1.4.?! 231?: , ‘- 5‘3,- .31??? rgt z;.9r!;fi:.;!. . i .. :in 21$” . (2.71:1 . . V ‘VAL; 515.191.711 ‘VIrllnAZaV, ...... hijrvuhslwltélitdfi'g‘ 3.. E '9' I . 5; u sifigvthuvvafia‘ It"... a - . .. 4 . . u r. éOnWizoltl Lair-«K .1. . i t III). a! fléitllyvuz . ‘ 9.. .5; 1:92.10! 53!. r . v. A r x... 1 .. Litt.$7‘..vir. )vrrvflyuvfdfir» :Jflwxufltlfiuwflnlt: .25 . u v 1 \.. tricnn. Iv . , o. a: rye. . ‘ 0.07 ‘hk‘VIi'W garih‘nlh fibfinihuvflr» cmfi,“vahvo.fl.akflwflrflprlvtdvhufi i . . f! . tit-4; I at. Q . Ila Ii for! vl‘Itbt. . x. "O and.“ bit!- ¢9070 . :1! it. .01! ‘1'»! u . r. 53:: r. e- 3!! 3. ‘ 5:992... {fir—.70.: \ N at}. r. v I?! .69 2., 0' ,v .0, Mi. .3 JP. 7. L. 1... 9' ti!!! ‘ . 1 3.0.5.: “‘0’.- _ of??? t... . 1'9 . ii‘fitifléé :E‘ 1. . 1 P. 223... $2; 2.36%.... : . t . . AZ)..." vlfi . . . . . l y .tpthlP: . stiflh . . ox - ALL)... I u ‘ I'M-Ski . .71‘.“ {I Y.- #5‘fii.tc’vac V . fftvvviulil _ . . x . . v. £131“... 9 q. 1a») 6". , 1 . flamingg . 92'. 3!! l 0 . Joel??? . A irrstvhnflfdtlll'A-‘MP‘anQ 1.x. girl-Irv... ‘ .‘ m - I ‘ . . . it £3.1E.§2.Pr(en. , “fl , t... 101E . in“? Viki}. iéfiMiWQ vii. «it... . v. 9.37%.. fibihfitt - gig?! ‘ . r I t.‘ 3.1%. L ‘54) VI}! ’(‘V0 .17... 4 g: , . urn: :r‘ i‘f‘ ‘»II‘ I n X... rstfitv‘tfv‘? ‘12:...” $1.. .unanW‘ihzxfiv nif: . m ‘ .7: rfinnyvilflnrg.r...flh§ (I .T. 5509‘ . i.....lll~.:... 3‘ ,2 .. ‘ r .X. ‘ .bx {$.Err’tvvil A. «zvfliili‘giix . .. vi“?! ‘fltE-f . if It; _ . a . . I , . L , K. ,, ,. ‘ fi ,. ,. c: v . ‘x.,. . ,. a. . .23: ..:.fi..,.. .....t.....V..‘. .5. T7». ,1 .....v.,..cd.~..w ‘ ,1." w'l’l1lv’K:levuu..l£u n.. l y . I‘I .... .. u. 1.. .59.”...1: $1.435“. 9113 qufilinaxv : ,.. 3. z T .V a.“ 3.1.22.2.3. .. ‘.n_it..., ... r." , .. .A o. I f ‘ . ’ y 2.. . . . , v . y ‘ u. ... .. , ., a L f?» l u: , , . . y . c... b y t . . .5. y V. v I '0. 9: ~ .. , n‘ , ,.. . ‘ . c ‘ Egtium.3m4§. may vunuhfiflzdz .. Qt ‘ 71! . .g-IJI‘I! 3 .1... Bart.» V... .- \£Kflflm2 .5 .m.«fo$.v.§.. . _ . a. r... infirlg’rii w: ‘ E: ‘ , . . .u. “Rafi 15.31.3134“. ‘ a . sié ‘ . . t .pnnnvnbflh P. VXLWMV¢ - .. u ‘ .. ‘3 I. eéuflunikvi .6‘...‘ “Kittie, .h‘arfr. :0va . p - . ”Jag Jwgflioa u~|¢h v9 5 A s .. 2 A ... I. a :66g1atthh fig... 1.; 1 a ‘ .5161...- Kfifii .if.§.+.wci‘l....x I . 2“”.‘vrgifnifi , , a O R Piié‘flu’ffll.’ . .r. r , , v2 ‘ V . s. 1. W1... 5' If : ..' \-’ . E ‘3 h g a {$056+- 1: ‘ , ¢ mun“... , éfibfli; f. if . . v . . Eyznihfrfhfi. z: I r 55.14.13» \ pun . A Pi. ‘ fs‘ ‘ EL «Ll-5;: 1» Kqflmphyflfi.§ .Wmvfi 1Y3}... Vt . ’5‘ m i L, .. , 1.9+)! ii; 533.354.? ,1?» “it L2“ "2 ‘ . . i .éf‘ . , fir .93.. 3.. afiv ‘90!qu glitz. , iii‘g. 24L. L51. at. .701. It}?! £51}? it‘s. .9: beélffilt molt: txtuuislbtvflefliti rifting e .r. ‘ E... ‘ ‘ . 4 . '11., A 5 0.»! 5.0.95. {71. . , ltffi-illcl‘l . . L ¥4§§$§.lavust?~ ., . . ton-Fig filit$#qif fiat: Lea-9”. \k’rt.;t..xt‘.l . so)... \. . A a}?! . ‘ $.13. ,. gig “mm? a 1.. T. a?! " .2 r ,‘ . , 'r m 1" a. -, -'.. « i a." ‘. ‘ 9 J V}... r 1 9 l, 3:..F 1. ‘3 I . i .3 3 b H a a H a z}- .- ‘ l" 1 t o" “Mat" 5_‘.;.::_':'}. ‘ ’2’"; (/‘réf’ ’ P A g 'u d_ ",i" l,- _‘ .‘ j J ,\’.\‘g ‘. 3‘34 u. .n An “2-) «A; .4 a «3.9 'h a; saw a;— 2' "A van-O" g a. )4; r- v 9’3 _ ' $1323 9 ca! J" 5 J This is to certify that the thesis entitled 711E Mm) PRO LYSI’S‘ OF CELLULZD'E AM) THE :5 75667 0f" K160; presented by K Fm / Nflek m0 has been accepted towards fulfillment of the requirements for \ . _ a, - /, ‘ MI 5 degree in (ééfip C 151 C 1? “had )7/2/ / z 4 Major prof or Date § A} //‘79 ( 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution MSU LIBRARIES RETURNING MATERIALS: —-_-_—_a_ Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. THE RAPID PYROLYSIS OF CELLULOSE AND THE EFFECTS OF KZCO3 By Kim, Nak Won A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemical Engineering 1986 ABSTRACT THE RAPID PYROLYSIS OF CELLULOSE AND THE EFFECTS OF K2CO3 By Kim, Nak Won 2C03 has been performed in a heated screen reactor to find the The rapid cellulose pyrolysis with and without K effects of temperature, pressure, and holding time. The electrodes and the shape of screen were modified from earlier experiments to heat a sample rapidly and uniformly, and aluminum foil was used to collect condensed tar. The tar yield (43 wt %) from pure cellulose pyrolysis at 6500C at 15 mm Hg total pressure is much larger than the yield (31 wt %) at 7500C. The product yields were independent of pressure in the range 10-760 mm Hg. Rapid pyrolysis with K200 as well as slow pyrolysis 3 with other additives resulted in drastically increased char and decreased tar yields. The tar obtained from cellulose pyrolysis was analyzed by G.C. and H.P.L.C. and found to contain levoglucosan, D-glucose, and unknown components. The unknown component at retention time 20 minutes in H.P.L.C. is increased with increasing weight of K2003. To my parents and wife ii ACKNOWLEDEGMENTS I would like to acknowledge the continuous and indispensable education support provided by Dr. Dennis J. Miller for the completion of this degree. TABLE OF CONTENTS Page LIST OF TABLES --------------------------------------- vi LIST OF FIGURES -------------------------------------- vii CHAPTER I. INTRODUCTION ---------------------------------- 1 A. The Need for Energy from Biomass ---------- 1 B. Pyrolysis from Wood ----------------------- 3 C. Requirements of Rapid Cellulose Pyrolysis - 6 D. Effects of Impurities in Pyrolysis -------- 8 E. Research Objective ------------------------ 10 II. EXPERIMENTAL APPARATUS ------------------------ 12 A. Pyrolysis Reactor ------------------------- 12 B. Electrical System ------------------------- 15 C. Gas Collection System --------------------- 21 III. EXPERIMENTAL TECHNIQUE ------------------------ 35 A. Cellulose Sample and Foil Preparation ----- 35 B. Wire Screen Preparation ------------------- 36 C. Sample Loading ---------------------------- 37 D. Preparation for Gas Collection and Analysis 37 E. Gas Calibration --------------------------- 37 F. Flash Pyrolysis --------------------------- 40 G. Collection of Pyrolysis Gas --------------- 41 iv CHAPTER Page H. Collection of Tar __________________________ 43 I. Analysis of Tar ____________________________ Ah J. Collection of Char _________________________ an IV. SAMPLE CALCULATION AND EXPERIMENTAL RESULTS —--- 46 A. Calculation of Weight of Calibration Gases - 46 B. Calculation of Weight of Pyrolysis Gases -—- 47 C. Calculation of Product Composition --------- A7 D. Tabulated Data and Calculated Results ------ A9 V. DISCUSSION AND CONCLUSIONS --------------------- 65 A. Temperature and Pressure Effects on Pure Cellulose Pyrolysis ------------------- 65 B Effects of K2003 --------------------------- 67 C. Analysis of Tar ———————————————————————————— 69 D Conclusions ———————————————————————————————— 74 E. Recommendation _____________________________ 75 REFERENCES ............................................ 76 APPENDIX A ____________________________________________ 78 APPENDIX B ____________________________________________ 86 APPENDIX C ____________________________________________ 89 TABLE Number Number Weight Weight Weight Number Weight Weight LIST OF TABLES of_Counts for Calibration Gases ------ of Counts of Pyrolysis.Gases --------- (CO, CHu, 002, H2) Of Pyrolysis Gases ................... Percent of Pyrolysis Gases ----------- and Weight Percent of Char and Tar --- of Counts of Pyrolysis Gases --------- and Weight Percent of CZHLL and H20 --- Percent of Pyrolysis Products -------- Average Weight Percent of Pyrolysis Products at Various Conditions -------- vi 50 52 54 56 58 6O 62 64 FIGURE H n ,_. O\O (EV O\UX 4Tb.) N H |-‘ . 13. 14. 15. LIST OF FIGURES Structure of 3 Components in Wood ——————————— Cellulose Degradation _______________________ Pyrolysis Reactor ___________________________ Electrical System of Pyrolysis Reactor —————— Electrical System of Gas Analysis ----------- Gas Collection System _______________________ Sorptometer _________________________________ Tar Analysis (0 wt % KZCOB) _________________ ) _________________ 3> ———————————————— Relation between Power and Temperature Tar Analysis (5 wt % KZCO3 Tar Analysis (10 wt % K200 ll—a, b, c, d, e, f ------------------------- Chromatogram of Water Calibration [a, b] ———- Chromatogram of Pyrolysis Gases _____________ 0, CH4, C02, H2, CZHN) Chromatogram of Pyrolysis Gas (H2) ---------- Chromatogram of Pyrolysis Gas (H20) --------- Page 2 A 14 17 2O 24 34 7O 71 72 83 87 90 91 92 CHAPTER I INTRODUCTION A. The Need for Energy from Biomass The increasing scarcity of and the rising costs for developing petroleum and natural gas oil stocks have stimulated interests in future sources of these commo— dities. Changeover to other sources of fuels and chemi- cals, including coal, oil shale, peat, and tar sands, are possible alternative fossil resources for the near future. The development of these resources remains an environ— mental and technical issue and fianl depletion is ultimate— ly inevitable. However, biomass, another alternative source of energy, is renewable and offers minimal environ- mental effects because of its low sulfur and nitrogen contents: many types also have very low ash. Therefore, biomass can provide a fraction of total energy requirements or nearly all chemical feedstock needs. A strict defini- tion of the term "biomass" refers to material produced by plants grown on land or in water [1]. In a broader sense, however, biomass is generally defined as the results of direct photosynthesis (terrestrial and aquatic plants) as well as indirect photosynthesis [2]. This definition enables us to include the recurring byproducts of life processes (animal residues) and civilization (municipal solid wastes, sewage, and industrial wastes) [2]. For a country without resources of fossil fuels, these alterna- tive and renewable biomass material can provide feedstocks = Cellulose = = Lignin = O l R, R' = H, OCH3 CH ““ CHCHZOH R" = H or Ar or OR" Carbohydrate = Hemicellulose = ( CsHsou )n ( C6H1005 )2 Figure 1. Structure of 3 Components in Wood LJIIEEEESEB for chemical and energy production. An example of such a country which could utilize biomass as fuel is Brazil. But the availability of biomass resources is obviously a function of its management. This management depends on a variety of facts including the cost of fossil—based fuels, chemical and economic climate, and available technology. Unfortunately, its effects are still very uncertain world- wide despite the development of biomass for chemicals, fuels, and energy. Anyway, it is obvious that biomass utilization for energy is likely in the future. So, as a contribution to biomass technology, research on cellulose pyrolysis has been initiated on the basis of much information related to coal gasification, wood combustion, and fire—proof fibres. Our research in this field emphasizes the effects of process conditions and impurities for rapid, high tempera— ture cellulose pyrolysis. B. Pyrolysis from Wood Wood consists of three kinds of components: cellulose, lignin, hemicellulose. It is known that cellulose composes about one-half of wood, and lignin and hemicellulose make up the rest, depending on the tree species. As shown in Figure 1, the structure of each component is quite different: lignin has a quite complex and irregular form. Therefore, in laboratory scale experi— ments, the consistent structure of model sample of ' L,,.2~g;: ~ COprUMMmoQ omoasaaoo .N ohswflm sopflm use made hhmcsooom Ll CoapmN COHpmNHMoezHom uflpmshaom pogo mezzomEoo UopMComzxo gospmNasofizaom Acwmoozamo>oqv I-li-illltlll'v pgwflog MdafiooHoEISOH . mane hamsflhm mam omhg omoasaamo mConmho . H mcflxowso mcflxomho mommw zhwsflsm momma zydecooom Al cellulose is recommended, and is commercially readily available. Pyrolysis of cellulose, generally in the absence of air in order to minimize the yield of water and CO2 from oxidation reactions, converts the solid polymeric composite into degradation products that can be in the solid state (char), the liquid state (condensible tar), and the gaseous state. The gas phase contains mostly carbon monoxide, carbon dioxide, and minor proportions of hydrogen, methane, ethene, and other light hydrocarbons. From a number of experiments, many researchers [3,4] have proposed the pathway of biomass pyrolysis, including cellulose pyrolysis, shown in Figure 2 as an overall simplified mechanism of the process. As shown in Figure 2, the pyrolytic process is composed of many parallel and consecutive reactions that yield different products. In this pathway, gases are produced directly or indirectly by depolymerization (cracking) reactions. Formation of levo- glucosan, which is believed to be a principal intermediate compound, takes place at somewhat high temperature (over 2800C) and leads to further decomposition reactions at elevated temperature. It is suggested that the main pyro- lysis products of levoglucosan from secondary reactions may be classified as fixed gases such as hydrogen, carbon monoxide, carbon dioxide, methane, acetylene, ethylene and propane, as well as semi-volatiles such as low molecular weight alcohols, ketones, aldehydes, and carbonyl compounds, aliphatic acids, hydrocarbons, and furan. Some investigators [6,10,11,13,14] reported the analysis of tar by various methods: Gas Chromatography (GC), Gas Chromatography and Mass Chromatography (GC / MS), Thin—Layer Chromatography (TLC), Liquid Chromatography (LC), and Infrared Spectroscopy. They confirmed that levoglucosan (1,6 anhydrojfi—D-glucopyranose) is the main component in tar and that the yield of tar (levoglucosan) depends on the employed conditions which will be described in the nest Section. C. Requirements of Rapid Cellulose Pyrolysis There are three major products (char, tar, gases) in rapid pyrolysis [4-9,16,17] as well as slow pyrolysis [10- 15]. The proportions of each of the three types of products are a function of the condition under which the pyrolysis is carried out. The most important factor influencing the pyrolysis yield is temperature. A lot of literature has shown the effects of peak temperature on the yields of char, tar, and gases. Hajaligol et al. [8] observed that low temperatures (<6OOOC) favor tar and char production, with the gas being dominated by water. Intermediate temperatures (70000) maxi- mize tar production, reduce char production, and augment gas evolution, primarily carbon monoxide. Higher tempera- tures (750 to 11000C) significantly increase gas formation via secondary cracking of the tar. Heating rate, which is another factor influencing the yields of tar, char, and gases for the pyrolysis, is directly connected with peak temperature. At the slow heating rate and high peak temperature, the pyrolysis will take place more slowly and result in secondary cracking at lower temperature. At the rapid heating rate and low peak temperature, the yield of tar will be dominant, as time is not allowed for secondary cracking to occur. Gases constitute the remaining products. The effect of sample holding time is also explained in various studies [3, 8]. It is generally agreed that hold— ing the sample at the final temperature in low temperature pyrolysis gives continuing decomposition, so that addition- al tar and gases are made and char yield is decreased. At high temperature, the effects of holding time are negligi— ble for the char yields, but the yield of tar is decreased and the yields of gases are increased, because the formed tar is believed to be decomposed by high temperatures of the screen. Only a few studies were found which addressed the effect of pressure on pyrolysis [8]. They showed that increasing inert gas pressure in pyrolysis gave decreased yield of tar and increased production of gases. From these results, it is understood that there is not much influence unless significant pressures are exerted during pyrolysis. Hajaligol et al. [8] observed the effect of sample dimension by using S N0. 507 Filter Paper (101 micrometers thickness) in the rapid pyrolysis. They showed that increasing the thickness of cellulose decreased the yield of tar and increased yields of the light gases such as hydrogen, methane, propylene, carbon monoxide, ethylene and carbon dioxide. The effects of inert gas [3] on the pyrolysis can be explained by comparing the effects of oxygen or air. Pyrolysis under oxygen or air is known to result in the fast depolymerization, large carboxyl compounds, carbon monoxide, carbon dioxide, and water. Thus, to minimize water and oxygenated compound production by oxidation reaction, usually oxygen or air is excluded from the pyrolysis. Another very important factor influencing the pro- portions of the products for rapid pyrolysis is impurites in the cellulose. This will be explained in Section D. D. Effect of Impurities in Cellulose Pyrolysis The pyrolysis of different cotton cellulose [10, 11] with and without flame retardants have been studied to develop better flame retardants. When any kind of biomass burns, the thermal degradation known as pyrolysis occurs and tar (mainly levoglucosan), water, carbon monoxide, carbon dioxde, char, and other inflammable volatile products such as hydrogen and methane are formed. Those volatiles react in the gas phase with atmospheric oxygen and result in a flame. Since it is difficult to prevent the combustion of these volatiles, many researchers have tried to find the better flame—retardants to prevent their formation. The major role of the flame retardants is to change the volatiles formed into the non—combustible materials and water vapor. Therefore, the flame retardant may be con- sidered to dehydrate the cellulose. Pictet, Sarasm and Venn pyrolyzed cotton cellulose under reduced pressure and at low temperature (3000C) found that the yield of tar was much greater from purified cotton cellulose than from raw cotton. Besides them, Madorsky et al. [14] fully pyro— lyzed, at low temperature in vacuum, samples of cotton and viscous rayon after they had been impregnated with sodium carbonate and with sodium chloride. Madorsky concluded that the salts caused a decrease in the yield of tar and an increase in the yields of residue (char) and gases. Recent— ly, Parks et al. [15] studied the effect of about 50 differ- ent impurities on the amounts of char, carbon monoxide, and carbon dioxide produced. In summary, it is generally believed that fire retardant as inorganic compounds and impurities increase the yield of water, carbon monoxide, carbon dioxide, and char, while decreasing the yield of the tar fraction and other inflammable volatile products. Despite many studies conducted about the effects of the impurity on pyrolysis, ther are none for the rapid pyrolysis at high temperature. In this study cellulose with and without potassium carbonate was pyrolyzed under 15 10 mm Hg and 760 mmHg pressure at 6500C, 7500C set point temperatures. The products from light gases to the components of the tar fraction were analyzed quantitatively using Gas Chromatography and High Performance Liquid Chromatography. E e_£8fla._0b 39913 Me. As shown in Section D, the pyrolysis of various biomass including cellulose with additives showed the role of those additives at low temperature and slow heating rate: additives increase the yield of char and gases and decrease the yield of tar. Researchers have investigated whether the additives influence mainly cellulose during the primary reaction, or levoglucosan and other degradation products during secondary reactions, since the efficiency of those additives depends on their availability at the appropriate stage or phase of the pyrolysis. They indicate that additives act mainly on cellulose in the solid phase, but they do not rule out the possibility of catalytic inter— action with the decomposition products in the vapor phase. The previous experiments using potassium carbonate perform— ed by Jonatan E. Trautz [17] might explain the possibility: he analyzed the remaining char after pyrolysis using neutron activation studies and concluded that 1 wt % K2003 on cellulose might play a role as catalyst. This is because the potassium carbonate was found in the residual char, but higher percentages of potassium carbonate (5,10 11 wt %) were lost during the pyrolysis and thus might not be a catalst. Therefore, as a continuing study, this work emphasizes the formation of tar and its analysis with and without K2C03 on cellulose under 15—760 mm Hg of helium and near 7500C peak temperature for the rapid pyrolysis. Recently, some literature [6] have shown that levo- glucosan is dominant in the tar analysis after rapid pyrolysis without additives. But nowhere has the tar analysis after rapid pyrolysis with some additives been reported. Whether pyrolysis will give the same tar composi— tion with and without additives is herein investigated. CHAPTER II EXPERIMENTAL APPARATUS The apparatus constructed for the study of the pyrolysis of cellulose at various condition is designed as follows. A. Reactor The reactor used in the experiments consists of sevel main parts: stainless steel tube, ceramic tube, thermo— couple, aluminum foils, copper electrodes, copper wire, and screen. A 6" in length by 2" 0.D. stainless steel tube, which is connected to a 1/4" copper tube and swaged to a 2" Swage— lock fitting, is made air tight so that it can be run at any pressure. Through a Swagelock connected to a 1/8" copper tube, helium is introduced. The power supply leads are insulated 8 gauge copper wire introduced into the reactor through air tight Conax fittings. These power supply leads connect to the electrodes. Inside the stainless steel tube is a ceramic tube to insulate the steel tube from electricity and heat and to provide a location on which to wrap the aluminum foil. Aluminum foils are used to collect tar in the reactor during pyrolysis: one of the aluminum foils is installed inside the ceramic tube, and the others are placed on the copper wire and copper electrode, particularly at the 12 Figure 3. Pyrolysis Reactor 1. Helium Gas Inlet 2. Electric Wire Inlets 3. Thermocouple Inlets a. Thermocouple on the wall b. Thermocouple between the Screen folded in half . Brass Cross Fitting . 2" Swagelock Fitting . Reactor Support [I 5 6 7. Stainless Steel Tube 8. Ceramic Tube 9. Copper Electrode 0 1 . 325 Mesh Wire Screen 11. Gas Outlet --\C'k ‘ 14 soposom mflmzaoszm .m oszmflm 15 outlet of reactor. Almost all tar after pyrolysis is collected near the outlet of reactor. Since levoglucosan decomposes over 28000, after pyrolysis almost all levoglucosan is supposed to condense in the reactor. One of the thermocouples, an ungrounded junction Chromel-Alumel (K) type, is placed near the wall to measure ambient temperature, and the other, an exposed junction Chromel-Alumel (K) type, is placed between the folded 325 mesh stainless steel screen. The exposed Chromel-Alumel thermocouple is connected to the temperature controller which regulates the screen temperature. The wire screen used in this experiment is folded in half and a small hole is made in which to insert the thermo— couple. The sample is placed between the screen which is then stretched between the electrodes. To heat the sample rapidly and uniformly, the copper electrode parts are designed as shown in Figure 3. When the screen is heated, the thin cellulose sample between the screen will be uniformly heated. Therefore, as mentioned in Section C in CHAPTER I, the cellulose sample will be decomposed satisfactorily and reproducibly. B. Electrical System The electrical system used in the pyrolysis experiment is drawn in Figure 4 and 5. Figure 4 is the electrical system for the pyrolysis reactor. Figure 5 is the electri- cal system for the product gas analysis. The pyrolysis 16 Figure 4. Electrical System of Pyrolysis Reactor 10. 11. Electron Arc Division Power Supply Magnetic Contactor Omega 4001 Single Set Point Proportional and On-Off Controller Omega Model 650 Thermocouple Thermometer Exposed Junction Chromel-Alumel (K) Type Thermocouple Ungrounded Junction Chromel—Alumel (K) Type Thermocouple Copper Electrode 325 Mesh Wire Screen Enlargement of Electrical System inside Reactor Vacuum Gas Line CENCO-MEGAVAC Vacuum Pump 17 1 ~02 2\f 3 4 / 5 6 ..__....— ----— ”bl Figure 4. Electrical System of Pyrolysis Reactor 18 reactor system will be discussed first. The actual power requirement for rapid pyrolysis (>2000C per second) was determined by repeated experiments, since the difference between the actual and the theoretical power requirement is somewhat large. The theoretical and actual power requirement will be shown in Appendix A. Figure 4 shows the electrical system for the pyrolysis reactor. The Electron Arc Division Power Supply (1) used in the experiment transfers alternating current into direct current. According to the previous experiments at 200°C per second heating rate, the maximum current was around 36 amperes. The exposed Chromel-Chromel thermocouple (5) placed between the screen is connected to an Omega 4001 Single Set Point PrOportional and On-Off Controller (3). The other thermocouple (6), used to measure the ambient temperature, is connected to an Omega Model 650 Thermo— couple Thermometer (4). The controller is connected to a Magnetic Contactor (2) which opens the circuit coming from the power supply when the temperature of the screen is above the set point temperature, and closes when the temperature is lower than the set point temperature. Because of the delay in response from the thermocouple, controller, and contactor, the temperature of the screen was found to oscillate around the set point. When the controller was set at 6500C, the highest peak temperature was around 720-76OOC, and when it was set at 750°C, the highest peak temperature was over 9500C. The results of 19 Figure 5. Electrical System of Gas Analysis 1. Power/Mate Corporation Power Supply 2. Model 154L Perkin-Elmer Vapor Fractometer Model XKR Sargent-Welch Recorder Sargent-Welch Electronic Integrator 6-Port Valve described in Figure 6 Hydrogen Transfer System '\1 O\\J‘\ PK» Omega Model 4001 Single Set Point Proportional and On-Off Controller 8. Model 40-200 GOW-MAC Power Supply 9. Sorptometer 2O 4:- ; L: ; 1 Figure 5. Electrical System of Gas Analysis 21 pyrolysis at 7500C and at 6500C set temperature will be shown in CHAPTER IV. The outlet of the reactor is connect- ed to the vacuum pump which is connected to 110 volts A.C. Figure 5 shows the electrical system for the gas analysis. The system is divided into two parts: one for the hydrogen gas analysis, and the other one for other pyrolysis gas (CO, CH“, C02, C2Hu, and H20) anlysis. The Model 154L Perkin-Elmer Vapor Fractometer gas chromatograph and the Sorptometer used for gas analysis are equipped with a thermal conductivity detector. The Wheatstone Bridge Circuit system and filaments which compose the detector are used to describe differences in the thermal conductivities of the carrier gas (reference) and the product gases. Therefore, both need a D.C. power supply (1). The Fractometer (2) uses 110 volts A.C. power to run the blower fan and heat the column. The electrical signal generated from the Fractometer (2) and the Sorptometer (9) is relayed to the Model XKR Sargent Welch Record (3) with integrator (4). Therefore, the gas analysis can be made quantitatively and qualitatively. An Omega Model 4001 Single Point PrOportional and On—Off Controller controls the temperature of the Hydrogen Transfer System (27). The controller is normally set at 5000C. C. Gas Collection System A schematic diagram of gas collection system for CO, CH“, 002, H20, CZH4’ and H2 in the pyrolysis experiments 18 22 Figure 6. Gas Collection System O\O (EV O\\J‘t «P'KO H H H O 12. 13. 14. 15. 16. 17. Helium Gas Tank Liquid Nitrogen Trap Whitey Sample Cylinder T—Connector 2-Way Nupro Metering Valve T-Connedtor and Pressure Gauge Pyrolysis Reactor 3-Way Whitey Ball Valve T-Connector and Vacuum Gauge 2-Way Whitey Ball Valve CENCO-MEGAVAC Vacuum Pump 3-Way Whitey Ball Valve Atmosphere Vent in the G.C. Line Bubble Flow Meter Model 154L Perkin-Elmer Vapor Fractometer 6-Port Valve Dry Ice / Acetone Trap 18 and 19. 3—Way Whitey Ball Valve 20. 21. 22. 23. 24. 25. 26. 27. Liquid Nitrogen Trap 6-Way Whitey Valve Sample Loop Calibration Gas Atmosphere Outlet Calibration Gas Tank (CO, CH4, C02) Calibration Gas Tank (H2) Desiccator Hydrogen Transfer System 28. 29. 3o. 31. 32. 33. 34. 35- 36. 37- 23 Thermal Conductivity Detector Bubble Flow Meter Gilmont Gas Rotameter Capillary Tube Fairchild Regulator Nupro Shut Off Valve Particle Filter Molecular Sieve Dryer Nupro Shut Off Valve Nitrogen Tank 24 Eopmzm SOHpooHHoo new .0 osswflm :33 I on. I Q on Hm mm mm :m mm mm mm om .liiillL mm fllJ mm U.%J an: L . IR mm - H 0 ma _ w Hm , mm u “ iiiaziui em .::|::‘ .11111: . i. cit RH 6H NH mg as. . # .3 m 25 shown in Figure 6. The sorptometer shown in Figure 6 as a part of the hydrogen analysis, is also shown in total in Figure 7. In this section, we will discuss the function of the various equipment used in the experiment. The numbers in the parenthesis refer to the equipment shown in Figure 6. Some of these descriptions are taken from J. E. Trautz [17]. (1) Helium Gas Tank : Helium (99.999%) was used as a purge gas in the pyrolysis experiments. As mentioned in Section C in CHAPTER I, helium is used as an inert to minimize oxidation reactions in pyrolysis as well as a carrier gas in the gas chromatograph. (2) Liquid Nitrogen Trap : This trap consists of a 14" by %" 0.D. U-shaped stainless steel tube. Inside the tube is 0.5 grams of silica gel desiccant (6-16 mesh) used to absorb impurites which might be in the helium. The U-shaped tube placed under the Dewar flask of liquid nitrogen (-196OC). (3) Whitey Sample Cylinder : A 500 cc sample cylinder filled with Linde 3A molecular sieves (1/8" pellets) is used to further absorb impurites missed by the upstream nitrogen cold trap. By passing the helium through the liquid nitrogen trap (2) and the sample cylinder, almost all impurites influencing the yields of products from pyrolysis will be removed. (4) T-Connector : The dried and purified helium is split into two lines: one to the gas chromatograph and the 26 other to the reactor. (5) 2-Way Nupro Metering Valve : The valve regulates the helium flow rate through the pyrolysis reactor. For the collection of hydrogen by the Hydrogen Transfer System (27), this valve regulates the lowest helium flow rate that will be more precisely explained at the Hydrogen Transfer System (27). (6) T-Connector and Pressure Gauge : This gauge is used to detect pressure changes caused by leakage in the reactor before running each sample. (7) Pyrolysis Reactor : This has been already discuss- ed in detail in the preceding section. (8) 3-Way Whitey Ball Valve : This valve has three functions. At the "off" position, it is used to check the leakage of the reactor before pyrolysis. One of the "on" position (a) allows gases to pass through the reactor so that the reactor can be evacuated by the vacuum pump (11), when pyrolysis is run at 15 mm Hg helium. This function is also to eliminate air after the cellulose sample is loaded and the reactor is sealed. Another "on" position (b) is used to purge the product gases after pyrolysis from the reactor through two loops, (17) and (20), and to the Hydrogen Transfer System (27). (9) T-Connector and Vacuum gauge : This gauge is used to measure pressure after evacuation. (10) 2-Way Whitey Ball Valve : This valve is used to isolate the reactor after evacuation to check for leakage. 27 (11) CENCO—MEGAVAC Vacuum Pump : This is used to operate the reactor at pressures below one atmosphere. It is also used to evacuate and to eliminate impurities in the reactor before pyrolysis. (12) 3—Way Whitey Ball Valve : This valve allows helium gas to flow through the G.C. or to be vented to the atmosphere at (13). (13) Atmosphere Vent in the G.C. Line : This vent is used to vent impurities to the atmosphere when the sample cylinder molecular sieves are being regenerated by heating at 2OOOF. (14) Bubble Flow Meter : This is used to determine the flow rate through the G.C. The flow meter is connected to the outlet of the G,C, downstream from the column side of the detector. (15) Model 154L Perkin-Elmer Vapor Fractometer This gas chromatograph is used to analyze the product gases from the pyrolysis experiments. The column used in this chromatograph is a 60/80 mesh Carbosieve S—II. The dimensions of this stainless steel column are 5 ft, in length by 1/8" 0.D. It was used successfully to separate CO, CH“, CO 02H”, and H O. 2’ 2 (16) 6-Port Valve : This valve has two position having three functions. The first position allows the product gases coming from the dry ice / acetone trap (17) to pass through the sample collection loop (20) and to the H.T.S. (27) for selectively analyzing hydrogen. In this 28 position, gaseous products from the reactor are collected in the sample collection loops, (17) and (20). The helium gas passing through the G.C. is vented out to the atmosphere (14). The second position allows the reference side helium from the G.C. (15) to pass through the sample collection loop (20) and return to the G.C. (15). In this way the gaseous products collected in the sample collection loop (20) pass through the G.C. for qualitative and quantitative analysis. In this position, the other gas line from the reactor through the dry ice / acetone trap (17) is connect— ed to the H.T.S. (27). (17) Dry Ice / Acetone Trap 3 This trap consists of a 22" in length by 1/4" 0.D. U-shaped stainless steel tube with fiberglass packed inside. The tube is placed in a Dewar flask containing dry ice and acetone (—77OC). The purpose of this trap is to collect any moisture (H20) or liquid products (propylene, low molecular alcohols) which are the products of the pyrolysis. (18) and (19) 3—Way Whitey Ball Valve : These two 3- way valves are used in combination for two purposes. In the first position (d), the gases from sample gas tank (24) or from pyrolysis reactor (7) are collected in the liquid nitrogen trap (20). In the second position (0), helium bypasses the trap. (20) Liquid Nitrogen Trap : This trap consists of a 12" in length by 1/4" 0.D. U-shaped stainless steel tube 29 placed in a Dewar flask filled with the liquid nitrogen. As the product gas from the reactor passes through this cooled tubing, they condense and are retained in this trap. Since the boiling points of CO (-192.OOC), CH4 (-161.4OC), CO2 (-78.500), C2H4 (~103.9OC) are lower than dry ice / acetone (—77OC) but higher than liquid nitrogen (-196OC), the pyrolysis gases (CO, CH4, C02, 02H4) pass through the dry ice / acetone trap (17) and are condensed in the liquid nitrogen trap (20). The trap contains 0.1 grams od silica gel to ensure complete collection of light gases (CO, CH4, 002, CZH4) which have higher boiling point than liquid nitrogen. After all gases are collected in the trap (20), the trap is placed again in boiling water so that the condensed gases become evaporated again. These gases are flushed into the G.C. for analysis. (21) 6—Way Whitey Sample Valve : This six port valve is used to keep the calibration gas in the known volume of loop (22) and to have it passed through the nitrogen trap (20) for the calibration. Like the six port valve (16), this valve has two positions. The first position allows the gases coming from the calibration gas tank (24) to flow through the 2.045 cc sample loop (22) and to exit to the atmosphere at the calibration gas outlet (23). In the second position, the collected calibration gases are injected to the liquid nitrogen trap (20) by the flow of helium gas, and are eventually analyzed at the G.C. (15) 30 Those gases are then exited to the atmosphere (14). (22) Sample Loop : This 36.8" length by 1/8" 0.D. tube has a volume of 2.045 cc and is used to measure the volume of calibration gases. (23) Calibration Gas Atmosphere Outlet : At the step of calibration gas, the gas from the tank (24) or (25) is exited to the atmosphere (23) after passing through 2 cc sample loop (22). (24) Calibration Gas Tank (CO, CHM, C0 This tank 2) contains a gas of known composition. The composition by volume is 4.9 percent CO, 4.8 percent CHu, 4.9 percent C02, and helium. (25) Calibration Gas Tank (H2) : Pure hydrogen (99.995 percent in volume) was used as a calibration gas. (26) Desiccator : In order to dry the cellulose sample, the desiccator with sodium hydroxide is used. Also, in the process of tar analysis, it is used to dry the tar solution in the votex tube with the vacuum pump after evaporation with the nitrogen stream. (27) Hydrogen Transfer System : The basis of the H.T.S. is the transfer of hydrogen from one carrier gas system into a second carrier gas system for measurement. Since hydrogen has non-linear and unpredictable behavior when measured by thermal conductivity in helium carrier, nitrogen gas is chosen as an alternative carrier. The reason that helium is not good as a carrier is because hydrogen has a thermal conductivity just slightly higher 31 than helium, and would be expected to produce peaks opposite in direction to those produced by other gases. Hydrogen Transfer System consists of two parts: an outer stainless steel chamber and an inner palladium alloy tube. The hydrogen carried in helium from the 6-port valve (16) enters the inner palladium capillary tube of the H.T.S. As the hydrogen moves down the tybe, it passes through the selectively permeable wall of the palladium alloy into a nitrogen carrier gas which is flowing in the opposite direction within the annular space formed by the two tubes. Hydrogen passes from the side with the higher partial pressure of hydrogen to the side with the lower partial pressure. Since the nitrogen carrier gas sweeps the hydrogen away to the detector immediately, the direction of hydrogen transfer is always from the helium carrier gas to the nitrogen carrier gas. (28) Thermal Conductivity Detector : The difference between the thermal conductivities of the carrier gas (nitrogen) and the components of interest (hydrogen) is detected in the detector cell using thermistor filaments. (29) Bubble Flow Meter : This is used to determine the flow rates of helium and nitrogen, respectively. The flow rate of helium is measured when the 6-port valve (16) is turned to the "Reactor through Loop" position. (30) Gilmont Gas Rotameter : This rotameter is used for measuring flow rate of nitrogen. (31) Capillary Tubing : This tube is 6 ft. in length 32 and 1/16" 0.D. (32) Fairchild Mode 10112V Regulator : This is used to control the pressure of nitrogen carrier gas in the sorptometer. (33) and (36) Shut Off Valve : A Nupro Shut Off Valve is used to control the flow rate of nitrogen. (34) Particle Filter : After passing through the Molecular Sieve Dryer, the nitrogen passes through this filter to further absorb impurities. (35) Molecular Sieve Dryer : Like the Whitey Sample Cylinder (3), this is used to absorb impurities including moisture. (37) Nitrogen Tank : Nitrogen (99.998%) is used as a carrier gas in the Hydrogen Transfer System (27). In the previous explanation of the H.T.S. (27), the utility and selectivity of nitrogen for the hydrogen analysis was illustrated. 33 Figure 7. Sorptometer (\D H O\O (EM O\\J\ Pb.) H H O 12. 13. 14. 15. 16. 17. 18. 19. 20. Thermal Conductivity Detector Bubble Flow Meter Sample Cell Exit atmosphere Valco Valve Flow Meter Control Valve Liquid Nitrogen Trap Dampener Nitrogen Tank Regulator NuproShut Off Valve Molecular Sieve Dryer Particle Filter Nupro Shut Off Valve Fairchild Pneumatic Pressure Regulator Moor Pressure Gauge Capillary Tubing Gilmont Gas Rotameter Helium Tank 3H 2 S wH _ E w a ofi 0H HmpoEononw .m whsmflm J8EE- ma :H mH NH HH om OH 3 3 9 NH 2 1%: CHAPTER III EXPERIMENTAL TECHNIQUE This chapter shows the procedure of the experiments in detail. The order of the sections in this chapter corresponds to that in the actual experiments. The numbers and letters in parentheses refer to the equipment and directions shown in Figure 6 in CHAPTER II, respectively. A. Cellulose Sample Preparation To study the effects of impurities in cellulose during the rapid pyrolysis, two kinds of samples are prepared: one is pure cellulose and the other is cellulose treated with K2C03. The pure cellulose used in the experiments is # 4 Whatman filter paper. Since the ash in the cellulose can act as a catalyst, low ash content filter paper ( 0.06 per- cent by weight) is chosen in this experiment. Also the thickness of the filter paper (0.008 inch) is such that heat penetration is effective and therefore the sample temperature is uniform during pyrolysis. The samples were prepared with 0, 5, and 10 weight percent K2003 in cellulose. Samples were placed in the desiccator at least for one week after preparation to dry entirely. The dried sample was cut and weighed to 14 mg. After that, the weighed sample is kept in the desiccator to prevent absorption of moisture. 35 36 B. Wire Screen and Foil Preparation Aluminum foils are prepared to collect the tar in the reactor. There are three foils used to collect the tar in the reactor : these are placed inside the ceramic tube wall, on the electrode, and on the copper wire. The foils are first cut to size, then cleaned with acetone to eliminate unknown impurities, and weighed. The wire screen used in the experiment is 325 mesh stainless steel screen. The screen is cut to A by 8 cm. and folded in half. A small hole is made to put the thermocouple in the screen. The screen is placed through the electrodes without thermocouple and then heated at the set point temperature for 10—20 seconds to remove impurities. The preheated screen is then weighed and placed through the electrode tightly. The advantages of the tightly stretched screen are explained in Section A in CHAPTER II. The thermocouple is cleaned with acetone to Iremove any tar which may be presented from the previous experiments. The uncleaned tar can give higher yield of gases by secondary reactions as shown in CHAPTER I. Finally, the screen is heated again with thermocouple to confirm the heating rate and to eliminate impurities which might be on the screen. The foils are then placed in the prOper locations. 37 C. Sample Loading The prepared cellulose sample is loaded between the folded screen. After loading the sample, the reactor is sealed with 2" Swagelock Fitting and the product gas outlet is connected. The biggest problem in this batch type reactor is the leakage of air. As shown in CHAPTER I, the thermal degrada— tion under air or oxygen has a different kinetic route and is undesirable. Moreover, elevated gas pressure in the reactor after pyrolysis at 1 atmosphere causes pyrolysis gas to leak unless the reactor is sealed completely. To test for leakage, the reactor is pressurized to 40 psig. If no change is found in pressure after 30 minutes, the vacuum pump (11) is switched on, the 2-Way Whitey Ball. Valve (10) is changed to the "on" position, and the 3-Way Whitey Ball Valve (8) is switched to route (a). In this step, the pressurized helium gas is vented and air in the reactor is eliminated. After 5 minutes, the 3-Way Whitey Ball Valve (8) is changed to the "off" position and the reactor is pressurized to 1 atmosphere pressure with helium. After that, the 3-Way Whitey Ball Valve (8) is switched again to route (b). D. Preparation for Gas Collection and Analysis The column pressure controller in the gas chromoto— graph is set to 14 psig and the column flow rate is established at 30 cc per minute using bubble flow meter 1’ 38 (14) and regulating the column pressure controller. Both power / Mate Corporation Power Supply (1) in Figure 5 and the detector voltage switch on the G.C. are turned to the "on" position and adjusted to 9 volts. The recorder and integrator are turned on and fixed at attenuation 32 and chart speed 2 cm per minute. The G.C. temperature controller is turned to the blower setting which activates the oven fan. In order to purify the column in the G.C. (15) and the three traps (2), (17), (20) which may have collected impurities during the previous experiments, boiling water is placed around each trap, and the 6—Port Valve (16) is turned to the "G.C. through Loop" position. The G.C. temperature controller is turned to 2200C and the column is heated. After the column temperature arrives at 2200C, boiling water around the trap (2) is replaced by liquid nitrogen to capture impurities in the helium carrier gas. Boiling water around the trap (20) is removed and the 3-Way Whitey Ball Valves (18) and (19) are changed to route (0) to keep the loop (20) clean. After the trp returns to room temperature, the trap (20) is placed in liquid nitrogen for 5 minutes before injection of the calibration gases (CO, CH4, C02). In this step, zero adjustment is made by regulating the zero point of the Sargent-Welch Recorder. Now the gas collection system is ready to be used. 39 E. Gas Calibration With the gas collection system ready, the 6-Way Whitey Valve (21) is turned to allow calibration gas to flow through the sample loop (22). The calibration gas from the tank (24), which has 4.9, 4.8, 4.9, volume percent of CO, CHu, C02, respectively, flows through 2.045 cc copper loop (22) and is vented to the atmosphere (23). After purging the calibration gases from the tank (24) into the sample loop (22) for 1—2 minutes, the 3—Way Whitey Ball Valves (18) and (19) are switched to route (d) to flow gas through the trap (20). The 6—Way Whitey Ball Valve (21) is switch- ed to inject calibration gas into the helium stream flowing through the trap (20). The 3—Way Whitey Ball Valves (18) and (19) are changed to route (0) to isolate the collected gases in the loop (20). The Model XKR Sargent—Welch Recorder and Integrator are adjusted to the "pen" position. To analyze the calibration gas, after boiling water is placed around the trap (20) and the 3—Way Whitey Ball Valves (18) and (19) are switched so helium gas flushes the gases to the chromatograph, the Model XKR Sargent—Welch Recorder and Integrator are adjusted to the "record" position. The temperature controller on the G.C. is turned to 1750C after 2 minutes. After arriving at 175°C, the temperature controller is turned to blower position to cool down the column, and the 3—Way Whitey Ball Valves (18) and (19) are changed to route (c) to keep the loop (20) clean. The Hydrogen Transfer System is calibrated according #0 to the following procedure. Analysis of hydrogen requires nitrogen carrier gas, because the thermal conductivities of hydrogen and helium are so close that analysis is difficult. In the sorptometer, the nitrogen gas from the tank (37) comes out by opening the Nupro Shut Off Valfe (36). Impurities in nitrogen gas are removed at the Molecular Sieve Dryer (35) and at Particle Filter (34). The flow rate of the filtered nitrogen gas is controlled at 30 cc per minute by the Fairchild Pneumatic Pressure Regulator (32). The flow rate is checked by the gas rotameter (30) and bubble flow meter (29). Hydrogen calibration (99.995% in volume) is passed through 2.045 cc copper loop (22) and vented to the atmosphere (23) for 1-2 minutes. To inject hydrogen for calibration, the 6—Port Valve (16) is switched to the "Reactor through Loop" position, and the 6-Way Whitey Ball Valve (21) is switched to inject hydrogen through route (0) of the loop (20) to the H.T.S. (27). Several pulses of hydrogen are used to clean the instrument. F. Flash Pyrolysis After the calibration gas analysis, the dry ice/ acetone trap (17), liquid nitrogen trap (20), and G.C. column do not have to be cleaned again. To run flash pyrolysis at 15 mm Hg with helium, the reactor must be evacuated. Therefore, after the 2-Way Nupro Metering Valve (5) is closed and the 3-Way Whitey Ball Valve (8) is opened 41 to the vacuum pump (11), the vaciim pump (11) is operated for 5 minutes. The 2-Way Whitey Ball Valve (10) is then closed and the pressure in the reactor is regulated to 15 mm Hg with helium using 2—Way Nupro Metering Valve (5) and pressure gauge (9). The 3-Way Whitey Ball Valve (8) is changed to close position. Before pyrolysis, the trap (17) is placed in the dry ice / acetone bath, and the trap (20) is placed in liquid nitrogen bath to collect pyrolysis gases during the purging step after pyrolysis. Pyrolysis experiments were run at two set point temperature (650°C and 7500C) and in two modes: one where the heater was shut off immediately after the peak tempera— ture was reached, and another where the peak temperature was maintained for about twelve seconds. As mentioned earlier, the delay in response from the thermocouple, controller, and contactor caused the actual thermocouple temperature to overshoot by 1000C at 6500C and 2000C at 7500C. The temperature of the screen oscillated around the set point in the extended experiments. G. Collection of Pyrolsis Gases After pyrolysis, the reactor is cooled down to room temperature to condense the tar sufficiently. After 5-10 minutes, the reactor is pressurized to one atmosphere with helium by regulating 2—way Nupro Metering Valve (5). The Sargent—Welch Recorder and Integrator are prepared to analyze hydrogen by connecting the output leads from the 42 sorptometer to the recorder. The attenuation on the power supply is changed to 64 and 6-Port Valve (16) is switched to "Reactor through Loop" (17), (20), and the flow rate of purge gas is adjusted to 30 [cc / minute] so that the fraction of hydrogen recovered is maximized. Analyzing the hydrogen gas component takes about 40 minutes and the peak of hydrogen comes out at 3.343.6 minutes. During this flushing step, the other gases, CO, CHM, CO C 4’ and 2’ 2 H20 are collected in the traps (20) and (17). Following hydrogen analysis, gases collected in the traps are analyzed by G.C. The recorder and integrator are prepared for analysis of these gases by adjusting the zero point with the G.C. zero knob while the column is at room temperature. To analyze the gases, the trap (20) is removed from liquid nitrogen, while the dry ice / acetone bath is kept around the trap (17) to hold product. Boiling water is placed around the trap (20) and the XKR Sargent- Welch Recorder is switched to the "record" position. Simultaneously, the 3—Way Whitey Ball Valves (18) and (19) are turned to allow helium to flow through the trap. After 2 minutes, the temperature controller on the G.C. is turned to 175°C. In this operation, the peak for each product appeared at 2.8-3.1 minutes for CO, at 5.7 minutes for CH4, 2, at 14.6 minutes for H20, and at 16.5 minutes for C2Hu. The typical Chromatogram will be at 8.3 minutes for CO shown in Appendix B. When the temperature in the G.C. column arrives at 1750C, the temperature controller on the 43 G.C. is switched to the blower position to cool the column. The components in the dry ice / actone trap (17) are next analyzed. Boiling water around the trap (20) is removed and the 3-Way Whitey Ball Valves (18) and (19) are turned to route (0). After 5 minutes, liquid nitrogen is placed around the trap (20) and the 3—Way Whitey Ball Valves (18) and (19) are changed to route (d). The 6-Port Valve (16) is changed to the "Reactor through Loop" position and boiling water is substituted for dry ice / acetone around the trap (17) to transfer contents (almost all water) of dry ice / acetone trap to the liquid nitrogen trap (20). After 30 minutes, the 3-Way Whitey Ball Valves (18) and (19) are changed to route (c). Boiling water is substituted for liquid nitrogen after adjusting the zero of the recorder. After the recorder is switched to the "record" position, the 3—Way Whitey Ball Vales (18) and (19) are turned to route (d). In the same way as before, after 2 minutes, the temperature controller on the G.C. is switched to 1750C. The peak of H20 appears at 12.6 minutes and slowly decays for 14 minutes. H. Collection of Tar After the quantitative and qualitative gas analysis, the reactor is opened and the tar collected on the three aluminum foils is recovered using rubber gloves to prevent contamination from bare hands. The tar with aluminum foil is weighed and then stored in a vial for future analysis. 44 The weight of tar is calculated as the difference of foils after and before pyrolysis. Most of tar was collected from the ceramic tube. I. Collection of Char In the same way as collection of tar, char is recover— ed after gas analysis. The char remaining between the folded screen is weighed. Since the char sometimes can not be separated from the screen, the weight of char is measur- ed by the difference of the weight of the screen and the weight including char and screen. J. Analysis of Tar Analysis of tar was done using Gas Chromatography and High Performance Liquid Chromatography. In order to separate the tar components in the G.C., the tar on the aluminum foil is trimethylsilylated. First, the solution of pyridine, trimethylchlorosilane, and hexamethyldisilazane, is made in the ratio 10 to 1 to 1. The solution is then sealed. Some part of the tar collected on the aluminum foil is dissolved with 250 microliter distilled water: the tar solution is then pipetted into a votex tube. The solution is evaporated to dryness under a stream of nitrogen and residue is further dried for more than 2 hours in a vacuum desiccator prior to trimethylsilylation. For trimethylsilylation, dried tar dissolved with 45 the three mixed reagents. To dissolve the tar and to activate the reaction, the mixture is stirred vigorously for 0.5-1 minutes. After that, accurately measured aliquots of about 5 microliter are drawn into a 10 microliter Hamilton Syringe and injected into a Varian 3700 Gas Chromatograph equipped with Flame Ionization Detector. The column used is 2 m in length by 1/8" 0.D. stainless steel column with packing 3 % OV - 101 on Chromosorb W — HP; the temperature is programmed from 1000C to 2300C at 50C per minute heating rate. Injector temperature is fixed at 1500C and the ionization temperature is controlled at 28000. CHAPTER IV SAMPLE CALCULATIONS AND EXPERIMENTAL RESULTS The experimental data for calibration gases, pyrolysis gases, char, and tar, were collected according to proce- dures in CHAPTER III. This chapter illustrates how those data are manipulated to give the final experimental results. A. Calculation of Weight of Calibration Gases The sample gas tank (24) consists of 4.9 %, 4.8 % CH4, 4.9 % CO2 by volume in helium. Hydrogen from tank (25) is 99. 995 volume percent pure. A 2 045 cc copper loop is and H used to calibrate for C0, CH4, C0 and a 0.056 cc 2’ 2’ loop 18 used for C2H4' Assuming the sample gases are ideal, the weight of each component is calculated by correcting to room tempera- ture in terms of the ideal gas law. Weight P x V x T' x 1 [mole] x Molecular wt [gram] of = (4-1) Interest P' x 22400 [cc] x 1 [mole] The recorder chart speed is used 2 [cm / minute] for all experiment. Table 1 shows the number of counts for the calibration gases (CO, CH“, CO and H at each run. 2’ 2) Since the calibration of H20 and 02H“ is difficult, the fixed values for H20 and C2H4 are used for all experiments. 47 B. Calculation of Weight of Pyrolysis Gases Table 2 and 6 shows the number of counts proportional to the peak area of pyrolysis gases (CO, CH“, CO H 2’ 2’ C2H4’ H20) from the experiments. Since the attenuations of calibration (32) for CO, CH“, CO and H are the 2' C2H4' 2) same as pyrolysis experiments, the weight of gases produced (CO, CH“, C02, CZH4’ H20) can be calculated by follow1ng Eq. (4-2). Weight of Weight of Number of Counts of Pyrolysis Gas Pyrolysis = Calibration x Gas Gas by Number of Counts of Calibration Eq. (4-1) Gas (1-2) The weight of H2 is one-sixteenth of the value by Eq. (4-2) since the attenuation of calibration (64) is sixteen times of pyrolysis experiment (4). Table 3 and 7 shows the weight of pyrolysis gases. C. Calculation of Product Composition The weight percent of each product from pyrolysis is based on the weight of pure cellulose and is defined as follows: Weight Percent Weight of Pyrolysis Component of = X 100 (4-3) Product Weight of Cellulose As a reference, the weight percent of K CO is defined as 2 3 follows: 48 Weight Percent wt. of K2003 * 100 of = (“-4) K2C03 wt. of Cellulose + wt. of KZCO3 Therefore, the weight of cellulose is wt. of Cellulose = wt. of Sample * (100 - wt % K2C03) (4-5) Table 4 shows the calculated values of weight percent for C0, CH4, CO 2, H2, and Tabel 7 for C2Hu and H 0. 2 Table 5 gives the calculated value of weight and weght percent of tar and char. 49 D. Tabulated Data and Calculated Results Table 1. Number of Counts for Calibration Gases # run Number of Counts ________-99_________§1{£+______992_______-_E[_2 ___________ 1 23.08 20.60 27.20 148.80 2 25.41 22.13 27.60 128.25 3 23.44 22.96 27.88 144.70 4 25 15 23 00 27.51 146.74 5 22.00 19.58 25.45 154.74 6, 7 23.24 20.99 26.99 154.74 8—11,13, 22.77 19.95 26.94 134.39 15-17 12, 14, 23.00 24.50 28.50 141.75 18-20 22.77 19.95 26.94 134.39 21 23.61 21.55 26.55 151.00 22 20.85 20.05 23.84 151.00 23—32 23.75 21.20 30.50 139.18 33 31.70 31.80 32.70 149.88 34-36 23.52 20.87 29.28 148.74 37-39 24.16 23.68 30.68 138.45 40 24.18 21.90 27.60 138.43 41 22.45 21.45 30.76 145.61 42 23.60 21.74 30.49 140.15 50 Number of Counts of Pyrolysis Gases # run T P K C0 2 _-__-E-SE_ETT‘E%E-_Y?_1-__€9 ______ 9:1. _____ €92 ______ :12 _____ 1 750 760 0 468.91 72.88 107.34 3286 11 2 477.44 79.76 92.25 - 3 350-74 45.73 75.41 1368.75 4 554.65 104.08 108.50 3974.50 5 750 760 1 346.05 54.25 172 03 2305.06 6 363.44 58.66 167.54 2374.78 7 451.25 83.66 178.10 2803 03 8 438.80 81.16 176 99 2931 08 9 750 760 5 373.83 54.15 261.88 4203 19 10 520 74 78.05 232.67 5534.78 11 446 15 52.60 237 44 4960.43 12 - - — - 13 535.38 65.20 223.50 7167 88 14 750 760 10 573.78 71.17 237.67 7822.90 15 616.40 65.24 191.23 7193.51 16 679.59 66.84 233.42 8912.00 17 556.45 70.45 240.00 7444.93 18 750 10 0 359-95 39.39 66.00 2395.70 19 514.10 61.90 93.85 5229.78 20 522.06 51.76 91.75 3912 60 21 306.70 75.60 100.00 — 22 469.38 53.00 81.81 4168.50 23 750 15 0 — 44.00 88.90 2233.90 24 396.50 43.80 84.88 2331 00 25 472.90 37.20 89.00 3015.75 26 446-79 31.65 84.43 2252.67 27 1371.50 24.60 71.50 286.50 28 - 32.30 80.30 818.00 29 397.90 36.35 77.00 2121.25 30 452.10 41.80 86.00 2415.25 31 — 25.40 91.80 486.10 32 518.38 9.60 88.65 1183.50 33 750 500 0 488.56 77.38 103 33 3127.50 34 650 15 0 307.07 13.70 81.27 2051.13 35 350.08 8.19 78.88 1832.58 35 339.00 11.29 76.74 2175.48 37 300.15 10.00 97.90 1324.50 38 348.26 11.79 110.07 2430 00 39 347.10 17.42 95.30 1546 35 40 650 15 5 327.50 24.40 215.00 3438 89 51 # run OT P K2C03 Number of Counts [ C][mmHg] wt % CO CHu CO2 H2 41 650 15 5 314 43 20.34 210.05 2268.30 42 - 25.37 — 1516.13 43 — 21.43 225.26 2564.05 44 342 83 21 96 246.93 2502 77 45 299 66 19 59 187 87 2068 88 46 348 34 19 14 211 68 2045 02 The symbol (-) indicates that the components was not analyzed. 52 Table 3. Weight of Pyrolysis Gases # run Weight of Gases [grams] CO (x 10-3) CH4 (x 10-5) CO2 (x 10-4) H2 (x 10"1‘L 1 2 331 22.724 _--_-7:116 ---‘-2-309__-- 2 2 156 23.150 6.027 - 3 1 717 12-793 4-877 O 989 4 2 531 29.066 7.112 2 832 5 1 808 17.797 12 189 1 557 6 l 795 17.951 11 193 1 624 7 2 228 25.601 11 899 1 894 8 2 211 26.131 11 847 2 280 9 1 884 17 434 17 529 3 270 10 2.625 25.129 15.574 4.305 11 2.248 16.935 15.893 3.859 12 - — — - 13 2.671 17.094 14.141 5.286 14 2.862 18.659 15.035 5.769 15 3.075 17.104 12.099 5.305 16 3.390 17.524 14/769 6.673 17 2.776 18.470 15.185 5.491 18 1.643 10.998 4.326 1.707 19 2.346 17.287 6.152 3.621 20 2.382 14.455 6.014 2.709 21 2.949 22.533 6.792 — 22 2.583 16.979 6.188 2.886 23 - , 13.331 5.256 1.678 24 1.916 13.271 5.018 1.751 25 2.285 11.171 5.262 2.265 26 2.159 9.588 4.991 1.692 27 - 7.453 4.227 0.215 28 - 9.786 4.747 0.614 29 1.874 11.013 4.552 1.593 30 2.184 12.665 5.085 1.814 31 - 7.696 5.428 0.366 32 1.104 2.909 5.241 0.890 33 1.769 15.628 5.698 2 181 34 1.498 4.216 5.005 1.449 35 1.708 2.521 4.858 1.288 36 1.654 3.475 4.726 1.529 37 1.426 2.740 5.754 1.000 38 1.654 3.198 6.469 1.835 39 1.649 4.725 5.601 1.168 53 # run Weight of Gases [grams] 00 (0 10'3) CH4 (x 10‘5) 002 (x 10‘”) H2 (x 10‘” 40 1.554 7.156 14.047 2.597 41 1.607 6.091 12.313 1.629 42 - 7.496 14.336 1 131 43 - 6.332 13.322 1.913 44 1.667 6.488 14.604 1.867 45 1.457 5.788 11.111 1.543 46 1.694 5.655 12.519 1.525 The symbol (-) indicates that the component was not analyzed. 54 Table 4. Weight Percent of Pyrolysis Gases # run Weight [grams] Weight Percent [wt %] ____§§§92-5f_i9:_3__?f}}?ffff__99_ _ 9H4____992___ H2 1 0 0 0145 16.08 1.57 4 91 1 59 2 0 0145 14.87 1.60 4 16 _ 3 0 0145 11.84 0.88 3.36 0.68 4 0 0146 17.33 1.99 4.87 1.94 5 1 682 0 0143 12.64 1.24 8.52 1.09 6 0 0143 12.55 1.26 7 83 1 12 7 1 311 0 0144 15.47 1.78 8 26 1 32 8 0 0144 15.36 1.81 8.23 1 58 9 9.665 0 0136 13.85 1.28 12.89 2 40 10 9.599 0 0135 19.44 1.86 11 54 2 19 11 9.918 0 0135 16.65 1.25 11.77 2 86 12 0 0135 — — — — 13 9.599 0.0135 19.99 1.55 11.08 4.13 14 18.415 0.0127 22.54 1147 11.84 4.54 15 0 0127 24.46 1.65 10.08 4 41 16 0 0127 26.96 1.69 12.30 5.46 17 18.230 0.0126 22.26 1.80 12.74 4.60 18 0 0.0141 11.65 0.78 3.07 1.21 19 0.0146 16.07 1.18 4.21 2.55 20 0.0146 16.42 0.99 4.12 1.91 21 0.0145 10.34 1.55 4.68 - 22 0.0146 17.69 1.16 4.24 1 98 23 0 0.0141 — 0.95 3.73 1.19 24 0.0141 13.59 0.94 3.56 1.24 25 0.0141 16.20 0.80 3.72 1.62 26 0.0141 15.31 0.68 3.54 1.20 27 0.0140 - 0.53 3.02 0.15 28 0.0140 — 0.70 3.39 0.44 29 0.0140 13.39 0.79 3.25 1.14 30 0.0140 15.60 0.90 3.62 1.30 31 0.0137 — 0.56 3.96 0.27 32 0.0140 17.89 0.21 3.74 0.63 33 0 0.0147 12.03 1.06 3.88 1.48 34 0 0.0136 11.01 0.31 3.68 1.06 35 0.0140 12.20 0.18 3.47 0.92 36 0.0139 11.90 0.25 3.40 1.10 37 0.0137 10.40 0.20 4.20 0.73 38 0.0139 11.90 0.23 4.68 1.32 39 0.0139 11.86 0.34 4.03 0.84 55 # run Weight [grams] K2003 (x 10‘ ) Cellulose CO 40 6.5458 0.0133 11.73 41 6.5932 0.0132 12 10 42 6.5458 0.0132 — 43 6.5659 0.0132 - 44 6.5932 0.0132 12.59 45 6.4035 0.0129 11.33 46 6.5458 0.0132 12.88 OOOOOOO The symbol (-) indicates that the component was not analyzed. 56 Table 5. Weight and Weight Percent of Char and Tar # run Weight Percent Weight [grams] Weight Percent ‘ 2 03 Char Tar Char Tar 1 0 0 0006 0 0044 4 14 30 34 2 0 0005 O 0045 4 14 31 03 3 0 0007 0 0048 4 83 33 10 4 0 0006 0 0045 4 11 31 03 5 1 0 0013 0 0037 9 07 25 82 6 0 0015 0 0038 10 47 26 51 7 0 0014 0 0021 9 74 14 61 8 O 0013 0 0028 9 05 19.49 9 5 0 0028 0 0025 20 54 18.34 10 0 0024 0 0025 17 73 18.46 11 0 0024 0 0020 17 77 14.81 12 0 0019 0 0016 24 37 11 84 13 0 0033 0 0028 24.37 20.68 14 10 0 0033 O 0030 26.07 23.70 15 0 0031 0 0029 24 49 22 91 16 0 0028 0 0031 22 12 24 49 17 0 0038 O 0034 30.21 27.03 18 0 0 0005 0 0050 3.55 35.46 19 0 0004 0 0042 2 74 28 77 20 0 0002 0 0054 1 37 37.00 21 0 0006 0 0044 4 11 30.14 22 0.0006 0.0044 4.11 30.14 23 0 — 0.0042 - 29.79 24 — 0.0051 — 36.17 25 0.0004 0.0045 2.84 31.91 26 0.0002 0.0042 1.42 29.79 27 0.0001 0.0064 0.71 45.71 28 0.0002 0.0051 1.43 36.43 29 0.0001 0.0048 0.71 34.04 30 0.0001 0.0043 0.71 30.71 31 0.0000 0.0052 0.00 37.96 32 - - - — 33 0 0.0090 0.0042 6.12 28.57 34 0 0.0003 0.0059 2.21 43.38 35 0.0003 0.0049 2.14 35.00 36 0.0002 0.0061 1.44 43.88 37 0.0004 0.0060 2.92 43.80 38 0.0003 0.0060 2.16 43.17 39 0.0004 0.0066 2.88 47.48 57 # RUN Weight Percent Weight [grams] Weight Percent T _ __________ 2-92 ________ 9??? _______ 35 ______ 9??? ______ 135--- 40 5 0 0033 0.0020 25.10 15.21 41 0 0037 0.0023 27.95 17.37 42 0 0032 0.0027 24.33 18.25 43 0.0037 0.0022 28.05 16.68 44 0 0036 0.0088 27.19 17.49 45 0 0034 0.0027 26.44 21.00 0 0 .0024 27.38 18.25 The symbol (—) indicates that the component was not analyzed. 58 Table 6. Number of Counts of Pyrolysis Gases (C2H4' H20) 7311?. """" iéiéfié'fiéiéééé """"" REESE—55.25;; """""" — K2003 ___C2H4 _ H39 _________ 1 0 46.29 332.43 2 28.52 344.20 3 14.04 207.21 4 37.05 318.11 5 1 17.14 399 59 6 20.28 419.88 7 28.78 315-35 8 30 75 292-75 9 5 12.25 _ 10 19.18 225.64 11 _ _ 12 - _ 13 11.38 226.80 14 10 7.98 171-17 15 6.24 216.00 16 8.00 179-35 17 11.30 328.15 18 0 13.35 262.05 19 15.10 258.85 3? 27-35 279-50 22 . 17.26 295.89 23 0 37.40 114.05 24 37.00 162.08 25 35.55 245.95 26 34.41 — 27 20 20 193.10 28 30 10 216 30 29 33.60 226.60 30 34.80 194.40 31 25 20 177.80 32 11.25 134.20 33 0 25.58 363-59 34 0 7.58 162.39 35 11.60 190 34 36 14.98 115.24 37 11.91 122.74 59 # run Weight Percent Number 0g Counts _______________ 2__2_______________9§11__________¥29_______ 40 5 4.85 242 39 41 0.87 270 05 42 4.75 297 95 43 3.16 255 04 44 3.78 111 74 A5 4.00 338 09 46 3.00 149 28 The symbol (—) indicates that the component was not analyzed. 60 Table 7. Weight and Weight Percent for 02H4 and H20 # run Weight [grams] Weight Percent -5 _ ______ 92¥1_£f_i9__1__1§29_£f_19__Z____f32i‘1_______1_*2‘3_______ 1 24.2427 18.7814 1.67 12.95 2 14.9605 19.4463 1.03 13.41 3 7.3648 11.7068 0.51 8.07 4 19.4349 17.9726 1.33 12.31 5 8.9910 22.5757 0.63 15.79 6 10.6381 23.7220 0.74 16.59 7 15.0968 17.8263 1.05 12.37 8 16.1302 16.5395 1.12 11.49 9 6 4259 - 0.47 _ 10 10.0611 12.7480 0 75 9 44 11 - - — — 12 — — — — 13 5.9695 12.8136 0 44 9 49 14 4.1860 9.6706 0.33 7.61 15 3.2733 12.2033 0.26 9.61 16 4.1965 10.1328 0.33 7.98 17 5.9275 18.5395 0.47 14.71 18 7.0029 14.8051 0.50 10.50 19 7.9290 14.6243 0.54 10.02 20 14.3467 15.7910 0.98 10.82 21 — — _ _ 22 9.0520 16.7170 0.62 11.45 23 19.6186 6.4435 1.40 4.57 24 19.4087 9.1571 1.38 6.49 25 18.6481 13.8955 1.32 9 85 26 18.0480 — 1.28 - 27 10.5961 10.9096 0.76 7.79 28 15.7893 12.2316 1.13 8.74 29 17.6252 12.7853 1.26 9.13 30 18.2547 10.9831 1.30 7.85 31 13.2189 10.0452 0.96 7.33 32 5.9013 7.5819 0.42 5.42 33 13.4182 20.5418 0.92 13.97 34 4.1834 9.1746 0.31 6.75 35 6.0849 10.7537 0.43 7.68 36 7.8579 6.5107 0.57 4.68 37 6.2475 6.9345 0.56 5.06 38 4.1178 12.2791 0.30 8.83 39 9.2323 13.3130 0.66 9.58 61 # run Weight [grams] Weight Percent -5 _ ______ 9214-12-19-3-3231219__Z_____‘325‘1_______§‘29______ 40 2.5441 13.6944 0 19 10.41 41 0.4524 15.2571 0 03 11.52 42 2.4917 16.8333 0.19 12.80 43 1.6589 14.4090 0.13 10.92 44 1.9802 6.3130 0 15 4.77 45 2.0982 19.1011 0 16 14.85 46 1.5737 8.4336 0 12 6.41 The symbol (—) indicates that the component was not analyzed. 62 Table 8. Weight Percent of Pyrolysis Products # run Weight Percent Based on the Weight of Pure Cellulose C0 _ CH1 C°g____¥g____<321 _____ 529----91‘?E_-_f‘:‘1 1 16.08 1.57 4 91 1 59 1 67 12.95 4.14 30 34 2 14.87 1.87 4 16 — 1 03 13.41 4.14 31 03 3 11.84 0.88 3.36 0.68 0 51 8.07 4.83 33 10 4 17.33 1.99 4.87 1.94 1 33 12.31 4.11 31 03 5 12.64 1.24 8.52 1.09 0 63 15.79 9.07 6 12.55 1.26 7.83 1.12 0 74 16.59 10.47 26 51 7 15.47 1.78 8.26 1.32 1 05 12 37 9.74 14 61 8 15.36 1.81 8.23 1.58 1 12 11 49 9.05 19 49 9 13.85 1.28 12.89 2.40 0.47 — 20.54 18.34 10 19.44 1.86 11.54 3.19 0.75 9.44 17.73 18.46 11 16.65 1.25 11.77 2.86 - - 17.77 14.81 12 — - - - - 14.07 11 84 13 19:99 1.55 11.08 4.13 0.44 9.49 24.37 20.68 14 22.54 1.47 11.84 4.54 0. 33 7.61 26 07 23 70 15 24.41 1.65 10.08 4.41 0. 26 9.61 24.49 22 91 16 26.96 1.69 12.30 5.46 0. 33 7.98 22.12 24 49 17 22.23 1.87 12.75 4.66 0. 47 14.71 30.21 27 03 18 11 65 0.78 3.07 1.21 0 50 10 50 3.55 35 46 19 16.07 1.18 4.21 2.55 0 54 10.02 2.74 28 77 20 16.31 0.99 4.12 1.91 0 98 10.82 1.37 37 00 21 10.34 1.55 4.68 — 4. 83 23 45 22 17.69 1.16 4.24 1 98 0 62 11.45 4.11 30 14 23 - 0.95 3.73 1.19 1.40 4 57 2.75 29 79 24 13.59 0.94 3.56 1.24 1.38 6 49 3.28 36 17 25 16.20 0.80 3.73 1.61 1.32 9 85 2.84 31 91 26 15.31 0.68 3.54 1.20 1.28 - 1.42 29 79 27 — 0.53 3.02 0.15 0.76 7.79 0.71 45 71 28 - 0.70 3.39 0.44 1.13 8.74 1.43 36.43 29 13.39 0.79 3.25 1.14 1.26 9.13 0.71 34.04 30 15.60 0.90 3.63 1.30 1.30 7.85 0.71 30. 71 31 - 0.56 3.96 0.27 0 96 7.33 0.00 37. 96 32 17.89 0.21 3.74 0.63 0 42 5.42 - — 33 12.03 1.06 3.88 1 48 0 91 13.97 6.12 28.57 34 11.01 0.31 3.68 1.06 0.31 6.75 2.21 43.38 35 12.20 0.18 3.47 0.92 0.43 7.68 2.14 35.00 36 11.90 0.25 3.40 1.10 0.57 4.68 1.44 43.88 37 10.40 0.20 4.20 0.73 0.56 5.06 2.92 43.80 38 11.90 0.23 4.68 1. 32 0.30 8.83 2.16 43.17 39 11.86 0.34 4.03 0. 84 0.66 9.58 2.88 47. 48 63 # run Weight Percent Based on the Weight of Pure Cellulose _____ 99____9¥1-___992112-1224___:Ig9____918:____2:_ 40 11.73 0 54 10.60 1.96 o 19 10 41 25.10 15 21 41 12.10 0 46 9.30 1.23 0 03 11 52 27.95 17 37 42 — 0 57 10.90 0.86 0 19 12 80 24.33 18 25 43 — 0 48 10.10 1.45 0 13 10 92 28.05 16 68 44 12.59 0 49 11.03 1.41 0 15 4 77 27.19 17 49 45 11.33 0 45 8.64 1.20 0 16 14 85 26.44 21 00 46 12.88 0 43 9.52 1 16 0 12 41 27.38 18 25 The symbol (-) indicates that the component was not analyzed. Table 9. Average 64 Weight Percent of Pyrolysis Products at Various Conditions Set Point [0C] Temperature H O 10.64 Char26.92 .05 750 760 10 13 1 14.49 16 1.61 1 8.34 4 1 33 1 0.93 1 13.22 12 9.29 4 19.98 30 CHAPTER V CONCLUSIONS AND DISCUSSION The results of the rapid cellulose pyrolysis with and without K2CO3 under various conditions were shown in Table 9 in CHAPTER IV. Those results indicated the effect of temperature and pressure on rapid cellulose pyrolysis, and what the role of KZCO3 was. This chapter gives an analysis based on those results. Table 9 represents the data at two different set point temperatures (6500C, 7500C). It is generally agreed that the tar yield is maximized at some critical tempera- ture near 7500C. Therefore, the experiments focused on obtaining the maximum yield of tar by setting the tempera— ture contrller to 65000. When the temperature controller was set at 6500C and 7500C, the highest peak temperatures were around 720—7600C and over 9500C, respectively. The experiments were run at three different pressures of helium (760 mm Hg, 10 mm Hg, 15 mm Hg). The experiments at 760 mm Hg and 19 mm Hg helium were done for 12 seconds duration time, while those at 15 mm Hg helium at both 7500C and 6500C were run only until the initial peak temperature was reached. A. Temperature and Pressure Effects on Pure Cellulose Pyrolysis It is generally believed that temperature and holding 65 66 time of rapid pyrolysis are the factors most influencing the pyrolysis product composition. The results of pure cellulose pyrolysis in Table 9 show the effects of tempera— ture. When the results of pure cellulose pyrolysis at 7500C and 15 mm Hg pressure are compared with those at 6500C and 15 mm Hg pressure, the yield of tar at 6500C (43 wt %) is much larger than the one at 7500C (30 wt %), while the yields of gas and char at 6500C are a little less than those at 750°C. The results are interpreted that the tar formed is decomposed to gases via secondary reactions at elevated temperature. It is generally agreed that secondary cracking of tar increases the yields of carbon monoxide, carbon dioxide, hydrogen and hydrocarbon gases such as CH” and C2H4' Hajaligol et al. [8] showed the maximum yield of tar in rapid cellulose pyrolysis occurred at 7000C and gradually decreased at higher temperature. Though the yield of tar in this experiment is less than theirs [8], the results are generally in agreement. The lower tar yield possibly results from loss of volatile liquids during purging. The effects on pressure in cellulose pyrolysis at 7500C are negligible, as the yields at 760 mm Hg and 10 mm Hg showed no difference at all. Therefore, the effects of pressure on cellulose pyrolysis at high temperature are not important unless possibly the pressure is increased to a very large value. 67 The results of pyrolysis at 7500C and 10 mm Hg and 15 mm Hg pressure in Table 9 show the effects of holding time. Only slight differences are found in product composition, indicating that almost all intermediate vaporized products diffuse out of the cellulose and the pyrolysis is complete when the peak temperature is reached. Therefore, it can be concluded that the yields of products for very thin cellulose pyrolysis at rapid heating rate ()ZOOOC per second) and at high peak temperature ()95OOC) are not dependent on holding time at the peak temperature. B. Effects of Impurities (K2003) The literature [10—15] shows the effects of additives on products from slow pyrolysis. These results conclude that the role of additives in pyrolysis is to reduce the yield of tar while increasing the yields of char, CO, C02, and H20. Though these results are from slow pyrolysis, results from our experiments indicate the same trends. From the comparison of pure cellulose pyrolysis with K2C03-treated cellulose pyrolysis, it is confirmed that K CO increases the char yield and decreases the tar yield 2 3 drastically regardless of the set point temperature. At 7500C set point temperature and 760 mm Hg pressure, the results of four different samples (0, 1, 5, and 10 weight percent KZCO cellulose samples) are shown in Table 9. As 3 the weight percent of K2CO3 on the cellulose sample is increased, the yields of C0, C02, and H2 show a monotonic 68 increase, while the yields of hydrocarbons (CH4, C2Hu) are reduced. Moreover, the yield of char is drastically increased while the yield of tar was abruptly decreased. The CO yields from the 1 wt % KZCO samples are scattered, 3 perhaps as a result of nitrogen impurity during the gas chromatographic analysis. Nevertheless, the yield of CO 2C03. When the cellulose pyrolysis are compared at shows a gradual increase with increasing K results of 5 wt % KZCO3 two different set point temperature (6500C, 7500C), it is found that gases and char yields at 7500C are larger than those at 6500C, while the tar yield at 7500C is smaller than that at 6500C. Even though the comparison is somewhat difficult because the actual weight percent of K CO for the sample at 7500C is a little larger than the 2 3 one at 6500C, the results could be explained by the relationship between the presence of K2CO3 on the cellulose and the temperature of the screen. It is generally agreed that the yield of tar is maximized around 7500C, that the yield of char is increased at lower temperature, and that KZCO3 as a flame retardant lowers the activation energy and threshold temperature of the pyrolysis. These facts explain the results in light of the higher rate of volatilization at lower temperatures and the increased yield of the residual char, which lessens the flaming combustion [3]. Therefore, the anlysis of the products suggests that KZCO3 primarily promote the dehydration and charring of 69 cellulose, and lessen the cleavage of intact pyranose monomers (as levoglucosan). This is in agreement with Madorsky et al. [14], who proposed in 1956 that sodium chloride and sodium carbonate as additives catalyze the dehydration of cellulose by scission of the C—0 bond and thus increase the yield of CO, CO H O, and char at the 2’ 2 expense of levoglucosan. This proposal seems to be valid CO for the interpretation of these results (Table 9) of K2 3 loaded cellulose in rapid pyrolysis. C. Analysis of Tar Figures 8, 9, and 10 show the results of analysis of tar obtained from the O, 5, and 10 weight percent KZCO3 loaded cellulose in rapid pyrolysis. The tar analysis was done by High Performance Liquid Chromatography [19] and by Gas Chromatography [18]. Since the results of tar analysis by G.C. showed only the presence of levoglucosan and D- glucosa, the results by H.P.L.C. only are shown in this chapter. The only calibrated components in the analysis were D—glucose and levoglucosan: therefore, the analysis focuses on the composition ratio for each tar sample. Figure 8 shows that levoglucosan and D-glucose are dominant components in tar analysis for pure cellulose pyrolysis. The ratio of levoglucosan to glucose is more than 2 to 1. The peaks at retention time 15.1 minutes and 29.7 minutes are D—glucose and levoglucosan, respectively. The peaks at retention time 12.6 and 13.7 minutes are 70 FILE: nn'mzszsxmea3 SCOLE: 1 110883 (NHL): 8.82 T0 48.88 25283 " M I- § 8 18229- ’ I I l I l I I s 18 1s 20 25 38 - 35 48 "1140128 ERROR: ZUnknoun command """"""""""" AREA—PEECENT_EEPOHT_"_ Peak H.T.(min) Area Percent ‘ Area Peak Ht BL 1 12 642 3 761 19095 366 BV 2 13.767 2.716 13789 515 VB 3 15.101 24.882 126324 3447 BB 4 16.785 0.646 3281 28 BB 5 22.055 0.558 2834 36 BB 6 29.722 67.437 342379 6828 BB TOTAL 100.000 507702 Figure 8. Tar Analysis (0 wt % K200 ) 3 71 FILE: DnTAZEfiSXXflBM SCGLE: 1 RnHCE (1118.): 8.82 To 48.88 15851] COUNTS 5. lmfi‘ W 11653.“fi ' ' I I I I I I I s 18 15 28 25 38 35 48 Peak R.T.(min) Area Percent Area . Peak Ht BL 1 14.950 32.865 7934 225 VB 2 20.432 12.613 3045 82 BB 3 29.500 54.521 13162 250 BB TOTAL 100.000 24141 Figure 9. Tar Analysis (5 wt % K2003) 72 FILE: nnrnzs:sxxnae4 SCnLE: 1 names (nxu.): 8.82 T0 36.88 187671 COUNTS 18119 - Peak R.T.(min) Area Percent Area Peak Ht BL 1 21.867 60.203 3269 134 BV 2 22.150 39-797 2161 122 VV TOTAL 100.000 5430 Figure 10. Tar Analysis (10 wt % KZCO ) 3 73 postulated to be dimers and cellubiose, but no confirmation is possible. Figure 9, which is the tar analysis of 5 weight percent K2CO3 cellulose sample, represents somewhat similar results to Figure 8, but it is seen that the area ratio of levoglucosan and D-glucose is decreases to around 1.7 to 1, while the yield of the unknown product at a retention time 20.4 minutes is drastically increased. Also in this experi- ment, the amounts of levoglucosan and D-glucose were about one-twenty sixth and one-sixteenth respectively of those for pure cellulose, even though the amount of tar analyzed was within a factor of two of the pure cellulose. Figure 10, which is the tar analysis of the cellulose pyrolysis with 10 weight percent of KZCO shows similar 3. results. It is very difficult to interpret each peak except levoglucosan and D—glucose. Some literature [6,13] has shown levoglucosan is the main component in tar in slow pyrolysis as well as in rapid cellulose pyrolysis. But the quantity of levoglucosan in tar depends on temperature, heating rate, residence time, additives, molecule structure, and even crystallinity. Therefore, it can be concluded from these figures that the amount of levoglucosan in tar decreases as the weighy of K2CO on cellulose is increased, while some unknown 3 components are drastically increased. 74 D. Conclusions The gaseous products of rapid cellulose pyrolysis with and without K2CO3 at high temperature were quantitatively analyzed by Gas Chromatography and char and tar yields were measured by weighing. The rapid pure cellulose pyrolysis at high temperature resulted in more tar yield and less char yield as the temperature on the screen was increased up to 7500C. Some secondary cracking of levoglucosan occurred at higher temperature. When K2C03—treated cellulose was pyrolyzed at the same conditions (tempera- ture, pressure, heating rate), the yield of tar was drastically decreased and char yield was increased. In addition, the yields of CO and C02 were increased, while the yields of hydrocarbons (CH4, C2H4) were slightly decreased. In the tar analysis, it was confirmed that levoglucosan is dominant component in tar for pure cellulose pyrolysis, but as the amount of KZCO3 was increased, the quantity of levoglucosan in the tar *decreased and the yield of the unknown components in tar increased. Therefore, the presence of K2003 in rapid cellulose pyrolysis as well as in slow pyrolysis has a catalytic effects on dehydration and bond scission, and alters the pyrolytic mechanism to yield more char, C0, C02, H20, H2, and other components in tar. E. Recommendation Although the contour of the influence of additives 75 and the predominance of levoglucosan in tar in pyrolysis experiments has been manifested, it still remains to identi- fy the unknown components and to manifest the reason why they are increased with increasing the additives. To yield more tar and other products (low molecular alcohol, furfuran, etc.), it is suggested to modify the reactor. In this experiment, aluminum foils were used to collect the condensed tar, so that some of tar and other products which have low boiling points did not condense on the foils and thus were lost. Thus, to spray water with high pressure might be more effective to collect volatile products without loss before the temperature in the reactor reaches at room temperature. This modified reactor, however, might require more complex equipment and procedure: water sprayer, liquid collector, heating tape, etc. LI ST OF REFERENCES [1] T. [2] I. [3] F. [4] P. 76 LIST OF REFERENCES A. McCLure, E. S. Lipinsky, "Handbook of Biosolar Resources" Vol. II (1981) S. Goldstein, presented in Symposium on Alternate Feedstocks for Petrochemicals, 18th Nat. Meet. Amer. 800., Las Vegas, Nev (1980), Abstract PETR 21 Shafizadeh, "Pyrolysis and Combustion of Cellulose Materials", Advances in Carbohydrate Chemistgy, 23, pp. 419—475 (1968) C. Lewellen, W. A. Peters, J. B. Howard, "Cellulse Pyrolysis Kinetics and Char Formation Mechanism" Sixteenth Symposium (International) on Combustion, The Combustion Institute Pittsburge, pp. 1471-1479 1977 [5] Toshimi Hirata, "Pyrolysis of Cellulose, An Introduc- [6] C [8] M. [9] X. [10] F. tion to the Literature", U. S. Department of Commerce, NBSIR 85-3218, August (1985) I. DeJenga, M. J. Antal Jr., "Yields and composi— tion of Sirups Resulting from the Flash Pyrolysis of Cellulosic Materials Using Radiant Energy", Journal of Applied Polymer Science, Vol. 27, pp 4313—4322 (1982) Diebold and J. Scahill, "Ablative Fast Pyrolysis of Biomass in the Entrained-Flow Cyclonic Reactor" at S.E.R.I., Fourteenth Biomass Thermochemical Con— version-ggntrag§9 s Masting. Arlington. Virginia June 23—24, (1982) U. S. Department of Energy Contract # DE—ACO6-76RLO 1830 R. Hajaligol, W. A. Peters, J. B. Howard and J. P. Longwell, "Product Compositions and Kinetics for Rapid Pyrolysis of Cellulose", Ind. Eng. Chem. Prod. Res. Dev., 21 (1982) I “— Deglise, C. Richard, A. Rolin and H. Francois, "Fast Pyrolysis/Gasification of Lignocellulosic Materials at short Residence Time", Energy from Biomass, Applied Science Publishers, pp. 548-553 1981 H. Holmes and C. J, G. Shaw, "The Pyrolysis of Cellulose and The Action of Flame-Retardants", J. Appl. Chem., 11, June (1961) [11] [12] [13] [14] [15] [16] [17] [18] [19] G. 77 A. Byrne, D. Gardiner, and F. H. Holmes, "The Pyrolysis—of Cellulose and The Action of Flame- Retardants", J. Appl. Chem., Vol 16, March (1966) P. C. Fung, Yoshio Tsuchiya, and Kikuo Sumi, "Thermal Degradation of Cellulose and Levo— glucosan The Effect of Inorganic Salts", Wood Scienge, 5, No. 1, July (1972) Yoshil Tsuchiya and Kikuo Sumi, "Thermal Decomposi- tion Products of Cellulose", Journal of Applied Polymer ScieQQE. Vol. 14, pp. 2003-2013 (1970) L. Madorsky, V. E. Hart and S. Straus, "Pyrolysis of Cellulose in a Vacuum", J. Res. Natl. Bur., 56, (1956) G. Parks, J. G. Erhardt, Jr., and D. R. Roberts, Amer. Dyestuff Reptr. 294 (1950) M. Suuberg, W. A. Peters, and J. B. Howard, "A Com— parison of the Rapid Pyrolysis of a Lignite and a Bituminous Coal", Thermal Hydrocarbon Chemistry pp. 239, Advances in Chemistry Series No. 183, A. G. Obald, H. G. Davis, and R. T. Eddinger, Ed., Am. Chem. 809,, Washington, D. C. (1979) E. Trautz, "The Flash Pyrolysis of Cellulose in the Presence of K CO ", . S. Thesis, Department of Chemical Enginéer'ng, Michigan §§§t9_HQiY§£: sity, East Lansing (1985) C. Sweeley, R. Bently, M. Makita and W. W. Wells, "Gas-Liquid Chromatography of Trimethylsilyl Derivatives of Sugars and Related Substances", J. Am, Chem. Soc., Vol. 85, 2497 (1963) Pecina, G. Bonn, E, Burtscher, and O. Bobleter, "High-Performance Liquid Chromatographic Elution Behavior of Alcohols, Aldehydes, Ketones, Organic Acids and Carbohydrates on a Strong Cation- Exchange Stationary Phase", Journal of Chromatof graphy 287 (1984) APPENDIX APPENDIX A Power Requirements for Heating the Screen Theoretical power requirement is total heat transfer by conduction, convection, and radiation. 1. Heat Flow by Conduction (00p _%%— = (v-va) --------------------- (A—l) If the thermal conductivity (k) is independent on tempera- ture or position, Eq. (A-l) becomes ’55- = 01V T ————————————————————— (A—2) where d = thermal diffusivuty (M = kflon) = fluid density = heat capacity By dimensionless form of temperature, 9 = (T-TO)/(T1—TO), the Eq. (A-2) for one dimension becomes 69 329 ---- = ----§ --------------------- M _Y_- A “ P _ ' * —.'_ —: “' “H t I ‘ I I ' I I I \ l\ .._._..._ 1 _ j .a . y I : I 1 l ! I ’ . § _ I ,_ ._ _ J l i 4 v - Figure 12—a(1). Chromatogram of water without loading water (liquid nitrogen trap (20)) I . ‘ I I r’ ‘I ___J_ I I _I I I. I . ! . I l I . ' I ! » , l ' ‘ ' 64' I An I ‘4 ’ ‘U 1 ‘ I l . I I 3 ‘\._ I I " I I ____. . I ‘ 3 -ii.,-..r- ___.._ _.__ _ “.7 A / l J 4/ L I) l . b. r ‘ ' | . I I I ~ — , ~—— ———e.~I—V—,HT— - . I II. I i , I I I II —_..— "3 r" I l l I | ‘ l I I '. I i ‘ _ l ,_3.- _. __.__ ___ \/I \A I I\I/ I 3. V Minnow as; = SARGENT-wéLCH SCIENTIFIC comm; I c Figure 12-a(2). Chromatogram of water without loading water (dry ice/acetone trap (17)) 88 . I l. l .. ,g. ‘ r I I , . ‘ I ‘ ' V ' 1 ‘ i : j I . : I l a I I A I. I I > - ' I I I ' an I . n '. , L“) \‘ ‘ ‘ i , |' I 5 ”I“ l I I i ’ ~ —- A-~-- , A~——~——r 7,~— I ”T f T I I 1 I x ‘. I l . i i l _ - l__ _ J___ l i __ --___ - __ _.._._._- l ____________ _~_ _I__.___ _ I ‘r I ‘ ‘ I I ‘ v i ‘7‘ 7 ! _ _ : i i , ‘ 1 h’ ‘ -——A — ~‘*” "1” > H — -—-*— *r- —- ,i, I | ‘ ‘ (I ; ' I i I | i _ I t { I I? i , I, z. I t ' ' ' ' ‘ I I ‘ I I . , I l I » } | y I 9 I l ‘ l I I I I ., I I K/\ A A A A A / v I,AAA;IV’AAnNAIIAQANy. u \ Figure 12—b(l). Chromatogram of 4.u[ distillated water (liquid nitrogen trap (20)) I I I ‘ 1 I I E 5 I‘ V l\. i l " I . I v . ' ' , . l I i ' I l I I . I _ I I 5 i L l 0 L4 /\/\7\/\/\ VVIVVI 'v‘; v‘, ‘ = snncam- WELCH scIsm-IFIc COMPANY ' ‘ R ' W CATALOG NO. s-72164-- Figure 12— b(2). Chromatogram of A.u1 distillated water (dry ice/acetone trap (17)) APPENDIX C CHROMATOGRAM OF PYROLYSIS GASES AND RAW DATA Figure 13 is the Chromatogram of pyrolysis gases (CO, CH“, C02, H20, CZHA) at 65000 set point temperature, 15 mm Hg helium. The chart speed is 2 cm/min and attenuation is 32. Figure 14 is the Chromatogram of pyrolysis (H2) using Hydrogen Transfer System. The chart speed is 2 cm/min and attenuation is A. Figure 15 is the Chromatogram of pyrolysis (H20) obtained from the dry ice / acetone trap (17). The chart speed and attenuation are the same as those in Figure 13. 89 9O Aimwo .omm .Noo .imo .oov momww mfimhaohhm wo anymopmsopso .mH ohsmflm .‘III I IIIIIIIII IIIIII. IIIIIII III 4 wI 91 Ammv was -iIII mama . H09 km 90 swgmopm 809: O I I I .II d.“ 92 Aomzv mmo mflmzaophm wo EMMwOpMEOMSU .mH mgsmflm 93 0% [Em/K- ,_ _,Isd-,p.f.1‘h7g (fibm @___,,W W '_ _ é‘f‘u—V afav a , __ ‘ rm AI?“ mom mo ALF)” Paw» M nag! OI/a7o am! (Tifffljjk / o, “(00‘ 0. ouch “_____‘__*_H _}'er .“v 1!; ~ I, ”I763; 0. 005 92. 9:33: «bx ., “flaky-4...”- era/31," . 7 _m 7. J y, 7 LL; ,i-_.9€ae.<~ , 7 .—) _ Co .__._....~, 15.5% . ofi4__._i...-.__)'o ..J’7 Cob --~~-~~ ’7‘ >9 4/334'1 / 3.1/ % AMI-‘1 I 707. 07 {3 .‘Io d‘z.>7 1% Ha- » ' ' I.Ll'P'\‘I‘f ( W64) vhf/‘13 If,» W1, ‘ "’ ('7). arr” " /6L . 3f 7' 7} Z _ --..._._ ._ _ ( ___._._. _.__v .i W- -J. _. ...——_.. m_.—...q._ N- Iw» ~M VIEW #1/ 94 H, mm, , 9%,}..127»: 7&0}. $145: , ,_ 5M .191 7V iv W m. ‘ q, 5W 5% {v WWII/7am. M 11%” “ “‘ ; .@._I_.__I_. bokW/I W”, , A M____ I II .. < Wm ..... __0 31¢.7 _~~w 0 loo/z__ .0. ”I7 [[HZEW MaQJ-w—AJ‘ FJk—w—Mmq xu a_— °\ “4}! . 9LM___ _2~_/2._// ___._ .I 0 °°/I_ , . 0mm . “or-..” __ Jam» ,___.,l9..,?70 »—-~n——— w< A”... I...‘ — (za‘xw . _-___,_ . 0""! i (I ova/4, rm)---_.m.l “bl/“W loo/y Z: t? ”A” ii...” a I _Megér¢fio~ ~__w m, :/Ml 1./A- . -I...._._I._..II..- 7 0° ., ,,J—3I>37r A .i, 411:.21’ _,____,l_.c_ 01¢“ . voflia—Ir ._- (01* w... ' do; ..... 77’6w7711r15' / _, In ..,I.S’¥\ 144m (w w) l. ago; °>r ( m/_ U H.» mm -_ an“ 73— ”I---“ CHI/“M,“an >f“ , 4... HS“ 113’ , _______ w, _...____ __ V. __ __ ‘ ._...._.__.._._. __ i .ii- .i m“...— -_._ 95 mm CA1 7$ra1£o€x _ F‘T‘PJN 256' 79$st /J‘. 3' 3'. 41. 6 “07; MW n fl KL 4+ IIIA>n1&II 56M ? flaw/Lt, W aoma/a. I-/% ”I 097% (V) x I. /m‘- M mixer mar/r»; AL 057 AkW/arc 47%» Wm; /'7f°c, 7914.» M And?» my filiaw‘ Cami AL mew {Mend/n. I’M fir .Ax,MmfluI/WA+M/7 [.4 IA)» 7937 MM Nil: arm- 49;» In}, W I III... I. @f.fidy do v57, m4» *0 UM?“ rrfiw» AIchJovfo {ya} Maid/Maw mlzu/ Mfed’. II III I M 60a} r I --__III aw. aux/776M [Myfv 7:1'c braced/n O‘W” (- W --7- »- )1! #4. in, WW I I II (5;? /V ”dd, M $796 MVAM hum 57* M/waJ-uym/ a} yM w”, M7, mac. 60M H7 W Af- /7f"¢. ...II-I-I-I_II__-, no”. WW /o;’/m~ M 6/942/ AL 957/449» M L/flf" Unawa/ M J— 709$ VAQMV fwl- A4- »«R 51" A. h-76\ . -WQ 6W. “fir/imam:[uujrflrgnanflnmmr