~ MICROWAVE PROCESSING OF LIGNIN AND CELLULOSE PART I-AN INVESTIGATION OF THE REACTIONS OF LIGNIN AND HYDROGEN USING A CONT INUOUS FLOW MICROWAVE DISCHARGE REACTOR PART II -THE EFFECT OF MICROWAVE RADIATION ON BATCH PROCESSING OF CELLULOSE IN WATER Thesis for the Degree of M. S. MICHIGAN STATE UNIVERSITY JACK D. JUNTIILA 1977 w-rvw:usww”’ 4 4 .. .9 "LW Y . .1 in} 13‘. ‘11 ix . -, ie'§:?,:_‘_-§.L‘:1 State Umvemcy ..l'fiw ‘i A I.. “MM, \l . — s... I A I I... AAAAA ABSTRACT MICROWAVE PROCESSING OF LIGNIN AND CELLULOSE By Jack D. Junttila Experiments have been performed investigating the application of microwave technology to chemical processing of waste products produced by the paper and pulp industries. The purpose of this study was to determine if use of microwaves as a catalyst could pro- mote production of useful chemical products from the decomposition of paper waste byproducts, lignin and cellulose. Two separate micro- wave processing schemes were utilized for this investigation: (I) con- tinuous flow microwave plasma processing and,(2) high pressure batch microwave processing. Part I of this study describes experiments which used microwave radiation for generating a low pressure hydrogen plasma in a quartz tube flow reactor. The reactions of this plasma with lignin contained ‘in Kraft Process Black Liquor were studied. The reaction product pro- duced in largest quantity in this reactor was a solid char. Small amounts of methoxy-phenols were also produced. These reaction products obtained are similar to those produced in conventional pyrolysis re- actions and indicates that high energy pyrolysis reactions dominate over radical reactions when lignin is injected into a hydrogen micro- wave plasma. Jack D. Junttila It was concluded from these experiments, that a single one step process of direct injection of lignin into a hydrogen plasma is not a viable method for producing useful chemicals from lignin. The most appropriate direction for further study of lignin processing should incorporate a two-stage process scheme which combines existing lignin technology with microwave technology. Described in Part II of this study are experiments which involved use of a batch reactor for investigating the effect of microwave radi- ation on high pressure and high temperature reactions. A Specially designed teflon-brass coaxial coupling was used to conduct the micro- wave radiation from the microwave source into the reactor interior. The effect of microwaves on the reactions of crude cellulose and water was investigated at 500 psig and 450°F using this apparatus. Microwave catalysis appeared to have little effect on this reaction and did not promote production of useful products. The product produced in great— est quantity, regardless of catalyst used, was a black, water soluble, tar-like decomposition product with a yield of approximately 35 percent. The results of this two part experimental study indicate that the techniques developed for the application of microwave radiation to chemical processing are feasible. These techniques have been success- fully demonstrated for both low pressure plasma processes and for high temperature,high pressure batch processes. Interesting and unusual results may be obtained if the techniques and apparatus developed for this study are applied to other types of process feeds. MICROWAVE PROCESSING OF LIGNIN AND CELLULOSE PART I - AN INVESTIGATION OF THE REACTIONS OF LIGNIN AND HYDROGEN USING A CONTINUOUS FLOW MICROWAVE DISCHARGE REACTOR PART II - THE EFFECT OF MICROWAVE RADIATION ON BATCH PROCESSING OF CELLULOSE IN WATER By ‘.;5 Jack D: Junttila A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemical Engineering 1977 To My Parents ii ACKNOWLEDGMENTS The author would like to express his appreciation to his academic advisor, Dr. Martin C. Hawley, for his guidance and assistance. Appreciation is also given to Dr. Jes Asmussen, Jr. for his advice and assistance. The author would also like to acknowledge the valuable assistance of Dr. John H. Cameron, Dr. Raghuveer Mallavarpu, Jim Pelkie and Steve Mertz. The financial support of the Kimberly-Clark Company is grate- fully acknowledged. iii TABLE OF CONTENTS INTRODUCTION PART PART I - AN INVESTIGATION OF THE REACTIONS OF LIGNIN AND HYDROGEN USING A CONTINUOUS FLOW MICROWAVE DISCHARGE REACTOR ABSTRACT LIST OF TABLES . LIST OF FIGURES INTRODUCTION EXPERIMENTAL RESULTS DISCUSSION CONCLUSIONS REFERENCES II - THE EFFECT OF MICROWAVE RADIATION ON BATCH PROCESSING OF CELLULOSE IN WATER . ABSTRACT LIST OF TABLES LIST OF FIGURES INTRODUCTION EXPERIMENTAL RESULTS DISCUSSION iv Page 28 29 31 32 33 35 4O 47 CONCLUSIONS REFERENCES RECOMMENDATIONS FOR FUTURE STUDY REFERENCES Page 49 51 52 54 I. INTRODUCTION Plants, primarily made up of cellulose and lignin, are a potential source of hydrocarbons. Each year large amounts of lignin and cellulose are produced as waste byproducts of the paper-making industry. For example, approximately 12.6 billion pounds of lignin are produced annually as byproduct waste, contained in Kraft Process Black Liquor. A significant quantity of cellulose is also produced as a waste byproduct. Presently, the primary use of these wastes is for fuel. However, as petroleum shortages increase and as prices rise, it may become economical to use cellulose and lignin from plants or from waste byproducts of paper production as sources of hydrocarbons for chemicals. The alternatives of cellulose and lignin are not well developed technologically. Cellulose can be converted to various chemicals by biochemical methods or pyrolysis techniques. Hydrocarbons produced from cellulose would be mainly chain compounds. Lignin is a complex structure containing aromatic rings and is essentially non-biodegrad- able. Thus, pyrolysis and hydrocracking technology are expected to result in the production of both chain and aromatic types of com- pounds. Significant technological development is required in order to place lignin processing and cellulose processing in the category of viable alternatives for producing hydrocarbon feedstocks. How- ever, it is expected that processes that are useful for converting coal to chemicals may also be useful for converting lignin to chemicals. Experiments have been conducted at Michigan State University studying the use of microwaves for catalyzing the decomposition reactions of the paper waste byproducts, lignin and crude cellu- lose. TWo process schemes were experimentally investigated and are described in this thesis: Part I) microwave plasma processing of lignin, and Part II) high pressure, microwave catalyzed, batch processing of cellulose and water. Both of these process schemes involve unique application of microwave technology to paper waste processing, but these techniques may be used for a variety of feeds. The method used for the lignin processing experiments and de- scribed in Part I involved using microwaves for generating a low pressure hydrogen plasma in a quartz tube reactor, and then continu- ously injecting a liquid feed solution containing lignin into this plasma. Lignin used in these experiments was extracted from Kraft Process Black Liquor supplied by the Kimberly Clark Corporation. The apparatus used for the lignin experiments was similar to the apparatus used for studies performed earlier at Michigan State University investi- gating the reactions of H2 and C0 in a microwave plasma]. One goal of this study of lignin processing was to develop lignin-microwave plasma experimental techniques and to compare the results of lignin-microwave experiments to other methods, including pyrolysis. Another major objective of this study was to determine what chemicals may be produced by reacting lignin in a microwave plasma and to determine a feasible microwave plasma processing scheme. Described in Part II are preliminary experiments which have been carried out to study the chemical reactions of cellulose waste in water and the effect of microwave radiation on this re- action. The feed used for these batch experiments was a cellu- lose containing "sludge", which is produced as a waste byproduct by the paper industry, and was supplied by Kimberly-Clark. Ex- periments were conducted at high pressure and temperatures (500 psig, 450°F) and included use of acid and base chemical catalysts. The major objective of this study was to determine if microwave radiation could catalyze the reaction of cellulose and water to volatile low molecular weight organic products. Parts I and II of this Thesis were prepared as manuscripts for publication purposes. Recommendations for future work are con- tained in a separate section following Part II. PART I - AN INVESTIGATION OF THE REACTIONS OF LIGNIN AND HYDROGEN USING A CONTINUOUS FLOW MICROWAVE DISCHARGE REACTOR Paper prepared for publication By Jack D. Junttila, Martin C. Hawley and Jes Asmussen, Jr. Michigan State University Departments of Chemical and Electrical Engineering ABSTRACT AN INVESTIGATION OF THE REACTIONS OF LIGNIN AND HYDROGEN USING A CONTINUOUS FLOW MICROWAVE DISCHARGE REACTOR By Jack D. Junttila, Martin C. Hawley and Jes Asmussen, Jr. Departments of Chemical and Electrical Engineering Michigan State University East Lansing, Michigan The reactions of lignin, contained in Kraft Process Black Liquor, and hydrogen in a continuous-flow microwave discharge reactor have been studied. The goal of this study was to determine what chemicals may be produced by reacting lignin in a microwave plasma and to suggest a feas- ible microwave plasma processing scheme. The reaction products were primarily carbonaceous char and small amounts of guaiacol, 4-hydroxy-3-methoxytoluene, and other methoxy- phenols. These reaction products are similar to those produced in con- ventional pyrolysis. This result indicates that high energy pyrolysis reactions dominate over radical reactions when lignin is injected into a hydrogen microwave plasma. It was concluded that a simple one step process of direct injec- tion of lignin into a hydrogen plasma is not a viable method for ob- taining useful chemicals from lignin. The most appropriate direction for further study of lignin processing should incorporate a two-stage process scheme which combines existing lignin technology with microwave technology. 6 In the conventional hydrocracking of lignin, low value methoxy- phenols are produced and the yield of useful products is low, (less than 10%). The microwave processing of methoxy-phenols has the potential to increase the yield of useful products from lignin to over 50%. The economics of a scheme which combines hydrocracking and micro- wave processing of lignin were examined. These economics are highly speculative, but they indicate that there is incentive for further study. LIST_0F TABLES Table Page 1. Summary of Experimental Conditions and Results . . . . . . . . . 15 2. Economic Summary of Hydrocracking and Microwave Pyrolysis of Lignin . . . . . . . 24 LIST OF FIGURES Figure Page 1. Microwave flow system . . . . . . . . I3 2. Two stage lignin processing . . . . . . . 21 l. Introduction Presently, the petrochemical industry obtains hydrocarbon feed- stocks from petroleum and natural gas. Increasing worldwide demand for energy is depleting the known petroleum and natural gas reserves, and is causing petroleum and natural gas costs to rise rapidly. Thus, other sources of hydrocarbons, such as coal and plant life, must be considered for the long-term as alternatives for the production of hydrocarbon feed- stocks. Coal processing for the production of fuels and chemicals has been studied extensively for many years and is at a fairly advanced stage of development throughout the world.1 Large coal gasification/methanation demonstration plants are under construction. Chemicals can be manu- factured from the synthesis gas produced in the gasification process. Production of acetylene from coal has been studied for a number of years in the United States, but has not been yet commercialized.2 Plants, primarily made up of cellulose and lignin, are also a potential source of hydrocarbons. Each year large amounts of lignin and cellulose are produced as waste byproducts of the paper-making in- dustry. For example, approximately 12.6 billion pounds of lignin are produced annually as byproduct waste, contained in Kraft Process Black Liquor. A significant quantity of cellulose is also produced as a waste byproduct. Presently, the primary use of these wastes is for fuel. However, as petroleum shortages increase and as prices rise, it may become economical to use cellulose and lignin from plants or from waste byproducts of paper production as sources of hydrocarbons for chemicals. l0 The alternatives of cellulose and lignin are not nearly as well developed technologically as coal processing. Cellulose can be con- verted to various chemicals by biochemical methods or pyrolysis tech- niques. Hydrocarbons produced from cellulose would be mainly chain compounds. Lignin is a complex structure containing aromatic rings and is essentially non-biodegradable. Thus, pyrolysis and hydrocrack- ing technology are expected to result in the production of both chain and aromatic types of compounds. Significant technological development is required in order to place lignin processing and cellulose processing in the category of viable alternatives for producing hydrocarbon feed- stocks. However, it is expected that processes that are useful for converting coal to chemicals may also be useful for converting lignin to chemicals. Reactions of coal and benzene in plasmas have been investigated experimentally. Bond, Ladner, and McConnell3 have investigated the reactions of powdered coal in a plasma jet. In their experiments, they obtained yields of 20% acetylene. The pyrolysis of coals in a microwave discharge has been studied by Fu and Blaustein.4 Again, the hydrocarbon product produced in largest quantity was acetylene. A con- tinuous-flow microwave discharge reactor was used to study the reactions of carbon monoxide and hydrogen to methane and acetylene.5 This re- action is of importance for coal processing schemes being considered whereby coal is first gasified to carbon monoxide and hydrogen; these reactants can in turn be used to produce a variety of hydrocarbons. Brooks and Sambrook6 obtained high yields of acetylene and butadiene in experiments in which they studied the reactions of benzene in a ll continuous-flow microwave discharge reactor. They reported that the reactions occurring in the microwave discharge are quite different from the reactions of benzene which take place in a lower frequency or d.c. reactor. Results of experiments of coal and benzene in microwave plasmas suggest that plasma technology be investigated for converting lignin to useful hydrocarbon products. The source of lignin for the experimental studies reported in this paper was Kraft Process Black Liquor. Kraft Process Black Liquor is produced as a byproduct of the paper and pulp industries. It typically consists of 28% lignin, l4% Na2804, and the balance is fixed solids and other residue. Experiments were conducted in which black liquor was continuously injected into a hydrogen plasma induced by a microwave discharge. The goal of this study was to develop lignin-microwave plasma experimental techniques and to compare the results of lignin-microwave experiments to other methods, including pyrolysis. There were two experimental objectives. First, it was necessary to determine if a microwave dis- charge reactor could be operated continuously when using a combination of liquid and gaseous feeds. It is relatively easy to operate a micro- Wave discharge reactor continuously when using gas feeds, but it was not known how liquid feed would affect reactor operation. A second objective of this study was the determination of the product distribution when lignin and hydrogen react in the microwave discharge. 2. Experimental A schematic flow diagram of the experimental apparatus used in this study is presented in Figure l. The experimental reactor system consisted of a vertically mounted quartz reactor tube which passed through two cylindrical microwave cavities in series. The quartz tube was 4 feet long and had an outside diameter of 25 mm. This tube was cooled by an air stream directed into the cavity. A specially designed gas-liquid feed nozzle was used to mix and inject the lignin solution and hydrogen into the reactor. Feed rates of lignin solution and hydrogen were measured and controlled using flowmeters. TWo cold traps in series were located at the reactor outlet. Product samples were collected in the first cold trap, which was cooled with ice water. and the second trap, which was cooled with liquid nitrogen. A vacuum pump connected to the second cold trap was used to evacuate the system and to control the experimental pressure between 10 and 60 mm Hg. The microwave apparatus includes two microwave sources coupled to two variable-length cylindrical cavities by a system of waveguides. The two cavities can be operated to produce one long reaction zone or two separate reaction zones. The cavities are water cooled by externally soldered cooling coils. The two microwave sources operate at a fre- quency of 2.45 GHz and maximum power levels of l500 and 600 watts. Only 50% of the input power is absorbed by the reactants. Heat produced by reflected power is removed by a water-cooled heat exchanger. Incident and reflected power levels are measured with power meters to determine absorbed power. .The microwave system has been used for experiments with gaseous feeds and has been described in detail previously.5’7 12 13 .amummm 30: 96383: .H muswfim £503 mun—Eda can» :02 vvuu 0.37:: umBOQ wCuu van £33.05 AE—a Eanum> oh _ = _ ‘ uvquocmz Amp cowumunmgmu noqumuom . no." “53 avenue. I'V J .2an >E§Q 9339—35 havoc—~30." j «Edevn ACNE .3 .u< .NZ .va can; wrung can $2353 , _. . . < 1 j 382 IVA oofigmu . o>m30uum2 W A 33.50.30 A 323 youuonnm comm Envfi\am0 . uaod nouwwuomwou >550 head: xumum 23> uwmwwfluou «.03 «.335 .3309 2 102m vouuvavu vcm «p.033: n3 condunfimu u 3053.5“ T4 The experimental procedure developed for investigation of the re- actions of lignin in a microwave plasma basically consisted of: (l) generation of a hydrogen plasma at l mm Hg (2) increasing the pres- sure to 20 mm Hg or more while maintaining the plasma, (3) injection of the lignin solution mixed with hydrogen, (4) adjustment of vacuum pumping rate and inlet flow rates for pressure control, and (5) col; lection of products for chemical analysis. The products which condensed in the ice water cold trap were analyzed using gas chromatography. A TO% Carbowax column was used for this purpose. Mass spectrometry was used to analyze the products which condensed in the liquid nitrogen trap. The solid products were not chemically analyzed, but were collected for estimating mass balances. 3. Results The experimental conditions and results of the investigation of the reactions of black liquor and hydrogen in the microwave plasma reactor are summarized in Table l. Lignin in a black liquor solution was used as the feed for runs l-3, whereas lignin in methanol was used for the remaining runs. It was discovered that using black liquor directly resulted in tube degradation due to deposition of sodium salts. Extracting the lignin from black liquor with methanol and utilizing this solution of lignin and methanol as the feed to the plasma eliminated the problem of reactor tube failure. Several observations were made during runs when aqueous solutions of lignin in black liquor were fed to the reactor. As this liquid feed first entered the reaction zone, the color of the plasma changed from a reddish color to a yellowish color. A few seconds later, the quartz reactor tube would begin to glow red hot and the experiment would be halted to prevent tube failure. When the tube was examined after an experimental run, the inner surface of the reactor tube was found to be coated with a black soot. This soot could be easily washed off, but there were also areas where the tube itself had started to de- compose and had become opaque and brittle. It was speculated that the presence of inorganic salts in the black liquor was largely responsible for the degradation of the reactor tube. When methanol was utilized for extraction of lignin from black liquor, the problem of tube failure was eliminated. Approximately 67% by weight of the black liquor dissolved in the methanol. Apparently, 15 6 1 .oauHmau Hague no .Somnmz "SH .avHHoa vuxHH Hoe .chwHH Ham ”HaaHuvv CoaaHH HaaHm Ho coHUHuoasoua .80 NH I HH one» ..80 «Huh I H anon .vouammma uo: umaon vonuouam vca Aucmum:00v 30w¢ I pagan unovHUCH mucoaHuuaxo HHa :H amuu vHou ammouuH=_vH=vHH osu :H vc20u mama ocmzuoa no muczoam uanHm .voumuuao mm: H ucou mammHm cho A "mucou acuumuu mo numCuH "HHucon nauummu cH .oucouumou wow wow: xmma Hocona noumm mxmoa N .cowuummu cu mac xmua Hoaosa .vmou mo Nun aN ow< me w.e .nuoa ca uoc m .uwnu :uwa voHHHu o: .xmwa Hocwauws unopw wcuucommumou ucon nauuwmm umumm mxmma uu mH .kuavoun umsu mm ooc me Nae .nuua :« Roe w .umzo .xmma Hocona oc .ooow mo Nnn ac nomN oq o.NH .auoa cw non N now: onHHu .xmma Hocmcuos mcuucmmouawu ocou neuomwm .3un 933 on MH .vuusvoun .55 ON 93 NN o.m :32: a.“ non o .unnu sud: vaHNm moaouon econ .uouum .333.» .9530: HHS—mom N 3:39 comm no new NN con 3 .OH .52: ca NOH m. nuu>o mASu uanHm Hocmnuma ouONon wawucmmoumou .mcHxaaam agom axaaa co m nauseoua wasp mm ooq HN o.w .nuaa cH ”OH H .aaeHu :OHuaauu “mums mH AN own on o.~ on aH «OH H uuosm .wcHumon uco>Hom can: Iuo>o can :oHu com: uo: uu muHmomov copumo wH ooe oe N.m 0N: cu NoH N Imvmuwwv moss . mH cog HN H.n o~= aH «OH H acoauu>uomno oHaEmm vaanH muusvoum vHHom mm as ASV H econ oom\uu :H8\ou aoHumuucoucou cam mo mammHmcm whammoun nosba mum» 30H“ many aon uauu you .u3 00 mo ouHsmom Hauomux conuomn< N: vanvaq muosuua xuon muHsmom can muowuuvcoo Hmuaoaauoaxm mo humaabm .H oHpmH I7 much of the black liquor salts (and some of the lignin) did not dissolve into the methanol because when this liquid feed was injected into the hydrogen plasma the problem of tube degradation did not occur. As before, when this liquid feed entered the reaction zone, the plasma would change color from red to yellow. As more feed reacted, a powdery black char would collect on the inner surface of the reactor tube. Occasionally, the char near the tube surface would emit sparks. This char was capable of absorbing large amounts of microwave power and the char in contact with the surface of the reactor tube would overheat the reactor tube. Despite the occasional problems with sparking, use of methanol as solvent allowed experiments to be carried out success- fully. Experiments were conducted with feeds of 10, 30, and 40% solutions of black liquor in methanol. The primary constituents of these black liquor solutions were lignin and fixed solids. In all of these experi- ments, the results were similar. The product produced in largest quantity was a solid char. Although difficult to measure, this char represented approximately 75% of the liquid feed. The reaction products were collected in the ice water cold trap and in the liquid nitrogen trap, and analyzed for chemical composition. The reaction products collected in the ice water cold trap were analyzed using gas chromatography. Thirteen different species were contained in this sample. Eleven_of these species were produced in very small quantities and show up in gas chromatograph analysis as GC peaks after a methanol peak and before a phenol peak. Larger quantities of two other species were found in the liquid product and a comparison of the chromatogram for the products of these microwave experiments was made l8 with a chromatogram of products of conventional pyrolysis reported by Watanabe and Kitaos. This comparison indicated that guaiacol and 4-hydroxy-3-methoxytoluene were produced in the microwave reactor. No phenol was produced in the microwave discharge, although phenol was used as a reference peak CG runs. The contents of the liquid nitrogen trap were analyzed by mass spectrometry and a small amount of methane was detected. A run was conducted with pure methanol feed, and the material collected in the liquid nitrogen trap was found to be methane. Therefore, methane formation was due, at least in part, to hydrogenation of methanol. 4. Discussion Complex reactions occur in a hydrogen discharge and several reactive species are available for reaction: hydrogen atoms, ionic species, and electrons. Vastola, Walker and Wightman9 studied the reactivity of these various species with carbon in a hydrogen dis- charge. They determined that hydrogen atoms were relatively non- reactive when they collided with a carbon surface. Hydrogen atoms recombine on the carbon surface instead of reacting with the solid phase carbon. They also theorized that reactions of carbon with hydrogen which do occur in the discharge were due to the action of high energy ionic species. The collisions of these energetic ionic species with the solid carbon produces gaseous carbon which can react with hydrogen. Char formation is the major reaction which occurs when lignin is injected into a hydrogen plasma. It is suspected that reactions of hydrogen atoms with lignin are of little significance in the discharge. Collisions of hydrogen atoms with lignin probably result in recombin- ation of hydrogen. It is likely that the large production rate of char in the discharge is due to the reactions of ionic species with lignin. The bombardment of lignin molecules by high energy ionic species probably strips the lignin of hydrogen and converts the lignin to char. 4.1 Alternative Methods of Processing Lignin A spectrum of products, which have no commercial use presently were produced in experiments when lignin was fed directly into a microwave plasma reactor. Lignin processing to produce chemicals may become of commercial significance if either a) chemical routes are developed I9 20 utilizing the various presently non-useful products, or b) alternative processes are used to convert lignin to present day usable products. A microwave discharge reactor could be useful if the lignin was -first processed using conventional methods. A study of existing lignin technology shows that lignin can be depolymerized without producing large amounts of char10’11. Controlled oxidation of lignin to vanillin and other products10 is a process used commercially, and a high pressure hydrogenation process which converts lignin to a variety of phenols has been patentedl]. These processes successfully depolymerize thelignin but they produce such a wide variety of products so that it is impossible to fully utilize the lignin feed. What is needed for economically successful processing of lignin, is some method of converting a wide variety of single ring aromatic compounds (mainly methoxy-phenols) to a select few useful products. Re- sults of experiments reported by Brooks and Sambrook6 suggest that use of a microwave discharge reactor may fulfill the above requirements. They studied the reactions of benzene in a microwave discharge reactbr and obtained large conversions of benzene to acetylene and butadiene. If it is possible to sustain a microwave discharge when the reaction feed is a mixture of methoxy-phenols, useful products may be formed in the discharge, and use of microwave processing could be the critical compo- nent of an overall scheme for converting lignin to useful hydrocarbons. 4.2 Two Stage Lignin Processing Approximately 200 million pounds per year of lignin are available from a typically~sized Kraft pulp mill. A two-stage processing scheme for converting 200 million pounds per year of lignin to useful products is illustrated in Figure 2. This scheme provides a basis for evaluating 21 .wcwmmuuoun caanH mwmum 039 .N unawam AITII. .u%\.AH :2 m.Nm «3363* T :53 2: RR mHofifiAxafiaz 4 5.3 :2 mom a .855 .omHm ‘muuws.woa Hm I umaoq .ome mcowu _ \ heuummm “CA—H z: nNe ocmvauam sauna—mm mammHm mHocmsmlbnofimz o>mzou0Hz I .um\.AH a: N.mN mcloumg IIWII III its: x: 2 263% m AI 2 .uééH 2: 8S 2538 o H £34: 2: RH aHoaozauasfimz w m .AII < Abs—H z: 93 H255 m m AIFIII m big: :2 93 Hoaaaoum A EH}: :2 H: H823: . HN\=HH =OH x m.m u a ooooe ‘lll Hma 000.0H umxomuoouvhm whammuum swam cHamHH 03:: .2 com chaHchou coauSHom :HGNHH (I 39¢QO z: 93 N: 22 the economic feasibility of including microwave plasma reactors as part of a two-stage lignin process. In the first stage, it is speculated that lignin could be hydrocracked to mainly low value methoxy- 11’12. Useful products which phenols using the process patented by Giesen may also be produced in the hydrocracking stage include phenol, propanol, and methanol. The methoxy-phenols would be fed to the second stage, a microwave plasma reactor, where the low value methoxy-phenols can be possibly upgraded to useful products, such as acetylene and butadiene. It is suspected that a microwave plasma may be unique in its ability to convert low value monocyclic aromatics to useful products. Several assumptions were required for this analysis because of a lack of directly related experimental information. However, the assump- tions for yield, product distribution, and energy requirements for each of the stages represent extrapolation of results of experiments closely related to each process. Extrapolation of published results of lignin 8’n’nprovided the information needed for designing the hydrocracking hydrocracking step. Lignin was assumed to be converted to 65% (by weight) methoxy-phenols, l9% residue, 8% phenol, 5% propanol, and 3% methanol. The yields of useful products (acetylene and butadiene) pro- duced in the second stage, were estimated by extrapolating the results of experiments with benzene in a microwave plasma performed by Brooks and Sambrook6. Thus, it was assumed that the molar conversion of the aromatic ring part of the feed to the microwave plasma system was 30% to butadiene, 20% to acetylene, 10% to residue, and the remaining feed did not react. It was also assumed that the side chain methoxy groups would be hydrogenated to methanol in the discharge with a molar conver- sion of less than 20%. The microwave energy requirement was estimated 23 to be 690,000 J per gram mole of aromatic feed, and was based on the work on benzene by Brooks and Sambrook. Results of an economic analysis based on these assumptions is con- tained in Table 2 which indicates that over $18 million for capital investment for a plant can be justified. These estimates were based on 1976 values for chemicals. The economics of this process are highly sensitive to yields of lignin to useful products and energy requirements. Microwave plasma processing appears to be most useful as part of an overall process to upgrade low value products to valuable products. These products would be, in general, different than those obtained from conventional pyrolysis and/or hydrocracking. It is important in micro- wave processing schemes to maximize the microwave reaction efficiency. This would minimize the capital investment and energy cost for a micro- wave generation plant. Many assumptions were made for this test case, but the results of the economic analysis indicate that application of microwave technology to lignin processing may be significant. Economic Summary for Hydrocracking and Microwave 24 Table 2 Pyrolysis ongignin Lignin: 200 MMllb./yr. Useful products: 110 MM lb./yr. Conversion: 55% Maximum Total Investment for 20% DCF: Raw Materials Unit Lignin I5. Hydrogen Gas MSCF Subtotal Utilities Fuel MM BTU Electricity M kWh Steam MM BTU Cooling water M gal. Subtotal Fixed Costs Labor and'Miscellaneous (20% of Investment) Depreciation (10% of Investment) Subtotal Total Manufacturing_Cost Revenue Methanol lb. Propanol lb. Phenol lb. Acetylene lb. Butadiene lb. Char and other lb. Subtotal Before Tax Profit Net Profit Annual Cash Flow 1 MM = million Quantity/ Year 200 MM .830 MM .851 MM .263 MM .226 MM 13.0 MM 10.0 MM 15.0 MM 29.2 MM 42.5 MM 52.3 MM $18,l00,000 gost gost /Unit / ear .0? ‘SUTPUDTUUU' "*0" TRASH 2.000 l,700,000 20.000 5,300,000 l.310 300,000 .050 - 773m 3,600,000 l,800,000 5,400,000 .07 900,000 .25 2,500,000 .28 4,200,000 .l8 5,300,000 .20 8,500,000 .02 l,000,000 5,000,000 2,500,000 4,300,000 5. Conclusions Experimental results show that lignin is degraded to char when lignin or black liquor is injected into a hydrogen plasma induced by a microwave discharge. The desired result of hydrogenation and de- polymerization of lignin in the discharge does not occur. Actually, the collisions of high energy ionic species with lignin dehydrogenate the lignin and overwhelm any radical reactions which may occur. There- fore, use of a one-step process involving a microwave plasma reactor is not applicable for production of commercially useful chemical products "; from lignin. A two-step process utilizing conventional hydrocracking to depolymerize lignin, followed by microwave plasma cracking of methoxy- phenols (to convert these monocyclic aromatics to acetylene and buta- diene, for example) may be a practical processing scheme. The results of a preliminary economic analysis show that use of microwave plasma processing as part of a two-stage system for converting lignin to useful chemicals might be economically feasible. Based on the results of this study, it is possible to identify the most important problems related to lignin processing which should be investigated fur- ther. The reactions of methoxy-phenols in a microwave discharge need to be investigated experimentally in order to determine the nature and quantities of reaction products. It is expected, however, that acety- lene and butadiene will be among the predominant products. Carrying out these additional experiments will provide the basic information needed to adequately design a system for processing the lignin and will permit a realistic evaluation of the process. 25 26 ACKNOWLEDGEMENT We are grateful to the Kimberly-Clark Corporation for financial assistance with this work. 10. 11. 12. 27 References White, P.C.; Zahradnik, R.L.; Cochran, N.P. Quarterly Report, Office of Fossil Energy, ERDA, July, 1975. Office of Coal Research. Annual Report. U. S. Government . Printing Office, Washington, 0.0. 1972, 26. Bond, R.L.; Ladner, W.R.; McConnel, G. Fuel, 1966, 45, 381. Pu, Y.C.; Blaustein, 8.0.; Wender, I. Chem. Engng. Frog, Symp; Sgr, 1971, 61(112), 47. Mertz, S.F.; Asmussen, J.; Hawley, M.C. IEEE Transactions on Plasma Science. 1974, 23 297. Brooks, B.W.; Sambrook, R.M. J. Appl. Chem. Biotechnol. 1972, 22, 9. Asmussen, J.; Mallavarpu, R.; Hamann, J.R.; Park, H.C. Proc. IEEE. 1974, 62, 109. Watanabe, Y.; Kitao, K. Wood Res., Kyoto. 1966, 38, 40. Vastola, F.J.; Walker, P.L.; Wightman, J.P. Carbon. 1963, l 11. Craig, 0.; Logan, C.D. U.S. Patent 3,054,659. September 18, 1962. Giesen, J. U.S. Patent 2,870,133. January 20, 1959. Sittig, M. Organic Chemical Process Encyglopedia. Noyes Develop- ment Corporation, Park Ridge, New Jersey. 1969, 2nd edition, 521. PART II - THE EFFECT OF MICROWAVE RADIATION ON BATCH PROCESSING OF CELLULOSE IN WATER Paper prepared for publication By Jack D. Junttila, Martin C. Hawley and Jes Asmussen, Jr. Michigan State University Departments of Chemical and Electrical Engineering 28 ABSTRACT THE EFFECT OF MICROWAVE RADIATION ON BATCH PROCESSING OF CELLULOSE IN WATER By Jack Junttila, Martin C. Hawley and Jes Asmussen, Jr. Departments of Chemical and Electrical Engineering Michigan State University East Lansing, Michigan 48824 A batch reactor system has been designed and constructed for investigating microwave radiation effects on gas or liquid phase reactions at high temperature and high pressure. Several batch ex- periments were conducted in this specially constructed reactor with cellulose in water. A fifteen percent solution of crude cellulose waste in water was placed in a high-pressure reactor vessel and heated to 450°F and a pressure of 500 psig. In some experiments microwave radiation was injected into the reactor interior to study the effects of microwave catalysis. Samples of both liquid and gas products were obtained for analysis. Reaction products were ana- lyzed using mass spectrometry of gas products and distillation of liquid products. The objectives of this study were to determine if microwave radiation would catalyze the hydrolysis reactions of cellulose and to determine the yield of reaction products for various experimental 29 30 conditions. Of particular interest was the yield of volatile, low- molecular weight organicproducts, and the effect of microwaves on their production. Several reaction products were identified in the experimental gas sample. Products present in large quantities were CO2 and C0. Also present in small quantities were methane, formic acid, and acetaldehyde. The liquid product samples consisted of two main reaction products. The major product found in the liquid sample was a watfr soluble, black sticky tar. This tar probably consisted of glucose and glucose decomposition products. Formic acid in small quantities was also identified. Based on the results of this study, it was concluded that microwave radiation has no positive effect on the hydroloysis re- actions of cellulose. The decomposition of cellulose seems to occur via the same reaction mechanism, cellulose+ glucose» decompo- sition products, with or without use of microwaves. 31 LIST OF TABLES Table Page 1. Summary of Experimental Program . . . . . 38 2. Identification of Mass Spectrosc0pic Peaks . . 43 32 LIST OF FIGURES Figure 1.. High Pressure Microwave Batch Reactor System . . . . 2. Mass Spectrogram of gas sample, run 4 - no catalysis . . . . . 3. Mass spectrogram of gas sample, run 10 - microwave and NaOH catalysis . 4. Reactor pressure as a function of temperature . . Page 36 41 42 45 1. INTRODUCTION Recent investigations in our laboratory of microwave plasma catalysis for chemical reactions have suggested the use of micro- wave processing of cellulose for the production of chemicals. De- tails of experiments on microwave processing of cellulose are re- ported in this paper. The use of cellulose and cellulose wastes as chemical, food, and energy resources has been widely studied.1 Many processes have been considered for converting cellulose to useful products: thermal pyrolysis, enzymatic hydrolysis, fermentation, high pressure water 1.2.3 Chemicals enzymatic, microbial. processing, and many others. and macrobial catalysts and their effects on cellulose processing have also been investigated.2 Products which have been produced in cellulose decomposition processes include glucose, char, tar, levo- glucosan, carbon black, organic acids, alcohols, aldehydes and ketones. Synthesis gas (C0 and H2) can be produced by partial oxidation of cellulose; this synthesis gas can in turn be used to produce a variety of chemical products]. Although a substantial amount of research has been performed investigating the problem of cellulose processing, most schemes for converting cellulose or cellulose waste to useful products are not yet economically feasible. 33 34 The paper and pulp industries produce large quantities of solid waste containing cellulose. Currently, the primary use of this waste product is for fuel. However, as petroleum shortages increase and prices rise, it may become economical to use cellu- lose from plants or from waste byproducts of paper production as sources of hydrocarbons for chemicals. The Kimberly-Clark Corporation has supported exploratory ex- periments at Michigan State University to study the processing of cellulose waste for the production of chemicals. This report summa- rizes the exploratory experiments which have been carried out to study the chemical reactions of cellulose waste in water and the effect of microwave radiation on these reactions. The feed used for these batch experiments was a cellulose containing "sludge", which is produced as a waste byproduct by the paper industry, and was supplied by Kimberly-Clark. Experiments were conducted at high pressure and temperatures (500 psig, 450°F) and included use of acid and base chemi- cal catalysts. The major objective of this study was to determine if microwave radiation could catalyze the reaction of cellulose and water to volatile low molecular weight organic products. 2. EXPERIMENTAL A batch system as shown in Figure l was used for the experimental studies. The pressure vessel used had a l-inch thick cylindrical shell. Removable flat circular plates 2 inches thick were used for the head and bottom of the vessel and were bolted down for high pressure operation. The vessel head was tapped to allow attachment of the sampling valve, pressure gauge, and the microwave coaxial coupling. The microwave coupling consisted of a variable length brass microwave probe inserted in a conical piece of Teflon. This allowed introduction of microwave radiation to the reactants in the interior of the vessel while at the same time permitting high pressure operation. The shell of the vessel was tapped to permit insertion of a thermocouple lead for temperature measurement. The heavy pressure vessel (approximately 200 pounds) was supported on a stack of cement blocks and direct fired heating was used for supplying the large amount of heat needed to bring the vessel and the reactants up to operating temperature. Fiber- glass insulation was wrapped around the vessel to decrease heat losses during experimental runs. A microwave source operating at a frequency of 2.45 GHz and a power level of 1,500 Watts generated the microwave radiation used in most experiments. The radiation generated at the microwave source passed through a series of wave guides which transmitted the radiation to the coaxial coupling connected to the vessel head of the reactor. The microwave radiation could then pass through the coupling into the reactor interior. Incident and reflected microwave power levels were measured with power meters. The microwave system has been used for 35 36 microwave (Ei:) coaxial sampling valve coupling afii>wzogu?s - op c3; .mpaEmm mom mo Emgmoguuwam mmmz .m mgzmwu Np «F up mp mm mm mm mm mm cc ‘ Ne «w _ : iii. A e :o moo La Nzfi N cu o N o :5 N8 om: Nz ou 43 TABLE 2 Identification of Mass Spectroscopic Peaks Chemica1 Species Mass Spec Peaks (descending_orde:)* C02 44, 28, 16, 22, 45 CO 28, 12, 16, 29, 14 O2 32, 16 N2 28, 14, 29, 14 H20 18, 17, 16, 19, 20 Ar 40, 20, 36, 38 CH4 16, 15, 14, 13, 17, 12 Formic acid 29, 46, 45, 28, 17 Aceta1dehyde 29, 44, 43, 42, 26 Propane 29, 28, 27, 44, 43 Propene 41, 42, 39, 27, 40 Dimethy1 ether 45, 29, 15, 46, 14 HC1 36, 38, 35, 37, 76 Methano1 31, 32, 29, 28, 18 Ethano1 31, 45, 29, 27, 43 3 *Information from At1as of Mass Spectra1 Data . '1: 44 sma11 quantities were methane, formic acid, aceta1ydehyde, and possib1y propene. Unexpected1y, there was no indication of methano1 or ethano1 in the gas samp1e. The presence of 02 and N2 in the gas samp1e, as shown in the mass spectrograms, probab1y was due to some 1eakage of air into the gas samp1e tube. The reactant pressure-temperature re1ationship for run 10 is shown in Figure 4. A1so shown in this figure is a comparison of the reactor pressure to the vapor pressure of water. The 1argest difference between the vapor pressure of water and the reactor pressure occurred at a temperature of 450°F. This difference in pressure was about 70 psi and indicates that a significant amount of gaseous product (main1y C02) was being produced. After an experimenta1 run when the vesse1 cover was first re- moved, severa1 pre1iminary observations cou1d be made. First, the reactor contents had a pronounced smoky odor. Usua11y there was a very thin 1ayer of f1at b1ack crysta11ine materia1 f1oating on the water surface. The 1iquid product was b1ack in co1or and had a viscosity that was s1ight1y greater than water. This b1ack product so1ution was produced whether acid, base or no cata1yst was used. It was easi1y observab1e that the ce11u1ose had undergone significant reaction. For examp1e, with the Type-1 ce11u1ose, which was very difficu1t to grind up, it was discovered that after reaction it was easi1y pu1verized. The Type-2 ce11u1ose seems to change co1or with reaction, as it was grey before reacting and brown after. Disti11ation was used for characterization of the aqueous 1i- quid product. It was expected that if there were significant quanti- Pressure (psia) 500 45 0; Pressure of Reactor Contents 400J a c, 300, . Vapor Pressure of water (7 200. o a, 100. ' to ¢;”’,iV 0 _ : gm 200 300 400 500 Figure 4. Temperature (0F) Reactor pressure as a function of temperature. 45 ties of any vo1ati1e organic product, it wou1d be possib1e to sepa- rate and identify these products using a batch sti11. Disti11ation of the fi1tered 1iquid product indicated that a11 vapor which boi1s off from the 1iquid product boi1s at about 100°C at 1 atmOSphere pressure. However, the first overhead out had a heavy odor and was definite1y not pure water. This behavior was due to the presence of formic acid, which a1so boi1s at 100°C. There was no indication of the presence of significant quantities of any other vo1ati1e or- ganic products in the 1iquid samp1e. After a11 vo1ati1e products were evaporated off during disti11- ation, what remained was a sticky, b1ack, sweet-sme11ing tar. This water so1ub1e materia1 was the product produced in 1argest quantity; approximate1y 30 to 40% of the ce11u1ose waste feed was converted to this tar. It is suspected that this tar consisted of g1ucose, 1evo- g1ucosan, and other carbohydrates. G1ucose and g1ucose decomposition products were reported to be obtained from high temperature ce11u1ose hydro1ysis reactionss. Levog1ucosan was reported as being the tar produced in ce11u1ose pyro1ysis experimentsz. The presence of sugars in the tar was a1so indicated by the rapid growth of mo1d on the air- exposed surface of the 1iquid product. Attempts were made to ana1yze the product tar using mass spectrometryz, but because of the presence of 1arge amounts of associated water, use of mass spectrometry was impossib1e. 4. DISCUSSION Exp1oratory experiments were conducted on the decomposition of ce11u1ose in water at high temperatures (480 to 660°F) by Appe11, Wender, and Mi11er at the Bureau of Miness. Their operating pro- cedure consisted of: (a) charging the reactor with a ce11u1ose s1urry and reactant gas, either H2 or C0, (b) heating to operating temperature, (c) maintaining this temperature for two hours, and (d) then quenching the reactor. They obtained a yie1d of about 40% char. 5 investigated the reaction mechanism for the hydro1ysis of Saeman ce11u1ose in acid mediums at temperatures ranging from 447 to 4630K. He determined that the decomposition of ce11u1ose proceeded via two steps: ce11u1ose + g1ucose 9 decomposition products. The reaction: ce11u1ose + decomposition products was found to be of 1itt1e importance. Amin, Reid, and Mode113 investigated the decomposition of g1ucose in an aqueous phase. They discovered that at temperatures be1ow the critica1 point of water, g1uc1ose decomposed to a sticky b1ack char, 1iquid products, and very 1itt1e gas. When they compared the compo- sition of the char they obtained to the char produced in the experi- ments performed by Appe11 (et a1), they found that the two chars were simi1ar. They conc1uded that ce11u1ose and g1ucose decompose to simi1ar products, and that the Saeman mechanism for ce11u1ose decomposition appeared to be correct. The most interesting resu1t of their study occurred when they injected g1ucose into water at its critica1 point; no so1id char was produced and gas production great1y increased. 47 48 The resu1ts obtained in this experimenta1 study of the re- actions of ce11u1ose waste with water were quite simi1ar to the 3’5’6. The reaction product produced in resu1ts obtained by others 1argest quantity was a b1ack, sticky, water so1ub1e tar. The re- action mechanism hypothesized by Saeman and Amin (et a1) for the decomposition of ce11u1ose a1so seems to correct1y describe the decomposition of ce11u1ose waste in water. The ce11u1ose waste first depo1ymerizes to g1ucose, then the g1ucose decomposes to char or tar, and formic acid. Use of microwave radiation had no positive effect on the re- actions investigated. It was determined that microwave radiation did not promote production of vo1ati1e organic products from ce11u- 1ose. The product species and the product yie1ds were the same for a11 runs, regard1ess of the type of cata1ysis used-~acid, base, or microwave radiation. However, it may be possib1e that if the microwave power 1eve1s were great1y increased, 1ow mo1ecu1ar weight organic products wou1d be formed. Microwave radiation is probab1y great1y absorbed by the 1arge amounts of water present and converted to heat before the radiation can interact with the ce11u1ose. Thus, an activation of the ce11u1ose by microwave radiation wou1d be great1y reduced in the presence of 1iquid water. Interesting resu1ts may possib1y be ob- tained if ce11u1ose were injected into high pressure water vapor irradiated with microwaves, especia11y if the water were at its critica1 point, as in the experiments by Amin3 (et a1). 5. CONCLUSIONS Microwave radiation appears to have no cata1ytic effect on the 1iquid phase reactions of ce11u1ose with water. G1ucose and g1ucose chars are the reaction products produced in greatest quanti- ty when ce11u1ose and water react at conditions be1ow the critica1 point of water. The high production rates of these 1ow-va1ue pro- ducts probably ru1es out the use of high-pressure water-ce11u1ose reactions for the production of chemica1s from ce11u1ose. This study has demonstrated that it is possib1e to investigate the effects of microwave radiation on high temperature and high pressure reactions. A1though use of microwaves was ineffective for ce11u1ose-water reactions, the effect of microwave radiation on other high-presSure and temperature chemica1 reactions may be of interest. 49 50 ACKNOWLEDGEMENT We are gratefu1 to the Kimber1y-C1ark Corporation for financia1 assistance with this work. 51 REFERENCES Wi1kie, C. R. Ce11u1ose as a Chemica1 And Energy Resource. John Ni1ey and Sons, New York. 1975, 321. Madorsky, S. L., Therma1 Degradation of Dr anic Po1ymers. John Ni1ey and Sons, New York. 1964, 2 9. Amin, 5.; Reid, R. C.; Mode11, M. "Reforming and Decomposition of G1ucose in an Aqueous Phase," presented at the Inter- society Conference on Environmenta1 Systems, Ju1y 21-24, 1975. Asmussen, J.; Ma11avarpu, R.; Hamann, J. R.; Park, H. C. Proc. IEEE. 1974, 62 109. Stenhagen, E.; Abrahamsson, 8.; Mc1afferty, F. w. At1as of Mass Spectra1 Data. Interscience, New York. 196 Saeman, J. F. Ind. Eng. Chem. 1945, 37, 43. Appe11, H. R.; Wender, I.; Mi11en, R. 0., Technica1 Progress Report. Bureau of Mines, May, 1970, 25. RECOMMENDATIONS FOR FUTURE STUDY Severa1 genera1 areas of research re1ated to microwave process- ing were suggested during the course of this study. These genera1 t0pics are out1ined be1ow a1ong with recommendations for improving experimenta1 procedures: 1. Studies shou1d be made to determine the nature of products obtained when 1ignin is processed in a high pressure hydrocracker. The aromatic methoxy-pheno1s produced by the hydrocracker may possib1y be upgraded to usefu1 products with use of a microwave p1asma. 2. Experiments shou1d be performed investigating the possi- bi1ity of generating a methoxy-pheno1 (e.g. guaiaco1) microwave p1asma. A1though a carrier gas (H2 or He) may be necessary for these experiments, it wou1d be preferab1e if no carrier gas were used. In past experiments with benzene feeds in a microwave p1asma, the most usefu1 resu1ts occurred when no carrier gas was usedz. 3. The decomposition reactions of ce11u1ose in water shou1d be studied at conditions at the critica1 point of water. Inter- esting resu1ts were obtained by Amin3 (et a1) when they studied the decomposition of g1ucose in water at the critica1 point. It is im- portant in these experiments that the ce11u1ose or g1ucose be in- jected into the water after the water has been heated to its critica1 point. 4. The apparatus for two microwave process schemes has been demonstrated to be experimenta11y viab1e: microwave p1asma pro- 52 53 cessing and microwave radiation cata1ysis of high pressure and temperature reactions. A1though these schemes are probab1y not usefu1 for paper waste processing, they may be usefu1 for other app1ications. A microwave p1asma cou1d be usefu1 for processing viny1 ch1oride or su1fur oxides, and microwaves may have a cata- 1ytic effect on high pressure vapor phase reactions. 5. The experimenta1 procedures and apparatus cou1d be improved -E’ for future experiments. There are better methods4 for removing 1ig- nin from Kraft Process B1ack Liquor than the methods that were used for this study. A better vacuum system is needed for the experimenta1 p1asma system, especia11y when studying a methoxy-pheno1 p1asma. Prob1ems were encountered during pre1iminary experiments with benzene and viny1 ch1oride because of incomp1ete air remova1 by the vacuum pumps. Fina11y, improvements shou1d be made in the reactor feed sys- tem for experiments in which it is necessary to vaporize a 1iquid feed. This may require use of a sma11 boi1er type heat exchanger and a source of steam, but this wou1d great1y faci1itate vaporized feed experiments. 54 REFERENCES Mertz, S. F.; Asmussen, J.; Haw1ey, M. C. IEEE Transactions on P1asma Science. 1974, g, 297. Brooks, 8. w.; Sambrook, R. M. J. App1. Chem. Biotechno1. 1972, 22, 9. Amin, S.; Reid, R. C.; Mode11, M. "Reforming and Decompo- sition of G1ucose in an Aqueous Phase," presented at the Intersociety Conference on Environmenta1 Systems, Ju1y 21-24, 1975. Wha1en, D. M. 13221; 1975, 5, 110.