure of the composites were observed in a field emission gun scanning electron microscope (FEG-SEM), model FEG QUANTA 250 (FEI), with accelerate voltage of 30 kV.The samples were placed coated with a platinum layer for avoiding charging.The compositions of samples were evaluated and confirmed by energy dispersive X-ray spectroscopy (EDS) analysis.
The mechanical properties were determined in a tensilecompression test machine (EMIC DL 10000), equipped with a 100 kN compression-tension cell.The tests were performed at 25ºC at a strain rate of 1.3 ± 0.3 mm/min.Cylindrical specimens were machined to the dimensions of 2.5 cm diameter and 5 cm height, required for plane strain compression tests by ASTM 695.The strain percentage of the stress-strain curves was normalized with respect to the initial sample dimensions 15 .

Results and Discussion
X-ray fluorescence data are presented in the Table 1.SiO2, CaO, Al2O3, Fe2O3 and K 2 O were the major components.According to the literature, these oxides are the main components of geopolymer concrete 16,17,18 .
FT-IR analyses allowed the identification of the functional groups present on PET polymer, U-GCW and T-GCW.Figure 1 shows the corresponding infrared absorption bands of PET polymer.The identified vibrational groups are summarized on the Table 2.The bands are in agreement with Edge et al. 19 and Silverstein and Webster 20 .
summarizes these results.
According to Figures 3 and 4, the PET/U-GCW and PET/T-GCW composites presented the absorptions corresponding to fillers of the geopolymer concrete waste and assigned to functional groups ascribed.The spectra of the PET/U-GCW, 80/20, 60/40 and 50/50 revealed the polymeric absorption bands with lower intensities than treated composites.It suggests that the compatibilization between the surface of the treated fillers and the polymer has occurred, providing a good adhesion between matrix and filler.As expected, the increase on filler amount was compatible with decreasing intensities of the absorption bands corresponding to the PET, for all composites, while increasing intensities of the geopolymer concrete waste bands were observed.
The thermal stability of the PET, neat and treated geopolymer concrete as well as hybrid composites were evaluated by thermogravimetric analysis.Figure 5 presents the data of TG/DTG corresponding to PET.The TG/DTG curves of the PET polymer showed one degradation step, with T onset around at 340ºC, indicating the loss of vinyl ester and acid end groups and the formation of cyclic oligomers.The DTG curve indicated T max at 434ºC, which represents the decomposition of cyclic oligomers and release of acetaldehydes groups and anhydrides oligomers 26 .The values of weight loss of all materials, T onset , T max and T end are presented in Table 5. Untreated geopolymer concrete waste and treated geopolymer concrete waste showed different behaviors (Figure 6).Initially, one peak of mass loss at 71ºC and 77ºC can be observed for each geopolymer material, untreated and treated, respectively.The effects can be attributed to Si-O bending 22,32 465 O-Si-O, bending vibration 32,33 liberation of free water present into pulverized materials, as C-S-H type or N-A-S-H gels and free water liberated by reaction between the oleic acid with hydroxyl groups onto the surface of the geopolymer materials during the modification synthesis 13,27 .According to curves of untreated geopolymer and treated material can suggest the efficiency of the surface chemical modification with oleic acid.The effect was more evident by elucidated event with high loss weight at 447ºC.Figures 7 and 8 show the TG/DTG curves for the PET/U-GCW composites.The results show the increase on the thermal stability of the composites (80/20, 60/40 and 50/50) with increasing filler content, and the degradation onset temperature was observed at 374ºC, with residue content of 66%, 51% and 44%, respectively.The first step presents the T max of the PET/U-GCW in the range of 431-437ºC, which corresponding to degradation of the polymer.The second step was detected at 580ºC and can be attributed to thermal decomposition of the calcium hydroxide, while the third step (around to 680ºC) can correspond to silica dehydroxylation, decomposition of aluminates and carbonates of the geopolymer concrete 13,14,18 .These behaviors reveal no chemical interaction between PET and filler content on untreated composites.Figure 9 shows the TG curve of the treated hybrid composites.According to results, the obtained esterification of the hybrid fillers favored the compatibilizing with the polymer matrix, what raised the thermal stability of the PET/T-GCW composites.The T onset of the treated composites (80/20, 60/40 and 50/50), were displaced to range of 380ºC and the amount of residues, 63%, 59% and 40%, were related to the efficient compatibilization between the fillers and polymer with the insertion of the OA.The DTG curves of the PET/T-GCW composites show one initial degradation temperature around to 263ºC that was related to ebullition of the OA and the three steps of degradation of the composite samples.DTG curves (Figure 10) of the PET/T-GCW composites presented the three steps of degradation with similar temperatures those were observed between the PET/U-GCW composites corresponding to events described.
The T g of the polymer was detected at 78ºC.The effects of the high proportions of geopolymer concrete   waste (untreated or treated) into the processed polymer matrix on melting and crystallization temperatures (T m , T c ), and degree of crystallinity were observed and analyzed (Table 6).The T m peaks of the all processed polymeric composites are slightly lower or higher than observed for PET.This effect can be attributed to lower crystallinity of the polymer in the composites by high amount of fillers embedded into the polymer matrix, which made recrystallization difficult of PET chains.The influence to 80/20 composites with and without treatment was lower, what can be explained by higher polymer proportion.This led to a slight increase in T m value.The crystallinity degree (X c ) of the PET matrix in the composites treated with oleic acid was lower than no treatment composites.Probably, the long chain of OA, anchored onto the fillers and polymer chain hampered the better crystallization 28,29 .This result is in accordance with the better processability and dispersion of fillers observed by FEG-SEM microscopy.
FEG-SEM electron micrographs (Figures 11 and  12) show embedded fillers into the polymer matrix, observed for both composites; this finding suggests that the melt processing of the PET was efficient.However, the treated composites showed the development of a more compacted structure, what can corroborate the efficiency of the surface treatment with OA compatibilizing both the inorganic surface of fillers and organic material.The electron micrographs of the external surfaces of all molded specimens show the polymeric matrix with several kinds of fillers disperses into the polymer, thereby the molding processing of higher amount of geopolymer concrete waste into the PET polymer was effective (Figures 13 and 14).The images showed better dispersions of the GCW to all the treated composites than no treatment composites with the increase in the amount of geopolymer concrete waste.The created porosity in the composites are the results of the releasing of the gases during the melt PET processing to molding of the specimens, however can exist some influence of the releasing of free water present into the filler, as observed by thermogravimetry results.The higher porosity of the composites 60/40 and 50/50 can suggest that the covered filler interfered in the output of gas liberated by melting of the PET.The EDS spectra of the composites showed the presence of high amount of elements presented into the aluminosilicates materials and carbon.These results corroborated with XFR and FT-IR analysis, which confirmed the composition of the geopolymer concrete waste and presence of the polymer in the composites.
The specimens with higher proportions of polymer exhibited a less pulverulent, brighter and smooth surface.This is due to the polymer present in the composites migrating to the surface of the specimen, promoting the appearance of a film, this effect being similar to the core-shell (Figure 15).Results of compression tests in the composite specimens molded under the melt state containing high geopolymer concrete waste contents.Figure 16 shows the compression tests results for the PET/U-GCW and PET/T-GCW composites.The 50/50 composites presented the highest mechanical performance,         and this finding is probably due to the low PET content, contributing to a rigid of the composite specimens.After the compression test, a high concentration of pores was observed for all samples, which corroborated with the FEG-SEM analyses.The porous structure of hybrid composites was assigned as results of the release of a high amount of CO 2 , in which a low amount was retained into the PET matrix.Fragility places of the composites were associated to the voids created and contributed to the rupture of all samples.Table 7 shows the compression resistance of the composites.The results presented the better compression properties for treated hybrid composites, what can corroborate with the efficiency

Conclusion
X-ray fluorescence and FT-IR analyses confirmed the geopolymer composition of the concrete waste, while the increase of the thermal stability of the polymer with higher amount of fillers in the composites was attested by thermogravimetric analysis (TGA).
DSC analysis permitted to verify the lower crystallization degree for treated geopolymer concrete waste composites than untreated geopolymer concrete waste composites and poly (ethylene terephthalate), which corroborate with the effect of the good dispersion of treated fillers and the difficulty to recrystallization of the polymer due the long chain of the oleic acid between polymer and filler surface.
The results of the study showed the melt processability and moldability of all composites by heat pressing.
The compatibilization between the geopolymer concrete waste and organic matrix was efficient and reflected in better compression resistance to the treated composites.
The good adhesion between filler and polymer was observed in the treated composites by FEG-SEM, what can be attributed to presence of oleic acid arrested onto the geopolymer concrete waste surfaces.Mechanical tests allowed to verify the good performance between all composites, however showed that the effect of the compatibilizing improved the better results between the composites.

Figure 2 .
Figure 2. Overlay of the FT-IR spectra of T-GCW and U-GCW.

Figure 3 .
Figure 3. Overlay of the FT-IR spectra for the PET/U-GCW composites.

Figure 4 .
Figure 4. Overlay of the FT-IR spectra for the PET/T-GCW composites.

Figure 5 .
Figure 5. TG and DTG curves for the PET samples.

Figure 6 .
Figure 6.Overlay of the TG and DTG curves of T-GCW and GCW composites.

Figure 7 .
Figure 7. Overlay of the TG curves of PET/U-GCW composites.

Figure 8 .
Figure 8. Overlay of the DTG curves of PET/U-GCW composites.

Figure 9 .
Figure 9. Overlay of the TG curves of PET/T-GCW composites.

Figure 10 .
Figure 10.Overlay of the DTG curves of PET/T-GCW composites.

Figure 16 .
Figure 16.Graphic of compression test with standard deviation of the PET/U-GCW and PET/T-GCW.

Table 1 .
Chemical composition of the U-GCW obtained by X-ray fluorescence.
Figure 1.FT-IR spectrum of the PET sample.

Table 2 .
Bands with assigned vibrational modes from FTIR specgtrum of the PET sample.

Table 4 .
Characteristics bands of the T-GCW FTIR samples obtained by FT-IR.

Table 5 .
TG/DTG properties of the specimens.

Table 6 .
Calorimetric properties of the specimens.