Unsaturated polyester resins from poly(ethylene terephthalate) waste for polymer concrete.
Unsaturated polyester (UP) resins are produced by reacting an unsaturated dibasic acid or a mixture thereof with a saturated dibasic acid with a polyhydric alcohol. The product is then dissolved in a vinyl monomer. These unsaturated polyester resins may be cured by an organic peroxide (1). A few years ago, some work described a technique for producing polyester resins based on recycled poly(ethylene terephthalate), PET (2, 3).
Polymer concrete (PC) formed from a resin and inorganic fillers is strong and durable and has a variety of applications (4). The excellent mechanical and durable properties of the material reduce maintenance and repair activities. The fast cure time of PC imparts an important advantage for this type of concrete in mending bridges and floors in industrial facilities. PC is excellent for precast components, because the short cure time enables the fast and efficient use of forms and other production facilities (2). In addition to adequate strength and bond, resins for bridge deck overlays require high elongation and low modulus to provide compatibility with thermal movements. Flexible resins should be used to meet this requirement.
However, the high cost of resins used in PC limits the growth of PC-based products. Poly(ethylene terephthalate), recovered from scrap textile or beverage bottles, may provide a potentially lower cost source of resin, and its recycling in PC will also help reduce an environmental problem (5, 6). This article deals with recycling of PET for the purpose of preparing flexible resins and evaluates the feasibility of using the recycled products in producing durable PC.
Diethylene glycol (DG), propylene glycol (PG), maleic anhydride (MA), sebacic acid (SE), and styrene (Sty) were reagent grade and used without further purification. All chemicals were obtained from Aldrich.
Methyl ethyl ketone peroxide (MEKP) and 10 wt% cobalt octoate solution in styrene were obtained from Saudi Industrial Resins (SIR), Jeddah.
Synthesis of Unsaturated Polyester Resins
The PET textile waste was depolymerized using eleven different mixtures of DG and PG in the presence of manganese acetate as a trans-esterification catalyst. The concentration of the catalyst was 0.5 wt% based on the weight of the PET. The glycol mixture used for glycolysis was 65 wt% of the weight of PET. The reaction was carried out at about 200 [degrees] C under reflux for 4 h in nitrogen atmosphere, and at 210-230 [degrees] C for 3 h. The temperature of the reaction system was then allowed to drop to 100 [degrees] C, where it was maintained for 1 h. The temperature of the reaction mixture was then allowed to drop to room temperature. The glycolyzed products were then analyzed for hydroxyl value and the amount of free glycols. The hydroxyl values were determined by the conventional acetic anhydride/pyridine method (7). To determine the amount of the free glycol, a weighed quantity of the glycolyzed product was extracted with water and filtered. The aqueous filtrate containing free glycol and some water-soluble oligomers was concentrated by evaporation of water. The water-soluble oligomers were separated by precipitation from the free glycol by cooling the filtrate. The precipitated water-soluble oligomers were filtered and added to the residue remaining after the first filtration and weighed together. The difference between the initial and the final weights represents the amount of free glycol removed by water extraction.
The UP resins were prepared by reacting the glycolyzed products with a mixture of saturated and unsaturated dibasic acids at a value of the hydroxyl to carboxyl ratio of 1.1/1. The hydroxyl number of the glycolyzed product before removing the free glycol was used to determine the amount of the dibasic acid required for the polyesterification reaction. A mixture of SE and MA was used as dibasic acid partner of the esterification reaction. The former component of the acid mixture (SE) was selected to impart flexibility to the synthesized resins as a result of increasing the spacing between the double bonds incorporated by the latter (MA). The reactions were carried out in an esterification reactor by heating the reactants from room temperature to 180 [degrees] C in nitrogen atmosphere for [similar to]1.5 h. Then the temperature was held at 180 [degrees] C for 4 h and finally raised to 200 [degrees] C and maintained until the acid value reached about 30 mg KOH/g. The acid value was monitored throughout the course of the reaction. The acid value was determined by titrating the solution of the weighed quantity of resin in acetone, with [similar to]0.2N standard alcoholic KOH solution using phenolphthalein indicator. The water of the reaction was removed throughout the course of the reaction. At the end of the reaction, the temperature of the entire contents of the reactor was allowed to drop to 100 [degrees] C, and then diluted with Sty containing hydroquinone as an inhibitor.
Determination of Peak Exotherm
The curing exotherms of unsaturated polyesters were measured by Digitron digital differential thermometer, type K, model 3202 with a resolution of 0.1 [degrees] C, as described elsewhere (8, 9). The formulation used for curing was resin/initiator/accelerator = 100/2/0.2 (parts by weight) (1).
Preparation of Polymer Concrete (PC)
There are no standard tests that are directly applicable to PC specimens. However, in the evaluation of mechanical properties of PC, an ASTM standard applicable to cement-based materials was used as guideline.
Different types of unsaturated polyester resins were used for preparing the cylindrical specimens. The PC mix design was optimized for workability, strength, and economy. The aggregate composition was 50% 10-mm pea gravel, 35 wt% sand, and 15 wt% fly ash (fly ash makes the mix both more workable and stronger). The aggregate-to-resin ratio was 9:1. MEKP initiator and CO activator were added to the resin immediately prior to mixing. Mixing of the aggregate with the activated resin was done for a period of [similar to]3 min. Specimens were then cast in a steel mold under mild pressure to maintain the desired height of the cylinder. The specimens were allowed to cure at room temperature. The age at testing was ten days to assure complete curing.
The compressive stress-strain diagrams were performed using a Zwick mechanical testing machine as described elsewhere (10, 11). A constant loading rate of 44500 N/min was applied. The specimens were cylinders of 25-mm diameter and 50-mm height.
RESULTS AND DISCUSSION
Preparation and Characterization of the Glycolyzed Products
Eleven glycolyzed products, coded G-1-G-11, were obtained by reacting PET with PG/DG mixtures having weight fractions of PG equal to 1.0, 0.8, 0.75, 0.7, 0.65, 0.60, 0.55, 0.50, 0.45, 0.40 and 0.35. The ratio of glycol mixture to PET waste was kept constant (0.65) in all depolymerization reactions of PET with different glycol mixtures. The DG was utilized among the glycol systems to impart flexibility to the final cured resin. The PG was combined with DG to increase the miscibility of the synthesized resins with styrene monomer (12). We presume that the incorporation of the hydroxy propylene enhances the solubility of the polymers based on these glycolyzed products in styrene monomer.
The free glycol in each of these glycolyzed products was measured according to the procedure described before. The resultant data for the glycolyzed products are tabulated in Table 1 in conjunction with the hydroxyl number before and after removing the free glycol. Mindful inspection of the measured hydroxyl numbers before removing the free glycols indicates that only 2-4% of the glycols are used up in the depolymerization. This finding agrees with that found in our previous publication (13). The hydroxyl numbers after removal of the free glycol show that the extent of depolymerization decreases with a decreasing amount of PG in the glycol mixture. These numbers reveal that there is a gradual decrease from 332 mg KOH/g to 318 mg KOH/g with a decrease in the weight fraction of PG in the glycol mixture from 1.0 to 0.35.
Table 1. Characterization of Glycolized PET Waste. Weight Fraction Free Glycol, -OH Value -OH Value Oligomer of PG % (1) (2) G-1 1.00 61.0 962 332 G-2 0.80 61.4 963 330 G-3 0.75 61.5 964 329 G-4 0.70 61.6 964 328 G-5 0.65 61.8 965 326 G-6 0.60 62.0 965 325 G-7 0.55 62.3 966 323 G-8 0.50 62.5 967 322 G-9 0.45 62.6 970 320 G-10 0.40 62.8 972 319 G-11 0.35 63.0 974 318 -OH value(1) = hydroxyl value before free glycol removal. -OH value(2) = hydroxyl value after free glycol removal. Amount of glycol mixture = 65% of the weight of PET waste.
The values of the hydroxyl numbers after removal of the free glycols indicate that the extent of depolymerization is considerable and the glycolyzed products are mainly terminated with hydroxyl groups.
Synthesis of UP Resins
The oligomers G-1-G-11 were reacted with the appropriate amount of SE/MA mixture composed of 1:1 mole % to produce the UP resins, coded UP-100-UP-35, respectively. The hydroxyl number of the glycolyzed product before separation of the free glycol was used for estimating the amount of SE/MA mixture. The results of the characterization of the prepared unsaturated polyester resins are given in Table 2. In this Table the measured values of acid and hydroxyl numbers are listed in columns 3 and 4, respectively. These values were used for determining the number average molecular weight for each of the synthesized unsaturated polyesters (7). In all cases, unsaturated polyesters of number average molecular weight ranging between 1206 and 1320 were obtained. This range of molecular weights is comparable with that obtained by Vaidya and Nadkarni (6), who prepared three polyesters having molecular weights ranging between 1045 and 1325. The data quoted in Table 2 show that the molecular weight increases with increasing the DG content in the glycolyzed products.
The curing exotherms of the prepared formulae were obtained by plotting the curing temperature as a function of time. Since the amount of heat evolved upon curing depends on the sample size (13), it was desirable to consider this parameter. For this reason it was very important to use glass bottles of the same volume in all measurements to achieve the repeatability of the measurements and afford a legitimate comparative study. Some of these plots, for brevity, are illustrated in Fig. 1. The maximum curing temperatures, [T.sub.max], and the times, [t.sub.max], required to reach these temperatures were obtained from these plots and tabulated in Table 3.
Table 2. Characterization of the UP Resins Obtained From the PET Waste. Glycolized Carboxyl Hydroxyl Molecular Product Value (mg Value (mg Weight Polyester Used KOH/g) KOH/g) <[M.sub.n]> UP-100 G-1 30 63 1206 UP-80 G-2 32 60 1219 UP-75 G-3 30 62 1219 UP-70 G-4 32 60 1219 UP-65 G-5 31 60 1233 UP-60 G-6 30 59 1260 UP-55 G-7 31 58 1260 UP-50 G-8 30 58 1275 UP-45 G-9 28 60 1275 UP-40 G-10 29 58 1289 UP-35 G-11 29 56 1320
In our previous publication (9), a series of UP resins based on phthalic anhydride, MA, and different glycol mixtures was prepared. The structure of formula No. 5a in this series is almost similar to the structure of UP resins prepared in the present study. The UP chains in the former formula contain o-phthalate group instead of the ethylene terephthalate group existing in the present formulae. On the other hand, another difference between formula No. 5a and the formulae of the resins under investigation is the inclusion of SE as unsaturation spacer. The characteristics of formula 5a are shown in the last row in Table 3. Comparing the curing characteristics of these types of formulae reveals that the incorporation of sebacic molecule within the resin chain and the existence of terephthalate group increase the curing reaction time by 24% to 50% and reduce the heat evolved during the curing by 3.5% to 17%. This may be attributed to the increase in the spacing between the curable double bonds. This increase in the spacing is afforded by the extra -C[H.sub.2]-C[H.sub.2]- moiety existing in the recycled formulae and the -[(C[H.sub.2]).sub.8]- component consolidated by SE as well as the position of the two ester groups in terephthalate moiety. The use of SE in conjunction with MA for preparing the present unsaturated polyester from the glycolyzed products of PET leads to an inclusion of -[(C[H.sub.2]).sub.8]- as well as a reduction of the number of the polymerizable double bonds. This certainly leads to an increase in curing time and a decrease in curing temperature. It can be seen also that [T.sub.max] decreases with increasing the content of DG, and hence an increase in [t.sub.max] is obtained. This finding contributes a great body of evidence that the structure of the UP chains affects the amount of heat liberated while curing as well as the curing time.
Preparation and Characterization of PC
The PC samples were prepared from the UP resins obtained from recycling of PET. The molding technique and the equipment used for casting the PC specimens were described in detail in the Experimental section. The PC mix design was optimized for workability, strength, and economy, as suggested by Rebeiz et al. (2). The optimum aggregate composition of 50% 10-mm pea gravel, 35% sand, and 15% fly ash (fly ash makes the mix more workable and stronger). The optimum aggregate-to-resin ratio was 9:1.
The measured values of compressive strength ([[Sigma].sub.u]) and Young's modulus (E), according to ASTM method [TABULAR DATA FOR TABLE 3 OMITTED] (D 695 - 44 T), of the PC based on the recycled resins are shown in Table 3. This Table shows that [[Sigma].sub.u] decreases from 110.2 to 89.2 MPa with an increase in the weight fraction of DG from 0.0 to 0.65, while the corresponding values of E decrease from 19.7 to 5.8 GPa. The range of the measured values of [[Sigma].sub.u] for PCs is slightly higher than the value obtained by Rebeiz et al. (91.8 MPa) (14) for a PC based on a commercial recycled UP from PET. On the other hand, the same authors reported (2) values of [[Sigma].sub.u] ranging between 42.1 and 87.3 MPa for six different PCs based on different commercial patterns of recycled UP resins. The difference between the present values of [[Sigma].sub.u] and that reported by Rebeiz et al. (2, 14) may be attributed to the difference in molecular weights of the resins. Industrial preparation of UP resins in large batches certainly leads to a wider molecular weight distribution than is found in UP resins prepared on a laboratory scale. A further reason that may influence the [[Sigma].sub.u] of the cured specimen is the procedure used for shaping this specimen. It is well known that the presence of voids is responsible for introducing weak points in the molded specimens. Rebeiz et al. (2, 14) used an ordinary casting procedure. In this work, the casting was performed under a tender pressure. This pressure certainly leads to a very low void content. This reduction in the void content enhances the mechanical properties. For the same PC specimens, Rebeiz et al. (2) reported Young's moduli ranging between 10.3 and 28.4 GPa. The lower values of E (higher flexibility) obtained in the present investigation seem due to the incorporation of DG and SE molecules in the UP chains. This enhanced flexibility may be attributed to the increase of the length of the repeating units in the UP chains, which results in a better local freedom of motion of these chains and hence increases the flexibility (15).
PCs made from virgin materials (2) were found to have compressive strengths ranging from 40 to 130 MPa. Thus, the properties of PCs made with resins using recycled PET are comparable with those using virgin resins.
The following conclusions may be drawn from these results:
1) The PET waste could be depolymerized by glycolysis with PG-DG mixtures.
2) The extent of depolymerization increases with increasing amount of PG.
3) The existence of a terephthalate group increases the curing reaction time and reduces the heat evolved during the curing.
4) A considerable increase in the cure reaction time is attained when SE is used in conjunction with MA for preparing the UP from the glycolyzed products of PET.
5) The reduction of the number of the polymerizable double bonds as a result of using SE in conjunction with MA leads to a decrease in the maximum curing temperature.
6) The structure of the UP chains affects the amount of heat liberated while curing as well as the curing time.
7) The UP resins designed from recycled PET may be used for making PC.
8) Embodiment of DG in the glycolyzed products and SE in the prepared resins increases the flexibility of the cured PC.
1. E. S. Nasr and A.-A. A. Abdel-Azim, Polym. Advanc. Technol., 3, 407 (1992).
2. K. S. Rebeiz, D. W. Fowler, and D. R. Paul, Plastics Engineering, February 1991, p. 33.
3. U. R. Vaidya and V. M. Nadkarni, Ind. Eng. Chem. Res., 26, 194 (1987).
4. K. S. Rebeiz, D. W. Fowler, and D. R. Paul, Proc. Third Intern. Symp. Mechanics of Polymer Composites, p. 367, Institute of Theoretical and Applied Mechanics, Prague (April 1991).
5. K. S. Rebeiz and D. W. Fowler, Polym.-Plast. Technol. Eng., 30, 809 (1991).
6. U. R. Vaidya and V. M. Nadkarni, J. Appl. Polym. Sci., 34, 235 (1987).
7. W. R. Sorenson and T. W. Campbell, Preparative Method of Polymer Chemistry, p. 155, Interscience, New York (1968).
8. A.-A. A. Abdel-Azim, I. A. Attia, and E. S. Nasr, Polym-Plast. Technol. Eng., 34, 79 (1995).
9. A.-A. A. Abdel-Azim, B. H. Mahmoud, and M. S. Farhat, Polym. Advanc. Technol., 5, 269 (1994).
10. A.-A. A. Abdel-Azim, E. S. Nasr, and M. S. Farhat, Polym. J., 26, 423 (1993).
11. A.-A. A. Abdel-Azim, Polym. Bull. (Germany), 35, 229 (1995).
12. S. N. Tong, D. S. Chen, C. C. Chen, and L. Z. Chung, Polymer, 24, 469 (1983).
13. A.-A. A. Abdel-Azim and I. A. Attia, Polym. Adv. Technol., 6, 688 (1995).
14. K. S. Rebeiz, D. W. Fowler, and D. R. Paul, J. Appl. Polym. Sci., 44, 1649 (1992).
15. P. Munk, Introduction to Macromolecular Science, Wiley, New York (1989).
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||International Forum On Polymers - 1996|
|Author:||Abdel-Azim, Abdel-Azim A.|
|Publication:||Polymer Engineering and Science|
|Date:||Dec 1, 1996|
|Previous Article:||Characterization of spherical microlatex particles made of linear uncrosslinked polystyrene chains.|
|Next Article:||Physical aging in polymers: comparison of two ways of determining Narayanaswamy's parameter.|