J Mater Cycles Waste Manag DOI 10.1007/s10163-014-0322-2
SPECIAL FEATURE: ORIGINAL ARTICLE
Chemical Feedstock Recycling 10
Pt/ITQ-6 zeolite as a bifunctional catalyst for hydrocracking of waste plastics containing polystyrene E. G. Fuentes-Ordo´n˜ez • J. A. Salbidegoitia • M. P. Gonza´lez-Marcos • J. L. Ayastuy • M. A. Gutie´rrez-Ortiz • J. R. Gonza´lez-Velasco
Received: 13 December 2013 / Accepted: 1 October 2014 Ó Springer Japan 2014
Abstract Thermal and catalytic hydrocracking of polystyrene to fuels was compared. The use of a bifunctional (platinum and acidic sites) catalyst such as Pt/Ferrierite not only increases conversion but also selectivity to a wider and more interesting variety of products in the gasoline range (C5–C12). As polymer molecules present steric hindrance to access internal active sites in the catalyst, Pt/ITQ6 was prepared by delamination to maximize the external surface of the catalyst while keeping its composition and type. Although Pt/ITQ-6 presented lower acidity than Pt/ Ferrierite, it was mostly external and, thus, accessible to the reactants. In this way, Pt/ITQ-6 significantly improved activity and selectivity of Pt/Ferrierite. The performance of Pt/ITQ-6 when recycled polystyrene was used as reactant proved this catalyst is very promising for this application, although catalytic activity decreased as a consequence of plastic additives and impurities. Keywords Plastic waste Polystyrene Hydrocracking Bifunctional catalyst Zeolites Delaminated
Introduction One of the consequences of economic development has been the incredible rise in plastic consumption around the world. In only six decades, production of these materials E. G. Fuentes-Ordo´n˜ez J. A. Salbidegoitia M. P. Gonza´lez-Marcos (&) J. L. Ayastuy M. A. Gutie´rrez-Ortiz J. R. Gonza´lez-Velasco Group of Chemical Technologies for Environmental Sustainability, Department of Chemical Engineering, Faculty of Science and Technology, The University of the Basque Country, UPV/EHU, P.O. Box 644, 48080 Bilbao, Spain e-mail:
[email protected]
has increased from 1.7 until 288 million tons. This dramatic increment, together with some plastic characteristics, such as low biodegradability, low density and production of harmful compounds during degradation, has made of plastic wastes a serious environmental problem [1]. Catalysis represents both a huge opportunity and a challenge for the conversion of waste plastics, as catalytic processes ensure lower operating temperature and pressure, and higher conversion and selectivity to the target products [2]. Catalysts have been studied in processes including chemical recycling and energetic recovery strategies, such as cracking [3–10], hydrocracking [11–15] and gasification [16–19]. In some of these processes, specifically cracking and hydrocracking, zeolites have demonstrated to be effective for waste plastics degradation. Additionally, zeolites are materials whose properties can be tuned according to reaction requirements, an important aspect to take into account to minimize the limitations on mass and heat transport due to the huge size of polymer macromolecules and their high viscosity during catalytic polymer degradation in a fluid phase [20, 21]. Several strategies have been studied to improve conversion in plastic degradation processes with catalysts based on zeolites, including their substitution by mesoporous ordered materials with large pores and a narrow distribution, such as Al-MCM-41 and SBA-15 in polyethylene and polystyrene cracking [22–24] and polyethylene hydrocracking [25], and by nanocrystalline (n-) or hierarchical (h-) zeolites with high external surface area, such as n-HZSM-5, n-Beta and h–HZSM–5 in polyethylene cracking [26–30]. The activity of the mesoporous catalysts improves for virgin polymer degradation in comparison with microporous materials, but it greatly decreases for waste post-consumer plastics degradation. Concerning catalysts based on nanocrystalline and hierarchical zeolites,
123
J Mater Cycles Waste Manag
the results show an improvement in the catalytic activity in both virgin and post-consumer waste plastics with respect to microporous and mesoporous ordered materials. In this paper, we study the catalytic behavior of a delaminated zeolite, ITQ-6, as support of bifunctional catalysts for polystyrene hydrocracking, pure and in mixtures with other plastics, to fuels in the gasoline range of molecular weight (C5–C12), in the liquid phase in a single stage. The use of delaminated zeolites is another strategy for increasing external surface area thus improving the access of polymer macromolecules to the active sites in the catalyst surface [31]. Although there are no studies in application of delaminated zeolites to plastics degradation processes, there are promising results of delaminated zeolites as supports of catalysts in the field of hydrotreating and hydrocracking of long-chain hydrocarbons [32, 33]. With respect to metallic sites, necessary in some hydrocracking reactions, platinum was chosen because noble metals, and especially platinum, are known to present the highest catalytic activity for this kind of reactions [34]. As we are working in catalyst design, the reaction conditions have been chosen so that no mass-transfer control occurs, although actual industrial processes should operate with mass-transfer limitations.
Experimental Catalysts preparation Zeolites were synthesized following indications of Corma and coworkers [31, 33]. First, the precursor material (PREFER) is prepared by mixing 10 g of silica, 2.3 g of alumina, 9.2 g of NH4F, 3.1 g of HF 49.8 % concentration, 26 g of 4-amino-2,2,6,6-tetramethylpiperidine, and 27.9 g of deionized water in an autoclave at 448 K for 5 days. The resulting product (PREFER) was filtered, washed and dried at 333 K. Then, ITQ-6 was obtained by swelling and subsequent delamination of the precursor, as follows: the PREFER was swollen in a water solution of cetyltrimethylammonium bromide (CTABr) and tetrapropylammonium hydroxide (TPAOH) and refluxed for 36 h at 353 K and, in the following step, the mixture was delaminated by ultrasonic treatment for 1 h. Finally, the solid phase was separated by centrifugation and washed, dried at 333 K, and calcined at 853 K for 7 h, yielding ITQ-6. For comparison purpose, Ferrierite was prepared by direct calcination of PREFER at 853 K. Bifunctional catalysts, to a metal content of 0.5 wt% Pt, were prepared by ionic exchange with each zeolitic support (ITQ-6 and Ferrierite), using tetraammineplatinum(II) nitrate as the metallic source. The exchanges was carried out under mixing and reflux (353 K, 24 h, pH 7).
123
Afterwards, the solid was filtered, washed, dried (383 K, 12 h), calcined (0.5 K min-1 up to 623 K, 2 h at 623 K) and reduced in hydrogen flow (0.5 K min-1 up to 723 K, 3 h at 723 K) in a tubular reactor. The catalysts have been coded as: Pt/Ferrierite and Pt/ITQ-6. Catalysts characterization Structure of synthesized materials was determined by Xray diffraction (XRD). The X-ray powder diffraction patterns were collected by using a PHILIPS X’PERT PRO automatic diffractometer operating at 40 kV and 40 mA, in theta-theta configuration, secondary monochromator with ˚ ) and a PIXcel solid state Cu-Ka radiation (k = 1.5418 A detector (active length in 2h = 3.347°). Data were collected from 5° to 50° 2h (step size = 0.026 and time per step = 150 s) at room temperature. A fixed divergence and anti-scattering slit giving a constant volume of sample illumination were used. Morphology of the prepared zeolites was studied by scanning electron microscopy (SEM) using a Schottky (JEOL JSM-7000F) equipment with a resolution of 20 kV. Surface area, external surface, and pore size distribution of the prepared catalysts, in the micropore and mesopore range, were determined by N2 physisorption at 77 K, using the conventional BET, t-plot, Horvath-Kawazoe and BJH methods, in a Micromeritics ASAP 2020 instrument. The sample was degassed (\1.5 Pa) at 573 K for 12 h and, afterwards, the isotherm was obtained. Actual platinum loading in the catalysts was determined by inductively coupled plasma (ICP), using a Varian (710ES ICP-Optical in radial position) equipment. An adequate amount of catalyst sample to obtain metal concentration in the linear range of measurement at 214.424 nm, was dissolved in nitro-hydrochloric acid and hydrofluoric acid, and heated until evaporation of the acids. Then, the solution was kept in an aqueous solution with 2 % nitric acid. Temperature-programmed desorption of ammonia (NH3-TPD) was performed to determine total acidity of the catalysts, using a Micromeritics AutoChem II equipped with a thermal conductivity detector (TCD). The sample was pre-treated in situ with 5 % H2 in Ar flow from room temperature to 823 K at 20 K min-1 and holding 30 min at 823 K. Then, it was cooled to 373 K under He flow. The adsorption of NH3 was carried out by pulses of 10 % NH3 in He until saturation, followed by He flow. NH3 desorption was measured while the temperature was raised from 373 to 823 K at 10 K min-1. Fourier transform infrared spectroscopy (FTIR) of adsorbed pyridine and adsorbed 2,6-di-tert-butylpyridine (DTBPy) was carried out to determine the distribution of Lewis and Brønsted sites and the external acidity of the catalysts, respectively, using a Nicolet Prote´ge´ 460
J Mater Cycles Waste Manag
Spectrometer equipped with vacuum cell and temperature controller Specac. The measurements were carried out over self-supported wafers previously treated at 623 K and high vacuum overnight. FTIR spectra were recorded at 423 K, after saturation of the samples by pulses of the respective probe molecules and cleaning with high vacuum. Metallic dispersion was evaluated by H2 chemisorption in a Micromeritics ASAP 2020C instrument, at 308 K. First, the sample was degassed (\1.5 Pa) for 12 h at 573 K, and then re-reduced in H2 flow at 623 K for 2 h. Afterwards, hydrogen was removed by a vacuum treatment (\1.5 Pa) at 633 K for 1.5 h and then the sample was cooled down to 308 K. Two successive isotherms were carried out, with intermediate evacuation (\1.5 Pa, 308 K, 30 min), and subtraction of the latter to the former allowed for evaluation of the irreversibly chemisorbed hydrogen. Metal cluster size in the reduced catalysts was calculated from the high-resolution images of transmission electron microscopy (TEM) technique. The equipment used was a Philips CM200 200 kV with LaB6 filament and energy dispersive spectrometer (EDS). Finally, the chemical state of platinum was studied by FTIR of adsorbed CO. The procedure and equipment used was the one already explained above for acidity measurements, and the spectra were recorded at 403 K. Polymer characterization Glass transition and melting temperatures of virgin and recycled polystyrene (PS) were measured by differential scanning calorimetry (DCS), using a Mettler Toledo DSC822e calorimeter within the range 263 to 573 K at a heating rate of 10 K min-1 under a continuous N2 flow. H, C, N and O content of virgin and recycled PS has been measured by elemental analysis, using an EuroVector Euro EA Elemental Analyzer (CHNS), with the combustion chamber operating around 1,293 K. The TCD signal for each element is translated to percentage content. FTIR of plastic materials was carried out in the same experimental equipment described above. The spectra were recorded at room temperature over plastic film samples supported on KBr wafers.
Virgin PS with an average molecular weight of 192,000 g mol-1 was supplied by Aldrich. Recycled PS, supplied as black pellets (Gaiker-IK4), was also used for comparison. Before reaction, the PS was milled to an average particle size of 1 mm in a mill (Retsh ZM 200) refrigerated with liquid nitrogen. The reactions were performed with 5 wt% of PS solved in decahydronaphthalene (DHN; Sigma-Aldrich) and catalyst in a slurry (2.36 gCat. L-1; average particle size 230 lm). The reaction mixture was introduced in the reactor, which was then closed and heated to operation temperature (598–698 K). At this point (t = 0), the system was pressurized under H2 (180 bar) and stirring (1,800 rpm) was started. The system was run for 40 min, and samples for analysis were taken at specific intervals. Finally, stirring was stopped and the reactor was allowed to cool down to room temperature. The degree of conversion of PS was evaluated by gel permeation chromatography (GPC, Waters 616) in line with Ultraviolet (Acquity) and Refraction Index (Waters 2410) detectors, by following PS molecular weight distribution and concentration in solution, after calibration. The GPC was equipped with Styragel HR1 and HR4 columns placed in series in an oven set at 308 K, to cover the molecular weight range of 100–500,000 g mol-1. Tetrahydrofuran (THF; 1 mL min-1) fed by a Waters 515 HPLC pump was used as mobile phase. To evaluate the degree of PS conversion, two calibrations were necessary. First, PS samples of different weight-average molecular weight (MW) were used for calibration of retention times in the GPC. A straight line is obtained when log(MW) is represented versus retention time, which allows for definition of the integration limits in the GPC. Then, a second calibration for concentration was carried out, by preparing solutions of PS in DHN of known concentration. Integrated area in the GPC is represented versus PS concentration, and a straight line is obtained from which concentration of polymer can be derived. Polystyrene is considered unconverted when its MW is above that corresponding to trimers. PS conversion (X) was calculated as: X ¼1
Catalytic activity The experimental system used for hydrocracking consists of a reactor (Autoclave Engineers, Hastelloy C-276), with a capacity of 300 mL, equipped with: a high speed rotary stirrer operated by a magnetic system; temperature, pressure, inlet and outlet flow controllers; inlet and outlet pipes; and a reflux condenser.
CPS CPS0
ð1Þ
where CPS0 represents the initial concentration of PS in the reactor, and CPS its concentration in the sample. Liquid and gas products were also analyzed by gas chromatography-flame ionization detector (GC-FID, Agilent 6890N), with PONA column (HP) 50 m 9 0.2 mm 9 0.5 lm. 1.5 mL min-1 of hydrogen was used as carrier gas to determine catalyst selectivity.
123
J Mater Cycles Waste Manag
Results and discussion Catalysts characterization Structural and textural properties of catalysts Figure 1 shows XRD patterns of the synthesized layered precursor (PREFER) and the final supports. Their patterns confirm the structure of the synthesized materials [31]. When the patterns for Ferrierite and ITQ-6 are compared in Fig. 1, the intensity of (0 k l) reflection planes remains unchanged in both structures, whereas that corresponding to (h 0 0) planes presents a significant decrease in ITQ-6 compared to Ferrierite, which indicates the loss of order along the a axis in ITQ-6 as a result of delamination. Figure 2 shows the SEM images of the zeolites. Comparing the images for Ferrierite and ITQ-6, changes in particle morphology due to the delamination process can be clearly seen; furthermore, calcination is observed to produce large particles in Ferrierite, whereas delamination produces an effect of size and thickness reduction to produce ITQ-6.
N2 isotherms at 77 K of the catalysts are presented in Fig. 3, where Pt/Ferrierite can be seen to present type I isotherms with a small H3 hysteresis loop, characteristic of microporous materials, whereas Pt/ITQ-6 presents isotherms with superposition of typical micro-, meso- and macroporous effects. Textural properties of the prepared catalysts are summarized in Table 1. Pt/ITQ-6 presents much higher external surface area and mesopore volume than Pt/Ferrierite, which can be associated to delamination. Figure 4 shows their pore size distribution for both micro- and mesopores. The average micropore sizes in Table 1, determined by Horvath-Kawazoe, are basically the same in both catalysts because the channels must be identical. On the contrary, a considerable change can be observed in the mesopore size distribution calculated by BJH in Fig. 4. A significant increase in the average mesopore size can be observed in Pt/ITQ-6 compared to Pt/Ferrierite, which can be attributed to interparticle condensation. Figure 5 shows the NH3-TPD profiles obtained for Pt/ Ferrierite and Pt/ITQ-6. Integration of areas under the experimental curves determines total acidity of the catalysts, which has been presented in Table 2 for Pt/Ferrierite and Pt/ITQ-6. Total acidity of Pt/Ferrierite can be observed to be much higher than that of Pt/ITQ-6. Although both zeolites were prepared from the same precursor and Pt/ ITQ-6 presents higher specific surface area according to Table 1, about 60 % of total acidity is lost in ITQ-6 synthesis from PREFER due to dealumination and destruction of acidic sites during delamination. FTIR spectra of pyridine adsorbed on the catalysts are shown in Fig. 6. Through this method, total acidity can be discriminated as Brønsted or Lewis type. Following the Emeis procedure [35], each acidic type has been quantified and the results are shown in Table 2, together with the Brønsted to Lewis ratio (B/L). The results show that, although the acidity loss caused by delamination occurs in both Brønsted and Lewis acid sites, Lewis acidity
Fig. 1 XRD patterns of synthesized materials: a PREFER, b ferrierite, c ITQ-6
Fig. 2 SEM images of synthesized zeolites: a ferrierite, b ITQ-6
123
Fig. 3 N2 adsorption isotherms of the prepared catalysts. Symbols: open circle Pt/Ferrierite; open square Pt/ITQ-6
J Mater Cycles Waste Manag Table 1 Chemical composition and textural properties of the prepared catalysts
Si/ Al
SBET, m2 g-1 Cat.
SExt., m2 g-1 Cat.
VMicro, cm3 g-1 Cat.
VMeso, cm3 g-1 Cat.
dMicro, nm
dMeso, nm
Pt/Ferrierite
60
339
158
0.087
0.143
0.522
5.71
Pt/ITQ-6
60
460
436
0.007
1.255
0.680
9.80
Fig. 4 Pore size distribution of the prepared catalysts: a Pt/Ferrierite, b Pt/ITQ-6
decreases much more (about 71 %) than Brønsted acidity (about 45 %), comparing Pt/ITQ-6 and Pt/Ferrierite, that is in agreement with a dealumination occurring during delamination. Because of this, B/L ratio is higher in Pt/ ITQ-6 than in Pt/Ferrierite. Due to the large size of DTBPy molecules, they cannot enter narrow micropores. Thus, FTIR studies with this probe molecule provide information related to acidity of external surface sites. Figure 7 shows the FTIR spectra of adsorbed DTBPy on Pt/Ferrierite and Pt/ITQ-6. Intensity of peaks at 3,370 and 1,670 cm-1 in Fig. 7 can be observed to be higher in delaminated Pt/ITQ-6 compared to Pt/Ferrierite, which means that although Pt/ITQ-6 presents less acidic sites, as shown above, their accessibility is higher. According to the procedure developed by Corma and coworkers [36], the fraction of acid sites accessible to DTBPy in each catalyst was estimated, and has been included in Table 2 as external acidity. Accessibility to the acidic sites in Pt/ITQ-6 amounts to around 90 % whereas the same property is almost insignificant in Pt/Ferrierite. Metallic characterization
Fig. 5 NH3-TPD profiles of the prepared catalysts: a Pt/Ferrierite, b Pt/ITQ-6
Actual platinum content for both catalysts is shown in Table 3. Platinum contents are near the theoretical 0.5 wt%. On the other hand, metal dispersion calculated by H2 chemisorption is significantly higher in Pt/ITQ-6 than in Pt/Ferrierite. To a certain extent, this is probably related to the higher surface area and better surface accessibility of ITQ-6 compared to Ferrierite. Table 3 also shows the average platinum size estimated from hydrogen chemisorption for both catalysts. Additionally, Fig. 8 shows TEM images obtained for both catalysts, where the small platinum nanoparticles supported
Table 2 Acidic properties of the prepared catalysts Total aciditya, lmolNH3 g1 Cat:
Brønstedb, lmolPy g-1 Cat.
Lewisb, lmolPy g-1 Cat.
B/L
Pt/Ferrierite
422
76
98
0.8
2.9
Pt/ITQ-6
162
42
28
1.5
93.5
a
External acidityc, %
Calculated by integration of NH3-TPD profiles
b
Obtained from FTIR spectra of adsorbed pyridine
c
Estimated by FTIR spectra of adsorbed DTBPy
123
J Mater Cycles Waste Manag
Fig. 6 FTIR spectra of adsorbed pyridine at 423 K: a Pt/Ferrierite, b Pt/ITQ-6
Fig. 8 TEM images and platinum particle size distributions of the prepared catalysts: a, c Pt/Ferrierite; b, d Pt/ITQ-6
Fig. 7 FTIR spectra of adsorbed DTBPy at 423 K: a Pt/Ferrierite, b Pt/ITQ-6
Table 3 Characterization of the supported platinum particles in the prepared catalysts Platinum, wt%
Pt dispersiona, %
daPt, nm
dbPt, nm
Pt/Ferrierite
0.50
54
2.0
2.1
Pt/ITQ-6
0.42
73
1.5
1.6
a
Estimated by hydrogen chemisorption
b
Estimated by TEM
on the prepared catalysts can be seen. The TEM images were used to determine the platinum particle size distributions that are also shown in Fig. 8. Particles between 1 and 7 nm are observed and the average platinum particle size estimated from TEM has been also added to Table 3. We can see that the calculated values estimated from TEM are almost identical to those estimated from chemisorptions, and confirm the higher platinum dispersion in Pt/ ITQ-6 compared to Pt/Ferrierite. Platinum species present on the reduced catalysts were studied by FTIR of adsorbed CO. Obtained FTIR spectra
123
Fig. 9 FTIR spectra of CO adsorbed on the platinum surface of the reduced catalysts: a Pt/Ferrierite, b Pt/ITQ-6
are shown in Fig. 9. All samples are shown to present a band between 2,080 and 2,099 cm-1, associated to linearly bonded CO on metallic Pt sites [37]. This result indicates that platinum is mostly present as reduced species in the prepared catalysts. Polystyrene characterization As commented above, both virgin and recycled PS was submitted to hydrocracking. Thus, both polymers were characterized to obtain information for an adequate interpretation of the reaction results. Table 4 shows the results for average molecular weights and polydispersivity in both PS samples, as determined by GPC. Recycled polystyrene used in this work can be seen to present a weight-average
J Mater Cycles Waste Manag Table 4 Molecular weight distribution of polystyrene samples and H/C ratio MW, g mol-1
Mn, g mol-1
Polydispersivity
H/C ratio
Virgin PS
198,229
120,018
1.65
0.95
Recycled PS
183,647
75,629
2.42
1.07
Fig. 11 DSC analysis of: a virgin PS, b recycled PS
Fig. 10 FTIR spectra of: a virgin PS, b reference PS, c reference PP, d reference HDPE, e reference LDPE, f recycled PS, g extracted from recycled PS by dissolution in DHN at room temperature
molecular weight (MW) only slightly smaller than that of virgin polystyrene, much lower number-average molecular weight (Mn) and, thus, higher polydispersivity. Table 4 also includes results for H/C ratio from elemental analysis of both polymers. Recycled PS also presents traces of N and O. FTIR spectra of virgin and recycled PS used in the hydrocracking reactions are shown in Fig. 10. Additional database spectra corresponding to: PS, where the characteristic peaks at 700, 750 and 3070 cm-1 corresponding to C–H bonds in mono-substituted benzene, and those at 1450, 1500, 1580 and 1600 cm-1 corresponding to C=C bonds in an aromatic ring can be observed; polypropylene (PP), with the characteristic peaks at 1380, 1460, 2870 and 2960 cm-1 corresponding to C–H bonds in methyl groups and those at 1470, 2850 and 2925 cm-1 corresponding to C–H bonds in methylene groups; high-density (HDPE) and low-density polyethylene (LDPE), also with their characteristic peaks at 1470, 2850 and 2925 cm-1 corresponding to C–H bonds in methylene groups; have been added to Fig. 10 for comparison.
We can see in Fig. 10 that spectra for virgin PS [curve (a)] and reference PS [curve (b)] are in good agreement. Curve (f) in Fig. 10, corresponding to recycled PS, presents a poor quality due to the difficulty to prepare a good sample for measurement, although it shows the typical peaks for PS. An additional sample, corresponding to curve (g) in Fig. 10 and named as extracted PS, was obtained by partially solving recycled PS in DHN at room temperature. We can see that extracted PS also presents good agreement with reference PS spectra. FTIR of recycled PS confirms the presence of PS in the recycled matrix, together with some additives and impurities. To identify the nature of, at least, some of the additives in recycled PS, Fig. 11 shows the DSC curves of virgin and recycled PS. The curve corresponding to virgin PS [curve (a)] presents one thermal event at 378.4 K, associated to PS glass transition temperature. The curve corresponding to recycled PS [curve (b)] presents multiple thermal events, at 369, 403.8, 435.9 and 532 K. The first event can be associated to the glass transition temperature of PS, whereas the second and third events can be associated to the melting temperatures of HDPE and PP, respectively, and the latter to polybutadiene crosslinking reactions. Therefore, recycled PS can be concluded to actually correspond to a mixture of high impact polystyrene (HIPS) with traces of, at least, HDPE and PP. Polystyrene hydrocracking Degradation of PS molecules due to hydrocracking can be followed by two parameters: conversion to products (low molecular weight hydrocarbons) through end-chain reactions, and residual PS Mn, mainly due to chain random scission reactions. No indication of coke formation was observed during the experiments. Figure 12 shows the dependence of both parameters with temperature after 40 min of reaction, for thermal and catalytic processes, comparing the performance of Pt/Ferrierite and Pt/ITQ-6; while Fig. 13 shows selectivity to different hydrocarbons.
123
J Mater Cycles Waste Manag
Fig. 12 PS conversion and residual number-average molecular weight (Mn) after 40 min of hydrocracking. 5 wt% of PS in DHN, 18 MPa of H2, 1,800 rpm and 2.36 gCat L-1. Symbols: filled diamond thermal; filled square Pt/Ferrierite; filled circle Pt/ITQ-6
We can see there is a significant conversion of PS through the thermal process, in the absence of catalyst. The presence of Pt/Ferrierite improves conversion only slightly, but Pt/ITQ-6 produces a very significant increase in conversion. Concerning Mn, the number-average molecular weight of residual PS significantly decreases by the presence of Pt/Ferrierite compared to the thermal process, and much more when Pt/ITQ-6 is used. All this indicates that the use of a catalyst significantly improves PS degradation. The higher activity of Pt/ITQ-6 compared to Pt/Ferrierite, both with the same composition, is associated to the better accessibility of PS macromolecules to the active sites in Pt/ITQ-6. As demonstrated above, although Pt/Ferrierite presented more acidic sites (associated to cracking reactions) than Pt/ITQ-6, accessibility of the acidic sites to big molecules such as those of PS is much higher in Pt/ITQ-6. Concerning metallic sites, platinum dispersion in Pt/ITQ-6 has been found to be higher than that in Pt/Ferrierite. In any case, even if dispersions were similar, platinum accessibility would be higher in Pt/ITQ-6 than in Pt/Ferrierite. However, the most important effect of using a catalyst and, more specifically, a bifunctional catalyst with metallic active sites, is associated to selectivity, as shown in Fig. 13. We can see that, in the absence of catalyst (thermal process), only aromatics are obtained as products (monomers, dimers and trimers) even though there is a high hydrogen pressure, as the presence of metallic sites is related to reactions where hydrogen is involved.
123
Fig. 13 Selectivity by lumps in PS hydrocracking, for the experimental conditions in Fig. 12: a thermal, b Pt/Ferrierite, c Pt/ITQ-6. Bar colors: white bar 598 K; light gray bar 623 K; medium gray bar 648 K; dark gray bar, 673 K
When Pt/Ferrierite and Pt/ITQ-6 catalysts are used, not only the yield to oligomers decreases, but hydrocarbons other than aromatics are also formed through cracking, hydrogenation-dehydrogenation, isomerization and ringopening reactions. Selectivity in Fig. 13 has been presented by grouping hydrocarbons produced through the different reactions in lumps: paraffins, isoparaffins, olefins (negligible), naphthenics and aromatics. We can see the wide variety of products obtained when Pt/Ferrierite and Pt/ITQ6 are used compared to the thermal process. Once aromatic monomers are formed through PS cracking, they are transformed into naphthenics on metallic sites and, then, on metallic and acidic sites, into paraffins and isoparaffins. Comparing the catalysts at different temperatures, temperature and accessibility of the active sites in Pt/ITQ-6 (compared to Pt/Ferrierite) favors
J Mater Cycles Waste Manag
Fig. 14 PS conversion and residual number-average molecular weight (Mn) after 40 min of hydrocracking on Pt/ITQ-6. 5 wt% PS in DHN, 18 MPa of H2, 1,800 rpm and 2.36 gCat L-1. Symbols: filled gray circle, virgin PS; filled square, recycled PS
transformation of aromatics (its yield decreases) into naphthenics, isoparaffins and paraffins. Pt/ITQ-6 was also tested as catalyst in hydrocracking of recycled PS. Figure 14 shows the comparison between catalyst performance with virgin and recycled PS. We can see that conversion and particularly residual numberaverage molecular weight indicate a decrease in catalytic activity when the recycled PS is used. The presence of other plastics such as HDPE or PP, and the crosslinking effect associated to polybutadiene (forming HIPS), as shown in Fig. 11, which change the molecular shape and can affect PS accessibility to the active sites, are thought to be the main reasons for this decrease. The presence of plastic additives that could inhibit plastic degradation is, probably, an additional factor, although their exact nature or concentration in the plastics has not been identified yet.
Conclusions Hydrocracking of polystyrene to produce automotive fuels in a single step was studied both in the absence and in the presence of a bifunctional supported Pt catalyst. Although the presence of a catalyst increased conversion to hydrocarbon products in the gasoline range (C5–C12) and decreased the number-average molecular weight of the residual polymer, it was selectivity which was mostly affected. In the presence of a bifunctional supported Pt catalyst, the yield to aromatics (and oligomers) was
significantly decreased in favor of naphthenics, isoparaffins and paraffins, as the catalyst promotes hydrogenationdehydrogenation, isomerization and ring-opening reactions. Two bifunctional catalysts with the same platinum content (ca. 0.5 wt%) supported on two zeolitic materials with the same composition and structure, Ferrierite (microporous) and ITQ-6 (delaminated), were tested for comparison. The results indicated that Pt/ITQ-6 was significantly more active than Pt/Ferrierite, and favored the decrease of aromatic content in the products. Pt/ITQ-6 was found to present much higher external surface area than Pt/Ferrierite, which is a crux point when the reactant molecules are big, such as those in PS, and cannot access the active sites, acidic and metallic, at the internal pores of the catalyst. Thus, although Pt/ITQ-6 presented much lower acidity, both Brønsted and Lewis, than Pt/Ferrierite, its accessibility to the reactants was much higher, as it was mostly external acidity. Concerning metallic sites, assuming uniform platinum distribution, they were also more accessible in Pt/ITQ-6 than Pt/Ferrierite, mostly because its external surface was higher, but also because its dispersion was higher. Thus, the strategy of increasing catalyst external surface area by delamination of a microporous zeolitic support to increase catalytic activity proved to be adequate for this kind of processes, with Pt/ITQ-6. In any case, platinum content and distribution in the final catalyst, as well as acidity, should be optimized to maximize accessibility to the reactants and selectivity to the desired products. In practice, catalytic hydrocracking would be focused on recycled PS and not virgin PS. Thus, a recycled PS where some HDPE, PP and polybutadiene crosslinking was detected was used for catalytic hydrocracking on Pt/ITQ-6, for comparison. Although conversion decreased and residual number-average molecular weight increased compared to the results with virgin PS, the results were very promising. In any case, more work is needed to associate the role of additives and impurities present in plastic wastes with catalytic performance. Acknowledgments The authors wish to thank the Spanish Ministry for Science and Innovation (CTQ2010-17277), the Basque Government (GIC-IT-657-13) and the University of the Basque Country, UPV/EHU, (UFI11/39) for their financial support, and Gaiker-IK4 for the sample of recycled PS. JAS wants to thank the Basque Government for the Research Grant (BFI-2010-150).
References 1. Association of Plastics Manufacturers in Europe (PlasticsEurope). Plastics—the facts 2013. http://www.plasticseurope.org. Accessed 12 Dec 2013 2. Bhaskar T (2011) Catalysts for the conversion of waste plastics: challenges for the development of industrial processes. In:
123
J Mater Cycles Waste Manag
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
Proceedings 6th international symposium on feedstock recycling of polymeric materials (ISFR 2011). Toledo (Spain), p 9–12 Ding W, Liang J, Anderson LL (1997) Thermal and catalytic degradation of high density polyethylene and commingled postconsumer plastic waste. Fuel Process Technol 51:47–62. doi:10. 1016/S0378-3820(96)01080-6 Kim JR, Yoon JH, Park DW (2002) Catalytic recycling of the mixture of polypropylene and polystyrene. Polym Degrad Stab 76:61–67. doi:10.1016/S0141-3910(01)00266-X Lo´pez-Urionabarrenechea A, de Marco I, Caballero BM, Laresgoiti MF, Adrados A (2012) Catalytic stepwise pyrolysis of packaging plastic waste. J Anal Appl Pyrol 96:54–62. doi:10. 1016/j.jaap.2012.03.004 Na JG, Jeong BH, Chung SH, Kim SS (2006) Pyrolysis of lowdensity polyethylene using synthetic catalysts produced from fly ash. J Mater Cycles Waste Manag 8:126–132. doi:10.1007/ s10163-006-0156-7 Nishino J, Itoh M, Ishinomori T, Kubota N, Uemichi Y (2003) Development of a catalytic cracking process for converting waste plastics to petrochemicals. J Mater Cycles Waste Manag 5:89–93. doi:10.1007/s10163-003-0086-6 Obuchi E, Suyama M, Nakano K (2001) Decomposition of mixed plastics consisting of polypropylene and polyethylene terephthalate into oils over titania/silica catalysts. J Mater Cycles Waste Manag 3:88–92. doi:10.1007/s10163-000-0043-6 Sekine Y, Fujimoto K (2003) Catalytic degradation of PP with an Fe/activated carbon catalyst. J Mater Cycles Waste Manag 5:107–112. doi:10.1007/s10163-003-0091-9 Serrano DP, Aguado J, Escola JM, Garagorri E, Rodrı´guez JM, Morselli L, Palazzi G, Orsi R (2004) Feedstock recycling of agriculture plastic film wastes by catalytic cracking. Appl Catal B-Environ 49:257–265. doi:10.1016/j.apcatb.2003.12.014 Balakrishnan RK, Guria C (2007) Thermal degradation of polystyrene in the presence of hydrogen by catalyst in solution. Polym Degrad Stab 92:1583–1591. doi:10.1016/j.polymdegradstab. 2007.04.014 Ding W, Liang J, Anderson LL (1997) Hydrocracking and hydroisomerization of high-density polyethylene and waste plastic over zeolite and silica-alumina-supported Ni and Ni-Mo sulfides. Energ Fuels 11:1219–1224. doi:10.1021/ef970051q Metecan IH, Ozkan AR, Isler R, Yanik J, Saglam M, Yuksel M (2005) Naphtha derived from polyolefins. Fuel 84:619–628. doi:10.1016/j.fuel.2004.10.006 Karagoz S, Karayildirim T, Ucar S, Yuksel M, Yanik J (2003) Liquefaction of municipal waste plastics in VGO over acidic and non-acidic catalysts. Fuel 82:415–423. doi:10.1016/S00162361(02)00250-8 Rodiansono, Wega Trisunaryanti (2007) Hydrocracking of plastic waste of polypropylene over Ni-Mo supported-mixed natural zeolite (Z)-Nb2O5, Z-c-Al2O3 and Z-ZCP-50. In: Proceedings international conference on chemical science (ICCS 2007). Yogyakarta (Indonesia) He M, Xiao B, Hu Z, Liu S, Guo X, Luo S (2009) Syngas production from catalytic gasification of waste polyethylene: influence of temperature on gas yield and composition. Int J Hydrogen Energ 34:1342–1348. doi:10.1016/j.ijhydene.2008.12.023 Mastral FJ, Esperanza E, Berrueco C, Juste M, Ceamanos J (2003) Fluidized bed thermal degradation products of HDPE in an inert atmosphere and in air-nitrogen mixtures. J Anal Appl Pyrol 70:1–17. doi:10.1016/S0165-2370(02)00068-2 Park Y, Namioka T, Sakamoto S, Min TJ, Roh SA, Yoshikawa K (2010) Optimum operating conditions for a two-stage gasification process fueled by polypropylene by means of continuous reactor over ruthenium catalyst. Fuel Process Technol 91:951–957. doi:10.1016/j.fuproc.2009.10.014
123
19. Wu C, Williams PT (2010) Pyrolysis-gasification of plastics, mixed plastics and real-world plastic waste with and without NiMg-Al catalyst. Fuel 89:3022–3032. doi:10.1016/j.fuel.2010.05. 032 20. Szekely T, Varhegyi G, Till F, Szabo P, Jakab E (1987) The effects of heat and mass transport on the results of thermal decomposition studies: part 2. Polystyrene, polytetrafluoroethylene and polypropylene. J Anal Appl Pyrol 11:83–92. doi:10.1016/ 0165-2370(87)85021-0 21. Fuentes-Ordo´n˜ez EG, Salbidegoitia JA, Gonza´lez-Marcos MP, Gonza´lez-Velasco JR (2013) Transport phenomena in catalytic hydrocracking of polystyrene in solution. Ind Eng Chem Res 52:14798–14807. doi:10.1021/ie401968r 22. Aguado J, Serrano DP, Romero MD, Escola JM (1996) Catalytic conversion of polyethylene into fuels over mesoporous MCM-41. Chem Commun 6:725–726. doi:10.1039/cc9960000725 23. Jeon JK, Park YK, Kim S, Kim SS, Yim JH, Sohn JM (2007) Catalytic degradation of polyethylene by Al-MCM-41: comparison of post-synthetic metal grafting and direct sol-gel synthesis methods. J Ind Eng Chem 13:176–181 24. Serrano DP, Aguado J, Escola JM (2000) Catalytic conversion of polystyrene over HMCM-41, HZSM-5 and amorphous SiO2Al2O3: comparison with thermal cracking. Appl Catal B-Environ 25:181–189. doi:10.1016/S0926-3373(99)00130-7 25. Hesse ND, White RL (2004) Polyethylene catalytic hydrocracking by PtHZSM-5, PtHY, and PtHMCM-41. J Appl Polym Sci 92:1293–1301. doi:10.1002/app.20083 26. Lee JY, Park SM, Saha SK, Cho SJ, Seo G (2011) Liquid-phase degradation of polyethylene (PE) over MFI zeolites with mesopores: effects of the structure of PE and the characteristics of mesopores. Appl Catal B-Environ 108–109:61–71. doi:10.1016/j. apcatb.2011.08.009 27. Lee YJ, Kim JH, Kim SH, Hong SB, Seo G (2008) Nanocrystalline beta zeolite: an efficient solid acid catalyst for the liquidphase degradation of high-density polyethylene. Appl Catal B-Environ 83:160–167. doi:10.1016/j.apcatb.2008.02.013 28. Serrano DP, Aguado J, Escola JM, Rodrı´guez JM, Peral A (2010) Catalytic properties in polyolefin cracking of hierarchical nanocrystalline HZSM-5 samples prepared according to different strategies. J Catal 276:152–160. doi:10.1016/j.jcat.2010.09.008 29. Serrano DP, Aguado J, Escola JM, Rodrı´guez JM, Peral A, Morales G, Abella E (2011) Synthesis of hierarchical ZSM-5 by silanization and alkoxylation of protozeolitic units. Catal Today 168:86–95. doi:10.1016/j.cattod.2010.12.040 30. Covarrubias C, Gracia F, Palza H (2010) Catalytic degradation of polyethylene using nanosized ZSM-2 zeolite. Appl Catal A-Gen 384:186–191. doi:10.1016/j.apcata.2010.06.034 31. Corma A, Dı´az U, Domine ME, Forne´s V (2000) AlITQ-6 and TiITQ-6: synthesis, characterization, and catalytic activity. Angew Chem Int Edit 39:1499–1501. doi:10.1002/(SICI)15213773(20000417)39:8\1499:AID-ANIE1499[3.0.CO;2-0 32. Corma A, Martı´nez A, Martı´nez-Soria V (2001) Catalytic performance of the new delaminated ITQ-2 zeolite for mild hydrocracking and aromatic hydrogenation processes. J Catal 200:259–269. doi:10.1006/jcat.2001.3219 33. Chica A, Dı´az U, Forne´s V, Corma A (2009) Changing the hydroisomerization to hydrocracking ratio of long chain alkanes by varying the level of delamination in zeolitic (ITQ-6) materials. Catal Today 147:179–185. doi:10.1016/j.cattod.2008.10.046 34. Venkatesh KR, Hu J, Tierney JW, Wender I (1996) Hydrocracking and hydroisomerization of long-chain alkanes and polyolefins over metal-promoted anion-modified transition metal oxides. US Patent 6184430 35. Emeis CA (1993) Determination of integrated molar extinction coefficients for infrared absorption bands of pyridine adsorbed on
J Mater Cycles Waste Manag solid acid catalysts. J Catal 141:347–354. doi:10.1006/jcat.1993. 1145 36. Corma A, Forne´s V, Forni L, Ma´rquez F, Martı´nez-Triguero J, Moscotti D (1998) 2,6-di-tert-butyl-pyridine as a probe molecule to measure external acidity of zeolites. J Catal 179:451–458. doi:10.1006/jcat.1998.2233
37. Nesterenko NS, Avdey AV, Ermilov AY (2006) FTIR study of the CO adsorption over Pt/MFI catalysts: Ab initio interpretation. Int J Quantum Chem 106:2281–2289. doi:10.1002/qua.20997
123