ARTICLE IN PRESS

0961-9534/$ - se

doi:10.1016/j.bi

�Correspondi22-2572-6875.

E-mail addre1The HF has

ash for the pres

using for comm

Biomass and Bioenergy 27 (2004) 445–457www.elsevier.com/locate/biombioe

Influence of pretreatment for deashing of sugarcane bagasseon pyrolysis products

Piyali Dasa, Anuradda Ganesha,�, Pramod Wangikarb

aEnergy Systems Engineering, Indian Institute of Technology Bombay, Powai, Mumbai 400076, IndiabDepartment of Chemical Engineering, Indian Institute of Technology Bombay, Powai, Mumbai 400076, India

Received 3 February 2003; received in revised form 4 December 2003; accepted 29 April 2004

Abstract

This paper reports the studies made on the vacuum pyrolysis of deashed sugarcane bagasse, on the pyrolysis

products. The present work is with an objective to understand the change in the quantity and quality of the oil fraction

obtained from pyrolysis, upon pretreatment for deashing of original biomass. Ash, in the entrained char is believed to

be catalyzing the polymerization reaction in the oils and thereby increases the viscosity. Three different pre-treatment

processes used for deashing are water leaching, mild acid treatment with HCl and mild acid treatment with HF.1 The

study indicates the remarkable influence of pretreatment process for deashing, by enhancing the total energy

distribution in oil fraction of the pyrolysis products. This is attributed to selective removal of ash elements along with

removal of extractives and hemicellulose in different proportions. However, it was found that the pre-treatments do not

improve the stability of oil. The water leachate, as expected, showed potential of making ethanol via fermentation.

r 2004 Elsevier Ltd. All rights reserved.

Keywords: Sugarcane bagasse; Leaching; Acid wash; Deashing; Stability; Vacuum pyrolysis

1. Introduction

Sugarcane bagasse—the residue left after juiceextraction is a waste available in abundanceworldwide. About 1.34Gt of sugarcane was

e front matter r 2004 Elsevier Ltd. All rights reserve

ombioe.2004.04.002

ng author. Tel.: +91-22-2576-7886; fax: +91-

ss: aganesh@me.iitb.ac.in (A. Ganesh).

been used only to achieve complete removal of

ent study and by no means a suggestion toward

ercial application.

produced globally in 1999, which equates toapproximately 375MT of bagasse, 50% of whichis typically burned [1]. India is the second largestproducer of sugarcane next to Brazil with aproduction of 300MT of sugarcane in1999–2000. In India, about 4 million hectares ofland is under sugarcane farming with an averageyield of 70 t ha�1. Besides Brazil and India,Australia, South Africa, Cuba, China, tropicaland subtropical countries also are major contri-butors to world production of sugarcane. Thus,

d.

ARTICLE IN PRESS

P. Das et al. / Biomass and Bioenergy 27 (2004) 445–457446

sugarcane bagasse has a strong potential indisplacing fossil fuels and can be extensively usedin boilers, turbines and furnaces for powergeneration.Generating power by direct combustion of

sugarcane bagasse in boilers has a maximumefficiency of about 26%. Combustion systems withlow efficiency are traditionally used in sugarcaneplants [2]. In populated areas, bagasse-fired boilerscan be one of the major health hazards due toairborne fly ash. Recently, the overall efficiency ofthe process is being greatly improved by cogenera-tion, wherein; the improvement comes from theproper utilization of heat and effective waste heatrecovery. On the other hand, pyrolysis offers aneffective utilization of the ‘‘fuel energy’’ itselfgiving energy-dense liquids (easier to handle, storeand transport), charcoal (improved solid fuel) andgaseous fuel. The ability to decouple the fuelproduction from the application is unique topyrolysis liquids and is a major advantage overgasification and combustion, which must use theenergy products immediately and cannot store ortransport them. Thus, transformation of bagasseinto high-density renewable fuels, like charcoaland bio-oil, can significantly increase the profit-ability of sugarcane plantations [3].Fast pyrolysis at medium temperature and low

vapour residence times are known to be mostsuitable conditions for maximizing liquid productsfrom biomass. To achieve the fast heating rates,fluidized and entrained bed reactors have beenextensively used. In case of fluidized bed pyrolysis,extensive particle entrainment with the vapourshas been reported. Vacuum pyrolyser—the lowturbulence inside moving and stirred-bed pyrolysisreactors developed by Pyrovac [4,5] has beenreported to have reduced carryover of particles incondensable products. The vacuum pyrolysisprovides the required low vapour residence timewith a slight compromise in the heating rateachievable and thereby the reduction in liquidproducts.Sugarcane bagasse pyrolysis has been referred to

by many authors. The role of parameters like peaktemperature and tar yield has been investigated[1,6]. Total condensates of the order of 40–60%have been reported on dry bagasse basis. Vacuum

pyrolysis of sugarcane bagasse has been reportedfirst by Pakdel et al. [7]. The recent studiesreported by Perez et al. [2] gives a very elaborateand extensive understanding about vacuum pyr-olysis of sugarcane bagasse particularly in theassessment of the yield and the product character-istics. The ageing tests reported therein are uniqueand gives an insight into the ageing process of thebagasse oil. This is of interest, particularly in itsapplication as a fuel. The major deterring factor inthe wide usage and acceptability of the bio oils areits change of physico-chemical characteristicsduring storage along with the corrosivity in thecommonly used storage medium. The oil under-goes polymerization thus resulting in an increase inviscosity with time. During ageing etherificationand esterification reactions occur between hydro-xyl, carbonyl and carboxyl group components[8–11]. The presence of condensation reactionsduring ageing is confirmed by increase in watercontent in the oil with time [2]. It is also mentionedin the literature [12,13] that this instability may beattributed to the presence of alkali metals in theash, which are being carried over/entrained by thechar particles with the vapours. These alkalimetals catalyse the polymerization reactions andthereby increase the viscosity. Moreover, thesealkali metals in ash, form deposits in combustionapplications, particularly in turbines, where thedamage potential is considerably high. Therefore,the present study aims at understanding theinfluence of this ash on the stability as well aspyrolysis product yields upon vacuum pyrolysis ofsugarcane bagasse.

2. Materials and methods

2.1. Biomass properties

The properties of the sugarcane bagasse used forthe present study are given in Table 1.

2.2. Preparation of sample

The oven-dried bagasse sample ground to passthrough 60-mesh sieve (p250 mm particle size) hasbeen taken for the investigation.

ARTICLE IN PRESS

Table 1

Properties of sugarcane bagasse

Proximate analysis Ultimate analysiswt% (on dry basis) wt% (on dry basis)

Volatile matter 84.83 C 56.32

Fixed carbon 13.28 H 7.82

Ash 1.89 N 0.89

O (By difference) 27.54

Table 2

Pretreatment processes and sugar analysis of leachates

Treatment

no.

Leaching process Soaking/leaching

time (h)

I Water leachinga (WL1) 1

II Water leachinga (WL2) 24

III Special water leachingb

(WL3)

24

IV Leachinga with 5(M)

HCl solution

1

V Leachinga with

(0.5–3)% HF solution

1

a150ml leachate corresponds to 12.5 g of bagasse.b150ml leachate corresponds to 25.0 g of bagasse.

P. Das et al. / Biomass and Bioenergy 27 (2004) 445–457 447

2.2.1. Deashing treatments

Bagasse has been subjected to three differenttypes of treatments in this study. The selection ofthe pretreatment processes is based on reportedliteratures, wherein, leaching of biomass [14,15]and coal/carbon deashing [16,17] processes havebeen elaborated. The pretreatment processes areaimed at maximum extraction of ash. The treat-ments are water leaching, leaching with 5(M) HClsolution and leaching with HF solution rangingbetween a concentration 0.5 and 3% of HF. Thedetails of proportion of leaching medium andsoaking time are given in Table 2. In case of HFtreatment, leaching with 3% HF solution wasfinally selected (after studying different concentra-tions) based on minimum concentration of HFsolution required for the reduction of ash to afairly negligible limit.For all the treatments, mixtures were leached

and occasionally stirred for predetermined time at2572 1C after which the samples were filtered offand washed with distilled water. The process ofwashing was continued until the wash-waterremained neutral. Finally, the samples were driedat 105 1C in oven. The leachates were collected andstored in refrigerator and subsequently analysedfor sugar content.

2.2.1.1. Effect of different deashing treatments on

sugarcane bagasse—sample analysis. Each of thedeashing treatments is associated with the massreduction in the original biomass feedstock. Thedegree of deashing as well as the mass reduction ofthe original biomass associated with each processis found to be different. The average massreduction on dry bagasse basis for treatments I,

II, and III was 18–20%, 26–27% and 26–27%,respectively. The acid leaching caused higher massreduction with as high as 50wt% with 5(M) HCltreatment and an average mass reduction of29–32% for treatment V. All biomass samples-treated and untreated, were ashed in a mufflefurnace to obtain ash, which is free of trapped orunburnt carbon. The degree of deashing was alsofound to be varying for different treatments. Thepercentage ash in the untreated and the treatedbagasse are presented in Table 3. It shows thattreatment with HCl leads to an apparent increaseof ash percentage, which is attributed to therelatively higher removal of other components inbagasse. This also corresponds to the highest massreduction during 5M HCl treatment. Waterleaching has a moderate effect on ash removal.The best result is achieved with HF treatment, inwhich case the ash percentage was reduced to aslow as 0.03% when treated with 3% HF solution.The elemental composition of ash was obtained

by Inductively Coupled Plasma Atomic EmissionSpectra instrument (ICP-AES) and presented inTable 4. In accordance to reported literature [18],simple water leaching washes out the alkalis likeNa and K wherein, 5M HCl leaching furtherremoves other alkali metals like Mg, Ca, Al etc.,but HF treatment removes almost all the ashelements.To understand the nature of change in the

chemical composition in sugarcane bagasse due todifferent pretreatments, the chemical composition

ARTICLE IN PRESS

Table 3

Percentage of ash in untreated and treated sugarcane bagasse

Normal

bagasse

Treated

bagasse

Water

leachinga at

2572 1C

Water

leachinga at

2572 1C

Water

leachingb at

2572 1C

5(M) HCl

leachinga at

2572 1C

0.5%HF

leachinga at

2572 1C

1% HF

leachinga at

2572 1C

2% HF

leachinga at

2572 1C

3% HF

leachinga at

2572 1CAsh (1.83) (1 h) (24 h) (24 h) (1 h) (1 h) (1 h) (1 h) (1 h)

Ash Ash Ash Ash Ash Ash Ash Ash

(1.27) (1.02) (1.09) (2.12) (0.5) (0.29) (0.08) (0.03)

a150ml leachate corresponds to 12.5 g of bagasse.b150ml leachate corresponds to 25.0 g of bagasse.

Table 4

Elemental composition of untreated and treated sugarcane bagasse

Elements Untreated bagasse (%

of biomass)

Water leached (1 h)

bagasse (% of biomass)

5(M) HCl-treated

bagasse (% of biomass)

3%HF-treated bagasse

(% of biomass)

Na 0.012 0.003 0.008 —

K 0.175 0.014 0.020 —

Ca 0.087 0.017 0.040 0.020

Mg 0.437 0.310 0.285 0.002

Al 0.003 0.002 0.004 —

Fe 0.004 0.003 0.005 —

Zn 0.001 0.001 0.014 —

Cr 0.004 — ND —

Co ND ND 0.001 —

Cu 0.001 0.001 0.030 —

Mn 0.141 — ND —

Ni — ND — —

P 0.014 0.003 0.002 —

Si 0.911 0.862 1.699 —

S 0.042 0.026 ND —

ND: not determined.

P. Das et al. / Biomass and Bioenergy 27 (2004) 445–457448

(extractives, hemicellulose, cellulose and lignin) ofall biomass (untreated and treated) was alsocarried out by Technical Association of Pulp andPaper Industries (TAPPI) methods. Table 5 givesthe chemical composition and higher heating valueof some of the untreated and treated bagasse.The investigations followed by the results and

discussion have been reported here in two parts—A, B. Part A deals with the pyrolysis experiments,product distribution and analysis of the pyrolysisproducts. Fermentation of leachate for ethanol hasbeen discussed in part B.

Part A: pyrolysis of untreated and treated

sugarcane bagasse—product characteristics

3. Experimental

Pyrolysis experiments were carried out in apacked-bed reactor of 3-in NB pipe made ofstainless steel. The reactor was electrically heatedat a maximum temperature of 500 1C under aninitial reactor vacuum of 5 kPA. At the end of eachpyrolysis run the reactor was cooled to room

ARTICLE IN PRESS

Table 5

Chemical composition and higher heating value of untreated and treated sugarcane bagasse

Biomass Extractivesa

(alcohol–benzene–water

soluble) wt%

Chemical composition HHV (Higher

heating value)

MJ/kg

Hemicellulosea

wt%

Cellulosea

wt%

Lignina wt%

Untreated sugarcane

bagasse

25.8 23.3 31.0 21.8 �20

Water-leached (1 h)

sugarcane bagasse

13.1 22.2 43.4 21.2 �18

5(M) HCl (1 h)-treated

sugarcane bagasse

14.0 4.4 61.8 19.9 �18

3% HF treated (1 h)-

sugarcane bagasse

14.6 16.2 40.4 23.3 �20

aTAPPY methods: T-9m, 11m, 13m, 17m and 201m.

P. Das et al. / Biomass and Bioenergy 27 (2004) 445–457 449

temperature under nitrogen flow and the char wascollected from the biomass basket hung inside thereactor. The volatiles removed on pyrolysis aregradually condensed in a pre-weighed condensingtrain. The condensates with dew point 60–65 1Cwere collected and the balance was condensed inice bath (5–71C). The total condensable collectedin the condensing train is termed as ‘total liquid’.Among the ‘total liquid’, the condensate collectedhaving a dew point upto 60–65 1C, are termed as‘bio-oil’ which have moisture content in the rangeof 8–12% with a calorific value X20MJ/kg and isdirectly combustible without any further treatment[19]. The condensate fractions other than the ‘bio-oil’ which are noncombustible, contain water andlight organics.

4. Results and discussion

4.1. Effect of deashing treatments on pyrolysis

product distribution of sugarcane bagasse

The effect of deashing treatments on sugarcanebagasse pyrolysis product distribution (char, gas,and total liquid including oil fraction), based ontreated bagasse basis, has been shown in Table 6columns A. Columns B in Table 6 represent thesame on original bagasse basis, i.e. all values areinclusive of mass reduction during leaching.

As expected, the pyrolysis product distributionrelates well to the chemical composition of bagasseas follows. In case of water leaching, the extrac-tives are being washed out and thereby reducingthe wt% of extractives from 25.8 to 13.3 (even for1 h water leaching). This is attributed to the factthat unlike woody biomass, extractives of sugar-cane bagasse comprise largely of starch, sugars,phenolic tannins, and are washed out in simplewater leaching. By virtue of above, though theactual amount of cellulose is not changed, therelative percentage of the same has increased in thewater leached bagasse. Leaching with 5(M) HClsolution hydrolyses the hemicellulose fraction to alarge extent leading to a drastic reduction ofhemicellulose and extractives with a resultantincrease in the apparent cellulose percentage inthe treated bagasse. In this case, the percentagemass reduction is so high (�50%) that thereduction in ash is not manifested; rather itactually seems to be increasing. Therefore, thepercentage increase in the oil yield is marginal andis not as high as it should have been, had the ashbeen removed completely.This aspect is confirmed by leaching with HF

solution. Leaching with HF not only increases therelative percentage of cellulose in the treatedbagasse by removing extractives and hemicellu-lose, but also completely removes ash elements.The increase in the oil percentage in this case is by

ARTICLE IN PRESS

Table 6

Pyrolysis product distribution of untreated and treated sugarcane bagasse

Biomass Product distribution

Total liquid Oil Char Gas

A B A B A B A B

Normal bagassea — 55.0 — 19.5 — 22.7 — 22.3

Leached bagassea (1 h) 62.4 49.9 32.1 25.6 18.5 14.8 19.1 15.3

Leached bagassea (24 h) 63.7 46.5 42.5 31.0 17.2 12.6 18.8 13.7

Leached bagasseb (24 h) 62.1 45.3 41.6 30.3 18.1 13.2 19.8 15.5

5MHCl-treated bagassea 61.0 30.5 42.0 21.0 20.4 10.2 18.6 9.3

HF(0.5%)-treated bagassea 67.3 47.8 41.6 29.5 15.6 11.1 17.1 12.1

HF(1%)-treated bagassea 68.7 48.1 43.9 30.7 15.1 10.6 16.2 11.3

HF(2%)-treated bagassea 71.3 49.2 45.2 31.2 12.4 8.6 16.3 11.2

HF(3%)-treated bagassea 72.9 49.6 47.9 33.0 12.4 8.4 14.7 9.9

A: wt% on treated bagasse basis, B: wt% on original bagasse basis.a150ml leachate corresponds to 12.5 g of bagasse.b150ml leachate corresponds to 25.0 g of bagasse.

P. Das et al. / Biomass and Bioenergy 27 (2004) 445–457450

about 69% (based on original untreated biomass)and by �145 (based on treated bagasse basis),while in case of leaching with least possible waterand longer period (Treatment III) the increase isby �55.4% (original basis) and by �113.3% (fortreated basis). A special reference has to be madeto explain the increase in the oil percentage fortreated bagasse. It is well reported that upondeashing, both the amount of volatiles and the rateof their evolution increase [20–23]. In the presenceof ash elements, the volatiles escaping undergosecondary cracking and form a soot deposit on theresidual char. The oil fraction, which consists ofmainly the condensates of primary vapours fromcellulose, lignin etc., increases only when theabsolute values of cellulose and lignin increaseand/or when the secondary cracking of oil to givelighter organic fraction occurs (catalysed by theash constituents). Thus, the increased devolatiliza-tion rates in combination with changed organicchemical composition are responsible for theincrease in oil fraction as well as oil to total liquidratios. Similar increase in oil percentages has beenreported by Subbarayudu [19] for other biomassalso wherein reduction of ash is considerableduring pretreatment.At this juncture, it is worth comparing the oil

percentages as well as the oil-to-liquid ratios for

two bagasse samples having similar composition.It is well known that the bagasse composi-tion changes with the source and origin. Samecomposition may be arrived by pretreatment ormay occur naturally. The bagasse compositionused by Perez et al. [2] matches with the composi-tion of 1 h water-leached bagasse in the presentstudy. The oil yield obtained in this study for theabove-mentioned pretreated bagasse is �32%,which is nearly similar to the oil yield of 34.2%reported by Perez et al. (see Table 7). Thus,the removal of extractives and the higher percen-tage of cellulose are capable of giving highyields of oil, which can be obtained by eitherwater leaching of the high extractive contentbagasse or by using a bagasse with high cellulosecontent.

4.2. Effect of deashing treatments on pyrolysis

product characteristics of sugarcane bagasse

The products obtained on pyrolysis of untreatedand treated bagasse have been subjected todifferent analysis for their properties. Theeffect of deashing treatments on the pyrolysisproduct characteristics is reported in the followingsection.

ARTICLE IN PRESS

Table 8

Properties of sugarcane bagasse pyrolysis oil

Biomass pyrolysis oil HHV (Higher heating value) MJ/kg pH Moisture content

Normal bagasse pyrolysis oil 23.3 2.6 12.0

Water-leached bagasse pyrolysis oil 22.2 2.5 11.2

HCl-treated bagasse pyrolysis oil 21.6 2.3 8.3

HF-treated bagasse pyrolysis oil 23.2 2.4 7.4

Table 7

Comparison of oil and liquid yield by two processes

Biomass components Total liquid

yield (% daf)

Oil yield

(% daf)

Ash (%

dry)

Hemicellulose+extractives Cellulose Lignin

(i) Present method (in situ

separation after

pretreatment)

35.3 43.4 21.2 62.4 32 1.89

(ii) Garcia Perez et al. [2]

(after processing of liquid)

35.8 43.1 21.1 62.0 34.2 1.6

P. Das et al. / Biomass and Bioenergy 27 (2004) 445–457 451

4.2.1. Higher heating value (HHV), pH and

moisture content of sugarcane bagasse

pyrolysis oil

Higher heating value, moisture content and pH ofbagasse and treated bagasse oil, is given in Table 8.There is no note-worthy variation in calorific value ofthe oils, however, it is seen that moisture content ofthe oil gradually decreases from untreated to water-leached and to acid-treated oil. This is supplementingthe observation discussed in Section 4.1, i.e. withleaching, the oil content increases and the charcontent decreases. It is appropriate to attribute thisto the removal of ash from bagasse, on pretreatment.The removal or and absence of ash reduces theoccurrence of ash catalysed lignin decompositionreactions forming char and water [20,21].

4.2.2. Miscibility characteristics of bagasse

pyrolysis oil

Characterization of pyrolysis oil in terms of polarand non-polar fractions has been carried out bymeans of solvent extraction. The percentage mis-cibility of untreated and pretreated bagasse pyr-olysis oil in different solvents ranging from non-polar hexane to highly polar methanol is presentedin Table 9. The comparative results of miscibility

show that moving from untreated to water leachingthe percentage solubility in non-polar solventsdecreases with gradual increase in solubility inpolar solvents. In case of acid-treated bagassepyrolysis oil, there is a drastic decrease in non-polar fraction with an equal increase in the polarfraction of oil. This may be attributed to theformation of low molecular weight sugars, alcohols,carboxylic acids resulting from the degradation ofcellulose and hemicellulose due to pretreatment.Miscibility characteristics of bagasse pyrolysis

oil with diesel were also studied. It is seen that amaximum of �15 g of diesel remain permanentlysoluble when mixed with 100 g of bagasse oil in5:5wt ratios. For higher diesel-to-oil ratio themiscibility was found to be broken down with timewith a clear separation of oil and diesel layer.

4.2.3. Stability characteristics of sugarcane

bagasse pyrolysis oil

The variation of viscosity was monitored foruntreated and treated bagasse pyrolysis oils, storedboth at room temperature (Fig. 1) as well as at 60 1C(Table 10). The viscosity is measured at 60 1C for theoils stored at room temperature, while for the oilstored at 60 1C, viscosity was measured at three

ARTICLE IN PRESS

0

10

20

30

40

50

60

70

80

90

0 14 21 30 35 42 60 365

time (days)

visc

osi

ty in

cS

t at

60˚

C

untreated bagasse pyrolysis oil

1 hour water leached bagassepyrolysis oil"

1 hour HF treated bagassepyrolysis oil

71 5

Fig. 1. Viscosity variation of untreated and treated bagasse pyrolysis oil with time (stored at room temperature).

Table 9

Solvent extraction of bagasse pyrolysis oil

Pyrolysis oil % Solubility

Hexane (Non-

polar aliphatic

and aromatic

compounds)

Benzene

(Aromatic

compounds)

Dichloromethane

(High proportion

of substituted

polar phenolic

compounds)

Ethyl acetate

(Phenolic

compounds+low

molecular wt.

carboxylic acids)

Methanol (Most

polar fractions

and difficult to

characterize.

Polyalcohols,

sugars and fatty

acids)

Untreated bagasse

pyrolysis oil

5.27 27.8 30.17 16.92 19.88

1 h water-leached

bagasse pyrolysis oil

3.59 17.70 34.39 22.44 21.88

1 h HF-treated

bagasse pyrolysis oil

0.9 8.89 39 28 24

Table 10

Viscosity of bagasse pyrolysis oil versus heating time and

heating temperature

Heating time Ht H Heating temperature (HT) 60 1C

Viscosity (cSt) (measured at)

30 1C 60 1C 80 1C

0 Thick 28.00 12.01

1 Thick 27.14 11.90

24 93.62 34.52 16.62

168 — 67.37 21.25

P. Das et al. / Biomass and Bioenergy 27 (2004) 445–457452

different temperatures like 30 1C, 60 1C and 80 1C. Itis seen that compared to untreated bagasse oil initialviscosity as well as rate of change of viscosity of

pretreated bagasse oil is more. This is attributed tothe more polar fractions present in treated oils asshown in Table 9. Pretreatment hydrolyses celluloseand hemicellulose which results in the increase ofmore acidic as well as polar fraction in the oil leadingto higher rate of increase of viscosity in thepretreated oil compared to untreated pyrolysis oil.The effect is more severe in case of acid pretreatmentwhich leads to accelerated polymerization givingmore viscous and more acidic oil having less watercontent.

4.2.4. Chemical characterization of bagasse

pyrolysis oil

The bagasse pyrolysis oil, treated and untreated,was analysed for their compounds using FTIR and

ARTICLE IN PRESS

Table 11

BET surface area of sugarcane bagasse pyrolysis char

Pyrolysis char BET surface area (m2/g) HHV (Higher heating

value) MJ/kg

Iodine no. (mg/g)

Untreated bagasse pyrolysis char 98 28.64 264

Water-treated bagasse pyrolysis char 115.77 26.27 332

HCl-treated bagasse pyrolysis char 153.72 27.32 —

HF-treated bagasse pyrolysis char 242.73 28.22 538

P. Das et al. / Biomass and Bioenergy 27 (2004) 445–457 453

GC/MS techniques. The oils were first separatedinto five fractions according to polarity as men-tioned in Section 4.2.2 and then analysed. There isan increase in the oxygenated–polar (dichloro-methane and ethylacetate-soluble fractions) asshown for treated bagasse pyrolysis oil.The major compounds present in the water-

leached bagasse are very similar to that reportedby Perez et al. for vacuum pyrolysis of bagasse ofsimilar composition under same reaction condi-tions. The moisture content and the calorificvalues are also comparable. Thus, it is seen thatnot only the product distribution but the nature ofpyrolysis oil composition also depend largely onthe relative composition of bagasse feedstock. Thishas been exploited and explained in detail byRavindran [20].

4.2.5. Effect of deashing treatments on

pyrolysis char properties

BET surface area, Iodine number (ASTM D4607-86) and higher heating value of untreatedand treated bagasse pyrolysis char is presented inTable 11. It shows that in each of the deashingtreatments there is an increase of BET surface areawith a maximum increase in HF-treated bagassepyrolysis char. The leaching leads to the elimina-tion of compounds containing metal cations. Theincrease in surface area occurs as a result ofopening of silica pores previously blocked by metalcations. The treatment with HF, which makes theobtained material practically ash-less, leads toelimination of silica causing a total increase ofsurface area [17,24].

Part B: ethanol fermentation with leachates

5. Introduction

As mentioned in Part A, pretreatment ofbagasse through leaching has been resorted tofor selective removal of ash. This is accompaniedby removal of other organic components likesugars, extractives, hemicellulose etc. The leachatehas enough carbon to support the growth oforganisms and to convert this carbon source intodesired chemicals like ethanol. In view of this, thefollowing study investigates the potential ofethanol production via fermentation of the lea-chates.

6. Experimental

6.1. Microorganism and culture media

In the present study, the leachates, which resultfrom different pretreatment processes of finelypowdered sugarcane waste, was fermented forethanol by yeast Saccharomyces cerevisiae Yeaststrain Saccharomyces cerevisiae (Distillers yeast)provided by Burns Philip India Ltd (Kegaon-Uran, India) was used for the study. The strainwas maintained on culture medium containingglucose 20 g/l, peptone 5 g/l, yeast extract powder5 g/l, and agar 25 g/l.

6.2. Ethanol fermentation with leachate

The ethanol fermentation was carried out in theleachate medium. The water leachates (pH 3–4)were first neutralized with 0.1N NaOH solution.Ethanol fermentation medium was prepared byadding peptone (Nitrogen source) and yeast

ARTICLE IN PRESS

P. Das et al. / Biomass and Bioenergy 27 (2004) 445–457454

extract powder, with a concentration of 5 g/l each,in the neutralized leachate medium and wasautoclaved. The culture maintained on slants wastransferred to 100ml (in 500ml flask) of culturemedia. This culture medium, containing 20 g/lglucose, 5 g/l peptone, 5 g/l yeast extract powder),was used to inoculate 50ml (in 250ml flask) offermentation media. This was used as a seedculture for experimental work. All experimentswere carried out at 30 1C on an orbital shaker at250 r.p.m. Biotransformation was initiated byaddition of 2ml of seed culture to the fermentationmedia and 1ml of sample was withdrawn atknown time intervals for analysis.Each of the samples taken in regular time

interval was then centrifuged at 6000 r.p.m. for10min. The supernatant was collected in Eppen-dorf tubes and stored at 4 1C for ethanol andglucose estimation. The centrifuged organismswere washed with saline water (0.9% aqueousNaCl solution) and finally were resuspended insaline water making the volume same as initial.

6.3. Estimation of cell concentration

Each sample (centrifuged organism suspendedin saline) was diluted with saline water to desiredconcentration range and mixed in a vortex to makehomogeneous solution. Then, the absorbance wasmeasured at 600 nm on UV/visible spectrophot-ometer (Jasco model V-530). For calibration ofdry cell weight of yeast, 5ml of cell suspensionswith known absorbance values were centrifugedand dried in an oven at 80 1C to constant weight.This procedure was repeated thrice and averagevalues were used for calibration.

6.4. Estimation of glucose concentration

For glucose estimation, 1ml of each sample(supernatant of each centrifuged sample diluted todesired range of concentration) was added to 3mlof o-toluidine reagent (8% v/v o-toluidine inglacial acetic acid). This was vortexed and placedin a boiling water bath for 20min. The resultingsolution was cooled and the absorbance was readat 630 nm. A series of standards was made withconcentration of glucose (dextrose) in the range

0.05–0.25 g/l and absorbance was measured. Cali-bration curve was obtained by plotting absorbanceagainst concentration of standard samples. Theequation for the line of best fit for glucosecalibration found is given in the following section.This method measures the reducing sugar glucoseexclusively among a mixture of saccharides. o-toluidine in the presence of glacial acetic acidreacts quantitatively with aldehyde [>CHO]groups of aldohexoses to yield glycohexyl aminesand Schiff’s base. Aldopentoses, maltose andlactose give similar reactions but are very lessreactive. A blue-green colour is obtained which isestimated spectrophotometrically at 630 nm.

6.5. Estimation of ethanol concentration

Ethanol concentration was estimated using aMAK series 911M Gas Chromatograph (GC) with0.5 in steel Porapak Q column of 2m length. Thecolumn has a mesh size of 80/100. The GCconditions are as follows: Oven temperature180 1C (isothermal), Injector temperature 220 1C,Flame Ionisation Detector (FID) temperature220 1C. 3 ml sample (supernatant of each centri-fuged sample) was injected for each run. Nitrogenwas used as the carrier gas with a flow rate of30ml/min. The calibration curve was obtainedusing standard ethanol solutions of known con-centration.

7. Results and discussion

All the leachates were analysed for glucosecontent and the results are presented in Table 12.It shows that, an average of 55% of total weightloss of biomass during leaching process is con-tributed by glucose and the rest might be due topentoses like xylose and other sugars. The glucoseconcentration of leachate being high (78%) in caseof pretreatment with higher biomass-to-waterratio.All the leachates were first neutralized with

0.1N NaOH solution and tested for the viability ofcell growth. In case of 5M HCl and HFpretreatment leachate, there was no growth ofcell, which may be attributed to higher salinity of

ARTICLE IN PRESS

0

5

10

15

20

25

30

35

40

0 2 4 6 8 10 12 14 16 22 24 26 30 32 34 42

Time (h)

con

cen

trat

ion

(g

/L)

cell(g/L)

glucose(g/L)

ethanol(g/L)

Fig. 2. Kinetics of treatment 3-leachate fermentation.

0

10

20

30

40

50

0 2 4 6 8 12 14 16 18 20 22 24 26 28 30

Time (h)

con

cen

trat

ion

(g

/L)

cell(g/L)

glucose(g/L)

ethanol(g/L)

Fig. 3. Kinetics of mixture of treatment 3-leachate and 2%

Dextrose fermentation.

Table 12

Sugar analysis of leachates

Treatment

no.

Wt. of glucose (g/l)

in leachate

% Sugar (glucose)

loss w.r.t total wt.

loss in leaching

I 7.5 50.00

II 12.90 57.30

III 35.31 78.00

IV 22.11 53.06

V 15.08 56.55

P. Das et al. / Biomass and Bioenergy 27 (2004) 445–457 455

the leachate on neutralization with NaOH solu-tion. On the other hand, NaF produced onneutralizing HF-treated leachate inhibits theTCA and glycolysis pathways, i.e. the centralmetabolic pathways of cell growth. Finally,leachate resulting from special 24-h water leachingof bagasse was taken for fermentation study.Fermentation tests were carried out with four

sets of substrates, viz. with control, 20 g/l glucosesolution, 24 h water leachate (Treatment II, WL2),water leachate of special water treatment (Treat-ment III, WL3), and with the leachate of Treat-ment III supplemented with additional 20 g/lglucose. The rate of production of ethanol as wellas the rate of consumption of glucose and rate ofcell growth for the leachate of Treatment III,and with the leachate of Treatment III supple-mented with additional glucose have been pre-sented in Figs. 2 and 3, respectively. It is veryinteresting to note that maximum conversion ofglucose to ethanol is �38–40% of original glucoseconcentration.

8. Conclusion

Pre-treatment of bagasse with water, dilute HClsolution and dilute HF solution shows a remark-able change in the pyrolysis product distributionby virtue of a combination of a change in theorganic constituents and the selective removal ofinorganic ash elements. Mild HF solution iseffective in reducing the ash content of the biomassto a negligible amount. Moreover, this treatmenteffectively increases the oil yield by �69% (on thebasis of original wt of bagasse before treatment)

compared to oil obtained from untreated bagasse.Thus, HF treatment removes the ash elementscompletely, yet does not help in improving thestability. Moreover, the leachates from this pre-treatment, causing environmental concerns, arenot advisable to use. Similarly, treatment with 5MHCl leads to a marginal increase in oil yield butresults in an increase in viscosity of pyrolysis oil.Acid pretreatment (both HF and HCl) hydrolysesthe hemicellulose and cellulose into smaller mole-cules resulting in the increase of acidic as well aspolar fractions in the oil and leads to higher rate ofincrease of viscosity in case of pretreated oilcompared to untreated pyrolysis oil.The leachates of both the acid pretreatment

processes have high sugar contents, yet ethanolfermentation are not feasible in acid-treatedleachates due to high salinity of neutralizedleachates, which inhibits the TCA cycle andglycolysis pathways of fermenting microorganism.

ARTICLE IN PRESS

P. Das et al. / Biomass and Bioenergy 27 (2004) 445–457456

On the other hand, simple water leaching (withleast possible water) for longer period is effectiveenough to reduce the extractives (function ofamount of extractives present originally, which inturn depend on bagasse generation process) alongwith some selected ash components in bagasse.Water leaching, by virtue of selective removal ofash and organic constituents increases the oil yield(increase in the oil yield vary with the reduction inextractives and total mass) by as much as 55.4%(on original basis) on vacuum pyrolysis.The calorific value of oil however is not affected

much and is in the range of 22–24MJ/kg. Thepretreatment also produces char with a higheradsorptive capacity, thereby adding value to thechar obtained from pyrolysis.The high sugar content of water leaching

process carried with least possible water and forlonger period (24 h) shows promising result with amaximum yield of ethanol 13.5 g/l and 38–40%conversion of sugar. The fermentation process canbe made more economic by supplementing theleachate with glucose or other ethanol makingsubstrate and hence increasing the ethanol con-centration in the final product. The ethanol inturn, can be used to stabilize the oil [20] producedfrom pretreated (water-leached) bagasse. Techni-cally based on simple calculations, i.e. if oil yieldfrom 24-h water-leached bagasse is 31% and 5%ethanol is required for stabilizing the oil, the waterleachate when fermented produces enough ethanolfor stabilization of the oil. This shows the potentialof integrated loop-type approach for stable, lessviscous bio-oil product.

References

[1] http://www.dynamotive.com/news.

[2] Perez M, Chaala A, Roy C. Vacuum pyrolysis of

sugarcane bagasse. Journal of Analytical and Applied

Pyrolysis 2002;65(2):111–36.

[3] Drummond AR, Drummond IW. Pyrolysis of sugar cane

bagasse in a wire mesh reactor. Industrial and Engineering

Chemistry Research 1996;35(4):1263–8.

[4] Roy C, Yang J, Blanchette D, Korving L, Caumia BDe.

Development of a novel vacuum pyrolysis reactor with

improved heat transfer potential. In: Bridgwater AV,

Boocock DGB, editors. Proceedings of Developments in

Thermochemical Biomass Conversion. London: Blackie

Academic and Professional; 1996. p. 15–9.

[5] Roy C, Blanchette D, Caumia BD. Horizontal moving and

stirred bed reactor. Canadian Patent claim number 2 196

841, US Patent number PET /IB98/00224, January 30,

1998.

[6] Brossard LE, Cortez LAB. Potential for the use of

pyrolytic tar from bagasse in industry. Biomass &

Bioenergy 1997;12(5):363–6.

[7] Pakdel H, Roy C, Chaala A. Production of useful products

by vacuum pyrolysis of biomass. In: Proceedings of the XV

Chemical Conference of the Oriente University, Santiago

de Cuba, Editorial Oriente, November 27–29, 1996.

[8] Czernik S, Johnson DK, Black S. Stability of wood fast

pyrolysis oil. Biomass & Bioenergy 1994;7(1–6):187–92.

[9] Sharma RK, Bakshi NN. Upgrading of biomass derived

pyrolytic oils over HZSM-5 catalyst. Report of Contract

File No. 058SZ-23283-8-6116. Bioenergy Development

Program, Energy Mines and Resources, Canada, 1989,

p. 79.

[10] Adjaye JD, Sharma RK, Ramesh K, Bakshi N, Narendra

N. Characterisation and stability analysis of wood derived

bio-oil. Fuel Processing Technology 1992;31(3):241–56.

[11] Darmstadt H, Perez GM, Chaala A, Roy C. Co-pyrolysis

under vacuum of sugarcanr bagasse and petroleum

residue: properties of the char and activated char products.

Carbon 2001;39:815–25.

[12] Bridgewater T, Peacocke C. Biomass fast pyrolysis. In:

Proceedings of the second Biomass Conference of the

Americas. Energy Environment, Agriculture, and Indus-

try, NREL/cp-2008098, DE 95009230, NREL, Golden

CO, August 21–24, 1995, p. 1037–46.

[13] Diebold JP. A review of the chemical and physical

mechanisms of the storage stability of fast pyrolysis bio-

oils. In: Bridgwater AV, editor. Fast pyrolysis of biomass:

a handbook, vol. 2. UK: CPL press; 2002.

[14] Turn SQ, Kinoshita CM, Ishimura DM, Jenkins BM.

Removal of inorganic constituents of fresh herbaceous

fuels process and costs. In: Overend RP, Chornet E,

editors. Proceedings of the third Biomass Conference of

the Americas. Making a Business from Biomass in Energy,

Environment, Chemicals, Fibers, and Materials, Montreal,

Canada; 1997, p. 401–14.

[15] Jenkins BM, Bakker RR, Wei JB. Removal of inorganic

elements to improve biomass combustion properties. In:

Proceedings of the second Biomass Conference of the

Americas. Energy, Environment, Agriculture, and Indus-

try, NREL/cp-2008098, DE 95009230, NREL, Golden

CO.; 1995, p. 483–92.

[16] Samaras P, Diamadopoulas E, Sakellaropoulos PG. The

effect of mineral matter and pyrolysis conditions on

gasification of greek lignite by carbon dioxide. Fuel

1996;75:1108–14.

[17] Rychlicki G, Terzyk AP, Majchrzycki W. The effect of

commercial carbon de-ashing on its thermal stability and

porosity. Journal of Chemical Technology and Biotech-

nology 1999;74:329–36.

ARTICLE IN PRESS

P. Das et al. / Biomass and Bioenergy 27 (2004) 445–457 457

[18] Jenkins BM, Bakker RR. On the properties of washed

straw. Biomass & Bioenergy 1996;10:133–42.

[19] Subbarayudu K. Liquid fuels from thermochemical con-

version of biomass. PhD thesis, Energy Systems Engineer-

ing, IIT Bombay, 2003.

[20] Ravindran K. Studies on influence of biomass composition

on pyrolysis. PhD thesis, Energy Systems Engineering, IIT

Bombay, 1995.

[21] Gray MR, Corcoran WH, Gavalas GR. Pyrolysis of wood

derived material: effects of moisture and ash content,

Industrial Engineering Chemistry, Process Design and

Development 1985;24:646–51.

[22] Shafizadez F. Pyrolysis and combustion of cellulosic

materials. Advances in Carbohydrate Chemistry

1968;23:419–65.

[23] Essig M, Richards GN, Schenk E. Mechanism of

formation of the major volatile products form pyrolysis

of cellulose. In: Schuerch C, editor. Cellulose and Wood-

Chemistry and Technology. New York: Wiley; 1989.

p. 841–62.

[24] Ravindran K, Ganesh A. Adsorption characteristics and

pore development of biomass pyrolysis char. Fuel

1998;77:769–81.