The effect of different zeolite based catalysts on the pyrolysis of poly butadiene rubber

Fuel 160 (2015) 544–548

Contents lists available at ScienceDirect

Fuel journal homepage: www.elsevier.com/locate/fuel

The effect of different zeolite based catalysts on the pyrolysis of poly butadiene rubber Sara Salam Zadeh Salmasi a, Mehrdad Seifali Abbas-Abadi a,b,⇑, Mehdi Nekoomanesh Haghighi a, Hosein Abedini a a b

Polymerization Engineering, Iran Polymer and Petrochemical Institute (IPPI), P.O. Box 14965/115, Tehran, Iran Chemical Engineering, Energy Department, Kermanshah University of Technology, P.O. Box 67178-3766, Kermanshah, Iran

h i g h l i g h t s  The effect of different zeolite based catalysts on the poly butadiene rubber is considered.  The catalysts can change and decrease the pyrolysis energy and control the product size.  The effect of heating rates on the poly butadiene rubber is considered.  The quality and quantity of the energy can specify the pyrolysis road map especially in the unsaturated polymers.  The degradation mechanisms changed by different heating rates in PBR pyrolysis.

a r t i c l e

i n f o

Article history: Received 24 October 2014 Received in revised form 24 July 2015 Accepted 28 July 2015 Available online 13 August 2015 Keywords: Pyrolysis PBR Catalyst Crosslink Stirred reactor

a b s t r a c t In this semi batch pyrolysis study, the effect of different zeolite catalysts containing used FCC, HZSM-5 and mordenite on the degradation of virgin poly butadiene (PBR) were considered. The liquid, gas and coke yield as well as composition of liquid product as a function of different catalysts effect in the stirred reactor were compared. Main products obtained were light hydrocarbons within the gasoline range. The formation of aromatic products depended appreciably on the catalyst type and catalyst/polymer ratio. The results showed that C6 was main component of pyrolysis product with up to 32%. The results also showed that crosslink mechanism played an important role in the rubber pyrolysis. A huge number of double bonds in the PBR chains could prepare the proper media for crosslink mechanism in slow pyrolysis while flash pyrolysis followed the chain scission mechanism in the most times of degradation. Crosslink nets could increase the activation energy and the resistance against the chain scission degradation by difficult chain mobility and heat transfer. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction The fast depletion of crude oil and the environmental issues have led to an intensive search for alternate 50 fuels for internal combustion engines. Waste to energy facilities provide significant economic benefit as a partially renewable energy source [1–4]. Rubber materials (poly butadiene rubber, styrene–butadiene rubber and poly isoprene or natural rubber) are the main components of general tires. It has high volatile and fixed carbon contents with heating value greater than that of coal. This makes rubber from old tire a good raw material for thermochemical processes and proper alternative for naphtha base fuels. On the other hand, ⇑ Corresponding author. Tel.: +98 2144580000x2168; fax: +98 2144580021x23. E-mail address: [email protected] (M.S. Abbas-Abadi). http://dx.doi.org/10.1016/j.fuel.2015.07.091 0016-2361/Ó 2015 Elsevier Ltd. All rights reserved.

scrap tire is not a biodegradable residue and therefore it is not possible to achieve its natural degradation in landfills. As a consequence, open dumping of scrap tire not only occupies a large space, presents an eyesore and could cause potential health and environmental hazards, but also illustrates wastage of valuable energy resource [5–7]. The comparison of different disposal options available shows that pyrolysis has a promising future, as it allows valuable materials to be recovered [8]. Pyrolysis, and gasification processes are considered to be more attractive and practicable methods for recovering energy from scrap tire, plastics and biomass. Pyrolysis of carbonaceous materials can be interpreted as incomplete thermal degradation, generally in the absence of air, resulting in char, condensable liquids or tars, and trace amount of gaseous products. Gasification refers

S.S.Z. Salmasi et al. / Fuel 160 (2015) 544–548

to pyrolysis followed by higher temperature reactions of the char, tars, and primary gases to yield mainly low molecular weight gaseous products [9–11]. The catalytic degradation of polymeric materials has been reported for a range of catalysts centered on the active components in a range of different model catalysts, such as amorphous silica– aluminas, zeolites Y, ZSM-5 and various acidic catalysts and particularly the new family of MCM materials [9–22]. However, these catalysts have been used that even if performing well, they can be unfeasible from the point of view of practical use due to the cost of manufacturing and the high sensitivity of the process to the cost of the catalyst. A very large number of double bonds in the poly butadiene rubber have a key role in the pyrolysis of the rubber. It can help us to better understand the degradation mechanisms and effective parameters on the pyrolysis. This paper studies the effect of zeolite based catalysts on the yield of pyrolysis, liquid composition, pyrolysis mechanism following two strategies: (i) The effect of different catalyst (catalyst/polymer: 15 w/w%) on the PBR pyrolysis, and (ii) used FCC catalyst as

Table 1 The specification of spent FCC catalyst used in the catalytic degradation of PBR. 235 m2/g 80.10% 13.40% 0.30% 1.54% 6 0.20% 450 180

Surface area (BET) SiO2 Al2O3 Na Ca Si/Al Fe V (ppm) Ni (ppm)

Table 2 The specification of HZSM-5 and mordenite catalysts used in the catalytic degradation of PBR. Specification

HZSM-5

Mordenite

Si/Al Surface area (m2/gr) Porosity (ml/gr) Porosity radius (Å)

20 381.3 432.68 22.69

20 449.1 367.92 14.39

545

commercially program by reactor pyrolysis and study of the mechanism in the degradation of PBR by reactor and TGA methods. 2. Experimental 2.1. Material PBR 1220 grade is supplied by Arak petrochemical company (Arak, Iran). Nitrogen gas (purity 99.99%) is supplied by Roham Co. Equilibrium FCC catalyst is supplied by Abadan FCC Refinery-regenerated at 750 °C, mild steaming – and its composition/properties are given in Table 1. HZSM-5 and mordenite catalysts are supplied by Sudchemi Company and Table 2 shows the specification. 2.2. Instruments 2.2.1. Analyzer instruments Identification of varied compounds in the condensed was carried out by a gas chromatograph mass spectrometer of model GC–MS-QP5000. The analysis was performed on a 60 m ⁄ 0.32 mm capillary column coated with a 1 lm film of DB 1. The oven temperature was programmed, 40 °C hold for 10 min to 300 °C at 5 °C/min hold for 10 min. Compounds were identified by means of the NIST12 and NIST62 library of mass spectra and subsets HP G1033A. The thermo gravimetric analysis was performed with a Netzsch TG 209 thermo balance. The PBR samples were studied at various heating rates between 15 and 45 °C/min. The initial mass of the sample was 12.0–15.0 mg. The experiments were carried out in a nitrogen atmosphere (99.99% minimum purity) with a flow rate of 30 ml/min. 2.2.2. Pyrolysis process Pyrolysis experiments were carried out in a 1 L stirred semi-batch reactor (bucchi pilot plant with a 120 custom built reactor) under atmospheric pressure and the schematic diagram is shown in Fig. 1. The fixed experimental conditions are as follows: The mixture of PBR rubber and catalyst in the reactor, carrier gas stream (300 ml/min), agitator speed (50 rpm) and heating rate (25 °C/min) up to the final temperature. The non-condensable products were vented after cooling through three condensers.

Fig. 1. Flow scheme of the laboratory stirred reactor.

546

S.S.Z. Salmasi et al. / Fuel 160 (2015) 544–548

Table 3 The effect of different catalysts on the products yield, specific density and the condensed product composition. Catalyst

Liquid

Gas

Coke

Specific density

Olefins (%)

Paraffins (%)

Naphthenes (%)

Aromatics (%)

No catalyst Used FCC HZSM-5 Mordenite

90.6 85.2 63.2 86.4

8.1 12.6 31.3 10.7

1.3 2.2 5.5 2.9

0.76 0.67 0.73 0.7

39 40.1 16.8 25.3

20.1 17.8 14.2 14.1

20.4 19.8 21.7 24.2

20.5 22.3 47.3 36.4

The condensed hydrocarbons products were stored in glass sampling bottles. The components of total condensed hydrocarbons (residue in the condensers contained C4 to C15) were quantified by gas chromatograph mass (GC–MS). The non-condensable products were not analyzed. The solid char yield was determined gravimetrically after completion of the reaction. The non-condensable yield was calculated by subtracting the weight of the condensed hydrocarbons and solid products from the sample weight.

The pyrolysis of PBR in the stirred reactor was undertaken at 450 °C in the absence of a catalyst and also with three zeolite catalysts. Table 3 shows the mass balance data for the pyrolysis and catalytic pyrolysis of the PBR in relation to different catalysts. The main product fraction is condensed hydrocarbons (yields up to 90.6%). The condensable products were characterized by GC–MS and composition as a function of catalyst is given in Tables 3 and 4. These reaction products were grouped into different classes i.e. naphthenes (cycloalkanes), paraffins (alkanes), olefins (alkenes) and aromatics (Table 3). The results show that products are

changed by type and amount of the catalyst. The distribution of C4 to C14 condensed hydrocarbon products and average molecular weight with pyrolysis temperature is given in Table 4. The pyrolysis in the absence of catalysis is shown as thermal pyrolysis and produced an oil yield of 90.6 wt.%, gas yield of 8.1 wt.% and char yield of 1.3 wt.%. With used FCC catalyst, the condensed product yield is 85.2% containing aliphatic hydrocarbons (olefins at 40.1% and paraffins at 17.8%) and cyclic hydrocarbons (aromatics at 22.3 and naphthenes at 19.8%). For light fuels, the C5 to C9 fraction is desirable for a gasoline feedstock. The C5-C9 fraction was shown to be the major product (87.9%) in the condensed product. In catalytic pyrolysis with mordenite, the condensed product yield is 86.4, gas yield of 10.7 and coke of 2.9%. The condensed product is composed of olefins (25.3%), paraffins (14.1%), naphthenes (24.2%) and aromatics as main component (36.4%). It clearly shows that the reduction in naphthene and paraffin yields was in favor of double bond formation indicating that unsaturation, cyclization and aromatization took place by mordenite catalyst. The results show that mordenite catalyst want to produce higher molecular weights hydrocarbons and the gasoline range is lower than the produced liquid by used FCC [12,23,24]. While with HZSM-5 catalyst, the condensed hydrocarbons decreased in comparison with the other catalysts and reached a yield of 63.2%. The highest amount of gaseous products was obtained with yield of 31.3%. With using all of the catalysts, the gasoline range is the main compounds of the product. The formation of aromatics in the pyrolysis of polymers is accomplished using the Diels–Alder reaction, followed by dehydrogenation. The detailed mechanism of the formation of BTX aromatics is presented in some papers [12,23,24].

Table 4 The effect of different catalysts on the carbon number distribution of the condensed product composition.

Table 6 The effect of used FCC/PBR ratio on the carbon number distribution of the condensed product composition.

3. Results and discussion Mass balance of PBR pyrolysis products (condensable products, solid residue and non-condensable by difference) were determined gravimetrically (Tables 3–6) for different catalysts (used FCC, HZSM-5 and mordenite) and used FCC catalyst/PBR ratio (0–60%). Finally the mechanism of rubber pyrolysis in a stirred reactor is considered. 3.1. Effect of the different catalysts

Carbon number

Used FCC

HZSM-5

Mordenite

Carbon number

0

15

30

45

60

C4 C5 C6 C7 C8 C9 C10 C10+ Gasoline range

4.6 13.2 31.6 12.9 15.4 14.8 1.8 5.7 87.9

7.9 16.7 33.4 19.2 11.1 6.4 4.2 1.1 86.8

0.5 1.1 31.7 12.4 11.9 23.0 6.6 12.8 80.1

C4 C5 C6 C7 C8 C9 C10 C10+ Sum (C5–C9)

0.1 1.3 9.7 7.8 17.0 26.7 19.8 17.6 62.5

11.6 6.2 31.6 12.9 15.4 14.8 1.8 5.7 80.9

2.8 5.1 19.5 14.2 17.9 15.2 9.1 16.2 71.9

5.3 5.2 16.3 17.2 18.1 15.2 10.2 12.5 72.0

9.3 5.6 13.2 13.6 17.5 17.8 9.1 13.9 67.7

Table 5 The effect of used FCC/PBR ratio on the product yield and liquid composition. Used FCC/PBR (%)

Liquid

Gas

Coke

Specific density

Olefins (%)

Paraffins (%)

Naphthenes (%)

Aromatics (%)

0 15 30 45 60

90.6 85.2 89.3 92.5 86.7

8.1 12.6 4.0 0.5 5.2

1.3 2.2 6.7 7.0 8.1

0.76 0.67 0.72 0.72 0.72

39.0 40.1 22.0 17.4 16.3

20.1 17.8 20.8 21.4 15.5

20.4 19.8 21.7 22.3 25.0

20.5 22.3 35.5 38.9 43.2

S.S.Z. Salmasi et al. / Fuel 160 (2015) 544–548

3.2. Effect of the used FCC catalyst content The effect of increasing the used FCC/PBR ratio from 0 to 0.6 on the condensable, non-condensable and coke product yields are given in Table 5. The maximum condensed product yield was achieved with a 45% FCC catalyst. The results show that coke yields increase with catalyst content. The coke yield may be attributable to aromatization and dehydrogenation on the catalyst surface. Non condensable products show no specific trend with catalyst increasing. The composition of paraffins, naphthenes, olefins and aromatics in the condensed fraction, as a function of catalyst content, are given in Table 5 too. The results show that the share of components were olefins (16–41%), paraffins (15–22%), naphthenes (19–25%) and amounts of aromatic (20–44%) and These data indicate that dehydrogenation increases with catalyst content as judged by the aromatics product patently. The product carbon number distribution and molecular weight of the condensed hydrocarbons at different catalyst ratios are given in Table 6. The results show that as the catalyst ratio has no specific effect on the molecular weight although it seems that as the catalyst ratio increases the product molecular weight decreases. Furthermore, in plastics as the catalyst/polymer ratio increases the molecular size selectively and the gasoline product (C5–C9) increases but in PBR it the results show an opposite trend and it maybe affect by different degradation mechanism in rubber pyrolysis [10,14]. 3.3. The mechanism of PBR pyrolysis For better consideration of the pyrolysis mechanism, pyrolysis is done in two conditions, fast and slow pyrolysis. The fixed experimental conditions for fast pyrolysis are PBR (100 g) and the used FCC catalyst (15 g). In slow pyrolysis, the rate of pyrolysis decrease up to half. For rate decreasing, we increase the PBR and used FCC catalyst amounts up to two times (200 g & 30 g). The reactor shows different behavior in these processes. In fast pyrolysis, the reactor have enough power and energy for the degradation of the rubber and the product leaves the reactor continuously but in the slow pyrolysis, quantity of the rubber in the reactor is very high and the inadequate power cannot degrade the rubber in short time. The results show that infinitive unsaturated double bonds and

Fig. 2. The rate of liquid temperature and temperature–time curves for slow pyrolysis of PBR.

547

the low energy – enough just for degrading the p bonds in low temperatures – can prepare a proper media for crosslink mechanism and reveal that the pyrolysis rate can have important role in the degradation mechanism. In the slow pyrolysis the products leave the reactor step by step (Fig. 2). In slow pyrolysis, rubber chains have enough time for crosslink creation in low temperatures and it can create the multilayer media with different degradation activation energy. The layers with different activation energy leave the reactor step by step although in fast pyrolysis the chains have no enough time for crosslink mechanism and the results confirm these phenomena. Fig. 2 shows the product rate and temperature of the reactor versus pyrolysis time in the slow pyrolysis. The results show a sinuous trend of the reactor temperature under pyrolysis time. It can obvious that the product leaves the reactor step by step because the temperature sensor is on the top of the reactor and the products with high temperature can affect on the sensor obviously. Among the steps, when the production rate decreases clearly, the temperature decreases because the hot products do not touch the temperature sensor and cold nitrogen as carrier gas decreases the temperature of the sensor. The results reveal that the pyrolysis time for the fast pyrolysis is less than 37 min and 82 min for the slow pyrolysis. For better understanding of these phenomena, PBR has considered by TG at the rates of 15, 30 and 45 °C/min (Fig. 3). TGA analysis of polymers generally show that by increasing the heating rate decreases the total heat transfer into the polymer bulk at the same temperatures, therefore polymers with low heat transfer coefficients will degrades at a higher temperature. If the heating rate and the other process parameters do not change the pyrolysis mechanisms and the share of each mechanism in the degradation, the polymer follows this rule [25–29]. Although the heating rate has different effects on the synthetic and natural polymers [30,31,8] but the pyrolysis of PBR shows the strange and reverse trend to common plastics and PBR degrades at lower temperatures under fast pyrolysis conditions (45 °C/min) as compared with lower heating rates (slow pyrolysis). Under slow pyrolysis, the double bonds in PBR can crosslink and the heating rate can influence the degradation mechanism. At low heating rates, the PBR chains have enough time and energy to crosslink. At high heating rates, such as fast pyrolysis, there is high energy available for chain scission and is the main degradation mechanism but insufficient time for PBR to crosslink. Under slow pyrolysis, PBR has resistance to chain scission degradation. The internal zones of bulk polymer have a temperature lag relative to the surface layer and thus there is enough time for the bulk polymer to crosslink.

Fig. 3. TGA curves of polybutadiene rubber at different heating rates (15, 30 and 45 °C/min).

548

S.S.Z. Salmasi et al. / Fuel 160 (2015) 544–548

4. Conclusion A laboratory catalytic stirred system has been used to obtain a range of volatile hydrocarbons by catalytic degradation of PBR with HZSM-5, used FCC and Mordenit, the used FCC/polymer ratio from 0 to 60. The catalytic degradation of PBR over zeolite based catalyst using cracking reactions was shown to be a useful method for the production of potentially valuable hydrocarbons. As can be seen in this work, used FCC and mordenite catalysts produce high liquid share although HZSM-5 produce high share of aromatics. The results show that the rate of pyrolysis can affect the degradation mechanism. The higher pyrolysis rates can decrease the crosslink density and decrease the pyrolysis time. The results show that in slow pyrolysis rates, crosslink mechanism creates 3-D nets and the products leave the reactor very difficult and in some steps although in fast pyrolysis, the products leave the reactor continuously. The quality and quantity of the energy can specify the pyrolysis road map especially in the unsaturated polymers. Sufficient energy with proper contact angle can follow the scission mechanism and insufficient energy or improper contact angle can carry out the process under crosslink mechanism. On the other hand, the catalysts can change and decrease the pyrolysis energy and with the different porosities can control the product size.

[11]

[12] [13]

[14]

[15]

[16] [17]

[18]

[19]

[20] [21] [22] [23]

References [24] [1] Murugan S, Ramaswamy MC, Nagarajan G. Assessment of pyrolysis oil as an energy source for diesel engines. Fuel Process Technol 2009;90(1):67–74. [2] Barbooti MM, Mohamed TJ, Hussain AA, Abas FO. Optimization of pyrolysis conditions of scrap tires under inert gas atmosphere. J Anal Appl Pyrol 2004;72:165–70. [3] Roy C, Chaala A, Darmstadt H. The vacuum pyrolysis of used tires end-uses for oil and carbon black products. J Anal Appl Pyrol 1999;51:201–21. [4] Zabaniotou AA, Stavropoulos G. Pyrolysis of used automobile tires and residual char utilization. J Anal Appl Pyrol 2003;70:711–22. [5] Aylo´n E, Calle´n MS, Lo´pez JM, Mastral AM. Assessment of tire devolatilization kinetics. J Anal Appl Pyrol 2005;74:259–64. [6] Islam MR, Haniu H, Alam Beg MR. Liquid fuels and chemicals from pyrolysis of motorcycle tire waste: product yields, compositions and related properties. Fuel 2008;87:3112–22. [7] Murillo R, Navarro MV, Stelmach S. Effect of acid catalysts on scrap tyre pyrolysis under fast heating conditions. J Anal Appl Pyrol 2008;82:199–204. [8] Leung DYC, Wang CL. Kinetic study of scrap tyre pyrolysis and combustion. J Anal Appl Pyrol 1998;45:153–69. [9] Seifali Abbas-Abadi M, Nekoomanesh Haghighi M, Yeganeh H. The effect of temperature, catalyst, different carrier gases and stirrer on the produced transportation hydrocarbons of LLDPE degradation in a stirred reactor. J Anal Appl Pyrol 2012;95:198–204. [10] Seifali Abbas-Abadi M, Nekoomanesh Haghighi M, Yeganeh H. Evaluation of pyrolysis product of virgin high density polyethylene degradation using

[25]

[26]

[27] [28]

[29]

[30]

[31]

different process parameters in a stirred reactor. Fuel Process Technol 2012;109:90–5. Zhou Q, Wang Y-Zh, Tang Ch, Zhang Y-H. Modifications of ZSM-5 zeolites and their applications in catalytic degradation of LDPE. Polym Degrad Stab 2003;80(1):23–30. Nisar J, Ali M, Awan IA. Catalytic thermal decomposition of polyethylene by pyrolysis gas chromatography. J Chil Chem Soc 2011;56(2):653–5. Jan MR, Shah J, Gulab H. Catalytic degradation of waste high-density polyethylene into fuel products using BaCO3 as a catalyst. Fuel Process Technol 2010;91(11):1428–37. Seifali Abbas-Abadi M, Nekoomanesh Haghighi M, Yeganeh H, McDonald AG. Evaluation of pyrolysis process parameters on polypropylene degradation products. J Anal Appl Pyrol 2014;109:272–7. Marcilla A, Beltra´n MI, Navarro R. Thermal and catalytic pyrolysis of polyethylene over HZSM5 and HUSY zeolites in a batch reactor under dynamic conditions. Appl Catal B 2009;86:78–86. Ahmad E, Chadar S, Tomar SS, Akram MK. Catalytic degradation of waste plastic into fuel oil. Int J Petrol Sci Technol 2009;3(1):25–34. Jung SH, Cho MH, Kang BS, Kim JS. Pyrolysis of a fraction of waste polypropylene and polyethylene for the recovery of BTX aromatics using a fluidized bed reactor. Fuel Process Technol 2010;91:277–84. Salman M, Rehman R, Shafique U, Mahmud T, Ali B. Comparative thermal and catalytic recycling of low density polyethylene into diesel-like oil using different commercial catalysts. Electron J Environ, Agric Food Chem 2012;11(2):96–105. Almustapha MN, Andrésen JM. Recovery of valuable chemicals from high density polyethylene (HDPE) Polymer: a catalytic approach for plastic waste recycling. Int J Environ Sci Develop 2012;3(3). Tiwari DC, Ahmad E, Kumar Singh KK. Catalytic degradation of waste plastic into fuel range hydrocarbons. Int J Chem Res 2009;1(2):31–6. Ali MF, Qureshi MS. Catalyzed pyrolysis of plastics: a thermogravimetric study. Afr J Pure Appl Chem 2011;5(9):284–92. Bajus M, Hájeková E. Thermal cracking of the model seven components mixed plastics into oils/waxes. Petrol Coal 2010;52(3):164–72. Marcilla A, Go´mez A, Reyes-Labarta JA, Giner A. Catalytic pyrolysis of polypropylene using MCM-41: kinetic model. Polym Degrad Stab 2003;80:233–40. Williams PT, Slaney E. Analysis of products from the pyrolysis and liquefaction of single plastics and waste plastic mixtures. Resour Conserv Recycl 2007;51:754–69. Abbas-Abadi MS, Haghighi MN, Yeganeh H, Bozorgi B. The effect of melt flow index, melt flow rate, and particle size on the thermal degradation of commercial high density polyethylene powder. J Therm Anal Calorim 2013;114(3):1333–9. Abbas-Abadi MS, Haghighi MN, Yeganeh H. Effect of the melt flow index and melt flow rate on the thermal degradation kinetics of commercial polyolefins. J Appl Polym Sci 2012;126:1739–45. Senneca O, Salatino P, Chirone R. A fast heating-rate thermogravimetric study of the pyrolysis of scrap tyres. Fuel 1999;78:1575–81. Aboulkas A, El harfi K, El Bouadili A. Thermal degradation behaviors of polyethylene and polypropylene. Part I: pyrolysis kinetics and mechanisms. Energy Convers Manage 2010;51(7):1363–9. Paik P, Kar KK. Thermal degradation kinetics and estimation of lifetime of polyethylene particles: effects of particle size. Mater Chem Phys 2009;113:953–61. Fushimi Ch, Araki K, Yamaguchi Y, Tsutsumi A. Effect of heating rate on steam gasification of biomass. 2. Thermogravimetric–mass spectrometric (TG–MS) analysis of gas evolution. Int J Chem Res 2003;42:3929–36. Mehrabian R, Scharler R, Obernberger I. Effects of pyrolysis conditions on the heating rate in biomass particles and applicability of TGA kinetic parameters in particle thermal conversion modelling. Fuel 2012;93:567–75.