Available online at www.sciencedirect.com

Bioresource Technology 99 (2008) 3737–3744

Epoxidation of cottonseed oil by aqueous hydrogen peroxidecatalysed by liquid inorganic acids

Srikanta Dinda, Anand V. Patwardhan *, Vaibhav V. Goud 1, Narayan C. Pradhan

Department of Chemical Engineering, Indian Institute of Technology, Kharagpur 721 302, India

Received 17 April 2007; received in revised form 9 July 2007; accepted 10 July 2007Available online 30 August 2007

Abstract

The kinetics of epoxidation of cottonseed oil by peroxyacetic acid generated in situ from hydrogen peroxide and glacial acetic acid inthe presence of liquid inorganic acid catalysts were studied. It was possible to obtain up to 78% relative conversion to oxirane with veryless oxirane cleavage by in situ technique. The rate constants for sulphuric acid catalysed epoxidation of cottonseed oil were in the range0.39–5.4 · 10�6 L mol�1 s�1 and the activation energy was found to be 11.7 kcal mol�1. Some thermodynamic parameters such asenthalpy, entropy, and free energy of activation were determined to be of 11.0 kcal mol�1, �51.4 cal mol�1 K�1 and 28.1 kcal mol�1,respectively. The order of effectiveness of catalysts was found to be sulphuric acid > phosphoric acid > nitric acid > hydrochloric acid.Acetic acid was found to be superior to formic acid for the in situ cottonseed oil epoxidation.� 2007 Elsevier Ltd. All rights reserved.

Keywords: Hydrogen peroxide; In situ epoxidation; Cottonseed oil; Kinetics; Liquid inorganic acids

1. Introduction

The epoxidation of alkenes and other unsaturatedhydrocarbon chains constitutes one of the most useful reac-tions in organic synthesis, as the epoxide group is an activeintermediate, which can be readily transformed to therequired functionality. Epoxidation of long chain olefins,and unsaturated fatty acid derivatives such as soybean oiland other plant oils is carried out on an industrial scale(Rios et al., 2005). Okieimen et al. (2002) have studiedthe epoxidation of rubber seed oil by peroxyacetic acid gen-erated in situ. Fats and oils are renewable resources thatcan be treated chemically or enzymatically to producematerials that can often act as replacement for materialsderived from petroleum (Gan et al., 1995; Rios et al.,

0960-8524/$ - see front matter � 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.biortech.2007.07.015

* Corresponding author. Tel.: +91 3222 283950; fax: +91 3222 255303.E-mail address: avp@che.iitkgp.ernet.in (A.V. Patwardhan).

1 Present address: Chemical Engineering Group, Birla Institute ofTechnology and Science (BITS), Pilani, Goa Campus, NH17B BypassRoad, Zuarinagar, Goa 403726, India.

2005). Due to the high reactivity of the oxirane ring, epox-ides can also act as raw materials for synthesis of a varietyof chemicals, such as alcohols, glycols, alkanolamines, car-bonyl compounds, olefinic compounds, and polymers likepolyesters, polyurethanes, and epoxy resin. Epoxidationusing soybean oil for the synthesis of commercially valu-able plasticisers and polymer stabilisers has already beensuccessfully implemented at plant scale (Wallace, 1978;Biermann et al., 2000; Klaas and Warwel, 1999). However,the cost effectiveness of the route greatly depends upon thelocal and cheap availability of the raw material. It is possi-ble that few agricultural belts may not be suitable for itscultivation but may be favorable for some other vegetableoils. Hence, large number of such oil systems have to bestudied for the cost effectiveness and easy availability ofraw materials.

Epoxidation methods vary from case to case depend-ing on the nature of reactants and catalysts used for epox-idation. To produce epoxides from olefinic type ofmolecules, the available methods are: (a) epoxidation withpercarboxylic acids (Guenter et al., 2003), can be catalysed

Nomenclature

A0 atomic weight of oxygen, g mol�1

AOH atomic weight of hydroxyl group, g mol�1

Ai atomic weight of iodine, g mol�1

Ea activation energy, kcal mol�1

[ ] concentration, mol L�1

[ ]0 initial concentration, mol L�1

[EP] concentration of oxirane or epoxide, mol L�1

Gexp experimentally obtained a-glycol, mol/100 gsample

Gthe theoretically obtainable maximum a-glycol,mol/100 g sample

h Planck’s constant, 6.626 · 10�34 JsIV0 initial iodine value, g of I2/100 g sample

Keq equilibrium constantKAcOH dissociation constant of acetic acidk rate constant, L mol�1 s�1

N Avogadro constant, 6.023 · 1023 mol�1

OOexp experimentally obtained oxirane oxygen, g/100 g oil sample

OOthe theoretically obtainable maximum oxirane oxy-gen, g/100 g oil sample

R universal gas constant, J mol� K�1

T temperature, KDF free energy of activation, kcal mol�1

DH enthalpy of activation, kcal mol�1

DS entropy of activation, cal mol�1 K�1

3738 S. Dinda et al. / Bioresource Technology 99 (2008) 3737–3744

by acids or by enzymes (Rios et al., 2005; Klaas and War-wel, 1999); (b) epoxidation with organic and inorganic per-oxides, catalysed by transition metal catalyst (Sharplesset al., 1983); (c) epoxidation with halohydrines, using hyp-ohalous acids (HOX) and their salts (Guenter et al., 2003);and (d) epoxidation with molecular oxygen (Guenter et al.,2003).

Epoxidation with molecular oxygen catalysed by silver isthe cheapest and greenest route. However, its use is mostlyrestricted to substrates like ethylene and butadiene owingto very low yields in the case of other alkenes (Guenteret al., 2003 and Sheldon and Dakka, 1994). Furthermore,degradation of the oil to low molecular weight volatilecompounds such as aldehydes, ketones, and short chaindicarboxylic acids makes the process less selective to epox-ide and therefore inefficient/unattractive. Therefore, it isnot an efficient method for epoxidation of vegetable oils.Epoxidation using halohydrines is highly unfriendly envi-ronmentally so it is obviously not our choice. Hence, fromthe above discussion, it is clear that for clean and efficientepoxidation of vegetable oils, the available technologiesthat need to be developed are epoxidation with percarboxy-lic acids and epoxidation with organic and inorganicperoxides.

The main purpose of this study was to develop value–added products form locally available renewable resourcesin India. Vegetable oils such as rice-bran, cottonseed,groundnut, sunflower, rapeseed, sesame, palm kernels,coconut, linseed, castor, sal, neem, karanja are easily avail-able in India. The potential availability of some oils andtheir current utilisation has been discussed by Srivastavaand Prasad (2000). Therefore, the objective of this workwas to study the epoxidation of cottonseed oil within situ generated peroxyacetic acid, with a view of obtainingvalue–added products from locally available renewablenatural resources. A further objective of this work was topropose a mathematical model for the reaction systemand estimation of the unknown kinetic parameters of theproposed model.

2. Experimental details

2.1. Materials

Refined cottonseed oil was obtained from a reputablefirm. Glacial CH3COOH (99–100%), aqueous H2O2

(50%), HCOOH (85%), H2SO4 (98%), HNO3 (70%),H3PO4 (85–87%), HCl (35%) were obtained from Merck(India) Limited. All other chemicals for analytical purposewere obtained from reputable firms. The average fatty acidcomposition (weight basis) of Indian cottonseed oil is in therange: myristic 0.5–1.5%, palmitic 20–23%, stearic 1–3%,oleic 23–35%, linoleic 42–54% (Chakrabarty, 2003).

2.2. Experimental setup

The epoxidation reactions were carried out in a fullybaffled mechanically agitated contactor or reactor madeof glass (6.5 cm i.d. and 250 ml capacity), equipped witha 2.5 cm diameter six bladed turbine type glass stirrer.The contactor assembly was immersed in a thermostaticwater bath, with a temperature control ±1 �C.

2.3. Epoxidation procedure

A suitable amount of cottonseed oil was taken in thereactor. Calculated amount of carboxylic acid (CH3-COOH/HCOOH) and liquid inorganic acid catalyst(H2SO4/HNO3/H3PO4/HCl) was added to the reactorand the mixture was stirred for about half an hour. Thenthe required amount of 50% aqueous H2O2 was addeddrop-wise at a rate such that the addition was completedin half an hour and the reaction was continued furtherfor the desired time duration. Samples were withdrawnintermittently, considering the completion of H2O2 addi-tion as zero time. The collected samples were then extractedwith diethyl ether in a separating funnel, washed with coldand slightly hot water successively to remove free acid, and

S. Dinda et al. / Bioresource Technology 99 (2008) 3737–3744 3739

then analysed for iodine value, oxirane content, and a-gly-col content.

2.4. Analytical techniques

Iodine value was determined according to Wijs method(Siggia and Hanna, 1979; Weiss, 1970). The percentage oxi-rane oxygen was determined by direct method using hydro-bromic acid solution in glacial acetic acid (Paquot, 1979).a-Glycol content was determined by the method reportedby May (1973), based on oxidation of glycol with benzylt-rimethylammonium periodate in nonaqueous medium. Theexcess of periodate and iodic acid formed are reduced withpotassium iodide and the liberated iodine is titrated withsodium thiosulphate. From the oxirane content values,the relative percentage conversion to oxirane was calcu-lated using the following expression:

Relative percentage conversion to oxirane

¼ ðOOexp=OOtheÞ � 100 ð1Þ

where OOexp (g/100 g sample) is the experimentally ob-tained oxirane oxygen and OOthe is the theoreticallyobtainable maximum oxirane oxygen, which was deter-mined from the following expression:

OOthe ¼ðIV0=2AiÞ

100þ ðIV0=2AiÞAo

� �Ao � 100 ð2Þ

where Ai (126.9) and Ao (16.0) are the atomic weights of io-dine and oxygen respectively and IV0 is the initial iodinevalue of oil sample.

The relative percentage conversion to a-glycol was cal-culated from the following expression:

Relative percentage conversion to a-glycol

¼ ðGexp=GtheÞ � 100 ð3Þ

where Gexp (mol/100 g sample) is the experimentally ob-tained a-glycol and Gthe is the theoretically obtainablemaximum a-glycol, which was determined from the follow-ing expression:

Gthe ¼ðIV0=2AiÞ

100þ ðIV0=2AiÞ2AOH

� �� 100 ð4Þ

where AOH (17.0) is the molecular weight of hydroxylgroup.

The properties of oil, as experimentally determined wereas follows: specific gravity 0.909 at 30 �C; iodine value 105–106 (g I2/100 g oil); acid value 0.5 (mg KOH/g oil) andsaponification value 192 (mg KOH/g oil).

Epoxidation was carried out with the parameters variedin the following ranges: stirring speed 600–2400 rev/min;temperature 30–75 �C; H2O2-to-ethylenic unsaturationmole ratio 1.1–2.5; CH3COOH-to-ethylenic unsaturationmole ratio 0.25–0.75. The liquid inorganic acids used ascatalysts for epoxidation were H2SO4, HNO3, H3PO4,and HCl. The catalyst concentration was expressed as apercentage of ‘‘total weight of H2O2 and CH3COOH added

together’’ and was 1–3% for H2SO4; 1.5% for the otherthree acids.

We have repeated some of the experiments under identi-cal conditions to find out the percentage deviation betweentwo experimental results and we found that the deviationwas between 1.5–3.5%.

3. Results and discussion

3.1. Effect of stirring speed

To investigate the effect of stirring speed, 0.2 mole oilwas treated with H2O2 (H2O2-to-ethylenic unsaturationmole ratio 2.0), CH3COOH (CH3COOH-to-ethylenicunsaturation mole ratio 0.5), and 2% concentrated H2SO4

at 60 ± 1 �C. The experiments were performed in the rangeof stirring speeds 600–2400 rev/min. It was observed fromthe experimental results that the relative percentage con-version to oxirane after 4 h was 69.3%, 72.2%, 77.3% and77.0%, when the stirring speeds were 600, 1200, 1800 and2400 rpm, respectively, under otherwise identical condi-tions. It can be seen that the oxirane formation rate wasnot substantially affected by stirring speeds beyond1800 rev/min, and hence, it was assumed that the reactionwas free from mass transfer resistance beyond 1800 rev/min under the given conditions. However, all subsequentexperiments were performed at a stirring speed of2400 ± 25 rev/min to ensure that the reaction was kineti-cally controlled.

3.2. Effect of temperature

To study the effect of temperature on epoxidation rate,H2O2 (H2O2-to-ethylenic unsaturation mole ratio 1.5),CH3COOH (CH3COOH-to-ethylenic unsaturation moleratio 0.5), and 2% concentrated H2SO4 were added with0.2 mole cottonseed oil under a stirring speed of2400 rev/min and the runs were taken at different tempera-tures in the range 30–75 �C. The oxirane value and a-glycolconversion are shown in Fig. 1 and the iodine value conver-sion is shown in Fig. 2. It was found that as temperatureincreased, the epoxidation rate increased. At lower temper-atures (30 and 45 �C), the relative percentage conversion tooxirane continuously increased within the experimentaltime limit. However, at higher temperature (60 �C), the rel-ative percentage conversion to oxirane attained a maxi-mum, after which it gradually decreased. The relativepercentage conversion to oxirane showed a continuousdecrease at 75 �C, after some increase during the initialphase of the reaction. This indicates that an increase intemperature not only increased the epoxidation rate butalso increased the rate of hydrolysis (oxirane cleavage) ofthe product. Reaction at lower temperatures gave lowerepoxidation rate but led to lesser ring opening. FromFig. 1, it is clear that time required for the attainment ofthe maximum relative conversion at different temperatureswas different. These results suggested that an optimum

0 10 12 140

20

40

60

80

for α-glycol 30°C 45°C 60°C 75°C

rela

tive

% c

onve

rsio

n to

oxi

rane

and

α-g

lyco

l

time (h)

for oxirane 30°C 45°C 60°C 75°C

2 4 6 8

Fig. 1. Effect of temperature on relative percentage conversion to oxiraneand a-glycol. Conditions: mole H2O2 per mole unsaturation = 1.5; moleCH3COOH per mole unsaturation = 0.5; H2SO4 = 2%; and stirringspeed = 2400 ± 25 rev/min.

100 120

20

40

60

80

100

30°C 45°C 60°C 75°Cpe

rcen

tage

con

vers

ion

of io

dine

val

ue

time (h)2 4 6 8

Fig. 2. Effect of temperature on percentage conversion of iodine value.Conditions: mole H2O2 per mole unsaturation = 1.5; mole CH3COOH permole unsaturation = 0.5; H2SO4 = 2%; and stirring speed = 2400 ±25 rev/min.

0 10 120

5

10

40

60

80

for α-glycol 1.1 mole 1.5 mole 2.0 mole 2.5 mole

rela

tive

% c

onve

rsio

n to

oxi

rane

and

α-g

lyco

l

time (h)

for oxirane 1.1 mole 1.5 mole 2.0 mole 2.5 mole

2 4 6 8

Fig. 3. Effect of H2O2 concentration on relative percentage conversion tooxirane and a-glycol. Conditions: temperature 60 ± 1 �C; moleCH3COOH per mole unsaturation = 0.5; H2SO4 = 2%; and stirringspeed = 2400 ± 25 rev/min.

3740 S. Dinda et al. / Bioresource Technology 99 (2008) 3737–3744

level of epoxidation could be obtained within a shortertime at moderate reaction temperature of about 60 �C.

3.3. Effect of H2O2-to-ethylenic unsaturation mole ratio

The effect of H2O2-to-ethylenic unsaturation mole ratioon the conversion to oxirane and a-glycol was studied inthe ratio range 1.1–2.5. Fig. 3 shows that the epoxidationrate increased as the concentration of H2O2 in the systemincreased. Although the maximum conversion to oxiranewas obtained for a mole ratio of 2.5, the stability of the oxi-rane ring was very poor at this high mole ratio. On theother hand, at low concentrations of H2O2, oxirane ringwas quite stable. It was also observed that the formationof a-glycol increased as this ratio increased (Fig. 3). Almostnegligible difference was observed in the final conversions

attained for mole ratios of 1.5 and 2.0. Therefore, withinthe experimental range explored in this work, the optimumH2O2-to-ethylenic unsaturation mole ratio appeared to be2.0.

3.4. Comparison of CH3COOH and HCOOH as oxygencarriers

To compare the effect of CH3COOH and HCOOH asoxygen carriers on epoxidation reaction, we carried outthe reaction under otherwise identical conditions. The rel-evant results are shown in Table 1. The results showed that,initially the oxirane conversion was comparatively higher(�67.5%) in the presence of HCOOH than for CH3COOH(�55.7%). However, with as time progressed, oxirane con-version decreased and iodine value conversion increasedgradually in the case of HCOOH, whereas withCH3COOH, iodine value conversion increased gradually.However, oxirane conversion first increased and reacheda maximum value (�77%) and then started decreasing withtime. Moreover, these results also showed that the differ-ence between iodine value conversion and oxirane conver-sion (at maximum oxirane value) was higher (�21%) in thecase of HCOOH than for CH3COOH (�14%). This meansthat more amount of undesired products were formed inthe presence of HCOOH. Therefore, in spite of the slowerrate of formation of oxirane in the presence of CH3COOH,the ultimate conversion to oxirane was higher and hence itwas chosen as the oxygen carrier for the present work.

3.5. Effect of CH3COOH-to-ethylenic unsaturation mole

ratio

To investigate the effect of CH3COOH concentration onthe epoxidation rate and glycol formation of cottonseedoil, the mole ratio was varied in the range 0.25–0.75.

Table 1Comparison of CH3COOH and HCOOH as oxygen carriers

Organic acid Time (h)

1 2 3 4 6 8

CH3COOH % Conversion of iodine value 69.6 83.0 88.6 91.2 93.1 93.9Relative % conversion to oxirane 55.7 65.2 72.4 77.0 72.2 66.0

HCOOH % Conversion of iodine value 88.5 90.0 91.0 92.2 94.0 94.6Relative % conversion to oxirane 67.5 66.7 65.9 63.5 59.9 53.1

Reaction conditions: mole H2O2 per mole unsaturation = 2; mole CH3COOH per mole unsaturation = 0.5; H2SO4 = 2%; temperature 60 �C; and stirringspeed = 2400 ± 25 rev/min.

time (h)0 10 12

0

5

10

40

50

60

70

80

for α-glycol 0.25 mole 0.35 mole 0.5 mole 0.75 mole

for oxirane 0.25 mole 0.35 mole 0.5 mole 0.75 mole

rela

tive

% c

onve

rsio

n to

oxi

rane

and

α-g

lyco

l

2 4 6 8

Fig. 4. Effect of CH3COOH concentration on relative percentageconversion to oxirane and a-glycol. Conditions: temperature 60 ± 1 �C;mole H2O2 per mole unsaturation = 2; H2SO4 = 2%; and stirringspeed = 2400 ± 25 rev/min.

0 1015

30

45

60

75

rela

tive

% c

onve

rsio

n to

oxi

rane

time (h)

H2SO

4(1.5 %)

H3PO

4(1.5 %)

HNO3 (1.5%)

HCl (1.5%)

2 4 6 8

Fig. 5. Effect of inorganic acid catalysts on relative percentage conversionto oxirane. Conditions: temperature 60 ± 1 �C; mole H2O2 per moleunsaturation = 2; mole CH3COOH per mole unsaturation = 0.5; andstirring speed = 2400 ± 25 rev/min.

S. Dinda et al. / Bioresource Technology 99 (2008) 3737–3744 3741

CH3COOH takes part in the overall reaction as a catalystin the formation of oxirane ring and as a reactant in thehydrolysis of the oxirane ring. Fig. 4 shows that the epox-idation rate increased as the concentration of CH3COOHin the system increased. It has been observed that at highmole ratio of CH3COOH, the stability of the oxirane ringwas very poor. It was also observed that, the formationof a-glycol increased as the CH3COOH mole ratioincreased (Fig. 4). Therefore, to obtain the maximum oxi-rane, the optimum level of concentration of CH3COOHshould be used where both the effects are optimised. Hence,within the experimental conditions, the most favorableconcentration of CH3COOH appeared to be 0.5 mole ratio.

3.6. Suitability of various inorganic acids as catalysts

The ability of some common and commercially availableinorganic acids (namely, H2SO4, H3PO4,HNO3, and HCl)to serve as catalysts was studied under identical kineticallycontrolled conditions. The relative percentage conversionsto oxirane oxygen in the presence of four acid catalystsare shown in Fig. 5. Fig. 5 shows that the oxirane conver-sions with HCl and HNO3 were significantly smaller thanthose obtained with H2SO4 and H3PO4 as catalysts, at

any particular time instant. Evidently, HCl and HNO3

were not very effective for the epoxidation of cottonseedoil. In the case of H3PO4, the oxirane value increasedmonotonically within the experimental time limit andabout 67.5% oxirane conversion was obtained after 10 h.In the case of H2SO4, the oxirane value first increasedand reached a maximum, and then decreased with time.The decrease in oxirane value was due to cleavage of theoxirane ring. Nevertheless, the highest conversion to oxi-rane (�70.4%) was in the case of H2SO4, obtained withinthe shortest time (after 6 h). Therefore, based on the exper-imental results, H2SO4 was chosen as catalyst for the pres-ent work. Goud et al. (2006a) have also reported similarobservation.

3.7. Effect of H2SO4 concentration

To investigate the effect of H2SO4 concentration on theepoxidation reaction, concentration of acid was variedfrom 1% to 3%. In general, with an increase in acid concen-tration, oxirane conversion increased. Fig. 6 shows that,when the acid concentration was increased from 1% to2% the reaction time taken to reach the maximum conver-sion of oxirane value, reduced. It was also observed that,

0 100

5

10

40

50

60

70

80

for α-glycol 1.0% 1.5% 2.0% 3.0%

for oxirane 1.0% 1.5% 2.0% 3.0%

rela

tive

% c

onve

rsio

n to

oxi

rane

and

α-g

lyco

l

time (h)2 4 6 8

Fig. 6. Effect of H2SO4 concentration on relative percentage conversion tooxirane and a-glycol. Conditions: temperature 60 ± 1 �C; mole H2O2 permole unsaturation = 2; mole CH3COOH per mole unsaturation = 0.5;and stirring speed = 2400 ± 25 rev/min.

0 10 123.100

3.125

3.150

3.175

3.200ln

{[H

2O2] 0 -

[EP

]}

time (h)

30°C

45°C

60°C

75°C

2 4 6 8

Fig. 7. Plot of ln{[H2O2]0 � [EP]} versus time for epoxidation ofcottonseed oil by peroxyacetic acid.

2.8 2.9 3.0 3.1 3.2 3.3 3.4-17

-16

-15

-14

-13

-12

ln (k

)

1/T x 103 (K-1)

Fig. 8. Arrhenius plot for epoxidation of cottonseed oil by peroxyaceticacid.

3742 S. Dinda et al. / Bioresource Technology 99 (2008) 3737–3744

with an increase in acid concentration, glycol formationincreased (Fig. 6). When the catalyst loading was increasedto 3%, higher oxirane cleavage was observed with a corre-spondingly lower oxirane value. Therefore, optimum con-version to oxirane was obtained at 2% loading of H2SO4.

4. Epoxidation kinetics

The reaction steps as suggested by Gan et al. (1992) areas follows:

CH3COOHþH2O2�CH3COOOHþH2O ð5ÞO

CH --HC CH3COOH-CH = CH - + CH3 COOOH +

ð6Þ

O

HC CH CH(OH) CH( OH ) -H2O H+++

ð7Þ

O

HC CH CH (OH) CH (OCOCH3)+ CH 3COOH

ð8Þ

In situ epoxidation reaction generally takes place in twosteps: (i) formation of peroxyacetic acid, and (ii) reaction ofperoxyacetic acid with the unsaturated double bond. If thefirst step is considered rate determining, and the concentra-tion of peroxyacetic acid is assumed constant throughoutthe reaction, then the following rate laws apply (Ganet al., 1992).

d½EP�dt¼ kf½H2O2�0 � ½EP�g � ½CH3COOH�0 ð9Þ

where subscript 0 denotes initial concentrations, [EP] is theconcentration of oxirane oxygen and [CH3COOH] is theconcentration of organic acids.

The concentration of acetic acid is assumed to remainconstant throughout the reaction with negligible degrada-tion of oxirane and is approximated to [CH3COOH]0 forthe following reasons: (i) the equilibrium constant (Keq)for the peracetic acid formation (Eq. (5)), Keq�2 (Jankovicand Fiser, 2005) and the dissociation constant of the aceticacid, KAcOH � 1. 75 · 10�5 (Pahari, 1990) are small; (ii)acetic acid is regenerated in the epoxidation process (Eq.(6)). Gan and Ng (1986) have reported similar type ofexplanation for the epoxidation of rubber latex by formicacid.

Integration of Eq. (9) yields,

lnf½H2O2�0 � ½EP�g ¼ �k½CH3COOH�0t þ ln½H2O2�0 ð10Þ

According to Eq. (10), a plot of ln{[H2O2]0–[EP]} versustime should yield straight lines for those reactions with neg-ligible degradation of oxirane. However, deviations from

S. Dinda et al. / Bioresource Technology 99 (2008) 3737–3744 3743

linearity were observed for the reactions with substantialring opening in some cases. Therefore, in such cases, theepoxidation rate constants were calculated from the initialslopes. By assuming that the specific gravity of the cotton-seed oil remained constant during the course of the reac-tion, [EP] was calculated from the experimentallydetermined oxirane value. Plots of ln{[H2O2]0–[EP]} versustime for in situ epoxidation of cottonseed oil at differenttemperatures are shown in Fig. 7. The rate constants valuesobtained for the reactions were 0.39 · 106 L mol�1 s�1,1.44 · 106 L mol�1 s�1, 2.32 · 106 L mol�1 s�1 and 5.42 ·106 L mol�1 s�1 at 30 �C, 45 �C, 60 �C, and 75 �C tempera-ture respectively. Activation energy (Ea) for the reactionwas calculated from the slope of the Arrhenius plot(Fig. 8) and was found to be about 11.7 kcal mol�1. Thisvalue was comparable with the values of 15.7, and15.1 kcal mol�1 reported for the epoxidation of rubber seedoil (Okieimen et al., 2002); and for the epoxidation ofmethyl esters of palm olein (Gan et al., 1992), respectivelyand 14.9 and 14.5 kcal mol�1 for karanja oil and for mahuaoil as reported by Goud et al. (2006a,b).

5. Thermodynamic properties of the cottonseed oil

epoxidation

The enthalpy of activation (DH), was estimated usingthe Eq. (11) (Kapoor, 1985),

DH ¼ Ea � RT ð11Þwhere ‘Ea’ is activation energy calculated from Arrheniusplot, ‘T’ is temperature and ‘R’ is universal gas constant.

The calculated value of enthalpy of activation at 60 �Cwas 11.04 kcal mol�1. The entropy of activation (DS) andfree energy of activation (DF) were calculated at the sametemperature using the Eqs. (12) and (13), respectively(Kapoor, 1985),

k ¼ RTNh

eDS=Re�DH=RT ð12Þ

DF ¼ DH � T DS ð13Þ

where ‘k’ is rate constant, ‘N’ is Avogadro constant, ‘h’Planck’s constant.

The calculated value of DS and DF were �51.4cal mol�1 K�1 and 28.1 kcal mol�1, respectively. The ther-modynamic properties show that, the present reaction isendothermic in nature because the enthalpy of activationis positive and hence an increase in the reaction tempera-ture may lead to an increase in the conversion to oxiraneoxygen. As the free energy of activation is positive, thereaction is non-spontaneous under the present experimen-tal conditions. With an increase of temperature, the non-spontaneity of the reaction should increase since DF willbe more positive and hence, at any particular time instant,the yield of oxirane will increase. However, the net conver-sion to oxirane will maximise if the in situ epoxidationoccurs at optimum temperature, which depends on the typeof catalyst used in such process.

6. Conclusions

The epoxidation of cottonseed oil using in situ generatedperoxyacid could be carried out at moderate temperatureof about 60 �C. CH3COOH was found to be more effectiveoxygen carrier than HCOOH, in the present work. Out ofall the liquid inorganic acid catalysts studied, H2SO4 wasfound to be the most efficient and effective. Higher temper-ature and higher acid concentrations reduced the reactiontime needed to reach the maximum conversion to oxiranevalue; however, it simultaneously increased the extent ofoxirane ring cleavage to glycols. The reaction was kineti-cally controlled beyond a stirring speed of about1800 rev/min. The rate constants obtained for the reactionwere in the range of 0.39–5.42 · 106 L mol�1 s�1 and fromthe relative percentage conversion data obtained for vari-ous reaction parameters, it could be concluded that it ispossible to develop value–added products from a locallyavailable renewable natural resource such as cottonseedoil. Maximum yield of oxirane with negligible amount ofoxirane cleavage could be obtained if the epoxidation ofcottonseed oil, using in situ generated peroxyacetic acid,is carried out at optimum conditions. These optimumconditions include a temperature range of 50–60 �C, H2-O2-to-unsaturation mole ratio range 1.5–2.0, CH3COOH-to-unsaturation mole ratio of about 0.5 and H2SO4 loadingof about 2% (by weight) of the aqueous phase.

References

Biermann, U., Friedt, W., Lang, S., Luhs, W., Machmuller, G., Metzger,J.O., Klaas, M.R., Schafer, H.J., Schneider, M.P., 2000. New syntheseswith oils and fats as renewable raw materials for the chemical industry.Angew. Chem. Int. Ed. 39, 2206–2224.

Chakrabarty, M.M., 2003. Chemistry and Technology of Oils and Fats.Allied Publishers, New Delhi, pp. 151–152.

Gan, L.H., Ng, S.C., 1986. Kinetic studies of the performic acidepoxidation of natural rubber latex stabilized by cationic surfactant.Eur. Polym. J. 22, 573–576.

Gan, L.H., Goh, S.H., Ooi, K.S., 1992. Kinetic studies of epoxidation andoxirane cleavage of palm olein methyl esters. J. Am. Oil Chem. Soc. 69,347–351.

Gan, L.H., Ooi, K.S., Gan, L.M., Goh, S.H., 1995. Effect of epoxidationon the thermal oxidative stabilities of esters of fatty acids derived frompalm olein. J. Am. Oil Chem. Soc. 72, 439–442.

Goud, V.V., Patwardhan, A.V., Pradhan, N.C., 2006a. Studies on theepoxidation of mahua oil (Madhumica Indica) by hydrogen peroxide.Bioresour. Technol. 97, 1365–1371.

Goud, V.V., Pradhan, N.C., Patwardhan, A.V., 2006b. Epoxidation ofkaranja (Pongamia Glabra) oil by H2O2. J. Am. Oil Chem. Soc. 83,635–640.

Guenter, S., Rieth, R., Rowbottom, K.T., 2003, 6th ed. Ullmann’sEncyclopedia of Industrial Chemistry, vol. 12 John Wiley and Sons,pp. 269–284.

Jankovic, M., Fiser, S.S., 2005. Prediction of the chemical equilibriumconstant for peracetic acid formation by hydrogen peroxide. J. Am. OilChem. Soc. 82, 301–303.

Kapoor, K.L., 1985. A Textbook of Physical Chemistry, vol. 4. MacmillanIndia Limited, New Delhi, Chapter 6.

Klaas, M.R., Warwel, S., 1999. Complete and partial epoxidation ofplant oils by lipase-catalysed perhydrolysis. Ind. Crops Prod. 9, 125–132.

3744 S. Dinda et al. / Bioresource Technology 99 (2008) 3737–3744

May, C.A., 1973. Epoxy Resins: Chemistry and Technology. MarcelDekker, New York, pp. 672–673.

Okieimen, F.E., Bakare, O.I., Okieimen, C.O., 2002. Studies on theepoxidation of rubber seed oil. Ind. Crops Prod. 15, 139–144.

Pahari, S., 1990. Physical Chemistry, first ed. New Central Book Agency,Kolkata, p. 569.

Paquot, C., 1979. Standard Methods for the Analysis of Oils, Fats andDerivatives Part-1, fifth ed. Pergamon Press, Germany, pp. 52–93.

Rios, L.A., Weckes, P., Schuster, H., Hoelderich, W.F., 2005. Mesopor-ous and amorphous Ti-silicas on the epoxidation of vegetables oils. J.Catal. 232, 19–26.

Sharpless, K.B., Woodard, S.S., Finn, M.G., 1983. On the mechanism oftitanium tartrate catalyzed asymmetric epoxidation. Pure Appl. Chem.55, 1823–1836.

Sheldon, R.A., Dakka, J., 1994. Heterogeneous catalytic oxidations in themanufacture of fine chemicals. Catal. Today 19, 215–246.

Siggia, S., Hanna, J.G., 1979. Quantitative Organic Analysis via Func-tional Groups. Wiley Interscience, New York, pp. 394–398.

Srivastava, A., Prasad, R., 2000. Triglycerides-based diesel fuels. Renew.Sustain. Energy Rev. 4, 111–133.

Wallace, J.C., 1978, third ed. Encyclopedia of Chemical Technology, vol.9 John Wiley and Sons, New York.

Weiss, F.T., 1970. Determination of Organic Compounds: Methods andProcedures. Wiley Interscience, New York, pp. 34–44.