Preparation of activated carbon from lignin by chemical activation

PERGAMON

Carbon 38(2000)1873–1878

Preparation of activated carbon from lignin by chemical

activation

a ,a a b

*Jun’ichi Hayashi ,Atsuo Kazehaya ,Katsuhiko Muroyama ,A.Paul Watkinson a

Department of Chemical Engineering ,Kansai University ,3-3-35Yamate -cho ,Suita ,Osaka ,564-8680Japan

b

Pulp and Paper Centre ,Department of Chemical and Bio -Resource Engineering ,The University of British Columbia ,Vancouver ,BC ,

Canada V 6T 1Z 4

Received 30September 1999;accepted 25January 2000

Abstract

Activated carbons were prepared from lignin by chemical activation with ZnCl ,H PO and some alkali metal 234compounds.The in?uence of carbonization and activating reagent on the pore structure of the activated carbon was investigated.It was found that the maximum surface areas were obtained at the carbonization temperature of 6008C in both ZnCl and H PO activation,and that the surface areas were as large as those of the commercial activated carbons.On the 234other hand,in alkali metal activation it was found that the maximum surface areas were obtained at the carbonization temperature of 8008C.Except for Na CO maximum surface areas were much larger than those of the commercial activated 232carbons.The activated carbon prepared by K CO activation showed a surface area of nearly 2000m /g.It was shown that 23ZnCl works effectively as dehydration reagent below 6008C.On the other hand,K CO works effectively in two 223temperature ranges,below 5008C and above 6008C.Below 5008C,the carbonization behavior was modi?ed by impregnation with K CO ,but the pore structure changes little.Above 6008C,carbon was consumed by K CO reduction and then the 2323surface area was increased.?2000Elsevier Science Ltd.All rights reserved.

Keywords :A.Activated carbon;B.Activation;C.BET surface area;D.Surface areas

1.Introduction

greatly desired and the production of activated carbons from wastes is an interesting possibility.Lignin is a waste Activated carbon is a high-porosity material which is which is generally used only for its fuel value.Therefore,useful in adsorption of both gases and solutes from it was of interest to prepare a higher value product such as aqueous solution.Therefore it has been widely used for the an activated carbon from lignin.

separation of gases,the recovery of solvents,the removal There are basically two methods for preparing activated of organic pollutants from drinking water and as a catalyst carbon:physical and chemical activation.Physical activa-support.As environmental pollution is becoming a more tion consists of two steps:the carbonization of the starting serious problem,the need for activated carbon is growing.material,and the activation of the char by using carbon Coals and ligno-cellulosic materials are commonly used dioxide or steam.In chemical activation both the carboni-as the starting material for preparing activated carbon.zation and the activation step proceed simultaneously.Recently,numerous attempts to prepare activated carbon Many studies for preparing activated carbon by chemical from solid wastes have also been made [1–10].The activation [1,4,5,9–16]have been done.Recently,acti-development of methods to re-use waste materials is

vated carbons with very high surface area have been prepared by chemical activation [11–14].Gas storage is a potential application for such high surface area materials.However,the purpose of most of these studies was to *Corresponding author.Tel.:181-6-6368-0913;fax:181-6-prepare the activated carbon,and little attention has been 6388-8869.

E -mail address :hayashi7@kansai-u.ac.jp (J.Hayashi)

given to the in?uence of the activating reagent on the pore

0008-6223/00/$–see front matter ?2000Elsevier Science Ltd.All rights reserved.PII:S0008-6223(00)00027-0

1874J .Hayashi et al ./Carbon 38(2000)1873–1878

Table 2

structure and to the mechanism of activation.Therefore it Speci?c surface areas and pore volumes of commercial activated is of interest to investigate these matters.

carbons

The purposes of the present work were to prepare activated carbons from lignin by chemical activation and to Speci?c surface Micropore Mesopore examine the in?uence of preparation conditions (carboniza-area volume volume 2(m /g)

(ml/g)(ml/g)tion temperature and chemical reagent)on the pore struc-ture of the activated carbon and to clarify the temperature Coconut shell AC 12670.4920.061range where the chemical reagents work effectively.

Coal AC

1089

0.428

0.179

2.Experimental

Chemical recovery 5

2.1.Preparation of activated carbon

(weight of sample before washing)2(weight of sample after washing)]]]]]]]]]]]]]].

(weight of impregnated chemical)

Strong black liquor from Kraft pulping of a mainly 2.2.Characterization of pore structure

spruce wood furnish from New Brunswick,Canada was acidulated with pure CO to precipitate the lignin,which 2The adsorption isotherm of N was measured on the 2was then washed in acid several times and dried.Ultimate activated carbon at 77K by use of a constant volume analysis of the lignin on a dry basis is given in Table 1.Six adsorption apparatus (BELSORP 28,BEL Japan Inc.).chemicals (K CO ,Na CO ,KOH,NaOH,ZnCl ,23232Before measuring the isotherm,the activated carbon was H PO )were used as activating reagents.

34heated at 2008C for 2h in vacuo to clean its surface.The Lignin was mixed with the activating reagent and water,surface area of the activated carbon was calculated by the and kneaded.This mixture was then dried at 1108C to BET method using the adsorption isotherm of N .The 2prepare the impregnated sample.The impregnation ratio micropore volume was calculated from the amount ad-was estimated from the following equation:sorbed at relative pressure of 0.1in the N isotherm,and 2Impregnation ratio 5

the mesopore volume was calculated by subtracting the amount adsorbed at relative pressure of 0.1from that at (weight of sample after impregnation)2(weight of lignin)]]]]]]]]]]]]]].

0.95[17].

(weight of lignin) 2.3.Thermogravimetric analysis

In this study the impregnation ratios were 1.0for all samples.

The weight loss behavior of the lignin impregnated with The impregnated sample was set on a ceramic boat chemical during the carbonization was measured using a which was then inserted in a stainless tube (i.d.535mm).thermobalance (TGA-50,Shimadzu).Samples were heated The impregnated sample was heated up to the carboniza-to 10008C at a heating rate of 108C/min under N ?ow.tion temperature under N ?ow at the rate of 108C/min 22In order to compare the prepared activated carbon with and was held for 1h at the carbonization temperature.The commercial products,the surface areas and the pore carbonization temperature was varied over the temperature volumes of two commercial activated carbons prepared range of 500to 9008C.After carbonization,the sample was from coconut shell and coal were measured.Results are cooled down under N ?ow,and the carbonized sample 2summarized in Table 2.

was washed sequentially several times with hot water,and ?nally with cold distilled water to remove residual chemi-cal.The washed sample was dried at 1108C to prepare the activated carbon.

3.Results and discussion

The yield was calculated based on the weight of lignin and the chemical recovery was estimated using the follow- 3.1.The in?uence of chemical reagent and carbonization ing equation:

temperature on the surface area

The surface areas of activated carbons prepared at different temperatures using various activating reagents are Table 1

shown in Fig.1.The lignin chars prepared at the carboni-Elemental analysis of lignin zation temperature of 500to 9008C without chemical 2[wt.%(d.a.f)][wt.%(d.b.)]activation had surface areas of 10–50m /g,and are not shown in this ?gure.

C H N O(diff.)Ash Using ZnCl and H PO as the activating reagent,it is 23459.8

5.3

0.1

34.8

4.0

2observed that the surface areas were around 1000m /g

J.Hayashi et al./Carbon38(2000)1873–18781875

commercial activated carbons.Thus activated carbons with

very high speci?c surface areas can be prepared from

lignin by chemical activation with K CO.

23

The surface area is a maximum value at the temperature

of8008C in alkali metal compound activation,and at the

temperature of6008C with ZnCl and H PO activation.

234

These results indicate that at temperatures greater than

6008C,alkali metal compounds work effectively as activat-

ing reagents,but ZnCl and H PO do not.This suggests

234

that there is a difference in the activation mechanism

between the alkali metal compounds and the ZnCl and

2

H PO.

34

3.2.The in?uence of chemical reagent and carbonization

temperature on the pore volume

The in?uences of chemical reagent and carbonization

temperature on the pore volume are shown in Fig.2.This

?gure indicates that the micropores are well developed in

the activated carbons prepared in this study.

The pore volumes(micropore volume1mesopore vol-

ume)of the activated carbons prepared by ZnCl and

2

H PO activation increase with an increase in temperature

34

over the range500to6008C,and decrease with an increase

in temperature over the range of600to9008C.At Fig. 1.In?uences of carbonization temperature and activating temperatures above6008C,the carbon structure shrinks, reagent on the surface area of prepared activated carbon.and the surface areas decrease as shown in Fig.1and the

pore volumes decrease as shown in Fig.2.

The pore volumes of the activated carbons prepared by 2

and700m/g at5008C,respectively.The surface areas of alkali metal compound activation generally increases with

these activated carbons are much larger than that of the carbonization temperatures up to about8008C.This indi-

lignin char.This indicates that ZnCl and H PO work cates that the pores enlarge up to this temperature.Above

234

effectively as activating reagents below5008C.The surface8008C,the excess enlargement induces combination of

areas increase between the temperature of500and6008C pores,resulting in an increase in mesopores for all alkali

2

and the maximum surface areas of more than1000m/g metal salts,and a decrease of micropore volume,and of

are observed at6008C.From comparison with Table2,the surface area.

activated carbons prepared from lignin by chemical activa-

tion with ZnCl and H PO have surface areas as large as 3.3.The mechanism of chemical activation 234

those of the commercial activated carbons.

In contrast,using the alkali metal compounds as activat-The mechanism of chemical activation can best be

ing reagent the surface areas of activated carbons prepared illustrated from the properties of activated carbons pre-

at a temperature of5008C are very small compared with pared by K CO and ZnCl,which showed different

232

those of activated carbons prepared by ZnCl and H PO tendencies in pore structural change.

234

activation.This indicates that the activation mechanisms of The effect of activating chemical on the weight loss

alkali metal compounds are different from those of ZnCl behavior during the carbonization is shown in Fig.3a

2

and H PO below5008C.The surface areas increase which indicates the experimental and the calculated weight 34

between the temperatures of500and8008C,and then loss rates for the lignin impregnated with ZnCl,and in

2 decrease with a further increase of temperature from800to Fig.3b which shows the weight loss rates of the lignin and

9008C.This shows that alkali metal compounds work as ZnCl measured independently.The calculated rate in Fig.

2

activating reagents effectively above6008C.3a was obtained by assuming that the weight loss for the

The surface areas of the activated carbon prepared by lignin impregnated with ZnCl was a summation of the

2

alkali metal compound activation(except for Na CO)at values for lignin and ZnCl A difference between ex-

23 2.

2

8008C are well over1000m/g,and are larger than those perimental and calculated rates was found in the tempera-

of the commercial activated carbons.In particular,the ture range of200to6008C in Fig.3a.This difference

carbon prepared by K CO activation has a surface area of indicates that ZnCl as an activating reagent modi?es the 232

2

about2000m/g,being much larger than that of the carbonization behavior of lignin in this temperature range.

1876J.Hayashi et al./Carbon38(2000)1873–1878

Fig.2.In?uences of carbonization temperature and activating reagent on the pore volume of prepared activated carbon.

In the temperature range of200to3508C,the ex-Therefore the experimental rate was somewhat smaller

perimental rate was larger than the calculated rate.This than the calculated rate.

shows the evolution of the volatile matter due to the In the temperature range of450to6008C,the ex-

impregnation of the ZnCl.Ahmadpour et al.[18]carbon-perimental rate is very much smaller than the calculated 2

ized bituminous coal impregnated with ZnCl and found rate.It appears that the restriction of ZnCl evaporation

22

that hydrogen was evolved below2008C.When lignin contributes to this difference.Little difference between the

impregnated with ZnCl was carbonized,it was deduced experimental and the calculated rate is observed above the 2

that hydrogen was evolved as reported by Ahmadpour et temperature of6008C.ZnCl works effectively only below

2

al.[18].This evolution of hydrogen could contribute to the a temperature of6008C.

difference of weight loss rate.Fig.4a shows the experimental and calculated weight

In the temperature range of350to4508C,the ex-loss rates of the lignin impregnated with K CO and Fig.

23 perimental rate is somewhat smaller than the calculated4b shows the weight loss rates of the lignin and K CO

23 rate,and this shows that the evolution of the volatile matter measured independently.In Fig.4a,around a temperature

is restricted by the impregnation of ZnCl.The volatile of4008C,the experimental weight loss rate peak shifts to

2

matter evolved in this temperature range is mainly tar[19].the lower temperature range compared with the calculated

It is generally claimed that ZnCl works as dehydration one.Beguin et al.[21],Mochida et al.[22,23]and

2

reagent[19,20]and then it is deduced that the dehydration Yamashita et al.[24]found that dehydrogenation started at

induces the charring and aromatization of the carbon and a relatively low temperature with the addition of alkali

restricted the formation of tar in this temperature range.metal compounds.Similarly,in this study the dehydroge-

J.Hayashi et al./Carbon38(2000)1873–18781877

Fig.3.(a)Experimental and calculated weight loss rate behaviors Fig.4.Experimental and calculated weight loss rate behaviors of

of lignin–ZnCl and(b)weight loss rate behaviors of lignin and lignin–K CO and(b)weight loss rate behaviors of lignin and 223

ZnCl separately.K CO separately.

223

nation started at a relatively low temperature with the pregnated with ZnCl scarcely change in the temperature

2

addition of K CO and the weight loss rate peak shifted to range of600to9008C.In this temperature range the 23

the lower temperature range.As described above,the surface area and the pore volume of the activated carbon

addition of K CO modi?ed the carbonization behavior of prepared by ZnCl decrease as shown in Fig. 1.This 232

the lignin below5008C.However,the surface area of the decrease appears to be due to heat shrinking.

activated carbon prepared by K CO activation at5008C is The yields of the chars impregnated with K CO

2323 very small,as shown in Fig.1.It can be presumed that the decrease between the temperatures of7008C and9008C,

micro-or mesopore structure was scarcely modi?ed.while the yields of other chars scarcely change.It is clear

A large difference in rates is observed around the that the char is gasi?ed by K CO.McKee[25]studied the

23

temperature of8008C.It is clear that this increase is not

due to the independent evaporation of K CO,because

23

K CO by itself evaporates above9008C as shown in Fig.

23

4b.Therefore it is concluded that this increase of the rate is

due to the evaporation of K CO and/or the volatile

23

matter from lignin,which is induced by the reaction of the

carbonized lignin and K CO.The results show that

23

K CO modi?es the carbonization behavior of the lignin

23

as a chemical reagent in two temperature ranges,below

5008C and above6008C,while ZnCl modi?es the be-

2

havior below6008C.

Fig.5shows the yields of the prepared activated carbons

based on the weight of the lignin.It is observed that the

yields of the chars impregnated with ZnCl are larger than

2

those of the chars with no impregnation in the temperature

range of500to9008C.This fact supports the premise that Fig. 5.Yields of lignin,lignin impregnated with K CO and

23 ZnCl promotes the charring and aromatization of the lignin impregnated with ZnCl at different carbonization tempera-22

carbon.It is observed that the yields of the chars im-tures.

1878J .Hayashi et al ./Carbon 38(2000)1873–1878

K CO worked as the dehydrogenation reagent in

23activation below 5008C.It was reduced by carbon above 6008C,consequently carbon was removed as CO and the surface area and the pore volume increased.However,with excess activation pores were combined,and mesopore volume increased and the surface area decreased.

References

[1]Kumamoto S,Takahashi Y,Ishibashi K,Noda Y,Yamada K,

Chida T.Trans Mater Res Soc Jpn 1994;18A(Ecomaterials):647–50.

[2]Simitzis J,Sfyrakis J.J Appl Polym Sci 1994;54(13):2091–

Fig.6.K CO recovery ratios at different carbonization tempera-9.

23tures.

[3]Simitzis J,Sfyrakis J,Faliagas A.J Appl Polym Sci

1995;58(3):541–50.

[4]Gonzalez-Serrano E,Cordero T,Rodriguez-Mirasol J,Rod-gasi?cation of graphite by alkali metal and found that riguez JJ.Ind Eng Chem Res 1997;36(11):4832–8.K CO was reduced in inert atmosphere by carbon as 23[5]Ahmadpour A,Do DD.Carbon 1997;35(12):1723–32.

follows:

[6]Tsai WT,Chang CY,Lee SL.Carbon 1997;35(8):1198–200.[7]Sun J,Brady TA,Rood MJ,Lehmann CM,Rostam-Abadi

K CO 12C →2K 13CO

23M,Lizzio AA.Energy Fuels 1997;11(2):316–22.

[8]Bacaoui A,Yaacoubi A,Bennouna C,Dahbi A,Ayele J,

Fig.6shows the recovery ratios of K CO over the 23Mazet M.Aqua (Oxford)1998;47(2):68–75.

temperature range 500to 9008C.The recovery ratio [9]Xia J,Noda K,Kagawa S,Wakao N.J Chem Eng Jpn

decreases with an increase of the temperature.This result 1998;31(6):987–90.

supports previous ?ndings that K CO is reduced by 23[10]Tsai WT,Chang CY,Lee SL.Bioresour Technol

carbon in inert atmospheres.

1998;64(3):211–7.

[11]Marsh H,Yan DS,Orady TM,Wennerberg A.Carbon

1984;22(6):603–11.

4.Conclusion

[12]Caturla F,Molina-Sabio M,Rodriguez-Reinoso F.Carbon

1991;29(7):999–1007.

[13]Hu Z,Srinivasan MP.Microporous Mesoporous Mater

Activated carbons have been prepared from lignin by 1999;27(1):11–8.

chemical activation.The maximum surface areas of the [14]Qiao W,Ling L,Zha Q,Liu L.J Mater Sci

activated carbons prepared by ZnCl activation and by 21997;32(16):4447–53.

H PO activation observed at 6008C,were as large as 34[15]Teng H,Yeh T.Ind Eng Chem Res 1998;37(1):58–65.values for commercial activated carbons.The micropores [16]Illan-Gomez MJ,Garcia-Garcia A,Salinas-Martinez LC,

were well developed.In contrast,the maximum surface Linares-Solano A.Energy Fuels 1996;10(5):1108–14.

areas of activated carbons prepared with alkali metal [17]Rodriguez-Reinoso F,Lopez-Gonzalez JD,Berenguer C.

compounds occurred at 8008C,and were generally larger Carbon 1982;20(6):513–8.

than the surface areas of the commercial activated carbons.[18]Ahmadpour A,Do DD.Carbon 1996;34(4):471–9.[19]Rodriguez-Reinoso F,Molina-Sabio M.Carbon

With K CO activation,the surface area reached about 2321992;30(7):1111–8.

2000m /g.The mesopore volumes were seen to increase [20]Balci S,Dogu T,Yucel H.J Chem Tech Biotechnol

with carbonization temperature.

1994;60(4):419–26.

ZnCl worked as dehydration reagent and restricted the 2[21]Beguin F,Settan R.Carbon 1972;10(5):539–51.

formation of tar and promoted the charring of the carbon [22]Mochida I,Nakamura E,Maeda K,Takeshita K.Carbon

below the carbonization temperature of 6008C.Conse-1975;13(6):489–93.

quently the carbon structure was modi?ed and the surface [23]Mochida I,Nakamura E,Maeda K,Takeshita K.Carbon

area increased.However,above 6008C ZnCl does not 1976;14(2):123–9.

2work as an activation reagent,and the surface area and the [24]Yamashita Y,Ouchi K.Carbon 1982;20(1):41–5.pore volume decreased because of heat shrinkage.

[25]McKee DW.Fuel 1983;62(2):170–5.