I" ‘11 ‘9 99:“; ; r3111 1:91“ . 9‘; "1'? g 931? I 9‘ 1; 7-!" 2.1- 133. ‘4" m f‘ 9.. :F 7 .~ 55:; {a7 7 .L R 1.. X .25}; 3! .r’é'rra ‘4 H “I ‘ {‘1 :ij q ‘2. 91.19 3*» r: ‘9; , , 5.3,; 9'9 19.31%.“ .2911 31.3, 7 . 531193.94 .11": 191'191‘2151hqw‘ ”,1, h- “1.11.19. 1‘9 .9 ‘9 9 99 -9§‘9 9. 7' :‘731 ' 1 9 “j ”:12?! 333”?“ . W 911‘; 9%; , 1 ; :1)? r' ,.-. W 9% . 5‘ 1 11‘5”." pd. 35? ‘32 a... 9 "4-1:. a“? , l 1’ )9 42429.5? ‘ “‘3'? :1" ;V:§ M'- 4, 1-. r: 3‘ ' ”3 Fag- :J’L 1;. 1'1“" I’d-m 31‘: Hfl1m955ji‘15 ! 11:13, .1" I39? . . ;5 ’9; 3’ [W 115529319 62‘” ”(95" ‘9.“ H149 WV? “2} 9 113 $111 .7 . I M1351: ~ ‘9 .39; “19:9 .39.91}w.*1‘3"§r w‘:% ~ ‘09 7 ‘Q1‘A"l7.‘ifi:‘ aw.“ .fipfif '39:? 1153; U. L11? {5‘53“ \‘51 9’61} ~",:’=21E;~‘R:9'Sf 919992-3999 , \' ‘v vet-1,1 \ $511.35 ; ‘19 ‘7 garb“. (‘5‘ $355 {:11 “1‘91 7.;"11 '3 .13" ‘19} , 139$ 15:.ij , .. 95%; 7.5373? ' ‘ 55:31: 99.1 efiy 9; P‘i ’* “9111,11: 1, . 9 Q ”giifiél 9 5," tn -’ , j . 912..., 9 .. ‘1 ““39“ $93M; 9: .9 ‘1. 4 “9%“ ‘1; ‘ - «g; {>- 3»"-)1<., .' 5‘5; . 5, "T , :1 .9? ~.‘ ‘ 5-.k.fg,';:,1‘, ‘91, ‘ '23; 5.19;...5k‘: 1%,.7’5 17; Klffliglfiwh 9.23317’ “-w..:.::‘.7- F 99.9%» I: "’""“‘*:; F.o7-;:-'p°-1“- guy-Si? «up g” $355-$135» rm ”1%, - tlg‘é‘ fig»? 2.1-. 9“. ‘19 9. 321 E7344" 91 11.9 vywhtéé‘zj‘fi “'7‘" 73:99”; 99:92:23 ‘59}; z .7 I ’93' 7.331%. ‘ ' “9.9.- ,5 ‘19 2,155 9953;. ‘Frkligb‘w 5::‘2' "’ 9 21:?” c - 1‘15: lira ‘ 1 ‘Auf {firm- 4 w. a u: 9‘ flag}, '9... .{ £3191: ‘ _ £9159. ‘93}: ,9,“ N T ,fig‘flvflm (Q); I» 7‘ $- ‘ ‘4 .1 2S. «flittg; 6' 7‘1.“‘."; : mag . 9n ‘le'vggv fi”ga§ - - 1 Km. 1—.» 5:33. :THESIS LIBRAIIY ~ Michigan State University This is to certify that the thesis entitled CATALYTIC DEHYDRATION OF SUGARS presented by Jinder Jow has been accepted towards fulfillment of the requirements for M.S. Chemical Engineering degree in 7777,77 Major professor Date Dec. 20, 1984 0.7639 MSU is an Affirmative Action/Equal Opportunity Institution MSU LIBRARIES v RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. CATALYTIC DEHYDRATION OF SUGARS By Jinder Jow A THESIS submitted to Michigan State University in partial Fulfillment of the requirements For the degree of MASTER OF SCIENCE Department of Chemical Engineering 1984 © 1985 JINDER JON All Rights Reserved ABSTRACT CATALYTIC DEHYDRATION OF SUGARS Jinder Jow The objective of this research was to explore the dehydration oF sugars to levulinic acid using solid acid catalysts (such as zeolites and heteropoly acids). Sugar dehydration via solid acid catalysts has several potential advantages relative to Fermentation and other acid- catalyzed sugar dehydration technologies For production oF chemicals. Primary advantages are (1) high carbon conversion (2) no dilution required (3) short reaction time (4) ease oF catalyst separation and recovery and (5) high Feasibility For continuous process. A lot oF work has been reported on the dehydration oF sugars to chemicals using inorganic acids (such as H3P04, H2804, HCl, etc.) and strong ion-exchange resins (such as Diaion PK-208, Amberlyst XN-lOlO. Dowex NSC—1H, etc.) as catalysts in a solvent system (such as water, dimethyl sulphoxide, etc.). Two major products are produced during catalytic dehydration oF sugars-—levulinic acid and 5— hydroxymethyl-2-Furaldehyde(HMF). Carbon conversion to major products is increased and reaction time is decreased relative to Fermentation. However, most work with previous acid catalyzed dehydration oF sugars studies resulted in low conversions to levulinic acid, the requirement oF a solvent For substrate. and requirement oF a homogeneous catalyst system. In this work, two basic experimental studies were conducted. They were designed to determine (1) iF one step catalytic dehydration oF sugars to gas phase hydrocarbons is possible and (2) the nature and yield oF the dehydration products in the liquid phase. In the First study. the aqueous sucrose solutions were reacted with several solid acid catalysts. The gas phase above the reaction system was analyzed to determine iF hydrocarbons were produced. It was shown that there were no organic gas products. even though the reactions were carried with zeolite or heteropoly acid. Only the carbon dioxide was produced in the gas phase. But the changes both in the value oF PH and in the color oF liquid along with C02 generation indicated that the dehydration reactions occured. In the second study. one gram oF Fructose was reacted with one gram oF LZY zeolite under a nitrogen atmosphere. Three temperatures (95 C. 120 0C, and 140 C) and several reactions time (0.5hr, 1.0hr, 2.0hr. 5.0hr, and 15hr) were investigated in this nonsolvent system. In addition, two runs were designed to study the catalyst eFFect by using heteropoly acid as a catalyst and the solvent eFFect by using water as a solvent For one hour at 95 OC. The liquid phase oF the reaction system was analyzed to determine the nature and the yield oF dehydration products. It was shown that high yields and selectivity oF levulinic acid were obtained by using LZY zeolite at moderate temperatures. There was an increase in yield oF levulinic acid by increasing either temperature or reaction time. HMF was observed as a reaction product For reaction times 0F 2 hours or greater at 140 OC. HMF was not observed at short reaction times. times less than two hours. This behavior is not characteristics oF a reaction intermediate as observed in some homogeneous catalyst systems as reported in the literature. The order oF the conversion rate oF Fructose in our work was higher than that oF Fructose in the solvent system as reported in the literature. Also, isomerization oF hexose occured in this dehydration reaction. This work demonstrates that the use oF solid acid catalysts such as LZY zeolite to dehydrate sugars in a non- aqueous medium is potentially superior to other acid— catalyzed sugar dehydration systems including Fermentation due to (l) the high yield and selectivity oF levulinic acid. (2) the ability to separate and recycle catalysts easily, (3) elimination oF the need For a high energy process to separate products. and (4) high Feasibility For development oF a continuous process. It is recommended that continued research be directed toward optimization oF catalyst systems and evaluation oF various overall process schemes For directly converting sugars to Fuels and chemicals. Tomy dear parents, ChingLLong Jow Lei-How Jow Wendie-Lexie L5,. 244: M91, $4144— ACKNOWLEDGEMENT The author wishes to express his appreciation to his academic advisor, Dr. Martin C. Hawley For his guidance and assistance. Appreciation is also given to Dr. Thomas J. Pinnavaia For his advise, Dr. Derek T. A. Lamport For his guidance in analytic methods, and Dr. Antonio Devera For his assistance in Gas Chromatography. The author is grateFul to Mr. E. Patrick Muldoon. Mr. James J. Smith, and Mr. Gregory L. Rorrer For their assistances in operating equipments. The author also wishes to thank Mr. Cheng-Liang Chang For his discussion and, especially, Shu-Chen For her encouragement. iii TABLE OF CONTENTS Page LIST OF TABLES -------------------- vi LIST OF FIGURES ------------------- vii LIST OF NOTATIONS ------------------ viii INTRODUCTION ————————————————————— I BACKGROUND ---------------------- B FUNDAMENTALS OF CHEMISTRY ———————————— B A. SUGAR DEHYDRATION B. LEVULINIC ACID REACTION PROPOSED MECHANISM OF SUGAR DEHYDRATION ------ 12 LITERATURE REVIEW ————————————————— 18 A. SUGAR DEHYDRATION B. LEVULINIC ACID CONVERSION KINETIC MODEL OF FRUCTOSE DEHYDRATION ------- 33 A. HOMOGENEOUS MODEL B. HETEROGENEOUS MODEL OBJECTIVES AND RESEARCH PLAN -------------- 38 EXPERIMENT ----------------------- 41 STUDY 1: DESIGN OF THE APPARATUS -------------- 41 MATERIAL PREPARATION ---------------- 43 EXPERIMENTAL PROCEDURE --------------- 4S ANALYTIC EQUIPMENT AND TECHNIQUES --------- 46 EXPERIMENTAL RESULTS ---------------- 48 STUDY 2: DESIGN OF THE APPARATUS --------------- 49 MATERIAL PREPARATION ---------------- SI EXPERIMENTAL PROCEDURE --------------- 52 ANALYTIC EQUIPMENT AND TECHNIQUES --------- 53 EXPERIMENTAL RESULTS ---------------- 69 DISCUSSION OF RESULTS ------------------ 80 CONCLUSIONS ---------------------- 88 RECOMMENDATIONS -------------------- 91 APPENDIX - DATA CALCULATION -------------- 94 REFERENCES ----------------------- 105 LIST OF TABLES Table I. 2. 10. 11. 12. 13. Products For acid—catalyzed reaction oF Fructose — — Page 10 EFFect oF a solvent (MIBK) For dehydration oF Fructose ------------------------- 23 Literature summary oF sugar dehydration ------ 25 A summary oF results For study 1 ---------- 48 Area percent report oF Run # 017 ---------- 66 A summary oF results For study 2 ---------- 69 Reaction time eFFect For study 2 at 140 0C with LZY -------------------------- 70 Temperature eFFect For study 2 For 1 hr with LZY -------------------------- 71 Temperature eFFect For study 2 For 1 hr without LZY -------------------------- 72 Catalyst eFFect For study 2 For 1 hr at 95 OC —————————————————————————— 73 Water eFFect For study 2 For 1 hr at 95 0C with LZY -------------------------- 74 Isomerization oF Fructose For study 2 ------ 75 Mass balance oF Run #017 For diFFerent reject areas —————————————————————————— 76 vi LIST OF FIGURES Figure Page 1. Schematic diagram oF research process ------- 39 2. Schematic diagram oF experimental apparatus For study 1 8. 9. 10. ll. 12. I3. 14. -------------------------- 42 Gas Chromatography oF Standard N2 and C02 ----- 47 Gas Chromatography oF Run # 001 ---------- 47 Schematic diagram oF experimental apparatus For study 2 -------------------------- 50 Chromatography oF Standard Fructose and Inositol - - 62 Chromatography oF Standard Glucose and Inositol - - 63 Chromatography oF Standard Sucrose and Inositol — - 64 Chromatography oF Run # 017 ———————————— 65 Chromatography oF Standard Fructose, LA. and HMF - - 67 Chromatography oF Run # 017 ------------ 68 Conversion percentage oF Fructose and Yield percentage oF Levulinic acid. Glucose, and HMF vs reaction time at 140 0C with LZY ------------------ 77 Conversion percentage oF Fructose with and without LZY vs temperature For 1 hr —————————————— 78 Yield percentage oF Levulinic acid with and without LZY vs temperature For 1 hr —————————————— 79 vii CAP DMF DMSO FA G. C. HFBA HMF HPA HPLC HUM LA LZY MIBK ODS Rc TCD .6 0. LIST OF NOTATIONS Chromatography application Program SoFtware Dimethyl Formamide Dimethyl Sulphoxide Formic Acid Gas Chromatography HeptaFluoro Butyric acid 5-Hydroxymethyl-2-Fura1dehyde Heteropoly acid High PerFormance Liquid Chromatography Humin Levulinic Acid Linde Zeolite Y type Methyl Isobutyl Ketone Octadecyl SulFate conversion Factor oF one compound to the internal standard Thermal Conductivity Detector Exponent notation viii INTRODUCTION Crude oil and natural gas are the primary resources For chemicals. The era oF low price and readily availabe petroleum will likely draw to an end in the Future. Chemical resources will eventually shiFt to coal and/or biomass. Even though coal has the potential to substitute For petroleum, it is nonrenewable and limited. In the very long term. biomass will be the main resource For hydrocarbons and organic chemicals. Biomass has several distinguishing Features relative to coal: composition. renewable. distributed source. and human control oF resource. A problem For biomass—to—chemicals system is whether biomass should be converted either to chemicals by destroying the chemical structure oF the original biomass. or by maintaining chemical Features in the products similar to the orginal biomass. There are several technologies available For converting biomass to useFul chemicals and Fuels. These technologies are summarized below : Thermal Conversion: gasiFication synthesis Biomass ---------- > synthesis gas --------- > hydrocarbons ————————————— > methanol Fast pyrolysis Biomass --------------- > oleFins pyrolysis Biomass --------------- > pyrolytic oils + char + gas Microbial Conversion: digestion Biomass --------------- > methane SacchariFication and Fermentation: enzyme hydrolysis Starch -------------------------- > sugars acid hydrolysis Cellulose or Hemicellose ------------------ > sugar + ligin Fermentation Sugar ------------------ > aqueous ethanol + C02 + Yeast : distillation ——————— > anhydrous ethanol dehydration/ZSM—S Anhydrous ethanol ------------------------ > hydrocarbons Mobil Process Thermal conversion processes For converting biomass to Fuels and chemicals involve energy intensive pyrolysis and gasiFication steps to produce products which can be converted to hydrocarbons and methanol. GasiFication processes basically destroy the chemical structure oF the original biomass to produce a synthesis gas which can be reacted over various catalysts to synthesize a spectrum oF hydrocarbons and oxygenated chemicals such as methanol. Microbial conversion involves a very slow digestive reaction eventually converting biomass to methane. SacchariFication and Fermentation processes convert biomass to sugars which are Fermented to dilute alcohol solutions. These processes are attractive since the technology is well developed. But. the separation oF ethanol From water by distillation is energy intensive. The Fermentation process is a traditional and well- known process to convert biomass into chemicals. But it has some disadvantages aFFecting the economics: (1) long reaction time. (2) low carbon conversion. (3) very dilute aqueous medium required, and (4) diFFiculty in continuous process development. Our goal is to explore a catalytically continuous process For converting sugars to useFul chemicals. which has short reaction time. high yield. high selectivity, high carbon conversion. and nondilute reaction medium. Instead oF Fermenting sugars to dilute alcohol solution. our research approach is directly to convert sugars to chemicals using a solid acid catalyst as the Mobil process did on methanol. A major diFFerence between our concept and the traditional process based on Fermentation is that the reactions may take place either in a concentrated solution or in a nonsolvent system. Only the water produced by the reaction can be easily removed by evaporation. Whereas, in a Fermentation process. dilute solutions oF water/ethanol (7 % ethanol) are separated to produce anhydrous ethanol by distillation which is energy intensive. Basically. our process has the potential oF converting sugars directly to chemicals without signiFicant dilution as required by Fermentation. Meanwhile. our research is stimulated by the desire to shorten the present long reaction time oF dilute solution Fermentation oF sugars. From Fundamentals oF chemistry. the dehydration oF hexoses in acidic media produces 5—Hydroxymethyl-2- Furaldehyde (HMF; also called as 5—hydroxymethyl—2- Furancarboxaldehyde) and Levulinic acid (LA; also called as 4-oxo—pentanoic acid). HMF, a ring structural Furan derivative, will be a good intermediate For the chemical industry due to the multiFunctional groups on its structure. LA has both keto and carboxyl groups to be a potential intermediate in producing various pharmacegticals, pesticides. dye. and plasticizes. Reid H. Leonard has investigated and suggested most 0F the reactions as well as applications oF levulinic acid as a basic chemical raw material. It. also. has attractive applications as a source oF three types oF lactone solvents and maleic anhydride. The salt oF levulinic acid could replace ethylene gycol as an antiFreeze in the automobile system. LA can be catalytically converted to 1.4— pentanediol by hydrogenation or to methyl ethyl ketone by decarboxylation. It seems possible to convert sugars directly either to ketone or to hydrocarbons through the Formation oF levulinic acid by using suitable catalysts and Favorable reaction conditions. In our research system. both HMF and LA. which are dehydrated From sugars. can be extracted by methyl isobutyl ketone From the production phase. The remains are water removed by evaporation and the unreacted sugar solution recycled back to the primary reactor. HMF or LA can be continuously converted to ketone or hydrocarbons in the secondary reactor by another reaction scheme. Basically. one mole oF HMF by dehydration oF one mole oF hexose produces three moles oF water. One mole oF HMF can be Further converted into one mole oF LA and one mole oF Formic acid (FA) with two moles oF water consumed. All the reaction schemes involve dehydration, hydrolysis. and decarboxylation. The ideal stoichiometric reaction scheme is shown below: Individual Reaction dehydration step acid ————————— > H O HMF + 3 H O C6H1206 (hexose) C6 6 3 ( ) 2 hydrolysis step acid C6H6O3 (HMF) + 2 H20 ------- > C5H8O3 (LA) + CHZOZ (FA) decarboxylation / hydrogenation step solid catalyst C5H8O3 (LA) ————————— > CQHBO (methyl ethyl ketone) + C02 Raney Nickel C5H803 (LA) + 3 H2 ————————————— > C5H1202 4’ H20 Total Stoichiometry For our process: without decarboxylation/hydrogenation: acid acid C6H1206 ————————— > HMF ———————— > C5H803(LA) + CH202 + H20 1 solid acid > C5H803(LA) + CH202 + H20 with decarboxylation/hydrogenation: acid/solid acid C6H1206 ——————————————— > LA —-> C4H8O + CH202 + C02 + H20 acid/solid acid 3H2 C6H1206 > LA —-> C5H1202 + CH202 + 2 H20 compared to Fermentation process: yeast/H20 C€h§6 >2<%%m4+2c% Apparently. our process has Four advantages over Fermentation in biomass-to-chemicals research: (1) high carbon conversion. (2) lower reaction time. (3) without signiFicant dilution For reaction. (4) availability oF continuous process development. BACKGROUND FUNDAMENTALS OF CHEMISTRY A. CHEMISTRY OF SUGAR DEHYDRATION: The dehydration oF carbohydrates in alkaline or acidic aqueous solution has been discussed by J. F. Harris et al . The Final dehydrated products are determined by the character oF the medium. the structures oF the carbohydrates reacting. and the conditions oF reaction. For sugar dehydration. the structures oF products depend on the character oF the medium. In an acidic solution. sugars produce Furan compounds. In a basic solution. sugars produce acyclic saccharinic acids. For various types oF sugars, the dehydration rate depends on their structures. The dehydration rate oF D—Fructose is about 40 times higher than that oF D—glucoseu. 1F sucrose dehydrates in an acidic solution. only the portion oF D— Fructose molecule reacts. and D—glucose is completely recovered. It implies that keto—structure is more reactive than aldo—structure in the sugar structure. Ring opening in the reaction mechanism is the First step For the acidic or basic reaction oF cyclic sugar. Then the acyclic sugar goes through the Lobry de Bruyn—Alberda Van Ekenstein transFormation. The transFormation results From the simultaneous occurence oF these three reactions : epimerization oF aldoses. epimerization oF ketoses. and aldose—ketose isomerization. The reactive acyclic species. principally 1.2- and 2.3— enediols will be Formed through structure rearrangement oF acyclic sugar. The rate oF sugar dehydration is naturally dependent on both the ease oF ring opening and the rate oF Formation oF the reactive acyclic species. which are 1.2- and 2.3-enediols. By isotope—exchange experiments ' . the Formation oF acyclic enediols. which are the intermediates in the isomerization. is apparently the crucial step that leads to dehydration products. The dehydration oF the enediols is the next step subject to general acid—base catalysis For sugar dehydration. There are three Forms oF dehydrated products oF sugars : (a) volatile products: carbon dioxide. acetone. water. etc. (b) nonvolatile soluble products : HMF. LA. FA. etc. (c) nonvolatile insoluble products : Humin, etc. The products oF dehydration oF D-Fructose in an acidic solution are given in Table 1 ’ . The major components oF the nonvolatile soluble products are 5-Hydroxymethyl—2- Furaldehyde (HMF) and Levulinic acid (LA). The Formation oF Humin (HUM) parallel to HMF has been supposed to result From the copolymerization between HMF and the acyclic intermediates 9' 10’ 11 . HUM is a nonvolatile insoluble solid whose color varies From brown to black and composition is C : 66.4% ; H : 3.9% determined by thermogravimetric analysis. 7.8 Tablel :products For acid—catalyzed reaction oF Fructose 5-hydroxymethyl- -2-Furaldehyde levulinic acid Formic acid acetic acid alpha-Angelica lactone beta-Angelica lactone isomaltol FurFural 4—hydroxy-2.3.5.- —hexanetrione S-methyl-Z-Furaldehyde 2-(2-hydroxyacetyl)Furan 2-(2~hydroxyacetyl)— ~Furan Formate 4-hydroxy-2-(hydroxymethyl) —5-methyl-3(2H)-Furanone humin carbon dioxide acetone c H o 5 8 3 CH O 2 2 C2H402 c H o 5 6 2 C5H602 C6H6O3 C5HI+O2 C6H704 C6H602 C6H6O3 C7H605 31-31.5 33-35 8.6 16.7 -38.7 110-120 245.8 100.8 118.1 161.7 —56.6 56.5 10 B. CHEMISTRY OF LEVULINIC ACID REACTION: Levulinic acid (LA) is obtainable From sugars via dehydration and hydrolysis. The theoretical yield oF LA From hexose is 64.5 1 (Reid H. Leonard: 1956). The reactive nature oF levulinic acid. which is a biFunctional intermediate. is shown by the keto and carboxyl groups. LA can be catalytically converted to 1.4—pentanediol by hydrogenation or to methyl ethyl ketone by decarboxylation. It also is very convertible to three types oF lactones as solvents. General reaction oF levulinic acid. the reduction oF levulinic acid by catalytic hydrogena—tion. the oxidation. halogenation. general application. and its reaction as ketone were investigated by Reid H. Leonard (1956). 11 MECHANISM OF SUGAR DEHYDRATION Consideration oF mechanism i these dehydrated products can simple combination oF hydrolys dehydration steps. Several mechanisms For the acid catalyzed shown as Follows: Hydrolysis: HOCH 2 H ZCOH O HOC O :———-O OH HO :— \ / OH \ / H0 \ / <-- \/ /\[____/ \ / \ \_/\ ~~> \ HOCH2 / / HOCH2 OH HO SUCROSE D— Mutarotation oF D—GLUCOSE: HOCH2 :-—-O <——— OH OHOH / OH \ --—> O=C—C—C—C—C—C—OH / \/___/\ OH OH H alpha-D(+)-GLUCOSE ACYCLIC GLUCOSE (36% at equilibrium) ndicates that most oF be Formed From sugars by a is. enolization. and workers have proposed diFFerent acid: dilute sulFuric acid dehydration oF sugars H2 HOCH2 O — O OH \ / \ OH H0\/ / H0 \ / ___\/ + \ __\/ \ / HZCOH OH OH GLUCOSE + D-FRUCTOSE HOCHZ <——- :---O OH ‘---> / 0H \/ / \/____/ OH OH beta-D(+)-GLUCOSE (64% at equilibrium) There is an evidence indicating that the amount oF open chain D-(+)—glucose in the solution is very small. Because the solution oF D-glucose gives no observable ultraviolet absorption band For a carbonyl group. and the solution oF D-glucose gives a negative test SchiFF's reagent. 12 A. Anet mechansim : Isomerization oF D-FRUCTOSE: HO O HOCHZ O \/HO \ —--> HO—C OH OH OH OH ---> \/HO \ /\___\/\ HOCH ------ > CH2 ------ > : :__: : HCOH HCOH HCOH HCOH HCOH HCOH CHZOH Cit OH Cit OH D—FRUCTOSE 1.2-ENEDIOL INTERMEDIATE HMF (Csthzosl (Csthzosl (Catioosl (@1503) 13 Featherlu used labelled sugars to conFirm the existence oF a cyclic precursor to HMF in the dehydration oF hexose. This gave a strong evidence For Scheme 1 and Scheme 2. B. Haworth and Jones 'mechansim: The Foramtion oF a cyclic precursor to HMF may be the rate determining step in the dehydration reaction. It was shown in Scheme 1. C. Anet and Moye mechansim: The Formation oF 1.2-enediol may be the rate determining step in the dehydration reaction ( H. Harry Szmant: 1981). It was shown in Scheme 2. 14 .a. , . oEoSOM Poppy one omoopam Mo COHthohcoo mo Emflpmpoms H omo 2020 O U.IOI LI: 30:55 . 0:0 9:0: \o :0 o: _ _ . a +5.5 :05 oio: 96 6:5: 3330 O O I. _ l. _ :olo = :o 10161: IO 15 3830 303:5. zoio __... .N QEGSOW some go smfleeeoo: SHW .HO COHPNHU smokeosp.H ops omoo Ioio :oio _# e2: :I.V Y / . \./ __ :oic ,o\r o:o_5;6 0 L96 xolo _ oHol: onolx s a _ ”m exaxo ICIU . x H OH M1\.MOI\. H. __ zozo ozo: :osho ozo :olul: DEHYDRATION and HYDROLYSIS of 5-Hydroxymethyl-2-furaldehyde if}: - H2O A—l \o HO I HOHC CH2 0 CH0 2 HMF *HZO +H 0 U . 11W) 0 o c H3 OH c HZ OH CEH3 <.:=O - F 2 _LEZO CH2 JL A CO2H CH3 0 o Levulinic Acid Scheme 3. Mechanism of dehydration and hydrolysis HMF 1? LITERATURE REVIEW A. SUGAR DEHYDRATION: I. EARLY WORK (1895 - 1966) Dull 38 discovered 5-hydroxymethyl—2-Furaldehyde using the oxalic acid to dehydrate Inulin in the aqueous solution. Bonner et allfsreported that HMF was obtained in a 71 mole % yield (based on the Fructose portion) by using iodine as a catalyst in N.N-dimethylFormamide to dehydrate sucrose. Wiggins]:7 used the oxalic acid to study diFFerent carbohydrate sources including glucose. Fructose. lactose. starch. and cellulose From wood pulp to dehydrate into levulinic acid. The yield was usually less than 25 % at the atmospheric pressure. Mckibbin18 used autoclaves to increase reaction temperature to 160 - 200()C. and then the yield increased. It was shown that a higher temperature would result in a higher yield oF levulinic acid. Moye used mineral acid to study the dehydration oF various ketohexose in the nonaqueous slovent For Five seconds at the solvent boiling point. A high yield oF HMF in excess 0F 80% was obtained. II. HOMOGENEOUS INORGANIC ACID AS CATALYST 18 (1). Dilute SulFuric Acid EL. 5. Amin8 has shown that the solution oF D-Fructose (10.09) in 500ml 0F 10% sulFuric acid was heated For 35 hr. The nonvolatile products were shown to be brown black water— soluble products (11.4%) and water-insoluble products (69 %). the rests oF the products (19.6%) were volatile materials which contained 5.6% acetone and 9% carbon dioxide. (2). Hydrochloric Acid The dehydration oF D-Fructose (0.25—1.0 M) to HMF and the dehydration oF HMF (0.25—1.0 M) to LA and Formic acid(FA) in 0.5-2.0 M HCl at 95 0C has been studied by B. F. M. Kuster et al19 . They indicated that more acidic conditions were needed For the Formation oF LA than that oF HMF. The decrease oF water concentration. equivalent to the increase oF the acidity oF the solution. highly increased the conversion rate oF D-Fructose. but slightly decreased the conversion rate oF HMF. The value oF PH apparently aFFected the reaction type. No HMF would be Formed. when the PH value oF the solution was greater than 3.9. No levulinic acid would be Formed. when the PH value oF the solution was greater than 2.7. D-glucose appeared to indicate the occurence oF isomerization while the PH value oF the solution was greater than 4.5. Weak-acid anion would lower the yield oF HMF and enhence the isomerization to glucose. The First order conversion oF D-Fructose and HMF l9 is Fit to the experi-mental data. The order oF the Formation rate ‘oF humin was 1.3 For the intermediate between D-Fructose and HMF; and 1.7 For the intermediate between HMF and LA. A kinetic model was proposed as Follows: KF K1 Kh K3 F -------- > X ----- > HMF ----- > Y --------- > LA + FA i i i K2 1 K4 ------ > HUM -——--—-—-—> HUM F: Fructose: HMF: 5—hydroxymethyl—2—Furaldehyde: HUM: humin: LA: levulinic acid FA: Formic acid; X.Y: intermediates KF. Kh. K1. K2. K3. and K4 : rate constants d(F)/dt = —KF*(F) d(X)/dt = KF"(F)-K1"(X)-K2"(X)l°3 d(HMF)/dt = K1*(X)-Kh*(HMF) d(Y)/dt = Kh"(HMF)-K3"(Y)-K4"(Y)l.7 d(LA)/dt = K3*(Y) —O.3 1.3 2.1: Kx = KF * Kl / K2 1.7: Ky = Kh-O'7 * K31'7 / K4 20 III. STRONGLY ACIDIC ION-EXCHANGE RESIN AS CATALYST (1). Water-Resin Biphase System H. F. Rasel5 obtained HMF and LA From sucrose by using highly acidic ion-exchange resins as catalysts. Four commercial resins (Dowex MSC-IH. Amberlyst 15. Amberlyst XN- 1010. and Amberlyst XN-1005 ) were used. The Former three resins achieved the selectivity 0F 83 % LA For 24 hr reaction time at 100 0C. but the yield percentage oF levulinic acid was less than 25 %. They indicated that the resin pore size had a strong eFFect on the selectivity. It was shown that HMF was Favored by a larger pore. but LA by a smaller pore. (2). Solvent-Resin Biphase System Nakamura 20used two types oF the strongly acidic ion- exchange resins with a low divinylbenzene (DVB) content as the catalyst and Dimethyl sulFoxide (DMSO) as the solvent. One was Porous type: Diaion PK-ZOB. PK-216. and PK-228. The other was Gel type: Amberlite 1R-118. IR—120. and Lewatit SC-IOB. A continuous dehydration oF D-Fructose was carried out under 60 0C. A 90 mole% yield oF HMF was obtained (basis on D-Fructose) For 8.3 hr reaction time at 80 0C with Diaion PK-216. The rate oF HMF Formation was proposed as Follows: 21 d(HMF)/dt = k * [ (F)’ - (HMF) ] (F)’: the initial concentration oF D—Fructose (HMF) : the concentration oF HMF k : the rate constant The rate constant k was reversely proportional to DVB content in the resin used. The rate constant oF the porous resin was greater than that oF the gel—type. when the DVB content in resins was the same. (3). Water—Solvent—Resin Triphase System 21, 22 1 explored new ways For the Luc Rigal et a synthesis 0F HMF which could lead to improve yields by ion- exchange resins as catalysts with an extractive sovlent (MIBK) in a triphasic system. The ratio oF the extractive solvent to water was 9. The reaction temperature was Fixed at the boiling point oF the water-methyl isobutyl ketone O azeotrope (88 C). The macroporous strong acid resins ( Lawatit SPC 108. SPC 118. NaFion—H. and Spherosil 5) gave the high yield and high selectivity oF HMF For 15 hr reaction time. but no reaction in the presence oF weakly acidic ion—exchange resins (e.g. Duolite CC3. Amberlite IRC 50). An increase in the average diameter oF the pores in the resins allowed higher yield oF HMF than oF LA. In Table 2. the conversion rate oF Fructose with MIBK as an extractive solvent was three-Fold greater than taht without MIBK. Some solvents ( MIBK. dichlor—ethylether. and benzonitrile) could promote the reaction to obtain high 22 yield oF HMF. Some solvents (alkane. diethyl ketone. t- butyl methyl ether) didn’t give high yields oF HMF due to their lack oF aFFinity toward the ion exchange resins. or due to the insolubility oF HMF ( a polar compound) in these non—polar solvents with the weak dielectric constants. The conversion rate oF Fructose was an increasing Fuction OF the amount oF MIBK introduced. A decrease in the ratio oF water to solvent gave rise to an increased conversion rate. It was a strong evidence 0F solvent Factor in this dehydration reaction. Table 2. EFFect OF a solvent (MIBK) (Luc Rigal:l981) concentration oF D—Fructose (Q/dm3) 222 222 water (cm3) 20 100 solvent MIBK (cm3) 180 none HMF ( % ) 63 I4 IV. LEWIS ACID AS CATALYST A 95 — 97% conversion oF Fructose to HMF was reported by H. Szmant23 by using 25 mole% (based on Fructose) boron triFluoride etherate catalyst (BF3 . Et2 O). and dimethyl sulphoxide (DMSO) as a solvent For reaction times 0F 30 minutes at 100 0C in an inert gas (N2 ) atmosphere. They indicated that the yield oF HMF was a Function 0F solvents. 23 reaction time. the ratio oF catalysts to sugars. reaction temperatures. and starting materials. The yield oF HMF increased very sharply to a maximum point as the reaction time increased. but then decreased sharply in all solvents except in DMSO. It implied that the DMSO would provide more stable reaction medium For sugar dehydration. Meanwhile. they indicated that the higher temperature would result in the higher yield oF HMF. V. LITERATURE SUMMARY Van Einenstein (1909) dehydrated Fructose using oxalic acid as a catalyst in the aqueous solution For 3-4 hr at 145 OC. HMF yield 0F 22-29 % was obtained. Haworth (1944) dehydrated sucrose using the same acid in the aqueous solution For 2-3 hr at 145 OC. Only HMF yield 0F 27 % (based on 12 carbons) was obtained and glucose was completely recovered From sucrose. Stone (1950) dehydrated glucose using phosphoric acid as a catalyst in the aqueous solution For 10 minutes at 190 (L. A low yield (15.5 %) oF HMF was obtained. Rice (1958) dehydrated Fructose using phosphoric acid in a water-ketone biphasic solvent up to 48 hrs at 200 (L. A high yield (65—85%) oF HMF was obtained. Moye (1966) dehydrated ketohexoses using mineral acid in nonaqueous solvents For 5 seconds at the solvent boiling point. A high yield oF HMF in excess 0F 80 % was obtained. Kuster (1977) dehydrated Fructose using hydrocholoric acid 0 in the aqueous solution For 24 hr at 95 C. A high yield 24 (65-80 mole %) oF levulinic acid was obtained. Szmant (1981) dehydrated Fructose using boron triFluoride etherate as a catalyst in the nonaqueous solvent about one hour at 100 0C. A high yield (78—97 %) oF HMF was obtained. Rase (1975) dehydrated sucrose using Amerlyst ion- exchange resins and Dowex resins in the aqueous solution For 24 hr at 100 0C. A low yield (less than 25 %) oF levulinic acid was obtained. Nakamura (1980) dehydrated Fructose using Diaion ion exchange resins in DMSO solvent For 8.3 hr at 80 0C. A high yield (90 %) oF HMF was obtained. Rigal (1981) dehydrated Fructose using Lewatit. Amberlite. and super-acid ion exchange resins in a water- MIBK biphasic solvent For 4 hr at 88 0C. A high yield (about 50 %) oF HMF was obtained. Table 3. Literature summary oF sugar dehydration (1909 to 1981) Year material catalyst solvents temp. time product reF. O ( C) 1909 Fructose oxalic water 145 3—4hr 29%HMF 24 acid 1944 sucrose oxalic water 145 2-3hr 27%HMF 13 acid 1945 sucrose PH=2-3 water 145 3-4hr 22%HMF 25 1950 glucose H3PO4 water 190 10min 15.5%HMF 26 25 (continuous) reF. 27 28 28 28 29 29 29 29 15 15 15 Year material catalyst solvents temp. time product 1958 Fructose H3PQQ Ketone/water 200 48hr 65—85%HMF =4:1 1962 glucose H3PQQ Water/dioxane 200 37min 23%HMF + NH3 1:1 HBPQ” 1:1 200 37min 30%HMF +N(CH3)3 H3PO Li 1 :1 200 _37min 44%HMF +pyridine 1966 Fructose H2500 2-methyl 126 Ssec 75%HMF ethanol tetrahydro— 78 Ssec 74%HMF FurFuralalcohol HCl methyl 193 Ssec 80%HMF carbinol 12 methyl 193 Ssec 80%HMF carbinol 1975 sucrose Dowex water 100 24hr 24%LA Amberlyst -15 water 100 24hr 23%LA Amberlyst XN-IDIO water 100 24hr 15%LA Amberlyst XN-1005 water 100 24hr 9%LA 26 15 (continuous) reF. 21 21 21 21 21 21 21 21 21 23 23 Year material catalyst solvents temp. time product 1977 Fructose HCl water 95 24hr 65—80%LA 1980 Fructose Diaion DMSO 80 8.3hr 90%HMF 1981 Fructose Lewatit MIBK/water 88 4hr 47%HMF SC—102 9:1 Amberlite 9:1 88 4hr 58%HMF IR-118 Duolite 9:1 88 4hr 54%HMF C-26 Amberlite 9:1 88 4hr 42%HMF A-200C Amberlyst 9:1 88 4hr 30%HMF A—15 Lewatit 9:1 88 24hr 51%HMF SPC -118 Lewatit 9:1 84 24hr 62%HMF SPC -108 Spherosil S 9:1 88 24hr 53%HMF NaFion-SOIH 9:1 88 15hr 50%HMF 1981 Fructose BF3.Et20 Carbitol 100 0.5hr 40%HMF Mecellosolve 100 1.0hr 78%HMF Cellosolve 100 2.0hr 63.5%HMF DMF 100 1.5hr 89.2%HMF DMSO 100 0.75hr 98.8%HMF 27 B. LEVULINIC ACID CONVERSION: Levulinic acid. which is a biFunctional compound. has both keto and carboxyl groups. It is a potential intermediate in producing pharmaceuticals. pesticides. dye and plasticizes. Levulinic acid can be catalytically converted to alcohols or ketones. The related researches For levulinic acid conversion are reviewed as Follows: I. METAL AS CATALYST (1). Hydrogenation Reid H. Leonard2 indicated that the catalytic hydrogenation oF levulinic acid over Ni and Cu-Cr above 200 C would yield the substantial amount oF 1.4-pentanediol. and the small amount oF alpha—methyltetrahydroFuran and l- pentanol. The reaction scheme is shown as Follows: H2 2H2 CH3C0C2H#COOH --> gamma-valeralactone -——-> 1.4 pentanediol Ni Cu-Cr C H O C H O C H O ( 5 8 3) ( 5 8 2) ( 5 12 2) + H O 2 28 : (2). decarboxylation Wilhelm F. Maier30 investigated decarboxylation oF a variety oF carboxylic acids in the gas phase over Ni/Al 203 + H2 between 150 0c and 280 0c and Pd/Sio2 + H2at 330 0c. The over-all reaction was considered as: catalyst RCOOH ------------ > RH + c02 They Found that the heptanoic acid was completely unreacted when N2 instead oF Hz was used as a carried gas . even though no hydrogen was needed in the over-all stoichiometry. It proved that a catalytic site might be a metal/H complex instead oF metal itselF. The reaction mechanism is shown as Follows: Zpd + H2 (============) 2 Pd—H RCOOH' + Pd-H ———————————— ) RCOOH'-Pd—H RCOOH"Pd‘H <=========> RH + Pd-H’ + C02 They showed that levulinic acid was completely decomposed not to 2-Butanone but to gamma-valerolactone by above reaction scheme in the same conditions. This implies that hydrogenation is more active than decarboxylation oF levulinic acid over metal catalysts. 29 (3). Hydrogenolysis or Hydrogenation by electrocatalysis Toshiro Chiba et al31 indicated that the Raney Nickel as a good catalytic electrode brought about the hydrogenation oF levulinic acid to gamma-valerolactone because oF a large surFace area and a high hydrogen- adsorption activity. The reaction mechansim For levulinic acid is shown as Follows: H20 + 2 Ra—Ni ----- > 2 Ra-Ni-H + 1/2 02 CHB cocgmcoon + Ra-Ni-H <====> erg COCZHICOOH-Ra-Ni-H c coc COOH-Ra-Ni-H <======> c H o + H o + Ra-Ni H3 2”” 5 8 2 2 (gamma—valerolactone) (73 % yield) II. CONDUCTIVE METAL OXIDE AS CATALYST Photocatalysis 32 H. L. Chum showed that the photocatalytic decarboxylation oF levulinic acid in slurries composed oF n— TiOg /Pt led to the major products: methyl ethyl ketone and carbon dioxide. The secondary products such as acetaldehyde. acetone. acetic acid. and propionic acid might be produced by cleavage and oxidation oF the relevant C-C bonds either oF levulinic acid or oF methyl ethyl ketone. The reaction scheme is shown as Follows: hv CH3COC2rlLlCOOH ------------ > CH3COC2H5 + C02 —T'O Pt n l 2/ 30 The mole yield oF C02 (based on levulinic acid) is quite low (about 0.4 to 1.4%). How to make these reactions occur at much higher rate and yield is important but not well-reported yet. III. SOLID ACID AS CATALYST (l). dehydration and decarboxylation The results by C. D. Chang 1 For the conversion oF acetic acid and acetone into hydrocarbons over ZSM-S (Famous Mobil catalyst) are shown below. The dehydration oF acetone at 399 9C. LHSV oF 8.0 hr.1 led to 95.3 % conversion and yielded 93.9 % hydrocarbons and 6.1 % C0 + O CO . The dehydration oF acetic acid at 371 C. LHSV oF 1.0 gl hr led to 29.9 % conversion and yielded 57.6 % hydrocarbons. 41.2 % COZiand 1 % C0 and 0.1 % acetone. It showed that the deoxygenation oF acetone and acetic acid occured via dehydration and decarboxylation. Levulinic acid with both keto and carboxylic group seems convertible to hydrocarbons by zeolite under certain Favorable reaction conditions. In general. reactivity oF Functional group into hydrocarbons over 25M— 5: alcohol > aldehyde > ketone > acid 31 (2). decarboxylation Masayuki Otake:x3investigated the gaseous decomposition oF primary. secondary. and tertial carboxylic acids at 200- 300 (3C over heteropoly acids. The main products are carbon monoxide and oleFins: ethylene From propionic acid(PRAC). propene From isobutyric acid(IBAC). butene From pivaric acid (trimethyl-acetic acid. TMAA). The conversions were above 90 %. But both butric and valeric acid were inactive on this catalyst below 300 C)C. Otake did not study levulinic acid. The reaction scheme is shown as Follows: H R' solid acid RC—C-COOH ———————————————— > RC=CR’ + C0 + H20 R‘R" R‘R" catalytic activity For decomposition oF carboxylic acid i-BPCL >15 [Pkfi-qu] or FLESIWLZ 910 ]>SIC£ -A12% >>SiC£ or A12% reactivity oF carboxylic acid tertial > secondary > primary From the above conclusions. it would logical that levulinic acid. primary carboxylic acid. could be less convertible to oleFins by heteropoly acid even at high temperatures. 32 KINETIC MODEL OF FRUCTOSE DEHYDRATION (A). Homogeneous Model The Following model and examples were presented by Ben F. M. Kuster (1976). KF K1 Kh K3 F -------- > X ----- > HMF ----- > Y --------- > LA + FA I I I I 1 K2 1 K4 ------ > HUM —---—----—> HUM F: Fructose: HMF: S—hydroxymethyl-2-Furaldehyde; HUM: humin: LA: levulinic acid FA: Formic acid; X.Y: intermediates KF. Kh. K1. K2. K3. and K4 : rate constants A typical example was the dehydration oF D-Fructose (0.25-1.0 M) to HMF and the dehydration oF HMF (0.25-1.0 M) to LA and FA in 0.5-2.0 M HCl. The First order conversion oF D-Fructose and HMF was in an agreement with the experimental results. So. the diFFerential equations could be derived: d(F)/dt = -KF*(F) ------------ (I) d(X)/dt = KF*(F)-Kl*(X)—K2*(X)‘Nx ------------ (2) d(HMF)/dt = KI*(X)-Kh“(HMF) ------------ (3) d(Y)/dt = Kh*(HMF)-K3*(Y)-K4*(Y)‘Ny ------------ (4) d(LA)/dt = K3*(Y) ------------ (5) 33 Because the concentration oF X and Y were very low. the steady-state concept could be applied. i.e. d(X)/dt = 0 : d(Y)/dt = 0 So. KF*(F) = K1*(X) + K2*(X)“Nx ————————————— (6) Kh*(HMF) = K3*(Y) + K4*(Y)“Ny ------------- (7) Let Sx : the Fraction oF D-Fructose reacting to HMF Sy : the Fraction oF HMF reacting to LA Sx -d(HMF)/d(F) = K1*(X)/[KF'(F)] ------------- (8) Sy -d(LA)/d(HMF) = K3*(Y)/[Kh*(HMF)] ------------- (9) Substituted (8).(9) into (6).(7) Sx‘Nx/[I-Sx] Kx*(F)‘[l-Nx] ------------- (11) Sy‘Ny/[l-Sy] Ky*(HMF)“[l—Ny] ------------- (12) KF‘[1-Nx]*K1‘Nx/K2 where Kx Ky Kh“[1-Ny]*K3‘Ny/K4 The diFFerential equations were Fully determined by the model parameters: KF. Kx. Kh. Ky. Nx. and Ny. First step. values For KF and Kh would be easily calculated From the conversion data For D-Fructose and HMF. Next step. Ky and Ny were calculated From LA data For the reaction starting with HMF. Several combinations oF Ky and Ny could be used equally well to Fit the experimental data. However. only For Ny = 1.7. Ky had an uniForm constant 34 value oF 1.7 For all experiments. So. Ky = 1.7 and Ny = 1.7 were chosen. ThereaFter. Kx. Nx. Kh. and KF were calculated From HMF and LA data For the reaction starting with D- Fructose. using the values oF Ky and Ny already obtained. Again. there were several combinations oF Kx and Nx which were able to be used equally well to Fit the experimental data and applied the same values oF Kh and KF obtained From First step. When Ky and Ny were Fixed at 1.7. only For Nx = 1.3. Kx had an uniForm constant value oF 2.1 For all data sets. Kx = 2.1 and Nx = 1.3 were picked up. ThereFore. d(F)/dt -KF*(F) d(X)/dt = KF*(F)-K1*(x)-K2.(x)l.3 d(HMF)/dt = K1*(X)-Kh*(HMF) d(Y)/dt = Kh* HMF on ---------- (1) inside catalyst oF Fructose outside catalyst Reaction: r.d.s. (HMF)/[K3*K4*K5] Ri K3 K4 K5 intermediates 36 For rate—controlling K1*(Fi)-K2*(X) ---(2) (XS)/(X)/(S) ----- (3) (HMFS)/(XS) ----(4) (HMF)*(S)/(HMFS)‘*(5) HMF : 5-hydroxymethyl-2-Fura1dehyde Ro : the global rate oF Fructose converted For steady state: Na = Ri = R0 ; * Na Kg [(Fo)-(Fi)] = R1 = K1*(Fi) - K2*(X) Fi Kg/[KI+Kg]*(Fo) + K2/[K1+Kg]/[K3*K4*K5]*(HMF) R0 = Na KQ*[(F0) - (Fi)] { Kg/[Kg+KI] } * { K1”(Fo) - K2/[K3’K4'K5]*(HMF) } It is in a good agreement with the Nakamura equation : R = d(HMF)/dt = -d(F)/dt = K ' [ (F0) - (HMF) ] 37 OBJECTIVES AND RESEARCH PLAN The long term goal oF this research is to develop a continuous process For the catalytic conversion oF sugars to chemicals such as ketones or alcohols by using a solid catalyst. This process would be an alternative to Fermentation oF sugars. The advantages oF our concept relative to Fermentation are: (1) high carbon conversion (2) short reaction time (3) no dilution required (4) ease oF catalyst separation and recovery and (5) high Feasibility For the continuous process. In this new process. there are two important reaction schemes: (1) catalytic dehydration oF sugars into levulinic acid. and (2) catalytic conversion oF levulinic acid into alcohols or ketones. These reaction schemes are illustrated as Follows: Reaction scheme 1: Dehydration: solid catalyst A sugar ---------------------------- > Levulinic acid Reaction scheme 2: Decarboxylation / Hydrogenation: solid catalyst B Levulinic acid ———————————————— > speciFied products (ketones or alcohols) 38 extractive solvent l i< --------- ‘ solid catalyst A g “-l---- sugar ---------------------------- l g ->: Fixed bed reactor :-> lextractor:->: Distil.: solution ---------------------------- g g . l l i l ------------ l -------- ------ : evaporator : <-—-------—-- __--_____?_- water < ----- solid catalyst B speciFied products <-—--: Fixed bed reactor 1 <- (ketones or alcohols) Figure 1. Schematic diagram oF research process 39 —lu\ The immediate goal is to explore the dehydration oF sugars into levulinic acid using solid acid catalysts. This work is signiFicant because previous studies (acid- catalyzed dehydration oF sugars) have not produced the high yield and high selectivity oF levulinic acid. The First- stage research is: (a) to design and develop speciFications oF solid acidic catalysts For preliminary experiments. and (b) to set up a small-scale batch experimental Facility to conduct preliminary sugar dehydration over solid acid catalysts. Two laboratory studies are conducted: Study 1: to determine iF one step catalytic dehydration oF sugars to gas phase hydrocarbons is possible by using solid acids. sugar: aqueous sucrose solution solid acids: Nax. ZSM-S. and 12-Tungstosilicic acid Study 2: to determine the nature and yield oF the dehydration products in the liquid phase by using solid acid. sugar: Fructose solid acids: LZY zeolite and 12—Tungstosilicic acid 40 EXPERIMENT STUDY 1: 1. DESIGN OF THE APPARATUS The reactor was a 500ml three—neck Flask. One oF its necks was connected by a distilling adapter which was extended to a 400 ml Liebieg condenser. The other necks were connected by reducing adapters which were connected by Flow adapters. One oF these two was extended to a nitrogen gas cylinder. The other was reserved to connect a manometer gauge to detect the reaction pressure. When the reaction occured. the vapor products would pass by a vaccum- type distillation adapter and go through the condenser to a 50 ml Flask liquid collector. The condensable vapor would be condensed in the liquid collector. The uncondensable gas would be pushed into a 500 ml gas-collected Flask by draining water out. The heat source was a Thermolyne Hot Plate. The temperature was measured by a ~10 0C to 260 OC thermometer. A schematic diagram oF the apparatus is shown in Fig 2. 41 H hoppm pow mapmpwgdw Heepoeapogxm no empwmflo oameoSOm .m opsmflm X __— III '1'” I'Hill ‘ II . IIIII' III I _— 'I llll fl :1 (D L) m Hommo> poems .m pompmopoo .m Hmmmm> poems .h oemam #0: .Q mesa soaeosm .m Mmmap xomcloopce as com .o popooaaoo new .0 pmpoannmSP .m popomaaoo owsofla .m poopflaho Cowopewp .< 42 . ..\\lL.r. II. MATERIAL PREPARATION A. Catalyst Preparation: Three types oF strong acid catalysts were used: (1) Heteropoly acid: H4[SiW 040] 7H 0 12 2 (2) Zeolite: ZSM-S and NaX (3) Inorganic acid: H2504 The 12-Tungstosilicic acid was prepared by the Following procedures ( Jolly. Willaiam L.: 1970): -2 -2 i 12 W0 + SiO + 26 H -—-> H Siw 0 7H 0 + 4 H O h 3 4t 12 40] 2 2 l. dissolve 509 sodium tungstate(+6) 2-hydrate in 100 ml H20 2. add 2.7 ml sodium silcate solution (40 0Be') 3. briskly stir and heat at boiling while adding 30 ml concentrated HCl drop by drop 4. cool. Filte. and add 20 ml concentrated HCl 5. shake the solution with 35 ml diethyl ether ( IF no three phase. add more little ether.) 6. collect the bottom layer and add 12 ml concentrated HCl and 38 ml H20 with 10 ml ether. 7. shake and collect the bottom phase into dish and stand in a draFty hood For two days. 8. dry the remaining crystal at 70 0C For 2 hrs. 9. yield 329 crystal avoiding contacting with anything metallic. The zeolites can be purchased From Mobil Company. 43 B. Reagent Preparation: The amounts oF reagent For each run were weighed on a Cahn eletrobalance and described as Follows: RUN #001: RUN #002: RUN #003: RUN #004: RUN #005: 20.0 with 20.0 with 20.0 2 20.0 with PH = 20.0 H 804 9 sucrose dissolved in 30.0 0.5 g NaX zeolite. PH = 8.0 9 sucrose dissolved in 30.0 0.5 g ZSM-5. PH = 7.0 9 sucrose dissolved in 30.0 . PH = 1.0 9 sucrose dissolved in 30.0 3 mi. 98.8% HzSouand 1.0 g 4.0 9 sucrose dissolved in 30.0 ml water added ml water added ml. 0.05 M ml water added NaX zeolite. ml. 0.015 M so added with 0.5 g H 5i ]7H , PH = 1.3 "b Ll ll; ”12%0 § 44 III. EXPERIMENTAL PROCEDURE The aqueous sucrose solution was prepared by the speciFied acidic medium. and the PH value oF the aqueous solution was indicated by PH paper. beFore the experiment run. AFter the apparatus was set up. water pump was used to suck the air out For 5 min. An inert atmosphere (99.9% nitrogen gas) was maintained throughout the reaction. The reactor was heated From room temperature to 950C. The gas was collected in 500 ml gas-collect Flask. Analyses For gaseous products were carried out using Varian-3700 Gas Chromatography. All reactions had the change values oF PH which were indicated by PH paper. but some turned clear liquid to yellow. 45 IV. ANALYTIC EQUIPMENT AND TECHNIQUES Varian 3700 Gas Chromatography via Hewlett Packard 3390A Reporting Integrator flifiliflflflflflflflflfllflfiflflfiI!*Iflflflfiflfifififlflflfiflfllfiflflflifli purpose: qualitative and quantitative determination For the gas products oF the sugar dehydration. G.C. speciFications: column: amorphous silica gel. 4m x 1/4 in carrier gas: helium. 30 c.c./min inject temperature: 110 0C detector temperature: 110 0C oven temperature: 41 0C (isothermal) detector type: gas volume injection: attenuation: Thermal Conductivity Detector 0.5 ml 4 Operating Procedure: 1. turn on the helium gas. 2. adjust Flowrate at 30 c.c./min For both the leFt (reFerence) and right column. turn on the main power. and wait For instrument warm-up and stabilty For 1 hr to 2 hr set the MODE to TCD turn on the detector power 6. adjust detector output level For zeroing baseline 7. inject the 0.5 ml gas sample. The standard calibration For N2and C02 is shown in Figure Run #001. is 4. One oF the experimental results in study 1. shown in Figure 5. 46 ll nwr m mus H men mm mm“ m .xcmmc hzxmc mm em mm vw\mm\u:q . mm+wmmmn m ucumq Sake» mm eemamw _. n Idm~ ms+ummm. r mm.m mdyh ammo Figure 4. Gas chromatography of standard N2 and 002 Nw/~_m.rll. lily no+wmmmm.v "cmmc achOP mmm.o moo.~ mm ommmvm vo.mu nom.mm m"—.o mmm no+wmmmm.v «m.o Ncwmc hzxac were «mac hm . Newmq mm menu“ vmxmdxusc m a 22m Fmfiu NAHVHva9- i. la: Figure 5. Gas chromatography of Run # 001 47 V. EXPERIMENTAL RESULTS Table 4: A summary oF results For study 1 run solvent catal. temp. react. PH gas prod. liquid(PH) no. (gram) (gram) (QC) time vol.yield color #001 water NaX 95 0.5hr 8.0 C02 3.4ml PH=6.0 30.0 0.5 yield=0.24% pale yellow #002 water ZSM-5 95 0.5hr 7.0 C02 3.0ml PH=6.0 30.0 0.5 yield=0.21% clear #003 water H2604, 95 0.5hr 1.0 C02 3.5ml PH=1.5 (30.09. 0.05M) yield=0.25% yellow #004 water NaX 95 0.5hr 4.0 C02 4.3ml PH=4.0 30.0 1.0 yield=0.30% yellow H2504 (98.8%. 39) #005 water HZSON' 95 0.5hr 1.3 CO2 4.0ml PH=1.6 (30.09. 0.015M) yield=0.28% brown HPA 0.5 9 note: 20.0 gram sucrose is used as sugar For each run. HPA: heteropoly acid. H4[SiW 0 ]7H 0 NaX: X typelgeSTIte2(Faujasite). Na56[(Al02)56(Si02)106] 264H20 ZSM-5: zeolite. n < 27. typically about 3 Na Al Si O 16H 0 n n 96-n 2 48 STUDY 2: 1. DESIGN OF THE APPARATUS The batch reactor was a 200 mm long glass test tube with an inner diameter 0F 25 mm. The reactor was dipped into a light paraFFin oil batch which was a 4000 ml beaker equipped with a 0 0C to 150 OC thermostat. The reactor was plugged by a #3 rubber stopper which contained with two 3mm- inner diameter tubes For nitrogen purging. The inlet tube was connected by a rubber tube to a nitrogen cylinder tank. while the outlet tube was extended to a water pump and an oil vessel. A schematic diagram oF the apparatus is shown in Fig 3. 49 N hoppm Hoe mSPMNHQQM HHHCmEHpomxo mo seaweed oapdEoSOm .m oppmflm 1 50 Hommm> Amen; .m sewn Hao .) mesa poaeosm .m Hmpmoanmpe .m made Pmoe Eo m.m R Go o.om .m newsflaho pomohefls .< II. MATERIAL PREPARATION A. Catalyst Preparation: Two types oF strong solid acid catalysts were used: (I) Heteropoly acid: HPA: H+[SiW120401 7H20 (2) Zeolite: LZY: Linde Zeolite Y type (Faujasite). Mes/h[(A102)t(5i02)z]mH20 with z/t > 2 The preparation oF catalysts was the same as study 1. B. Reagent Preparation: The amounts oF reagent For each run were weighed on a Cahn eletrobalance and described as Follows: RUN #006. RUN #007. and RUN #008: 1.000 9 Fructose RUN #009: 1.000 g Fructose added with 1.000 g HPA RUN #010: 1.000 9 Fructose dissolved in 2.50 ml water added with RUN #011. RUN #012. RUN #013. RUN #014. RUN #015. RUN #016. and RUN #017: 1.000 g Fructose added with 1.000 g LZY zeolite LZY zeolite 51 III. EXPERIMENTAL PROCEDURE Fructose. LZY zeolite. and heteropoly acid were weighed on a Cahn eletrobalance. To purge air out and to maintain an inert atmOSphere were the same procedures as the Study 1 did. A steady state temperature was obtained in the oil bath. beFore the apparatus was set up. The reaction proceeded For a desired period oF time. AFter the end oF reaction. the solid residues along with the catalyst were added with 5.0 ml deionized. distillated water and briskly stirred until the solid residues were completely dissolved to be a dirty solution. A clean solution was obtained by Filtrating the dirty solution out oF the solid catalyst. A small amounts oF clean solution were diluted to a suitable concentration For each HPLC operating requirement. The determination oF Fructose in the aqueous solution was made by HPLC with the LDC 1107 reFractometer detector. The determination oF both HMF and levulinic acid in the aqueous solution was made by HPLC with the SF 770 UV detector at 220 nm wavelength. which was chosen From the UV spectrum on Perkin—Elmer Lambda 3 UV/VIS spectophotometer For the standard solution oF Fructose. HMF. and LA. The chromatogram and chromatographic data were automatically acquired and analized on IBM—9000 microcomputer system. 52 IV. ANALYTIC EQUIPMENT AND TECHNIQUES A. Perkin-Elmer Lambda 3 Spectrophotometer via Perkin-Elmer R-lOOA Chart Recorder an»usnneeueee.unenanossnnaeueuaneeouenneneeeaeeeee purpose: determination oF bestabsorbance wavelength For Fructose. LA and HMF principle: double-beam. UV-Visible spectrophotometer with a microcomputer control. which programs changes Tungsten—bromine lamp For visible light and Deuterium lamp For UV light. detector: side-window photomultiplier Operating procedure: 1. turn on the main power and turn on the record power 2. turn on the UV or VIS power 3. allow at least 30 minutes For instrument warm-up 4. select MODE button to select reading mode ( usuallly ABS. ABS means Absorbance.) select the desired SCAN SPEED (usually 60 nm/min) 5. press SAFE MEM until a " C " appears in the display 6. place solvent blank in both the reFerence and the sample cuvettes 7. press RUN For correction oF diFFerence in cuvettes 8. choose the wavelength limit 53 a. press " Lambda LIM " button. and enter the maximum wavelength limit you want b. press " Lambda LIM " button. and enter the minimum wavelength limit you want 9. choose the Full scale limit a. press " 0RD LIM " button. and enter the maximum ordinate limit you want b. press " ORD LIM " button. and enter the minimum ordinate limit you want 10. press " AUTO ZERO " button 11. select the chart speed (usually 60 mm/min) 12. adjust the pen position by pressing " PEN LIFT " a. LEFT or RIGHT press " L/R " For coarse adjustment press " ZERO ADJUST " button and Thumbwheel For Fine adjustment b. FORWARD or BACKWARD turn the thumbwheel For adjustment 13. place the sample in " SAMPLE " cuvette 14. press " RUN " Routine operation: 1. clean the sample cuvette and place the another sample into it 2. press " RUN " Shut-down I. clean the cuvettes and put them back 2. turn OFF the UV or VIS light 54 3. turn OFF the Record power 4. turn oFF the main power B. HPLC on BIORAD 42A and 87 P in series via LDC Model 1107 diFFerential reFractometer eenueoeeeueeseneuoeeaenunnnoguess.seasonaneenuunneueaeeuee detector principle: monitoring the quantitative diFFerence in the reFractive index between two liquids HPLC speciFications: column: Aminex HPX—42A and HPX—87P Heavy Metal in series. 300 x 7.8 mm For each mobil phase: deionized. distillated water (HPLC water) Flowrate: 0.6 ml/min temperature: 85 0C (isothermal) pressure: 450 to 600 psi (less than 1000 psi. sensitive to temperature oF column) inject volume: 30 ul to 50 ul suitable standard quantities: less than 0.5 mg Detector speciFications: attenuation: 2.0 trasmittance: 0.5 Operating Procedure: 1. Fill the eluant reservoir with degassed HPLC water 2. check the Haake water level 3. turn on the Haake column jacket circulator 55 9. turn on the Haake column jacket heater check whether a steady state temperature is 85 (L aFter 30 minutes switch on the reFractometer and allow at least 1 hr For instrument warm-up set the ReFractometer range at 2 and the transmittance to 0.5 via the Fine adjustment. turn on the HPLC pump ( already set at 23 Flowrate about 0.6 ml/min) allow 30 min to 1 hr to achieve stable baseline Routine Operation: 1. Open the data File on channel #2 with the method File: SUGARCOL For CAP OF IBM 9000 system neutralize the sample to be PH = 5.0 - 7.2 weigh the equal volume OF 1.0 mg/ml Inositol as the internal standard Flush the sample loop with 50 ul HPLC water inject 30 - 50 ul sample pull the manual inject bar From right to leFt switch channel #2 From ready to run End OF Run. pull the manual inject bar From leFt to right Shut-down: I. 2. Flush the sample loop with 50 ul HPLC water turn OFF the HPLC pump. the Haake circculator. and the Haake Heater 56 3. switch OFF the reFractometer The standard calibration For Fructose and inositol is shown in Figure 6. The standard calibration For glucose and inositol is shown in Figure 7. The standard calibration For sucrose and inositol is shown in Figure 8. One OF experimental serults. Run #017. is shown in Figure 9. Note Typical sugar retention time on this HPLC is shown as Follows: trimer ——————————— 18 min dimer ----------- 21 min sucrose ---------- 23 min glucose --------- 25 min xylose ----------- 27 min mannose ---------- 29 min Fructose --------- 30 min inositol ————————— 36 min C. HPLC Spectra-Physics SP 8000 via SchoeFFel SF 770 SpectroFlow Monitor {uuuaulniuuun*nuununnlunnuunuuanuunau purpose: qualitative and quantitative determination OF levulinic acid. HMF principles: SP-8000 is microprocessor controlled high perFormance liquid phase chromatograph. which programs runs OF parameter sets. temperature and mobil phase program. 57 HPLC speciFications: column: Zorbax ODS (Octadecyl SulFate) mobil phase: A: 0.13 % HeptaFluoro-Butyric acid (HFBA) B: 0.13 % HFBA + 80 % (v/v) acetonitrile programming: Time A % B % 0.0 100.0 0.0 15.0 70.0 30.0 20.0 70.0 30.0 25.0 100.0 0.0 temperature: room temperature (isothermal) pressure: above 1000 psi (retention time sensitive to press.) Flowrate: 0.5 ml/min inject volume: 30 to 50 ul suitable standard quantity: less than 0.5 mg UV detector speciFications: absorbancy: 0.4 wavelength: 220 nm Operating Procedure: start-up: 1. check solvent A. 8 level (reservoir A. B should be at least halF Full) 2. connect channel #4 box to IBM 9000 system 3. turn on the main power 4. turn on the UV/VIS detector power and allow 30 sec in START position. beFore switching in ON position 5. turn on the helium gas For degass solvents at 2 - 5 psi 58 .iaigi 6. sparge briskly the solvent For 5 min. then adjust to less than 10.0 ml/min 7. set UV/VIS absorbancy at 0.4 8. set UV/VIS wavelength at 220 nm 9. type M: and press RETURN (create the mobil phase no. 1) 10. type AB and press RETURN (select solvent A and B) 11. type 100 and press RETURN 12. type 15 and press RETURN 13. type 70 and press RETURN 14. type 20 and press RETURN 15. type 70 and press RETURN 16. type 25 and press RETURN 17. type 100 and press RETURN 18. type EX and press RETURN 19. type M11 and press RETURN 20. type F:0.5 and press RETURN 21. type QG and press RETURN 22. waiting until the constant Flowrate and ready light on 23. type GB and press RETURN to check the baseline type GX and press RETURN to end the baseline Routine Operation: 24. open the data File on channel #4 with the method: SUGAR For CAP OF IBM 9000 system 25 Filte the 70 ul sample solution by micropore Filter (0.45 um) 26. type "50" and press RETURN 27. Flush the sample loop with HFBA 59 LL __:‘_?i 28. type "SK" and press RETURN 29. type "50" and press RETURN 30. inject 30 to 50 ul sample 31. type "SK" and press RETURN 32. waiting For "pump marker " light to come on. then manually lower injection handle 33. end OF run . type EX.and press RETURN 34. pull the injection handle bar back aFter hearing twO clicks Shut-down 35. type "50 ". and press RETURN 36. Flush the sample loop with HPLC water 37. type "SK". and press RETURN 38. type F:0.0. and press RETURN 39. Turn OFF the main power and the detector power 40. Turn OFF the helium gas The standard calibration For levulinic acid and HMF is shown in Figure 10. One OF experimental results. Run #017. is shown in Figure 11. 0. IBM Instrument’s Chromatography Application Program on the microcomputer 9000 system *fliflflfllfl'fillill§*****I*I*****l§***§*Iflflfiifllflflfififlflli”l”Gil: purpose: acquiring. storing. and analyzing chromatographic data automatically 60 Operating Procedure: I. turn on the Disk Drive Power 2. insert Operating System Diskette. press Ctrl/Alt & Del 3. insert CAP diskette. type CAPMC I and press RETURN 4. insert the Data File diskette and press RETURN (into Chromatography mode) 5. press soFtkey EDIT (into edit channel) 6. create a method File For noncalibration. Fill the pages 1. 2. 3. 4. and 7 out For calibration. Fill the pages 5. 6 and Conc. Table out 7. press pad on the screen EXIT 8. press soFtkey READY 9. choose the channel number 10. Fill the data File name and the speciFications out 11. press pad on the screen EXIT 12. the channel will be automatically ready to acquire and storethe chromatographic data 61 Time 06:26:16 DIII'HON 26 NOV 04 RECONSTRUCT SCREEN DUHP Data Acquisition T:ne:20:07:08 Date UED 14 NOV 64 Hothod'SUCARCOL RANGE (MIM.)2 8.83 TO (5.88 FILE: DATAIBIJOHBBG SCALE: 1 473213 N V) E D K u-l O P M O t 9) 0-1 [.4 Z 3 O U T T I F ‘r i T I S 18 IS 28 25 38 35 48 15 MINUTES Inverse Response Factor: Fructose 1.92 x E-7 (ms/area) Inositol 1.62 x E-7 Figure 6. Chromatography of standard Fructose and Inositol 62 Time 06'14 25 Date HON 16 NOV 84 RECONSTRUCT SCREEN DUMP Data Acquisition Time;04:54:50 Dato:HON 26 NOV 84 Method : SUGARCOL FILE: onrnixzcznsasa SCALE: 1 RANGE (MIM.): 0.03 TO 45.00 .. “I 54304 8 g 5‘ A S ‘9 in O z ‘0 e z D o U 23070 W I M l l l l r l I l s 10 IS 20 25 30 as 40 45 MINUTES Figure 7. Chromatography of standard Glucose and Inositol 63 Tinni06:OI:JQ Dole MON 26 NOV 04 RECONSTRUCT SCREEN DUMP 'OQ:03.J7 Date NON 26 NOV 84 Data AcquIIIIion Tine. Method25UCARCOL PILC: onrnxszowosr aanz: I name: (flifl.): 0.03 70 15.00 67413‘ '4 o F m o z u U) 3 W t ‘3’ z w a O U a m o F o a m E 23290 lure-.1“ _. —~ L rg I I I T I I I F 5 18 15 28 25 36 35 48 45 MINUTES Figure 8. Chromatography of standard Sucrose and Inositol 64 T|n0:00:22:33 Deto:HON 26 NOV 84 RECONSTRUCT SCREEN DUHP Data Acquisition Tine;08:48:25 DatctTHU 15 NOV 00 Hoth0d35UCARCOL rlLE: onrnxa:Jou016 seats: 1 RANGE (MIM.): 0.03 T0 45.00 407341 4 O p w 0 Z W e Z 3 O U FRUCTOSE 23375 5 18 IS 28 25 38 35 48 45 MINUTES Figure 9. Chromatography of Run #017 65 Table 5. Area percent report from 183 9000 system A. Run #017 Channel 0 ...... REINT TIao:06:47:43 Doto:HON 26 NOV 84 Sample none ......... FRUCTOSEOLZY.I4O'C.I5hr 0616 III. ........... DATAIO:JOV016 Method name ......... SUGARCOL Author ......... JINDER JOU Instrument ..... HPLC SUGAR REFRACTOHETER Column ......... BIORAD 42A and 87? IN SERIES Notes .......... INOSITOL STD;INJ: 50uI;300-500uq;PH-5-7 INITIAL FRUCTOSE:.4nq; INOSITOL:..294mg Run time.......45.00 min. DoIay t1m0...0.00 min. Acq. (In. ..... 08:48:25 Acq. date....THU 15 NOV 84 Start PU ....... 20.00 sec. End PU ....... 20.00 soc. Slope sons ..... 3.00 uv/roc. Area roioct....50000 4 peaks Iound..23 AREA PERCENT REPORT Polk R.T.(m1n) RIS Peak none Area 5 Area Peak Ht. EL 1 9.614 X1 (9.6 min) 10.058 246888 251 88 2 20.611 X2 (20.6 min) 2.955 72530 577 VV 3 25.393 x: (GLUCOSE) 6.174 15153? 2015 BV 4 29.796 FRUCTOSE 3.418 83889 646 BB 5 35.258 INOSITOL 77.395 1899698 24520 BE TOTALS 100.000 2454542 66 Tlml:07136:35 DItQZHON 26 NOV 84 RECONSTRUCT SCREEN DUMP 0A1: ACQUI£|IlOn Tiue'05:05;24 Dot-:FRI 16 NOV 84 HethOdISUCAR FILE: DAT918:JOU828 SCALE: 1 RANGE (NHL): 8.82 To 45.88 168891 ¢ p-l 4. m M O I- Q 2 a 3 ‘- g 0 U 8 I I I I I I I l 5 18 15 28 25 36 35 48 4S HINUTES Figure 10. Chromatography of standard Fructose, LA. and HMF 67 Ttmn.07:l4 II 08!: HON 26 NOV 84 RECONSTRUCT SCREEN DUHP Ont: Acquillllon T1mc.06:57:36 Date VED 21 NOV 84 Hothod:SUCAR FILE: DATRIIZJOH854 SCRLEI I RANGE (flIN.)I 8.82 To 45.88 191355 ¢ 4 6 u M O F U D I L m k I D O U 8 I I I I I I I I S 18 IS 26 25 3B 35 46 4S NINUI‘ES Figure 11. Chromatography of Run #017 68 V. EXPERIMENTAL RESULTS Table 6: A summary of results For study 2 __----_-—------—-—----—---—-------—_---—-—_-----_—-—-----— run :cata.:solvent:t mp.:time: F : F 1 LA : HMF 1 X no I I II C) :(hr):conv.: 1 I 1 I 1 I 1 IBSETSEISSSS""SEmITBmIfi"TESTS"?TEXT-'77}; #007 none none 120 1.0 25.4 74.6 5.3 0.0 20.1 #008 none none 140 1.0 32.2 67.8 32.0 0.0 0.2 #009 HPA none 95 1.0 60.9 39.1 47.0 0.0 13.9 #010 LZY water 95 1.0 9.9 90.1 5.3 0.0 4.6 #011 LZY none 95 1.0 36.3 63.7 35.4 0.0 0.9 #012 LZY none 120 1.0 42.2 57.8 39.2 0.0 3.0 #013 LZY none 140 0.5 47.3 52.7 16.8 0.0 30.5 #014 LZY none 140 1.0 55.3 44.7 25.1 0.0 30.2 #015 LZY none 140 2.0 70.3 29.7 33.5 1.2 35.6 #016 LZY none 140 5.0 87.4 12.6 66.8 2.0 18.6 #017 LZY none 140 15.0 96.0 4.0 43.2 4.4 48.4 ISEQZ‘E’ESSCTTQ‘EEQ'ESSC2;;SSS-S;EZQSEQQZSFEEJEESQQT"’— F 1 is the component percentage of Fructose. LA 1 is the yield percentage of Levulinic acid. HMF 1 is the yield percentage of HMF X 1 is the yield percentage of unidentified products. All percentages are based on the initial Fructose of 1.0 gram. The ratio 0F catalyst to Fructose is 1.0. Water as a solvent is added 2.5 ml in run #010. HPA: heteropoly acid, H4[Siw On J 7H20 LZY: Linde Zeolite Y type (Faujasite). Mes/n[(Al02)t(SiOZ)Z] mHZO Wlth z/t > 2 69 Table 7: Reaction time eFFect For study 2 at 140 C with LZY run :catalyst:solvent:t mp.:time: F 1 LA : HMF no I I I( u) I(hr)I 1 I 1 I 1 #61572? """ 333;""IZBWSTEW’ZIT‘755""???- #014 LZY none 140 1.0 55.3 25.1 0.0 #015 LZY none 140 2.0 70.3 33.5 1.2 #016 LZY none 140 5.0 87.4 66.8 2.0 #017 LZY none 140 15.0 96.0 43 2 4.4 Note: F 1 is the conversion percentage oF Fructose. LA 1 is the yield percentage 0F Levulinic acid. HMF 1 is the yield percentage 0F 5—hydroxymethyl-2- Furaldehyde. The ratio 0F catalyst to Fructose is 1.0. All percentages are based on the initial Fructose OF 1.0 gram. HPA: heteropoly acid LZY: Linde Zeolite Y type (Faujasite). 70 Table 8: Temperature eFFect For study 2 For 1 hr with LZY run :catalyst:solvent:t mp.:time: F 1 LA I HMF no I I I( C) I(hr)I 1 I 1 I 1 #011 LZY none 95 1.0 36.3 35.4 0.0 #012 LZY none 120 1.0 42.2 39.2 0.0 #014 LZY none 140 1.0 55.3 25.1 0.0 Note: F 1 is the conversion percentage oF Fructose. LA 1 is the yield percentage oF Levulinic acid. HMF 1 is the yield percentage oF 5-hydroxymethyl—2— Furaldehyde. The ratio oF catalyst to Fructose is 1.0. All percentages are based on the initial Fructose oF 1.0 gram. LZY: Linde Zeolite Y type (Faujasite). 71 Table 9: Temperature eFFect For study 2 For 1 hr without LZY run :catalystIsolventhemp.:time: F : LA : HMF no I I I( C) I(hF)I 1 I 1 I 1 #006 none none 95 1.0 0.0 0.0 0.0 #007 none none 120 1.0 25.4 5.3 0.0 #008 none none 140 1.0 32.2 32.0 0.0 Note: F 1 is the conversion percentage oF Fructose. LA 1 is the yield percentage oF Levulinic acid. HMF 1 is the yield percentage oF S—hydroxymethyl—Z— Furaldehyde. The ratio oF catalyst to Fructose is 1.0. All percentages are based on the initial Fructose OF 1.0 gram. 72 Table 10: Catalyst eFFect For study 2 For 1 hr at 95 C run :catalyst:solvent:temp.:time: F 1 LA : HMF noI : :(Q) I(hr)I 7. : 1. : 7. I662‘73;SWEET"??-TYNE?""32'6"???" #011 LZY none 95 1.0 36.3 35.4 0.0 #009 HPA none 95 1.0 60.9 47.0 0.0 Note: F 1 is the conversion percentage oF Fructose. LA 1 is the yield percentage oF Levulinic acid. HMF 1 is the yield percentage oF S-hydroxymethyl—Z— Furaldehyde. The ratio oF catalyst to Fructose is 1.0. All percentages are based on the initial Fructose oF 1.0 gram. HPA: heteropoly acid LZY: Linde Zeolite Y type (Faujasite) 73 Table 11: Water eFFect For study 2 For 1 hr at 95 C with LZY run :catalystIsolvent1t6mp.:time: F 1 LA : HMF : no I I I( C) I(hr)I 1 I 1 I 1 I #010 LZY water 95 1.0 9.9 5.3 0.0 #011 LZY none 95 1.0 36.3 35.4 0.0 Note: F 1 is the conversion percentage oF Fructose. LA 1 is the yield percentage oF Levulinic acid. HMF 1 is the yield percentage oF 5-hydroxymethyl—2— Furaldehyde. The ratio oF catalyst to Fructose is 1.0. Water as a solvent is added 2.5 ml in run #010. All percentages are based on the initial Fructose oF 1.0 gram. LZY: Linde Zeolite Y type (Faujasite). 74 Table 12: Isomerization oF Fructose For study 2 reaction condition 1 yield percentage E3372;73;;12373232353232?""EILESEE’ZEETSS'QIET no I I I( C) I(hF)I 1 1662573.???33; """ EMITTM’MMSTS '''''''''''' #007 none none 120 1.0 4.3 #008 none none 140 1.0 4.8 #009 HPA none 95 1.0 0.0 #010 LZY water 95 1.0 9.8 #011 LZY none 95 1.0 3.3 #012 LZY none 120 1.0 8.3 #013 LZY none 140 0.5 2.3 #014 LZY none 140 1.0 11.3 #015 LZY none 140 2.0 7.0 #016 LZY none 140 5.0 6.8 #017 LZY none 140 15.0 5.9 Note: The ratio oF catalyst to Fructose is 1.0. Water as a solvent is added 2.5 ml in run #010. All percentages are based on the initial Fructose oF 1.0 gram. HPA: heteropoly acid LZY: Linde Zeolite Y type (Faujasite) 75 Table 13: Mass balance oF run #017 For diFFerent reject area retention time 1 1 (min) I (1) mg 7. I (2) mg ”I. LA 0.1732 43.3 0.1732 43.3 HMF 0.0176 4.4 0.0176 4.4 2.01 ----- 0.003 0.8 3.63 ————— 0.002 0.5 4.71 ————— 0.002 0.5 5.76 ————— 0.002 0.5 6.70 ----- 0.002 0.5 9.61 0.038 9.5 0.038 9.5 11.98 ————— 0.001 0.3 12.7 ————— 0.001 0.3 13.4 ----- 0.001 0.3 14.2 ----- 0.002 0.5 14.96 ————— 0.001 0.3 15.72 ----- 0.001 0.3 16.72 ————— 0.002 0.5 17.16 ————— 0.002 0.5 18.4 ————— 0.005 1.3 20.6 0.011 2.8 0.011 2.8 21.7 ----- 0.007 1.8 glucose 0.023 5.8 0.023 5.8 Fructose 0.016 4.0 0.016 4.0 38.7 ————— 0.003 0.8 42.5 ————— 0.006 1.5 total amount ( mg ) 0.279 0.324 1 70 1 81 1 Note: Compounds except HMF and LA were determined by HPLC ReFractometer by two diFFerent reject area: (1) reject area on HPLC with the reFractometer was set 50,000 and(2) reject area on HPLC with the reFractometer was set 5,000. HMF and LA in both (1) and (2) were determined by HPLC with the UV detector whose reject area was set 10,000. Total amounts oF initial Fructose in these sample oF run #017 are 0.4 mg. 76 m> mzm 62m .mmoosao opflaomu MNQ spas o 02H Pm asap powpommm mwmpcoopmm powmuo>coo .NH opswflm Ufio< aflcflap>mq Mo mwMPsmopmm eHme ppm mmovosnm mo 7 _ _ _ _ _ _ _ _ _ _ _ _ _ ILI IEII.I m. . . V. . O 4"— mmoosHo .: men .m mace oacfias>mq .m mmopospm .H 77 l. with LZY zeolite °k F 2. without catalyst - 1 (Z) . 05550:- (I L1>J I. 2 Z O O __ / U / ._ / 20 I / | / ‘l 'T I/ A /o J Lz’i ' 1 1 J 1 J 1 100 150 °C TEMPERATURE Figure 13. Conversion percentage of Fructose with and without LZY zeolite vs temperature for 1 hr reaction time 78 °h 5C) YIELD 2C) l. with LZY zeolite 2. without catalyst Figure 14. Yield percentage of Levulinic Acid with and without LZY zeolite vs temperature for 1 hr reaction time 79 DISCUSSION OF EXPERIMENTAL RESULTS 1. An evidence oF the dehydration reaction oF sugars over zeolites and heteropoly acids For study 1, only carbon dioxide was obtained in the gas phase. even though we modiFied the acidic aqueous sucrose solution with a NaX zeolite For Run #004 or a heteropoly acid For Run #005. The experimental results in study 1 showed that the value oF PH and the color oF the aqueous sucrose solution were due to the decomposition oF sucrose. From the literature reviews, it was observed that the solution oF sugars heated under the acidic medium would produce a yellow, Followed by a brown, and Finally a black viscous product. A. M. Taher35 had clearly reported that the yellow color was due to the Formation oF gamma- unsaturated, dicarbonyl compounds For the dehydration oF sugars (such as HMF For hexose and 2—Furaldehyde For pentose). It could be explained that the clear aqueous sucrose solution turned out to be yellow and viscous. It might be explained that a change in PH was due to the Formation oF soluble acidic products (such as levulinic acid and Formic acid). 80 II. Parameters oF the dehydration oF Fructose over solid acid catalysts The parameters in the dehydration oF sugars are:(l) characteristics oF catalysts, (2) types oF sugars. (3) the ratio oF catalyst to sugar, (4) temperature. (5) pressure. (6) reaction time. (7) solvent eFFect, and (8) an air or an inert atmosphere. This experiment was designed and run in twelve diFFerent conditions to explore the behavior oF the Following parameters: catalyst eFFect. temperature eFFect, reaction eFFect, and water as a solvent. The other reaction parameters were Fixed: an inert (nitrogen gas) atmosphere. the ratio oF catalyst to sugar = 1, atmospheric pressure, and Fructose as the sugar. 1. Temperature eFFect and catalyst eFFect: In Figure 13, it was shown that the conversion oF Frucose with LZY zeolite was higher than that without a catalyst, and both were proportional to temperature. But there was a drastic diFFerence below the melting point oF Fructose. Fructose conversion 0F 36 1 was obtained For the Former, but zero For the latter at 95 0C. In Figure 14, it was indicated that the Formation oF levulinic acid without a catalyst was proportional to temperature. But the Formation oF levulinic acid with LZY 81 zeolite, which was a concave curve, would pass a maximum about 110 0C - 120(3C. Two reasons For this are: (1) the Formation oF dehydrated products (such as Humin) parallel to levulinic acid would be much enhanced over a certain temperature (e.g. 110 (C - 120 (DC). and (2) the Further conversion oF levulinic acid might be much signiFicant over a certain temperature. Meanwhile, it was also showed that a drastic yield oF levulinic acid occured below the melting point oF Fructose using LZY zeolite in this case. As can be seen in Table 9, the conversion oF Fructose with heteropoly acid was higher than that with LZY zeolite. The yield oF levulinic acid From Fructose with heteropoly acid was higher than the yield oF levulinic acid From Fructose with LZY zeolite. But the selectivity oF levulinic acid with heteropoly acid was less than that with LZY zeolite. This could be explained by the two reasons mentioned in the above paragraph. It is implied that the more -acidic catalyst enhances much more the side reaction than the Formation oF levulinic acid. since the acidity oF heteropoly acid is higher than that oF LZY zeolite. 2. Reaction time eFFect: The conversion oF Fructose and the yield oF HMF would be proportional to the reaction time, but the yield oF HMF was only observed For reaction times 0F 2 hours or greater at 140 OC. The reaction time aFFected the yield oF levulinic acid in the same way as the temperature did. 82 3.Water as a solvent: It was apparently shown that water would reduce the conversion oF Fructose and the yield oF levulinic acid. Two reasons For this are: (1) water would reduce the acidity oF the reaction medium which caused low reactivity, and (2) the Fructose had high aFFinity toward water rather than toward the surFace oF the catalyst. The latter reason could apply to the dehydration oF Fructose using a strongly 21, 22 acidic ion—exchange resin in literatures. The same tendency in the homogeneous catalyst systems was reported by B. F. M. Kuster 19. Generally, sugars has insolubility in organic solvents but high solubility in 37 reported water. There are some good nonaqueous solvents For sugars, which are pyridine. N,N-dimethyi-Foramide, sulpholane, dimethylsulphoxide (DMSO), morpholine, r— butyrolactone, FurFuryl alcohol, tetrahydroFurFuryl alcohol, monoallyl ethers oF ethylene glycol, 2-methoxy ethanol, methyl carbionol, and dimethyl Formamide (DMF). But only DMSO as a solvent provided the stable yield oF HMF in the dehydration oF sugars. This phenomena has been 2 2 1' ,and Szmant 3 . who use 20 proven by Nakamura , Rigal ion—exchange resins and boron triFluoroide etherate as catalysts, separately. it seems that the solvent For sugar dehydration will reduce the reactivity oF the acidic catalyst. 83 Ill. SigniFicant discoveries The high yield oF levulinic acid was obtained by using LZY zeolite as a catalyst and Fructose as a sugar at the moderate temperatures. Specially, there was a drastic diFFerence For both the conversion rate oF Fructose and the yield rate oF levulinic acid below the melting point oF Frucotse with LZY zeolite. The isomerization oF Fructose to glucose occured in this nonsolvent dehydration reaction. There were three diFFerent kinetic models used in the isomerization reaction oF hexose . which were either an enolate-ion mechanism or a hydroxyl-ion dependent mechanism. The results might be a good explanation For the enolate-ion mechanism in the isomerization oF hexose due to the lack oF the hydroxyl ion in the reaction. The conversion oF Fructose in this nonsolvent system was not Fit to the First order conversion which was obtained in the solvent system. The order oF the conversion rate oF Fructose in this nonsolvent system oF our work was higher than that oF the solvent system reported in the literature using either an inorganic acid or a strongly acidic ion-exchange resin as a catalyst. For example, the conversion oF Fructose was carried out at 140 0C with LZY zeolite under an inert atmosphere. 84 Kinetic model: F -------- > products d[F]/dt = —K x [F]” {[F]1_n — [FoJI‘ni/[Foil'n = {(n—l)/[Fo]I'“} x K x t Let B = (n-l)/[Fo]l—n c = [F] / [Fo] ; [Fo] = 1.0 / 180.0 01'" — 1.0 = a x t No. t (hr) c (1) For 01'” = B x t + A 1 0.0 1.0 r = 0.9832 2 0.5 0.527 A = 1.000705 3 1.0 0.447 B = 0.07092 4 2.0 0.297 n = 2.648 5 5.0 0.126 6 15.0 0.04 so, K = 0.06226 mole-1‘65 / sec HMF did not exhibit the behavior oF a reaction intermediate in our work. but levulinic acid did. Also. isomerization oF hexose occured during the reaction. There might be another reaction scheme than that discussed in the literature reviews to explain the Formation oF levulinic acid and the isomerization in this nonsolvent system oF our work using solid acid catalysts. 85 A variation oF reaction scheme is proposed as Follows: F -------- > x <====> G -------- > 21 I (3) (4) I _________ > HMF --—--—--> Y —--——--> LA I (5) (6) --------- > LA -—-———--—> 22 I (7) ————————— > 23 F: Fructose, X: intermediate. G: glucose, Y: intermediate HMF: 5-hydroxymethyl—2-Furaldehyde. LA : levulinic acid 21, 22, and Z3 : either insoluble products or unidentiFied products (2) and (5) are more Favorable reactions than (3) in this nonsolvent system. (7), (6), and (3) will be promoted aFter increasing either temperature or reaction time. But (4) seems to be prohibited in this non-solvent system. 86 1V. Material Loss and Uncertainties in Data It was seen that there was still 20 1 weight loss For Run #017, even though all the trace unknown products were taken into account in Table 11. Two Facts could be explained For this: (1) there were brown to black undetermined insoluble residues (such as Humin. carbon, and copolymer oF Fructose and HMF) deposited upon catalyst, Filtrating out oF the yellow solution and (2) some soluble dehydrated products might be unable to be determined by these two HPLC’s. The amount oF levulinic acid and HMF was calculated by the area percentage method, but the amount oF glucose was calculated by the internal standard method. Material balances did not close For Runs #008, #010, #011, and #012, since the mass oF products including glucose exceeded the mass oF reactant. Two possible reasons For this are: (l) the isomerization oF Fructose to glucose was overestimated due to the overestimated conversion Factor oF glucose to the internal standard and (2) the yield oF levulinic acid was overmeasured due to the incorrect sample concentration. 87 CONCLUSIONS Sugars are major products in the sacchariFication oF iignocellusic materials. The eFFicient conversion oF sugars to high-value products is a very important step in the biomass-to—chemicals concept. A major question is : what is the "best" use oF sugars in production oF Fuels and chemicals. Today, Fermentation is used to convert sugars to alcohol. There are Four disadvantages oF Fermentation all adversely aFFecting economics: (1) long reaction time, (2) high energy requirements to separate the dilute product water system, (3) low carbon conversion, and (4) batch processing because oF inability to control all reactions. In past years, related research work has been done on the catalytic dehydration, in place oF Fermentation, oF sugars to chemicals (such as levulinic acid and HMF) using either an homogeneous acid or an ion-exchange resin as a catalyst. Some signiFicant improvements relative to Fermentation have made: (1) high carbon conversion, (2) lower reaction time. (3) no dilute medium required. (4) high yield and selectivity oF the intermediate dehydrated product (HMF), and (5) availability oF continuous process development. But the rate oF levulinic acid, the Final dehydrated product. was quite low in these researches. Our 88 research has positively shown that the high yield oF levulinic acid was obtained by using an LZY zeolite For a short reaction time at moderate temperatures. Especially, there was a drastic diFFerence in the conversion rate oF Fructose and the yield rate oF levulinic acid below the melting point oF Fructose with and without LZY catalyst. The conversion and yield rates were zero without catalyst below the melting point oF Fructose. The subsequent catalytic hydrogenation oF levulinic acid to alcohol may be Feasible due to the high reactive nature oF the carboxyl and keto groups. The modiFication oF the solid catalyst to catalyze the levulinic acid to alcohol reaction in a hydrogen atmosphere is an interesting topics For Further research. The inFluence oF water as a solvent highly decreased both the conversion rate oF Fructose and the yield oF levulinic acid. The same tendency was reported by B. F. H. Kusteéua. Generally. sugars has insolubility in organic solvents but high solubility in water. There are some good nonaqueous solvents reported For sugars. But only DMSO as a solvent For the dehydration oF sugars provided the stable 20 yield oF HMF. This phennomenon waizproven by Nakamura , 21, 22 Rigal , and and Szmant , who used ion-exchange resins and boron triFluoroide etherate as catalysts, separately. It seems that the solvent For sugar dehydration will reduce the reactivity oF the acidic catalyst. 89 HMF did not exhibit the behavior oF a reaction intermediate in our work. but levulinic acid did. Also. isomerization oF hexose occured during the reaction. There might be another reaction scheme than that discussed in the literature reviews to explain the Formation oF levulinic acid and the isomerization in this nonsolvent system using solid acid catalysts. The order oF the catalytic conversion rate oF Fructose was about 2.65 and the rate constant was 0.0623 mole-l / sec at 140 0C with LZY. The order oF the conversion rate oF Fructose in this nonsolvent system oF our work was higher than that oF the solvent system reported in the literatures using either an inorganic acids or strongly acidic ion- exchange resin catalysts. The increase oF temperature, reaction time, and acidity oF the catalyst highly enhances the Formation rate oF side reaction products such as humin and carbon over the Formation rate oF levulinic acid. But the yield oF levulinic acid may be optimized with either temperature or reaction time. This implies that at the maximum yield oF levulinic acid, there is a minimum oF side products such as humin and carbon. Further work on the optimization 0F these parameters is required For process development. 90 RECOMMENDATIONS This research has conFirmed the Feasibility and advantages oF the dehydration oF Fructose using solid acid catalystse. Further work is suggested as Follows in order to Fully develop the continuous catalytic dehydration oF sugars into chemicals. (1). chemical engineering Feasible study: (a). process Flow design and synthesis (b). material and energy balance (c). economic analysis (2). best catalyst selection: (a). diFFerent types oF solid acid catalysts (specially, zeolites and heterolpoy acids) (b). acidity eFFect For the same type oF catalyst (c). pore eFFect For the same type oF catalyst (3). examination oF various sugars For this system (a). Hexose: glucose, mannose, and galactose (b). Pentose: xylose and arabinose (c). Dimer: sucrose and maltose (4). examination oF starch, hemicellulose and cellulose 91 (5). kinetic investigation to evulate rate constant and rate expression (a). reaction time eFFect For a wild range oF temperature (b). eFFect oF pressure at a diFFerent temperature (c). the diFFerent ratio oF catalyst to sugar (d). the diFFerence between an air and an inert atmosphere (6). derivation and veriFication oF reaction scheme and kinetic model (7). determination oF insoluble products in the reaction (8). determination oF moles oF water produced to indicate the degree oF dehydration (9). examination oF various solvent eFFects (10). modiFication oF solid acid catalysts and reaction medium to Further convert levulinic acid to alcohols or ketones (a). Hydrogen as a carrier gas and reactant to proceed hydrogenation over metal catalyst (b). High temperature decarboxylation over solid acids 92 _H .. I I .1911. 1 II I 1M} . . .. i 1. . I iiIIIHIIuiI MILIiUHH The equipment which is required to execute this research is listed as Follows: (1). HPLC to determine the liquid products (2). good separating ability oF packed column For HPLC (3). TCD G.C. to determine the gas products (such as water, carbon dioxide, etc.) (4). Thermogravimeter to determine the insoluble products (5). PH meter to determine the acidity oF reaction (6). Automated data acquisition system 93 APPENDIX Calculations and procedures For experimental results are presented as Follows: 1. Calculation oF the response Factor For both standard Fructose and inositol on HPLC via LDC 1107 reFractometer. 2. Calculation For conversion oF Fructose and yield oF glucose on HPLC via LDC 1107 reFractometer. 3. Calculation oF the response Factor For standard Fructose, levulinic acid. and HMF on HPLC SP-BOOO via SF 770 UV detector. 4. Calculation For yield oF levulinic acid and HMF on HPLC SP—8000 via SF 770 UV detector. 5. Reintegration oF Run #017 on HPLC via LDC 1107 reFractometer set reject area: 5000 6. Properties oF Catalyst Note: All data Files are storedd in CAP oF IBM 9000 microcomputer system, MSU—DOE Plant Research Laboratory, Michigan State University 94 1. Calculation oF the response Factor For both standard Fructose and inositol on HPLC via LDC 1107 ReFractometer reject area: 50,000 set in the method Files (SUGAR) oF CAP oF IBM 9000 microcomputer system I FRUCTOSE I INOSITOL I I(l) (2) (3) I (4) (5) (6) I (7) Files amount area amount area I DATAIO: (mg) amount/area (mg) amount/area: JON006 0.373 1935885 1.93E—7 0.252 1585003 1.59E-7 0.824 JOW019 0.403 2080774 1.94E-7 0.250 1525430 1.64E-7 0.845 JOWOZO 0.050 263155 1.90E-7 These standard runs were designed to calculate the response Factors oF Fructose and inositol. (l) and (4) are known From the preparation. (2) and (5) are obtained From chromatographic data For each run. RF: the response Factor oF Fructose Ri: the response Factor oF inositol (3)=(1)/(2) : the inverse response Factor oF Fructose. (6)=(4)/(5) the inverse response Factor oF inositol. (7)=(6)/(3) R0; the conversion Factor For RF to Ri. For area percentage method, we average the inverse response Factor For Fructose, and inositol. -1 RF = 1.92E-7 and Ri—l=1.62E—7 They will be used in next part to calculate our experimental data. 95 For the internal standard method, we average the conversion Factor (Rc = 0.835), which is used to compare with area percentage method and calculate For unknown products. 2. Calculation For the conversion oF Fructose and yield oF glucose on HPLC via LDC 1107 ReFractometer reject area: 50000 set in the method File (SUGAR) oF CAP oF IBM-9000 system. Run No. Data File area amount(mg) conversion (1) ISSEmBXIXIBIJSQSS7'2651'232"“636I’m-mm???” #007 DATA10:JOWOO9 1553331 0.298 25.4 #008 DATA10:JON010 1411958 0.271 32.2 #009 DATAlO:JOW008 814896 0.159 60.9 #010 DATAIO:JOW012 1876441 0.360 9.9 #011 DATAIO:JOH018 1326204 0.255 36.3 #012 DATAII:JOW028 1203029 0.231 42.2 #013 DATA11:J0w053 1171950 0.189 47.3 #014 DATA10:J0w017 931131 0.179 55.3 #015 DATAII:JOW055 618335 0.119 70.3 #016 DATAlO:JOH015 263156 0.051 87.4 #017 DATAIO:JOW016 83889 0.016 96.0 (8) is obtained From the chromatographic data. 96 1 (9) = (8) x RF- (10) = [0.400 - (9) ]/ 0.400 x 100 1 IFRUCTOSEI INOSITOL I FRUCTOSE ESE-’63:}?72"”;§;;"";;;; ”””” ;;SSSE';;;SEE'ESSC;QISE No. (8) (11) (12)(mg)(13)(m9) (l4) 1 41632“BEES13613687268125;'IEQESSI’ITEE’ITSQE """ £76" #007 DATA10:JOW009 1553331 1748548 0.273 0.291 27.3 #008 DATA10:J0w010 1411958 1684892 0.264 0.263 34.3 #009 DATAIO:JOW008 814896 none none none none #010 DATAIO:JOW012 1876441 1573563 0.248 0.354 11.5 #011 DATAIO:JOW018 1326204 1325509 0.21 0.252 37.0 #012 DATA11:J0w028 1203029 none none none none #013 DATA11:J0w053 1171950 none none none none #014 DATA10:J0w017 931131 1753830 0.276 0.174 56.5 #015 DATAII:JOW055 618335 none none none none #016 DATA10:JOWOIS 263156 1601295 0.259 0.51 87.3 #017 DATA10:JOWOI6 83889 1899698 0.294 0.015 96.3 (8) and (11) are obtained From the chromatographic data. (12) Rc: (13) (14) is known by preparation. [ 0.400 ~ (8) x (12) / (11) / RC : conversion 1 oF Fructose 97 conversion Factor oF Fructose to inositol (12) I / 0.400 x 100 1 3 amount oF Fructose Run Data File area area amount 1 No. (mg) I inositol: X3 (glucose) I (11) I (15) (16) (17) #006 DATA10:JOWOO7 1685997 none none none #007 DATA10:JOW009 1748548 110983 0.017 4.3 #008 DATA10:J0w010 1684892 123724 0.019 4.8 #010 DATAIO:JOW012 1573563 246086 0.039 9.8 #011 DATAIO:JOW018 1325509 84980 0.013 3.3 #014 DATAIO:JOW017 1753830 285476 0.045 11.3 #016 DATAIO:JOW015 1601295 166351 0.027 6.8 #017 DATA10:J0w016 1899698 151537 0.023 5.9 (11) and (15) are obtained From the chromatographic data. (12) is the same as the above case. Rc: conversion Factor oF glucose to inositol; assumed 1. (16) (12) x (15) / (11) / Rc : the amount oF glucose (17) (16) / 0.400 x 100 1 ; yield 1 oF glucose 0.400 mg: total amount oF initial Fructose in these samples AREA PERCENT METHOD FOR YIELD CALCULATION OF UNKNOWN PRODUCT Run Data File area amount 1 No. (m9) #009 DATA10:JOW008 none #012 DATA11:JON028 216688 0.035 8.8 #013 DATAII:JOW053 57802 0.009 2.3 #015 DATA11:JOW055 186147 0.028 7.0 (18) x Ri-1 : amount oF glucose A p—A to v 11 (20) (19) / 0.400 x 100 1 ; Yield 1 oF glucose 0.400 mg: total amount oF initial Fructose in these samples 98 3. Calculation oF the response Factor For standard Fructose, levulinic acid. and HMF on HPLC SP-8000 via SF 770 UV detector reject area: 10,000 set in the method File (SUGARCOL) oF CAP IBM 9000 microcomputer system Data File amount area amount/area (mg) (21) (22) (23) DATA10:ZEPM026 0.05 9154698 5.46lE-9 DATA10:ZEPM027 0.05 8704727 5.744E-9 DATA10:J0ww021 0.05 7903008 6.327E-9 (23) = (21) / (22):the inverse response Factor oF Fructose I (23) RF1_ For the area percentage method, the inverse response Factor oF Fructose is the average oF summation 0F (23). That is RFl‘l =5.844E-9. data File amount area amount/area (mg) """mm'"7227"""7237"-"mmm'Eééi """""" BXIXISZEEEESEE‘BTSEm‘"’IESSEQIE """""" 3353;; """" DATA10:JOWOOI 0.05 12005791 4.164E-9 (26) = (24) / (25); the inverse response Factor oF levulinic acid ; Ri.-l 99 For the area percentage method, the inverse response Factor oF levulinic acid is the average oF summation 0F (26). 1 That is Rl— 4.019e-9. data File amount area amount/area (mg) m"""""'""IE§T"‘""Z£§§”“"“""7233' """"" SEEIEEEESEE‘ITBSE """""" éSEEIEE'""""ETQZEEIIS """ DATA10:ZEPM027 0.005 8132644 6.148E-10 DATA10:ZEPM028 0.005 8767719 5.703E-10 DATA10:JOWOO3 0.003 5114199 5.866E—10 DATAIO:JOW004 0.001 1605398 6.229E—10 DATA10:JOWO43 0.005 8856750 5.645E-10 (29) = (27) / (28): the inverse response Factor oF HMF:Rh-— 1 For the area percentage method, the inverse response Factor oF HMF is the average oF summation 0F (29). 1 That is Rh- = 6.042E-10. 100 4. Calculation For yield oF levulinic acid and HMF on HPLC via LDC 1107 ReFractometer. area reject: 10.000 set in the method File (SUGARCOL) oF CAP oF IBM 9000 microcomputer system Run Data Fil yield No. DATAIl e Total area area amount yield area amount #006 JOWOSO #007 J0w056 #008 Jow045 #009 Jow044 #010 Jow047 #011 J0w048 #012 J0w046 #013 JOWOSZ #014 Jow031 #015 Jow054 #016 Jow029 #017 JON027 15242128 7030498 9834814 9209231 8365761 9857799 9824472 7544762 6950958 6721986 9401160 5722627 0 656579 4038398 5851160 665725 4403608 4883616 2090570 3122329 4176696 8310321 5380403 0.0 0.0027 0.0162 0.0235 0.0027 0.0177 0.0196 0.0084 0.0125 0.0168 0.033 0.0216 (1) (mg) (1) 0.0 0 0.0 0.0 5.3 0 0.0 0.0 32.5 0 0.0 0.0 47.0 0 0.0 0.0 5.4 0 0.0 0.0 35.4 0 0.0 0.0 39.2 0 0.0 0.0 16.8 0 0.0 0.0 25.1 0 0.0 0.0 33.5 990361 0.0006 1.2 66.8 1711646 0.001 2.0 43.2 3672184 0.0022 4.4 (30) and (34) are obtained From the (10) is given From part two (31) = subtotal area oF LA 1 (32) = R1" x (31); 1 (35): (34) x Rh" ; (33) = [ 0.05 — (36) = [ 0.05 — 0.05mg: total amount oF the amount oF LA (32) ] / 0.05 : (35) J / 0.05 : initial chromatographic data. : conversion 1 oF Fructose 101 I (30) — [1.0 — (10)] x 0.05 / RFI‘ ; the amount oF HMF the yield 1 oF LA the yield 1 oF HMF Fructose in these samples 5. Reintegration oF Run #017 on HPLC via LDC 1107 ReFractometer area reject: 5.000 set in the method File (SUGAR) oF CAP oF IBM 9000 microcomputer system Retention time (min) area amount yield (1) (37) (38) (39) (40) 2.01 22492 0.003 0.8 3.63 10284 0.002 0.5 4.71 15312 0.002 0.5 5.76 15634 0.002 0.5 6.70 15432 0.002 0.5 9.61 246888 0.038 9.5 X1 11.98 8596 0.001 0.3 12.7 8828 0.001 0.3 13.4 9356 0.001 0.3 14.2 9946 0.002 0.5 14.96 9542 0.001 0.3 15.72 8508 0.001 0.3 16.72 9754 0.002 0.5 17.16 13303 0.002 0.5 18.4 32698 0.005 1.3 20.6 72530 0.011 2.8 X2 21.7 42339 0.007 1.8 25.39 151537 0.023 5.8 glucose (X3) 27.57 14937 0.002 0.5 29.8 83889 0.016 4.0 Fructose 38.7 22265 0.003 0.8 42.5 39541 0.006 1.5 (37) and (38) are obtained From the chromatographic data. (39) = (38) x Ri.l ; amount oF each compound (40)=[0.05 - (39)]/0.05 : yield percentage oF each compound 0.05 mg: total amount oF initial Fructose in these samples 102 6. Properties oF Catalyst A. Heteropoly acid General Formula: Hq [xmnoy] (usually x < m) X: center atoms (hetroatoms) : P, Si, Te, As. Mn M: coordinated atoms (polyatoms) ; Mo. w. V, Nb x:m = 1:12. 1:11. 1:10, 1:9, and 1:6 H replaced by the metal ion called salt oF heteropoly acids General properties: high molecular weight electrolytes over 4000. signiFicantly soluble in water and organic solvents. strong acid and protons have the same dissociation constant. strong oxidizing agents which change to blue color upon reduction Free acids as well as salts contain many molecules oF water oF crystallization. decomposed by strong base. heteropoly acids show brilliant colorations. 103 B. Zeolite: General Formula: M x/n [ (A102)x (Si02)y] w H O n the charge oF the cation w numbers oF hydration oF water on the structure y/x : From zero to inFinite General properties: 1. reversible dehydration: dehydration oF Bronsted acid to Lewis acid 2. ion exchange property. 3. molecular seiving catalyst. 4. high selectivity 0F reactant, product and restricted transition state. 5. high stability (i.e. high Si/Al,high stability .low acidity) 6. highly crystalline and high surFace area 104 8. 9. 10. II. 12. 13. 14. 15. REFERENCES Chang, Clarence D. and Silvestri. Anthony J., Journal oF Catalysis, 47. 249, (1977) Leonard. Reid H., Ind. Eng. Chem.,48 (8), 1331. (1956) Harris, John F. and Feather, Milton 5., Advan. Carbohydr. Chem. Biochem., 28. 161, (1973) McKibbins. S. w.. Ph. D. Thesis, Univ. oF Wisconsin (1958) Speck, J. C., Jr., Advan. Carbohydr. Chem. 13, 63, (1958) Feather. M. 5.. Carbohydr. Res.. 7. 86. (1968) Shaw, Philip E.. Carbohydr. Res.. 5, 266. (1967) Amin. EL 5., Carbohydr. Res.. 4, 96, (1967) Heyns, K., Carbohydr. Res. 6. 436. (1968) Krol. B., Acta Aliment. Pol., 4, 287. (1978) Bonner. T. G., J. Chem. Soc.. 1960. 787, (1960) Anet, E. F. L. J., Advan. Carbohydr. Chem.. 19. 181, (1964) Haworth. w. N. and Jones. w. G. M.. J. Chem. Sco.. 1944. 667. (1964) Harris, D. w. and Feather. M. 8., J. Org. Chem.. 39. 724, (1974) Rase, Howard F., Ind. Eng. Chem. Prod. Res. Dev. 14. 40. (1975) 105 l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. Bonner. T. G., Bowrne, E. J. and Ruszkieuier. M.. J. Chem. Soc., 1944, 667. (1944) Wiggins, L. F., Advan. Carbohydr. Chem.. 4. 293, (1949) McKibbin, 5.:Harris, J.: Saeman, J. and Neil. H., Forest Product J., Jan. 1962, 17. (1962) Kuster, Ben F. M. and Feather, L. M.. Carbohydr. Res.. 54. 159. (1977) Nakamura,Yoshio, and Morikawa, Shunichi, Bull. Chem. Soc. Jpn, 53. 3705. (1980) Rigal, Luc, J. Chem. Tech. Biotechnol., 31. 489. (I981) Rigal. Luc. Ind. Eng. Chem. Prod. Res. Dev., 20. 719. (1981) Szmant.H. H. and Chundury, D. D.. J. Chem. Tech. Biotechnol.. 31. 135. (1981) Van Einenstein, w. A. and Blanksma. J., J. Chem. Weeklad, 6. 717. (1909) Wiggin, L. F. and Montgomery. R., J. Soc. Chem. Ind. Lond..66. 31. (1945) Stone, J. E. and Blundell. M. J., Can.. J. Res.. 28. 676. (1950) Rice. F. A. H., J. Org. Chem.. 23. 465. (1958) Mednick, M. L.. J. Org. Chem.. 27, 398. (1962) Moye. C. F. and Goldsck, R. F., J. Appl. Chem.. 16, 207. (1966) Maier, N. F.; Roth, w.; Thies I. and Schleyer P. R., Chem. Ber., 115, 808, (1982) 106 31. 32. 33. 34. 35. 36. 37. 38. Chiba. T.; Okimoto, M. and Nagai. H..Bull. Chem. Soc. Jpn..56. 719. (I983) Chum. H. L.: RatcllFF M.: Posey F. L.: Turner J. A. and Nozek, A. J., J. Phys. Chem.. 87, 3089, (1983) Otake, Masaayaki and Onoda, Takeru. J. Catalysis, 38. 494, (1975) Jolly, William L.. " the systhesis and charaterization oF Inorganic Compounds", 32. 460. (1970) Taher, A. M. and Cutes. D. M.. Carbohydrate Research, 34, 249. (1974) Kooyman. C.: Vellenga. K. and Dewilt. H. G. J., Carbohydrate Research, 54. 33. (1977) Moye, C. J., and Smythe, B. M.. Carbohydrate Research, 1. 284. (1965) Dull, G. Chem. Ztg.. 19. 1003, (1895) 107 STAT 11111111 1293 0 an/ERSIHI lelriAR’ms 3142 3928 “7111111311