V . £5 .. J I. .. .fimfiamwfigwa. .2. .hwwmaiu A. . u. .. u... .Mwmfl rg. . $§¢,%m $3.2, flyingfinwg .mfifiw, .42... I. ‘ . 5 .i .. .er7 Ev :- 3!}... .r: as )9. v . « p 1 r. .u. a x...» . It.» .whyyfl». 3 .1 . 3 2 “WV“ .11. . a «thbnmwant lasnuhunuwsfi . A. .i '1 r . I ~ 1 .1... {0| 4 . .237»; v... ‘2???" i9, 5. «2.62... 2.... 7) . .1... 1; a 1 v van.” fififlmafi. 31F; x a 2.1..- ‘ .‘3'1I0‘l’ ‘0 1:... I W! . iiifi 1:...» 5?... .3... a... .. 351$ a»: .x ‘ ‘35 n5 1- ain”. , e P. i 4?..an z. r .l... on... . . cx?‘ .. 1‘ tuna...) .. . {I x r Inaugural . .C a, 3n . 3. BS.) Ill titan: ‘1 A c b ~IO.... Halt....t‘ ‘n i... a)": , .. 3.1.1.1112: A 24‘ .JWfiWii g“. 2:; 3.52 1:! ,1. .111: ii). tof‘lvvu‘ .3 .510». . .3 I q. . l- .. .21 .n. 15,990. k Dru! ‘Ln:u.\?. _ lllllllllllllllllIlll'lllllllllllllllll 3 1293 01020 1576 This is to certify that the thesis entitled Yield and Quality of Onion Flavor Oil Obtained by Supercritical Fluid Extraction and Other Methods. presented by Aditi Dron has been accepted towards fulfillment of the requirements for MS degree in Agricultural Engineering wow-wit ' Major professor Date Ma. 9) H97 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution .r ‘4Vf __.4——-—--' LIBRARY Mlchlgan State University PLAcE "RETURN BOXtorunmthI-chockomfiomymm To AVOID FINES Mum on or bdoro «Id-duo I DATE DUE DATE DUE DATE DUE YIELD AND QUALITY OF ONION FLAVOR OIL OBTAINED BY SUPERCRITICAL FLUID EXTRACTION AND OTHER METHODS By Aditi Dron A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Engineering 1994 ABSTRACT YIELD AND QUALITY OF ONION FLAVOR OIL OBTAINED BY SUPERCRITICAL FLUID EXTRACTION AND OTHER METHODS By Aditi Dron Methods of extraction of onion flavor oil were studied including supercritical fluid extraction using carbon dioxide (C02), liquid C02 extraction and steam distillation-solvent extraction. The effect of using entrainers with supercritical fluid extraction was also studied. The yield and the quality of onion extracts obtained from the different methods were compared. The maximum yield of 0.0324% was obtained by supercritical C02 extraction at 3600 psi (24.5 MPa), 37°C at a C02 flow rate of 0.5 1/min. Ethyl alcohol used as entrainer enhanced the yield of onion oil over that obtained by supercritical C02 experiment without entrainer at the C02 flow rate of 1.0 l/ min. Gas chromatography and combined gas chromatography-mass spectrometry of the extracts indicated that the flavor profiles were different for extracts obtained by different methods. Supercritical and liquid C02 extracts had fresh onion-like flavor as opposed to a cooked flavor of the extract obtained by steam distillation-solvent extraction. DEDICATED TO MY PARENTS iii ACKNOWLEDGNIENTS Sincere thanks to Dr. Daniel E. Guyer, my major professor for guidance, support and encouragement. I am also grateful to my guidance committee, Dr. Fred W. Bakker-Arkema, Dr. Douglas A. Gage and Dr. Carl T. Lira for their suggestions and contributions. The research was funded by the Research Excellence Fund through the Crop and Food Bioprocessing Center and by the Michigan Onion Industry. Thanks for their financial support that made this work possible. I would like to thank Dr. Jerry N. Cash and Dr. J. Ian Gray for allowing me to use their lab facilities. I also want to thank Dr. Ninnal K. Sinha for his help and suggestions. I want to thank the Michigan State University Mass Spectral Facility for facilitating the MS analysis. I want to thank my parents, brother and sister for their love and encouragement. Thanks to my friends: Krunali, Umashankar, Shanti Swaroop, Sondes, Noi, Jeff, Philip and many others who have made my stay at MSU very enjoyable. I want to specially thank my husband, Vijay, for his love, support, help and patience. iv TABLE OF CONTENTS LIST OF TABLES ..................................................... viii LIST OF FIGURES ..................................................... ix 1 INTRODUCTION ..................................................... l 1.1 Overview ..................................................... l 1.2 Objectives .................................................... 3 2 LITERATURE REVIEW ................................................ 4 2.1 Onion Chemistry ............................................... 4 2.1.1 Development of Flavor in Onion ........................... 4 2.1.2 Properties of Onion Oil ................................... 5 2.2 Supercritical Fluid Extraction ..................................... 7 2.2.1 Properties of Supercritical Fluids ........................... 7 2.2.2 Supercritical Fluid Extraction Process ....................... 9 2.2.2.1 How the Process Works ........................... 9 2.2.2.2 System Components ............................. 9 2.2.2.3 Advantages of Supercritical Fluid Extraction ......... 11 2.2.2.4 Historical Developments and Applications ........... 12 2.3 Use of Entrainer with Supercritical Fluid Extraction .................. 14 2.4 Extraction with Liquefied Gases .................................. 16 2.5 Traditional Methods of Extraction of Flavor Concentrates .............. 17 2.5.1 Steam Distillation ...................................... 18 2.5.2 Classical Solvent Extraction .............................. 19 2.6 Analytical Methods Used for Assessment of Quality of Extracts ......... 19 2.6.1 Gas Chromatography and Combined Gas Chromatography-Mass Spectrometry .......................................... 19 2.6.2 Headspace Volatiles Analysis ............................. 20 2.6.3 Other Methods ........................................ 21 3 MATERIALS AND METHODS ......................................... 22 3.1 Onions and Juice Preparation ..................................... 22 3.2 Extractions ................................................... 22 3.2.1 Supercritical C02 Extraction ............................. 22 3.2.2 Supercritical C02 Extraction with Entrainer ................. 23 3.2.3 Liquid C02 Extraction .................................. 27 3.2.4 Steam Distillation-Solvent Extraction ...................... 27 3.3 Analysis ..................................................... 29 3.3.1 Estimation of Onion Oil Yield ............................ 29 3.3.1.1 Gravimetric Method ............................. 29 3.3.1.2 Quantitative Gas Chromatographic Analysis .......... 29 3.3.2 Estimation of Quality of Extracts .......................... 31 3.3.2.1 Headspace Volatiles Analysis ..................... 31 3.3.2.2 Thiosulfinate Analysis ........................... 31 3.3.2.3 Gas Chromatography ............................ 32 3.3.2.4 Gas Chromatography-Mass Spectrometry ............ 32 3.3.2.5 Experiments on Post-Extraction Residue and Fresh Juice 33 4 RESULTS AND DISCUSSION .......................................... 34 4.1 Quantitative Analysis ........................................... 34 4.1.1 Gravimetric Yield of Onion Oil from Different Methods ....... 34 4.1.2 Quantitative GC Analysis for Comparison of Onion Oil Yield from All Methods .......................................... 35 4.1.3 Effect of Volume of C02 Passed Through the Extraction System on the Yield of Onion Oil .................................. 37 4.1.4 Effect of Entrainer ...................................... 39 4.1.4.1 Effect of Polar Entrainer ......................... 39 4.1.4.2 Effect of Non-polar Entrainer ..................... 41 4.1.4.3 Effect of Entrainer at Different Pressures ............ 41 4.2 Qualitative Analysis ............................................ 42 vi 4.2.1 Quality of Extracts Obtained from Various Methods ........... 42 5 SUMMARY & CONCLUSIONS ......................................... 52 6 SUGGESTIONS FOR FURTHER RESEARCH ............................. 54 7 BIBLIOGRAPHY ..................................................... 55 APPENDIX A ......................................................... 66 APPENDIX B ......................................................... 74 APPENDIX C ......................................................... 79 APPENDIX D ......................................................... 82 vii Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 LIST OF TABLES Gravimetric Yield of Onion Oil Obtained from Different Methods ........ 34 Onion Flavor Compounds Identified by GC-MS in Supercritical C02 Extracts and Headspace Samples ......................................... 44 Results of Thiosulfmate Analysis. ................................. 48 Area Percent Report of SFE-C02 + Ethanol (75 m1, 3600 psi, 37°C) Extract obtained by GC ............................................... 70 Peak Areas of Sulfur Compounds and Internal Standard for SEE-€02 + Ethanol (75 ml, 3600 psi, 37°C) ......................................... 72 Mass Spectral and Retention Data for Onion Flavor Compounds Detected in Various Extracts Studied ........................................ 74 viii Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 LIST OF FIGURES Pressure-Temperature Diagram of a Pure Component. From McHugh and Krukonis (1994). .............................................. 8 Supercritical Fluid (SCF) Extraction System ....................... 10 Supercritical C02 Extraction System ............................. 24 Supercritical C02 Extraction System with Modifications for Adding Entrainer .................................................... 26 A Modified Likens-Nickerson Apparatus for Distillation-Solvent Extraction ............................................................ 28 Total Amount of Sulfur Compounds in the Extracts Obtained by Quantitative GC Analysis ................................................. 36 Yield of Onion Oil versus Volume of C02 passed at 3600 psi, 37°C. . . . . 38 Yield of Onion Oil at Different Volumes of C02 Passed at Different Flow Rates of C02 ................................................. 40 Gas Chromatogram of Onion Oil Obtained by Supercritical C02 Extraction ............................................................ 50 Gas Chromatogram of Commercial Onion Oil Obtained by Steam Distillation. ................................................. 50 Gas Chromatogram of a Dichloromethane Extract of Fresh Onion Juice . . 51 Gas Chromatogram of a Dichloromethane Extract of Onion Juice after SFE— C02. ....................................................... 51 Total Ion Chromatogram of SEE-€02 + Ethanol (75 ml, 3600 psi, 37°C) Extract ..................................................... 67 Gas Chromatogram of SFE-COZ + Ethanol (75 ml, 3600 psi, 37°C) Extract Showing the Selected Peaks for Quantitative GC Analysis ............. 68 Figure 15 Gas Chromatogram of SFE-COZ + Ethanol (75 ml, 3600 psi, 37°C) Extract. Peak Numbers correspond with Table 6 ............................ 78 1. INTRODUCTION 1.1 OVERVIEW Onions (Allium cepa L.) and other members of the genus Allium are among the oldest of the cultivated plants. Their origins, most likely in Central Asia, predate written history (Block, 1985). They possess strong, characteristic aromas and flavors which have made them important ingredients of food. A remarkable property of this genus is that most members have no odor unless the plant tissue is cut or otherwise damaged. Stoll and Seebeck (1951) found that these characteristic volatiles are absent from intact tissues and that the volatiles are enzymatically produced when injury occurs. Apart from their use in food, certain extracts of onions and garlic have been found to be antibacterial, antifungal and antithrombotic (Block, 1985). Onion and onion flavors (onion oil) are important seasonings widely used in food processing. Currently the majority of onion oils used in US. food production are imported. These products, which are usually steam distilled, lack the fresh onion flavor. Their quality varies depending on the origin of production and is sometimes undesirable or inconsistent. Thus, for production of onion flavor oil on a commercial scale, a detailed study is required into the various possible methods for extraction and their comparative evaluation for yield and product quality. Extraction using carbon dioxide (C02) as solvent is gaining attention because it is nontoxic, easily separated from the extract, non-flammable, inexpensive and available in high purity. C02 in liquid state has been used for extraction of many natural 2 products (Schultz and Randall, 1970). In addition, C02 in supercritical state is also being used for extraction of natural products. The critical temperature and pressure of C02 are 31°C and 1070.7 psi (7.4 MPa). Thus, supercritical C02 extractions can be carried out under relatively moderate conditions with minimal degradation of thermally labile flavor components (Rizvi et al., 1986). This technique has the ability to change and "fine tune" its solubilizing power by controlling pressure and temperature. Supercritical C02 is generally a more powerful solvent than liquid C02. However, liquid C02 tends to be more selective (Grimmett, 1981). While supercritical C02 has many desirable properties, its polarizability is very low. Therefore, small amounts of co-solvents, which are referred to as modifiers or entrainers, may be added to modify the polarity and solvent strength of supercritical C02 to increase the solute solubility (and/or selectivity). The entrainers used are commonly polar or non-polar organic compounds which are miscible with supercritical coz. This study was undertaken to research various methods for extraction of onion flavor oil and to compare the yield and quality of the product obtained by each method with the goal of evaluating SFE-C02 (supercritical fluid extraction using C02) feasibility in extracting a unique fresh onion flavor. 1.2 OBJECTIVES 1. Compare the yields of onion flavor oil obtained by Supercritical C02 Extraction, Liquid C02 Extraction, and Steam Distillation-Solvent Extraction under the conditions studied. 2. Investigate the effect of using entrainers with Supercritical Fluid Extraction on yield of onion flavor oil. 3. Compare the quality of extracts obtained by the various methods. 2. LITERATURE REVIEW 2.1 ONION CHEMISTRY 2.1.1 Development of Flavor in Onion The characteristic flavor of onions comes primarily from volatile organic sulfur compounds released enzymatically by the action of allinase (alliin alkyl-sulfenate- lyase; EC 4.4.1.4) on several nonvolatile, odorless amino acid precursors, namely (+)—S- methyl-, (+)-S-propyl-, and trans-(+)-S-l-propenyl-L—cysteine sulfoxides, when the onion bulbs are chopped or crushed (Whitaker, 1976). The primary reaction products of these amino acids are thiosulfinates, which dissociate to produce various sulfides containing methyl, propyl, and propenyl groups, thiophene derivatives, and other sulfur- containing heterocycles (Carson, 1987). The formation of various sulfur compounds can be summarized as follows: Allinase S-alkyl cysteine sulfoxide % NH3 + CH3COCOOH (amino acid precursor) + RS OH 2 R S O H ——’ RSSOR (Thiosulfinate) RSSR + R8802R (Thiosulfonate) 2 R S S O R ——< RSSR + RSR + 802 2 RSSR ———> RSSSR + RSR S The alkyl thiosulfonates (methyl methane-, propyl methane-, and propyl propanethiosulfonates) have been associated with fresh onion-like flavors, while propyl- and propenyl- containing di- and trisulfides have been associated with cooked onions or steam distilled onion oils (Boelens et al., 1971). Previous studies (Brodnitz et al., 1969; Boelens et al., 1971; Mazza et al., 1980; Kallio and Salorinne, 1990; Kuo et al., 1990) have reported flavor components of head space, solvent extracts, and distilled oils. Block (1985) reviewed the chemistry of garlic and onion. Block et al. (1986) reported synthesis of antithrombotic organosulfur compounds from garlic. 2.1.2 Properties of Onion Oil Fenaroli (1971) described the onion oil as a yellowish liquid with a characteristic onion odor whose main constituents are di-n-propyl and methyl n-propyl disulfide with a specific gravity of 1047-1098 and solubility of 1:10 (in 90% ethyl alcohol). Fenwick and Hanley (1985) described onion oil as a brown-amber liquid obtained in 0.002 to 0.03% yield by the distillation of minced onions which had been allowed to stand for some hours prior to distillation. It was reported that the oil comprises a complex mixture of (mainly) sulfur containing volatiles. The product possesses (on a weight basis) 800 to 1000 times the strength of odor of fresh onions, but its commercial value may be many thousand times that of the onion. The product is used for its solubility, lack of color, and strong aroma. 5 The alkyl thiosulfonates (methyl methane-, pr0pyl methane-, and propyl propanethiosulfonates) have been associated with fresh onion-like flavors, while prOpyl- and propenyl- containing di- and trisulfides have been associated with cooked onions or steam distilled onion oils (Boelens et al., 1971). Previous studies (Brodnitz et al., 1969; Boelens et al., 1971; Mazza et al., 1980; Kallio and Salorinne, 1990; Kuo et al., 1990) have reported flavor components of head space, solvent extracts, and distilled oils. Block (1985) reviewed the chemistry of garlic and onion. Block et al. (1986) reported synthesis of antithrombotic organosulfur compounds from garlic. 2.1.2 Properties of Onion Oil Fenaroli (1971) described the onion oil as a yellowish liquid with a characteristic onion odor whose main constituents are di-n-propyl and methyl n-propyl disulfide with a specific gravity of 1047-1098 and solubility of 1:10 (in 90% ethyl alcohol). Fenwick and Hanley (1985) described onion oil as a brown-amber liquid obtained in 0.002 to 0.03% yield by the distillation of minced onions which had been allowed to stand for some hours prior to distillation. It was reported that the oil comprises a complex mixture of (mainly) sulfur containing volatiles. The product possesses (on a weight basis) 800 to 1000 times the strength of odor of fresh onions, but its commercial value may be many thousand times that of the onion. The product is used for its solubility, lack of color, and strong aroma. 6 Sinha et al. (1992) reported extraction of onion oil using supercritical carbon dioxide extraction. They described the oil as having the characteristic fresh onion- like flavor. Several new components were reported, including diallyl thiosulfinate and propyl methanethiosulfonate. Onion oil has been granted the status "GRAS" (generally recognized as safe) by the Food and Drug Administration (Fenaroli, 1971). Onion oil forms part of the "White list" of the FDA which means that in order for a product to be safe, the conditions of intended use must be known and reasonable assurance given that the actual use conforms to the intended conditions. Besides being commonly used as a food ingredient, onion oil shows other potential uses. Wit et al. (1979) found that onion oil (or garlic oil), when used in the proportion of 1.5 mg/g in meat slurry, inhibited toxin production by Clostridium botulinum type A (strain 73A). However, the inhibition of toxin production was not complete since the oil solution did not inhibit toxin production by Clostridium botulinum type B (strain R1V1) and type B (strain R1V2). Some studies were conducted on the inhibition of aflatoxin-producing fungi by onion extracts by Sharma et al. (1979), who reported the effect of an onion extract (2% v/v), onion oil solution (10% v/v), lachrymatory factor solution (1% v/v), and dipropyl disulfide on the growth rate of cultures of Aspergillus flavus and Aspergillus parasiticus. They found that the onion extract, the onion oil solution, and the lachrymatory factor solutions possess similar fungi-growth inhibitory properties and that the dipropyl disulfide solution hardly 7 inhibited the culture growth. Apparently the lachrymatory factor was the main component responsible for the inhibition of the aflatoxin-producing fungi growth. 2.2 SUPERCRITICAL FLUID EXTRACTION 2.2.1 Properties of Supercritical Fluids Supercritical fluid extraction makes use of a supercritical fluid (SCF) as a solvent. A SCF is at a temperature above its critical temperature and at a pressure above its critical pressure. The supercritical region for a pure component is shown in Figure 1. This compressed gas has characteristics of both gases and liquids. It has the density of a liquid and functions like a liquid solvent, but it diffuses easily like a gas. A SCF has viscosity and diffusivity lying between that of a gas and a liquid. It is particularly attractive as an extracting agent because the solvent power can be manipulated by small changes in temperature and/or pressure. Another important feature of extracting aroma concentrates with supercritical gases is that both the enhancement of vapor pressure and phase separation play a role in the process. In other words, two unit operations are carried out simultaneously, namely, distillation and solid—liquid or liquid-liquid extraction. Additionally, the zero surface tension of SCFs allows facile penetration into microporous materials. A reduction in solvent density with changes in temperature and/or pressure allows the recovery of solute and solvent. The result is a highly efficient extracting solvent. SUPERCRITICAL FLUID 7/'/ / REGION/g/ /// \\\\\ to I: D (D m SOLID LIQUID w c: o. GAS TEMPERATURE Figure 1. Pressure-Temperature Diagram of 3 Pure Component. From McHugh and Krukonis (1994). 9 2.2.2 Supercritical Fluid Extraction Process 2.2.2.1 How the Process Works Supercritical extraction is a unique process that uses the special properties of fluids above their critical temperatures and pressures. In short, the solvent fluid is pressurized and heated to its supercritical state. Then it is introduced into the extraction vessel at the selected extractor operating conditions. The material in the extractor can be in either solid or liquid phase. In the extractor, the supercritical solvent extracts one or more components from the source material. The solute—rich supercritical fluid exits the extractor and undergoes a temperature and/or pressure change. This change decreases the solubility of the solute in the solvent fluid and, due to the change in solubility, a solute/ solvent separation takes place in the separator vessel. Solute is removed and solute-lean fluid is pressurized and recycled in a continuous flow (Cohen, 1984). 2.2.2.2 System Components 0 A supercritical fluid supply system 0 An extraction vessel - A pressure/ temperature control system 0 A separation system 0 A recycling system Figure 2 shows a flow chart of a supercritical fluid extraction system. 10 “8.5m > .EBmzm :ouonbxm 05% Bin BEEP—03m .N Esmi -————— Seam coguaom K Eofizm 3.550 0585““th 3:535 p——— 833m :ouoabxm lllll , Eogmzm >396 ”Um 823m 2050M 11 2.2.2.3 Advantages of Supercritical Fluid Extraction The advantages of supercritical fluid extraction are (Cohen, 1984): l. Heat-sensitive compounds with low vapor pressure may be more successfully separated without damage or contamination. The extracted material may be separated from the supercritical fluid by changing the temperature or pressure, an energy efficient process. High selectivity can be achieved in the removal of specific components of multicomponent mixtures. By selecting the proper extractor and multiseparator operating conditions, a fractionation of multicomponent solutes can be made. Many supercritical fluids are more environmentally acceptable than organic solvents, thereby reducing workplace and environmental hazards. Many supercritical fluids do not leave solvent residue. This is an important consideration in food applications. Supercritical extraction processes often work where other separation techniques fail. However, this method of extraction presents two basic drawbacks: the physicochemical principles involved in the theory of this process are highly sophisticated and the technology to develop the high pressure system necessary to carry out the extraction operation is expensive. 12 2.2.2.4 Historical Developments and Applications The first observations that supercritical fluids dissolve unexpectedly large quantities of relatively non-volatile materials were reported in the literature about 100 years ago (Hannay and Hogarth, 1879; Andrews, 1887) when it was noted that metal halides became soluble in supercritical tetrachloromethane and ethanol. Buchner (1906) subsequently reported that the solubilities of low volatility organic materials in C02, under supercritical conditions, were orders of magnitude higher than would be expected from vapor pressure considerations alone. Beginning in the 19303, the effort to improve petroleum refining technology led to acquisition of vapor-liquid equilibrium data on hydrocarbon mixtures at high pressures (Sage et al., 1936; Kay, 1938). Messmore (1947) obtained a patent for deasphalting of oils using SFE. In the 19505 the Residuum Oil Supercritical Extraction process was developed for the removal of lighter products from the residue of commercial distillation of crude oil (Basta, 1984). In studies on solubilities, Francis (1954, 1955) established the technical feasibility of using liquid C02 just below the critical point as a solvent for organic materials. Elgin and Weinstock (1959) presented a method for separating a number of mixtures into water-rich and organic-rich fractions. Intensive study of supercritical fluids for extraction of food components began in the early 19705. Many patents resulted from these first studies covering the SFE of h0ps, coffee, tea, tobacco, and spices (Roselius et al., 1972a, b; Vitzthum and Hubert, 13 1976, 1979; Zosel, 1972), among others. Built by HAG AG. (Germany) and currently operated by General Foods Corp., the first large-scale production plant using SFE for food systems was designed to remove caffeine from green coffee beans. The plant, which uses supercritical C02, has been operating since 1979 (Rizvi et al., 1986). Supercritical fluid extraction has been applied to a wide variety of foods. Research and applications include: decaffeination of coffee and tea (Vitzthum and Hubert, 1979; Zosel, 1982), fractionation of fish oils, hops extraction, oleoresin and essential oil extraction from spices and herbs (Hubert and Vitzthum, 1978), deodorization and hydrogenation of fats and oils, flavor extraction (Nguyen et al., 1991; Sinha et al., 1992), food coloring extraction from plant material (Degnan et al., 1991), citrus oil extraction from peels (Calame and Steiner, 1982; Copella and Barton, 1987), oil extraction from snacks, oilseed extraction (Stahl et al., 1980), and cholesterol removal from eggs (Rossi et al., 1989). Other applications include: 0 Separation of organic-water solutions (Paulitis et al., 1981; Moses et al., 1982; Kuk and Montagna, 1983). 0 Polymers and monomers processing (Krase, 1945; Cottle, 1966; Wild et al., 1982; DeSimone et al., 1990; Guckes et al., 1990). 0 Regeneration of activated carbon (Modell etal., 1978). The design of commercial supercritical fluid extraction plants has been discussed by several authors (Eggers and Tschiersch, 1978; Schneider et al., 1980; Stahl et al., 1988; Novak and Robey, 1989). According to Novak and Robey, important design 14 considerations are: raw material preparation, extraction conditions, separation conditions, and supercritical solvent recycle and treatment. 2.3 USE OF ENTRAINER WITH SUPERCRITICAL FLUID EXTRACTION Small amounts of coosolvents, which are referred to as modifiers or entrainers may be added to modify the polarity and solvent strength of the primary supercritical fluid to increase the solute solubility (and/or selectivity) and to minimize operating costs. The entrainers are commonly polar and non-polar organic compounds which are miscible with the supercritical fluid solvent (Dobbs et al., 1986; Dobbs et al., 1987). The increase in solvent power when a co-solvent is added has been noted by several authors: Brunner and Peter (1982); Kumik and Reid (1982); Ely and Baker (1983); Paulaitis et al. (1983); Dandge et al. (1985); Gopal et al. (1985); and Kim et al. (1985). For many systems the increase in solvent power is due to an increase in the density of the solvent mixture, and does not lead to improved selectivity. VanAlsten et al. (1984) and Schmitt and Reid (1986) measured the solubility of a single solid compound in a pure supercritical fluid and in a mixture of the supercritical fluid with a small amount of co-solvent. The study by VanAlsten et al. focused on the functionality of the solute while the study by Schmitt and Reid focused on the functionality of the co- solvent. The addition of a small amount of entrainer into a primary supercritical fluid tends to increase both the critical temperature and pressure of the resulting solvent 15 mixture. To ensure operation in the critical region of a binary solvent mixture of defined composition, knowledge of the gas-liquid critical point for the mixture is essential (Gurdial et al., 1993). These authors have presented data for binary mixtures containing C02 as primary supercritical fluid and many different entrainers. The majority of the information in the literature dealing with entrainers focuses on the increased solubility of solids in supercritical fluids containing small amounts of co-solvents. There are few examples of liquid-fluid equilibrium data for liquid-entrainer-supercritical fluid systems. Most deal with the separation of two organic compounds with similar boiling points that are difficult to separate by conventional distillation. Peter and Brunner (1978) increased the concentration of glycerides in propane by adding small amounts of acetone. They also found that the distribution coefficient of palm oil in C02 doubled with the addition of ethanol (Brunner and Peter, 1982). Brunner (1983) studied the effect of entrainers on the separation factor of hexadecanol, octadecane, and salicylic acid phenyl ester in several supercritical fluids including C02. He found that, depending on the temperature and pressure, the solubility as a function of entrainer concentration may decrease, increase, or run through a maximum. Roop and Akgerman (1989) proposed a method of predicting the effect of adding a small amount of an entrainer to a supercritical fluid for the extraction of organic compounds from aqueous systems. 16 2.4 EXTRACTION WITH LIQUEFIED GASES If the pressure is raised sufficiently, many substances which are gaseous at ambient pressure either liquefy or begin to behave like liquids in that they exert appreciable solvent power, even for solutes of low volatility. For example, at temperatures up to 31°C, the critical temperature, C02 can be liquefied by raising the pressure and this liquid can be used to dissolve natural oils and quite a wide range of non- polar or slightly polar materials (King and Bott, 1993). Liquid carbon dioxide has been found to be a very selective solvent for the extraction of flavor compounds such as terpenes, aldehydes and ketones, while other components of foods such as sugars, fruit acids, salts, amino acids, fats and water are practically insoluble (Schultz and Randall, 1970). Early research on liquid C02 was carried out by Francis (1955), who devised a process using liquid carbon dioxide to increase the dissolving power of conventional solvents and also conducted an extensive study on the mutual solubilities of liquid carbon dioxide with 261 different substances (Francis, 1954). One of the first uses of liquid carbon dioxide was on the extraction of coffee aroma (Sivetz, 1963). Schultz and Randall (1970) carried out a liquid C02 extraction of aroma components from apple, pear and orange juices, orange pieces and roasted ground coffee. They also listed data about the distribution of alcohol and esters between liquid carbon dioxide and water. The authors claimed that they obtained highly concentrated essence extracts, i.e., up to 100,000 fold in the case of apple juice. 17 The greatest use of the liquid C02 extraction method is probably in the preparation of hop extracts for the brewing industry. Currently hop extracts are prepared mainly by direct solvent extraction. The solvents are generally methylene chloride, trichloro-ethylene, hexane or methanol. Laws et al. (1977) carried out an extensive study on the extraction of a solvent-free isomerized concentrate from hops, and claimed a recovery of extractables of up to 90%. In general, the aromas and flavors of extracts obtained by liquid C02 extraction bear a closer resemblance to the original material than those obtained by organic solvent extraction. This is ascribed to the very mild conditions of the process and to the lack of oxygen in the extraction system (Grimmett, 1981). 2.5 TRADITIONAL METHODS OF EXTRACTION OF FLAVOR CONCENTRATES Teranishi et al. (1971) provided a comprehensive review of the methods for the isolation and concentration of volatile food constituents. Distillation is by far the most widely used method. The authors described two most utilized variations of this method: flash distillation and high vacuum (steam) distillation. These methods of extraction of flavor components are most commonly used commercially. Flash distillation is used mainly in the recovery of essences in the fruit industry (Milleville and Eskew, 1946) while steam distillation is widely used in the extraction of concentrates and essential oils of seasoning and aromatic herbs (Heath, 1973). Actually, in commercial applications more than one method of flavor extraction is practiced in a process. For instance, in steam distillation the distillate is subjected to a liquid-liquid 18 extraction in order to separate the flavor components from the aqueous solution or emulsion. In the food industry, solvent extraction is an important method for isolation of soluble spice essences and other oleoresins. It has to be followed by distillation to remove the solvent. Spiro and Kandiah (1989) described the kinetics of extraction of ginger oleoresin by solvent extraction. 2.5.1 Steam Distillation Steam distillation is the oldest and still the most important method for obtaining essential oils, i.e., the characteristic smelling volatile oils contained in plant material. These relatively volatile oils are separated from other material with the help of the carrier, steam. Live steam is injected into the liquid or solid mixture which contains the flavor volatiles. When the steam separates from the mixture, it carries the flavor volatiles with it. The steam is subsequently condensed and subjected to a liquid—liquid extraction with a solvent or is further concentrated (Teranishi et al., 1971). In steam distillation, the starting material is subjected to a temperature of 100°C. This can lead to artifacts of the flavor oil components which are often thermolabile. In addition, water can exert a hydrolytic influence, bringing about chemical changes in the oils (Stahl et al., 1988). Steam distillation can also be conducted under reduced pressure (vacuum distillation). It has the advantage of a lower temperature, and thus yields a higher oil quality. Distillation at a pressure above atmospheric, using superheated steam, is 19 sometimes applied to plant material containing oils which are difficult to distill. The higher temperature leads to large amounts of decomposition products (Stahl et al., 1988). 2.5.2 Classical Solvent Extraction Extraction with classical organic solvents is an important procedure for obtaining lipophilic plant components. The selection of a suitable solvent is a critical decision. Each technical step in the extraction, separation and recovery of the solvent from the solution of the extract, and from the extraction residue, has to be considered. The extractions need to be followed by an energy-expensive solvent- stripping stage. The last traces of solvent are difficult to remove completely, and even small amounts of organic solvents in foods are now coming under critical scrutiny for health reasons. 2.6 ANALYTICAL METHODS USED FOR ASSESSMENT OF QUALITY OF EXTRACTS 2.6.1 Gas Chromatography and Combined Gas Chromatography-Mass Spectrometry Gas chromatography (GC) is the most widely used analytical method in flavor research because of its ability to separate compounds. It has been a favorite analytical method of researchers working on flavor components of onion oil either alone or in combination with mass spectrometry (Brodnitz et al., 1969; Brodnitz and Pollock, 1970; Boelens et al., 1971; Block et al., 1992; Kuo and Ho, 1992). 20 Mass spectrometry (MS) has been considered an important method of compound identification. Since GC is not a direct identification technique and only provides component separation, researchers directly link the GC to MS. This is successful because of the compatibility of sample size requirements of both methods (Teranishi et al., 1971). The method is imperfect in that it measures secondary compounds of the enzymatic action, and the relative contribution of these compounds to overall flavor and aroma is not known. It could give a distorted picture of the actual sequence of events, since the compounds found are those that are thermostable, thus surviving the separation process, and those that arise as a result of the heat (Whitaker, 1976). 2.6.2 Headspace Volatiles Analysis Headspace analysis has been the tool of several researchers (Saghir et al., 1964; Bernhard, 1968; Boelens et al., 1971; Bandyopadhyay et al., 1970; Freeman and Mossadeghi, 1970; Freeman and Whenham, 1974; Tewari and Bandyopadhyay, 1977; Mazza and LeMaguer, 1979; Mazza et al., 1980; Yagami et al., 1980) who worked with Alliums. Mazza et al. (1980) assessed several procedures suitable for qualitative and quantitative determination of the volatiles of fresh and dehydrated onions. They emphasized the fact that for obtaining detectable amounts of minor constituents important to the aroma profile, concentration of the headspace vapor is required by means of external cold traps or adsorbents. They found that adsorbents are suitable for this purpose in the case of onions. 21 2.6.3 Other Methods Appraisal of flavor or pungency of Alliums, such as onion and garlic, can be based on either subjective sensory analysis or detection of compounds generated by cysteine sulfoxide lyase (C-S lyase; EC 4.4.1.4) activity after tissue disruption. The typical flavor of Allium species is due to the conversion of endogenous alk(en)yl-L- cysteine sulfoxide flavor precursors to pyruvate, ammonia, and thiosulfinates by OS lyase (Nock and Mazelis, 1987). The determination of pyruvate as an indicator of pungency is perhaps the most established method of appraisal although it is cumbersome and time consuming. An alternative method for the evaluation of pungency in Allium species involved the determination of the thiosulfinates (Carson and Wong, 1959; Nakata et al., 1970). The procedure involved derivatizing the thiosulfinates with N-ethylmaleimide and measuring the absorbance of the conjugate at 515 nm. This procedure was quite specific for thiosulfinates. The method has been further modified by Thomas et al. (1992). 3. MATERIALS AND METHODS 3.1 ONIONS AND JUICE PREPARATION Onions (Allium cepa L.) MSU experimental variety 3506 grown at the Michigan State University (Muck Farm) were obtained in November 1993. They were stored at 2.8°C. The pH of the onion juice was 5.45. It was measured using a Corning pH meter (Model: 610A). The soluble solids content, measured using a hand held refractometer with automatic temperature compensation (Kemco Instruments Co., Inc.), was 7.7 °Brix. The onion bulbs were peeled, cut and immediately processed with an Acme Juicerator (Model: lllEZl) which uses filter paper to separate the pulp from the juice. The juice was stored in a covered container and kept at ambient temperature (26 — 29°C) for one hour to facilitate enzymatic action for flavor development. 3.2 EXTRACTIONS 3.2.1 Supercritical C02 Extraction Supercritical C02 extractions were conducted at 3600 psi (24.5 MPa) and 37°C. The density of C02 under these conditions is 0.89 g/cm3 (Angus et al., 1976). 22 23 Industrial grade C02 (AGA gas, 99.5% purity) from a gas cylinder was compressed with a gas booster (Haskel, Inc.) and stored in a 2.0 L reservoir. A pressure regulator positioned between the reservoir and the extraction vessel controlled the extraction pressure. A two-liter stirred autoclave (Model: Magnedrive H bolted closure autoclave, Autoclave Engineers) was used as the extraction vessel. The vessel was filled with 800 g onion juice prior to pressurization. Onion oil was collected in a 25 m1 test tube. C02 was monitored with a flow meter and a dry test meter. The extraction was commenced by slowly raising the pressure in the extraction vessel while the system outlet was closed. After reaching the extraction pressure, the heated micrometering outlet valve was opened to commence flow. Figure 3 shows the supercritical C02 extraction system. The collection trap was weighed after 1100 liter C02 (STP) had passed through the system. The experiments described in the next paragraph had shown that passing larger volumes of C02 do not increase the yield appreciably beyond 1100 liter. Hence, this amount was chosen for the experiments. For this study, C02 was vented and not recycled. To study the effect of the C02 flow rate on the extraction process, similar experiments were conducted in triplicate at two different C02 flow rates, namely 1.0 l/ min and 0.5 l/min. The collection trap was weighed periodically during the extractions. 3.2.2 Supercritical C02 Extraction with Entrainer The following additions were made to the supercritical C02 extraction system: 0 A high pressure metering pump (Model: A-30-S, Eldex Laboratories Inc.) was 24 .Eoemxm :ozugxm NOD Eoutocoasm .m 23E Comuo> to w an... :3 O).~UOH3< VOL... mum J a...» 20> 838. .8 .8330.— c392 3.... 8.... o>_o> act-nest: .w P eaten. 531.3601 «583. 0.50:...“ o _L NHL e5>couom 028cm IS 95009;. fl vie xv ta? counoom you—u o>.a> xuozu 8 21.... 25 attached using a tee fitting in the line between the pressure regulator and the C02 inlet of the extraction vessel. 0 A static mixer, consisting of a steel tubing filled lightly with glass beads restricted to remain inside the tubing by closing both ends with steel filings, was installed between the extraction vessel and the high pressure metering pump. Its purpose was to ensure that C02 and entrainer entering the extraction vessel were mixed properly. Figure 4 shows the supercritical C02 extraction system with modifications mentioned above for adding entrainer. Ethanol (200 proof dehydrated alcohol, U.S.P. punctilious, obtained from Quantum Chemical Corporation, USI Division) was used to evaluate the effect of a polar entrainer on the extraction of onion oil. Octane (obtained from IT Baker, Inc.) was used to evaluate the effect of a non- polar entrainer. Exu'actions were conducted in triplicate at 3600 psi (24.5 MPa) and 37°C. Two different amounts of each were used, namely 50 ml & 75 ml, to evaluate the effect of different concentrations of entrainer. These amounts were chosen on the basis of preliminary experiments. In each case, 10 ml entrainer was added initially to the juice before extraction. The remaining was added continuously during the time when the system was at the desired extraction pressure. Information regarding critical conditions of supercritical COz-entrainer mixtures at various concentrations of entrainer was obtained from Gurdial et al. (1993). It was used to ensure that the extractions were conducted within the supercritical region of the binary mixtures of C02 and entrainer. .BEEEm w:§o< .8 23.3%qu £3» Eemxm cozugxm NOD 3.52.235 .v Bswi attic-om 95m New .5: _ 9. new 0290...; Luanoom in. no: 1&- 6 9:0) cozuozou 95a 5:230: an 0.5» m a a —. 0:3... 6 c... H1 . 0530005.; no» «Suez 8.9%.: conaauob Laue: 30.... 30C o>.u> mScoueEobf 27 In addition, an experiment was conducted at a lower pressure, i.e. 2600 psi (17.7 MPa), with 75 ml ethyl alcohol as entrainer to evaluate the effect of entrainer at a lower pressure (though still in the supercritical region). The temperature was kept the same at 37°C. 3.2.3 Liquid C02 Extraction The same equipment as for supercritical C02 extraction was used. Extraction was conducted at 2900 psi (19.7 MPa) and 27°C. The density of C02 is 0.90 g/cm3 under these conditions. The above conditions were chosen such that the density of C02 for the supercritical C02 and liquid C02 experiments was nearly equal so that a meaningful comparison of the two methods could be made. 3.2.4 Steam Distillation-Solvent Extraction 200 m1 onion juice was mixed with 100 ml distilled water. The volatile components were extracted by 20 ml dichloromethane for one hour in a modified Likens- Nickerson apparatus (Likens and Nickerson, 1964, Schultz et al., 1977). The apparatus is shown in Figure 5. Trace water in the extracted volatile solution was removed by anhydrous sodium sulfate and excess solvent was removed by nitrogen purging. 28 - whom-mm "'“\ 0-“ --m-” O .5” I dichloromethane g .,..,,.. Figure 5. A Modified Likens-Nickerson Apparatus for Distillation-Solvent Extraction. 29 3.3 ANALYSIS 3.3.1 Estimation of Onion Oil Yield 3.3.1.1 Gravimetric Method Onion oil yield was estimated gravimetrically in the case of supercritical C02 extraction and liquid C02 extraction. The collection trap was weighed after passing 1100 liter of C02 under desired conditions. However, in the case of (supercritical C02 + entrainer) and steam distillation-solvent extraction this was not possible because the extract contained entrainer and moisture (in some cases). The entrainer present in the extracts could have been removed by nitrogen purging, and then the gravimetric yield determined. However, nitrogen purging could result in a loss of some volatile flavor components. Hence, an alternative method, quantitative GC analysis, was used for estimating the yield of extracts containing entrainer. 3.3.1.2 Quantitative Gas Chromatographic Analysis Gas chromatography was done in duplicate on the samples of extracts using GC (Model: Hewlett-Packard 5890 Series H). A 30 m HP-l methyl silicone column of 0.53 mm inner diameter was used to separate the flavor components. A 1 tr] dichloromethane dilution (containing 0.0045 g extract/ml dichloromethane) of each extract was injected for analysis. The Operating conditions were as follows: injector temperature, 220°C; helium carrier gas flow rate, 6 ml/min; oven temperature, 35 to 200°C at a linear rate of 5°C/min and 15 min holding time at 200°C. A flame ionization 30 detector at temperature 240°C was used. HP 3365 Series II Chemstation Version A.03.21 by Hewlett Packard was used to control the GC and to record and integrate the data. Sample preparation for GC analysis involved dilution of the extract with methylene chloride. In addition, the samples from (supercritical C02 + ethanol) experiments had to be dried using anhydrous sodium sulfate. Abraham et al. (1976) and Sinha et al. (1992) reported that the characteristic flavor of onions is due to the volatile oil, which consists chiefly of sulfur compounds. Sulfur-compound peaks were identified using gas chromatography-mass spectrometry (GC-MS). Summation of these peak areas was done for each sample. A known amount (0.005 g) of an internal standard, benzyl disulfide, was added to the juice before extraction. The peak area of the internal standard, the amount of internal standard added and the sum of peak areas of sulfur peaks were used to calculate the amount of sulfur compounds present in each sample using the following equation: (ZPeak Area of 8 Compounds) x Wt. of IS . Peak Area of IS Total Wt. of S Compounds = The following simplifying assumptions were made: 1. The internal standard and the onion flavor compounds have similar extraction properties and GC and MS responses. 31 2. The total weight of sulfur compounds present in the extracts was considered as an indicator of the yield of onion oil. The quantitative GC analysis will give information regarding the yield obtained by different extraction methods on a relative basis and not the absolute yield of onion oil. 3.3.2 Estimation of Quality of Extracts 3.3.2.1 Headspace Volatiles Analysis 30 ml onion juice was purged with purified nitrogen gas at a flow rate of 75 m1/min for 2.5 hours. The headspace volatiles were adsorbed onto an activated coconut charcoal trap. For gas chromatography analysis, the onion headspace volatiles were eluted from the trap using 1 ml carbon disulfide. This analysis was done to evaluate the efficiency of the various extraction methods studied with regard to the quality of extract obtained. Headspace samples contain most of the volatile compounds responsible for fresh onion flavor. By comparing the profiles of extracts obtained from the various methods with the headspace volatiles profile, it can be determined if the specified method is capable of extracting the components found in the headspace of fresh onion juice. 3.3.2.2 Thiosulfinate Analysis Thiosulfinate analysis was done by the method described by Thomas et al. (1992). The analysis was performed on the following samples: (1). Onion oil obtained 32 from SFE-002 with ethyl alcohol as entrainer, (2). Onion oil obtained from SFE-C02 with octane as entrainer, and (3). Commercial onion oil obtained by steam distillation. The samples were purged with nitrogen to remove any traces of solvent/ entrainer present. 5 ml isopropyl alcohol was added to equal amounts of the samples. 1 ml of 0.05 M N-ethylmaleimide in isopropyl alcohol, 1 ml of 0.25 M KOH in isopropyl alcohol, and 1.5 ml of 10 g liter’1 ascorbic acid in distilled water were added. After vortexing the solution for approximately 10 seconds, the absorbance at 515 nm was recorded using a LKB Biochrom spectrophotometer (Model: Ultrospec II). The purpose of this analysis was to get information about the quality of the extract obtained from supercritical C02 extraction and compare it with commercial onion oil obtained from distillation. In addition, it is also a way of evaluating whether the extracts possess a fresh- like or cooked flavor. The samples containing high level of thiosulfinates tend to have a fresh-like flavor. 3.3.2.3 Gas Chromatography Extracts obtained from the extraction methods were analyzed in duplicate by GC. The chromatograms were compared. 3.3.2.4 Gas Chromatography-Mass Spectrometry Extracts from different methods were analyzed by GC-MS to identify the sulfur compounds present. Analysis by GC-MS was carried out on a JEOL AX-SOSH double-focusing mass spectrometer coupled to a Hewlett-Packard 5890] gas 33 chromatograph via a heated interface. GC separation employed a SPB-l fused silica capillary column (30m length, 0.25mm i.d. with a 0.25pm film coating) supplied by Supelco, Inc. Direct (splitless) injection was used. 3 it] methylene chloride solutions of different extracts were injected for analysis. Helium gas flow was approximately 1 ml/ min. The GC temperature program was initiated at 35°C, held at this temperature for 5 min then heated at 5°C/min to 200°C and held for 15 min at this temperature. MS conditions were as follows: interface temperature 280°C, ion source temperature ca. 220°C, and the scan rate of the mass spectrometer was 1 s/scan over the m/z range 35/ 500. The mass spectra were obtained by electron ionization at 70 eV. 3.3.2.5 Experiments on Post-Extraction Residue and Fresh Juice Dichloromethane extracts of onion juice that had undergone supercritical extraction were analyzed using GC to establish how much of the flavoring material remained in the post-extraction residue. GC analysis of dichloromethane extract of fresh onion juice was also done. This provides information regarding the amount of flavoring material that was present in the juice before extraction. Comparison of these two profiles would provide information regarding the efficiency of extraction of flavor compounds using supercritical C02. 800 ml fresh onion juice was kept in a covered beaker for one hour before doing solvent extraction with dichloromethane. The extract was concentrated using nitrogen purging and analyzed by GC. The dichloromethane extract of post- extraction residue was also concentrated. 4. RESULTS AND DISCUSSION 4.1 QUANTITATIVE ANALYSIS 4.1.1 Gravimetric Yield of Onion Oil by Different Methods Gravimetric yield of onion oil from SFE-C02 with a C02 flow rate 1.0 l/ min, SFE-C02 with a C02 flow rate 0.5 llmin and liquid C02 extraction was obtained by weighing the extract collected in the collection trap after passing 1100 liter C02 through the extraction system. Table 1 shows the gravimetric yield for these three cases. SFE- C02 extraction with the C02 flow rate of 0.5 llmin gave the maximum onion oil yield. Table 1: Gravimetric yield of onion oil obtained by different methods. % on Yield° Method Oil Yield8 (3) (wt basis) SFE-C02, C02 flow: 1.0 llmin 0.22870 0.0286 SFE-C02, C02 flow: 0.5 llmin 0_2592d 0.0324 Liquid C02 extraction 0.17196 0.0215 a. Mean of triplicate samples b. On basis of weight of onion juice used c, d, e. Indicate that the yields are significantly different at the 1% level. Gravimetric yield data were subjected to analysis of variance and Fisher’s protected least significant difference analysis. It was found that the yields for all three methods (Table l) were significantly different at the 1% level. The ANOVA table and least significant difference calculations are presented in Appendix C. Gravimetric yield 34 35 could not be evaluated for the other methods due to the presence of small amounts of entrainer/solvent in the extracts. Fenwick and Hanley (1985) reported 0.002 to 0.03% yield of onion oil by distillation. In this study a maximum of 0.0324% yield of onion oil was obtained. 4.1.2 Quantitative GC Analysis for Comparison of Onion Oil Yield from All Methods The total weight of sulfur compounds contained in a sample was calculated as follows: (ZPeak Area of S Compounds) x Wt. of IS Peak Area of IS Total Wt. of S Compounds = Where ’18’ is the abbreviation for the internal standard and ’S’ for sulfur. The total weight of sulfur compounds present in a sample was assumed to be representative of the yield of onion oil. Abraham et al. (1976) and Sinha et a1. (1992) reported that the characteristic flavor of onions is due to the volatile oil, which consists chiefly of sulfur compounds. Figure 6 shows the total weight of sulfur compounds present in the extracts from all the methods studied. The standard deviation of yield obtained from triplicate experiments is shown on top of each bar in Figure 6. The relative proportions of yield obtained by SFE-C02 at 0.5 l/min and by SFE-C02 at 1.0 l/min were similar for both the gravimetric and the quantitative GC methods for estimation of yield. However, in the case of Liquid C02 Extraction, quantitative GC method indicated less yield than the gravimetric method. The difference between the gravimetric and the quantitative GC methods for estimation of yield could be due to the presence of small amounts of moisture in the extract, which may affect the gravimetric yield data but have no effect on .m_mba=< U0 02:33:30 .3 3:330 muombxm 05 E mtcsanoU Sham mo EsoE< .38. .o Eswi Susaxm fizfiégaaama 28% a can .5ch 238.3 .532 a 8 6595:... 8 .0525 _E 2. .23 ”00-9.5 a 352:... a 355m .5 fl 5? ”8-9; a. 3585 a .225 E. on 5? ”00.95 o 3588 a 0530 .e we 53 ”8.9.; m 3585 a 238 .5 8 53 N8.3m v 858$ Now 2.55 m 5:: no u as .5: N8 .23 60.95 N 5:: S u as 38 N8 53 ”8-9.5 _ 36 m m h o m w m N P o x I x x x x x x 0 Ed . V I 86 mod -. L I 86 u t m. 86 I I I 86 m i m V0.0 in it V0.0 M t m 85 I I I 86 . m 85 I I 86 m. t I 3.0 I I 8.0 ® 8.0 I t I mod 85 I I 8.0 37 quantitative GC data since water is not detected by a flame ionization detector. The extract obtained by SFE-C02 with a C02 flow rate of 0.5 llmin contained the maximum quantity of sulfur compounds thus indicating that the onion oil yield would be maximum in this case. Liquid C02 extract was found to contain the minimum quantity of sulfur compounds. Statistical analysis showed that the onion oil yield obtained by the different methods was significantly different at the 5% level except in the case of the following pairs: SFE-C02 with 50 m1 octane and SFE-C02 with 75 ml octane; SFE-C02 with 75 ml octane and steam distillation-solvent extraction; and SFE- C02 with 75 ml ethyl alcohol at 3600 psi and at 2600 psi (Appendix C). The steps followed in the calculation of the total weight of sulfur compounds in the different extracts are explained in Appendix A. 4.1.3 Effect of Volume of C02 Passed Through the Extraction System on the Yield of Onion Oil Figure 7 shows the effect of volume of C02 passed through the extraction system on the yield of onion oil under the conditions of SFE-C02 at 3600 psi (24.5 MPa) and 37°C with C02 flow rate 0.5 llmin. The yield increases at a higher rate during the initial part of the extraction as indicated by the initial steepness of the curve. This is followed by a region of lower extraction rate as indicated by the flattening of the curve during the latter part of the extraction. The lower extraction rate during the latter part of the extraction may be explained in terms of depletion of the solute (onion flavor oil) in the substrate (onion juice). Yield of Onion Oil (g/800g onion juice) 38 0.3 r I 000.0... .08" 0.25- ’0, . z" o .I’ o 0.2b ,8 . o I O o .I / 015 ’- . " lo “ 09 d O l 0 ~/ 0.1- .1 .. /. / I. .. / . 0.05 "(8 / 0 L 1 0 500 1000 1500 Volume of Carbondioxide Passed (lite) Figure 7. Yield of Onion Oil versus Volume of CO2 passed at 3600 psi, 37°C. 39 Figure 8 shows the yield of onion oil obtained at different volumes of C02 passed through the system at two different C02 flow rates, namely, 0.5 llmin and 1.0 llmin. The data points in the initial part of the curves indicate similar extraction rates. A possible reason for this could be the presence of headspace (more than 1 liter) in the extraction vessel, which was the same during both the experiments. In the later part of extraction, 0.5 l/min C02 experiment resulted in a higher yield than 1.0 llmin experiment. The higher yield in the case of the experiment with 0.5 llmin flow rate could be due to greater residence time of the solvent (C02) in the extraction system. The two curves also indicate that the process does not come to an equilibrium state under the conditions tested. In Figures 7 and 8, polynomials of second degree were used to fit the experimental data. 4.1.4 Effect of Entrainer The effect of entrainer was studied under the following subheadings: - Effect of polar entrainer 0 Effect of non-polar entrainer 0 Effect of entrainer at different pressures 4.1.4.1 Effect of Polar Entrainer The polar entrainer, ethyl alcohol, was added to the SFE-C02 system in two different amounts, i.e. 50 ml and 75 ml. As indicated in Figure 6, 26.4% greater yield was obtained in the case of 75 ml ethyl alcohol as compared with 50 ml ethyl alcohol. Yield of Onion Oil (g/BOOg onion juice) 0.3 0.25 :- 0.2- I 0.15 {7137f if 7 Solid: 1 Vmin 0.05 - , - ”Q“ Dashed: 0.5 llmin L 0 500 1000 1500 Volume of Carbondioxide Passed (liter) ‘ Figure 8. Yield of Onion Oil at Different Volumes of C02 Passed at Different Flow Rates of C02. 41 The 75 ml ethyl alcohol experiment enhanced the yield of onion oil over the experiment without entrainer (at same C02 flow rate, i.e. 1.0 llmin) by 16.6%. In the case of the 50 ml ethyl alcohol experiment, the yield was lower than SFE-C02 without entrainer (Figure 6). 4.1.4.2 Effect of Non-polar Entrainer The non-polar entrainer, octane, was added to the SFE-C02 system in two different amounts, i.e. 50 ml and 75 ml. The yield of onion oil obtained from experiments with 75 ml octane as entrainer with SFE-C02 extraction was higher than that obtained from 50 ml octane experiments. However they were both less than the yield of SFE-C02 experiments without any entrainer. Thus, under the conditions studied, octane failed to enhance the yield of onion oil from SFE-C02. 4.1.4.3 Effect of Entrainer at Different Pressures The SFE-C02 system was found to behave almost identically in the case of 75 ml ethyl alcohol used as entrainer at 3600 psi (24.5 MPa) and at 2600 psi (17.7 MPa) in terms of the onion oil yield (Figure 6). Although the solvent power of supercritical C02 is known to be higher at higher pressures, in these experiments the yield of onion oil was 2.7% higher at 2600 psi than that at 3600 psi. However, statistical analysis showed that the yields at 2600 psi and at 3600 psi were not significantly different at the 5% level (Appendix C). 42 This is a very interesting result because it indicates the possibility of lowering the extraction pressure from 3600 psi to 2600 psi with no reduction in the yield of onion oil by using ethyl alcohol as entrainer in the above mentioned quantity. This could have positive effects on the economics of this process. This result supports the findings of Brunner and Peter (1982) that supercritical fluid extraction can be conducted at a lower pressure in the presence of an entrainer. This happens because the solubility of the solute in the SCF under the same temperature and pressure conditions is greatly enhanced in presence of an entrainer. Lower operating pressures are also desirable for safety reasons. Experiments were not done at 2600 psi without entrainer. Thus, it cannot be proved that the yield enhancement in the case of 2600 psi experiment over the 3600 psi experiment was solely due to the presence of entrainer. The enhancement of yield could also be due to the solvent characteristics of supercritical C02 at 2600 psi and 37°C. 4.2 QUALITATIVE ANALYSIS 4.2.1 Quality of Extracts Obtained by Various Methods The onion extracts produced through SFE-C02 (with or without entrainer) and liquid C02 extraction had characteristic fresh onion-like smell in contrast to the rather unpleasing cooked onion-like smell of steam distilled extracts. Fresh onion- like aroma of supercritical C02 extracts can be confirmed by the report of a trained panel organoleptic evaluation of similar extracts (Appendix D). Gas chromatography—mass spectrometry was used to identify the chemical components of the various extracts and 43 headspace volatiles of onion juice. The emphasis was on sulfur compounds which are the main flavor components of onions (Abraham et al., 1976, Sinha et al., 1992). These results are presented in Table 2. The identification of the flavor compounds was based on the comparison of mass spectral data with published works. In one case tentative identification was made by direct interpretation of mass spectral data. Where definitive characterization could not be made, the mass number is indicated. Considering the C02 extraction methods studied (SFE-C02 without entrainer, SFE-C02 + Ethanol, SFE-C02 + Octane and Liquid C02 extraction), the maximum number of flavor compounds were identified in SFE-C02 + Ethanol extracts. SFE-C02 + Ethanol and Liquid C02 extracts contained all the compounds identified in the headspace of onion juice except molecular sulfur (31), which is most likely an artifact resulting from the processing or GC-MS analysis, and dipropyl trisulfide (24). SFE-C02 (without entrainer) extracts contained all the compounds identified in the headspace of onion juice except molecular sulfur (31), dipropyl trisulfide (24) and 1- propenyl propyl trisulfide (25). SFE-C02 + Octane did not contain other compounds that were identified in headspace, like methyl trans-propenyl disulfide (9) and dirnethyl trisulfide (10). This indicates that SFE-C02 + Ethanol, SFE-C02 without entrainer, and Liquid C02 Extraction produce more fresh onion-like flavor than other methods studied as indicated by the presence of the headspace flavor compounds. The SFE-C02 without entrainer, SFE-C02 + Ethanol and SFE-C02 + Octane extracts contained diallyl thiosulfinate (22) or its isomer, di-l-propenyl \. \. \. \. 52 .mmaEJ 92:25: :33... 12.2: c. \. \. \. \o c: Amazou u:o-m-_.==..~._222.25% m. \. v2 5069.2 :90 22:02:82: 222.52. 13:... v— x. x. x. x. 5: Am: 2.0 5.53... .22.. .582... m. x. x. on. £22.“. 2.53.... .52.... N. \. \. \ Om. 2393:: 5:28.25”. 2255.2: : x. \. x. 8. $.26 2.53.2. .5595... 5. \a \. \. ou— .~m..:..U 3536152825255 .25....E a x. \. x. x. 8. 55.26 85...... .5255... 252.. w x. \. um. .chfvu 3.53:. .325 .252: N. x. x. x. 3. 552.0 2.5... 3...... o x. x. \. x. a: 5.2.0 22.5.5. 5.53.5-2.” m x. x. x. N: 5.26 2.2.8.... 5.5.55-3 v x. x. x. 5. 5.2.6 2.2.5.5 55.5.5-3 M x. x. x. 5.. 55.26 2.53... .55.. :52: N x. x. 2.. $.26 2.55... 5525.. . 89:53: Now .23 wmsuowww ”MW—«Hm Noomum $.22 .2258". 5:28:50 2.: muEEmm coax—memo: 2:2 Busbxm NOU 322.2053 E 22-00 .3 coca—BE mcczanoU .0522 2220 .m as“... 45 \. x. \. EN n2.5.0.2....5..522...2525? Nm \. 8N 2m =5... 5:885 .N. x. N.N 8.5m 2.52.285.555......m-.b..oe...-...N cm \. N.N 25.53.“..-nnN..2523... N.N \. x. x. x. 8. @526 2.53... .Eo._...N-.>.aE=.-v.m .52.. MN \. \. S S: .wNINJ 0223253236 N.N \. x. x: .552... 2.53... 3...... 0N \. x. x. x. 5.. $22.. Sm. 0553.5 .32.. 3.2.2.... MN \. x. N... 25.5.“. 2.53... .32.... .N x. x. x. x. x. o... .x: 2.9 GE. 85...... .32... .582... .N x. \. \. no. .Nmoofeu 2.2.5325 3.2:. mm \. 5N. 50.2.. 2.5... €555.56 .52.. .N x. x. x. 3.. if... 55.5-...7535-55.22.55 cN \. x. x. x. x. 8. $52.0 05-15...-N..-.2=2.5-m 2 x x. x. x. x. Nm. .mmm2d SE. 2.5%... .232... .32: w. x. x. x. x. \. Nm. 3.2.0 5.3 2.53... 3.2.2..-. :5...= : BESS: Nov .75 WWNMWW ”MWSNWM Noummm $22 .2225”. 52:02:50 .2. 92.5.80. N 2...? 46 thiosulfmate. Sinha et al. (1992) reported the presence of diallyl thiosulfinate or its isomer, di-l-propenyl thiosulfinate, in supercritical C02 onion extracts for the first time. However, they mentioned that the identification of diallyl thiosulfmate was tentative due to the absence of a synthesized reference compound or supporting NMR or IR spectra. Hence, the identification of diallyl thiosulfinate (22) is tentative. Brodnitz et al. (1971), using IR and NMR analysis, reported diallyl thiosulfinate to be a major constituent of fresh garlic extracts. The authors indicated that during gas chromatography diallyl thiosulfinate undergoes dehydration, forming two isomeric disulfides: 3-vinyl-1,2-dithi-5-ene and 3-vinyl-l,2-dithi-4-ene. Nishimura et al. (1988) reported 3,4-dihydro-3—vinyl-1,2-dithiin in Allium victorialis L. The authors indicated that the decomposition of diallyl thiosulfinate in methanol at room temperature for 7 days produced 3,4—dihydro-3-vinyl-l,2-dithiin, 2—viny1—4H-l,3-dithiin, diallyl sulfide and diallyl trisulfide. In the headspace of crushed onions, Kallio and Salorinne (1990) reported the tentative presence of 3—ethenyl-1,2-dithi—4—ene and 3-ethenyl-l,2— dithi—S-ene. In the present study, diallyl thiosulfinate (or its isomer) (22), 3-ethenyl— 1,2-dithi-4-ene (19), 3-ethenyl-1,2-dithi-5-ene (15), 3,4-dihydro-3-vinyl-1,2—dithiin (20) and diallyl trisulfide (26) have been found in each of the supercritical extracts (SFE-C02 without entrainer, SFE-C02 + Ethanol, SFE-C02 + Octane). The presence of the above mentioned compounds in the extracts supports the identification of diallyl thiosulfinate since these compounds have been reported to be produced by decomposition of diallyl thiosulfinate. 47 Thiosulfonates are secondary products arising from thiosulfinates (Abraham et al., 1976). The presence of methyl methanethiosulfonate, propyl methanethiosulfonate, and propyl propanethiosulfonate in dichloromethane extracts of freshly chopped onions has been reported by Boelens et al. (1971). These authors suggest that the absence of these compounds in steam-distilled onion oil is a result of their being soluble in water; thus, they are present in the water layer and do not get transferred to the organic phase during the steam distillation process. In this study, only propyl methanethiosulfonate (14) was identified and only in SFE-C02 + Ethanol extracts (Table 2). Molecular sulfur (31) is being reported in headspace of onion juice for the first time. The identification is based on spectral comparisons with data from Heller and Milne (1978). It is expected to be an artifact resulting from rearrangement of sulfur atoms during processing or GC-MS analysis due to heat. The identification of 4,6-diethyl-l,2,3,5-tertathiane (29), 2,4-dimethyl-5,6- dithia-2,7-nonadienal 5-oxide (30), and 5,7-diethyl-l,2,3,4,6-pentathiepane (32) was based on mass spectral data from Kuo and Ho (1992). An unknown compound of molecular weight 130 was detected in SFE- C02, SFE-C02 + Ethanol and SFE-C02 + Octane extracts. The MS data is provided in Appendix B. The SFE-C02 + Ethanol sample contained the maximum amount of thiosulfinates as indicated by the maximum absorbance recorded in this case (Table 3). 48 Its value was almost five times that for commercial onion oil. The absorbance for SFE- C02 + Octane was also many times higher than that for commercial onion oil. This justifies the fresh-like flavor of supercritical C02 extracts of onion (with the use of entrainer) as opposed to cooked flavor in case of commercial oils (which are steam distilled). Table 3: Results of Thiosulfinate Analysis. Sample Absorbancea SFE-C02 + Ethanol L673" SFE-C02 + Octane 1.573b Steam distilled commercial oil 0.308“ a Measured at 515 nm in triplicate b, c Numbers followed by different letters are significantly different at the 1% level. By smelling the extracts it was found that SFE-C02 extracts (with and without entrainer) had a similar, fresh-onion like, smell. On the other hand, steam distilled extracts had the unpleasant smell of cooked onions. Hence, on the basis of this paragraph and the previous one, we can say that SFE-C02 extracts (with or without entrainer) have the unique and pleasant smell of fresh onions. Figures 9 and 10 show typical gas chromatograms of SFE-C02 extract of onion and of steam distilled commercial onion oil respectively. The two flavor profiles 49 are very different There are more peaks in the Chromatogram of SFE-C02 extract than in the Chromatogram of steam distilled extract. This is an evidence of the presence of extra compounds in SFE-C02 extracts which are either absent or present in very small amounts in the steam distilled extracts. Figures 11 and 12 show gas chromatograms of the dichloromethane extract of fresh onion juice and the dichloromethane extract of onion juice after it has undergone supercritical fluid extraction. The sample that has undergone SFE-C02 has very small peaks indicating very small amounts of the compounds present. This indicates that supercritical fluid extraction using C02 is an efficient method of isolating onion flavor oil. 50 1.6.6 1 t 1.495- 1 1.2.54 4 1 .0054 -0..- ) 6.004— ;:: M J UL WMM I 1 '0 2'0 30 4Y0 a 0 Time (min) Detector ReSponse Figure 9. Gas Chromatogram of Onion Oil Obtained by Supercritical C02 Extraction. Detector Response 9 o o r e - 4.0a4— 2.0a4n J,“ “I" l ‘ J .. 0 1'0 are 3'0 4'0 go Tune (11in) Figure 10. Gas Chromatogram of Commercial Onion Oil Obtained by Steam Distillation. 51 Detector Response Tum (min) Figure l 1. Gas Chromato gram of a Dichloromethane Extract of Fresh Onion Juice. 1.2.64 1 1 .005 4 6.000 4 Detector Response L - l l 1.1 L; .l .lagdaag WW fifl Time (min) Figure 12. Gas Chromatogram of a Dichloromethane Extract of Onion Juice after SFE-COL 5. SUMNIARY & CONCLUSIONS A comparison between various methods of extracting onion flavor oil has been made with respect to the yield and the quality of the extracts. Supercritical fluid extraction with C02 resulted in higher yield and better quality extracts than steam distillation—solvent extraction (the latter is the most widely used method for commercial production of onion flavor oil). The use of polar entrainer, ethyl alcohol with SFE-C02 was found to enhance the yield of SFE-C02. Ethyl alcohol is considered a food-grade solvent. Thus, onion flavor oil produced by SFE-C02 with ethyl alcohol as entrainer will result in a high yielding, safe (free from organic solvent residues) and unique fresh onion- like flavor. The commercial potential of the unique fresh onion-like flavor extracted by supercritical C02 extraction can be supported by the results of a trained panel organoleptic evaluation of similar extracts (Appendix D). 1. Supercritical fluid extraction, using carbon dioxide as solvent, is a promising alternative to steam distillation-solvent extraction which is the most widely used method for extraction of onion flavor oil. It is possible to produce high yielding and better quality extracts using supercritical fluid extraction. 2. Lower flow rate of carbon dioxide in supercritical fluid extraction of onion oil from onion juice, under the conditions studied, resulted in a higher yield of onion oil. 3. The use of 75 ml ethyl alcohol as entrainer in supercritical C02 extraction enhanced the yield of onion oil obtained from the extraction without entrainer 52 53 under similar conditions (3600 psi (24.5 MPa), 37°C, 1.0 llmin C02 flow rate). . The yield of onion oil was greater at 2600 psi ( 17.7 MPa) than at 3600 psi (24.5 MPa) in experiments with 75 m1 ethyl alcohol as entrainer. This indicates that the supercritical extractions can be conducted at lower pressures. . Supercritical and liquid C02 extracts had fresh onion like flavor as opposed to cooked flavor obtained by steam distillation-solvent extraction. 6. SUGGESTIONS FOR FURTHER RESEARCH This study has indicated that supercritical fluid extraction using carbon dioxide as solvent is a promising alternative to the traditional methods used for extraction of onion flavor oil. There is a need for a detailed study to optimize the solvent power of C02 with respect to extraction of onion oil in which experiments are conducted at various temperatures and pressures. Use of ethyl alcohol and octane as entrainers in the SFE-C02 process have been studied in this work. Information on the use of other solvents as entrainers could help understand the effect of entrainers on the process more clearly. In addition, the possibility of obtaining same or higher yields at lower pressures in the presence of an entrainer needs further study. GC and GC-MS techniques may preferably be replaced by high performance liquid chromatography (HPLC), cryogenic GC, SCF chromatography or other analytical techniques which employ lower temperatures. This is desirable because some onion flavor components are therrnolabile and there is a possibility of their degradation or rearrangement under high temperatures which are usually employed in GC injectors and detectors (Block et al., 1992; Block, 1993). 54 BIBLIOGRAPHY 7. BIBLIOGRAPHY Abraham, K.O., Shankaranarayanana, M.L., Raghavan, B., and Natarajan, GP. 1976. Alliums-varieties, chemistry and analysis. Lebensm. Wiss. Technol., 9: 193-200. Andrews, T. 1887. On the properties of matter in the gaseous and liquid state under various conditions of temperature and pressure. Phil. Trans. Roy. Soc. (London), 178:45. Angus, 8., Armstrong, 8., and de Reuck, K.M. 1976. International thermodynamic tables of the fluid state: Carbon dioxide. IUPAC Project Centre, Imperial College, London. Bandyopadhyay, C., Srirangarajan, AN. and Sreenivasan, A. 1970. Studies on flavour components of onion (Allium cepa) 1. Thin layer chromatographic investigation of onion oil. J. Chromatogr., 47: 400-407. Basta, N. 1984. Supercritical Fluids. High Technology, June: 75. Bernhard, RA. 1968. ComparatiVe distribution of volatile aliphatic disulfides derived from fresh and dehydrated onions. J. Food Sci., 33: 298-304. Block, E. 1985. The chemistry of garlic and onion. Sci. Am., 252: 114—119. Block, E., Ahmad, S., Catalfamo, J.M., Jain, MK. and Apitz-Castro, R. 1986. Antithrombotic organosulfur compounds from garlic: Structural, mechanistic and synthetic studies. J. Am. Chem. Soc., 108: 7045—7055. Block, E., Putrnan, D. and Zhao, SH. 1992. Allium chemistry: GC—MS analysis of thiosulfinates and related compounds from onion, leek, scallion, shallot, chive and Chinese chive. J. Agric. Food Chem., 40: 2431-2438. Block, E. 1993. Flavor artifacts. J. Agric. Food Chem, 41: 692. 55 56 Boelens, M., de Valois, P., Wobben, H]. and van der Gen, A. 1971. Volatile flavor compounds from onion. J. Agric. Food Chem., 19: 984-990. Brodnitz, M.H., Pollock, CL. and Vallon, RP. 1969. Flavor components of onion oil. J. Agric. Food Chem., 17: 760-763. Brodnitz, M.H. and Pollock, CL. 1970. Gas chromatographic analysis of distilled onion oil. Food Tech., 24 (79): 78-80. Brunner, G., and Peter, S. 1982. On the solubility of glycerides and fatty acids in compressed gasses in the presence of an entrainer. Sep. Sci. and Tech., 17: 199- 214. Brunner, G. 1983. Selectivity of supercritical compounds and entrainers with regard to model substances. Fluid Phase Equil., 10: 289-298. Buchner, EH. 1906. Liquid carbon dioxide as a solvent. Zeitschr. Phys. Chem., 54: 665. Calame, J .P. and Steiner, R. 1982. Carbon dioxide extraction in the flavor and perfumery industries. Chem. Ind., June: 399. Carson, J.F. and Wong, RF. 1959. A colour reaction for thiosulfinates. Nature, 183: 1673. Carson, J.F. 1987. Chemistry and biological properties of onions and garlic. Food Rev. Int., 3(1,2): 71-107. Cohen, RS. 1984. Supercritical extraction: New processing technique. Prepared Foods, September: 115-116. Copella, SJ. and Barton, P. 1987. Supercritical carbon dioxide extraction of lemon oil. In: Supercritical Fluids: Chemical Engineering Principles and Applications. Squires, T.G. and Paulitis, M.H. (Editors). ACS Symposium Series No. 329. 57 Cottle, J .E. 1966. Supercritical polymerization of olefms. US Patent 3,294,772. Dandge, D.K, Heller LP. and Wilson K.V. 1985. Structure solubility correlations: Organic compounds and dense carbon dioxide binary systems. Ind. Eng. Chem. Prod. Res. Dev., 24: 162-166. Degnan, A.J., von Elbe, J.H. and Hartel, RW. 1991. Extraction of annatto seed pigment by supercritical carbon dioxide. Journal of Food Science, 56:1655-1659. DeSimone, J.M., Hellsten, AM., Siochi, E.J., Smith, S.D., Ward, T.C., Gallagher, P.M., Krukonis, V.J. and McGrath, J .E. 1990. Homogeneous and multiphase poly (methyl methacrylate) graft copolymers via the macromonomer method. Makromol. Chem., Macromol. Symp., 32: 21. Dobbs, J.M., Wong, J.M. and Johnston, KP. 1986. Nonpolar co-solvents for solubility enhancement in supercritical fluid carbon dioxide. J. Chem. Eng. Data, 31: 303- 308. Dobbs, J.M., Wong, J.M., Lahiere, RJ. and Johnston, KR 1987. Modification of supercritical fluid phase behavior using polar co-solvents. Ind. Eng. Chem. Res., 26: 56-65. Eggers, R. and Tschiersch, R. 1978. Development and planning of a plant for high- pressure extraction of natural products. Chem. Ing. Tech., 50: 842. Elgin, J.C. and Weinstock, J .J . 1959. Phase equilibrium at elevated pressures in ternary systems of ethylene and water with organic liquids. Salting out with a supercritical gas. J. Chem. Eng. Data, 4:3. Ely, J.F. and Baker, J.K. 1983. A review of supercritical fluid extraction. National Bureau of Standards Library, US. Government Printing Office, Washington. 58 Fenaroli, G. 1971. Onion, In: Handbook of Flavor Ingredients. The Chemical Rubber Co. Fenwick, GR. and Hanley, AB. 1985. The genus Allium. CRC Crit. Rev. Food Sci. Nutr., 22: 199-271. Francis, A.W. 1954. Ternary systems of liquid carbon dioxide. J. Phys. Chem., 58: 1099. Francis, A.W. 1955. Solvent extraction with liquid carbon dioxide. Ind. Eng. Chem., 47: 230. Freeman, G.G., and Mossadeghi, N. 1970. Effect of sulphate nutrition on flavour components of onion (Allium cepa). J. Sci. Food Agric., 21: 610-615. Freeman, G.G., and Whenham, R]. 1974. Changes in onion (Allium cepa L.) flavour components resulting from some post-harvest processes. J .Sci Food Agric., 25: 499-515. Gopal, J.S., Holder, G.D. and Kosal, E. 1985. Solubility of solid and liquid mixtures in supercritical carbon dioxide. Ind. Eng. Chem. Process Des. Dev., 24. Grimmett, C. 1981. The use of liquid carbon dioxide for extracting natural products. Chemistry and Industry, 16 May: 359-362. Guckes, T.L., McHugh, M.A. Cozewith, C. and Hazelton, R.L. 1990. Phase separation processes. US. Patent 4,946,940. Gurdial, G.S., Foster, N.R., Yun, S.L.J. and Tilly, K.D. 1993. Phase behavior of supercritical fluid-entrainer systems. In: Supercritical Fluid Extraction Science, Fundamentals and Applications. Kiran, E. and Brennecke, J.F. (Editors), ACS, Washington, DC. Hannay, J.B. and Hogarth, J. 1879. On the solubility of solids in gases. Proc. Roy. Soc. (London), 29: 324. 59 Heath, H. 1973. Herbs and spices-A bibliography. VI. Flavor Ind., 4: 346-348. Heller, SR, and Milne, G.W.A. 1978. EPA/NIH Mass Spectral Data Base. US. Department of Commerce/National Bureau of Standards. Hubert, P. and Vitzthum, OF. 1978. Extraction of hops, spices and tobacco with supercritical gases. Agnew. Chem. Int. Ed. Engl., 17: 710-754. Kallio, H. and Salorinne, L. 1990. Comparison of onion varieties by headspace gas chromatography- mass spectrometry. J. Agric. Food Chem., 38: 1560-1564. Kay, W.B. 1938. Liquid-vapor phase equilibria in the C2H6 n-heptane system. Ind. Eng. Chem., 30:459. Kim, S., Wong, J.M., and Johnston, K.P. 1985. Theory of the pressure effect in dense gas extraction. In: Penninger, J.M.L., Radosz, M., McHugh, M.A. and Krukonis V.J. (Editors), Supercritical Fluid Technology, Elsevier Science Publishers, Amsterdam, The Netherlands, 45-66. King, MB. and Bott, T.R. (Editors) 1993. Extraction of Natural Products Using Near- Critical Solvents. Blackie A & P. London. Krase, NW. 1945. Separating ethylene polymers. US. Patent 2,388,160. Kuk, MS. and Montagna, J.C. 1983. Solubility of oxygenated hydrocarbons in supercritical carbon dioxide. In: Chemical Engineering at Supercritical Fluid Conditions, Paulitis, M.E., Penninger, J.M.L., Gray, JD. and Davidson, P. (Editors), Ann Arbor Science, Ann Arbor, MI. Kuo, M.C., Chien, M. and Ho, CT. 1990. Novel polysulfides identified in the volatile components from Welsh onions (Allium fistulosum L. var. maichuon) and scallions (Alliumfistulosum L. var. caespitosum). J. Agric. Food Chem., 38: 137 8- 1381. Kuo, MC. and Ho, CT. 1992. Volatile constituents of distilled oils of welsh onions (Allium fistulosum L. Variety Maichuon) and scallions (Allium fistulosum L. Variety Caespitosum). J. Agric. Food Chem., 40: 111-117. Kumik, RT. and Reid, RC. 1982. Solubility of solid mixtures in supercritical fluids. Fluid Phase Equil., 8: 93-105. Laws, D.R.J., Bath, N.A., Pickett, J.A., Ennis, CS. and Wheldon, A.G. 1977. Preparation of hop extracts without using organic solvents. J. Inst. Brew., 83: 39- 40. Likens, ST. and Nickerson, GB. 1964. Detection of certain hop oil constituents in brewing products. Proc. Am. Soc. Brew. Chem., 5-13. Mazza, G. and LeMaguer, M. 1979. Volatiles retention during the dehydration of onion (Allium cepa L.). Lebensm.-Wiss. Technol., 12: 333-337. Mazza, G., LeMaguer, M. and Hadziyev, D. 1980. Headspace sampling procedures for onion (Allium cepa L.) aroma assessment. Can. Inst. Food Sci. Technol. J., 13: 87-96. McHugh, M.A. and Krukonis, V.J. 1994. Supercritical Fluid Extraction, Principles and Practice, 2nd edition. Butterworth-Heinemann, Stoneharn, MA. Messmore, HE. 1947. Asphaltic materials. US. Patent 2,420,185. Milleville, HR and Eskew, R.K. 1946. Recovery of volatile apple flavors in essence form. Western Canner and Packer, 38: 51-54. Modell, M., de Filippi, RP. and Krukonis, V.J. 1978. Regeneration of activated carbon with supercritical carbon dioxide. Paper presented at the ACS Annual Meeting, 61 Moses, J.M., Goklen, KB. and de Filippi, RP. 1982. Pilot plant critical fluid extraction of organics from water. Paper presented at the Annual AIChE Meeting, Los Angeles, November. Nakata, C., Nakata, T. and Hishikawa, A. 1970. An improved colorimetric determination of thiosulfinate. Anal. Biochem., 37: 92-97. Nguyen, K., Barton, P. and Spencer, J .S. 1991. Supercritical carbon dioxide extraction of vanilla Journal of Supercritical Fluids, 4: 40-46. Nishimura, H., Wijaya, H. and Junya, M. 1988. Volatile flavor components and antithrombotic agents: vinyldithiins (from Allium victorialis). J. Agric. Food Chem., 36: 563-566. Nock, LP. and Mazelis, M. 1987. The C-S Lyases of higher plants. Direct comparison of the physical properties of homogeneous alliin lyase of garlic (Allium sativum) and onion (Allium cepa). Plant Physiol., 85: 1079-1083. Novak, RA and Robey, RJ. 1989. Supercritical fluid extraction of flavoring material: Design and economics. In: Supercritical Fluid Science and Technology. Johnston, K.P. and Penninger, J .M.L. (Editors), ACS Symposium Series 406, American Chemical Society, Washington, DC. Paulaitis, M.E., Gilbert, ML. and Nash, CA. 1981. Separation of ethanol-water mixtures with supercritical fluids. Paper presented at the 2nd World Congress of Chemical Engineers, Montreal, Canada, October. Paulaitis, M.E., Krukonis, V.J., Kumik, RT. and Reid, RC. 1983. Supercritical fluid extraction. Reviews in Chem. Eng., 1: 179-250. 62 Peter, S. and Brunner, G. 1978. Separation of nonvolatile substances by means of compressed gases in countercurrent processes. Agnew. Chem. Int. Ed. Engl., 17: 744. Rizvi, S.S.H., Benado, A.L., Zollweg, J.A. and Daniels, 1A 1986. Supercritical fluid extraction: Fundamental principles and modeling methods. Food Technol., 40: 55- 64. Roop, R.K. and Akgerman, A. 1989. Entrainer effect for supercritical extraction of phenol from water. Ind. Eng. Chem. Res., 28: 1542-1546. Roselius, R., Vitzthum, O. and Hubert, P. 1972a. Nicotine removal from tobacco. German Patent 2,043,537, May 31. Roselius, R., Vitzthum, O. and Hubert, P. 1972b. Recovery of coffee oil containing aroma ingredients from roasted coffee. German Patent 2,106,133, August 31. Rossi, M., Schiraldi, A. and Spedicato, E. 1989. Supercritical fluid extraction of cholesterol and its oxidized products from eggs: Preliminary experiments. In: Flavors and Off-Flavors, Charalambous, G. (Editor), Proceedings of the 6th International Flavor Conference, Rythmnon, Crete, Greece. 855-863. Sage, B.H., Webster, DC. and Lacey, W.N. 1936. Phase equilibria in hydrocarbon systems. XVI. Solubility of methane in four light hydrocarbons. Ind. Eng. Chem., 28: 1045. Saghir, A.R., Mann, L.K., Bernhard, RA and Jacobsen, J.V. 1964. Determination of aliphatic mono- and disulfides in Alliums by gas chromatography and their distribution in the common food species. Proc. Amer. Soc. Hort. Sci., 84: 386- 398. 63 Schmitt, W.J. and Reid, RC. 1986. The use of entrainers in modifying the solubility of phenanthrene and benzoic acid in supercritical carbon dioxide and ethane. Fluid Phase Equilib., 32: 77. Schneider, G.M., Stahl, E. and Wilke, G. 1980. Extraction with Supercritical Gases. Verlag Chemie. Deerfield Beach, Florida. Schultz, T.H., Flath, R.A., Mon, T.R., Eggling, SB. and Teranishi, R. 1977. Isolation of volatile components from a model system. J. Agric. Food Chem., 25: 446-449. Schultz, W.G. and Randall, J.M. 1970. Liquid carbon dioxide for selective aroma extraction. Fd. Tech., 24(11): 1282-1286. Sharma, A., Tewari, G.M., Shrinkhande, A.J., Padwal-Desai, SR, and Bandyopadhyay, C. 1979. Inhibition of aflatoxin producing fungi by onion extracts. J. Food Sci., 44: 1545-1547. Sinha, N.K., Guyer, D.E., Gage, DA. and Lira, OT. 1992. Supercritical carbondioxide extraction of onion flavors and their analysis by gas chromatography-mass spectrometry. J. Agric. Food Chem., 40: 842-845. Sivetz, M. 1963. Coffee Process Technology, Volume 2. AVI Publishing Co., Westport, Connecticut. Spiro, M. and Kandiah, M. 1989. Extraction of ginger rhizome: Kinetic studies with acetone. International Journal of Food Science and Technology, 24: 589-600 Stahl, E., Erwin, S. and Mangold, H.K. 1980. Extraction of seed oils with liquid and supercritical carbon dioxide. J. Agric. Food Chem., 28: 1153. Stahl, E., Quirin, K.W. and Gerard, D. 1988. Dense Gases for Extraction and Refining. Springer-Verlag, New York. 64 Stoll, A and Seebeck, E. 1951. Chemical investigations on alliin, the specific principle of garlic. Advan. Enzymol., 11: 377-400. Teranishi, R., Flath, RA and Sugisawa, H. 1971. In: Flavor Research: Principles and Techniques. Marcel Dekker, Inc. Tewari, GM. and Bandyopadhyay, C. 1977. Pungency and lachrymatory factor as a measure of flavor strength of onions. Lebensm.-Wiss. Technol., 10: 94-96. Thomas, D.J., Parkin, KL. and Simon, P.W. 1992. DeveIOpment of a simple pungency indicator test for onions. J. Sci. Food Agric., 60: 499-504. VanAlsten, J.G., Hansen, RC. and Eckert, CA 1984. Supercritical enhancement factors for non—polar and polar systems. Paper 84a presented at the AIChE 1984 annual meeting, San Francisco. Vitzthum, O. and Hubert, P. 1976. Spice extracts. Canadian Patent 989,662, May 25. Vitzthum, O. and Hubert, P. 1979. Process for the manufacture of caffeine-free black tea. U.S. Patent 4,167,589. Walsh, J.M., Ikonomou, G.D., and Donohue, MD. 1987. Supercritical phase behavior: The entrainer effect. Fluid Phase Equilibria, 33: 295-314. Whitaker, JR. 1976. Development of flavor, odor, and pungency in onion and garlic. Adv. Food Res., 22: 73-133. Wild, L., Ryle, T.R. Knobeloch, DC. and Peat, LR. 1982. Determination of branching distributions in polyethylene and ethylene copolymers. J. Polym. Sci. Polym. Phys. Ed., 20: 441-445. Wit, de J.C., Noterrnans, G.N. and Kampelmacher, EH. 1979. Effect of garlic oil or onion oil on toxin production by Clostridium botulinum in meat slurry. J. of Pd. 65 Prot., 42 (3): 222-224. Yagami, M., Kawakishi, S. and Namiki, M. 1980. Identification of intermediates in the formation of onion flavor. Agric. Biol. Chem., 44 (11): 2533-2538. Zosel, K. 1972. Caffeine from crude coffee. German Patent 2,221,560, November 30. Zosel, K. 1982. Process for the direct decaffeination of aqueous coffee extract solutions. U.S. Patent 4,348,422. APPENDICES APPENDIX A Quantitative GC Analysis to Estimate Onion Oil Yield Addition of an appropriate internal standard in the beginning of the analysis is an important step in quantitative GC analysis. 0.005 g of benzyl disulfide (internal standard) was added to onion juice before extraction in each experiment. The following simplifying assumptions were made: 1. The internal standard and the onion flavor compounds have similar extraction properties and GC and MS responses. 2. The total weight of sulfur compounds present in the extracts was considered as an indicator of the yield of onion oil. GC-MS was used to identify the sulfur compound peaks. This was done by matching corresponding peaks on the total ion chromatogram (TIC) obtained from GC-MS and the chromatogram obtained from GC. Figure 13 shows the TIC for SFE- C02 + Ethanol, 3600 psi, 37°C, extract. Figure 14 shows the GC chromatogram of the same extract. Twenty large sulfur peaks were chosen for the analysis. One of these was the internal standard peak. These twenty peaks were identified on GC chromatograms of all extracts and their peak areas were obtained from the Area Percent Reports. The Area Percent Report for SFE-C02 + Ethanol at 3600 psi (24.5 MPa), 37°C is shown in Table 4. Each extract was analyzed in triplicate.Table 5 shows the peak areas of the twenty selected peaks for experiment SFE-C02 + Ethanol at 3600 psi (24.5 MPa), 37°C. 67 Scam COR .aa 88 .E 3 353 + «8.9.; .6 sewageoau :2 as... .2 use". CMUm ooov ooon ooom Good 0 ONM.Omm O.~c > > u u . . > . . . r . . r p p r r . 1 lo T a row 1 r rov f v 0 row 0 v C v m . u 40¢ a . s . n o: . < 4 4 4 4 q 1 4 4 4 a 4 4 4 4 . 4 4 4 4 q 4 4 4 4 4 4 4 4 4 OOH .UCQDG .OmEow Om Ov Om ON OH Acfiev oEfiB m _ a I I n H O A o — H m + F— 00 ON a A. vane... ON m: CH p F ’ P r b b > n L p D b h Li » r b l r 2: 333;: j 1531474 53 :44 nvnhw AMC ON 82.558 3 2&5 ES 25 0 «v on p > [r u r b ('3 Aha Av Mn 10 000. 70 Table 4: Area Percent Report of SFE-C02 + Ethanol (75 ml, 3600 psi, 37°C) Extract Pk! Ret Tine 11.381 11.690 12.257 15.749 15.917 16.795 17.929 18.146 18.348 18.599 19.111 19.427 19.557 19.727 20.241 20.954 21.047 23.034 23.536 23.693 24.063 24.449 24.563 24.663 24.943 25.130 26.292 26.459 26.609 27.298 27.752 28.050 28.946 29.607 30.095 30.562 30.750 30.980 31.083 31.196 31.382 31.653 31.756 31.877 32.168 32.297 32.386 32.816 33.003 934847 151642 239716 205729 2336415 666972 811847 216169 308840 1401112 290850 306386 407259 534907 433759 281759 240565 577784 326196 211044 237371 233582 408606 196847 386788 1906242 703588 651115 1016782 167926 167346 537771 231060 477919 268521 195117 455163 216656 1219514 464663 299098 477812 354552 169203 229706 182923 196808 193982 230845 627609 obtained by GC 117442 31051 54178 34694 388523 136856 155937 44874 69181 158269 58570 61727 78805 115321 95480 41145 56839 115266 53113 31335 36526 23990 65377 46187 84492 282319 97232 104729 182740 21229 20049 41925 15960 88107 32241 23491 42452 35105 123281 93400 57770 70685 41365 30331 37073 25515 28330 29209 26717 114185 22522555222552222522522552555222322222225622222$223 2.4638 0.3996 0.6318 0.5422 6.1575 1.7578 2.1396 0.5697 0.8139 3.6926 0.7665 0.8075 1.0733 1.4097 1.1432 0.7426 0.6340 1.5227 0.8597 0.5562 0.6256 0.6156 1.0769 0.5188 1.0194 5.0238 1.8543 1.7160 2.6797 0.4426 0.4410 1.4173 0.6090 1.2595 0.7077 0.5142 1.1996 0.5710 3.2140 1.2246 0.7883 1.2593 0.9344 0.4459 0.6054 0.4821 0.5187 0.5112 0.6084 1.6540 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 71 Table 4 (Continued) Pk! Ret Tine Area Height ITypeI Width | Area 4 33.116 560045 103693 VV 0.079 1.4760 33.274 154937 20447 VV 0.101 0.4083 33.431 199617 22085 VV 0.123 0.5261 33.649 216390 18938 VV 0.160 0.5703 34.593 168730 16923 VV 0.134 0.4447 34.938 272775 19221 VV 0.185 0.7189 35.264 301184 43609 VV 0.097 0.7938 35.387 602018 102595 VV 0.085 1.5866 35.492 812547 143575 VV 0.083 2.1414 35.589 1061836 140390 VV 0.102 2.7984 35.846 301263 37324 VV 0.110 0.7940 36.106 668814 76495 VV 0.118 1.7626 36.201 203183 43249 VV 0.069 0.5355 36.355 268015 36765 VV 0.107 0.7063 36.506 432865 48381 VV 0.129 1.1408 36.741 525072 45016 VV 0.159 1.3838 37.099 175498 27333 VV 0.090 0.4625 37.264 364710 35193 VV 0.139 0.9612 37.401 158133 29680 VV 0.078 0.4168 37.489 232993 31219 VV 0.104' 0.6140 37.701 286684 32065 VV 0.122 0.7555 37.854 514187 54921 VV 0.128 1.3551 38.642 210090 13658 VV 0.199 0.5537 39.013 287220 19045 VV 0.198 0.7570 39.416 433173 48733 VV 0.129 1.1416 39.590 256805 27611 VV 0.129 0.6768 40.396 2481299 223121 VV 0.147 6.5394 41.227 313073 25111 VV 0.170 0.8251 42.403 1491929 199127 VV 0.118 3.9319 79 area - 3.7944E+007 72 (75 ml, 3600 psi, 37°C). Table 5: Peak Areas of Sulfur Compounds and lntemal Standard for SFE-C02 + Ethanol Peak No. Peak Area 1 Peak Area 2 Peak Area 3 1 934847 3620787 821680 2 2336415 2726697 2250953 3 666972 580099 712804 4 1401112 1421560 1369923 5 - 816678 1028352 6 534907 866853 672291 7 577784 1199303 781626 8 326196 1008938 442970 9 1906242 3447582 2206183 10 703588 1112474 811043 11 651115 840105 832081 12 1016782 1287419 1206924 13 537771 797538 538039 14 477919 719733 476052 15 1219514 2128427 140915 16 464663 749909 639012 17 627609 1378408 585982 18 560045 1087334 816499 19 1061836 3110072 1373058 20(IS) 1271368 2103342 1364128 2 Peak Area (1- 19) 16005317 28899916 17706387 73 The following equation was used to evaluate the total weight of sulfur compounds present in the extracts and thus provide an estimate of the onion oil (ZPeak Area of S Compounds) x Wt of IS Peak Area of IS Total Wt. of S Compounds = For Peak Area 1: 2 Peak Area of S Compounds = 16005317 Peak Area of IS = 1271368 Weight of IS = 0.005 g Thus, Total Wt. of S Compounds = 16005317 * 0.005/1271368 = 0.0629 g Similarly, For Peak Area 2: 2 Peak Area of S Compounds = 28899916 Peak Area of IS = 2103342 Weight of IS = 0.005 g Thus, Total Wt. of S Compounds = 28899916 "' 0.005/2103342 = 0.0687 g Similarly, For Peak Area 3: 2 Peak Area of S Compounds = 17706387 Peak Area of IS = 1364128 Weight of IS = 0.005 g Thus, Total Wt. of S Compounds = 17706387 * 0.005/1364128 = 0.0649 g Total Wt. of S Compounds (Average) = 0.0655 g Similarly, the total weight of sulfur compounds (average of triplicate experiments) were calculated for the other extracts and these data were presented in the form of a bar chart in Figure 6. Table 6: Mass Spectral and Retention Data for Onion Flavor Compounds Detected in APPENDIX B Various Extracts Studied. No. Compound, Mass MS Datab dimethyl disulfide, 94 706 9603.1), 95(5.7), 94000), 79(48.0), 64(3.0), 61(9.2), 48(20.4), 46(41.6), 45(62.3) methyl ethyl disulfide, 108 831 110(8.0), lO9(4.2), 108(30.0), 7903.8), 6605.0), 64(3.2), 4705.8), 4501.1), 43(100) 2,5-dimethy1thiophene, 112 862 1140.5), 11302.3), 11204.6), 111000), 97(45.0), 79(5.4), 7702.6), 6900.2), 6700.9), 51(3.8), 45(27.7) 2,4—dimethylthiophene, 112 863 114(3.1), 11302.3), 112(100), 111(91.5), 97(68.2), 79(8.0), 7705.0), 6900.2), 6700.0), 51(6.3), 45(29.0) 3,4—dimethylthiophene, 1 12 866 114(4.8), 11302.9), 112(100), 111(97.2), 9700)), 79(9.4), 7705.0), 69(482), 67(9.l), 51(21.0), 45(84.7) diallyl sulfide, 114 870 116(4.8), 1150.0), 114(33.2), 11308.4), 101(8.6), 99000), 84(4.9), 73(9.5), 7203.0), 7101.9), 6505.3), 5903.8), 5505.1), 45(35.4), 41(24.7) methyl propyl disulfide, 122 910 12405.3), 123(5.0), 122000), 80(88.4), 6403.1), 60(6.2), 4702.2), 4602.2), 45(44.0), 43(36.8), 41(29.4) methyl cis-pmpenyl disul- fide, 120 933 122(4.2), 121(6.1), 12006.8), 105(3.9), 80(26.l), 75(322), 72(26.5), 6106.6), 4707.5), 4601.0), 45000), 41(29.4) methyl trans-propenyl disul- fide, 120 942 12205.0), 12102.6), 120000), 105(9.7), 8000.0), 75(44.8), 72(34.1), 61(5.1), 4704.8), 46(24.7), 45(902), 4104.2) 10 dimethyl trisulfide, 126 953 128(9.8), 127(4.7), l26(38.3), 1110.9), 800.0), 79(42.5), 6400.0), 61(5.1), 47(32.1), 4609.8), 45000) 74 75 Table 6: Mass Spectral and Retention Data for Onion Flavor Compounds Detected in Various Extracts Studied. No. Compound. Mass 11‘ MS Data” 11 unknown, 130 1017 132(4.8), 13100.3), 130000), 11503.0), 114(2.3), 11302.2), 101(83.1), 8504.9), 69(26.8), 6804.0), 67(39.6), 5903.0), 5303.9), 45(20.0), 41(31.1), 3908.2) 12 dipropyl disulfide, 150 1057 1520.2), 151(4.0), 15008.3), 10807.9), 7500.6), 730.8), 66(6.4), 61(8.9), 4707.3), 45(68.5), 4305.0), 41000) 13 1-propenylpropyldisulfide, 1062 15001.4), 149(8.8),l48(100), 10600.1), 148 73(40.1), 72(42.2), 7101.0), 61(22.5), 47(22.4), 4506.3), 43(35.0), 41(672) l4 3-etheny1 1,2-dithi-5-ene, 1068 147(4.6), 14607.5), 113000), 112(8.0), 146 11101.8), 9706.2), 850.9), 7909.6), 7702), 59(5.7), 45(8.4), 41(6.0) 15 methyl propyl trisulfide, 154 1072 15602.0), 15508.3), 154000), 114(49.0), 11303.8), 112(32.8), 7905.6), 6406.1), 47(69.0), 4608.5), 4503.0), 4300.3), 4101.2) 16 propyl methane thiosul- 1075 15603.8),155(5.7),15408.0),1390.5), fonate, 154 1380.4), 11203.1), 970.0), 7903.8), 64(43.2), 4700.0), 45005). 43000), 41(99) 17 methyl l-propenyl trisulflde 1079 1530.0), 152(100), 8809.8), 7300.0), (E/Z), 152 640.6), 4701.3). 4604.5), 4508.2), 410.8). 400.1), 3900.0) 18 methyl l-propenyltrisulfide 1081 1540.0), 1530.6),152000), 1110.1), (Ii/Z), 152 88(67.2), 7305.0), 640.3), 470.1), 46(9.4), 4506.8), 4101.0), 3905.8) 19 3-ethenyl 1241101441», 1084 1480.8), 1470.0), 146(35.0),113(100), 146 1120.1), 11104.4), 9806.1), 9706.1), 85(6.9), 7908.3), 77(8.2), 59(5.1), 4504.0), 41(6.3), 390.5) 20 3,4 dihydro-3-vinyl-l,2- 1091 146(8.5), 14503.8), 144000), dithiin, 144 12900.0), 11100.4), 9902.8), 85(6.7), 7709.2), 70(11.5),69(18.0), 6806.3), 6700.0), 59(8.3), 4109.1) 21 methyl 5-methylfuryl sul- 1220 130(O.8), 1290.1),128000), 11300.0), fide, 128 10008.3), 9906.4), 85(98.0), 6707.0), 6605.0), 6504.7), 5908.1), 55(8.8), 5108.2), 45(67.8), 4107.0) 76 Table 6: Mass Spectral and Retention Data for Onion Flavor Compounds Detected in Various Extracts Studied. No. Compound, Mass 11“ MS Data” 22 diallyl thiosulfinate, 162 1225 1640.6), 16204.6), 12900.0), 990.1), 87(4.2), 86(4.6), 8501.0), 69(100), 5905.4), 5502.1), 4508.7), 4104.3) 23 l-propenyl propyl trisulfrde 1249 18208.5), 18102.0), 180000), (M), 180 164(11.8),151(5.1), 11602.2), 115(40.0), 106(41.0), 101(6.3), 8(9.1), 8304.2), 750.0), 74(47.3), 7305.7), 6404.6), 5903.8), 470.4), 4503.0), 41(52.5) 24 dipropyl uisulfide, 182 1254 18405.2), 183(X).3), 182000), 140(8.2), 117(5.0), 98(6.2), 890.3), 7506.0), 47(9.1), 4500.8), 4301.8), 4101.7) 25 l-propenyl propyl uisulfide 1261 18202.6), 181(8.1), 180000), 1380.3), (E/Z), 180 11607.2), 10505.0), 75(8.8), 7409.6), 7304.7), 6107.5), 5905.2), 47(9.8), 45(57.0), 4302.2), 41(48.4), 3904.3) 26 diallyl trisulfide, 178 1263 18203.0), 1810.2), 180(65.1), 1790.5), 178000), 14708.0), 1310.0), 11404.8), 1130.3), 99(61.6), 790.2), 73(58.3), 6107.7), 4705.8), 4501.0), 43(42.5), 4200.0) 27 dibenzothiophene, 184 1265 18502.8), 184000), 16905.0). 15505.9), 15105.0), 140(50.8), 139(47.6), 12509.7), 11200.4), 9103.2), 8503.3), 7902.4), 6902.0), 6802.7), 6407.5), 5901.4), 5401.0), 4500.4), 41(49.1) 28 methyl 3,4-dimethyl-2-thie- 1429 19204.3), 191(4.8), 190(69.0), nyl disulfrde, 190 143000), 11(9.8), 97(5.5), 850.7), 69(5.4), 6700.1), 650.8), 590.8), 550.6), 4500.0), 410.4), 3900.8) 29 4,6-diethyl-1,2,3,5- 1484 213(4.2), 21207.3), 19202.0), tetrathiane, 212 14602.8), 1390.0), 1380.1), 11500.7), 9905.0), 74000), 73(62.6), 64(45.8), 590.4), 5502.2), 4607.0), 45(35.4), 430.7), 4202.1), 4104.0) 30 2,4-dimethyl-5,6—dithia-2,7- 1688 2180.5), 12900.6), 7302.5), 71(9.6), nonadienal-S-oxide, 218 70(6.0), 69000), 5907.5), 470.3), 45(44.5), 41(59.2) 77 Table 6: Mass Spectral and Retention Data for Onion Flavor Compounds Detected in Various Extracts Studied. No. Compound, Mass 1k. MS Datab 31 5,7—diethyl 1,2,3,4,6 pen- 1817 18005.8), 17001.9), 13902.0), tathiepane, 264 106(63.2), 99(54.0), 7400.2), 7307.7), 6404.6), 5908.1), 45(67.0), 41000) 32 molecular sulfur, 256 1878 257(41.2), 256000), 2240.0), 1940.4), 1930.8), 19201.3), 16203.2), 1610.8), 16002.2), 13001.0), 129(0.8), 12806.4), 980.7), 970.7), 9609.5), 660.2), 650.8), 6404.2) a. Kovats retention indices 1). m/z (intensity) 78 .c 2%... £3, tcoamotoo EonEzz x8; .Sabxm COR .Vm coon ._E mt 6525 + NOU-mn_m mo EEonEoEU .80 .2 8:9”. 2:5 25... :o_.:20x h h. 22 m m. 8 v. N. = APPENDIX C Statistical Analysis of Gravimetric Yield Data Gravimetric yield data: SFE (1) SFE (0.5) Liq C02 0.225 0.2607 0.1843 0.2341 0.2579 0.164 0.227 0.259 0.1675 ANOVA Table: Anova: Single-Factor Summary (alpha = 0.01) Group: Count Sum Average Variance SFE (1) 3 0.6861 0.2287 2.29505 SFE (0.5) 3 0.7776 0.2592 1.99E-06 Liq 002 3 0.5158 0.171933 0.000118 ANOVA Source of Variation ss df MS F Between Gr 0.011768 2 0.005884 123.7684 VVIthin Gro 0.000285 6 4.756-05 Total 0.012053 8 Fisher’s Protected Least Significant Difference (LSD) Analysis: LSD= 0.0209 P-velue 1 .335-05 F crit 10.92485 Since all the pairwise differences of means were found to be greater than the LSD. hence. the population means for the three methods are different at the 1% level. 79 80 Statistical Analysis of Quantitative GC Data Yield Data from Quantitative GC Analysis: 1 2 3 Avg StDev 1 0.0525 0.0613 0.0592 0.0577 0.0046 2 0.0854 0.0813 0.0796 0.0833 0.003 3 0.0228 0.0148 0.0214 0.0197 0.0043 4 0.0231 0.0254 0.0292 0.0259 0.0031 5 0.0262 0.0329 0.0297 0.0296 0.0034 6 0.0516 0.0476 0.0454 0.0482 0.0031 7 0.0629 0.0687 0.0649 0.0655 0.003 8 0.0645 0.0677 0.0698 0.0673 0.0027 9 0.0396 0.0334 0.0314 0.0348 0.0043 Analysis of Variance: Single-Factor (alpha = 0.05) Groups Count Sum Average Variance 1 3 0.173 0.057667 2.11E-05 2 3 0.2463 0.0821 8.89E-06 3 3 0.059 0.019667 , 1.83E-05 4 3 0.0777 0.0259 9495-06 5 3 0.0888 0.0296 1.125-05 6 3 0.1446 0.0482 9.88E-06 7 3 0.1965 0.0655 8.6815-06 8 3 0.202 0.067333 7.12E-06 9 3 0.1044 0.0348 1.835—05 ANOVA Source of Variation SS df MS F P- value F crit Between Gr 0.01122 8 0.001403 111.7572 9.47E-14 2.510156 \Mthin Gro 0.000226 18 1.25E-05 Total 0.01 1446 26 Fisher’s Least Significant Difference Analysis: LSD = 2.101 All the pairwise differences were found to be significantly different at the 5% level except the following pairs: 4-5, 5-9, and 7-8. 81 Statistical Analysis of Thiosulfinate Analysis Data Thiosulfinate analysis data: SC-EtOH SC-Oct Stm Dst. 1.784 1.556 0.322 1.524 1.447 0.286 1.711 1.717 0.315 ANOVA Table: Anova: Single-Factor Summary (alpha = 0.01 1 Groups Count Sum Average Variance SC-EtOH 5.019 1.673 0.017983 3 SC-Oct 3 4.72 1.573333 0.01845 Stm Dst. 3 0.923 0.307667 0.000364 ANOVA Source of Variation SS df MS F P-value F crit Between Gr 3.475981 2 1.73799 141.693 8.91E-06 10.92445 \Mthin Gro 0.073595 6 0.012266 Total 3.549576 8 Fisher’s Protected Least Significant Difference (LSD) Analysis: LSD: 0.3352 The first two means were not significantly different from each other but they both were found to be significantly different from the third mean at 1% level. APPENDIX D Organoleptic Evaluation of Supercritical C02 Onion Extracts KALSECvanc. P.O. Box 511, KALAMAIOO, MICHIGAN W11 0 USA PHONE16161 349-9711 I TELEX: 295161 ' FAX1616) 3620060 1' SPICEX TO: Dr. Daniel Guyer Department of Agricultural Engineering. HSU FROH: James A. Guzinski Kalsec, INC. DATE: February 26. 1993 FAX a; 1-517-353-8982 RE: Onion flavor extract Dear Dr. Guyer: Thank you for sending the sample of onion oil you obtained through carbon dioxide extraction. 1 submitted the material to our sensory division for comparison to a typical. commercially available oil manufactured by steam distillation of onions. The sample was diluted by 1 part in 10,000 and rated on our standard flavor attributes. The results of the organoleptic evaluation show that the sample, which was described as having 2.52 onion oil, was weaker than a commercial onion oil tasted at the same level. However, the flavor profile was distinctly different. It was described as having less of the burnt and metallic flavor notes of typical onion oil. As one of the panelists, I thought it was a very good flavor. I am going to resubmit the sample for testing at 502 higher concentration. I think there would be a market for a better, fresher onion flavor. If possible, I would like a larger sample to test in some flavor applications, possibly in an uncooked application such as a salad dressing. I encourage you to continue in this line of research. We are definitely interested in a fresh onion flavor and are willing to evaluate any samples you can provide. Regards, Dr. James A. Guzinski Senior Chemist