b. National Synchrotron Radiation Laboratory, University of Science and Technology of China, Hefei 230029, China;
c. Key Laboratory for Power Machinery and Engineering of Education, Shanghai Jiao Tong University, Shanghai 200240, China
Plastics are one of the dominant components in municipal solid waste. According to the data supplied by the World Bank, the amount of plastic waste in the world will be 2.2 million tons by 2025 . Landfill and incineration are the most widely used methods for solid wastes processing. However, landfill is not a proper way to deal with plastic wastes for its vast amount and long term of degradation in the environment. While, incineration of plastic wastes produces lots of poisonous products including dioxin .
Catalytic pyrolysis is an appropriate way for plastic recycling [3, 4], in which zeolite is a kind of widely used catalyst and shows powerful functions . The strength of acid sites of zeolites is critical in catalytic reactions and greatly influences the reaction pathways and selectivity of catalytic products [6-8]. Selective poisoning with basic molecules can be used as a method to investigate the performance of zeolites with varied acid sites [8-17]. To study the effect of selective poisoning on the catalysts and corresponding catalytic reactions, gas chromatography (GC) and gas chromatography mass spectrometry (GC-MS) are generally used to identify the catalytic products, and evaluate the selectivity and conversion efficiency of catalyst. However, GC-MS is time-consuming, and the nascent products and their dynamics information cannot be revealed in real time. On-line photoionization mass spectrometry is a kind of experimental method which has great advantages in detecting reaction intermediates in real time [18-21]. In comparison to the conventional "hard" electron ionization method used in commercial mass spectrometers, fragment-free mass spectra can be obtained with near-threshold photoionization, which is conducive to characterize the products from catalytic reactions in real time .
Recently, a pyrolysis photoionization time-of-flight mass spectrometer (Py-PI-TOFMS) using a portable Krypton discharge lamp as ionization source has been constructed and utilized for the studies of polymer, coal pyrolysis [18, 19, 21, 23]. In this work, Py-PI-TOFMS was applied in the study of polypropylene (PP) pyrolysis over various H-form ultra stable Y (HUSY) zeolites after ammonia poisoning. Online photoionization mass spectra for the pyrolysis products of PP and HUSY with various acid strength were obtained, and the formation curves of various pyrolysates of PP/HUSY with the increase of temperature were determined. Our results show that the acid strength affects the pyrolysis temperature, products selectivity and yields significantly.Ⅱ. EXPERIMENTS A. Sample preparation
The PP particle (isotactic,
The experiment was carried out using a homemade Py-PI-TOFMS apparatus that has been previously reported in detail [19, 21, 24], so only a brief introduction is given here. The experimental setup comprises a tubular furnace, a transfer line, and a PI-TOFMS (see Fig. 1).
The pyrolysis temperature of the tubular furnace was controlled by a temperature controller (SKY Technology Development Co., Ltd., China). The feedback of the furnace temperature and the sample temperature were measured by K-type thermocouples 1 and 2, respectively. Prior to the experiment, the system was purged with nitrogen for 15 min. The flow rate of the carrier gas (nitrogen) was maintained at 200 standard cubic centimeters per minute (SCCM). When the temperature reached the set value, the sample was introduced into the middle position of the furnace using a quartz sample boat. The pyrolysis products were transferred through a deactivated fused-silica capillary (Internal diameter of 250 µm) inside the heated transfer line (250 ℃) to reach the ionization chamber (0.75 Pa), where the products were ionized by ultraviolet light emitted from a Kr lamp with a photon energy of 10.6 eV (PKS106, Heraeus, Ltd., Germany). To remove the fine particles from the product gas stream, a glass fiber filter with a pore size of 1.2 µm was placed between the transfer line and the furnace. The formed ions were mass analyzed by TOFMS. The ion signal was amplified with a VT120C preamplifier (ORTE C, Oak Ridge, USA) and recorded by a P7888 multiscaler (FAST Comtec, Oberhaching, Germany).
The pyrolysis/photoionization mass spectrometric measurements were conducted in both temperature-fixed mode and temperature-programmed mode. In the temperature-fixed mode, the furnace was hold at a specific value in the range of 300-600 ℃. In the temperature-programmed mode, the heating rate of the furnace was set to 10 ℃/min and the acquisition time for each spectrum was 10 s.C. NH3-temperature-programmed desorption
The NH3-temperature-programmed desorption (TPD) experiment was also fulfilled using Py-PI-TOFMS. The heating rate was set to 10 ℃/min with 200 SCCM nitrogen as purge gas. Figure 2 shows the normalized NH3-TPD profile of HUSY, which contains two broad bands. The peak at approximate 200 ℃ is due to desorption of weak acid, and the peak at approximate 400 ℃ corresponds to the desorption of strong acid . From Fig. 2 we can deduce that HUSY pretreated at ambient temperature and 200 ℃ (ID: 1 and 2) only contain weak acid sites. HUSY pretreated at 300 ℃ (ID: 3) contains part of moderate acid sites. When HUSY was pretreated at 400 and 500 ℃ (ID: 4 and 5), strong acid sites were liberated.Ⅲ. RESULTS AND DISCUSSION A. Mass spectra
The photoionization mass spectra of pyrolytic products from PP/50% HUSY are presented in Fig. 3. Due to the near-threshold photoionization character, nearly all the mass peaks could be attributed to parent ions. The pyrolysis products can be separated into several categories, i.e., alkanes (m/z=72, 86, 100, 114, etc.), alkenes (m/z=28, 42, 56, 70, 84, 98, 112, etc.), dienes (m/z=68, 82, 96, etc.) and aromatics (m/z=78, 92, 106, 120, 128, 134, etc.) . The major products of PP pyrolysis are identified and listed in the supplementary materials (Table S1). Varying the pretreatment conditions of HUSY and pyrolysis temperatures will not change the products distribution.B. Temperature-fixed experiments
The influences of the acid strength and amounts of acid sites in catalyst (ID: 1, 2, 3, 4, and 5) on the products pyrolyzed at four temperatures (300, 400, 500, and 600 ℃) are shown in Fig. 4. The catalyst's acid strength influences the yields of all kinds of products obviously. The yields of alkanes with low carbon numbers increase as the acid strength gets higher (Fig. 4(a) and (b)). However, alkanes with carbon numbers higher than 7 show a totally different trends. With the increase of acid strength, the yields firstly increase to some extent, and then drop to a relatively low level (Fig. 4(c)). The yields of most alkenes and dienes decrease as the strength and amounts of acid sites become higher, except for low carbon number alkenes (C4H8 and C5H10 in Fig. 4(d) and (e)). Aromatics products larger than benzene (C7H8) also show great dependence on strength and amounts of acid sites of HUSY.
Temperature is a key factor in PP pyrolysis, and will greatly affect the yields of different pyrolysates . As shown in Fig. 4, higher yields will be observed for low carbon number products by increasing the pyrolysis temperature from 300 ℃ to 600 ℃ (Fig. 4(a), (d), (g), (j)). As for larger carbon number products, such as C9H20, C8H16, C9H16, and C12H18, higher temperature will reduce their yields due to the secondary cracking reactions (Fig. 4(c), (f), (i), (l)).
An interesting phenomenon is that the extent of enhanced yields of PP pyrolysates with increased temperature will be reduced by improving the strength of acid sites. For example, the relative intensities of C7H8 in Fig. 4(j) changed from 1.02 (300 ℃) to 3.84 (600 ℃), with HUSY only having weak acid sites (ID: 1). However, when acid sites were replaced by stronger ones, the relative intensities of C7H8 varied from 1.29 to 2.79 (ID: 6). Obviously, the difference between the yields at two pyrolysis temperatures is greatly reduced by using high acidity HUSY. From the point of view of activation energy, the strong acid sites can lower the cracking energy more than the weak ones. Therefore, the increase of pyrolysis temperature affects little to the samples with strong acid sites. The result shows that 300 ℃ is enough for cracking PP, but the weak acid samples need more energy to get equivalent results than the strong ones.C. Temperature-programmed experiments
The formation profiles of pyrolysates can be obtained from temperature-programmed mode of Py-PI-TOFMS. Figure 5 shows the profiles of four kinds of C7 products from the pyrolysis of PP/HUSY with different acid strength (ID: 1, 2, 3, 4, 5). Pyrolysis products with other carbon numbers are supplied in the supplementary materials (Fig.S1-Fig.S3).
The initial formation temperatures of these pyrolysates can be distinguished from the baseline. As shown in Fig. 5, the appearance temperatures of pyrolysis products with weak acidity (ID: 1 and 2) are relatively high and nearly the same. However, with the improvement of acid strength (ID: 3, 4 and 5), the formation temperatures of these C7 products are gradually decreased and separated. Pyrolysis with mild acidic catalyst (ID: 1, 2, 3) prefers to produce alkene (C7H14) initially, while catalysts with strong acidity (ID: 4, 5) result in the production of alkane (C7H16) first.
To make the information in Fig. 5 more reasonable, all the appearance temperatures of alkanes, alkenes, dienes and aromatics with carbon numbers from 3 to 14 were recorded. Figure 6 shows the relationship between the initial appearance temperatures of these pyrolysis products and carbon number, with HUSY at different ammonia desorption levels. As can be seen from Fig. 6, acid strength is a dominant factor to decrease the appearance temperatures (ID: 1, 2, 3). When acid strength of HUSY reaches a relatively strong level (ID: 3, 4, 5), the formation temperatures of pyrolysis products will not change too much, indicating the strength of acid sites is not predominant. At the same time, the temperature needed for the formation of long chain pyrolysates becomes higher, especially for alkanes and alkenes. In comparison with alkanes and alkenes, the formation temperatures of dienes and aromatics are higher, indicating that the activation energies of the corresponding reactions are higher.Ⅳ. CONCLUSION
In this work, Py-PI-TOFMS was applied to study the behavior of ammonia poisoning on HUSY for the catalytic pyrolysis of PP. Secondly, online photoionization mass spectra for the pyrolysis products of PP and HUSY with various acid strength were recorded at different pyrolysis temperatures. Finally, the formation curves of various pyrolysates of PP/HUSY with the increase of temperature were determined. Our results indicate that the formation temperatures, yields and selectivity of the pyrolysis products of PP demonstrate obvious relationship with the acid strength of HUSY.
The yields of product and pyrolysis temperature effect were investigated at temperature-fixed experiments. The yields are significantly affected by the strength of acid sites. The strong acid sites (ID: 3, 4, 5) are more efficient on generating C4 to C6 products. But for other products, the samples with HUSY which only have weak (ID: 1, 2) acid sites get higher yields. The temperature effect of samples with weak acid sites is obvious owing to the high cracking activation energy of these samples.
The formation temperatures of four kinds of products are studied in temperature-programmed experiments. Stronger acid sites lower the reaction temperature more obviously. A critical strength of acid sites can be infered: once the valid acid strength is higher than the critical strength, the much stronger acid sites contribute little to cracking temperature. The promotion effect on formation temperature of strong acid sites for alkanes is better than that for alkenes.
Supplementary materials: The identification of major pyrolysis products of PP are shown in Table S1. The temperature-programmed pyrolysis results for C2-C5, C6-C9, and C10-C13 species are shown in Fig.S1, Fig.S2, Fig.S3, respectively.Ⅴ. ACKNOWLEDGMENTS
This work was supported by the National Natural Science Foundation of China (No.91545120 and No.U1432128), the National Basic Research Program of China (973 Program) (No.2013CB834602 and No.2012CB719701), Chinese Academy of Sciences, and Chinese Universities Scientific Fund.
|||Gloable Urban and DRM Unit, Results-based Financing For Municipal solid Waste, Washington D. C.:World Bank, (2014).|
|||U. S. EPA, Estimating Exposures to Dioxin-like Compounds, Washington, D. C.:Environmental Protection Agency US, (1994).|
|||S. L. Wong, N. Ngadi, T. a. T. Abdullah, and I. M. Inuwa, Renew. Sust. Energ. Rev. 50 , 1167 (2015). DOI:10.1016/j.rser.2015.04.063|
|||S. Kumar, A. K. Panda, and R. K. Singh, Resour. Conserv. Recycl. 55 , 893 (2011). DOI:10.1016/j.resconrec.2011.05.005|
|||J. H. Clark, Accounts. Chem. Res. 35 , 791 (2002). DOI:10.1021/ar010072a|
|||A. Corma, and H. Garcia, Catal. Today. 38 , 257 (1997). DOI:10.1016/S0920-5861(97)81500-1|
|||Y. Ono, and T. Baba, Catal. Today. 38 , 321 (1997). DOI:10.1016/S0920-5861(97)81502-5|
|||L. Garcia, G. Giannetto, M. R. Goldwasser, M. Guisnet, and P. Magnoux, Catal. Lett. 37 , 121 (1996). DOI:10.1007/BF00813529|
|||P. A. Jacobs, and C. F. Heylen, J. Catal. 34 , 267 (1974). DOI:10.1016/0021-9517(74)90036-0|
|||C. M. Fu, and A. M. Schaffer, Ind. Eng. Chem. Prod. Res. Dev 24 , 68 (1985). DOI:10.1021/i300017a013|
|||T. C. Ho, A. R. Katritzky, and S. J. Cato, Ind. Eng. Chem. Res. 31 , 1589 (1992). DOI:10.1021/ie00007a002|
|||R. Hughes, G. Hutchings, C. L. Koon, B. Mcghee, and C. E. Snape, Studies in Surface Science and Catalysis, Ostend, Belgium:Elsevier, 377(1994)|
|||V. R. Choudhary, and A. K. Kinage, Zeolites 15 , 732 (1995). DOI:10.1016/0144-2449(95)00082-H|
|||A. Borgna, J. Sepulveda, S. I. Magni, and C. R. Apesteguia, Appl. Catal. A 276 , 207 (2004). DOI:10.1016/j.apcata.2004.08.007|
|||G. Caeiro, P. Magnoux, J. M. Lopes, and F. R. Ribeiro, Appl. Catal. A 292 , 189 (2005). DOI:10.1016/j.apcata.2005.06.016|
|||C. Breitkopf, S. Matysik, and H. Papp, Appl. Catal. A 301 , 1 (2006). DOI:10.1016/j.apcata.2005.11.009|
|||G. Caeiro, A. F. Costa, H. S. Cerqueira, P. Magnoux, J. M. Lopes, P. Matias, and F. R. Ribeiro, Appl. Catal. A 320 , 8 (2007). DOI:10.1016/j.apcata.2006.11.031|
|||F. Qi, Proc. Combust. Inst. 34 , 33 (2013). DOI:10.1016/j.proci.2012.09.002|
|||Y. Wang, Q. Huang, Z. Y. Zhou, J. Z. Yang, F. Qi, and Y. Pan, Energy Fuels 29 , 1090 (2015). DOI:10.1021/ef502529w|
|||Y. Pan, Y. H. Hu, J. Wang, L. L. Ye, C. Y. Liu, and Z. X. Zhu, Anal. Chem. 85 , 11993 (2013). DOI:10.1021/ac402955k|
|||Y. N. Zhu, X. M. Chen, Y. Z. Wang, W. Wen, Y. Wang, J. Z. Yang, Z. Y. Zhou, L. D. Zhang, Y. Pan, and F. Qi, Energy Fuels 30 , 1598 (2016). DOI:10.1021/acs.energyfuels.5b02286|
|||F. Jiao, J. Li, X. Pan, J. Xiao, H. Li, H. Ma, M. Wei, Y. Pan, Z. Zhou, and M. Li, Science 351 , 1065 (2016). DOI:10.1126/science.aaf1835|
|||Y. Wang, Y. N. Zhu, Z. Y. Zhou, J. Z. Yang, Y. Pan, and F. Qi, Energy Fuels 30 , 1534 (2016). DOI:10.1021/acs.energyfuels.5b02234|
|||Z. X. Zhu, J. Wang, K. Q. Qiu, C. Y. Liu, F. Qi, and Y. Pan, Rev. Sci. Instrum. 85 , 3 (2014).|
|||N. Y. Tops, K. Pedersen, and E. G. Derouane, J. Catal. 70 , 41 (1981). DOI:10.1016/0021-9517(81)90315-8|
b. 中国科学技术大学国家同步辐射实验室, 合肥 230026;
c. 上海交通大学机械与动力工程学院, 上海 200240