Effect of pyrolysis temperature and hold time on the characteristic parameters of adsorbent derived

Journal of Encironmersal Sciences Vol .16.No.4.p.683- 686 .2004Artcle ID: 10.0742(200)04.06830CLC mumber: X703; 0647.3Document codle: AEffect of pyrolysis temperature and hold time on the characteristic parametersof adsorbent derived from sewage sludgeZHAI Yun-bo，WEI Xian- xun, ZENG Guang- ming(Departnent of Environmental Science and Engineering. Hunan University, Changsha 410082. China. E-mail: ybzhai @ 163 . com)Abstract: According to the Doehlert' s matrix method, the adsorbent derived from sewage sludge was preparedthrough chemical activation under controlling the pyrolysis temperature and hold time. The characteristic parametersincluding the total yield, adsorption of methylene blue, adsorption of iodine, BET surface area, micro- pore volumeare 35%- -49%， 16.5- -38 mg/g, 285- -362 mg/g, 185- -359 m'1g, and0.112- -0.224 m /g, respectively. Accordingto the experimental data， the multi-linear regression method was adopted to fit the relations between thecharacteristic parameters and influential factors. At final, through optimization method, the optimal adsorbent isobtained when using 62 min as hold time and 1105K as pyrolysis temperature. Under the conditions, the adsorbentwas produced and compared the characteristic parameters with model forecast value, the coherence is satisfied.Keywords: adsorbent; sewage sludge; pyrolysis temperature; hold time; Doehlert' s matrixthe pyrolysis temperature is reached. It is another importantIntroductionparameter affecting the characteristic of the adsorbent derivedSewage sludge amount increased due to the rapidurbanization in recent decades, at the same time the effluenttemperature and hold time varied from 973K to 1173K andcriteria is stringent, so the treatment of the sewage sludge isfrom 40 to 80 min, respectively .attention widely ( Chen，2000). The agricultural use of1.3 Analysis method of the characteristic parameter ofsewage sludge, land flling and incineration, which are theadsorbentmost common methods to got rid of sewage sludge( Inguanzo,The totaul yield ( M.) was analysed through weight2002). In recent years, the sewage sludge was pyrolysis tomethod. The adsorption ( M2 ) of methylene blue wasproduce the adsorbent which removal the pollutions in thedetemmined spectrophotometrically at the maximum absorbancewastewater and gas is attention ( Naozumi，2002; Andrey ，wavelength(680 nm). The adsorption( M;) of iodine wa2002). The preparation of adsorbent derived from sewagedetermined using the sodium thiosulfate volumetric method .sludge is influenced by many factors. The most influentialThe adsorption capacity was calculated by applied thefactors were found to be hold time ( U ) and pyrolysisLangmuir equation ( Stumm，1981). The BET surface areatemperature(U2). In the present work, a series of adsorbent(M.) was obtained using N2 adsorption isotherm at 77Ksing BET equation. The micro-pore volume ( Ms ) was .characteristic parameter of adsorbent derived from sewagedetermined through Quantachrome Autosorb- 1 analyzer .sludge was analysed as a function of two factors. A Doehlert1.4 Methodology of experimental designmatrix was used to represent the responses of two factors in allUsing the response surface methodology ( Rivera- Urila,experimental regions studied.1991; Box, 1978; Wachter, 1999; Bacaoui, 2001), it is .1 Experimentalpossible to systematically study many processes. This is asingle- factor- at- a time method that studied the phenomenon.1 Raw materialsby varying one factor while keeping all other conditionsThe experimental sewage sludge was obtained from theconstant. However, the effect of each factor is not necessarilywastewater plant, which adopted the activaled sludge methodaddition. It is, therefore, necessary to take into account theto treatment wastewater. Thesewage sludge containsinfluence of each factor and the interaction between theseapproximately 35 % inorganic matter mainly metal and theirfactors that may be synergistic or antagonistic( Box，1978) .oxidations that are considered as catalysts in the process ofAs indicated in the introduction， the most infuentialcertain pollution gas removal, and 60% organic matter, andfactors on the characteristics of adsorbent derived from sewage5% water, and with a particle size between 0.2 and 2.0sludge are hold time( U ) and prolysis temperature(U2). Amm. .Doehlert matrix was used to represent the responses studied in1.2 Pyrolysis and hold timeall experimental regions of these two factors. The factors arePyrolysis was carried out in an isothermal reactiongiven in the form of coded variables( X;) with no units inheated by a vertical fumace, high-purity nitrogen was used asorder to permit comparison of factors of different natures. Ththe punging gas flowing through the sample bed at certainlos ( U|) into codedflux. Then the reactor by then heated to the desired中国煤化工deon the basis of thetemperature at certain increasing temperature ratio. The holdcofollltime is the residence time of the sample in the fumace afterYHCNMHGFoundation item: The National HighTech R&D Progam(863) of China( No.20AA64020) , the Natioal Foundation of China for Oustanding Young( No.50225926),and the Science Foundaion of Edueu tion Minietry of Chin for Ontstanding Young Teacere(2000)684ZHAI Yun-bo et al.Vol. 16U,-Ui2 Results and discussionwhere X; is the value of coded variable j in experiment i;2.1 Response analysis and interpretationUj is the value of natural variable j in experiment i; U, is2.1.1 Adsorbent total yield( M )According to Table 2，the muliple linear regressionthe value of natural variable j in the center of the domain ofinterest; it corresponds to X =0; OU; is the variation of themethod was adopted to fit the relation between the hold time，pyrolysis temperature and the total yield ( M;) Hy usingnatural variable j corresponds to a variation of the codedMatlab software, and then excluded the non- marked items,variable j equal to + 1.Doehlert' 8 experimental matrix and the correspondingthe total yield ( M) of. the adsorbent derived from sewageexperimental conditions are given in Table 1. Thissludge model was obtained as follows:combination is chosen in such a way that the factors are notM, = 44.5-4.2X - 6.05X2 -4.8X + 2.15X, Xx2.(3)correlated with each other and that modeling can be done withBetween the theoretical and experimental response, theuniform accuracy over the whole parameter space. Thiscoefficient of correlation( R?) is 0.923. This showed that theanalysis was achieved by response surface methodology thatgood coherence existed between the experimental value andgave satisfactory results. Each response are described bythe model forecast value.second order model adequate for predicting the responses inAccording to the nodel to draw three dimension responseall experimental regions :suface as follow Fig. 1. Analysis of the results shows， asM=bo+b.X+b2X2+bX+bX+bp2X.Xx.expected, that the pyrolysis temperature(X2) has a highly(2)negative influence(b2 = -6.05) on yield( M, ) comparedWhere X, is the coded variable related to the natural variablewith other factors. The hold time(X) also has a negativei ( activation time ); Xz is related to T ( activationinfluence(b = - 4.2). The interaction influence(b12 =temperature); bo is the intercept tem， a constant that2.15) is positive and nol important than the hold time andpyrolysis temperature. A maximum yield is obtained at lowcorresponds to the response when.X, is zero for each factor;bi determines the influence of activation time; br is thehold time and activation temperature. According to Fig. 1，interaction efect between the activation time and activationwith the pyrolysis temperature increasing, the adsorbent yieldtemperature; br and bn can be regarded as a curve shapedecreased. This is as expected because volatilizationincreasesrith pyrolysis temperature. The yield of theparameter.adsorbent derived sewage sludge resulted from pyrolysis atTabhe Debient's epertmeatal matxs Aand the corepodn eperinental cndfitios850C was found to be 35% - -37 % of the original weight ofNo. ex. Time( XTeperanure(X )Tme(U). min Tmprnre(U2), Kthe raw sewage sludge. And the yield of the adsorbentdecreased faster at the high temperatures， this trend is0different from the carbonization of coals. This esentially).s0.866 .70173indicated that there are two volatilization stages : primary and0.5-0.866msecondary volatilization during the pyrolysis of the sewagesludge. The result is similar to the many literatures studied0.866results. And for all samples during experimental time, the5073effect of the hold time on the yields was no significant thanFor each response( M), the coffcients(b, by) of thepyrolysis temperature .postulated model were calculated on the basis of theexperimental responses by lease squares regression and fttedmodel coefficient, and then venfied the marked regression55cofficient, and the non-marked items were excluded andobtained the simplified model. The detailed calculation' sprocesses refer to literatures ( Bacaoui, 2001; Myers, 1995 ;Wu, 2001; Doehlert, 1970; Zhu， 1993). The Doehlertmatrix consists of N experiments with N=K°+K+1,40|where K is the number of variables studied. In this case K ;3s{70.2 and, therefore, the matrix comprised of seven experiments3ok(Table 2)..5~No.exp. Canditions M. % Mz,mg/g_ M,mog/g_ M.m'/g M, .cm'/g80/1073 4031362470.14940/1073916.5302850.125中国煤化工，T70/1173389590.15350/9732852150.112MCNMH(blue( M)70/97322.3330.132uL1Is wuik, 111 urucn w ucscribe the characteristics of50/117329.434100.150he adsorbent derived from sewage sludge in the organic601073 .31986.0.22compound adsorption, because the methylene blue meloculeNo.4Efect of pyrolysis temperature and hold time on the characleristie parameters of adsorbent derived from sewage sludge685diameter is very ft to response the mescropore characteristicsopened with the increased pyrolysis temperature and holdof the adsorbent， the methylene blue was chosed astime, the reasons are that the volatile matter is liberale duringadsorbate. According to Table 2，through regress thpyrolysis. And without any decrease in the adsorption ofexperimental data, the model is as follows :iodine in the domain studied. But, the final temperature isM2 = 33.75 + 2.373X + 7.543X, - 1.343X}raised, a shrinking of the paricle and a narrowing on the+ 0.791X号+ 1.2329X Xz.(4)pore entrances occur.The eoffcient of correlation( R2 ) obtainedin this caseis 0.991. The efects of the pyrolysis temperature(b2 =7.543) and the hold time(b, = 2.373) are more important380han the interaction ( b2 = 1.233). The increase in370temperature entails an opening and enlargement of the pores,360which enhances the adsorption of methylene blue ( M2;350Fig.2). Indeed the methylene blue molecule has a minimum340{molecular cross-section of about 0.8 nm, and it has been330estimated that the minimum pore diameter it can enter is 1.33200.nm. Therefore, it can only enter the large micropore, but31okmost of it is likely to be adsorbed in mesopores .-0.21σ-0.845Fig.3 Veriation of iodine adorption capeciy(M,) versust. Ts 352.1.4 BET surface area( M.)According to Table 2, this regression model is described0by the fllwing equation:M。= 325 + 34.5X + 49X2 - 13.6X}20<+ 8.12X2 - 15.6X; X2.(6).80.4The cofficient of correlation( R2) is 0.923. Analysis ofthis response shows that the direct efects of hold time(b, =34.5) and pyrolysis temperature( b2 = 49.0) are positive and1.0-0.8more important than their interaction(br = - 15.6). TheFig.2 Varation of methyene blue adorapio( M) versus, rapparent BET surface area( Ms) increased with inereasing ofX and X;(Fig.4). Fig.4 shows the BET surface area versusTherefore, it can only enter the largest micro-pore, butpyrolysis temperature and hold time relationship. It is notedmost of it is likely to be adsorbed in mesopores. Thus, thethat the highest surface area obtained is from the pyrolysismethylene blue testing can be used to predict organictemperature at 850C and hold time 80 min. The generalmesopores and macropore volumes. In general, increasingtrend is that BET surface area increases with pyrolysispyrolysis temperature led to improve pore volume， but whentemperature，which does not correspond to the adsorbent yieldthe temperature reached to 850，the adsorbent did notrend. However， this increased trend will weaken. Twoimprovement adsorption capacity obviously. Theffects can interpret phenomena: the first efet is that newmethylene blue adsorption increased with the hold time，whenpores developed when volatile matter is liberated duringthe hold time reached to 60 min, the effect is not obvious .pyrolysis, when the certain temperature was achieved, most2.1.3 Adsorption of iodine( M,)organic solids began to lose hydrogen and produce free-The iodine molecule, compared with methylene blue, isradicals, which are then condensed into larger moleculargreatly adsorbed due to itse smaler size permitting itunits and lead to a microporous texture , hence the accessiblepenetration into micro-pores. According to Table 2，thmicroporisity and BET surface area increase. Second, as theregression model is 88 follows:heat trealment temperature is raised， a shrinking of theM; = 332 + 16.9X, + 11.3X2 + 3.6X} +particle and a narrowing of the pore entrances occur or some8.92X + 8.77XX2.(5)of pores can be sealed off， resulting in a decrease of theThe cofficient of correlation( R2 ) between the maximumacessible BET surface area. About the efeet of the holdcapacities of adsorption calculated by the model and thosetime on the BET surface area, generally, the BET surfacedetermined experimentally in this case is 0.939. As in the .increases with hold time, about 65- 75 min the increasingevels nff I Grnm_the mndel. the nnclusion was draw that thecase of response Mz，it is clear that the pyrolysis temperature中国煤化工surface area were less(b2=11.3) and hold time( b, = 16.9) have a strong impacton development of the porous texture, the interaction(b12 =2.1IYHCNMHGemperature .8.77) efet on the adsorption of iodine is not important thanFor this response, according to Table 2, the regressionsingle factor ( pyrolysis temperature and hold time ) .model is as follows:According to Fig.3, it seems that more micropores are being686ZHAI Yun-bo e al .Vol. 16through using the lease squares method. The optimal pointindicated by the model corresponds to a hold time of 62 minand pyrolysis temperature of 1105K. And under the400380optimization conditions， the adsorbent was prepared， and3601utilizing the previous models to calculate the characteristicparameters，then comparing the experimental values anc34calculated values, it is important to note the good agreement300found between the experimental values and calculated from280the models .260170.84013 Conclusions220-1.0The response surface methodology using Doehlert ’s-0.matrix is an appropriate tool to study optimization of the0.51.0-0.8activation process to prepare adsorbent to be used in a giventechnological process. In general, among the characteristicig.4 Variation of BET surface area( M.) verus I. Tparameters of the adsorbent derived from sewage sludge，theyield decreased with pyrolysis temperature and hold timeMs = 0.137 + 0.036X + 0.052X; - 0.021 xincreasing, the other parameters including methylene blue- 0.049X} - 0.085X X2.adsorption, iodine adsorption, BET surface area and micro-The coefficient of correlation(R?) is 0.923. As in thepore volume are increasing with the pyrolysis temperature andcase of response Ms, it is clear that the pyrolysis temperaturehold time, but the increasing rate is obvious diferent.(b2= + 0.052) and holdtime(b,= + 0.036) have a strongAccording to the experimental data， adopted the multiplepositive effect . The Fig.5 shows that the relationship betweenlinear regression to simulate the correlation between ththe micro-pore volume and different hold times and pyrolysischaracteristic parameters and pyrolysis temperature and holdtemperature. It clearly shows that for all hold time， thetime，the obtained equation is good coherence to experimentalmicropore volume has developed with the pyrolysisdata.temperature increased. This means that the micro-poreReferences :structure evolved in this temperature stage. But with theAndrey B, Tereal B, 2002. H2S adsorption on malerials obrained using sulfuriepyrolysis temperature increasing, the rate of the micro- poreacid activation of sewage sludge-derived ferilizer[J]. Jourmal of Collid andvolume increasing will weaken. The reason is that theInterface Science. 252: 188 - 194.pyrolysis temperature increased, the larger sheets can burringBacaoui A, Yaacoubi A，Dahbi A, 2001. Opimization of conditions for thepreperation of activated carbons from odive-wasle cakes[J]. Carbon, 39:out as the walls between the adjacent pores converting425-432.probably more and more of the micro-pores into macropores.3ox G, Hunter w G, Huner J s. 1978. Stalics for experimenters, anBut in the experimental pyrolysis temperature， thisintroduction 10 desigo, data analysis and model building[ M]. New York :phenomenon can not been observed .Viley. 306- * 604_ben X G, Jeyaseelaun S, Grnham N, 2002. Phyical and chemical propertiesstudy of the actireated earbon made from sewage sludge [J]，WasteManggenent, 22: 755- -760Doehlert D H, 1970. Uniform shell deigslJ]. Applied Saistie, 19: 231-0.182390.16Inguanao M. Dnningguvez A, Blanco C G, 2002. 0n the prolyis of sewage0.1410.12|0.10{Myers R H,0.08product optimisation using designed experiments[ M]. New York: John Wiley0.061.0.041Naozuni K, Aki M, Yashinoni 1, 2002. Adorpion removal of putios by.021/70.active cokes produced from sludge in the energy reeycle procese of wastee[J ]; /-0.4of activated earbons prepared from agriculural raw materials and Spanishlignites when Temoving chorophenols from aqueous solutions[J]. Catbon,0.3.0-0.829: 613- -619.Stumm w，Morgan J J, 1981. Aquatic chemisty[M]. 2nd ed. New York:Fig.5 Variation of miroporty volume(Ms) versus 品TWiley, Inter-Science.Wecher R, Cordery A. 1999 Response surface methodoloay modeling ofdiamond-like curbon flm depositin[J]. Carbon, 37: 1529-- 1537.2.2 OptimizationWuDF, u Y D, ZhangJP et al., 2001. Efets of the number of tesingTo optimize all responses under the same conditions isspeciments and the estimation methods on the weibul paraneters of soliddificult because the interest region of factors are diferent,56(24): 7035- 7044.thus, when certain response reached the optimization, all the中国煤化工ie pepetn of he piainothers responses decreased probably; therefore, in order toYHCNMHG8 Sciene & Technology Press.find a compromise，the“function of desirability”is resorted( Received for review July 16, 2003. Acepted Septenmber 24, 2003)