s opa

logykka

s heffd tre oeatirioue liJ/k

ing anith a

ural are majoalmosadultcle for

be compacted. Dumped scrap tire in massive stockpiles is one ofthe possible causes of ideal breeding grounds for disease carryingmosquitoes and other vermin with the aid of rain water, which isdeposited in the free space of the tire wall. Also, landlling is a po-tential danger because of the possibility of accidental res withhigh emissions of hazardous gases. Moreover, different alternativesare often used for tire recycling such as retreading, reclaiming,

high value light hydrocarbon. Limonene has extremely fast grow-ing and wide industrial applications including formulation ofindustrial solvents, resins, and adhesives, as a dispersing agentfor pigments, as a fragrance in cleaning products and as an envi-ronmentally acceptable solvent [79,13]. It is very common incosmetic products and also used as aming combustible liquid.Furthermore, the biological activity of limonene, such as its chemo-preventive activity against rat mammary cancer, has been recentlyinvestigated [7]. In order to enhance the marketability of tirepyrolysis liquids, Stanciulescu and Ikura [14,15] separated

* Corresponding author. Tel./fax: +88 0721 750319.

Fuel 87 (2008) 31123122

Contents lists availab

ue

: wE-mail address: mrislam1985@yahoo.com (M. Roqul Islam).[1]. It is estimated that 1.72 million (5160 metric ton) of motor-cycle tires become scrap every year and wait for disposal. The esti-mated value is about 6.16 wt.% of total tire waste generation inBangladesh. Moreover, motorcycle is also a common carrier for fas-ter movement all over the world.

The disposal of solid tire wastes from human activity is a grow-ing environmental problem for the modern society, especially indeveloping countries. This organic solid waste is non-biodegrad-able. One common way of disposal is landlling. Landlling for dis-posal of used tires is connected with some problems: it needs aconsiderable amount of space because the volume of tires cannot

liquid products can be stored until required or readily transporta-tion to where it can be most efciently utilized. Tire pyrolysis liq-uids (a mixture of parafns, olens, and aromatic compounds)have been found to have a high gross caloric value (GCV) ofaround 4144 MJ/kg, which would encourage their use as replace-ments for conventional liquid fuels [210]. In addition to their useas fuels, the liquids have been shown to be a potential source oflight aromatics such as benzene, toluene, and xylene, whichcommand a higher market value than the raw oils [24,8,11,12].Similarly, the liquids have been shown to contain monoterpenesuch as limonene [1-methyl-4-(1-methylethenyl)-cyclohexene], a1. Introduction

Bangladesh is one of the developlated (914 persons/km2) countries, wmillion. About 77% people live in rland; agriculture is remaining as thconditions in the rural areas arestructed, and hence more than 2% ofpopulation (43 million) use motorcy0016-2361/$ - see front matter 2008 Elsevier Ltd. Adoi:10.1016/j.fuel.2008.04.036d most densely popu-total population of 135eas owing to its fertiler occupation. The roadt narrow and un-con-(1564 years old) maletheir faster movement

incineration, grinding, etc. with signicant drawbacks and/or limi-tations [2].

Pyrolysis for energy recovery from organic solid wastes basi-cally involves the decomposition of the wastes at high tempera-tures (300900 C) in an inert atmosphere. Three products aretypically obtained from the organic solid wastes: liquids, solid charand gases. The pyrolysis of solid tire wastes has received increasingattention since the process conditions may be optimized to pro-duce high energy density liquids, char, and gases. In addition, theCompositions 2008 Elsevier Ltd. All rights reserved.Liquid fuels and chemicals from pyrolysiProduct yields, compositions and related

M. Roqul Islam a,*, H. Haniu b, M. Raqul Alam BegaDepartment of Mechanical Engineering, Rajshahi University of Engineering and TechnobDepartment of Mechanical Engineering, Kitami Institute of Technology, Kitami City, Ho

a r t i c l e i n f o

Article history:Received 28 May 2007Received in revised form 22 April 2008Accepted 22 April 2008Available online 2 June 2008

Keywords:Motorcycle tire wastesLiquid fuels and chemicals

a b s t r a c t

In this study, experimentdetermine particularly thepyrolysis product yields ana nal pyrolysis temperatuin a xed-bed re-tube helemental analysis and vaspectroscopic studies on tha caloric value of 42.00 M

F

journal homepagell rights reserved.f motorcycle tire waste:roperties

, Rajshahi 6204, Bangladeshido 090-8507, Japan

ave been conducted on the sample of solid motorcycle tire wastes toect of temperature, feed size, and apparent vapor residence time on theheir compositions. The maximum liquid yield of 49 wt.% was obtained atf 475 C, feed size 4 cm3, with a residence time of 5 s under N2 atmosphereng reactor system. The pyrolysis liquid products were characterized bys chromatographic and spectroscopic techniques. Chromatographic andquids show that it can be used as liquid fuels and chemical feedstock, withg and empirical formula of CH1.27O0.025N0.006.

le at ScienceDirect

l

ww.fuelfirst .com

tional literature for the effect of feed size andvapor residence time intire wastes pyrolysis regime. Dai et al. [6] and Barbooti et al. [10]

Fuellimonene-enriched fraction using vacuum distillations and reactedwith methanol to produce limonene ethers. Methyl limonene etheris a high value product and has a very pleasant fragrance. It can beused alone or as an odor-improving additive. The former research-er groups [79,13] extensively investigated separation of limonenefrom tire derived pyrolysis liquids using various techniques. Pyro-lytic char may be used as a solid fuel or as a precursor for activatedcarbon manufacture [2,8,10,16]. Roy et al. [9] investigated that an-other potentially important end-use of the pyrolytic carbon blackmay be used as an additive for road bitumen. Furthermore, activecarbons were prepared from used tires and their characteristicswere investigated by Roy et al. [9], Zabaniotou and Stavropoulos[17], and Zabaniotou et al. [18]. They reported that active carbons,produced from tire chars, possess surface areas comparable withthose of commercially available active carbons (areas around1100 m2/g). Some of the previous researcher groups [2,4,8,11,19]studied the composition of evolved pyrolysis gas fraction andreported that it contains high concentrations of methane, ethane,butadiene, and other hydrocarbon gases with a GCV of approxi-mately 37 MJ/m3, sufcient to provide the energy required by thepyrolysis process.

A variety of scrap tires are available in the modern society.These are bicycle and rickshaw tires, motorcycle and auto-rick-shaw tires, car and taxi tires, microbus and jeep tires, tractor tires,bus and truck tires. Tires contain vulcanized rubber in addition tothe rubberized fabric with reinforcing textile cords, steel or fabricbelts, and steel-wire reinforcing beads [20]. Other components inthe tire are: carbon black, extender oil, which is a mixture of aro-matic hydrocarbons, sulphur, accelerator, typically an organo-sul-phur compound, zinc oxide, and stearic acid [2022]. There aremany different manufacturers and countless different formulationsavailable all over the world; the composition of the tire variesdepending on the tire grade and manufacturers. Consequently,the tire pyrolysis products may also vary in terms of yield andchemical composition depending on the source and grade of thetires. Ucar et al. [23] determined the polymer types in rubber con-tent of scrap truck and car tire wastes and reported that truck tirewaste contained natural rubber (NR) 51 wt.%, styrenebutadienerubber (SBR) 39 wt.%, and butadiene rubber (BR) 10 wt.% whilecar tire waste contained NR 35 wt.% and BR 65 wt.%. They alsofound signicant variations in proximate and ultimate analysesof the two different solid tire wastes and variations in pyrolyticproduct yields and liquids and gaseous products compositions.Kyari et al. [24] studied pyrolysis analysis of seven different brandsof used car tires from several countries throughout the world andcharacterized the product liquids obtained from individual andmixture of seven categories of tire wastes. They reported that therehad been signicant variation in concentration of different com-pounds presented in the derived liquids and gaseous products. Cy-pres and Bettens [21] have shown that pyrolysing different brandsof automotive tires results in signicant differences, of the order of10%, in the yields of solid, liquid, and gaseous products. For thedevelopment of the pyrolysis processes, there is a need to under-stand the inuence of different types of tire waste ranging frombicycle to truck tires available in different communities, on theyields and compositions of the derived products. There have beenmany pyrolysis woks [253] in international literature for automo-tive tire wastes but there is no work for the inuence of rubbercontent of motorcycle tires on the product yields and composi-tions. Thus, it is crucial to investigate the pyrolysis behaviors ofthe motorcycle tire wastes including pyrolytic product yields andproduct characteristics.

Very different experimental procedures have been used to obtain

M. Roqul Islam et al. /liquidproducts fromautomotive tirewastes bypyrolysis technologyincluding xed-bed reactors [3,4,8,11,2332,34], uidized-bedpyrolysis units [6,33,35], vacuum pyrolysis units [7,9,3639],investigated the effects of feed size and vapor residence time onproduct yields and compositions with respect to operating temper-ature for circulating uidized-bed and xed-bed reactor, respec-tively but poorly detailed. However, more understanding foroptimum operating conditions in relation to the range of reactortemperature, larger feed size and vapor residence time, whichmakeinuence on the yields and composition of the derived products areessential to develop a more efcient pyrolysis unit.

In thepresent study, toprovide anewapproach inheating systemfor xed-bed pyrolysis technology to the recovery of hydrocarbonsfrom used tires, we have investigated xed-bed pyrolysis with are-tube heating reactor. This process of heating has proved veryeffective in the technology for re-tube steam boilers. The thermalrecyclingofmotorcycle tirewastes bypyrolysis technologyhas beencarried out in the internally heated re-tube heating reactor systemunder N2 atmosphere. The effects of operating temperature, feedsize, and vapor residence time on the yields and compositions ofproduct liquids were investigated. A detailed characterization ofthe whole pyrolysis liquids obtained at optimum operating condi-tions has been carried out including physical properties, elementalanalyses, GCV, FT-IR, 1H NMR, GCMS analysis and distillation.

2. Materials and methods

2.1. Feed materials

The Indian made MRF brands of motorcycle tires, which aremostly consumed in Bangladesh, has been taken into considerationas feedstock throughout the experimental investigations and it wascollected locally from a dumpsite of the Rajshahi City Corporation.The main components of tires such as rubber, llers like carbonblack, steel, sulfur, zinc oxide, processing oil, vulcanization acceler-ators, etc. are heterogeneously distributed over the cross-section.Therefore, in order to maintain uniformity of the components inthe representative samples, very same tires were chopped cross-section wise into four different sizes of 8 cm 1 cm 0.25 cm = 2 cm3; 8 cm 1 cm 0.50 cm = 4 cm3; 8 cm 1 cm 1 cm = 8 cm3 and 8 cm 1 cm 1.50 cm = 12 cm3. The cross-sec-tion pieces, which were representative of the whole tire, containedno steel cords but the textile fabrics.spouted-bed reactors [40], etc., ranging from laboratory to commer-cial scale plants. Pyrolysis yields and characteristics of the productsobtained from tirewastes depend on type and size of feedstock, size,and system conguration of reactor, efciency of heat transfer, va-por residence time, etc. A xed-bed consists of individual feed parti-cles of different shapes and sizes, which are in contact with eachother or with the void in between. Maximum liquid yields are ob-tained with high heating rate, at reaction temperature around500 C and with short vapor residence time. The heat transfer raterequirement imposes particle size limitations on the feed. The costof size reduction in nancial and energy terms have been clearlyrealized and similarly liquid yields from the very smaller size feedis low due to the fact that too quick devolatilization occurs and sec-ondary reaction takes place in the reactor. Moreover, longer vaporresidence time contributes to secondary reactions, which leads toless oil and more gas product. A number of research works[3,4,8,23,25,27,31,32,37,5153] have been performed for the effectof reactor temperature on the product yields and product composi-tions but there have been very limited studies [6,10] in the interna-

87 (2008) 31123122 31132.1.1. Major composition of motorcycle tire rubberA representative sample of the whole tire waste has been pre-

pared grinding the steel cords and textile fabrics free rubber for

characterization of the tire pieces used in the pyrolysis experiment.The rubber powder was then screened by 9 mesh (or 2 mm). Pyrisdiamond thermogravimetric/differential thermal analyzer (TG/DTA) was used to obtain thermogravimetric analysis (TGA) and dif-ferential-thermogravimetric (DTG) data for the prepared samples.The samples (1520 mg) were heated over the temperature rangeof 30800 C at constant heating rates of 10 and 60 C/min in ahigh purity N2 atmosphere with a ow rate of 100 ml/min. Themost common rubbers used for tires are NR, SBR, and BR. The rub-bers mostly consist of blends of two or three rubbers together withtire additives. Several authors [22,4245] reported that the ther-mal degradation behavior of a solid tire waste gives informationabout the type of rubber contents itself. Williams and Besler [22]reported after exhaustive investigations that during thermaldecomposition, NR gives a peak at 375 C, SBR gives a peak at

3114 M. Roqul Islam et al. / Fuelaround 450 C whereas BR gives two peaks at 400 and 475 C fora heating rate of 5 C/min. Seidelt et al. [42] proved that at heatingrate of 10 C/min tire rubber DTG curves are characterized by theirpeak temperature: 378 C for NR, 430458 C for SBR and 468 Cfor BR. Similar results were also found by Liu et al. [45]. They inves-tigated the pyrolysis of NR, BR, and SBR, and reported that the max-imum weight loss rate of NR occurs at a temperature of 373 C, BRat 372 C and 460 C, SBR at 372 C and 429460 C. Little variationin peak temperatures for respective rubber component among dif-ferent studies is due to the variation in test apparatus and environ-ment used. Thus, it can be concluded that when a rubbercomposition contains NR, SBR, and BR components, peaks ataround 370 C, between 400 and 460 C and at around 460 C mustbe found in their DTG curve, respectively. The DTG plots obtainedat heating rates of 10 and 60 C/min for the representative samplesof motorcycle tire waste are presented in Fig. 1. There is a peak ataround 372 C and a dominant peak over 420450 C in the DTGplot for heating rate of 10 C/min indicating the presence of NRand SBR, respectively, in the investigated motorcycle tire waste.

2.1.2. Proximate and ultimate analyses and GCVs of motorcycle solidtire waste

The proximate and ultimate analyses and GCV of the motor-cycle tire feedstock compared to those of passenger car and trucktire wastes obtained by other authors are presented in Table 1.The table shows that volatile content is lower and ash content ishigher for the motorcycle tire waste compared to those of carand truck tire wastes. Ucar et al. [23] also found higher ash andlower volatile contents of their car tire waste. The variation ofash and volatile contents is due to the variation of inorganic mate-rials contents of the tire rubber such as zinc, clay, silica, etc. as rub-ber additives, which magnitudes are greatly depend onformulations and manufacturers. The ultimate analysis shows that

0 200 300 400 500 600 7000

102030405060708090

100

TG, (

1-X)

(wt%

)

1000

200

400

600

800

1000

1200

1400

TG & DTG at heating rate of 10C/min

TG & DTG at heating rate of 60C/min

DTG

, dX/

dt (

g/m

in)

Temperature(C)Fig. 1. TG and DTG plots for motorcycle tire wastes at hearting rates of 10 and60 C/min.the carbon present in the motorcycle tire wastes is comparativelylower than that in car and truck tire wastes. The oxygen content inthe motorcycle tire waste is not too high; the value presented inTable 1 include high amount of ash. Due to high presence of inor-ganic compounds, the GCV of the present motorcycle tire waste islower than that of passenger car and truck tire wastes.

2.2. Experimental section

The schematic diagram of the experimental set-up is presentedin Fig. 2. The experimental unit consists of eight major compo-nents: (1) a xed-bed re-tube heating reactor chamber with apower system; (2) a gravity feed type reactor feeder; (3) two ice-cooled condensers, each of them having a liquid collecting glassbottle; (4) a N2 gas cylinder with a pressure regulator, a ow con-trol valve and a gas ow meter; (5) a N2 gas pre-heater with LPGburner; (6) an air compressor; (7) char collecting bag; and (8) K-type (chromelalumel) thermocouples, whose measurement accu-racy is 2.5 C with a temperature controller. At a distance of30 mm from the closed bottom of the reactor, a distributor platewas tted to support the feedstock. The distributor plate was madeof stainless-steel plate having 150 holes of 3 mm diameter. The N2gas inlet was 20 mm below the distributor plate. Eight equallyspaced stainless steel, 10 mm diameter re-tubes containing insu-lated electric coil of a total capacity 1.60 kW were xed inside thereactor. The re-tubes and pre-heated N2 gas provided uniformheating across the cross-section of the reactor chamber. The reac-tor was thermally insolated with asbestos cylinder. The reactorheight from the distributor to the gas exit was 270 mm and itsdiameter was 100 mm, which provided an apparent vapor resi-dence time of 5 s. The sweeping gas ow rate or vapor residencetime for the re-tube heating reactor system was calculated bythe following equation:

Free space in the reactor for the flow of sweeping gas N2;V fsp

pd2l

4 npd

21l1

4

" # 1 Vm

100

pd

22l24

661 cm3 0:661 L

and the sweeping gas ow rates, vf = vfsp/t were 8, 4 and 2 L/min forresidence time, t = 5, 10 and 20 s, respectively. Where for the pres-ent reactor system: internal diameter of the reactor, d = 10 cm;effective length of the reactor, l = 27 cm; length of each re-tube,l1 = 27 cm; diameter of re-tube, d1 = 1 cm; number of re-tubes,n = 8; reactor volume occupied by feed materials, Vm = 70%; diame-ter of vapor outlet pipe, d2 = 2.54 cm; and length of vapor outletpipe (from reactor to condenser), l2 = 15 cm.

The vapor residence time (5, 10, and 20 s) was calculated atroom temperature. During experiments the pressure in the owmeter and reactor chamber was same but slightly higher than thatof atmospheric just to maintain continuous ow of N2 gas. Temper-ature in the ow meter was near about ambient and its values in-side the reactor chamber were measured by the thermocouples.According to the gas law (Boil and Churls combined law) the owrate in the reactor varies with pyrolysis temperature. When theow meter readings show calculated ow rates: 8, 4, and 2 L/min, the ow rates in the reactor chamber become about 2.5 timesof the calculated values at a reaction temperature of 475 C andconsequently the apparent vapor residence time reduces to 2, 4,and 8 s, respectively. The ow rate inside the reactor chamber alsoincreases slightly due to addition of pyrolysis vapor products fromtire sample to the N2 gas during decomposition of tire bed. A owcontrol valve varied the N2 ow rate from the cylinder as well as

87 (2008) 31123122vapor residence time for different test.In each pyrolysis run, by the action of gravity force a quantity of

750(2.0) gm of tire sample was supplied from the feeder into the

tho

Cu

162297

FuelTable 1Proximate and ultimate analyses and GCV of solid motorcycle tire waste compared to

Type of used tire Present study (motorcycle tire)

Proximate analysis (as received, wt.%)Moisture content 1.53Volatile 57.50Fixed carbon 20.85Ash content 20.12

M. Roqul Islam et al. /reactor chamber by opening the feed control valve. Then the reac-tor was purged before experiments by the ow of N2 gas of 4 L/minfor 5 min to remove the inside air. The reactor heater (power sys-tem) and LPG burner were switched on, and the temperature of thereactor was increased to a desired value of 375, 425, 475, 525 or575 C. During pyrolysis of tire mass a reddish/bright brown visiblevapor usually ared into the atmosphere. When the decompositionwas completed colorless (no visible) N2 gas came out from thereactor. The colorless aring was the signicance of the completionof the thermal decomposition of the tire bed. The reaction time was50 min for every of all pyrolysis runs after which usually no further

Ultimate analysis (dry, wt.%)Carbon (C) 75.50 86Hydrogen (H) 6.75 8Nitrogen (N) 0.81 0Sulfur (S) 1.44 1Oxygen (O)a /othersb 15.50b 3Ash 2GCV (MJ/kg) 29.18 40

H/C molar ratio for motorcycle tire waste is 1.07.a Calculated by difference (included: oxygen only).b Calculated by difference (included: oxygen + ash).

N2

gas

Liquidcollector 1

Aircompressor

Condenser 2

Condenser 1

Liquidcollector 2

Flaring

N2 G

as cylinder

Reactor feeder

Flow m

eter

Reactor cham

ber

LPG burner

Pressure regulatorwith flow control valve

Charcollection bag

Fig. 2. Schematic diagram of the xed-bed re-tube heating pyrolysis system.visible vapor product came out. Nitrogen gas was supplied in orderto maintain the inert atmosphere in the reactor and also to sweepaway the pyrolyzed vapor product to the condensers. Pyrolysis va-por product was passed through two sets of condenser pipes toquench into liquid and then collected in the glass bottles. Theuncondensed gases were ared to the atmosphere. The bottleswere completely lled up with liquid so that no air could betrapped into the bottles. When pyrolysis of the feed material inthe reactor was completed, the vapor exit port was closed andthe reactor heater and LPG burner were switched off. N2 gas supplywas also stopped. After cooling down the system, the char productwas pushed out from the reactor chamber with the aid of com-pressed air supplied from the air compressor by opening char exitport. Char was collected in the char collection bag and weighed.The liquid was then weighed and gas weight was determined bysubtracting the liquid and char weight from the total weight offeedstock. Afterwards, the system had been made ready for thenext run just repositioning the valves. Initially the pyrolysis exper-iments were performed by varying the temperature within therange of 375575 C at every 50 C for a particular feed size and va-por residence time. Once the temperature of maximum liquid yield

se of passenger car and truck tire wastes

nliffe and Williams [8] (car tire) Ucar et al. [23]

Car tire Truck tire

.30 1.60 1.40

.20 58.20 66.10

.40 21.30 27.50

.10 18.90 5.00

.40 74.30 83.20

.00 7.20 7.70

.50 0.90 1.50

.70 1.71 1.44

.40a 15.89b 6.16b

.40

.00 30.50 33.40

87 (2008) 31123122 3115(475 C) was selected, additional tests were conducted at the opti-mum temperature by varying the feed size and the vapor residencetime to nd out the optimum process conditions. Before analyzing,the liquid product was centrifuged at 3000 rpm for 15 min to re-move heavy condensate and impurities.

2.3. Pyrolytic product liquid analysis

Pyrolytic liquids obtained under the maximum liquid yield con-ditions were well mixed and homogenized prior to analysis beingmade. Some physical properties of pyrolytic liquids: density, vis-cosity, ash point, pour point, and GCV were determined by usingthe following standard methods: ASTM D189, ASTM D445, ASTMD92, ASTM D97, and ASTM D240, respectively. Elemental analysis(C, H, N, and S) of liquids was determined with an elemental ana-lyzer of model EA 1108, which followed the quantitative dynamicash combustion method. The boiling point range distribution ofhydrocarbons in the pyrolytic liquids was determined by usingThermo-gravimetric Analyzer of model SHIMADZU TGA-50 accord-ing to ASTM D2887-89 standard test method. The sample (1520 mg) was heated from ambient temperature to 600 C at a heat-ing rate of 10 C/min in a high purity helium atmosphere at a owrate of 100 ml/min. The data obtained from TGA was used to eval-uate the simulated distillation curves. The functional group com-

positions of the product liquids were analyzed by FT-IR spectros-copy. The FT-IR instrument of model PERKIN ELMER FT-IR 2000was used to produce the IR-spectra of the derived liquids. Identi-cation of compounds in pyrolytic liquids was carried out by a gaschromatograph mass spectrometer of model GCMS-QP5000. Theanalysis was performed on a 60 m 0.32 mm capillary columncoated with a 1 lm lm of DB-1. The oven temperature was pro-grammed, 40 C hold for 10 min to 300 C at 5 C/min hold for10 min. Compounds were identied by means of the NIST12 andNIST62 library of mass spectra and subsets HP G1033A. The 1HNMR analyses of pyrolytic liquids were recorded at a frequencyof 500 MHz with an instrument of model JEOL A-500 using CDCl3as solvent.

3116 M. Roqul Islam et al. / Fuel3. Results and discussion

3.1. Pyrolysis product yields

Three types of products are usually obtained from pyrolysis oftire rubber: solid char, liquid, and gas. The product distributionsobtained from pyrolysis of prepared representative sample fortemperature range of 375575 C at every 50 C, feedstock size of4 cm3 and vapor residence time of 5 s are presented in Fig. 3. Thepyrolysis of the experiments was repeated three times. The pyroly-sis yields presented are the mean value of three equivalent exper-iments and the standard deviation for the liquid yields data were2.3, 2, 1.3, 1.5, and 1.4 for ve corresponding temperatures.Fig. 3 shows that when the temperature increases from 375 to575 C, the yield of liquid increases rst from 42 2.3% to a maxi-mum value of 49 1.3% of feed weight at 475 C, and then de-creases to 42 1.4%. The gas yield increases from 8 1.5 to18 1.5 wt.% over the whole temperature range, while char yielddecreases from 50 2.8 to 41 1.5 wt.% at 475 C and then remainsalmost constant. It is apparent that a fairly sharp optimum exists intemperature at which maximum yield of liquid was achieved prob-ably due to strong cracking of tire rubber at this temperature. Tirerubber is not totally decomposed (pyrolysis is not complete) at atemperature less then 475 C. The thermal decomposition of thetire rubber increases (i.e. solid yield decreases) up to a temperatureof 475 C and hence increasing the liquid and gas yields. In thetemperature range of 475575 C, solid yields are almost equiva-lent to those of the values obtained by TGA of tire rubber(37 wt.%) at 475 C presented in Fig. 1. The solid char consists ofcarbon black, some solid hydrocarbons, and lower amounts of tirerubber additives such as zinc, sulfur, clays, and silica, or metal oxi-des [8,46]. Therefore, there is no obvious mechanism for char losswith increasing temperature, excepting only the higher tempera-ture volatilizes some of the solid hydrocarbons content of the char

0

10

20

30

40

50

60

375 425 475 525 575Temperature (C)

Prod

uct y

ield

s (w

t%) Liquids

Gas

CharFig. 3. Effect of temperature on product yields for feed size of 4 cm3 and vaporresidence time of 5 s.[8,27]. Competing with char loss reactions, certain amount of charor coke like carbonaceous material is formed in the pyrolysis ofmany polymeric materials, due to secondary repolimerizationreactions among the polymer derived products [4750]. Since solidyields do not decrease in the temperature range of 475575 C, itmay be concluded that tire decomposition is completed and carbo-naceous material has been formed. The decrease in liquid yieldsand increase in gas yields above the temperature of 475 C areprobably due to the decomposition of some oil vapors into perma-nent gases [6], and secondary repolimerization and/or carboniza-tion reactions of oil hydrocarbons into char [8]. The increase ingas yields in higher temperatures is also provided by the char lossreactions. Thus, at higher temperatures the gas yields becomegradually dominating. However, 475 C seems to be optimum tem-perature to obtain liquid product from thermal recycling of motor-cycle tire rubber by pyrolysis technology, since decomposition iscomplete and the liquid yields become maximized at thistemperature.

Pyrolysis product yields and their distributions over the wholerange of temperature depend not only on the feedstock composi-tion and operating temperature used for the experiments, but alsoon the specic characteristics of the system used, such as size andtype of reactor, efciency of heat transfer from the hot reactor sur-face to and within the tire mass, feed particle size, vapor residencetime, etc. Therefore, results from different authors are many timesdifcult to compare. Moreover sometimes contradictory data canbe found in the literature. Some of the former research groups[4,6,32,34] have been found very similar trends in product distri-butions to those of the present study, that is, solid yield decreasedwhile gas and liquid yields increased up to an intermediate tem-perature and then solid yield remained almost constant whilegas yield increased and liquid yield decreased with further increasein temperature. Gonzalez et al. [4] pyrolyzed automobile tirewastes of particle size of 0.21.6 mm in diameter in N2 atmospherefor both isothermal and non-isothermal cases. The temperaturerange studied was 350700 C in isothermal regime and the non-isothermal experiments were performed with a heating rate of520 C/min until a nal temperature of 600 C. Run times were30 min for their isothermal experiments. The maximum liquidand char yields were 55.6 and 37.6 wt.% at 550 C for isothermalrun, and 55.4 and 39 wt.% at heating rate of 15 C/min for non-iso-thermal run. Dai et al. [6] studied pyrolysis of automotive tirewaste in a circulating uidized-bed reactor for a temperature rangeof 360810 C, feed size of 0.320.8 mm and residence time of 15 s and reported that the optimum conditions were: 500 C,0.32 mm and 1 s, respectively, with liquid yields of 50% and charyields of 32%. Chang [32] pyrolyzed tire wastes of particle size2 mm in diameter with N2 atmosphere over a temperature rangeof 200800 C in a reactor, whose structural construction was sim-ilar to the Dupont TGA of model V2.2A-9900. He reported that theyield of liquid had a maximum level up to 40 wt.% at 350 C with achar yield of 17 wt.%. Islam et al. [34] pyrolyzed automotive scraptire in an externally heated xed-bed reactor with N2 atmospherefor feed sizes of 01 cm, 12 cm and 35 cm, a temperature rangeof 350550 C and reaction time of 25110 min. A cylindrical bio-mass source heater was used to heat the reactor and the gas-pre-heating chamber. They controlled the temperature of the reactorby varying the supply of air by means of an air blower. A maximumamount of liquid product was obtained at 450 C for a sample sizeof 35 cm and reaction time of 75 min. Similarly Cunliffe and Wil-liams [8] found that oil yield decreased from 58.1 to 53.1 wt.% andgas yield increased over a temperature range of 450600 C whilechar yields remains almost constant with a mean value of

87 (2008) 3112312237.8 wt.%. The decreases in oil yield with increasing temperatureand corresponding increase in gas yield have also been found byother workers [5153].

consequently the vapor residence time inversely decreased, in-creased the rate of removal of pyrolysis vapor products from tirefeed from hot zone. When the vapor residence time increases from

0

10

20

30

40

50

60

2 4 6 8 10 12

Feed size (cm3)

Prod

uct y

ield

s (w

t%) Liquids

Char

Gas

0

10

20

30

40

50

60

5 10 15 20Residence time (sec)

Prod

uct y

ield

s (w

t%) Liquids

Char

Gas

Fuel 87 (2008) 31123122 3117On the other hand Ucar et al. [23] pyrolyzed passenger car tireand truck tire separately in a xed-bed reactor at the temperaturesof 550, 650, and 800 C under N2 atmosphere and found that prod-uct distributions for both feed did not change signicantly withincreasing temperature. They obtained maximum liquid yields of48.4 2.4 and 56 1.8 wt.%; with char yields of 41.7 2.7 and33.8 2.4 wt.% at 650 C for car and truck tire feeds, respectively.Similar results have been reported by other researchers. Laresgoitiet al. [3,25] carried out pyrolysis tests in an unstirred stainless steel3.5 dm3 autoclave from 300 to 700 C every 100 C, and found thatsolid yield decreased while gas and liquid yields increased from300 to 500 C, but not from 500 to 700 C. In their studies, the max-imum liquid yield was 38 1.8 wt.% at 500 C with a char yield of44.8 0.6 wt.%. Roy et al. [37] studied tire pyrolysis in the range350700 C, and found no inuence of temperature on pyrolysisyields over 500 C. The thermochemical conversion of rubber fromscrap tires by pyrolysis and hydropyrolysis has been studied byMastral et al. [31]. They observed that neither the total conversionnor liquid yield increased with increasing temperature above500 C for the pyrolysis at a heating rate of 300 C/min. On the con-trary Williams et al. [27] studied pyrolysis from 300 to 720 C andobserved that char and gas yields decreased while liquid yield in-creased in the whole temperature range, although the effect oftemperature was slighter over 600 C, and varied depending onthe heating rate used. Barbooti et al. [10] pyrolyzed scrap automo-tive tire particles of size 220 mm, under a temperature range of400460 C and with a N2 gas ow rate of 0.20.5 m3/h in axed-bed batch reactor. They also observed that the solid yieldssharply increased and liquid yields decreased for all size particlesuntil a temperature of 430 C. The char yield for smaller particlesdecreased while the yield for relatively larger particles (16 and20 mm) almost remained constant with further increase in tem-perature (450460 C). In the temperature range of 430450 C,the yield of liquid was almost constant, but further increase oftemperature from 450 to 460 C caused fast increase in liquid yieldfor all particle size. The gas yields decreased over the whole rangeof temperature. Under their system conguration the optimumconditions were: reactor temperature 430 C, gas ow rate0.35 m3/h and particle size 10 mmwith product yields of solid charand pyrolytic liquid 32.5 and 51 wt.% respectively.

The liquid yields obtained in the present study are lower andchar yields are higher compared to those of the previous studies[4,6,8,10,27,34,37,52,53] due to mainly the compositional differ-ence between motorcycle tire and heavy automotive tire feeds.The motorcycle tire waste, which was used as feedstock in thepresent study, has lower volatile and higher ash content than thoseof heavy automotive tire wastes result in lower liquid yields. Ucaret al. [23] also found lower amount of oil and higher amount ofchar from pyrolysis of car tire waste than those from truck tirewaste due to the same compositional difference between car andtruck tire wastes. In the same way, the authors of this paper havealso found higher amount of liquid and lower amount of char frompyrolysis of heavy automotive tire waste compared to those frommotorcycle tire waste due to variation in composition of the twotire feeds [30]. Although the feed used in the present study con-tained higher amount of ash and lower amount of volatile, the li-quid yields were equivalent or somewhat higher and char yieldswere equivalent or somewhat lower than those obtained by otherresearch groups [3,25,31,32,51] because of variations mainly in thedesign of reactors and operating conditions used.

The effect of feed size on product yields under optimum reactortemperature and for a vapor residence time of 5 s is presented inFig. 4. The gure shows that liquid yield rst slightly increases

3

M. Roqul Islam et al. /up to a value 49 1.3 wt.% for feed size of 4 cm and then decreasesfor larger feed size while the char yield increases and gas yield de-creases through all the piece sizes from 2 to 12 cm3. The thermalconductivity of organic solid pieces is very poor (0.1 W/m2 C alongthe grain, 0.05 W/m2 C cross grain). During pyrolysis, a high heat-ing rate (claimed up to 10,000 C/s) may be achieved at thin reac-tion layer but the lower thermal conductivity of organic solid pieceprevents such heating rate through the whole piece [54]. Smallerfeed size provides more reaction surface causes high heating rateand too quick decomposition of the rubber feed occurs. The prod-uct oil vapors comparatively get enough time for secondary reac-tion in the reactor and consequently increase in gas yields anddecrease in liquid and char yields. On the other hand, the heatingrate in larger tire feed is low due to its lower thermal conductivityand heat can ow only to a certain depth in the available pyrolysistime compared to almost complete thermal decomposition of thesmaller pieces. Thus, the rubber core of the larger pieces becomescarbonized and/or cannot be decomposed completely resulting in-crease in char yields and decrease in liquid and gas yields. Thepresent results are in agreement with the results of Barbootiet al. [10], where the core of large pieces was protected from theheating source. Similarly Dai et al. [6] also found that the char yieldof the 0.8 mm tire sample was higher, and the gas and oil yieldswere lower, than those of the 0.32 mm sample. In the presentstudy it may be concluded that the optimum feed size is 4 cm3

for which decomposition of tire pieces is complete and has lesspossibility of secondary cracking at the optimum reactor tempera-ture and for a vapor residence time of 5 s.

The effects of vapor residence time on gas, char, and liquidyields for optimum reactor temperature and for optimum feed sizeare shown in Fig. 5. Increasing the ow of N2 gas from cylinder,

Fig. 4. Effect of feed size on product yields for 475 C and vapor residence time of5 s.Fig. 5. Effect of vapor residence time on product yields for 475 C and feed size of4 cm3.

5 to 20 s (i.e. N2 gas ow decreased from 8 to 2 L/min), the liquidand char yields decrease while the gas yield increases slightly.The increase in gas yields with increasing vapor residence timein the present investigation is due to the decomposition of someoil vapor into secondary permanent gases. Primary vapors are rstproduced from pyrolysis of tire rubber at optimum reactor temper-ature and the primary oil vapors then degrade to secondary gases{for instance: oil vapors? heavy hydrocarbons + light hydrocar-bons (CH4 + C2H4 + C3H6 + ) + CO + CO2 + H2} within the periodof higher vapor residence time, which leads to less oils and moregaseous products. Besides, long contact time between the volatilesand the char leads to another parallel secondary pyrolysis reaction(for instance: C + CO2? 2CO) and hence reduces in char yields[55]. The results of the present work, that is, the decrease in liquidand char yields, and increase in the gas yields with increasing va-por residence time are in good agreement with those of the somepublished papers [6,10,21]. It should be noted that the true resi-dence time of the volatiles is shorter than the values given inFig. 5 because the total gas ow increases as the tire mass decom-poses and as N2 gas expands at the reactor temperature.

3118 M. Roqul Islam et al. / Fuel3.2. Analysis of product liquids

3.2.1. Fuel properties of the liquidsThe pyrolytic liquids obtained from pyrolysis of motorcycle tire

wastes, which are oily organic compounds, appears dark brownwith a strong acrid smell. Careful handling of the liquids is requiredsince it reacts easily with human skins, leaving permanent yellow-ish brown marks and an acrid smell for a few days, which is dif-cult to remove by detergent. No phase separation was found totake place in the storage bottles. The liquids were characterizedin terms of both fuel properties and chemical compositions. Thefuel properties of the pyrolytic liquids in comparison to commer-cial automotive No. 2 diesel, which is mostly consumed in Bangla-desh, are shown in Table 2. It shows that the density of pyrolyticliquids was found higher than that of the commercial diesel fuelbut lower than that of heavy fuel oil (980 kg/m3 at 20 C). The vis-cosity of liquid products from motorcycle tire wastes was slightlyhigher than that of the No. 2 diesel but too much lower than thatof heavy fuel oil (200 cSt at 50 C). Low viscosity of the liquids of4.75 cSt at 30 C is a favorable feature in the handling and trans-porting of the liquid. The ash point of a liquid fuel is the temper-

Table 2Characteristics of the pyrolytic liquids in comparison to petroleum products

Analyses Motorcycle tirederived liquids

Commercial automotiveNo. 2 diesel

Elemental (wt.%)C 85.86 8487H 9.15 12.8015.70C/H 9.38 5.356.80N 0.65 653000 ppmS 1.25 11007000 ppma

Ash 0.22 0.0O 2.87 0.0H/C molar ratio 1.27 1.762.24O/C molar ratio 0.025 Empirical formula CH1.27O0.025N0.006 Density (kg/m3) 957 820860Viscosity (cSt) 4.75b 2.04.5c

Flash point (C) 632 >55Pour point (C) 6 40 to 30Moisture (wt.%) N/A 80 ppmpH value 4.40 GCV (MJ/kg) 42.00 44.0046.00a Valid for legislation requirements of Bangladesh.b At a temperature of 30 C.c At a temperature of 40 C.ature at which the oil begins to evolve vapors in sufcient quantityto form a ammable mixture with air. The temperature is an indi-rect measure of volatility and serves as an indication of the rehazards associated with storage and application of the fuel. Theash point of the tire-derived liquids was 632 C. The ash pointis low when compared with petroleum-rened fuels; for example,kerosene has a required minimum ash point of 23 C, diesel fuelof 55 C and light fuel oil 79 C. The low ash points of the tire-de-rived liquids are not surprising since the product liquids representsun-rened liquids with a mixture of components having a widedistillation range. The pour point of the tire-derived liquids is com-paratively low compared to the automotive diesel fuel but the lab-oratory experience of the authors of the present paper shows thatit is not problematic even at 7 C. The pH value of the pyrolytic liq-uids is 45, which is in weak acidic nature. It is found that there isvery little contamination of the liquids with metals (V, Mn, Mg, Ba,Ni, Ti, Cu, Cr, Cd, Co, Fe, Al, and Zn) [9], and does not contaminatewith glass and PET plastic and/or other plastics [30]. The pH valueof soft drinks like Cola and Pepsi of Coca Cola company is 2.5 andthey use PET plastic bottles for its storage and handlings. Thus,storage and handling of the liquids are little problematic in indus-trial usage in this regard.

3.2.2. Chemical composition of the liquid productsCondensable liquids are the major products of solid motorcycle

tire wastes pyrolysis. Since the liquids consist of numerous and di-verse components, it is difcult to quantify them. All most all of theresearchers have been using elemental analysis, FT-IR, NMR, andGCMS modern analytical techniques to identify and to quantifypossible compounds in the pyrolytic liquids derived from differenttypes of organic solid wastes.

Elemental analysis, H/C and O/C molar ratios, empirical formulaand GCVs for the liquids are listed in Table 2. The average chemicalcomposition of the pyrolytic liquid has been analyzed asCH1.27O0.025N0.006. Obviously, the product liquids contain a smallamount of oxygen content, with a higher H/C ratio than that of so-lid tire wastes (Table 1). The H/C ratio of the pyrolytic liquids (Ta-ble 2) indicates that such oils are a mixture of aliphatic andaromatic compounds [3]. The C/H ratio is somewhat higher thanthat found for petroleum derived fuels. The long chain diesel likestructure attributes to high C/H ratio as reected by very highGCVs, comparable to fuel oil and diesel, as well as indicates thehigh miscibility with the diesel fuel as per like dissolves like prin-ciple. The nitrogen content is however, rather higher than the No.2diesel but similar to a heavy fuel oil in the range of 0.30.5 wt.%.The sulphur content in the derived liquids is signicantly highcompared to the legislation requirements of commercial diesel fuelin Bangladesh (5000 ppm or more) to safe environment. Almostcoincidental results concerning physical properties, ultimate anal-ysis and GCVs of tire pyrolysis liquids were obtained by previousstudies [3,8,9,23,30,34] for heavy automotive tire wastes. Someof the previous studies [3,8,30] also show that there is no effectof pyrolysis temperature on the physical properties, ultimate anal-ysis and GVCs of the product liquids. The important requirementsfor diesel fuel are its ignition quality, viscosity, water, sediment,and sulfur contents. Therefore, the pyrolytic liquids require preli-minary treatments such as decanting, centrifugation, ltration,desulphurization, and hydrotreating to be used as fuels. The trea-ted pyrolysis oil could be used directly as fuel oils or blended withdiesel fuels, which will reduce the viscosity and, increase the pHvalue and ash point of the resulting blends. Consequently, theatomization will be improved, ensuring a complete burnout ofthe fuel [9]. Based on its fuel properties, tire-derived pyrolytic liq-

87 (2008) 31123122uids may be considered as a valuable component for use with auto-motive diesel fuels. Moreover, the liquids may be directly used asfuels for industrial furnaces, power plants, and boilers.

The FT-IR is not the most appropriate analytical tool to deter-mine saturated, aromatic, and polar components. Nevertheless, itallows functional group analysis to reveal the chemical propertiesof the liquids. The FT-IR analysis for pyrolytic liquids derived frommotorcycle tire waste have been carried out and the results ob-tained from the transmittance spectrums compared to those of dif-ferent authors are presented in Table 3. The data shows thereforethat the present liquids contain mainly aliphatic and aromaticcompounds.

The 1H NMR has been performed for the pyrolytic liquids andthe hydrogen distribution obtained from the 1H NMR spectrumsis given in Table 4, indicating that no aliphatic carbon is still boundto oxygen (peaks in 3.34.5 ppm chemical shift range). Clearly, the

ent motorcycle tire composed of mainly NR and SBR. The truck tirewaste [23] was also contained mainly NR and SBR with smallamount of BR. It is known that limonene is the most important

Functional groups Class of compounds

OH stretching Alcohols, phenols or carboxylic acidsCH stretching Aromatic compoundsC@C stretching AlkenesCH stretching AlkanesC@O stretching Aldehydes or ketonesC@C stretching AlkenesCarboncarbon stretching Aromatic compoundsCH bending AlkanesCH in-plane bending Aromatic compounds

Table 41H NMR results for the product liquids

Type of hydrogen Chemicalshift (ppm)

Mol% (% oftotal hydrogen)

Aromatic 9.06.5 13.31Phenolic (OH) or olenic proton 6.54.5 9.74Aliphatic adjacent to oxygen/hydroxyl group 4.53.3 Aliphatic adjacent to aromatic/alkene group 3.31.8 17.55Other aliphatic (bonded to aliphatic only) 1.80.4 59.40

M. Roqul Islam et al. / Fuel 87 (2008) 31123122 3119main structure of the liquids seems to be aliphatic bonded to ali-phatic only (0.41.8 ppm chemical shift range), and as a resultthe carbon aromaticity of the liquids is comparatively low. The al-kanes and long alkyl spectrums are probably largely derived fromsolid tire wastes.

GCMS analysis were carried out with the pyrolysis liquids ob-tained at 475 C temperature for feed size of 4 cm3, and apparentvapor residence time of 5 s to get an idea about the nature and typeof compounds of such liquids. The NIST search software was usedto analyze the peaks provided by the chromatogram, from whichmore than the half was not properly identied. The peaks whoseidentication results of mach quality P90% were considered validand their tentative assignment were conrmed where in agree-ment the published GCMS data for similar products [3,23,24].

Table 5 shows the tentative compounds assigned and their per-centage area compared to the total area of the chromatogram,which gives an estimate for their relative concentration in thepyrolytic liquids. It can be seen that, tire pyrolysis liquids are a verycomplex mixture, containing many aliphatic and aromatic com-pounds with their total concentration of 49.54% and 16.65%,respectively. The GCMS results support well the results obtainedfrom FT-IR (Table 3) and 1H NMR (Table 4) analyses. The aliphaticcompounds are mainly of alkane and alkene groups but the secondis predominant (43.23%). The aromatic compounds are only singlering alkyl aromatics. In addition to the main hydrocarbons, smallpercentage of nitrogen containing compound, 1-azido-2-methyl-benzene (C7H7N3); nitrogen, oxygen, and sulphur-containingcompound, 5-methyl-6-phenyltetrahydro-1,3-oxazine-2-thione(C11H13NOS); and chlorine-containing compound, parachlorophe-nol (C6H5ClO); octanoyl-chloride (C8H15ClO) and 1-chloro-dode-cane (C12H25Cl) were also identied. Other oxygen-containingcompound in the form of acid such as 3-methyl-2-pentanoic acidis also present in the motorcycle tire derived liquids. The mostabundant compound present in the pyrolytic liquids is limonenewhose total concentration is 29.54%. Ucar et al. [23] also foundsimilar results from their truck tire pyrolyis liquids obtained at650 C, where the concentration of limonene was 28.78%. The pres-

Table 3The FT-IR functional groups and the indicated compounds of pyrolysis liquids

Frequency range (cm1)

Present studies Gonzalez et al. [4] Williams et al. [27]

36003250 3250-3100 31003005 30002800 30002800 30002800 1750165016751605 16701580 1675157516001545 16001500 1625157515201115 14801360 14751350 11501000

1020845 960870 950875810660 900675 900675product form pyrolysis of polyisoprene [7,31]. Beside formationof limonene depends on operating conditions such as pyrolysispressure, temperature, vapor residence time and sample size. Pak-dal et al. [7] studied in detail the optimum operating conditions forthe production of limonene from used tires by vacuum pyrolysis.They reported that low pyrolysis pressure and temperature, andshort vapor residence time increase the limonene yields. At thereaction conditions, the polyisoprene part of the rubber thermallydecomposes through b-scission mechanism to an isoprene inter-mediate radical. It is then transformed to isoprene (depropaga-tion). The isoprene molecules in the gas phase dimerise to dl-limonene following a low energy reaction mechanism. The forma-tion of dl-limonene by intermolecular cyclization is also possible.Mastral et al. [31] also reported that at mild reaction conditionsthe polyisoprene is depolymerised forming dimeric species. As thisactivation giving the diradical species takes place in the absence ofoxygen, a cyclization could be produced. This dimmer species, ashort-life radical, could be stabilized through a two-step processdriving to limonene by pyrolytic isomerization.

Rodriguez et al. [2] carried out the pyrolysis of automotive tiresin a xed-bed reactor at 500 C and reported that product oils con-sisted of 62.4 wt.% aromatic compounds, 31.6 wt.% aliphatic com-pounds, 4.2 wt.% nitrogen-containing compounds, and 1.8 wt.%sulfur-containing compounds. Laresgoiti et al. [3] have been givenan exhaustive effort to characterize total pyrolytic product oilsfrom car tire wastes obtained at every 300, 400, 500, 600, and700 C and reported that tire pyrolysis oil is mainly aromatic andaliphatic in composition but with nitrogen, sulfur, and oxygenatedspecies present. They have been found that at every temperaturethe tire pyrolysis oils were predominately aromatic. They alsofound a range of compounds in tire pyrolysis oils, showing thatthe oil was composed mainly of alkylated benzenes, naphthalenes,and phenanthrenes, n-alkanes from C11 to C24, and alkenes from C8to C15. Kyari et al. [24] carried out GCMS of total pyrolysis liquidsobtained form pyrolysis of a mixture of seven brands of used cartire feeds at 500 C. They reported that the oil contained mainlyaromatic and aliphatic compounds such as alkylated benzenes,C@C stretching AlkenesCH out-of-plane bending Aromatic compounds

compounds but Lersgoiti et al. [3] suggests that when the reactortemperature is raised, non-aromatic compounds link to aromaticstructures yielding as a result a higher proportion of aromatic com-pounds. In the present study, motorcycle tire feed contains signif-icant amount of NR and mild reaction temperature and apparentvapor residence time, and larger feed size have been used duringpyrolysis operation. Thus it can be concluded, taking into consider-ation the above discussions that the present feed material andoperating conditions are very favorable for formation of alkenesin large quantities, especially limonene. The signicant percentageof aromatic compounds in the present study is most probably dueto the presence of SBR in the tire feed and ciclation of product ole-nic compounds through dehydrogenation reactions at the pyroly-sis conditions.

The present pyrolytic liquids as well as the previous studies[2,3,24] contain small amount of oxygenated, nitrogenated, and ni-

Fuel 87 (2008) 31123122Table 5Tentative GC/MS characterization of motorcycle tire-derived pyrolytic liquids

Retention time(min)

Tentative assignment Peak area(%)

7.58 2-Methyl-1,3-butadiene 1.397.98 2-Methyl-2-butene 0.2913.36 Benzene 0.1318.71 Toluene 6.0323.08 2-Methyl-2-hepten-4-yne 0.3823.57 p-Xylene 3.1423.78 m-Xylene 0.9223.82 o-Xylene 1.2424.62 1-Azido-2-methyl-benzene 3.4527.81 1,5-Dimethyl-1,5-cyclooctadiene 4.8528.24 (E,Z)-1,5-Cyclodecadiene 2.1429.03 1,1-Dimethyl-2-(2,4-pentadienyl)-

cyclopropane3.28

29.15 b-Myrcene 2.2829.46 (1S)-6,6-Dimethyl-2-methyl-bicyclo 3.1.1

heptane0.61

29.54 (Z)-3,7-Dimethyl-1,3,6-octatriene 0.4429.62 3,7-Dimethyl-1,3,7-octatriene 0.4129.66 2,6-Dimethyl-1,5,7-octatriene 0.9630.11 2,5-Dimethyl-3-methylene-1,5-heptadiene 0.8130.82 Limonene 21.7631.12 Limonene 1.5231.25 Limonene 6.2631.38 (E)-3-Nonen-1-yne 0.4031.61 1-Propynyl-benzene 0.96

3120 M. Roqul Islam et al. /alkylated naphthalenes, alkanes, and alkenes. They found ethyl-,propyl-, butyl-, pentyl-, hexyl-, and heptylbenzene and ethenyl-,propenyl-, and butenylbenzenes in high concentrations. Naphtha-lene and methyl, dimethyl, and trimethyl naphthalenes were pres-ent in their oils in signicant concentrations. The identied alkaneswere pentadecane, heptadecane, octadecane, nonadecane, eico-sane, and some other high molecular weight alkanes. In additionto the main hydrocarbons, sulfur-, and nitrogen-containing com-pounds were also identied in their product oils. Moreover, somexygenated compounds in the form of acids such as hexadecanoic,heptadecanoic, and octadecanoic acids were identied.

The previous research groups [2,3,8,23,24] have been found aro-matic compounds in higher concentrations in their car tire-derivedliquids than aliphatic compounds but in the present study the ali-phatic compounds are found predominately. This is due to the dif-ferences in tire rubber compositions and operating conditionsused. The most common tire rubber formulations use 60% SBRfor manufacturing of car tire [17] and 5060% NR for manufactur-ing of truck tire [7]. The aromatic nature of the tire oils is due: (i) tothe aromatic nature of the source polymeric materials SBR, whichalready contains aromatic rings and hence chain splitting may eas-ily lead to the formation of aryl chain fragments and (ii) to ciclationof olen structures through dehydrogenation reaction, which hap-pen during pyrolysis process. Several of the former research groups[2,3,7,8,21,52] have been proved that the proportion of aromaticsincreases with pyrolysis temperature. It is known that formationof aromatic at high temperature is due to the expense of aliphatic

32.81 5-Methyl-6-phenyltetrahydro-1,3-oxazine-2-thione

0.80

32.89 2-Butenyl-benzene 0.7835.28 1,3-Decadiyne 0.4135.34 1,4,5,8-Tetrahydro-naphthalene 0.1235.40 3-Undecene-1,5-diyne 0.5736.55 3-Meththyl-2-pentenoic acid 0.4336.70 Parachlorophenol 0.5837.80 (Z)-3-Decen-1-yne 0.2138.04 (Z)-3-Dodecen-1-yne 0.2739.29 Octanoyl-chloride 0.3539.46 1-Chloro-dodecane 0.65

Total identied 69.00tro-sulphurated compounds. The oxygenated compounds are prob-ably derived from the thermal degradation of oxygenatedcomponents of the tire, such as stearic acid, extender oils, etc.The presence of nitrogenated and nitro-sulphurated compoundsmay be explained by thermal degradation of the accelerators usedin tire compounding, which are frequently sulfur- and/or nitrogen-based organic compounds, such as 2-mercaptobenzothiozoe, ben-zothiozolyl disulphide, etc. The presence of chorine containingcompounds in the present pyrolytic liquids may be explained bythermal decomposition of sulphur monochloride, which sometimeis used as vulcanizing agent with main vulcanizing agent, sulphur.The greater non-identied peaks in the chromatogram are due tothe presence of heavier and more complex products, which areproduced during pyrolysis and hence are more difcult toidentify.

3.3. Boiling point distribution of the pyrolytic liquids

The boiling point distribution of hydrocarbons in pyrolytic liq-uids from pyrolysis of motorcycle tire at the temperature of475 C is presented in Fig. 6. For comparison purpose, the simu-lated distillation curves of commercial gasoline and diesel fuelare also presented in Fig. 6. The data shows that the pyrolytic liq-uids have a wide boiling point range. Pyrolytic liquids derived frommotorcycle tire waste contained about 40% gasoline fraction (boil-ing point range

Fuel3.4. Stability characteristics of the pyrolytic liquids (viscosityvariation)

The variations of viscosity of pyrolytic liquids over a time periodfor 12 months and over a temperature range of 30100 C are pre-sented in Table 6. It is interesting to note that the pyrolytic liquidswhen stored at room temperature and in the absence of air andlight show almost no change in viscosity with time. Table 6 showsthat the viscosity of the pyrolytic liquids decreases obviously withthe increase of temperature. Over the same temperature range of30100 C the reduction rates of viscosity of the pyrolytic liquids,diesel fuels and a typical heating oils (fuel oil No. 2) are4.2 102, 5 102 and 10 102 cSt/C, respectively. It canbe seen that there is a mild difference between the values of vis-cosity reduction rates of pyrolysis liquids and diesel fuels andhence the pyrolytic liquids would be less problematic for atomiza-tion when blending with diesel fuels.

3.5. Miscibility study of the pyrolytic liquids with diesel

The tire derived pyrolytic liquids have been found completelymiscible in commercial diesel. The liquids have been mixed withdiesel in different ratio and the physical properties like pH and vis-cosity of each mixture have been reported in Table 7. From the re-sults it is seen that the liquids can be upgraded in terms of increaseof pH and decrease of viscosity on mixing with diesel as required.

Table 6Variation of viscosity over time and temperature

Time (Months) Viscosity (cSt) Temperature (C) Viscosity (cSt)

3 4.79 30 4.756 4.77 50 3.869 4.75 75 2.7012 4.76 100 1.75

Table 7Miscibility study of pyrolytic liquids with diesel

Liquids:diesel pH Viscosity in cSt

9:1 4.67 4.504:1 4.96 4.351:1 5.75 3.89

M. Roqul Islam et al. /4. Conclusions

The main conclusions for pyrolysis of motorcycle tire waste areas follows:

The main rubber components in the present motorcycle tirewaste are NR and SBR. The motorcycle tire rubber formulationcomparatively use larger amount of inorganic materials as addi-tives consequently energy content of the solid tire waste islower than that of car and truck tire wastes.

The optimum liquid yield conditions for the xed-bed re-tubeheating reactor system are: operating temperature 475 C, feedsize 4 cm3 and apparent vapor residence time 5 s. The maineffects of operating conditions on the product distributions arethat: (i) the lower temperature and larger feed size favor incom-plete decomposition, which increase in char yields and decreasein the liquid and gas yields, (ii) the higher temperature andlonger residence time contributes to secondary reactions resultsin more gaseous products with the expense of liquids while charyields remain almost constant. The fuel properties of the pyrolysis liquids such as density, vis-cosity, GCV, carbon and hydrogen contents are found almostcomparable to those of the commercial automotive diesel fuelsbut higher sulphur content and lower ash point are problem-atic. The pyrolytic liquids may be used as diesel fuel or heatingoils after the upgrading such as desulphurization and dehydro-genation or blending them with petroleum renery streams.

The pyrolytic liquids abundantly contain olens, specially limo-nene and light aromatics, which have higher market values aschemical feedstock than their use as fuels.

However, further studies are necessary to utilize pyrolytic liq-uids as liquid fuels or chemical feedstock.

Acknowledgement

The rst author (M. Roqul Islam) would like to express his sin-cere gratitude and thanks to the Japan Society for the Promotion ofScience (JSPS) for nancial support during his research period in Ja-pan under ID No.: UGC-10632. The technical assistance for all ofthe chemical analyses supported by the technical staff, especiallyMr. Matsuda, of Instrumental Analysis Center at Kitami Instituteof Technology of Japan, is gratefully acknowledged. The authorsalso would like to thank Professor Dr. Entazul Haque, Departmentof Chemistry, University of Rajshahi, Bangladesh for useful discus-sion regarding chemical analyses during preparation of themanuscript.

References

[1] Bangladesh Bureau of Statistics. Statistical year book of Bangladesh 2004. 24thed. Dhaka: Government of Peoples Republic of Bangladesh; 2005.

[2] Rodriguez IM, Laresgoiti MF, Cabrero MA, Torres A, Chomon MJ, Caballero BM.Fuel Process Tech 2001;72:922 [and references therein].

[3] Laresgoiti MF, Caballero BM, De Marco I, Torres A, Cabrero MA, Chomon MJJ. JAnal Appl Pyrol 2004;71:91734.

[4] Gonzalez JF, Encinar JM, Canito JL, Rodriguez JJ. J Anal Appl Pyrol 2001;5859:66783 [and references therein].

[5] Diez C, Martinez O, Calvo LF, Cara J, Moran A. Waste Manage 2004;24:4639.[6] Dai X, Yin X, Wu C, Zhang W, Chen Y. Energy 2001;26:38599.[7] Pakdel H, Pantea DM, Roy C. J Anal Appl Pyrol 2001;57:91107.[8] Cunliffe AM, Williams PT. J Anal Appl Pyrol 1998;44:13152 [and references

therein].[9] Roy C, Chaala A, Darmstadt H. J Anal Appl Pyrol 1999;51:20121.[10] Barbooti MM, Mohamed TJ, Hussain AA, Abas FO. J Anal Appl Pyrol

2004;72:16570.[11] Williams PT, Brindle AJ. J Anal Appl Pyrol 2003;67:14364.[12] Williams PT, Brindle AJ. Fuel 2003;82:102331.[13] Pakdel H, Roy C, Aubin H, Jean G, Coulombe S. Environ Sci Technol

1992;9:1646.[14] Stanciulescu M, Ikura M. J Anal Appl Pyrol 2006;75(2):21725.[15] Stanciulescu M, Ikura M. J Anal Appl Pyrol 2007;78:7684.[16] Cunliffe AM, Williams PT. Energ Fuel 1999;13:16675.[17] Zabaniotou AA, Stavropoulos G. J Anal Appl Pyrol 2003;70:71122.[18] Zabaniotou AA, Madau P, Oudenne PD, Jung GC, Delplancke MP, Fontana A. J

Anal Appl Pyrol 2004;72:28997.[19] Murena F, Garu E, Smith RB, Gioia F. J Hazard Mater 1996;50:7998.[20] Williams PT. Waste treatment and disposal. 2nd ed. London: Wiley and Sons;

2005.[21] Cypres R, Bettens B. In: Ferrero GL, Maniatis K, Buekens A, Bridgwater AV,

editors. Pyrolysis and gasication. London: Elsevier Applied Science; 1989.[22] Williams PT, Besler S. Fuel 1995;74(9):127783.[23] Ucar S, Karagoz S, Ozkan AR, Yanik J. Fuel 2005;84:188492.[24] Kyari M, Cunliffe A, Williams PT. Energ Fuel 2005;19:116573.[25] Laresgoiti MF, De Marco I, Torres A, Caballero B, Cabrero MA, Chomon MJ. J

Anal Appl Pyrol 2000;55:4354.[26] de Macro I, Laresgoiti MF, Cabrero MA, Torres A, Chomn MJ, Caballero BM.

Fuel Process Technol 2001;72:9.[27] Williams PT, Besler S, Taylor DT. Fuel 1990;69:147482.[28] Napoli A, Soudais Y, Lecomte D, Castillo S. J Anal Appl Pyrol 1997;40

41:373.[29] Morillo R, Aylon E, Navarro MV, Callen MS, Aranda A, Mastral AM. Fuel Process

Technol 2006;87:1437.[30] Islam MR, Haniu H, Beg MRA. J Environ Eng 2007;2(4):68195.

87 (2008) 31123122 3121[31] Mastral AM, Murillo R, Callen MS, Garcia T, Snape CE. Energ Fuels2000;14(4):73944.

[32] Chang YM. Resour Conserv Recy 1996;17:12539.

[33] Zailani RB. Fluidized-bed pyrolysis of organic solid waste. Mech Eng Thesis,Universiti Teknologi Malaysia, Malaysia, 1995.

[34] Islam MN, Islam MN, Beg MRA. Int Energy J 2004;5(1):118.[35] Kaminsky W, Mennerich C. J Anal Appl Pyrol 2001;5859:80311.[36] Mirmiran S, Pakdel H, Roy C. J Anal Appl Pyrol 1992;22:205.[37] Roy C, Rastegar A, Kaliaguine S, Darmstadt H. Plast Rubber Compos Proc Appl

1995;23:21.[38] Benallal B, Roy C, Pakdel H, Chabot S, Poirier MA. Fuel 1995;74(11):1589.[39] Roy C, Labrecque B, de Caumia B. Resour Conserv Recy 1990;4:203.[40] Aguado R, Olazar M, Arabiourrutia M, Alvarez S, Bilbao J. In: Proceedings of the

VIII Congreso de Ingenier and Ambiental PROMA03, Bilbao, Spain; 2003.[41] Chaala A, Roy C. Fuel Process Technol 1996;46(3):227.[42] Seidelt S, Muller-Hagendorn M, Bockhorn H. J Anal Appl Pyrol 2006;75:118.[43] Leung DYC, Wang CL. J Anal Appl Pyrol 1998;45:15369.[44] Yang J, Kaliaguine S, Roy C. Rubber Chem Technol 1993;66:21329.[45] Liu ZR, Tang HY, Zheng YL. Hand book of rubber industry. China: Chemical

Industry Publishing House; 1992 [p. 4778].

[46] San Miguel G, Fouler DG, Sollans CJ. Ind Eng Chem Res 1998;37:24305.[47] De Macro I, Legarreta JA, Laresgoiti MF, Torres A, Cambra JF, Chomn MJ, et al. J

Chem Tech Biotechnol 1997;69:187.[48] Grittner N, Kaminsky W, Obst G. J Anal Appl Pyrol 1993;25:293.[49] Kaminsky W. Adv Polym Technol 1995;14:337.[50] Van Krevelen DW. Thermal decomposition, properties of polymers. The

Netherlands: Elsevier; 1990 [Chapter 21].[51] Kaminsky W, Sinn H. Pyrolysis of plastic waste and scrap tires using a uidized

bed process. In: Jones JL, Radding SB (editors), Thermal conversion of solidwastes and biomass, ACS Symposium Series 130. Washington, DC: AmericanChemical Society Publishers; 1980.

[52] Lucchesi A, Maschio G. Conserv Recycl 1983;6(3):8590.[53] Williams PT, Besler S, Taylor DT. Proc Inst Mech Eng 1993;207:5563.[54] Bridgwater AV. J Anal Appl Pyrol 1999;51:322 [and references therein].[55] Herbst FM. Low temperature pyrolysis of waste. In: Proceedings of the eiegth

European conference on biomass for energy, environment, agriculture andindustry, Vienna; October 1994.

3122 M. Roqul Islam et al. / Fuel 87 (2008) 31123122

Liquid fuels and chemicals from pyrolysis of motorcycle tire waste: Product yields, compositions and related propertiesIntroductionMaterials and methodsFeed materialsMajor composition of motorcycle tire rubberProximate and ultimate analyses and GCVs of motorcycle solid tire waste

Experimental sectionPyrolytic product liquid analysis

Results and discussionPyrolysis product yieldsAnalysis of product liquidsFuel properties of the liquidsChemical composition of the liquid products

Boiling point distribution of the pyrolytic liquidsStability characteristics of the pyrolytic liquids (viscosity variation)Miscibility study of the pyrolytic liquids with diesel

ConclusionsAcknowledgementReferences