Jhlt . $.51.-- .. “Tunas: . 95.» V .‘ , 5 . .. 1",}0h'1 V... ‘1 "I l Infill. I‘lll’lfifiunullil 1.3.?" .Illulll'nulul‘fill 'Ilc’ . . . fl...“ 9‘. I 0’? ~I\ . . Infra. . .II I . v . . c . 1 9 . t . g, . I“! . 5'3 .0.“ v}. t 102 ..,~ 0 v .v c I.. n A n. .5 .o . a [2.012.315 ‘Vlmv n I 1""? l?!.... t . I.II-V"1n01.i‘lll\n.‘ll’"t . él‘ bibliilllll.‘ ' . ~Oc ”m‘ I. v. 5 . . \I it‘ .II .4 1].". I. . v .IIII.|115...5 I. 2.5.... 1.5.0.. . V I... ‘ o ‘ .‘i unlL....fl..1..-.u .I. 55:... 5 . .5. . Ho. . ah‘ 0v . . .. 9%.? . . . v... .qu1 t I-.. . .. . .a......~...u.......u.muu.,....w. u -54»... - madman“ 4 C O . .IV.C.~ . .. Q . .. 3-. {\u‘du ' v.0; v .5»... 5.x»? Ad I no '0. c .99 v. .........v_ .. . w o .35! ~ ' 513$?" .I 5'» - é :15 I X a I ‘ ‘ r ‘ « . o a . .J 5". v o 9- .‘F . . ‘ £1 ‘ . '. ‘- . . "I. _ ' 1. a . E I . l at 7‘ 1- .z.‘ ‘. I. I w. ‘ " I . ‘0 ‘ v I . ‘ a "' . . . . a o ‘ ' ‘ .- z ' ‘ I u ‘ ... I .2 (It. -.1:D5.I.|\..u...§utll:fl. In... 1.19. .1.» I. 1 -1uu2...u..r.....l!. . .. 5. l. . .11....» .1J2451MIJ.-. 2 giiihnvzxuttut...“ 1 . . fivfiv5§lh Ego. ii... . .1 I. lrn‘ll-lla“ -u b. a. . . . Nd}; .|.. II. A\ 13:. '13... v [3314... .flgldlru...:2. . . z I Q ‘ , u‘wlx. . I» A 5 . . I .0 .A .0 v. Ilvtrl u 3.59... |.~ ..............u.h..dl....uz. Lu... 3... : .V...1.|.v.\|| .v‘fNAJ .. 3... . 5... I... n.-. . .. 3:. . .In. {D I.» v . 5..- . I... N...u..d..........-..... \o" I .l‘ - Q . . v ‘ In. V . -l . 4|"- . . - .cno ahfllnl‘ofinlnifl .u r «111.155... 0101.15.12. i . it! 5 I. I .5 {35.51“ I...‘ log... (I. ’0‘ 5.4L 1" 5.15; 2 .4: 51”.... 5.2....01 -l . o a 503%.. 1. -55. 1%.! 1143.03.13! - 3 7 - , . 1‘11"“ TH" .';;‘o I . + w . 5555555555555555 5555555555555555555555555 L..._.-.w-. .. .1 a 3 1293 10691 53?. 297-77": Eda-7"- a——-’—--——.¢,-__-a Uh¢.¢-§; 4 [33,: .1,— 4 £4-;‘g £4.11. This is to certify that the thesis entitled Development and Packaging of a Soynut Butter presented by David K. Elsinger has been accepted towards fulfillment of the requirements for M. S . degree in Food Science W Major professor Date JUIY 83 1986 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution RETURNING MATERIALS: )V1531_J Place in book drop to remove this checkout from w your record. FINES will ' be charged if book is returned after the date ~ stamped below. 'AAQ O [’5'99’3 ‘ x.) I- 4.36 ' DEVELOPMENT AND PACKAGING OF A SOYNUT BUTTER By David Karl Elsinger A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science and Human Nutrition 1986 J «1,7 ABSTRACT DEVELOPMENT AND PACKAGING OF A SOYNUT BUTTER By David Karl Elsinger Expanding upon the work of Pichel and Weiss (1967), the develop- ment of a commercially viable soynut butter was pursued. The objectives were to determine the required processing parameters, optimum packaging and antioxidant combinations needed during distribution. Sensory evaluation studies indicated that consumers preferred smooth soynut butter containing 3.0% fructose. Stabilization of processing temperatures was achieved through the addition of 25% vegetable oil per batch, a 5-minute pre-cutting step, and a consistent rate of delivery of the pre-cut mix into the final grinding process. Product stability was assessed by conjugated diene absorbance, peroxide and thiobarbituric acid (TBA) values. Results showed that samples containing 0.02% TBHQ/CA had slower rates of autoxidation than control samples within the first two months of storage. High density polyethylene tubs generally provided protection equal to that afforded by glass upon the addition of antioxidant to the product. ACKNOWLEDGMENTS The author would like to express his sincere appreciation to Dr. Charles M. Stine, his major advisor for his friendship, guidance, support of the author's assistantship, and advice concerning the research as well as the many hours spent reviewing this manuscript. Appreciation is extended to the following people: Mr. Leonard Stuttman of INARI, Ltd. for the contribution of his time, experience and the use of the equipment in the INARI, Ltd. pilot plant; to Dr. Bruce Harte for his encouragement, guidance, critical review of this manuscript, and help with the author's assistantship; to Dr. Robert Brunner for his participation on the author's committee and critical review of this manuscript; and to Dr. Mary Ann Filadelfi for her great assistance in the design and interpretation of the sensory evaluation portion of my research. The author also wishes to express sincere gratitude to: Dr. Leroy Dugan, Jr., Dr. Susan Cuppett, Mr. Michael Stachiw, Mr. George Chen, and Ms. Mary Schneider, who helped make this thesis possible. The author is especially grateful for the love and support of his parents and grandparents during his graduate program. But most of all my life and thanks belong to God through Jesus Christ, who has given me His love, patience, peace and the strength to stand in any situation. Romans 8:32 (NIV) "He who did not spare His own son, but gave Him up for us all, how will He not also, along with Him, graciously give us all things?" ii TABLE OF CONTENTS Page LIST OF TABLES ......................... V LIST OF FIGURES ........................ VI INTRODUCTION .......................... 1 LITERATURE REVIEW ....................... 3 Soybeans as Food ...................... 3 Nutritive Value of Soybeans ................ 4 Peanut Butter Technology .................. 10 Soynut Butter Technology .................. I4 Mechanism of Oxidative Rancidity .............. 16 Measurement of Lipid Oxidation ............... 17 Peroxide Value ....................... 18 Anisidine Value ...................... I9 Thiobarbituric Acid (TBA) Test ............... l9 Distillation of Malonaldehyde ............... 21 Kreis Test (Rancidity Index) ................ 21 Conjugated Diene Absorption Method ............. 22 Antioxidants ........................ 23 Organoleptic Evaluation .................. 26 Packaging of Legume Spreads ................ 28 MATERIALS AND METHODS ..................... 3l Soybeans .......................... 31 Oils ............................ 31 Stabilizer ......................... 32 Antioxidants ........................ 32 Batch Production Procedures ................ 32 Packaging of Soynut Butter ................. 33 Analysis of Samples Prepared in Study #1 .......... 33 Lipid Extractions ..................... 33 Peroxide Value ....................... 35 Thiobarbituric Acid Values ................. 35 Analysis of Samples Prepared and Stored in Study #2 . . . . 35 Conjugated Diene Absorption ................ 35 Purification of TBA .................... 37 TBA Reagent Preparation .................. 37 TBA Distillation Method .................. 37 1'11 Page Sensory Analysis of Optimum Levels of Sweetener ....... 38 Methylation of Fatty Acids ................. 39 Gas Chromatography ..................... 42 Oxygen Transmission Studies ................. 43 Evacuation of Containers .................. 45 Light Transmission Studies ................. 45 Measurement of Water Vapor Transmission Rates ........ 46 RESULTS AND DISCUSSION ..................... 48 Processing ......................... 48 Results of Study #1 ..................... 58 The Effect of Antioxidant Type Upon Lipid Stability ..... 58 Effect of Light and Heat on Shelf Life ........... 65 Results of Study #2 ..................... 7T The Effect of Varying Packages on Lipid Autoxidation . . . . 71 Sensory Analysis of Optimum Levels of Sweetener ....... 85 Degree of Preference Test .................. 91 SUMMARY ............................. 94 RECOMMENDATIONS ......................... 96 BIBLIOGRAPHY .......................... 97 iv Table 10 ll 12 LIST OF TABLES Page Experimental Design of Chemical Analyses ......... 34 Soynut Butter Comparison to Peanut Butter, Based on 32 9 Samples .......................... 57 InfOrmal Laboratory Sensory Evaluation of Soynut BUtter Flavor after 13 Weeks Time ................ 57 Informal Laboratory Sensory Evaluation of Soynut Butter Flavor after 24 Weeks Time (for Selected Conditions) . . . 53 Light Transmission Characteristics of HDPE Tubs ...... 70 Fatty Acid Composition of Original Oils. . .. ....... 86 Fatty Acid Composition of Extracted Oil at 24 Weeks, 37°C, Dark ........................... 87 Fatty Acid Composition of Extracted Oil at 24 Weeks, 37°C, Dark ........................... 87 Fatty Acid Composition of Extracted Oil at 24 Weeks, 37°C, Dark ........................... 88 Degree of Sweetness (ANOVA) (n=45) ............ 89 Comparison of Sample Means (Tukey's test (Snedecor, l956)) .......................... 90 Degree of Preference (% of 45 Responses) ......... 92 LIST OF FIGURES Figure Page 1 Decomposition of linolenate dimer hydroperoxides (Frankel, E.N. Prog. Lipid Res. l9:l (l980), p. 16). . . . 9 2 Proposed TBA reaction (Sinnhuber, l958) .......... 20 3 Sensory evaluation instrument (multiple comparisons of relative soynut butter sweetness) ............. 40-41 4 Product flow for production of soynuts .......... 49 5 Product flow chart for production of soynut butter . . . . 51 6 Peroxide value vs time for soynut butter during storage in dark, 22°C (ambient) condition .............. 59 7 Peroxide value vs time for soynut butter during storage in fluorescent light, 22°C (ambient) condition ........ 60 8 Peroxide value vs time for soynut butter during storage in dark, 37°C condition ................... 51 9 TBA value, absorbance 532 nm, vs time for soynut butter during storage in fluorescent light, 22°C (ambient) condition ......................... 53 10 TBA value, absorbance 532 nm, vs time for soynut butter during storage in dark, 37°C condition .......... 54 ll Conjugated diene absorbance, absorbance 233 nm. vs time for soynut butter during storage in dark, 37°C condition . 72 l2 TBA, absorbance 532 nm. vs time for soynut butter during storage in fluorescent light, 22°C (ambient) condition in glass jars ........................ 73 13 TBA, absorbance 532 nm. vs time for soynut butter during storage in fluorescent light, 22°C (ambient) condition in HDPE tubs ....... . ................. 74 vi Figure Page 14 TBA, absorbance 532 nm. vs time for soynut butter during storage in fluorescent light, 22°C (ambient) condition in HOPE in (nylon/saran pouches) ............... 75 l5 TBA, absorbance 532 nm. vs time for soynut butter during storage in dark, 22°C (ambient) condition in HOPE tubs . . 77 l6 TBA absorbance 532 nm. vs time fOr soynut butter during storage in dark, 37°C condition in glass jars ....... 78 l7 Conjugated diene absorbance 233 nm. vs time for soynut butter during storage in dark, 37°C condition in HDPE 79 tubs ........................... 18 Conjugated diene absorbance 233 nm. vs time for soynut butter during storage in dark, 37°C condition in HDPE in (nylon/saran pouches) ....... . ........... 80 19 Conjugated diene absorbance 233 nm. vs time fOr soynut butter during storage in dark, 37°C condition in glass Bl jars ......................... . . 20 Net weight gain of water (mg) vs time in hours for HOPE tubs .............. . ............ 33 vii INTRODUCTION Like the peanut the soybean is a legume, however the soy bean is richer in various amino acid residues. It was thought that the food- buying public would be receptive to a nutritious soynut spread, and that such a development might have a significant impact on the utiliza- tion of Michigan-grown soybeans. The production of a nut-butter from soybeans was described by Pichel and Weiss (l967), but the questions of optimum antioxidant and package combinations were not studied. Optimum processing parameters for small-scale batch production, and optimum sweetener levels also remained unanswered. The present study was designed to gain knowledge and to propose future courses of action for the processing and packaging of a commercially acceptable soynut butter. Peroxide value, TBA value, conjugated diene and TBA (distillation) absorbances were measured to follow autoxidative changes in the sample Spreads. Four antioxidant blends were evaluated: BHA/BHT, BHA/PG, TBHQ, and TBHQ/CA. Three package systems - glass jars, high density polyethylene (HOPE) tubs, and tubs in evacuated nylon/SaranR pouches - were used to study effects of differing barrier on product stability. Sensory evaluation of added sweeteners according to type and concentration was accomplished using two tests. The first evaluation utilized a degree of preference test to measure the ability of respondents to differentiate statistically significant differences in concentrations of various sweeteners. The second evaluation was a degree of preference test which is a hedonic measurement. LITERATURE REVIEW Soybeans as Food The soybean is one of the earliest crops cultivated by man, some 4,800 years ago in China. The legendary emperor of China, Shang Nung, taught his subjects how to use a plow and sow grain for harvesting (Shih, 19l8). The first historical reference of soybean cultivation is 2,207 B.C., as there was no accurate dating system prior to this time (Smith, Circle, 1978). Soybeans were prescribed in a materia medica around 450 A.D., as a drug (Smith, Circle, l978) and are Closely associated with Buddhism. They constitute a major protein source in the diet of Chinese today. The soybean of commerce is the variety Glycine Max (L.) Merr., although there are approximately 100 named varieties which have been registered with the Crop Science Society of America. There are three major U.S. markets for soybeans: raw beans, soybean oil, and soybean meal. In l979 there were 61 million metric tons of soybeans produced by more than 650,000 farmers in 27 states (Ray, l98l). America has become the major exporter of beans to world markets, and is responsible for 80% of the beans involved in international commerce (Smith, Circle, l978). Most of the soybean crOp in the U.S. is solvent extracted since soybean oil is in great demand for use in margarine, shortenings and as a salad oil. The residual cake is processed into high protein animal feed. In Asia soybeans are an important ingredient in many foods for human consumption and are often utilized fOllowing fermentation. The young green pods are sometimes eaten as a vegetable (Elsevier, 1981). Defatted soy-flour is the starting material in the production of soy protein isolate, the collection of soy proteins precipitated at low pH. Soy-protein isolate may subsequently be textured and processed to simulate meat and poultry foods in chewiness and flavor (Smith, Circle, 1978). The protein is "spun" and the emerging fibers are coagulated in an acid bath, and stretched to a desirable size and strength (Campbell, 1981). Upon the addition of various binders, fats, flavorings and colors, the product is heat set to fOrm the finished meat analog (Campbell, 1981). The concept of a total soy-protein replacement for meat has failed because the analogs were as expensive as the meat they replaced, and lacked the flavor of real meat. Nutritive Value of Soybeans The soybean seed consists of proteins, lipids, carbohydrates, minerals and crude fiber (Krober and Carter, 1962). Sixty percent of the seed consists of protein and lipid. The majority of the protein resides in the aleurone bodies, which are subcellular structures between 2-20 microns, and which account for 60-70% of the protein in the seed (Orthoefer, 1978). The lipid components are found in subcellular structures called spherosomes, and are 0.2-0.5 microns in diameter. The soybean has the fOllowing approximate whole bean composition: (Krober and Carter, 1962). Protein 40% Lipid 20% Cellulose and hemicellulose 17% Sugars 7% Crude fiber 5% Ash (dry weight basis) 6% The proximate composition of soybean components as fOund in seed parts (Wolf and Cowan, 1971) are: Seed Percentage Percent composition of soybean components part of whole bean Protein Fat Carbohydrate "ash" Cotyledon 90 43 23 29 4.9 Hull 8 9 1 86 4.3 Hypocotyl 2 41 11 43 4.4 Defatted soya flour is very high in most of the essential amino acids. Soy protein isolate is rich in isoleucine, leucine, lysine, and valine as compared to peanut isolate. 9/100 9 protein (W.H.0. Tech. Rep. Ser., No. 37 (1965)) AMINO ACIDS: - Soy Isolate Peanut Isolate Tryptophan 1.2 1.0 Threonine 3.8 2.5 Isoleucine 4.7 4.3 Leucine 7.5 6.7 Lysine 6.2 3.0 Methionine 1.4 1.0 Cystine* 1.0 1.2 Phenylalanine 4.8 5.6 Tyrosine* 3.6 Not reported Valine 5.0 4.5 *Not essential but often limiting or lacking. Protein Efficiency Ratio (PER) is one method of determining protein quality. PER = weight gain E weight protein consumed and is the only method currently approved by the Food and Drug Administration for nutritional labeling of protein quality of foods (Labuza, 1977). Two procedures commonly used to measure the degree of protein denaturation in soya are: the nitrogen solubility index (N51), and the protein dispersi- bility index (POI). The N51 value measures the percentage of the total nitrogen in the sample that is soluble; the P01 value, the percentage of total protein soluble in a sample. In the evaluation of a food protein, anti-nutritional factors such as trypsin inhibitors, hemagglutenins and phytic acid must also be considered (Nelson, Steinberg, Wei, 1978). The primary function of a protein is to furnish the essential amino acids and nitrogen required for the normal function and growth of an organism (Hopkins, Steinke, 1981). There are many ways to measure this "protein quality" in humans such as weight changes, nitrogen balance, serum protein concentrations in the blood and growth rates. Nutritional studies have shown that there is a favorable nutritional impact of soybean protein as a component of a mixed diet or as a protein supplement (Ibid). When combined with cereal grains in the diet, a more favorable amino acid complement is achieved. The overall protein intakes of lower socioeconomic groups in develop- ing countries are often deficient for proper nutrition. Calorie intake may also be deficient among such groups. In developing countries that have cereal-based diets, there has been a marked decrease in the availa- bility and consumption of legumes. Soybeans have contributed to food systems as sources of calories, as supplementary protein and aS complementary protein because of their relatively good essential amino acid pattern (Bressani, 1981). The future of soy protein supplements in the diets of developing populations is still in doubt. The major obstacles are not only political, but alSo cultural, economic and technical. Many countries import U.S. soybeans for animal feeding but do not yet have the available technology to make better economic use of the protein for human consump- tion. Soy products are just beginning to become popular in the U.S. and may become more accepted worldwide if people will accept the flavor of soy products and alter their social and cultural habits of eating. The lipid components of soy include triglycerides, phospholipids, lipid-soluble pigments, tocopherols and sterols. Soybean oil contains a Significant level of linolenic acid (Pryde, 1980). This polyunsaturated acid is readily oxidized by enzymatic, thermal, photo- and metal catalysed autoxidation with the production of highly undesirable reversion or rancid flavor and odor. The development of flavor reversion is characteristic of soybean oil and other linolenate-containing oils and occurs at low levels of oxidation. The flavor of reverted soybean oil is described as beany and grassy at the early stages and as fishy or painty at the more advanced stages (Frankel, 1980). Chang gt 11. (1966) attributed the flavor of reverted soybean oil to a-pentyl furan, which they assumed to be derived from linoleate oxidation. Another linolenic acid derivative implicated as a cause of flavor reversion are oxidative polymers. Oxidative polymers are a complex mixture of oxygen-containing compounds formed by polymeric decomposition of linolenate hydroperoxides. These high molecular weight compounds can easily decompose and generate volatile aldehydes and other compounds contributing to flavor deterioration occurring even in the absence of oxygen and low temperatures (Frankel, 1980). Dimeric compounds with carbon-carbon linkages between fatty acid molecules were produced by thermal decomposition of linolenate hydroperoxides (Frankel §t_al,, 1960). A mixture of allylic 14-, 15—, 16- and l7-hydroperoxides would be produced from the free radical oxidation in a fatty acid with a single double bond at carbon 15. Decomposition of these hydroperoxides would produce a mixture of aldehydes and alcohols (Frankel, 1980) (Figure 1). These compounds have been cited among the more than 70 compounds identified in oxidized soybean oil (Selke gt_gl,, 1970). Crude soya oil contains substantial amounts of natural antioxidants such as tocopherols and phospholipids (Buck, 1981). The principal phospholipids are phosphatidylcholine, phosphatidylethanolamine, phosphatidylinositol (Orthoefer, l978). Soybeans contain approximately one-third carbohydrate which consists primarily of the sugars sucrose, raffinose, and stachyose (Ibid). The insoluble polysaccharides consist of galactomannans, acidic poly- saccharides, xylan hemicellulose, and cellulose (Ibid). The seed coat is the major site of insoluble saccharides and lignins. Also present in soy are various enzymes, of which lipoxygenases are of importance from the aspect of flavor deterioration. Lipoxygenases catalyze the oxidation of unsaturated fatty acids containing 1,4-pentadiene systems in the ciS configuration (Ibid). Other lipid related enzymes present are lipoperoxidase and lipases. Proteases are also present, and they cleave interior peptide bonds of proteins (Ibid). Alpha and beta amylases are present, but have an unknown effect as mature soybeans reportedly do not contain starch. .‘ll 1. {‘15 1' Cll3-3'-CIl—fi-CII=CII— 0fl3—CII=cll-’r-CII-;-01lz— :' on“. i on). i z i 1 Nathan Acetaldohyde Propioual Z-lltenal (methanol) ,‘162 5'14 3 CN3—0flleill-‘5'CN=0N- Cll3-CII2—CII=Clljf-illll-fcllz i coax f coal ; i i 1, [than Propional Intalal Z-Pontenal (ethanol) Figure l. Decomposition of linolenate dimer hydroperoxides (Frankel, E.N. Prog. Lipid Res. 19:1 (1980) p. 16) lO Peanut Butter Technology Before discussing the manufacture of soynut spread, it is helpful to look at the processing of peanuts into peanut butter. Peanuts are the fruit or pod of Arachis hypogaea of the Leguminosae family. The annual per capita consumption of peanuts in the United States exceeds eight pounds in terms of a farmers' stock sold in the shell (Woodroof, 1981). Consumption of peanut butter per capita has increased from about 2% lb in 1950, to 3% lb in 1970, and about 4 lb in 1980 (Ibid). The U.S. Department of Agriculture has set standards for grades of peanut butter in Section 52.306l-.3073. Legal peanut butter is a "cohesive, comminuted food product prepared from clean, sound, shelled peanuts by grinding properly roasted mature peanut kernels from which the seed coats have been removed and to which salt is added as a seasoning agent", U.S. standard for grades (1962). There are three defined textures of peanut butter: smooth, regular, and chunky. "Smooth" peanut butter means that there is a very fine and even texture with no perceptible grainy peanut particles. "Regular" peanut butter has a definite grainy texture with perceptible peanut particles not more than 1/16 in. in any diameter. "Chunky" peanut butter has partially fine and partially grainy particles with substantial amounts larger than 1/16 in. in diameter. Three grades of peanut butter exist: U.S. Grade A, U.S. standard, and substandard, which fails to meet the requirements of U.S. standard. Peanut butter manufacture consists of shelling, dry roasting, and blanching the peanuts, followed by fine grinding (Woodroof, 1981). Peanut butter may be produced from any variety of peanuts, but for ll Optimum consistency a blend of two parts Spanish or Runner peanuts with one part Virginia peanuts is suggested (Ibid). By federal regulations, 90% of peanut butter must be peanuts. Artificial flavors, artificial sweeteners, chemical preservatives, natural or artificial color, purified vitamins, or minerals are prohibited. The fat content may not exceed 55% including the natural oil in the peanuts. Hydrogenated vegetable oils are added in small amounts aS emulsifiers to prevent oil separation, and to improve spreadability. Dextrose, sugar, or honey are the commonly used sweeteners (Ibid). Peanuts may be dry roasted by one of two methods: batch or continuous. Batch-roasted peanuts are roasted in 400-1b lots in a revolving oven heated to a temperature of 800°F. The peanuts are heated to 320°F (Beattie, 1936) and held for one hour to achieve a satisfactory roast. All the nuts in the batch must be uniformly roasted so that there is complete development of color from the center to the surface of each kernel, without scorching, excessive oiliness, or decomposition of surface fats (Woodroof, 1981). Continuous roasting involves a conveyor- fed batch of peanuts with a counter-current flow of heated air. The peanuts are constantly agitated to improve the transfer of heat and extraction of moisture and volatiles (Ibid). The moisture content is reduced from an average of 5.0% to 0.5%. The color change is due to absorption of lipid by the cell walls. The roasted peanuts leave the roaster and are discharged into a perforated metal cylinder, where cooled air is forced through the mass by suction fans. The next step in the peanut butter making process is dry blanching, which removes the hearts and Skins. The peanuts are passed through 12 the blancher in a continuous flow and brushes or ribbed rubber belting remove the skins. The Skins are blown into porous bags and the hearts are separated from the cotyledons by screening (Woodroof, 1981). The blanched nuts are inspected and screened to remove scorched and rotten nuts. Discolored peanuts are removed by electric eye, and metal parts by magnets (Ibid). Peanut butter is usually made by two grinding operations. The first grinding reduces the nuts to a medium ground texture. The final grind gives a smooth and fine texture to the peanut butter. The nuts are forced between the grinding surfaces by an impeller mounted on a rotor shaft. Openings between the rotor and stator are from 3-5 mils for regular peanut butter (Weiss, 1983). The peanuts must be kept under a constant pressure from the start to the finish of the grinding process to assure uniform grinding and protect the product from air bubbles. Swept-surface heat exchangers such as the VotatorTM, are used to cool peanut butter from 1700 to 120°F, before it is packaged. Oil separation is determined primarily by the nature and amount of crystals present in the peanut butter. One procedure for preventing separation involves shock-chilling the hot peanut butter to produce finely divided crystals, followed by a Slow tempering process (Woodroof, 1981). Heat processing of peanuts will generally improve the flavor, aroma and texture, but will reduce the shelf-life of the oil by destroying natural antioxidants such as tOCOpherols. At the time of manufacture the oil in peanut butters remains relatively stable to oxidative rancidity (Willich gt_gl,, 1954), and seems to remain stable after storage in the dark at 80°F for two years. It has been observed that l3 autoxidation initially proceeds rapidly for about 3 months, or until available internal and headspace oxygen is depleted, whereupon stability remains relatively constant for more than two years (Freeman gt 31., 1954). Methods for preventing autoxidation in peanut butter include evacuating the headspace above the product or by addition of an approved antioxidant or antioxidant blend to function as a free radical terminator. Light, especially shorter wavelengths, also catalyzes lipid oxidation in peanut butter. Experimental data (Woodroof, 1981) indicated a measurable difference in flavor of peanut butter stored in the dark, or in amber, green, or blue colored jars. The differences noted are reduced oil separation, better aroma and flavor, but no major differences in peroxide or other chemical determinations of lipid oxidation. In 174 samples, free fatty acid values (FFA) averaged 0.373 for samples stored in the light and 0.345 for those in dark; the peroxide values averaged 4.06 for samples in the light and 4.05 for those in the dark (Ibid). Most peanut butter is packed in glass jars for retail distribution. Opaque polyethylene and transparent XT polymer jars and polystyrene cups with polyethylene lids have also had limited use (Anon., 1950; Anon, 1965; Russo, 1968). The shelf-life of peanut butter in plastic containers is about 9 months to one year. In glass jars the normal shelf-life is about two years (Weiss, 1983). The decreased stability is due to permeation of oxygen through the plastic jar wall which results in increasing off-flavors. Another defect in peanut butter known as "pull-away" occurs when the butter loses its ability to adhere to the wall of a glass jar. Moisture in the processing area will cause jars to become damp and accentuate the pull-away. Jars for peanut l4 butter are designed without a shoulder, which would act as a starting point for pull-away. The bead at the rim of the jar must be above the peanut butter surface for the same reason as stated above (Ibid). Soynut Butter Technology The roasting of soybeans to make a "soynut" product began in the 1940's. Soynuts became a specialty item on U.S. candy counters (Heller and McCarthy, 1944). Soynuts are prepared at home by oven-roasting soy- beans in soybean oil, and with the addition of a little salt they have been claimed to compare favorably in taste and texture with more expen- sive nuts. Compared to the peanut, the soynut has substantially less fat, more protein and are less expensive than peanuts. Soybeans generally have a characteristic undesirable "beany" flavor due to early low level development of unusual compounds from autoxidation (reversion) unlike peanuts which have a desirable "nutty" flavor and aroma. Pichel and Weiss (1967) attempted to overcome these problems in the process for preparing a nut butter from soybeans. Roasting or frying alone naturally does not modify this off-flavor since the reactions will have occurred prior to heating. Pichel discovered that by treating the dehulled soybeans with hot water, either by soaking or by steaming, practically eliminated the "beany“ or "grassy" flavor without rendering the product tastless. This may be due in part to thermal inactivation of lipoxy- genases. After moisturizing the beans, they were fried in oil and reduced to a fine paste. Pichel also found that he could achieve a greater throughput rate and smoother butter by comminuting whole fried beans with sufficient added oil in an Urschel Micro-cut mill with a shaving head (Weiss, 1983). The palatability of the soybean spread may 15 be enhanced by incorporating sugar, salt, hardfat stabilizer and oil. The enhanced spread is smooth, homogenous, with no noticeable "beany" or "grassy" flavor. The final soynut butter has a consistency similar to that of peanut butter (Pichel, 1963). Commercial soynut butter spreads to date have not been widely accepted due to continuing tactual and flavor defects. Consumer demand for a soynut butter spread appears to be greatest in the health-food and Specialty market. The use of soynut butter as an ingredient for confectionary items has begun recently. The idea for the research reported in this thesis originated from early interest of a Michigan- based soybean processor to create various products based on the attrition or grinding of Michigan produced soybeans and flavored with various additives. The reasoning behind such research is that the food-buying public would be receptive to a nut butter spread with a more complete amino acid array and that soybeans are lower in cost than peanuts. If a commercially acceptable, flavored, soy spread could be developed, it could have a significant impact on the utilization of Michigan-grown soybeans. 16 Mechanism of Oxidative Rancidity The oxidation of lipids, both mechanisms and oxidation products, has been well reported in numerous scientific papers. For this research a brief review of mechanisms of autoxidation, oxidation products, and antioxidants is presented. Rancid off-flavors are produced through autocatalytic processes with oxygen in a refined fat or oil. This type of rancidity is known as oxidative rancidity. Off-flavors produced through reactions catalyzed by lipases from food or from microorganisms is known as hydrolytic rancidity. Oleate, linoleate, linolenate and other polyunsaturated fatty acids are the lipid moieties responsible for the formation of products of lipid oxidation (Labuza, 1971). Lipid autoxidation proceeds through a free-radical chain mechanism involving initiation, propagation, and termination steps. These steps can be schematically represented by: Initiation initiators RH+02 ———a-R + OOH (fat molecule) (free radical) Propagation R. + 02 —>ROO° (peroxy free radical) ROO' + RH —>RO0H + R' (hydroperoxide) Termination R' + R°—> RR R. + ROO°——.ROOR R00. + ROO°_..ROOR + 02 17 In this presentation RH refers to any unsaturated fatty acid in which the H is labile by reason of being bound to a carbon atom alpha to a double bond. Various agents such as radiation and heavy metals like copper ion are the principal initiators of autoxidation. The primary products of lipid oxidation are hydroperoxides (ROOH) which are generally known Simply as peroxides. The oxidation process becomes extremely complex as the peroxides undergo scission, dismutation and other inter- actions through thermal instability or reactions with other materials to form more free radicals (Dugan, 1961). Secondary reaction products involving hydroperoxides include alkoxy free radicals which dismutate to give numerous aldehydes, ketones, and alcohols varying in carbon chain length. It is these Secondary reaction products which are primarily responsible for the off-flavors in oxidized polyunsaturated lipid systems. Different hydroperoxides are formed when light and photo- sensitized oxygen molecules are present. The production of singlet oxygen is believed to be the mechanism for the production of allylic hydroperoxides in which the double bond has been shifted (Hamilton, 1983). The off-flavors produced from photooxidation are sometimes characterized as having a "grassy" aroma and flavor. Measurement of Lipid Oxidation The acceptability of a food product depends on the extent to which-3 the oxidative deterioration has occurred. There are five known stages of autoxidation of a fat; the induction period, peroxide formation, peroxide decomposition, polymerization, and degradation stage. In the induction period hydroperoxides increase very slowly in amount and are not measurable. At the end of the induction period, there is a sudden 18 increase in peroxide content. Hydroperoxide concentration will reach a maximum, then decrease as a result of peroxide decomposition. As oxygen is absorbed total carbonyl content increases as well the viscosity of oils. Monitoring the various precursors and products of autoxidation at each of these stages is accomplished with different chemical tests. Susceptibility tests measure the stability of a lipid under conditions favor oxidative rancidity and include tests such as the Schaal oven test and active oxygen methods. Some of the tests which measure the extent of oxidation in a lipid system include: peroxide value, anisidine value, thiobarbituric acid test, Kreis test, conjugated diene absorption method and sensory evaluation. Peroxide Value Measurement of peroxide value is useful up to the stage at which extensive decomposition of hydroperoxides begins. Many analytical procedures for the measurement of peroxide value have been developed. In all cases the accuracy of the test depends on the specific experi- mental conditions, as the method is highly empirical (Rossell, 1983). The most common procedures are iodometric methods developed by Lea (1931) and Wheeler (1932). The method is based on the measurement of the iodine liberated by oxidation of potassium iodide by the peroxides present in the oil. The iodine is liberated in a stoichiometric ratio of two atoms of iodine for each atom of active oxygen in the system. The iodine is titrated with sodium thiosulfate in the presence of soluble starch indicator. The peroxide value (P.V.) is expressed in milliequi- valents of "peroxide oxygen" per kilogram of fat. The assumption is that a high P.V. indicates that oxidation has begun, and the fat is in or 19 past the induction period and oxidation will accelerate rapidly (Gunstone, Norris, 1983). Mehlenbacher (1961) has suggested the two principal sources of error in these methods are (a) the absorption of iodine at unsaturated bonds of the fatty acid, and (b) the liberation of iodine from potassium iodide by oxygen present in the solution to be titrated. Lea (1931) attempted to eliminate this latter error by filling the sample tube with nitrogen at the beginning of the test and making the assumption that the vaporization of chloroform would prevent the re-entry of oxygen into the tube. Other possible sources of error in iodometric methodolo- gies include variation in weight of sample, the type and grade of solvent used, and variation in reaction conditions such as time, temperature, and reactivity of the peroxides being titrated (Gray, 1978). Anisidine Value This method is used for the measurement of high molecular-weight carbonyl compounds, which indicate the prior oxidative deterioration of a fat. The IUPAC standard method 11.0.26 defines anisidine value (A.V.) as 100 times the absorbance of a solution resulting from the reaction of l g of fat or oil in 100 ml of a mixture solvent and p-anisidine, measured at 350 nm in a 100 mm cell under the conditions of the test. A fat that had undergone extensive oxidation could Show a low P.V. after deodorization, but the A.V. would presumably be little changed, and would warn of this past history. The A.V. is seldom used in the U.S. because most oils are domestically produced and fresh (Gunstone, Norris, 1983). Thiobarbituric Acid (TBA) Test The Thiobarbituric Acid (TBA) test is a common method for the detection of lipid oxidation. Early studies by Sinnhuber et_al, 20 helped to clarify the nature of the colorimetric reaction that occurs during the TBA test. Two moles of 2-thiobarbituric acid condense with one mole of malonaldehyde to yield a red pigment. The intensity of the red (pink) color produced is prOportional to the amount of malonaldehyde in the oxidized oil and within a certain range can be linearly related to the extent of oxidation of the fat (Figure 2). The main absorption maximum is at 532N535 nm. Proposed TBA reaction (Sinnhuber, 1958) "fad T" \ —c=c\ 011 II “a\ 011 zuzo TBA Malonaldehyde TBA chromagen (pink-red) Figure 2. Proposed TBA reaction (Sinnhuber, 1958). Malonaldehyde is a dicarbonyl compound formed in oxidized polyunsaturated lipid systems. A disadvantage of the TBA test is that malonaldehyde can combine with amino acids such as lysine to fOrm a Schiff's base, an intermediate step in non-enzymatic browning. The obvious limitation is that not all the malonaldehyde produced will be available to react with the thiobarbituric acid to produce the pink chromagen. Another limitation is that oxidative rancidity in lipids containing little or no fatty acids of the linolenate or higher unsaturation would not be expected to show Significant TBA values even though these lipids gave a high peroxide value (Sinnhuber, Yu, 1977). Pohle et_al, (1964) found that flavor score 21 could not be estimated for any given fat from the TBA value since the relative level varied from product to product. The change in flavors in an oil and their relationship to TBA value would have to be established before the TBA value could be used as an index of off-flavor development (Gray, 1978). Distillation of Malonaldehyde In 1955 Sidwell et 31. described a steam distillation procedure for dried milk in which the malonaldehyde was distilled from the acidified milk. An aliquot of the distillate was then reacted with TBA, and the color was read directly. Tarladgis et 31, (1960) described a simplifi- cation of Sidwell's technique by directly heating samples (meat slurries) on a Kjeldahl distillation rack. This procedure allows the simultaneous distillation of multiple samples with equipment generally available in most food laboratories. The distillation procedure offers several important advantages over other methods. The malonaldehyde is obtained in a clear aqueous solution so that its reaction product with thiobarbi- turic acid does not require a long solvent extraction procedure. The distillation heat treatment uses acid to effect the liberation and distillation of malonaldehyde from the sample, which means that there is less likelihood of fat oxidation occurring during the test itself (Ibid). Kreis Test(Rancidity Index) The Kreis Test was one of the first tests used to evaluate the oxidation of fats, and is based on the production of a red color when phloroglucinol reacts with epoxyaldehydes and their acetals in acid solutions (Rossell, 1983). The standardized method involves reacting the sample with phloroglucinol in diethyl ether solution. The products 22 are next extracted with HCl, and a red aqueous solution is obtained if the sample material is rancid. The red color is quantitated with a Lovibond colorimeter in a.l in. glass cell. Color readings up to 3 red units indicates incipient rancidity. A color reading between 3 and 8 units indicates that the rancidity is occurring towards the end of the induction period. Color readings of over 8 units indicates definite rancidity. One limitation of the Kreis test iS that some food additives, such as vanillin, can interfere with the test (Ibid). Conjugated Diene Absorption Method Oxidation of polyunsaturated fatty acids is accompanied by an increase in ultraviolet absorption due to the formation of conjugated diene and triene hydroperoxides. Conjugated unsaturations of fatty acids absorb strongly in the region 230 to 375 nm. The magnitude of change in absorbance is not directly related to the degree of oxidation because the effects upon various unsaturated fatty acids can differ in magnitude and quality (Holman and Burr, 1946). Oils which contain linoleate or higher unsaturated fatty acids are oxidized to produce conjugated diene systems that can be quantitated by ultraviolet absorption at 233 nm. Absorption will increase proportion- ately to the uptake of oxygen and to the formation of peroxides in the early stages of oxidation (Farmer, Sutton, 1943). St. Angelo et_al, (1975) studied the autoxidation of peanut butter by measuring the peroxide value as subsequent increase in absorption at 234 nm due to diene conjugation. Samples were analyzed by modifications of both the peroxide value and conjugated diene method A.O.C.S. official method Ti 1a-64. St. Angelo concluded that the modified conjugated diene 23 hydroperoxide (CDHP) method can be used as an index of progressive staling in place of or in addition to, the peroxide value. The (CDHP) method requires smaller samples and is more rapid, and simpler than the peroxide value method. The (CDHP) method does not require additional reagents and does not depend upon chemical reaction (Ibid). This method is applicable for the analysis of conjugated diene production in soynut butter which contains a high amount of linoleic and lower, but Signifi- cant, levels of linolenic acid. Antioxidants An antioxidant is a substance that is added to fats or fat-contain- ing foods to retard oxidation and thereby prolong their wholesomeness and palatability. Ideally an antioxidant should: (1) have no harmful physiological effect; (2) not contribute any objectionable odor or taste to the fat or food in which it is used; (3) be fat soluble; (4) be effective at low concentrations; (5) be readily available; (6) be economical; (7) be legal; (8) persist following processing to provide effective protection to food in which it exists, (i.e. "carry-through" properties) (Dugan, 1976). Primary antioxidants function by inhibiting or interrupting the free radical chain mechanism. Their ability to interrupt the free radical is usually based on the phenolic configuration within their molecular structure (Sherwin, 1976). Antioxidants such as ascorbic acid function by being preferentially oxidized and they afford relatively poor protection. 24 An antioxidant AH apparently reacts with radicals produced during autoxidation according to the scheme: RH—->R° + H. R' + AH—>RH + A. RO° + AH—>ROH + A. R00. + AH—>ROOH + A. (termination reactions) R' + A—>RA R0. + A'——>ROA A' + Al—>AA The preceding reaction diagram shows how antioxidants interfere with the free-radical mechanism. Boozer gt_al, (1955) proposed a different mechanism, involving complex formation, as follows: (ROZAH2)°+R001--stab1e product Peroxide decomposers act as catalysts to decompose peroxides initially present as well as those that are formed during further Oxidation. An important feature of this scheme is that the primary stable products are not free radicals (Tranggono, 1978). This naturally rules out the decomposition of peroxides by metals such as copper, cobalt and iron (Dugan, 1963). Some antioxidants provide increased protection as the concentration increases whereas others have optimal levels and higher levels are sometimes prooxidant. The correct balance must be achieved to provide maximum stabilization without intensifying oxidation. Among the most widely used commercial antioxidants are: (BHA) butylated hydroxyanisole (a mixture of isomers of 2-t-butyl-4- methoxyphenol and 3-6-butyl-4-methoxyphenol); (BHT) butylated 25 hydroxytoluene (2,6-di-t-butyl-4-methylphenol); esters of gallic acid; and (TBHQ) di-tert-butyl-hydroquinone. In the U.S. the Federal Food and Drug Administration requires that the phenolic antioxidant content may not exceed 0.02% of the fat or Oil content of a food including the essential (volatile) oils. When various mixtures of antioxidants are used, the concentration of any single primary antioxidant may not exceed 0.01% of the fat or oil content of the food. When antioxidants or synergists are added, the combined total may not exceed 0.025%, with no Single antioxidant exceeding 0.01% of the fat or oil content of the food. The amount of antioxidants permitted in a food lipids is never greater than one hundredth of the L050 established for that antioxidant. In some cases it is found that a specific combination of two or more antioxidants is more effective in inhibiting oxidation than the equivalent quantity of a single antioxidant. This phenomenon is known as positive synergism. Not all combinations of antioxidants display this synergistic effect. Although BHA and BHT are synergistic and BHA and propyl gallate (PG) are synergistic, the combination of BHA with PG results in a decreased stability of a fat than expected from the sum of the effectiveness of each antioxidant if used alone. This effect is known as a negative synergism. TBHQ (di-tert-butyl-hydroquinone) was first permitted as an anti- oxidant for food use in 1972 in the U.S. (Coppen, 1983). TBHQ has been shown to be very effective as an antioxidant for vegetable oils and is more effective than PG. Unlike some of the gallic acid esters, TBHQ is quite soluble in oil and will not discolor in the presence of iron and water. Crude soybean oil without added antioxidant possesses 26 a high degree of oxidative stability and treatment with TBHQ increases the stability of the crude oil considerably (Sherwin, Luckadoo, 1969). The high degree of stability in the crude soybean oil may be due to naturally occurring antioxidants in the oil such as tocopherols. The activity of the various tocopherols has been compared and found to vary inversely with the order of vitamin E activity. Studies by Dugan and Kraybill (1956) observed that cooking may destroy or modify tocopherols, and that ganma tocopherol, which is a better antioxidant than alpha tocopherol, is also a better carry-through antioxidant. Tocopherols found in soybean oil are mostly gamma tocopherols. Lecithin or mixtures of phosphatides have some antioxidant activity, as do several flavones, sterols and sulfhydryl compounds. Warner and Frankel (1985) found that the most effective combination of antioxidants in soybean oil was achieved with TBHQ and citric acid (CA). The combination of TBHO and CA increased the induction period from 1 day in the soybean oil without antioxidants, to 9 days, as measured by direct gas chromatography of head-space volatiles. TBHQ/CA is an important antioxidant combination used in soya oil, and should be just as important in products containing soybeans or soybean oil as ingredients. Organoleptic Evaluation Sensory evaluation plays a critical role in the development of foods and beverages for human consumption. Consumers will often reject fOods from which they could derive nutritional benefits because the fOod has a poor taste or because the particular consumers have not had previous experience with the food item (Moskowitz, 1984). Most aspects of quality are measurably only through trained sensory panels that evaluate foods 27 by the senses of taste, smell, touch, and hearing when a food is eaten (Larmond, 1977). Sensory evaluation panels can be grouped into three types: highly trained experts, routine laboratory panels. and large consumer panels. Preparation for sensory panel analysis involves setting up a special training area so that distractions can be minimized and conditions can be controlled (e.g. light, sound). Selection of the proper test, development of a statistically consistent language for odor, texture, and appearance, and proper sample preparation are all part of the difficulties involved in obtaining accurate sensory data. In the case of the trained expert panel, a difficult and long process occurs when educating and screening each panelist so they can properly identify a particular threshold of a certain stimulus consistently over an extended period of time. When people are used as a measuring instrument, it is necessary to rigidly control all testing methods and conditions to overcome errors caused by psychological factors. "Errors" may include all kinds of extraneous influences (Ibid). The area of sensory analysis concerned with the judgements peOple make such as "good-bad" or "like-dislike" is known as hedonic measure- ment. Hedonic characteristics determine our behavior when it comes to choosing foods for the first time and repeating an initial choice upon continued exposure (Moskowitz, 1984). Hedonics are concerned with: attitudes versus behavior; measuring likes by classifying responses; measurement of degree of liking; ratio scales; and time preference measures of liking. The degree of overall "liking" of a food product represents the key evaluative criterion against which the researcher judges all other variables (Ibid). An order of preference test as used 28 in this thesis is a hedonic measurement tool. In a multiple comparisons test, a known reference or standard sample is presented to the panelist and compared to the reference on the basis of some named characteristics. Multiple comparisons may be used to evaluate the effects of replacing or changing an ingredient, of packaging material, of changing a specific process, or of storage. Small differences between the reference sample and control can be detected. Information about the direction and magnitude of the change is also obtained (Larmond, 1977). Packaging of Legume Spreads The legume spread of major importance is peanut butter. Recently almond butter, cashew butter, and sesame butters have also received more attention in consumer markets. As previously discussed most peanut butter is packed in glass jars for retail distribution. In glass jars the shelf-life of peanut butter is about two years. In opaque poly- ethylene containers, the decreased shelf-life of 9 months may be extended by incorporating an antioxidant into the plastic. Since packaging a fat or fatty food allows fat from the food to be dispersed over the surface of the packaging material, there is an increased tendency for oxidative rancidity to occur. The use of BHA or BHT in packaging materials increases the shelf-life of lard or butter by a factor of 2 to 3 and has an improving effect on margarine stability (Eastman Chemical Products, Inc., 1953). Antioxidants such as BHA or BHT have sufficient volatility to migrate into the package atmosphere and contact the material in the package for Significant improvement in stability. Till et_al, (1982) studied the migration of BHT from high density polyethylene (HOPE) in a variety of foods and food simulants. 29 Migration of BHT was found to be more rapid into oils and fatty foods than into aqueous materials. Another important consideration in packaging of legume spreads iS the effect of light tansmission, especially U.V. light, upon the stability of the product. As previously mentioned, when peanut butter is stored in the dark or in colored jars, there is reduced oil separation and better aroma and flavor. Polyethylene containers for milk that have been pigmented with titanium dioxide results in substantial reduction in U.V. light transmission (Nelson, Cathcart, 1983). The barrier characteristics of a plastic container are critical when packaging a high-fat legume spread. Shelf-life can be defined as the length of time that a container or a material will remain in a saleable or acceptable condition under Specified conditions of storage. When discussing the influence of barrier on shelf-life the term "permeability" is important to understand. The permeability of a material is the flux or the rate at which a quantity of permeant gas or vapor pasSes through a unit surface area in unit time, and is dependent upon the partial pressure of the permeant, the film thickness, the surface area, and often the temperature at which permeation occurs. The mechanism of permeation is a complex topic and is based on mass- transfer theory. The permeation mechanism is based on: the size of the permeant molecule; the molecular structure of the permeant molecule; the thickness and density of the package material; the concentration gradient of the permeant molecule, and temperature of the environment. Water vapor transmission rate, as expressed in terms of weight gain, is defined as mg/package x 24 hours x mm Hg. The oxygen transmission rate 30 through a packaging material is expressed in cc/m2x24 hours. From each rate a permeability constant may be calculated for a certain barrier material under specific test conditions. The advantages of measuring the permeability are that the integrity of the closure, and the influen- ces of machine processing and distribution can be quantified. Once a standard container's permeability constant is calculated over a range of temperature and humidity values, the packaging engineer has a protective tool which can help eliminate ineffective container design and selection. MATERIALS AND METHODS Soybeans Soybeans procured for soynut production were dehulled, cleaned, and obtained from Diehl Fields, Inc. of Dansville, MI. The soybeans are packaged in 3-ply paper and polyethylene bags. The moisture content of the incoming soybeans was between 13 and 15%. le§_ The pre-boiled soybeans were roasted in Gordon Food Service (GFS) Red LabeiTM solid vegetable fat. The GFS fat consists of 100% hydro- genated soya oil, 0.25% lecithin, .002% B-carotene, 0.02% TBHQ, .0005% citric acid, and 0.001% antifoam (dimethyl polysiloxane). The TBHQ in the roasting vegetable oil means that there will be a small level of antioxidant in every sample batch. This implies that the "control" samples may have contained some antioxidant, and that some unplanned synergisms, negative or positive, may have occurred with other anti- oxidants. The oil added to the soynut butter to control body and spreadability was a commercial soybean oil (Bunge Edible Oil Co.) and did not contain antioxidants (AH). The antioxidant-free oil was specifically produced for this project for INARI, Inc. and shipped to their production site. Production of an antioxidant-free oil is difficult and expensive as most oil is produced with antioxidants in it, and to remove all residual antioxidant from the machinery is very time-consuming. Therefore traces of antioxidant might be present in the oil. 31 32 Stabilizer A peanut butter stabilizer (hardener) composed of semi-solid partially hydrogenated palm oil, was used to harden the soynut butter and prevent oil separation. Antioxidants Four antioxidants were used in this study. Commercial antioxi- dant or blends were obtained from UOP, Inc. and consisted of Sustane ZOTM, a commercial name for di-tert-butyl hydroquinone with a citric acid synergist; Sustane 3T", a commercial name for a mixture of mono-tertiary-butyl-4-hydroxyanisole (BHA) and n:propyl-3,4,5-tri- hydroxybenzoate (propyl gallate) and citric acid; Sustane 6T", a commercial name for a mixture of BHA and 2,6-di-tert-butyl-pgrgfcresol (BHT); Sustane ZOATM, a commercial name for a mixture of TBHQ, citric acid, and propylene glycol. Batch Production Procedures Pilot plant batches for chemical analysis for studies one and two consisted of 11 and 15 pounds of roasted soybeans, 4 and 6 pounds of antioxidant-free vegetable oil, 0.18 and 0.55 pounds of dextrose, 0.045 and 0.022 pounds of salt and 0.179 and 0.176 pounds of "hardener" (partially hydrogenated vegetable fat) respectively. A control was made for each by these formulations and each succeeding batch contained a pre- weighed amount of a particular antioxidant which was dissolved in the oil. The amount of antioxidant that was added was the legal maximum of 0.02% by weight of the total lipid in the final spread. The ingredients were "pre-cut" in a Hobart VCM 2-speed grinder at 1,800 r.p.m. for approximately 5 minutes. The premix had a "gritty" 33 "soupy" consistency. The precut was next sent through an Urschel Comitrol 1700TM microcut grinder for the final grind. The gritty precut is spun inside the microcut blade head at speeds up to 13,000 r.p.m. The primary grind (pre-cut) has a final temperature of about 80°F, while the final grind temperature Should be about 140-180°F. Temperatures in excess of l95-200°F will cause the outcoming soynut butter to scorch, forming dark brown streaks in the product. Packaging_of Soynut Butter Two different packaging studies were undertaken; in the first soynut Spread was packed in wide mouth pint jars and stored under the following conditions: dark/100°F, dark/ambient 22°C and under fluorescent light/ ambient temperature. The second study involved packaging of soynut butter in wide-mouth pint glass jars with metal dome lids, high density polyethylene (HDPE) pint tubs with pry-off HOPE lids. The third package R plastic pouches system consisted of the same tubs placed in nylon/Saran which were evacuated prior to sealing. Storage conditions were identical to those used in the first study. Each container was filled to approxi- mately the same level in which the glass jars contained an average of 255 g soynut butter per container; while the HDPE tubs contained an average of 275 g soynut butter per container. The average headspace was contrOlled by filling to the same point when filling each container. A sufficient amount of spread was packaged in each of the containers for both study one and study number two. The experimental design for the studies is shown in Table l. 34 . Amongmm Page“ wepv Am .LP maze maaxum ”maze uao2u< ”new: msmumxm mcwmmxoma my me xenpm Ppomno mmwgozoa Amo_a5am .aoou oev A.wa\<=m .aU\ozme .ozme .e=m\.av mxmmz Np mxmmz v mxwmz N Popuwcfi xcmo\mooop mcopu_ccou «museum Amgmn mmmpm guzosiocwz ppmm cw mm_asmm PPom eoea.e.a mongoum .mpzcxom eo cowuusnoea Low Zap; uuzuosm .q mc:m_d oc—mmwuoga coausn uncxom so m:_m~xoaq so» awn :_ mongoum co.uooqm=. uuanoca umsmpc_u m:__oou uo.paau ac—commmm go mcvu—am ammo; __o mac::_ucou magma gmumzioo mecca acmgxo use meouvnvccp :.mg»eu muo>wuoooo mcooaxom so; we =o_uoaam:_ PNMQ’LDONQ zap; uoauoca l b “ecu go>o>cou |_l go>m>=ou — soumaom maoacpucou — ‘Logoumzoa Locomaom —\\\Illl# Agm—_onv Emoum Aocaumwos nosv .mcoon ucunxo ecu \\\\\\L\. 1 mgou.a_g=_ :Fmaxsu mua>vuuaoa ogzumpoe xm_-xm_ «museum comm gum 50 the blades on the microcut head. Temperatures in excess of l95-2000C will cause the outcoming soynut butter to scorch, forming dark brown streaks in the product. Three factors appeared to be critical in maintaining constant, reproducible temperature: a minimum of 25% vegetable oil to achieve desirable viscosity, a primary grinding time of five minutes, and the rate of delivery of the pre-cut mix to the Urschel Comitrol. The proposed product flow would next enter a scraped-surface heat-exchanger, likeia VOtatorTM. Small amounts of moisture must be removed before the butter is packaged to reduce high viscosity and pastiness. The moisture in the wann butter can be stripped off along with residual air in a deaerator. This could be used in addition to a swept-surface heat-exchanger. A proposed flow chart for soynut butter can be seen in Figure 5. The cooled soynut butter would then be filled automatically into jars or tubs, moving on a conveyor belt, and the lids placed and tightened on each container. Filling temperatures should be maintained below the point at which the polymer becomes soft 9 250-300°F. The containers would pass through a metal detector before the lids were in place, be code dated, and once closed could be packed into corrugated shippers with partitions and stapled shut. Corrugated trays are another possibility with a saran-coated low density polyethylene shrink-wrap over the containers and tray. Adams Foods, a division of International Multifoods, Inc., produces old-fashioned peanut butter and found leakage of oil from plastic containers to be a major problem (ihass, 1985). The peanut butter, without stabilizers, released oil to the product surface where it worked its way through the snap-on lid during shipment. This caused rejection 51 mcwgums Loamuwpaam uw_\mcwpomm \o=e_P.e oeoaeoe=< Aconcaguxm new; mumuczm uqmzmv st cmpdwm .mcwmoxuma Low :_n ommCOHm ou gang .x_e ;m_cme o» n:_co .pogu_Eou 0» vow; .emucmpn oc_u_o; 0» gang .st acmncoOOm o» ucmeo .Foeuweou ou Luann ucm ucmpm .mucmmnmgmcp acmewcg name use zm_m: . 3oAm buzooma emuuaa uzcxom mo :o_uu=noea go; acmco 20pm pounced .m meamwm —_- __ gang .m *u m>~m> m d / cemp zh—oguwsou szumsa _U _ L i U xv: xemewea xv: xcmucoumm coccum connwm mechm cona_m a=_a.o= .5 \\\II|//J \\\\\IIIIIIJ .o .m .c .omem:~ .m —_o muzcxom xeo .N — mpmum 52 of individual containers and surrounding intact containers in the shipping case which were stained by the leakage. A two-head automatic filling and sealing machine was used to apply a clear polyester film innerseal over the container tops before lidding. Adams also used a tamper evident feature on its plastic pails, in which two tabs on the shoulder rim must be broken in order to lift the lid. In the case of soynut butter, oiling- off is reduced by the addition of hardener in the formulation, but high temperature storage at 100°F over a period of 1 week did cause some oil to leak to the surface. The system used by Adams Foods would therefore be equally as effective in the packaging of soynut butter, especially in the reduction of leaking containers. Adams projected product shelf-life at 6 months to 1 year maximum, because of the high oil content of the product and the oxidative, thermal, and light transmission stresses applied during the distribution cycle. A swept-surface heat exchanger, such as the VotatorTM , provides a system in which there is rapid cooling under agitation. Such a system is used by poanut butter manufacturers to crystallize the fat in the mix providing improved filling capabilities while limiting air bubbles in the final product. Initial soynut butter production runs indicated that air bubble formation was a persistent problem at normal processing tempera- -‘ tures of 180°F. Some measure of control was achieved by allowing containers of warm butter to stand for 30 minutes with occasional shaking to force the trapped air bubbles out of the product. This procedure would be highly impractical for high-speed filling operations TM and would necessitate the use of a heat exchanger such as the Votator . Also critical was the ability to control the rate of addition of the 53 pre-cut mix into the Urschel unit. A constant flow over a short period of time gave the coolest processing temperatures, below 160°F. Among the objectives of processing soynut butter was to develop a commercial spread with sufficient added sweetener to mask any bean flavor. In considering improvement of flavor the related changes in palatability were also observed. With the addition of 3.0% powdered dextrose there was no significant change in the viscosity of the soynut butter with the added vegetable oil composing 25% of the formula, and the soynuts composing 71% of the same batch. Added salt and partially hydrogenated palm oil respectively provided additional flavor and helped prevent excess oil-off. When dextrose was added in excess of 7.0%, fructose in excess of 3.0%, or pear concentrate in any amount, a corresponding percentage of vegetable oil had to be added to prevent clogging of the Urschel Comitrol grinder. Product with a 4.0% fructose of 7.5% dextrose and containing only 25% vegetable oil would exit the finish grind "scorched" at temperatures of 200-220°F. Dextrose, fructose, and corn syrup solids are reducing sugars and can react with free amine groups in the soy protein. This reaction is accelerated by heat and is especially pronounced in peanut protein at 180°F (Weiss, 1983). This same effect was seen in sample batches of soynut butter in excess of 180°F and resulted in increased viscosity and frictional heat. Another undesirable effect occurs when hydrated dextrose in a nut butter dehydrates during grinding. The moisture is transferred to the soy protein and causes the butter to thicken considerably. Dextrose can then rehydrate forming larger crystals and a grainy texture (Ibid). The solution may be to use an anhydrous dextrose, increasing the pre-mix 54 grinding time, or adding normal dextrose monohydrate to the butter after it has cooled following the final grind(Ibid). The same logic would apply when adding other reducing sugars to warm soynut butter. It was observed that when pear extract was added before final grinding, even in small quantities, excess of 40% added vegetable oil per batch was necessary to achieve appropriate processing temperatures and palata- bility. Fructose added before grinding at a concentration of 3% required approximately 35% added vegetable oil per batch. The implications of these processing parameters will be critical in future product scale-up for commercial production runs. It must be emphasized that through the optimization of the amounts of these necessary ingredi- ents, i.e. vegetable oil, sweetener, soynuts, stabilizer and/or emulsifier and salt, a spread of good consumer acceptability has been produced. The discussion of consumer preference is discussed in the Results of Sensory Evaluation section. The high percentage of added vegetable oil required to achieve palatability will be of major concern ‘UJ health-conscious consumers who seek to avoid consumption of high-fat foods. However, both peanut or soy spreads are by nature high fat foods. The high fat content is also of major concern in protection of the product from off-flavor development and the necessary packaging to protect the product from autoxidation. In the work of Pichel and Weiss (1967) a Similar process was used which included the step of conditioning of the soybeans by moisturizing them with an amount of water which was reported to affect removal of "beany" or "grassy" flavor when the beans were properly heated. The added moisture appeared to penetrate and permeate the bean 55 to an extent that the "grassy" or "beany" flavor constituents of the bean were volatilized along with the water, and most likely lipoxy- genases were inactivated. In the subsequent heating step Pichel fried the moisturized bean in hydrogenated vegetable oil at 140-180°C for about 2-5 minutes. Subsequent grinding in a high Speed attrition mill with an impeller speed of 7,500 r.p.m. was found to give the most efficient subdivision of soybeans. Results obtained in processing procedures paralleled much of Pichel's findings with the exception that the soynut butter was grainy despite the addition of 25% vegetable oil. When the added vegetable oil content exceeded 25%, based on the total batch weight, the soynut butter could be processed through the Urschel grinding unit at tempera- tures below 180°F. The product still had a grainy mouth-feel. With the addition of 35-40% extra vegetable oil the spread had very little grainy texture. The high oil content of the soynut butter made it feel smooth on the palate. Pichel added enough bland, edible oil to increase the oil content to about 35-60% based on the weight of beans (Ibid). Separation of oil is a major concern so that various mono- glycerides, partially hydrogenated palm oil, and other commercial _hardeners must be added. Although beany flavor is substantially reduced after boiling the raw, dehulled beans, the bean-flavor was still present in the soynut butter. Hawley (1966) realized this problem and stated that it is Sometimes necessary to add 10% to 15% of flavoring materials based on total product weight. Among the flavoring components were peanut oil extract, sesame seeds and various sweeteners. Experimentation with 56 various sweeteners has shown that the addition of 3.0% fructose signi- ficantly improved the flavor of the spread and ameliorated off-flavors. Addition of fruit flavored extracts also improved the flavor but required the addition of even greater amounts of added vegetable oil to achieve the same smooth consistency in the product. An independent laboratory report on the differences between peanut butter and soynut butter is seen in Table 2. These data indicate that the base amounts of carbohydrates, protein and fat are virtually identical, but as previously seen, soy protein contains a richer amino acid array than peanut protein. 57 Table 2. Soynut Butter Comparison to Peanut Butter, Based on 32 g Samples. Soynut butter Peanut butter Calories 189 200 Protein 9.6 gm 9.0 gm Carbohydrates 5.0 gm 6.0 gm Fat 14.5 gm 16.0 gm 58 Results of Study #1 The Effect of Antioxidant Type Upon Lipid Stability The type of and concentration of antioxidant plays a significant role in the stabilization of fats and oilS, depending on the chemical composition of the lipid system. Sherwin and Luckadoo (1969) reported that .02% TBHQ provided greater protection than .02% BHA and .02% PG in soybean oil stored at 140°F over a 40 week time period as measured by peroxide value. Reported peroxide values were less than 10 meq/kg after 12 weeks of storage in both control samples, and those containing antioxidants. In this study, control soynut butter samples were compared to samples containing four different antioxidants: BHA/BHT, BHA/PG, TBHQ and TBHQ/cirtic acid. The results are shown in Figures 6, 7 and 8 were based on peroxide values obtained from extracted oil in a 5 9 sample of soynut butter. All of the figures indicate increasing peroxide values in each storage condition which suggests that the extent of peroxide decomposition was not achievable. Samples stored in the dark at ambient conditions had the slowest rate of increase in peroxide value. The dark, 37°C environment proved to be the most detrimental, with storage in fluorescent/ambient conditions were of intermediate severity. Samples containing TBHQ and TBHQ/CA provided greater protection to soynut butter than did BHA/PG or BHA/BHT at dark, 37°C and fluorescent, ambient (22°C). BHA/BHT provided protection approaching that of TBHQ only in a dark ambient environment. BHA/PG provided the least protection of any of the antioxidants and control samples had the highest rate of increase of peroxides over 12 weeks total 59 .:o_u_u:oo Agcmwnsmv uomm .xemv :_ ommeopm mcwezn emppan uzczom go» we?“ m> mzpm> muwxoema .o we:m_u exam: cm asap .___e_ a. m w m Mama“: 0x/'bau 'an[eA apIXOJad 1m .Im a! o L. 335 n :32: a 233 4 33:8 4 o hzmHmz< xm0m sow mew» m> mapm> mupxoema .m mezmwa exec: e“ came a; a J .A ._ w a. ._ ._ as ._ : Dx/°bau 'an[eA apIXOJad rm in a asap 0 .In 33.2 a 533 a 233 a. 235 s a #2392 Hzmommmoa... .cowuwucoo comm .xemv cw mmmcoum mcwesu emupaa uzcxom so» ms_u m> m=Fm> onwxoemm mxomz cm asap .3. a. a. ._ ._ ._ S ._ .0 ._ a. a; ~ \\__.~ .1 a \ Tn I\\uuuuu\\\i. _\‘\\. r. “M .1m Ta 15 To o;:.o ¢axazmh I. hzmsszu.u .1m 8254. 2.34.39... 2 ohm zm .5: Nmm mocmaeomam .mzpm> 0m com ms_u m> .e: Nmm mocmneomnm .m=_m> .E: mmm mocmagomnm .wocmneomnm wcm_u woummzacou axon: ca «5 a: see... a. ._ e a. ._ A .Lx axmh «no. mess conga a azmh_nufi unshmrx—_as 33:8 22. 8.8 a 3:8. 5.35;: 3.: as _. 2:89 5.55.; Be as a nogacau mash man: a. 9 IA. IN. ohm v.52 Foo. 1mg. ‘uu 392 v 'auagg ‘fuoo .__ mesmea 73 .memn mmopa cw :o_uwvcoo Au:m_nsmv uoNN .ugmwF ucmommeoapw cw mameoum a:_g:u Leeann uzcxom gee ws_u m> .e: Nmm mocmoeomam .om go: ms_u m> .5: Nam mocmngomam . .E: Nmm mu:m::omne . .5: Nmm 3:33QO £3 Ho; 3:3... axoou :m numb 2. m m P m a. d_ a e cl. 8 V A 0. mm ”a v re. Mn 7.. U mu 5N. 2:65.938... 553.55%... a. hszmz< ¥m .5: «mm mucmngomom .5: mmm mucmngomna acmwc cmammzncou .NF mg:m_u axooz cw gawk we. w ._ J w ._ : rm: . 1mg. 'uu as; v 'auagn 'fuoo film. 9.2 fig. 33 was! a. 85:8 23 wk! C 13. ohm xm .5: mmm mucmngomam mcmpu umummancou .mWP mgzmwu mxwoz cm camp a ._ a. m. _ m w ._ .9 98. 'W 882 V 'auam TUOO Imp. 559.5532: .2 w... 4 25.8 5.53:: .2 as q r8. chm xm’a 81 .mgmn mmmpm cw cowuwccou uoum .xgmu cm mangoum m:_gzc gmuuza uzcxom go» mswu m> .E: mmm mucungomam acmwu umacmawcou .m_ mgzmwu mxooz c“ osgb a - M...“ ._ m- L_—o O 4 ""- 253. '“U 882 V '3U9EU 'fUOO a! Q. a... 8.3 4 25.3 a... 8.3 « 1mm. “Km xm Amsv Loam; do :_mm acmwmz umz 28: E 2.: mmw ojm cmw mam vvm ~* 9 was: cowmmwemcmep Loqm> amym: Fe .Ioc foo" ..|o~— lbw" (0m) UEES auogan 13M Jaaen .cm mesa.“ 84 headspace (Ibid). As previous1y seen in Tab1es 3, 4 and 5, there were more pronounced 1emony and grassy off f1avors in those samp1es stored in g1ass jars in f1uorescent 1ight. HDPE tubs with .02% TBHQ/CA added to the samp1e provided better stabi1ity despite on1y s1ight differences in conjugated diene absorbance va1ues between packages (Figures 12, 13, 14). Storage in opaque or co1ored containers, as compared to c1ear jars, can be seen in terms of reduced oi1 separation, better aroma and f1avor, but is not significant for chemica1 detenninations (Ibid). The she1f—1ife of peanut butter without antioxidants in po1yethy1ene jars has been determined to be 9 months to 1 year (Ibid). It may be projected that soynut butter with .02% TBHQ/CA in HDPE tubs, can be stored in a contro11ed ambient environment from s) to 12 months as we11 due to the antioxidant protection. Soynut butter stored in g1ass jars in a dark, ambient environment may be stored from 12 to 16 months. Soynut butter stored in p1astic (po1yethy1ene) pai1s for 1 year in the dark was tasted by an infbrma1 pane1 a1ongside month 01d samp1es stored in g1ass jars in the dark. Pane1ists cou1d not DEPCETVE any major differences between the two samp1e soynut butters in terms of rancid f1avor. Additiona1 sensory work, in terms of optimum sweetener 1eve1s, was studied to further product acceptabi1ity by consumers. The importance of these resu1ts is that a significant cost-savings can be rea1ized by using Tighter HDPE tubs for packaging of soynut butter. Fatty acid composition of extracted oi1 samp1es from soynut butter showed that after 24 weeks of storage in 37°C, dark unsaturated fatty acids 18:1, 18:2, and 18:3 were most protected with .02% TBHQ/CA 85 added (Tables 6, 7, 8 and 9). Linolenic (18:3) acid was 9.0% of the total fatty acid array measured in the soynut butter in HDPE tubs with antioxidant, 6.7% in tubs in nylon/saran pouches without antioxidant, and 5.0% in glass jars, control and antioxidant samp1es. Increased linolenic acid protection may be related to decreased production of secondary lipid oxidation products; aldehydes, ketones, alcohols. HDPE tubs are slightly poorer in protection of oleic (18:1) and linoleic (18:2) acids which are also responsible for secondary oxida- tion products. Sensory Analysis of Optimum Levels of Sweetener 0f the fifty volunteers, 45 responses were accepted in both the degree of difference test and the degree of preference test. An analysis of variance was perfonmed on the data received from the degree of difference test, and can be seen in the ANOVA Table 10. The calculated F value (33.60) exceeds the F values at the 5% and 1% levels of significance. The conclusion is that consumers can detect signifi- cant differences in the relative sweetness between soynut butter samples at the 1% level of significance. Since there was a significant difference among the samples, the ones that were different were determined using Tukey's Test. Sample means were arranged according to magnitude and the least significant difference was calculated as 0.756 (Table 11). Any two sample means that differ by 0.756 or more are significantly different at the 5% level. Results indicate that sample containing 9.6% pear concentrate is significantly sweeter than the other seven samples. Samples with 8.1% pear concentrate and 3.0% fructose were significantly sweeter than samples containing dextrose 86 Table 6. Fatty Acid Composition of Original 0ils Fatty acid (%) Fatty Acid** "Hi-Tone" Extracted Extracted veg. oil* Hi-Tone Hi-Tone Control (.02% TBHQ) 16:0 10.0 11.6 12.2 18:0 6.0 7.5 7.9 18:1 49.0 36.5 38.9 18:2 31.0 33.3 33.9 18:3 2.0-5.0 3.8 4.4 20:0 - - 0.6 22:0 - - 3.7 2.0 24:0 -' 2.8 - *Values CourteSy Bunge Edible Oil Co. **The notation used to describe fatty acids is number of carbon atom: number of double bonds. 87 Table 7. Fatty Acid Composition of Extracted 0i1 at 24 Weeks, 37°C, Dark. Fatty acid (%) Fatty Acids** HDPE tubs (control) HDPE tubs (.02% TBHQ) 16:0 11.1 11.5 18:0 7.4 8.1 18:1 35.6 38.3 18:2 32.3 33.0 18:3 5.5 9.0 22:0 8.0 - Table 8. Fatty Acid Composition of Extracted 0i1 at 24 Weeks, 37°C, Dark. Fatty Acid (%) Fatty Ac‘d** Glass jars Glass jars HDPE tubs (control) (.02% TBHQ) nylon/saran (control) 16:0 12.1 11.6 12.1 18:0 8.7 7.6 8.7 18:1 38.9 40.3 38.5 18:2 34.0 35.4 33.9 18:3 5.0 5.0 6.7 22:0 - - - 88 Table 9. Fatty Acid Composition of Extracted Oil at 24 Weeks, 37°C, Dark Fatty Acid** Fatty acid (%) HDPE tubs HDPE tubs (control) (.02% TBHQ) 16:0 11.1 11.5 18:0 7.4 . 8.1 18:1 35.6 38.3 18:2 32.3 33.0 ‘18:3 5.5 9.1 22:0 8.0 - 89 Table 10. Degree of Sweetness (ANOVA) (n=45) Source of variance df 55 ms f Samples 7 318.92 45.56 33.60* Judges 44 126.29 2.87 2.12 Error 398_ 417.58 1.36 Total 359 862.79 13.05 (7/303) = 2.0392, ram (7,303) = 2.7084, (SE) = m-35/45) = 0.174 (*33.60 > 2.7084, The difference between samples is significant at the 1% level) 90 Table 11. Comparison of Sample Means (Tukey's test (Snedecor, 1956)) n=45 Sample 9.6% P 8.1% P 3.0% Fructose 1.50% Fructose Sample avg. 7.82 6.58 6.44 5.89 Sample 3.0% Dextrose 4.0% C.S.S. 2.0% C.S.S. 1.5% Dextrose (R) Sample avg. 5.55 5.13 4.95 4.89 C.S.S. = Corn Syrup Solids Significant studentized range at the 5% level (8/308) = 4.347 Least significant difference = 4.347 x 0.174 = 0.756 ( Any two sample means that differ by 0.756 or more are significantly different at the 5% level.) 91 and corn syrup solids. The reference sample (R) was 1.5% dextrose and was judged to be the least sweet of all the samples. Degree of Preference Test The degree of preference test is a hedonic measurement used to determine the likes and dislikes, and the degree to which the consumer registers the sensory qualities of the product being analyzed. Consumers may say they prefer a certain product, but when it comes to behavior they may or may not follow what they say. For simple tastes, most individuals rated sweet as pleasant, salt as pleasant at low and middle levels but unpleasant at high levels, and sour and bitter as unpleasant at most concentrations (Engel, 1928). In most applied sensory research, overall "liking" represents the "bottom line" or key evaluative criterion against which the researcher judges all other variables (Moskowitz, 1984). The data in Table 12 show percentages of the 45 person sample liking and disliking the soynut butter samples with different levels of sweetener. The words "would not purchase" were added to the dislike category, and the words "would purchase" were added to the category of like responses. Samples containing 3.0% fructose showed that one third of the 45 people polled indicated they liked it moderately and would purchase it. 17.8% of the respondents said they liked the product very much or more and would purchase it. In contrast, the sample with 1.5% dextrose only rated 11.1% like responses at the "moderate" level and 17.8% at the "slight" level. The most disliked product contained 1.5% dextrose with 71% of the 45 consumers indicating dislike for the product. The 3.0% fructose sample was least disliked with only 31% 8.8 8.8 8.88 8.88 8.88 F... 8.8 8.8 888_88 88888 8888 88.8 ~.N N.N 8.8 m.w~ m.mm m.mp m.m v.8 mu_Pom aagxm :gou Ro.m o.o o.mp m.m o.o~ e.¢~ 8.8N o.o n.o L888 no.m o.o 8.8 m.m_ p.88 8.8m 8.8. N.~ ~.N 8888 88.8 ~.N o.mp m.mm . m.sp N.- m.m o.o o.o mmouuzgu no.m o.o 0.0 m.m m.mp o.o¢ 5.0 8.8 o.o wmouuzs$ Nm.~ o.o o.o m.- m.w~ w.mm m.m 8.8 ~.~ mmoguxmo wo.m o.o o.o 8.88 m.- o.ov N.- v.¢ 8.8 wmoguxwo am.F 83888x8 cuss Ng8> 88888882 .mmmmam .mmmflflm 88888882 88:5 xg8> .mammmmm 88888838 8888: 8:8 838; 88888888 882 88:83 8:8 8xppmwo mun: Am8m=omm8c m8 mo my 8888888888 88 88888: .N8 8.888 93 disliking it. Textural parameters were not analyzed in these sensory tests but ultimately played a part in the respondents' evaluation of each sample. A higher amount of added vegetable oil, and presentation of soynut butter on saltless crackers was quite effective in registering the consistent responses. Initial soynut butter production for consumer markets should use a sweetener like fructose at the 3.0% level. The disadvantages of fructose are its high cost, and its reducing sugar characteristics which will affect Maillard browning. Citric acid may help to slow the browning reactions. Dextrose is used by peanut butter manufacturers as well as corn syrup solids (Weiss, 1983). SUMMARY Development and packaging of a soynut butter was studied based on the work of Pichel and Weiss (1967). Optimum antioxidant combinations and optimum package systems were evaluated in two separate studies. In the initial study the peroxide value and TBA value were utilized to follow the oxidative changes in soynut butter samples containing various antioxidants. Results of the first study indicated that 0.02% TBHQ/CA was the most effective of the antioxidants used in terms of lower peroxide and TBA values. Control samples without antioxidant, or containing BHA/PG, were least effective in inhibiting autoxidation in the soynut butter. In a subsequent study conjugated diene absorbance and TBA (distilla- tion) absorbance methods were employed to evaluate lipid autoxidation in soynut spreads in various packaging systems during storage for 24 weeks. Results of the best straight lines at eight weeks indicated increased conjugated diene and TBA absorbance numbers for control samples without TBHQ/CA. There was little difference in packages as measured by the two chemical determinations, but significant differences were detected in tenns of light catalyzed flavor for samples stored in fluorescent light. Soynut butter in Opaque high density polyethylene (HDPE) tubs had decreased light catalyzed off-flavor production as compared to glass jars. A substantial cost savings may be realized if soynut butter is packaged in HDPE tubs, with a resultant shelf-life 94 95 of S? to 12 months under controlled storage. Savings would result from lower distribution costs due to decreased weight per pallet and decreased package costs. Sensory evaluation indicated that consumers prefer a spread containing 3.0% fructose as a sweetener, and they they are able to differentiate various levels of sweetener in soynut butter at the 1% level of significance. Production studies showed that with a minimum of 25% added vegetable oil, a pre-cut time of 5 minutes, and a steady rate of addition of the pre-cut mix to the ComitrolTM grinder, smooth throughput and temperatures below 180°F were achieved. RECOMMENDATIONS Further work in the study of autoxidation rates in soynut butter is recommended. Headspace volatiles such as hexanal, and reversion compounds such as a-pentyl furans could be concentrated on a TenaxTM GC trapping system and quantitated over time during storage in various environmental conditions (Doi gt_gl,, 1980). A full production scale-up study should be undertaken, with an uninterrupted flow from raw ingredients to final processing and packaging. A swept-surface TM is recommended to give a smooth heat exchanger such as the Votator and cool product without air bubbles. A distribution study with soynut butter packed in various containers to the points of retail sale would help to further confinn the final package selection. Further market testing as to ethnic and geographical preferences is also recommended to assure that the perceived market actually exists. 96 BIBLIOGRAPHY BIBLIOGRAPHY Angelo, A.J. St., R.L. Dry, and L.E. Brown. 1972. A comparison of minor constituents in peanut butter as possible sources of fatty acid peroxidation. J. Amer. Peanut Res. and Educ. Assoc. 4:186. Angelo, A.J. St., R.L. Dry, and L.E. Brown. 1975. Comparison of methods for determining peroxidation in processed whole peanut products. J. Am. Oil Chem. Soc. 52:34. ”Annual Book of ASTM Standards, Part 35", ASTM, Philadelphia, PA., 1981, D-l434. "Annual Book of ASTM Standards, Part 18", ASTM, Philadelphia, PA., 1972, E-93-99. Bligh, E.G. and R.J. Dyer, Can. J. Biochem. Phys. 37:912. Quoted in Melton, S.L., S.L. Moyers,, and C.G. Playford. 1979. Lipids extracted from soy products by different procedures. J. Am. Oil Chem. Soc. 56(4):489. Boozer, C.E., G.S. Hammond, C.E. Hamilton, and J.N. Sen. 1955. Air oxidation of hydrocarbons. II. The stoichiometry and fate of inhibitors in benzene and chlorobenzene. J. Am. Chem. Soc. 77: 3233. Bressani, R. 1981. The role of soybeans in food systems. JAOCS 58(3):392. Buck, D.F. 1981. Antioxidants in soya oi1. J. Am. Oil Chem. Soc. 58(3):275. Campbell, M.R. 1981. Processing and product characteristics for textured soy flours, concentrates and isolates. J. of Am. Oil Chem. Soc. 58(3):336. Chang, S.S., T.H. Smouse, R.G. Krishnamurthy, B.D. Mookherjee, and B.R. Reddy, 1966. Chem. Ind. 1926. Coppen, P.P. 1983. The use of antioxidants. Chapter 5. In: Ran- cidity in Foods. Eds. J.C. Allen and R.J. Hamilton. Applied Science Publishers Ltd., Essex, England, pp. 67-87. Dahle, K., E.G. Hill, and R.T. Holman. 1962. The thiobarbituric acid reaction of polyunsaturated fatty acid methyl esters. Arch. Biochem. and Biophys. 98:253. 97 98 Day, E.A. 1962. Discussion; Meat products. Chapter 11. In: Symposium on foods: Lipids and their oxidation. Ed. H.R. Schultz, A. Eds., E.A. Day, and R.O. Sinnhuber. AVI Publishing Co., Inc., Hestport, CT, p. 212. Doi, Y., T. Tsugita, T. Kurata, and H. Kato. 1980. Changes of head- space volatiles of soybeans during roasting. Agric. Biol. Chem. 44:1043. Dugan, L.R., Jr., and H.R. Kraybill. 1956. Toc0pherols as carry- through antioxidants. J. of Am. Oil Chem. Soc. 33(11):527. Dugan, L.R., Jr. 1961. Development and inhibition of oxidative rancidity in foods. Food Technology 15(4):10. Dugan, L.R., Jr. 1976. Lipids. In: Food Chemistry. Ed. 0.R. Fennema. Marcel Dekker Inc., New York, Part I, p. 182. Eastman Chemical Products, Inc. 1974. “Effectiveness of TBHQ Antioxi- dant in Refined Vegetable Oil.," Publication No. 2F-204A. Eastman Chemical Products, Inc., 1980. "Tenox, TBHQ Antioxidants for Oils, Fats and Fat Containing Foods", Kingsport, TN. "Effect of Mo-Con Oxtran 100 Operating Variables on Polyester Film Oxygen Transmission Rate", 1978, Ontario Research Foundation Report No. Eng. R-79-32. Elsevier, 1981. Oil and protein crops. Chapter 6. In: The agricultural compendium for rural development in the tropics and subtropics. p. 481, Elsevier Scientific Pub. Co., N.Y. Farmer, E.H., and D.A. Sutton. 1943. The course of autoxidation reactions in polyisoprenes and allied compounds. Part IV. The isolation and constitution of photochemically-formed methyl oleate peroxide. J. Am. Chem. Soc. 48:392. Fore, S.P., L.A. Goldblatt, and H.P. Dupuy. 1973. A simplified technique used to study the shelf-life of peanut butter. J. Amer. Peanut Res. and Educ. Assoc. 5:39. Frankel, E.N., C.D. Evans, and J.C. Cowan. 1960. Thermal dimerization of fatty ester hydroperoxides. J. Am. Oil Chem. Soc. 37(9):418. Frankel, E.N. 1980. Lipid oxidation. Prog. Lipid Res. 19:1-22. Frankel, E.N. 1980. "Analytical methods used in the study of autoxida- tion processes. Chapter 9. In Autoxidation in Food and Biological Systems", Eds. M.G. Simic, and M. Karel. Plenum Press, New York, pp. 142-149. 99 Frankel, E.N., and v.5. Neff. 1983. Formation of malonaldehyde from lipid oxidation products. Biochimica et Biophysica Acta. 754:264. Frankel, E.N. 1984. Lipid oxidation: mechanisms, products and biological significance. J. Am. Oil Chem. Soc. 61(12):l908. Freeman, A.F., N.J. Morris, and R.K. Hillich. 1954. Peanut butter. USDA AIC-370. Gray, J.I. 1978. Measurement of lipid oxidation: A review. J. Am. Oil Chem. Soc. 55(6):539. Gunstone, F.D. and F.A. Norris (Eds.). 1983. "Lipids in Foods, Chemistry, Biochemistry and Technology". Pergamon Press, Inc., Oxford, England. Hawley, R.L. 1969. High protein product and process for producing _same. U.S. Pat. 3,469,991, Sept. 30. Hamilton, R.J. 1983. The chemistry of rancidity in foods. Ch. 1. In "Rancidity in Foods", J.C. Allen and R.J. Hamilton (Eds.). p. 11. Applied Science Publishers Ltd., Essex, England. Heller, A.N. and J. McCarthy. 1944. "Soybeans From Soup to Nuts", Vangard Press, Inc., New York, p. 18. Hess, L. 1985. Double tamper-guarding also solves leakage problem. Food Processing 46(4):124. Holman, R.T. and 0.0. Burr. 1946. J. Am. Chem. Soc. 67:562. Quoted in Holman, R.T. 1954. "Progress in the Chemistry of Fats and Other Lipids, Vol. II", Pergamon Press, London and New York. Hudson, B.F. 1983. Evaluation of oxidative rancidity techniques. Ch. 3. In "Rancidity in Foods", pp. 47-56. Applied Science Publishers Ltd., Essex, England. LaBuza, T.P. Protein. 1977. Ch. 5. In "Food and Your Nell-Being". T.P. Labuza (Ed.), p. 92. West Publishing Co., St. Paul, MN. LaBuza, T.P. 1971. CRC Crit. Rev. Food Technol. 2:355. Larmond, E. 1977. "Laboratory Methods for Sensory Evaluation of Food". Publication No. 1637, Communications Branch, Agriculture Canada, Ottawa, Canada. Lea, C.H. 1962. Oxidative deterioration of food lipids. Chapter 1. In "Symposium on Foods: Lipids and Their Oxidation", H.N. Schultz, (Ed.), E.A. Day and R.O. Sinnhuber. pp. 5-15. AVI Publishing Co., Nestport, CT. 100 Lundberg, N.O. 1962. Mechanisms. Chapter 2. In "Symposium on Foods: Lipids and Their Oxidation", H.R. Schultz (Ed.), E.A. Day and R.O. Sinnhuber. pp. 39-46. AVI Publishing Co., Nestport, CT, p. 40. Mehlenbacher, V.C. 1960. "The Analysis of Fats and Oils", Garrard Press, Champaign, IL. Min, 0.8. and J. Hen. 1983. Qualitative and quantitative effects of antioxidants on the flavor stability of oil. J. of Food Sci. 48(4): 1172. Moerck, K.E. and H.R. Ball. 1974. Lipid oxidation in mechanically deboned chicken. J. of Food Sci. 39(5):876. Morrison, H.R. and L.M. Smith. 1964. Preparation of fatty acid methyl esters and dimethylacetals from lipids with boron fluoride- methanol. J. of Lipid. Res. 5(5):600. Moskowitz, H.R. 1984. Sensory analysis, product modeling, and product optimization. Ch. 2. In "Analysis of Foods and Beverages: Modern Techniques", G. Charlambous (Ed.), pp. 13-48. Academic Press, Inc. Orlando, FL. Nawar, H.N. and A. Hitchwoot. 1980. Autoxidation of fats and oils at elevated temperatures. Ch. 13. In "Autoxidation in Food and Biological Systems", M.G. Simic and M. Karel (Eds.). Plenum Press, New York, p. 217. Nelson, A.I., M.P. Steinberg, and L.S. Wei. 1978. Development of whole soybean foods for home use: rationale, concepts, and examples. In "Whole Soybeans for Home and Village Use", Nelson, A.I. (Ed.) International Ag. Publications, Intsoy Series #14, pp. 5-11. Nelson, K.H. and N.M. Cathcart. 1983. Transmission of light throu h pigmented polyethylene milk bottles. J. of Food Protection. 47(5): 346. "Official and Tentative Methods of the American Oil Chemists' Society", AOCS, Champaign, IL, 1973, Method T: la-64. Orthoefer, F.T. 1978. Processing and utilization. Ch. 7. In "Soybean Agronomy, Processing, and Utilization", A.G. Norman (Ed.) Academic Press, New York, pp. 220-246. Pichel, M.J. and T.J. Weiss. 1967. Process for preparing nut butter from soybeans. U.S. Pat. 3,346,390, Oct. 10. Pohle, N.D., R.L. Gregory, and B. Van Giessen. 1964. Relationship of peroxide value and thiobarbituric acid value to development of undesirable flavor characteristics in fats. J. Am. Oil Chem. Soc. 41(10):649. 101 Porter, R.L. 1980. Recent trends in food applications of antioxi- dants. Ch. 19. In "Autoxidation in Food and Biological Systems", M.G. Simic and M. Karel (Eds.), Plenum Press, New York, pp. 301-310. Pryde, E.H. 1980. Composition of soybean oil. Ch. 2. In "Handbook of Soy Oil Processing and Utilization", American Soybean Associa- tion and the American Oil Chem. Soc., p. 13. Ray,(F3 1981. Soy, the world's miracle. J. of Am. Oil Chem. Soc. 58 3 :123. Rossel, J.B. 1983. Measurement of rancidity. Ch. 2 In "Rancidity in Foods", J.C. Allen, and R.J. Hamilton. Applied Science Publishers Ltd., Essex, England, pp. 25-30. Russo, J. 1964. Stabilized peanut butter. U.S. Pat. 3,129,103. April 14. Selke, E, H.A. Moser and R.K. Rohwedder. 1970. Tandem gas chromato- graphy - mass spectrometry analysis of volatiles from soybean oil. J. of Am. Oil Chem. Soc. 47(10):393. . Selke, E. and N.K. Rohwedder. 1983. Volatile components of trilino- 1enin heated in air. J. of Am. Oil Chem. Soc. 60(11):1853. Sherwin, E.R. and B.M. Luckadoo. 1970. Studies on antioxidant treatments of crude vegetable oils. J. of Am. Oil Chem. Soc. 47 1 :19. Sherwin, E.R. and J.H. Thompson. 1967. Tertiary-butylhydroquinone-an antioxidant for fats and oils and fat containing foods. Food Tech. 21 6 :106. Sherwin, E.R. 1976. Antioxidants for vegetable oils. J. Am. Oil Chem. Soc. 53(6):430. Sidwell, C.G., H. Salwin, M. Benca, J.H. Mitchell, Jr. 1954. The use of thiobarbituric acid as a measure of fat oxidation. J. Am. Oil Chem. Soc. 31(12):603. Sidwell, C.G., H. Salwin, and J.H. Mitchell, Jr. 1955. Measurement of oxidation in dried milk products with thiobarbituric acid. J. Am. Oil Chem. Soc. 32(1):13. Sinnhuber, R.O. and T.C. Yu. 1958. Characterization of the red pigment in the 2-thiobarbituric acid determination of oxidative rancidity. Food Res. 23:626. Sinnhuber, R.O. and T.C. Yu. 1977. The 2-thiobarbituric acid reaction, an objective measure of the oxidative deterioration occurring in fats and oils. Lipid Chemistry 26(5):259. 102 Smith, A.K. and S.J. Circle. 1978. Historical background. Ch. 1. In ”Soybeans and technology, Vol. I: Proteins", AVI Publishing Co., Inc., Westport, CT, p. 11. Smith, A.K. and S.J. Circle. 1978. Processing soy flour concentrates, and protein isolates. Ch. 9. In "Soybeans: Chemistry and Technology, Vol. I: Proteins”, AVI Publishing Co., Inc., Westport, CT, pp. 294-338. Steinke, F.H. 1979. Measuring protein quality of foods. In "Soy Protein and Human Nutrition", H.L. Wilcke, D.T. Hopkins and D.H. Waggle (Eds.). Academic Press, Inc., New York, pp. 310-311. Steinke, F.H., E.E. Prescher, and D.T. Hopkins. 1980. Nutritional evaluation (PER) of isolated soybean protein and combination of food proteins. J. Food Sci. 45(2):323. Tarladgis, B.G., A.M. Pearson, L.R. Dugan, Jr. 1962. The chemistry of the 2-thiobarbituric acid test for the determination of oxidative rancidity in foods. I. Some important side reactions. J. Am. Oil Chem. Soc. 39(1):34. Tarladgis, B.G., A.M. Pearson and L.R. Dugan, Jr. 1964. Chemistry of the 2-thiobarbituric acid test for determination of oxidative rancidity in foods. II. Formation of the TBA-malonaldehyde complex without acid-heat treatment. J. of Sci. of Food Agric. 15(9):602. Tarladgis, B.G., B.M. Watts, M.T. Younathan and L.R. Dugan, Jr. 1960. A distillation method for the quantitative determination of malon- aldehyde in rancid foods. J. of Am. Oil Chem. Soc. 27(1):44. Till, D.E., D.J. Ehntholt, R.C. Reid, P.S. Schwartz, K.R. Sidman, A.D. Schwope and R.H. Whelan. 1982. Migration of BHT antioxidant from high density polyethylene to foods and food simulants. Ind. Eng. Chem. Prod. Res. Dev., 21(1):106. Tranggono. 1978. Studies with selected antioxidants in vegetable oils. M.S. thesis, Michigan State University, East Lansing, MI. Vanderveen, J.E. 1979. Needs and concerns by federal regulatory agencies on measuring protein quality. In, I'Soy Protein and Human Nutrition", Academic Press, Inc., New York. p. 303-305. Warner, K. and E.N. Frankel. 1985. Flavor stability of soybean oil based on induction periods for the formation of volatile compounds by gas chromatography. J. of Am. Oil Chem. Soc. 62(1):100. Weiss, T.J. 1983. Peanut butter. Ch. 11. In: "Food Oils and Their Uses", 2nd Ed. AVI Publishing Co., Inc., Westport, CT. pp. 247-267. Wheeler, D.H. 1932. Peroxide formation as a measure of autoxidative deterioration. Oil and Soap 9:89. 103 Willich, R.K., N.J. Morris and A.F. Freeman. 1954. Peanut butter. V. Effect of processing and storage of peanut butter on the stability of their oils. Food Technol. 8, 101-104. Wolf, W.J. and J.C. Cowan. 1971. "Soybeans as a Food Source”, CRC Press, Cleveland, OH. pp. 38-39. Woodroof, J.G. 1983. Peanut butter. Ch. 9. In "Peanuts: Production, Processing, Products", 3rd Ed., AVI Publishing Co., Inc. pp. 181-225. ICH GAN STATE UNIV I “I 3 . LIBRARIES I" III I111I111111111111111 13498 12931069