1

Indian Journal of Textile Research

Vol. 7, June 1982, pp. 49-55

Thermal Degradation of Cellulose and Its Esters in Air

RAJESH K JAIN, KRISHAN LAL & HARI L BHATNAGAR *

Department of Chemistry, Kurukshetra University, Kurukshetra 132119

Receil'ed 9 May 1982; accepted 25 May 1982

The thermal degradation of cellulose and its esters (fJ-chloropropionate, acetate) in air has been studied by dynamicthermogravimetry and differential thermal analysis from ambient temperature to 600'C. The values of the energy of activationof thermal degradation, Ea, for cellulose and its esters lie in the range 155-225kJ mol -, . However, the values of free energy ofactivation for the degradation of cellulose and its esters are almost same and lie in the range 152-185kJ mol -, . This indicatesthat the basic step in the thermal degradation of cellulose and its derivatives is the same. IR spectra of the pyrolysis residues ofcellulose fJ-chloropropionate gave indication of dehydration and formation of a compound having C =0 groups.

Cellulose and its esters are used widely in the textileindustry, and their thermal degradation enables us tounderstand the mechanism which renders them more

resistant to heat and flame. The thermal degradationof cellulose itself has been investigated both in l'acuoand in an inert atmosphere and a mechanism for

decomposition of cellulose has been proposed 1 • Themajor weight loss in the pyrolysis of cellulose occursbelow 400°C with the formation of water, carbon

mOl)oxide, carbon dioxidez and levoglucosan3.Esterification of cellulose with halogen-substitutedacids provides an additional advantage, as its has beenobserved that certain elements like nitrogen andhalogens when incorporated into the matrix ofcellulose impart flame resistance to it. In the presentstudy, cellulose samples treated with {3-chloropropionic acid have been obtained and thekinetics of the thermal degradation of cellulose and itsesters in air atmosphere from ambient temperature to600°C has been studied using dynamic TG, DTG andDllA techniques. Different kinetic parameters havebeen obtained adopting the methods of Freeman andCarroll 4 , Broidos, Dave and Chopra6 andKissinger 7 •

Materials and MethodsCellulose supplied by Mis Schleicher and Schull,

Dassel, West Germany and dried to constant weight invacuo at 60°C was used. Xylene (BDH) was freed fromthiophene, distilled and dried over sodium wire. {3-Chloropropionic acid (Fluka, Switzerland) andperchloric acid (60%) (Riedel-De-Haenag, Germany)were used as such.

The following samples of cellulose and itsderivatives having different degrees of substitutionwere used: (i) cellulose; (ii) - (iv) cellulose {3-

chloropropionate samples obtained by treatingcellulose (1.62 g) with {3-chloropropionic acid (1.0853g, 2.1706 g and 3.2559 g in I: I, 1:2 and 1:3 molar ratiorespectively) in xylene, mixed with perchloric acid (8%on the weight of cellulose) and refluxing the mixtureunder nitrogen atmosphere for 2 hr; the products werefiltered through G-4 sintered glass funnel, washedthoroughly with water followed by ethanol and driedover PzOs in l'acuo; and (v) cellulose acetate (BDHChemicals Ltd, Poole, England) extracted with ethanolin a soxhlet apparatus, dried in air, washed with hotdistilled water, and again dried in air. The purifiedmaterial was dried in vacuo at 60°C. The degrees ofsubstitution of the different samples are given inTable I.

Thermogravimetric analysis, derivative thermogra-vimetric analysis and differential thermal analysis-The TG and DTG thermograms were obtained using aStanton Redcroft TG 750. The TG and DTG curves

were run under a static air atmosphere and a scalesensitivity t)f 0.25 mg/cm and 0.1 (mg/min)/cmrespectively. The TG measurements were made at aconstant heating rate of 10°C/min.

The DTA-02-Universal model (Mis VEBLaborelektronic Halle, GDR) was used for recordingthe thermograms. The DT A curves were run under astatic air atmosphere with the use of calcined aluminaas a reference. The DT A measurements were made at a

heating rate of 7.9°, 12.3°, 18.8° and 24.6°C/min.Infrared studies-For the IR studies (Beckman

spectrophotometer IR-20, USA), 2% charred samplesof cellulose {3-chloropropionate (sample iii only)obtained after heating it to different temperatures wereprepared by the KBr technique. The charred sampleswere prepared by heating them in a DT A cell in staticair atmosphere. Heating was stopped at the desired

49

INDIAN J TEXT RES, VOL 7, JUNE 1982

temperature and the residues were allowed to cool andquickly transferred to a sealed sample container.

Results and Discussion

The DT A thermograms of pure cellulose (i),cellulose p-chloropropionate (ii-iv) (thermograms ofonly sample iii shown in the figures) and celluloseacetate (v) were recorded in air and are shown in Figs 1-3. The peak temperatures of the curves at differentheating rates were measured and are presented inTable I. The DT A curves for pure cellulose shows an

endotherm loop in the temperature range 700 -80°C. Atabout 250°C, the exothermic effect starts with peakingat 322°C, which may be due to auto-oxidation of

carbonyl group and C-H bond. The endotherm at336°C following the exotherm is presumably due to theoxygen in the sample atmosphere being exhaustedbefore the combustion is complete, so that theremainder decomposes in an inert atmosphere. Thisendotherm seems to be associated with the pyrolysis ofcellulose into water, carbon monoxide, carbondioxide2 and levoglucosan 3 •

Table I-Peak Temperatures in the DT A Thermograms and Heats of Pyrolysis (d If) for Cellulose and Its Derivatives in Air

JAIN et at.: THERMAL DEGRADATION OF CELLULOSE & ITS ESTERS IN AIR

... (l)

dW Z (E )- dT=RT= RH exp - RT . W' (ref. 11) .. (la)

retardant, as the role of the flame retardant is to lower

the decomposition temperature, so that a lowerpercentage of flammable volatiles is produced and acorrespondingly larger amount of char is formed 8,9.

The values of the heat of pyrolysis 10, I1H, of purecellulose and treated cellulose samples are presented inTable 1. It is seen that the exothermic combustion

value, I1H, for pure cellulose is 408 J g -1, whichdecreases considerably on the introduction ofchloropropionyl group, indicating that the burningvelocity of gases decreases. The values of I1H furtherdecrease as the chlorine content of the samplesincreases. This may be due to the replacement of theexothermic carbon-hydrogen bond in pure cellulose bythe endothermic carbon-chlorine bond in the treated

cellulose. Further, for the endothermic reaction, theI1H values increase for the treated cellulose samples.This is possibly due to the scission of the carbon-chlorine bond involving further absorption of energy.This energy is probably not compensated by theexothermic polymerization and aromatization of thepyrolytic products.

Kinetic parameters-The weight versus temperaturecurves for cellulose and treated cellulose samples (iiiand v) are shown in Figs 1-3. The initial rapid but smallweight loss is attributed to the loss of sorbed moistureand has been neglected. Kinetic parameters forthe decomposition of cellulose and treated cellulosesamples were determined using the proceduresdescribed by Freeman and Carro1l4, Broid05, Daveand Chopra 6 and Kissinger 7 •

Freeman and Carro1l4 derived the followingequations for the calculation of energy of activationand frequency factor from the TG curve:

l1(l\I1I0gRT_n_ Ea ~11 log W' 2.303 R . 11 log W

51

and the rate of decomposition,

where W is the weight fraction of the materialundergoing degradation at time t; RH is the rate ofheating CCfmin); and Z, the frequency factor.

From the slopes of the TG curves in Figs. 1-3, thevalues of 11 log RT and 11 log Wwere calculated (Eq. I)

corresponding to an arbitrarily chosen 11 (~ ) interval(in this case 0.2 x 10-4) and plots of 11 log RTf 11 log W

versus 11 (~ ); 11 log W were drawn (Fig. 4). The energyof activation, Ea, was determined from the slope. Theorder of reaction, n, was determined from the intercept

7000,0

o

6,0,- TG

5'01-

\i::;:

a::w:r:•...0xw

~ •...

INDIAN J TEXT RES, VOL 7, JUNE 1982

and is found to be unity for all the cases, Theseparameters have been evaluated using the method ofleast squares for all the samples and for all themethods.

Broid05 used Eq. (2) for determining kineticparameters from TG curve:

... (4)R

-2,0

-"2

-,>-

.;. - 08

JAIN et uf.: THERMAL DEGRADATION OF CELLULOSE & ITS ESTERS IN AIR

and

... (6)

... (7)

... (5)

and !1G* = !1H* -- T !1S* .

The values of !1S* and !1G* for the pyrolysis ofcellulose and cellulose derivatives are given in Table 3.The entropy of activation for thermal degradation ispositive for pure cellulose sample and is lower fortreated cellulose samples. This is due to the fact thatwith increasing degrees of substitution of chlorop-ropionyl or acetyl group in cellulose, the magnitude ofthe degrees of freedom accompanying the formation ofthe activated complex goes on decreasing, thus makinga more tightly bound matrix. The values of free energyof activation for the process at 6000K are found to bealmost same, suggesting that the mechanism ofthermal decomposition of all the chloropropionylatedsamples is basically the same.

samples (ii-iv) and cellulose acetate determined usingEqs. (1)-(4) are presented in Table 2. It is seen that thevalues of the energy of activation for the pyrolyticdegradation of cellulose and treated cellulose samplesobtained using different methods are not in closeagreement. Further, the value of energy of activationfor treated cellulose samples is less than that for purecellulose. This means that the charring temperature for

cellulose {I-chloropropionate samples is lower thanthat for cellulose. The char yield obtained from treatedcellulose samples is higher (Table 3), indicating that thetreated cellulose samples are better flame retardants.

Assuming Ea = !1H* + RT for a solid system 12.13,the values of !1S* and hence !1G* were obtained at the

average peak temperature of DTG (6000K) using thefollowing equations 12 -14:

.,. (4a)

1'68

(a)

1-721·88 1'8' 1-80 1-76.1.... Xl03Tm

1'92

E•.RH n 1 (E.)~ = Zn(1-x)m - exp - -r- ,R.Tm R.~ m

where RH is the heating rate CC/min); T m, the peaktemperature CK); x, the fraction reacted at time t; and

n, the reaction order. Plots of In ~~ versus -f- yieldedm mstraight lines (Fig. 7) and the values of the energy ofactivation were obtained from the slopes of the curves.

Eq. (4a) was used for determining the frequency factor.l;he energy of activation, E., and the frequency

factor, Z, for cellulose, cellulose {I-chloropropionate

-9'8

-11'01'92

-10'2

:x L..ea:r;....£: LI0·6

RH IFig.'7-Plols of In - vs - using Kissinger equalion for pyrolysis

T;' Tmof cellulose (e), cellulose P-chloropropionate sample (ii, A), (iii, D),(iv, 0) and cellulose acetale (6) for (a) exolherm, and (b) endotherm

Table 2-Energy of Activation, Frequency Factor and Reaction Order for Pyrolysis of Cellulose and Its Derivatives in Air

Sample E./(kJ mol-') Frequency factor Z/s-'ReactionNo.·

order, nFreeman

BroidoDaveKissingerFreemanBroidoDaveKissingerand

methodandmethodandmethodandmethodCarroll

Chopra CarrollChopramethod

methodexo-endo-method methodexo-endo-

thermtherm thermtherm

178.9

198.2184.7224.4197.97.16x 10'39.48 x 10'46.11 X 10137.80 x 10.71.24 x 10'51.019", Iii

184.8172.1 -5.26x1O'51.79 x 10'4) iii

172.3181.9169.7178.1160.61.58 x 10.43.18 X 10'42.14xlO'32.50 X 10'52.59 X 10.30.960",1iv'

-- 161.2155.1 .-8.75 x 10131.24 x 10.3v'

163.0183.4166.7201.8184.76.44 x 10"9.41 X 10'22.99 X lO"4.45 X 10'55.47 X 10.30.991 "'I

·Compounds referred to as Samples No. i to v are Ihe same as in Table 1.

53

INDIAN J TEXT RES, VOL 7, JUNE 1982

Table 3-DTG Maxima, Entropy and Free Energy of Activation and Char Yields for Pyrolysis of Cellulose and ItsDerivatives in Air

Sample DTG d S*/(J deg-I mol-I) d G* (kJ mol -I) at 6000KCharNo.*

maximal at 6000K yield at°C

700'KFreeman

BroidoDaveKissingerFreemanBroidoDaveKissinger%and

methodandmethodandmethodandmethod

CarrollChopra CarrollChopra

method

methodexo-endo-method methodexo-endo-

therm

therm thermtherm

344.0

6.227.74.983.529.9170.2176.6178.8169.3175.023.50

II

- -41.913.8 -154.6158.8iii

301.012.818.6-3.835.7-2.3159.7165.8167.0151.6157.027.42

IV

- -7.9-8.4 -151.5155.2v

362.0-33.0-10.7-39.340.54.0177.8184.8185.3172.5177.330.92

*Compounds referred to as Samples No. i to v are the same as in Table 1.

Fig. 8-IR spectra of (a) cellulose {J-chloropropionate sample (iii)and (b-e) chars of cellulose II-chloropropionate sample (iii) at

different temperatures (b) 260°, (c) 290°, (d) 320°, and (e) 350°C

Infrared examination of chars-An inspection of theIR spectra of the chars (Fig. 8) obtained at differenttemperatures from the pyrolytic degradation ofcellulo&e j1-chloropropionate (sample iii) confirms theabove observations. The spectrum for the charobtained at 200°C showed no change in the absorptionbands compared to that of the initial sample (curve a),while at 260°C (curve b), the spectrum showed evidenceof dehydration. There is a decrease in the intensity ofthe hydroxyl stretching vibration bands (3405, 3350,3305 cm -I). Also, at this temperature, the bands due tocellulose j1-chloropropionate at 2900 (C-H str), 1720 (C=0 str), 1430(CHz sym bending), 1370(C-H bending),1335(0-H inplane bending), 1315(CHz wagging), 1280

zainIIIiIIIZ..0:>-

>-zwU0:W0.

3000 2000 1500

WAVENUMBER, cm-1

1000 500

(C-H bending), 1160 (antisym bridge C-O-C str), 1125,and both 1060 and 1035 cm -1 (skeletal vibrationsinvolving C-O str), etc. showed slight decrease,indicating minor degradation of cellulose j1-chloropropionate along with dehydration. At 290°C,all the above-mentioned bands due to cellulose j1-chloropropionate almost disappeared and theabsorption due to C =0 stretching (1720 cm -1)became intense, indicating that the skeletal rearrange-ment and the evolution of volatile products commenceonly at higher temperatures. At 350°C, the conjugatedC = C band at 1640 shifted to 1600 cm -I, suggestingthe extension of conjugation of the C = C bonds in theresidue from cellulose j1-chloropropionate.

Pyrolytic mechanism-Recently, Scotney 15 pro-posed a mechanism for the thermal degradation ofcellulose triacetate in vacuo, suggesting deacetylationin the polymer chain and chain scission at the 1,4-glycosidic linkages between the pyranose rings. Theformer process results in the formation of acetic acidand the latter in the formation of the tarry material.During pyrolysis of the samples of cellulose j1-chloropropionate, most of the chloropropionic acid, ifnot all, is probably formed by cis-eliminationinvolving a j1-hydrogen atom, as suggested byScotney 15. A detailed mechanism for the thermaldegradation of cellulose derivative has been proposedelsewhere 14 .

AcknowledgementOne of the authors (R.K.J.) is thankful to

autohorities of Kurukshetra University, Kurukshetra

for providing a research fellowship.

References

1 Fairbridge C, Ross R A & Sood S P, J appl Po/ym Sci, 22 (1978),497, and the references cited therein.

54

JAIN et al.: THERMAL DEGRADATION OF CELLULOSE & ITS ESTERS IN AIR

2 Madorsky S L, Hart V E & Straus S, J Res natn Bur Stand, 56(1956) 343.

3 ShaflZadeh F & Fu Y L, Carbohydr Res, 29 (1973) 113.

4 Freeman E S & Carroll B, J phys Chem, 62 (1958) 394.

5 Broido A, J polym Sci, Part A-2, 7 (1969) 1761.6 Dave N G & Chopra S K, Z phys Chem, 48 (1966) 257.

7 Kissinger H E, Analyt Chem, 29 (1957) 1702.

8 Schuyten H A, Weaver J W & Reid J D, cited in Fire retardantpaints, Advances in chemistry series I, No 9, (AmericanChemical Society, Washington), 1954, 7.

9 Schuyten H A, Weaver J W & Reid J D, Ind Engng Chem, 47(1955) 1433.

10 Tang W K & Neill W K, J Polym Sci, polym Symp, Part C, 6(1964) 65.

II Reich L & Stivala S S, Elements of polymer degradation,(McGraw-Hili Book Co, New York), 1971, 102.

12 Laidler K J, Chemical kinetics, (Tata McGraw-Hili, New Delhi),1976, 86-90.

13 Moore W J, Physical chemistry, (Longmans Green & Co,London), 1966, 297.

14 Jain R K, Lal K & Bhatnagar H L, Makromolek Chem, 183,(1982),000.

15 Scotney A, Eur Polym J, 8 (1972), 163, 175, 185.

55